Unicompartmental

Knee Arthroplasty

A New Paradigm?
Arnaud Clavé
Frédéric Dubrana
Editors

123
Unicompartmental Knee Arthroplasty
Arnaud Clavé • Frédéric Dubrana
Editors

Unicompartmental
Knee Arthroplasty
A New Paradigm?
Editors
Arnaud Clavé Frédéric Dubrana
Service de Chirurgie Orthopédique Service de Chirurgie Orthopédique
Clinique Saint George CHU Cavale Blanche
Nice, France Brest, France

ISBN 978-3-031-48331-8    ISBN 978-3-031-48332-5 (eBook)

https://doi.org/10.1007/978-3-031-48332-5

Original French edition published by Sauramps Medical, Montpellier, France, 2020

Translation from the French language edition: “La prothèse unicompartimentale de genou - Vers
un nouveau paradigme ” by Arnaud Clavé and Frédéric Dubrana, © Sauramps Médical,
Montpellier, France, 2020. Published by Sauramps Médical. All Rights Reserved.

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature
Switzerland AG 2024
This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher,
whether the whole or part of the material is concerned, specifically the rights of reprinting, reuse
of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical
way, and transmission or information storage and retrieval, electronic adaptation, computer
software, or by similar or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are
exempt from the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors, and the editors are safe to assume that the advice and information in
this book are believed to be true and accurate at the date of publication. Neither the publisher nor
the authors or the editors give a warranty, expressed or implied, with respect to the material
contained herein or for any errors or omissions that may have been made. The publisher remains
neutral with regard to jurisdictional claims in published maps and institutional affiliations.

This Springer imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

Paper in this product is recyclable

Foreword

Unicompartmental knee arthroplasties (UKA) have a long history in the treat-

ment of degenerative knee disease. Initially described in the early 1970s as an
alternative to total knee arthroplasty, the concept of a resurfacing option lim-
ited to one of the three knee compartments continues to play an essential role
in managing knee osteoarthritis (OA). The current renewed interest in UKA
is completely justified first for their delivery of a less invasive procedure with
faster recovery and, secondly, their improved results and patient satisfaction
compared to total knee arthroplasty. They represent a significant proportion
of the so-called forgotten knees that every surgeon and patient dreams of
obtaining after a surgical procedure, and modern prosthetic knee surgery can-
not be considered without precise knowledge of mono- or bicompartmental
arthroplasty. Of course, we must not forget technological advances, such as
computer-assisted and robotic surgery, which have made this procedure even
more reliable and reassured its most reluctant opponents.
As we enter the fifth decade of its use, Arnaud Clavé and Frédéric Dubrana
have sought in this work entitled Unicompartmental Knee Arthroplasty
towards a new paradigm to present a modern vision of UKA, supported by
long-term results that equal, and in many cases are better than total knee
arthroplasty. In this book, they have brought together a group of French and
international experts in the field, asking them to present the state of the art in
monocompartmental knee surgery without, of course, overlooking the his-
torical aspects, which make it possible to better comprehend the current
strategies.
The book naturally starts with the history and biomechanical concepts of
OA and monocompartmental arthroplasty. The conventional indications and
modern approach to them will then be detailed before the principles for per-
forming fixed and mobile plateau arthroplasty, as well as the different align-
ment philosophies, are comprehensively reviewed. Ambulatory management
and complications will be described, as will UKA revision. Innovative tech-
nologies will receive special attention before specific situations such as bilat-
eral arthroplasty, external UKA, or resumption of athletic activities are
examined. Lastly, the registers will provide exclusive insights into current
objective data.
From indications for the surgical technique to the results, the reader will
have access to the latest reviews and opinions on this fascinating topic, and
we must thank Arnaud Clavé and Frédéric Dubrana as well as the authors for
all their hard work summarising them.

v
vi Foreword

I know that this book will be an invaluable resource for anyone interested
in knee surgery and I hope that you will enjoy reading it.

Lyon, France Sébastien Lustig

Preface

After almost half a century of reflection, hesitation, and research, unicom-

partmental knee arthroplasty is finally reaching maturity. It was the Oxford
school that bravely carried the torch for its resurrection. A renaissance,
because orthopedists, distracted from their history, had, for a time, forgotten
the very origin of modern TKAs: quite simply, two unicompartmental knee
implants! This book is not a new paradigm because the truth it defends may
be false tomorrow. However, it is a change of references, an opening of the
mind, and a hope for many patients.
Through the photography of their activities, we must thank the forty (or
so) authors who participated in its creation and shared their knowledge and
expertise.

Nice, France Arnaud Clavé

Brest, France Frédéric Dubrana

vii
Contents

1 
History of Unicompartmental Prostheses��������������������������������������   1
Frédéric Dubrana and Hoel Letissier
2 
The Disappearing Unicompartmental Knee Prostheses�������������� 19
Samuel Laurent, Baptiste Montbardon, Arnaud Clavé,
and Frédéric Dubrana
3 
Anteromedial Osteoarthritis and UKA������������������������������������������ 29
Samuel W. King, Bernard H. Van Duren,
and Hemant Pandit
4 
Conventional Indications for Unicompartmental Knee
Arthroplasty ������������������������������������������������������������������������������������ 37
Caroline Vincelot Chainard and Henri Robert
5 
The Modern Indications for Medial UKA
the “Oxford Philosophy” Deciphered�������������������������������������������� 47
T. Gicquel, J. C. Lambotte, F. X. Gunepin, and Arnaud Clavé
6 How to Deal with a Fixed-Bearing Medial
Unicompartmental Knee Arthroplasty Implant?������������������������� 69
Camille Steltzlen and Nicolas Pujol
7 Principles of the Oxford® (Zimmer Biomet)
Unicompartmental Knee Arthroplasty (OUKA)�������������������������� 79
François Hardeman and Arnaud Clavé
8 Lateral Unicompartmental Knee Arthroplasty���������������������������� 95
Axel Schmidt, Matthieu Ollivier, and Jean-Noël Argenson
9 Kinematic Alignment Technique for Medial
Unicompartmental Knee Arthroplasty������������������������������������������ 107
Charles C. J. Rivière, Philippe Cartier, and Cédric Maillot
10 Computer-Assisted and Robotic Unicompartmental
Knee Arthroplasties ������������������������������������������������������������������������ 117
Constant Foissey, Cécile Batailler, Elvire Servien,
and Sébastien Lustig
11 
Full Polyethylene or Metal Back?�������������������������������������������������� 131
Hubert Lanternier and Arnaud Clavé

ix
x Contents

12 Recovery
 After Partial Knee Arthroplasty and Daycare
Surgery���������������������������������������������������������������������������������������������� 145
A. Sharma, H. A. Wilson, C. O’Neill, A. Alvand, N.
Bottomley, A. J. Price, and W. F. M. Jackson
13 Utility
 of Bilateral Single-Stage Unicompartmental
Knee Arthroplasty���������������������������������������������������������������������������� 149
Quentin Nicolas, Arnaud Clavé, Fabien Ros,
and Frédéric Dubrana
14 Sports
 and Functional Activities Following Unicondylar
Knee Arthroplasty���������������������������������������������������������������������������� 163
David A. Crawford and Keith R. Berend
15 Complications
 of Unicompartmental Knee Replacement������������ 169
Stefano Campi
16 What
 to Do If a Medial Unicompartmental Knee
Arthroplasty Fails���������������������������������������������������������������������������� 177
F. -X. Gunepin, L. Tristan, G. Le Henaff, O. Cantin,
and T. Gicquel
17 Results
 and Registry Data for Unicompartmental Knee
Replacements������������������������������������������������������������������������������������ 191
A. Rahman, A. D. Liddle, and D. W. Murray
 History of Unicompartmental
Prostheses 1
Frédéric Dubrana and Hoel Letissier

1.1 Arthroplastic Resection of a 16-year-old boy with infectious necrosis of

Eighteenth–Nineteenth his left shoulder [3, 4]. However, primacy is not
Century certain, because a year earlier Professor
Barthelemi Vigarous of Montpellier would have
It was not until the eighteenth century that the made the same intervention on a young man of
first descriptions of arthroplastic resections were 17 year old. This is described in a posthumous
described and taught. This intervention was not book published in 1820 by her son Professor
without risk for the patient, yet there are two his- Joseph-Marie Vigarous [5, 6]. However, there is
torical evocations, one of Hippocrates (460–377) an even older description of 1730 (Fig. 1.1). This
and the other of Paul D’Égine (VIIth century). If is a clinical case published by Johanne Daniele
Hippocratic corpus is vague and cautious, Paul of Schlichting. This surgeon removed the carious
Aegina mentions it and recommends it without head of the femur in a 14-year-old girl by dilating
specifying the indication and the technique: a fistulous opening on the hip (Fig. 1.2).
The resection by the saw of the protrusion of the Schlichting mentioned that his patient recovered
bone is controlled by the following conditions: if it in 6 weeks: [7]
cannot be reduced, if it is only a little needed that
In 1973. A 14-year-old girl’s hip joint is swollen,
it does not fit in, and if it is possible to remove it; it
painful, suppurative and disturbed. The surgeon,
is still a case of resection when it causes inconve-
due to the nature of the large hole, removes the
nience, injures the wattles in some way, makes the
entire head of the femoral bone, then inserts into
position of the limb bad, and at the same time is
the bone cavity a tincture of myrrh, and a juice...
stripped naked. In other circumstances, it does not
Finally, he binds the wound with a tight tie, and
matter whether or not to resect; because it is nec-
secures her for 6 weeks, so that after that the girl
essary to know that all the bones, which are com-
can walk freely. Here is the figure roughly sketched
pletely stripped [1]
by this surgeon:
Similarly, if the tip of the bone near a joint is sick,
1. It designates an unnamed dish.
it must be resected [2]
2. of the head, which must be removed from the
It was in 1768 that the first surgical descrip- ulceration.
3.the bone cavity, etc.
tion of arthroplastic resection was reported. It
was Charles White (1728–1813) who took care The first arthroplastic resection of the knee
was made in 1781 by Dr. Henry Park of Liverpool:
“Suffice it to note that the case caused him a lot
F. Dubrana · H. Letissier (*)
of problems and was accompanied by many
Department of Orthopaedic & Trauma Surgery,
University Hospital La Cavale Blanche, embarrassing circumstances, resulting mainly
Brest cedex, France from the difficulty of keeping the limb in a fixed

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2024 1

A. Clavé, F. Dubrana (eds.), Unicompartmental Knee Arthroplasty,
https://doi.org/10.1007/978-3-031-48332-5_1
2 F. Dubrana and H. Letissier

Fig. 1.2 Carious head of the femur in a 14-year-old girl

ends of the femur and tibia. Three months later the

limb was very solid, and the operated on was good
enough to no longer need care. He was still living
in 1782. ‘The person still lives,’ filkin Jr. said in his
letter to Binns (of Liverpool), and sometimes goes
Fig. 1.1 Front page of philosophical transactions. Royal to Liverpool, where, if I may, I will ask her to go
Society London see you [10].

This intervention was widely distributed, in

position [8]...” This intervention was published 1862 Dr. Heyfelder counted more than 176 cases
20 years later by Samuel Cooper; following this of arthroplastic knee resections in the world
publication Dr. Filkin claimed the anteriority of (Fig. 1.3). He devotes to it in his book Treatise on
the technique (Northwich, Cheshire), in a letter Resections the tenth chapter: Joint resections or
he wrote to Park he specified its anteriority of in contiguity [11]. For the hip, the arthroplastic
20 years (1762) [9]: resection procedure was less frequent, in 1860
Professor Léon Le Fort made an exhaustive
Filkin operated on a subject who had been carry- inventory, he found 86 publications of arthroplas-
ing a white tumor of the knee for several years, and
who, in a fall from a horse, fractured his kneecap. tic hip resections. It was not until the work of
The result was a suppuration of the article for Léopold Ollier [12, 13] at the end of the nine-
which amputation of the thigh was proposed. teenth century that a scientific approach to arthro-
Despite the subject’s dilapidated health, Filkin plastic resection developed. Léopold Ollier
proposed the resection of the decayed parts. After
practicing on the corpse, he performed resection developed the concept of arthroplastic resection
on 23 Aug. 1702. He found the ligaments very sub-capsulo-periosteal. This concept interested
affected, the cartilage very compromised, and the the shoulder, elbow, hip and knee. But, unlike the
articular extremities severely impaired, especially elbow the goal for the knee was not to obtain a
those of the tibia. He removed the patella and the
1 History of Unicompartmental Prostheses 3

result in human beings... Instead of cutting the

lateral ligaments, we carefully preserved them,
and we also spared everything that was healthy
from the perio-capsular sheath to accumulate
around the healing line of the bones as much
ossifiable tissue as possible. In this way, we con-
siderably increased the chances of bone healing,
and, in the event that mobility persisted, we
would have retained the tendon-muscular belt of
this new joint, that is, its means of resistance and
its organs of movement.” [15]
We will then have the semi-articular resections,
either femoral or tibial, and the partial resec-
tions... Partial resections of the knee will include,
according to the general division that we have
given of the resections, operations in which one
will remove either a condyle of femur or a condyle
of the tibia, or even a part only of the two opposite
ends by maintaining the contact of the two bones
by a certain extent of their normal articular sur-
faces. In this classification the total removal of the
patella will constitute a partial resection of the
knee [16].

Fig. 1.3 Treatise of resections. O. Heyfelder: 1863. fig.

1. Mr. Billroth's apparatus for resection of the knee (p. 93). 1.2 From Osteotomy
fig. 2. Apparatus of M. Esmarch for resections in general. to Arthroplastic
fig. 3. Apparatus of M. Bceckel for the resection of the
instep (p. 127)
Interposition

The first osteotomies of relaxations are due to

neo joint, but a bone fusion or even a stable J. Rhea Barton who performed two osteotomies,
fibrous ankylosis. However, Ollier specified that one for the hip in 1827 (Fig. 1.4) and one for the
the surgical risk (death, complications) was knee in 1837 (Fig. 1.5). These straightening oste-
important during resection of the knee: otomies were done in a very short time, 7 min for
The frequency of surgical shock after knee resec- the hip and 5 min for the knee. For the femur, it
tion has long been reported (Holmes). Even before was a subtrochanteric osteotomy and for the knee
the absorption of toxic substances could be a supracondylar osteotomy [4, 17, 18]. Barton
blamed, it was considered more dangerous from
this point of view than other joint resections [14]. hoped that after creating a neo joint, the bone
fusion would take place. For the hip, it was not so
In Ollier’s work, two notions are important to and the patient resumed his work with his neo
remember the scientific basis based on experi- joint:
mental studies and the notion of partial resection
The patient, upon whom this operation was per-
of the knee: “We had well demonstrated that it is formed, enjoyed the use of his artificial joint for 6
possible in young animals to reconstitute, after a years; during which period he pursued a business
subperiosteal resection, distinct femoral con- (trunk- making) with great industry, earning for
dyles, which can be articulated with a tibial pla- himself a comfortable subsistence, and a small
annual surplus [18].
teau of new formation, and play on it in flexion
movements and of more or less extensive exten- On the other hand, for the osteotomy of the
sions. But we did not propose to pursue the same knee, he obtained a consolidation by keeping a
4 F. Dubrana and H. Letissier

Fig. 1.4 J. Rhea Barton hip osteotomy in 1827

splint for 4 months. Six months before the opera- am on a visit to my Parents... Letter of the eighth is
tion, the patient (Mr. Seaman Deas) sent a long the first information I have had of your return. I
have the satisfaction and pleasure of saying to you
letter to his surgeon: now, that the operation you performed on my leg
Charleston, November sixth, 1837. My dear sir, — has been completely successful and has more than
Your letter of the eighth October, directed to me at realized my most sanguine anticipations. The small
Mobile, has just reached me at this place, where I abscess, which you dressed the day before we
1 History of Unicompartmental Prostheses 5

Rizzoli between 1853 and 1857. These interven-

tions consisted of a mandibular resection osteot-
omy without interposition during the initial
resection: “It has been several years since chance
provided Mr. Rizzoli with the opportunity to sur-
gically treat the immobility of the jaw. His first
observation dates back to 1853.” [20]
The notion of tissue interposition is found at
the end of the nineteenth century in Ollier’s book
(Fig. 1.6) [21]:
We can, after extracting it, reconstitute the shape
of the region by suturing the palatine periosteum at
the periosteum of the outer side of the bone. By
thus joining the horizontal palatine plane to the
external vertical plane, the separation of the nasal
Fig. 1.5 J. Rhea Barton osteotomies of the knee in 1837 and oral cavities is restored, which is very impor-
tant from the point of view of the functioning of the
organs of phonation and swallowing.
parted at Norfolk, continued open, and threw out,
from time to time, small pieces of bone, until the
August after, when the last piece was discharged;
the orifice then closed, and I have suffered no
material in convenience from it since. I am at pres-
ent well; the wound sound; and I feel no other
inconvenience in riding or walking, than what
arises from my knee joint being stilf, which was the
case before you performed the operation. I walk
without a stick or other aid, with the sole of the foot
to the ground, and my friends tell me, with but a
slight limp; and I have great pleasure in adding
that the leg and foot have increased considerably
in size, so as now to be nearly equal to the other....
Adieu and I am, my dear sir, very sincerely, your
friend. Seaman Deas. To Dr. J. Rhea Barton [19].

Certainly resecting, allowed to give mobility

in the neo joint of the hip, but quickly the sur-
geons wanted to put an interposition tissue to
facilitate the movement while limiting the risk of
bone ankylosis. Aristide Verneuil (1823–1895) is
falsely credited with the first peripheral interposi-
tion during the resection of the temporomandibu-
lar joint. This ambiguity is linked to the concept
of anaplasty and autoplasty that Verneuil is devel-
oping and where he can consider one of the
founders of modern plastic and reconstructive
surgery. For Verneuil, anaplasty is synonymous
with reconstructive surgery by natural means and
autoplasty with the help of prosthesis. However,
one does not find his writings the description of
tissue interposition in the temporomandibular
joint. In his text of 1860, it is as Verneuil specifies Fig. 1.6 Resection specimen on tuberculosis of the knee:
the description of the clinical cases of Professor 1883. Pr L. Ollier
6 F. Dubrana and H. Letissier

1.3 Tissue Interposition function has been, on the whole disappointing,

although the aim was lessened, the results were not
such as to awaken my enthusiasm.
In 1886, Ollier proposed tissue interposition by
periosteum for the hip and a few years later Gluck Willis Campbell in 1922 also focused on the
proposed to use skin. In 1918, Erich Lexer of the different techniques of interposition, he advised
University of Jena took stock of the interposi- against interpositions by animal tissue in favor of
tions of tissue, especially fatty in the hip and knee pediculated shreds (Fig. 1.7) [26]:
joints. Concerning the interpositions of fascia Pedunculated fascial flaps have been extensively
lata, the author specifies that the indications are employed between the articular surfaces, after
numerous [22]: remodelling or carving out a new joint. The proce-
dure has been discarded by a majority of experi-
To prevent adhesions: In this connection, from my enced operators in this field, interposition of
own experiences, fat transplantation plays a very animal membranes specially prepared, such as the
important role. Fat insertion to prevent rigidity of fragile membrane, Baer’s pig’s bladder, Allison’s
the joint after operations for ankylosis succeeds fascia, etc. While successes have been reported,
with best results in the loose joints of the arm, the disadvantage is that foreign body irritation
although favourable results. I have also been invites infection and the material is often excluded.
obtained in the hip and knee (Murphy, Lexer, Transplantation of free fascia lata, extensively
Ropke). In operations on the knee, fat pads pre- used by Putti, of Italy, and Russell Mac Ausland, of
vented recurrence of the fixation of the patella. Boston.
Likewise, fat implantation on the freshened acetab-
ulum has relieved the ankylosis of congenital dislo- In the same year, Campbell published a series
cation of the hip due to haemorrhage (Lexer). of 24 cases [flap of pediculated and free fascia
What changes take place in the flap of fat intro-
lata, pig bladder] (Figs. 1.8 and 1.9), of the 13
duced into the joint is not known. There was no
sign of the oft-mentioned watery-like fluid. Whether patients who could be assessed only five patients
or not it will make its appearance later, I cannot
say… The indications are numerous ... application
of fascial flaps between articular surfaces after
postoperative injury to the synovial membrane; in
mobilisation of joints; as a base for haemostatic
sutures in organs...

John Murphy stated in an article in 1913 that

he and his team had performed more than 60
arthroplasties including all joints, including 28
knees [23]: “The knee is the most difficult joint in
which to secure the perfect restoration of func-
tion and restoration of nearly normal joint
anatomy.”
In another article, John Murphy describes sur-
gical techniques and publishes many photographs
[24]. In 1918, Melvin Henderson of the Mayo
Clinic grouped cases from several centers and
published results for 121 patients. At a longer set-
back, 80 patients were evaluated as successful.
Nevertheless, at the end of his career, Henderson
was reserved about this type of surgery about
conservative arthroplastic surgery he wrote [25]:
I am free to confess that my own experience leaves
me still far from satisfied with my efforts along
these so-called reconstructive lines... I have used Fig. 1.7 Elbow interposition flap for arthrolysis. Willis
all the operations mentioned, with the result that C. Campbell. 1922
1 History of Unicompartmental Prostheses 7

Fig. 1.8 Knee mobilization device after arthrolysis of the

knee joint. Willis C. Campbell. 1924

Fig. 1.10 Fascia lata flap. ipsilateral: William

C. Campbell. 1924

fer (Figs. 1.9 and 1.10) and [27] stated that he had
operated on 16 patients in recent years with homo
or contralateral transplants. Only 12 patients
could be evaluated, he had ten good results and
two failures, his conclusion is as follows [28]:
“My first report was by no means encouraging
but from results obtain, especially during the last
year, arthroplasty of the knee is justifiable in
well-selected cases, with an excellent chance of
obtaining satisfactory motion.”
These results led him to develop another con-
Fig. 1.9 Fascia lata flap. contralateral: William cept, the joint interposition of inert materials.
C. Campbell. 1924 Following Smith-Petersen’s work on the hip, he
opted for vitallium. Smith-Petersen had made
had a mobility considered as good. In his last many tests before vitallium with the following
publication of 1924, Campbell described an orig- materials: glass, pyrex glass, viscaloids, Bakeites.
inal technique for contralateral lata fascia trans- In 1940, Campbell published his first two clinical
8 F. Dubrana and H. Letissier

cases of interposition of vitallium plate also performed in 1952 for a villonodular synovi-
called cap. It was a cast of the lower end of the tis, it was a single tibial piece fixed by a blade
femur whose size was evaluated by [29] X-rays. [32]:
The fixation was made by two posterior hooks He had a restoration of both tibial plateaus by a
and an anterior screw. He specified in his publica- prosthesis, a patellar prosthesis and an extensive
tion that he is also working on a tibial plate. joint debridement. Cellophane was interposed to
These interposition trays [30] had always been restore the periarticular gliding surfaces and the
suprapatellar pouch. The conclusions are as fol-
used in 1970 by Ranawat and Sbarbaro. From the lows: “With this prosthesis it is possible to restore
1950s, many types of arthroplasty will see the satisfactory function to most of the badly damaged
light of day and some anecdotal models testify knee joints that ordinarily would be subjected to
however to the medical reflection engaged, an arthrodesis. If this prosthesis will function sat-
isfactorily in these severely damaged knee joints, it
C. Rocher in 1952 proposed an arthroplasty of will function in any case other than that with an
the knee by two femoral heads in Judet acrylic infection.
[31]. However, all these attempts were
2. Mac Intosh published his first cases in 1966
disappointing.
[33], then in a second publication in 1972
about 130 surgeries. Mac Intosh [34] [Toronto]
was the first prosthesis which was implanted
1.4 Synthetic Interposition
in 1954. He made an oral presentation in 1965
at the annual meeting of the British
From the 1950s, three surgeons modified
Orthopaedists’ Association, on 58 cases, 51 of
Campbell’s concept by proposing interposition
which were bilateral. For Mac Intosh, the
trays such as Mc Keever in 1953, Mc Intosh in
ideal indication is rheumatoid arthritis, for
1954 and Spotarno.
osteoarthritis he prefers to do arthrodesis of
1. Mc Keever. the knee.
In1949, Mc Keever set up two patellofem- In a 1972 publication, Mac Intosh described
oral prostheses (left and right), and in 1955, how in 1954 he made his first case: “A
he published 40 cases. In 1960 in a posthu- 73-year-old woman was operated on at
mous article, Robert Elliot published the work Toronto General Hospital for knee arthrode-
of Mc Keever on the placement of unicom- sis.” During surgery, she realized that the val-
partmental tibial prosthesis about 76 interven- gus deformity could be reduced and the
tions (Fig. 1.11). The first surgery was tension of the lateral collateral ligament
improved stability to the knee. He decided
intraoperatively to put a hemiprosthesis, for
this he saws in two a knee prosthesis and
implanted only the external part. It was an
acrylic prosthesis from Dr. Sven Kiaer, Kund
Jansen from Copenhagen. The patient lived
12 years with this hemiprosthesis. Six other
patients were operated, four patients on six
had a result considered as good at 10 years.
The acrylic initially used for hip prostheses
following the work of the Judet brothers was
abandoned after reactions to foreign bodies at
the hip. The conclusion of the article is in
rheumatoid arthritis, hemi arthroplasty was
the procedure of choice, since tibial osteot-
omy was not a reasonable alternative.
Fig. 1.11 McKeever prosthesis
1 History of Unicompartmental Prostheses 9

All these implants were based on the con- dissapointing: Cartier et al. [35]. performed
cept of joint improvement related to the resto- six surgical revisions on the 17 surgeries, so
ration of joint line and the tension of collateral they did not recommend the use of this
ligaments. The prostheses of Mc Keever and implant and commercialization was stopped
Sbarbaro were stabilized by a keel or blade (Fig. 1.13).
inserted into a groove of the tibial plateau.
Mac Intosh’s device was free placed on a pre-
pared tibia, the shape of the prosthesis allowed
a ligament tension stabilizing the joint
(Fig. 1.12). The procedure for implanting the
devices was demanding and relatively lengthy.
However, two major problems persisted: the
lack of secondary fixation of the implants and
femoral cartilage wear. The last experiments
with this type of implant date back to the
2000s with the development of the Unispacer™
(Zimmer, Warsaw, USA). The results were Fig. 1.12 Mac Intosh prosthesis

Fig. 1.13 Gunston Polycentric Knee Prosthesis

10 F. Dubrana and H. Letissier

1.5 The Unicompartmental ity, 7% incidence of loosening, 3% incidence of

Prosthesis infection, and 4% incidence of tibiofemoral joint
pain. This double prosthesis was abandoned;
1.5.1 Cement however, surgeons at the Mayo Clinic used it for
unicompartmental damage. During surgery if
Frank H. Gunston, (Winnipeg, Canada) was only one side was worn, the prosthesis was
awarded a travel grant to study hip replacement implanted on the worn side. Bryan et al. pub-
in Wrightington with Sir John Charnley in 1966. lished a series of 207 knees, with 3 years of fol-
During this period, he worked on a concept of low-­ up, 83% of patients were satisfied [37].
arthroplasty of the cemented knee and published However, at the same time, other cemented uni-
in 1966 a biomechanical and clinical work on the compartmental arthroplasties were developed,
polycentric prosthesis (prosthetic simulation of such as Geomedic, Savastano, and Marmor.
normal knee movement). This prosthesis was
cemented, it was two cemented unicompartmen-
tal prostheses for femur and tibia. Upon his return 1.5.2 Insall’s Articles
to Canada, he worked with Peterson of the Mayo
Clinic to develop a polycentric knee prosthesis John Insall in the 1970s and 1980s was in a
(Howmedica [36], Rutherford, New Jersey, USA) dilemma, his position about unicompartmental
in 1970. This prosthesis was technically difficult prostheses was ambiguous while remaining open
to implant, and the clinical results were unsatis- [38]: “Unicompartmental replacement for osteo-
factory in long term (Fig. 1.14). In 1984, Lewallen arthritis of the knee is an attractive concept. It
published the results of the Mayo Clinic with seems reasonable that limited replacement would
10 years of follow-up, the survival rate was 66%. come closest to normal functional restoration.”
Patients had 13% instability due to ligament lax- The confusion was total, because the same year

Fig. 1.14 Unispacer™

1 History of Unicompartmental Prostheses 11

he published two contradictory articles! In –– Reducible frontal deformity: in varus less than
October 1976, an article with P. Walker [38], in 10° and in valgus less than 15°.
this article, his conclusion about the unicondylar
prosthesis was as follows: These extremely restrictive criteria carried by
The best results were seen in the lateral compart- the Robert Breck Brigham hospital team will be
ment arthroplasties. Such deformities may be the taken up by John Insall, then relayed in France by
only future indication for the use of this operation the Lyon teams, including Gérard Deschamps
as these knees do not do well when treated by tibial and Chol [43]:
osteotomy. However, when only the medial com-
partment is involved, osteotomy may still remain in Summary: Unicompartmental knee arthroplasty –
the treatment of choice. UKA – is designed for patients presenting arthritic
wear limited to a single medial or lateral tibio-
In the second article of 1976 [39], where he femoral compartment. The indication is based on
compared four models of prostheses his conclu- strict criteria. Wear must stem from degenerative
osteoarthritis or be secondary to aseptic necrosis
sion was without appeal: “We now think that
of the medial condyle. Inflammatory rheumatism is
there is no indication for this type of prosthesis a contraindication. Age and activity level should
and that the tibial osteotomy or the bicondylar be compatible with an indication for arthroplasty.
prosthesis should be chosen preferentially.” The body mass index should be less than 30 kg/m2.
The ligament system must be intact, particularly
In 1980, Insall et al. confirmed the impres-
both cruciate ligaments. Any pre-existing axis
sions of 1976 and in a study of a series of 32 uni- deformity should be moderate and the residual axis
compartmental prostheses at 5 years of hindsight, deformity, after correction of wear with a unicom-
they showed that despite the good results of the partmental tibial augmentation spacer, should not
exceed 7–10° varus or valgus.
initial clinical results deteriorate over time [40]:
“Unicompartmental prostheses are used in the For more than four decades, these criteria will
least advanced cases that give the least compli- become paradigmatic, defining a new global
cations, but the clinical results are not superior standard of indications.
to other prostheses.”
Following these publications, even if Insall
moderated his remarks in his book Surgery of the 1.5.4 The 1980s: The Awakening
knee. Total knee replacement [41] we can say that
for Insall that a good knee prostheses is a total Léonard Marmor for the Anglo-Saxon countries
knee arthroplasty. then Philippe Cartier in France extracted the
UKA from ostracism where John Insall and his
collaborators had locked it up. But it is the Oxford
1.5.3 Kozinn and Scott’s Criteria team that is pulling the UKAs out of the New York
rut while reviewing, thanks to a continuous and
In an article that will serve as a reference: Stuart scientific work spanning more than 30 years, the
Kozinn, Clare Marx, and Richard Scott proposed indications, the surgical technique, and the pros-
an algorithm of indications. In their [42] series, thetic models.
they reported 92% excellent and good results at
5.5 years of follow-up by respecting the follow- 1.5.4.1 Marmor
ing selection criteria: Léonard Mamor developed a unicompartmental
prosthesis that he implanted in 1974. The first
–– Over 60 years. publications were encouraging with 88% of
–– Less than 67 kg. patients satisfied at 2 years. However, consider-
–– Moderate activity. ing the wear and tear he advised against using
–– Little pain at rest. polyethylenes with a thickness of less than six
–– Flexum less than 5° and flexion greater than millimeters [44]:
90°.
12 F. Dubrana and H. Letissier

A follow-up of 2 years or more on 105 patients with and the Genesis. In 2007, in an oral publication
the Modular – Marmor – knee replacement [GECO] on 2170 cases, he detailed the factors of
revealed that 88 per cent of the patients had a suc-
cessful result. The complications and failures are recovery of UKAs. In his experiment, the failures
analysed in depth. Late loosening of the compo- of the models of the 1970s had [47, 48] essen-
nents were not observed except with the 6 mm tib- tially a technical cause [instrumentation and sur-
ial plateau. Pain relief was dramatic as well as geons], for the models of the 1980s, the failures
improved function, stability and motion.
were essentially mechanical linked in particular
His second major publication dates from to the poor quality of polyethylene (sterilization
1988, about 60 prostheses more than 10 years of with gamma rays, polyethylene too fine, metal
follow-up. Marmor in introduction clarified the back too rigid ...).
scope of the UKAs [45]: In his practice from the 90s complications
In the past decade, two concepts have caused con- were rare, in total 46 surgical repetitions out of
siderable controversy in orthopedic surgery of the 1170 UKAs, the main causes of failures are the
knee. Some orthopedic centers contend that oste- defects of initial indications.
otomy of the tibia is the procedure of choice for
unicompartmental gonarthrosis of the knee and
resist the concept of unicompartmental arthro- 1.5.4.3 The Oxfordian Revolution
plasty. "The other concept is that if unicompart- In a remarkable study of the biomechanics of
mental arthroplasty is necessary, the entire joint the Knee, John Goodfellow laid the foundations
should be replaced, since the uninvolved compart- of the prosthesis with a movable plate in 1978.
ment may develop arthritis in the future.
He first observes, as Aldabert Kapandji had
At 11 years of hindsight, he had in his series done before him, that the tension of the cruciate
70% satisfied patients and 87 painless knees. ligaments constrains the flexion/extension
However, a resounding trial will damage the movement of the knee. Goodfellow completed
image of Marmor’s UKA. In June 1983, Richard Kapandji’s model by associating the menisci
was ordered to pay Dr. Marmor $25,000 on the movement. He attributed to the menisci a double
prosthesis patent and $500,000 in personal dam- role: stabilization of the condyles and increase
ages. The Richard company was condemned fol- of the contact surface between the femur and the
lowing the manufacture between January 1973 tibia. To illustrate his point, he made a model
and April 1973 of 4000 medium tibial parts not (Fig. 1.15) that served as a two-dimensional
corresponding to the sizes of the concept. This model and introduced between the femur and
modification may lead to difficulties in surgical the tibia a “meniscal washer.” Meniscus substi-
placement and affect the clinical results, and a tutes by increasing the contact surfaces decrease
patient complaint will be filed [46]. the stresses. Starting from this model and fol-
lowing his articles on hip and elbow constraints
1.5.4.2 Philippe Cartier [49], he developed the concept of the Oxford
A few years later, Philippe Cartier in France [50] prosthesis and filed a patent [1977, US, pat-
reported a positive experience, with more than ent 21,905] (Fig. 1.16) jointly with John
90%, excellent or good results at less than 5 years O’Connor of Oxford and Nigel Shrive of
of decline. This author will successively use the Calgary [51]. For these designers, the Oxford
Marmor, the Mod III condylar, the Mansat Uni, prosthesis met the following specifications: con-
1 History of Unicompartmental Prostheses 13

Fig. 1.15 Biomechanical model: importance of the mobile meniscal wedge between the femur and the tibia

gruence during bending movements, little stress The mean wear rate of 0.02 mm/year measured in
at the interfaces, and reduced wear. the vivo study compares favorably with the pub-
lished results of polyethylene penetration for other
Initially, this prosthesis is implanted in bilat- forms of arthroplasty which use a metal-on-­
eral osteoarthritis, but from 1982 it was implanted polyethylene bearing. The value is approximately
in isolated unicompartmental osteoarthritis. In ten times less than the penetration rates of 0.1–
1988, Goodfellow et al. published their first clini- 0.2 mm/year reported for total hip arthroplasty
[54].
cal results, 36 months of follow-up [52], and then
10 years of follow-up. The authors confirmed the At 15 years of decline, the clinical results of
very good clinical results and show that pros- the Oxford 3 prosthesis remain excellent [55]
thetic wear was minimal, thus standing out from confirming the results of Price and Svard [56]
other prostheses, for example, the Lotus [53]. For and Liddle [57] who found a survival of the pros-
these authors, the rate of wear of the Oxford pros- theses of 92% at 20 years of follow-up for the
thesis remained negligible, well below other Oxford models 1 and 2 prosthesis (Figs. 1.17 and
arthroplasties: 1.18).
14 F. Dubrana and H. Letissier

Fig. 1.16 Oxford prosthesis: 1978 patent drawings

1 History of Unicompartmental Prostheses 15

Fig. 1.17 Oxford 3 prosthesis

Fig. 1.18 Oxford 3 prosthesis

1.6 Conclusion the surgeons wanted to be correctors, functionals,

and humans.
The concept of unicompartmental prosthesis was At the dawn of this twenty-first century, the
built over the past century, it is the emergence of unicompartmental prosthesis has just acquired its
an inventive, prudent, and respectful surgery, as letters of nobility, it is safe and reproducible. We
evidenced by the clinical and philosophical can say with Ahmadou Kourouma looking at the
descriptions of our elders. path traced and the finality:
We must not forget the ancient interventions, When you don’t know where you’re going, let you
all of which aimed at function, with empathy as a know where you’re coming from [58].
corollary; from Anthony White to Léopold Ollier,
16 F. Dubrana and H. Letissier

References 20. Aristide Verneuil. The creation of a false joint by sec-

tion or partial resection of the lower maxillary bone,
as a means of remedying a true or false ankylosis of
1. Hippocrates. Complete works. Fractures. Literary and
the lower jaw. Ed Labé. Arch Gen Med 15, Paris.
artistic union, Paris. 1955, p. 44.
1860; vol. 1, pp. 174–88.
2. Surgery of Paul of Aegina. Trad. by M. René Briau.
21. Treaty of resections and operations preservatives that
Paris. Ed. Victor Masson. 1855, pp. 315–7.
can be performed on the bone system. Leopold Ollier.
3. Philosophical Transactions. 1769, vol. LIX,
Ed. G. Masson, Paris. 1891; T3, l. p. 777.
pp. 39–45.
22. Lexer E. Free transplantation. Ann Surg.
4. Barton JR. On the treatment of anchylosis by the
1914;60(2):77–182.
formation of artificial joints. North Am Med Surg J.
23. Murphy JB. Original memories. Arthroplasty. Ann
1827;3:279–92.
Surg. 1913;57:611.
5. Works surgery—practical, civil and military.
24. Murphy JB. The classic: ankylosis: arthroplasty—
Barthelemi Vigarous. Ed. Tournel, Montpellier. 1812,
clinical and experimental. Clin Orthop Relat Res.
pp. 431–3.
2008;466:2573–8.
6. The analysis of Vigarous Sr.’s book, however, raises
25. Henderson MS. Role of fusion operations as applied
doubts about his son’s words, because Barthelemi
to the hip-joint. Br Med J. 1933;19(2):3789.
Vigarous perfectly described all the clinical cases he
26. Campbell WC. Arthroplasty of the elbow. Ann Surg.
cites and arranges them by chronology. In his book
1922;76(5):615–23.
he quotes and describes White's description before his
27. Campbell WC. Arthroplasty of the knee. Report of
own case.
cases. Am J Orth Surg. 1921;3(9):430–4.
7. Observations variœ medico-chirurgicœ a Johanne
28. Campbell WC. Arthroplasty of the knee. Ann Surg.
Daniele Schlichting, Med. & Chir. Doctore.
1924;76(5):615–23.
Obserrationis rarie medico chirurgica, Philosophical
29. Smith-Petersen MN. Arthroplasty of the hip. A new
Transactions, London. 1742, vol. XLII, p. 274.
method. J Bone Joint Surg. 1939;21:269. (We were
8. A Dictionary of Practical Surgery – Samuel Cooper.
unable to view the original).
Ed. J. Harper, New York. 1818, pp. 53–5.
30. Campbell WC. Interposition of vitallium plate
9. Barton J. On the treatment of anchylosis by the
sin arthroplasties of the knee. Report. Am J Surg.
formation of artificial joints. North Am Med Surg
1940;47:639–41.
Newspaper. 1827;3:279–92.
31. Rock. Knee arthroplasty by femoral heads in acrylic.
10. Treaty of resections and conservatives operations that
Bordeaux Elit. 1952;1:48. (We were unable to view
can be practiced on the system bony. T3, l. Ollier. Ed.
the original).
G. Masson. Paris. 1891, p. 207.
32. Dc. MC Keever. Tibial plateau prosthesis. Blinking.
11. Treatise on resections. O. Heyfelder Ed Baillère,
Orthop Relat Res. 1960;18:86–95. This is a post-
Translated from German, with additions and notes, by
humous article that was written by Dr. Robert
Dr. Eug. Boekel. Strasbourg. 1863, pp. 45–58.
B. Elliott (Houston), a year after Mc’s death. Keever.
12. Hip resection in cases of coxalgia. Leon the Strong.
Republished in 2005: The classic. Tibial plateau
Dissertation read at the Imperial Academy of
prothesis. Duncan C. Mc Keever. Blinking Orthop
Medicine. 1860, vol. XXV, pp. 445–583.
Relat Res. 2005;440:4–9.
13. Hip resection in cases of coxalgia. Leon the Strong.
33. Mac Intosh D. Arthroplasty of the knee in rheumatoid
Ed. Baillière, Paris. 1862, pp. 3–5.
arthritis. J Bone Joint Surg. 1966;48B:179.
14. Treated with resections and conservative operations
34. Myc Intosh D. The use of the hemiarthroplasty proth-
that can be performed on the bone system. T. Ollier.
esis for advanced osteoarthritis and rheumatoid arthri-
Ed. G. Masson, Paris. 1891; T3, l. p. 243.
tis of the knee. J Bone Joint Surg Br. 1972;54:244.
15. Treated with resections and conservative operations
35. Catier C, Turcat M, Jacquel A, Baulot E. The
that can be performed on the bone system. Leopold
Unispacer Tm unicompartmental knee implant: it’s
Ollier. Ed. G. Masson, Paris. 1891; T3, p. 210.
outcomes in medial compartment knee osteoarthritis.
16. Leopold Ollier. Treatise of resections and conserva-
Orthop Traumatol Surg Res. 2011;97:410–7.
tive operations that can be performed on the bone
36. Gunston F. Polycentric knee arthroplasty. Prosthetic
system. T3, l. Ed. G. Masson, Paris. 1891, pp. 220–1.
simulation of normal knee movement. J Bone Joint
17. J. Rhea Barton. A new treatment in a case of anchy-
Surg Br. 1971;53:272–7.
loses. The American Journal of the Medical Science.
37. Jones W, Bryan R, Peterson LF, Ilstrup
Philadelphia. Ed Carey, Lea, Blanchard. 1837,
D. Unicompartmental knee arthroplasty using poly-
pp. 332–40.
centric and geometric hemicomponents. J Bone Joint
18. J. Rhea Barton. A new treatment in a case of anchy-
Surg Am. 1981;63(6):946–54.
loses. The American Journal of the Medical Science.
38. Insall J, Walker P. Unicondylar knee replacement.
Philadelphia. Ed Carey, Lea, Blanchard. 1837, p. 333.
Clin Orthop Relat Res. 1976;120:83–5.
19. J. Rhea Barton. A new treatment in a case of anchy-
39. Insall J, Ranawat C, Aglietti P, Shine J. A comparison
loses. The American Journal of the Medical Science.
of four models of the total knee replacement proth-
Philadelphia. Ed Carey, Lea, Blanchard. 1837,
eses. JBJS. 1976;58-A(6):754–65.
pp. 337–8.
1 History of Unicompartmental Prostheses 17

40. Insall J, Aglietti P. A five to seven-year follow-up 51. https://patentimages.storage.googleapis.com/1a/94/

of unicondylar arthroplasty. J Bone Joint Surg Am. a9/71e1f4ea6a25cc/US4085466.pdf.
1980;62(8):1329–37. 52. Goodfellow J, Kershaw C, D’A Benson M, Connor
41. J. Insall. Surgery of the knee. Total knee replacement. JO. The Oxford knee for unicompartmental osteoar-
Ed. Churchill Livingstone, New York. 1984, p. 615. thritis. The first 103 cases. J Bone Joint Surg (Br).
42. Kozinn S, Marx C, Scott R. Unicompartmental knee 1988;70(5):692–701.
arthroplasty. A 4.5-6-year follow-up study with 53. Witvoët J, Peyrache M, Nizard R. Single-compartment
a metal-backed tibial component. J Arthroplast. “Lotus” type knee prosthesis in the treatment of later-
1989;4(Suppl):S1–S10. alized gonarthrosis: results in 135 cases with a mean
43. Deschamps G, Chol C. Fixed-bearing unicompart- follow-up of 4.6 years. Rev Chir Orthop Reparatrice
mental knee arthroplasty. Patients’ selection and Appar Mot. 1993;79(7):565–76.
operative technique. Orthop Traumatol Surg Res. 54. Price A, et al. Ten years in vivo wear measurement
2011;97(6):648–61. of a fully congruent mobile bearing unicompart-
44. Marmor L. The modular (Marmor) knee. Clin Orthop mental knee arthroplasty. J Bone Joint Surg (Br).
Relat Res. 1976;120:86–94. 2005;87-B:1493–7.
45. Marmor L. Unicompartmental knee arthroplasty. Ten- 55. Pandit H, Hamilton T, Jenkins C, Mellon S, Dodd C,
to 13-year follow-up study. Clin Orthop Relat Res. Murray D. The clinical outcome of minimally inva-
1988;226:14–20. sive phase 3 Oxford unicompartmental knee arthro-
46. Nellie Vossler, Plaintiff and respondent, v. Richards plasty: a 15-year follow-up of 1000 UKAs. Bone Joint
Manufacturing Company, Inc., Defendant and J. 2015;97B:1493–500.
Appellant. Docket No. 6436. Court of Appeals of 56. Price A, Svard U. A second decade lifetable sur-
California, Fifth District. June 15, 1983. vival analysis of the Oxford uni-compartmental knee
47. Cartier P, Cheaib S. Unicondylar knee arthroplasty: arthroplasty. Clin Orthop Relat Res. 2011;469:174–9.
2–10 years of follow-up evaluation. J Arthroplasty. 57. Liddle A, Judge A, Pandit H, Murray D. Determinants
1987;2:157–62. of revision and functional outcome following uni-
48. https://docplayer.fr/128712875-­R eprise-­d e-­p -­u -­c -­ compartmental knee replacement. Osteoarthr Cartil.
geco-­2007-­chirurgie-­de-­reprise-­janvier-­2007.html. 2014;22:1241–50.
49. Bullough P, et al. The relationship between degen- 58. Ahmadou Kourouma : Ivorian writer (1927/1994)
erative changes and load-bearing in the human hip. J born in Côte d'Ivoire, he comes from the Malinké
Bone Joint Surg Br. 1973;55B:746. ethnic group. He was an infantryman in Indochina
50. Goodfellow J, O’Connor J. The mechanics of the from 1950 to 1954. “Waiting for the vote of the wild
knee and prosthesis design. J Bone Joint Surg (Br). beasts” Seuil, 2000.
1978;6G-8:358–69.
 The Disappearing
Unicompartmental Knee 2
Prostheses

Samuel Laurent, Baptiste Montbardon,

Arnaud Clavé, and Frédéric Dubrana

2.1 Introduction Unfortunately, the literature on this subject is

very poor and unlike Australia or the Scandinavian
The improvement in the functional results of uni- countries, we do not have quality registers on
compartmental knee surgery over the last three unicompartmental implants.
decades is closely linked to the development of We will present some of these unicompart-
prosthetic implants, but also to the reproducibil- mental prostheses (UKA) that have disappeared,
ity of this surgery thanks to more efficient ancil- while trying to understand if their disappearance
lary equipment. is linked to a design defect or to the tumult of the
These improvements have been made possible world economy, punctuated by the takeover bids
by the existing demand for these new prostheses, and counter-takeovers of the world giants pro-
but also because of financial competition between ducing our orthopaedic equipment.
the different laboratories involved in the market. This chapter does not pretend to be exhaustive
The first versions of these prostheses have or to be a peremptory judgement on the implants
now given way to new-generation implants that mentioned.
more faithfully reproduce the anatomical and
biomechanical characteristics of the knee joint.
This chapter will focus on the discontinued 2.2 History of Major Chip
unicompartmental prostheses. Companies
There are many reasons for the discontinua-
tion of these devices: poor clinical results, lack of The choice of these different companies was
financial profitability, technological innovation, made on the basis of data that some laboratories
laboratory consolidation or restructuring, etc. have communicated to us (Table 2.1).

S. Laurent · B. Montbardon · F. Dubrana 2.2.1 Depuy Synthès

Department of Orthopaedic Surgery and
Traumatology, Brest University Teaching Hospital At the beginning of the 90s, several laboratories
“La Cavale Blanche”, Brest, France were producing unicompartmental prostheses of
A. Clavé (*) French design. Among them, the LANDANGER
Department of Orthopaedic Surgery and laboratories which produced the Goeland, and
Traumatology, Saint George Private Hospital,
Nice, France the Roannais MEDINOV AMP which produced
the Gonometric.
LaTIM, INSERM-UBO UMR 1101, Brest, France

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2024 19

A. Clavé, F. Dubrana (eds.), Unicompartmental Knee Arthroplasty,
https://doi.org/10.1007/978-3-031-48332-5_2
20 S. Laurent et al.

Table 2.1 UKA status in 2020 II and the ZUK (Zimmer Uni Knee), bought the
UKA CENTERPULSE company.
NK-Uni Then allegretto Now Alpina Uni In order to rationalize its ranges, the produc-
Repicci tion of the uni NK2 and Allegretto has been
Miller Now ZUK Commercialized stopped, these will be replaced by the ZUK.
Galante by LIMA In 2014, ZIMMER bought out BIOMET,
Uni
Oxford Commercialized by
which produced two unicompartmental the
UKA Zimmer Biomet Oxford and the Alpina Uni, and the creation of
Persona the giant ZIMMER BIOMET.
UKA Following the takeover of BIOMET and due
Alpina to anti-trust laws, in some countries ZIMMER
UKA
had to cede the exploitation rights of ZUK,
acquired in 2015 by the Italian company LIMA.
These two laboratories were bought in 1996, From now on, ZIMMER BIOMET markets 3
by the American-Swiss group SYNTHES, which unicompartmentals:
then decided to stop the production of Gonometric
in 1997 to rationalize its range. –– Oxford.
With the acquisition of the Depuy Synthès –– PPK (evolution of the ZUK through the
laboratory by the American pharmaceutical Persona range).
group Johnson and Johnson, the unicompartmen- –– Alpina Uni.
tal prostheses of French design have been aban-
doned, the project was to create a new prosthesis Thus, with globalization, we are witnessing a
with American surgeons, the Preservation. progressive takeover of companies producing
French and European unicompartmentals, which
are gradually coming under the American flag.
2.2.2 Zimmer Biomet American flag. But their conception remains
for the most part European (Oxford of English
At the origin of this giant we find 2 companies conception, Alpina uni France, and PPK of
belonging to the Swiss Group SULZER: French design).
ALLOPRO which produced the Uni NK2 and For example, Zimmer manufactures most of
PROTEK: which produced the Uni Allegretto. its EMEA products in Winterthur, Switzerland
These two companies were merged in 1995 to with subcontractors in France.
create the Swiss group: SULZER MEDICA.
At the beginning of the year 2000, SULZER
MEDICA decided to create the company 2.2.3 Smith & Nephew
CENTERPULSE in order to manage only its
Orthopaedic branch. This was also a way to make Among the laboratories that still produce uni-
disappear SULZER’s name in the United States, compartmental implants in Europe, we find
which had become embarrassing due to lawsuits SMITH AND NEPHEW, which produced the
over defective hip prostheses and for which the Marmor Modular Knee, the MOD 3 and then the
group was condemned by the American justice Genesis, and recently the Genesis and more
system and suffered a loss of 793 million euros in recently the Accuris Uni and the Journey Uni for
2001. which the technique, the ancillary, and the shape
Two years later, the ZIMMER company, of the implant have been revised in a new version
which was already producing the Miller Gallante to be more guiding and reproducible.
2 The Disappearing Unicompartmental Knee Prostheses 21

2.3 Unicompartmental Devices The Gonometric line included:

Withdrawn from the Market
–– 5 sizes of internal and 4 sizes of external fem-
The choice of these examples was based on the oral condyles
existence of articles in the scientific literature, –– 5 sizes of internal or external tibial bases
allowing us to support our discussion. This list is –– 5 polyethylene thicknesses ranging from 8 to
not exhaustive. 12 mm

2.3.1 Goeland 2.3.3 Preservation

The Goeland was a cup prosthesis, produced Preservation was marketed in the United States
from 1988 to 1998. The femoral implant could be from 2001 to 2007 by DePuy (Johnson &
cemented or unsealed, and consisted of an Johnson).
anchoring pin associated with an anti-rotational This prosthesis was composed of a femur
fin and a flat polyethylene insert fixed on a tibial made of a chrome-cobalt alloy associated with
metal-back. tibial implant with a full polyethylene or metal-­
back with a polyethylene insert that could be
either fixed or mobile.
2.3.2 Gonometric In the series by Marini [1], out of 38 Uni
Preservation cases, 15 had to be rehabilitated for
The Gonometric (Fig. 2.1) was a cemented pros- loosening of the femoral component. No correla-
thesis with a cup made of a Chrome- Cobalt alloy, tion was found between failure and preoperative
produced from 1991 to 1997. or postoperative flexion, the difference obtained,

Fig. 2.1 Gonometric UKA

22 S. Laurent et al.

the postoperative tibial slope, the alignment of 2.3.5 PCA

the femoral component, postoperative range of
motion, gender or BMI of the patient. Designed by two Swedish surgeons, A. Lindstrand
and A. Stenstrom [4], it was commercialized by
Stryker Howmedica Osteonics in 1983.
2.3.4 Miller-Galante II This cut prosthesis could be inserted with or
without cement. The bone-prosthesis interface
Marketed by Zimmer (Warsaw, Ind. USA) was equipped with a PCA microbead blasting
between 1992 and 2008, this was a cemented, system using 800-micron balls.
fixed-plate, cup prosthesis (Fig. 2.2). The femoral component, thanks to its anatom-
The Miller-Galante II was designed to be ical shape, ensured good compatibility between
cemented only. For this purpose, it had benefited the prosthesis and the patella in a range of flexion
from two studs on the femoral component and from 0° to 130°.
three on the tibial component, to improve stabil- There were 4 sizes for the femoral, tibial, and
ity. The thinnest polyethylene that could be used polyethylene implant.
was 8 mm thick and was sterilized by gamma The femoral and tibial parts were designed to
irradiation. be compatible with each other regardless of their
The series of Koskinen [2], on the medium-­ size.
term survival of Miller-Galante II showed poor The series by Gacon [5] showed a 9.5% revi-
results with 86% survival at 7 years. In fact, 8 out sion rate (65/772) at 2 years follow-up. The fail-
of 46 prostheses had to be revised for premature ures were mainly due to femoral loosening (35
wear of the polyethylene, with the hypothesis of cases) and premature wear of the polyethylene
a poor quality of the polyethylene due to steril- (20 cases).
ization by gamma irradiation. Skyrme’s series [6], showed a revision rate of
However, Berger et al. [3] showed very good 42% at 4 years, with also as main failures main
results in their prospective series with 98% sur- failures: loosening of the femoral implant and
vival at 10 years average follow-up. rapid degradation of the polyethylene (Figs. 2.3
The marketing of the Miller-Gallante II was and 2.4). The hypothesis of this wear was the
stopped in 2008 in favour of the Physica ZUK® poor quality of the polyethylene which was hot-­
prosthesis (Zimmer and nowadays LIMA pressed during its manufacture to obtain a smooth
Corporate. UD, Italy). polyethylene. This manufacturing process caused

Fig. 2.2 Miller-Galante II

(courtesy of Zimmer Biomet)
2 The Disappearing Unicompartmental Knee Prostheses 23

Fig. 2.3 Loosening of the

femoral component of a PCA Uni

Fig. 2.4 Polyethylene wear of tibial component found during a revision procedure

early delamination of the polyethylene. The continuation of the PCA prosthesis due to the
thickness of the polyethylene has also been wear rate of the polyethylene and the resulting
blamed. high revision rate.
Other series, notably those of Christensen [7]
and Lindstrand [8], have recommended the dis-
24 S. Laurent et al.

Fig. 2.5 Marmor UKA

2.3.6 Marmor Modular Knee

Between 1970 and 1972, Dr. Leonard Marmor, in Fig. 2.6 MOD3 UKA
association with Richards Manufacturing
Corporation, developed a prosthesis known at the According to Dr. Leonard Marmor, the main
time as the Marmor Modular Knee (Fig. 2.5). causes of failure of the Marmor Modular Knee
This was a pure resurfacing prosthesis, with [14] were as follows:
minimal condylar femoral resection. The
implants were cemented, and the femoral implant –– At the level of the femoral implant, the evi-
had a central stud [9]. dence of a stress on the posterior edge of the
The tibial tray was made of 6 mm thick PE. implant at 90°, as well as a posterior gap most
This implant was developed with the philoso- often filled by cement.
phy that “a unicompartmental prosthesis is not –– In the tibial implant, the cemented full poly-
half of a total prosthesis” [10]. ethylene component was only 6 mm thick and
The widespread use of the Marmor Modular was also a source of loosening.
Knee was halted in 1973 by an unfortunate engi- –– Given Dr. Leonard Marmor’s contributions
neering error, as the final implants were larger and innovation in the design of the compo-
than the trial ones. This led to lawsuits. nents and surgical technique, he is still consid-
In 1976, Marmor published a series with a ered by many to be the father of the modern
2-year follow-up with a stable and pain-free joint CUP [9].
in 88% of cases [11], and a few years later a –– The Marmor CUP was replaced in 1984 by the
series at 13 years follow-up with 86% good MOD3 (Fig. 2.6).
results [12].
Cazanave and Cartier [13], in their series of 69
Marmor PUCs, showed a survival rate of 93% at 2.3.7 MOD 3
12 years’ follow-up. The functional scores at the
same follow-up were excellent in 57% of cases, This was the evolution of the Marmor Modular
good in 20%, fair in 7%, and poor in 7%. Knee (Fig. 2.7 and Table 2.2), marketed by Smith
2 The Disappearing Unicompartmental Knee Prostheses 25

OLD NEW However, the results obtained were not as

expected. Out of 790 implants of this type, an
early failure rate of 6% has been demonstrated.
90°
The real rate would probably be even closer to
130°
probably even closer to double if we took into
account the information provided by other
Stress on the posterior edge Squatting possible centres.
at 90° up too 130° However, Cohen [16] reported a 95% survival
rate at 8 years, with only one case of early malpo-
sition requiring revision by total prosthesis.
The main causes found for these failures were
mechanical with accelerated wear of the
polyethylene.
Gap fille with cement Posteriorly self-locking
The primary reasons put forward and high-
lighted at the time were gamma ray sterilization
Fig. 2.7 Evolution of the Marmor design towards the (instead of ethylene oxide) [17], a polyethylene
MOD3 UKA that was too thin (7.5 mm) and a metal-backed
CrCo too rigid.
Table 2.2 Differences between Marmor and MOD3 These failures led to new reflections concern-
UKAs ing the design of the implants with the replace-
Evolution of the Marmor UKA to the MOD3 ment of the MOD3 by the Genesis (Fig. 2.8):
Implant MARMOR MOD3
Femoral Constraint on the Flexion >130°
posterior wall of the without any
implant at 90° constraint 2.3.8 Genesis
Posterior gap between Posterior design
implant and bone was with an auto Marketed between 1991 and 2006 by Smith &
filled up by cement locking system Nephew, (Watford, UK). At the femoral level, the
Central keel with
a press-fit main differences with MOD3 were as follows:
fixation.
Tibia 9 mm cemented full Chrome cobalt –– Between each size, a proportional increase
PE metal back between the width and length of the implant.
Tibial cut has to be Tibial tray of
–– A tapered and superiorly contoured compo-
≥9 mm 7.5 mm
nent appearance to minimize the risk of patel-
lar impingement.
& Nephew (France) from 1984 to 1994. It was –– A wide range of sizes, offering seven possi-
bought by Richards Manufacturing Corporation bilities, each with two implant thicknesses
in 1986. femoral implants: standard 4 mm and 7.5 mm
Modifications to the femoral implant were for lateral condylar dysplasia or post-­traumatic
intended to allow for more than 130° of uncon- reconstruction.
strained flexion, as well as a self-locking poste-
rior design with a central keel with a central pin At the tibial level, the main evolution was the
for press-fit fixation (Fig. 2.7 and Table 2.2). use of titanium for metal-back where the advan-
In the tibial implant, the main change was an tages over chrome-cobalt were greater strength
improved fixation with a metal-back support in and less rigidity [18].
chrome-cobalt, and a fixed 7.5 mm polyethylene The aim was to limit the thickness of the
(Fig. 2.7 and Table 2.2). metal-back in favour of polyethylene [15].
According to Cartier [15], MOD3 could be Cazenave [13] reported a survival rate of 94%
used in younger and more active patients. at 10 years, followed by a progressive decrease to
26 S. Laurent et al.

THE RANGE

Fig. 2.8 Genesis (courtesy of Smith and Nephew)

88% at 12 years in his series of 1173 knees oper- It represents a considerable advance in the tech-
ated on from 1991 to 2006. nical reproducibility of unicompartmental pros-
The main complications could be divided into thesis implantation. This prosthesis is still on the
two categories: problems of technical origin and market.
inappropriate surgical indications. Notable advantages over previous ancillaries:
The study of polyethylene wear, on the other
hand, was quite reassuring, with wear not exceed- –– Pre-balancing of the knee joint is performed
ing of 1 mm at any time during the revision. before the bone cuts with the help of intra-­
All of these complications had in common the articular wedges.
simplicity of their revision, contrary to the opin- –– Posterior femoral and tibial resection is per-
ion expressed by Douglas, Padgett, and Stern formed at the same time and at the ideal
[19], due to the use of a resurfacing implant, the level.
very small amounts of cement used, and the early –– The parallelism between the tibial cutting sur-
diagnosis of loosening follow-up. face and the femoral resurfacing level is
The Genesis was replaced in the early 2000s respected, by using an electric femoral resur-
by the Accuris (Fig. 2.9), developed for a mini- facing drill that can be adapted to the tibial
mally invasive approach by Leo Pinczewski [15]. trial base.
2 The Disappearing Unicompartmental Knee Prostheses 27

Human Factors
For the patient and his surgeon

–– Functional results
–– Importance of the revision rate

For the surgeon

–– Simplicity in its use

–– Reliability
–– Reproducibility
–– Revision rate

Economic factors
Probably the most important for companies.
The great paucity of literature in this field
leads us to wonder about the lack of serious eval-
uation by these different companies.
In addition, we note an impoverishment of
Fig. 2.9 Accuris UKA (courtesy of Smith et Nephew) design companies in the field of orthopaedics in
France and in Europe.
Concerning the implants, the femur is made of The acquisition of French and European
oxidized zirconium, which seems to cause less SMEs with technological innovations has been a
wear of the polyethylene than cobalt-chromium strong trend for several years.
[20]. There are many examples of this: the takeover
by an American firm of the Montpellier- based
company Medtech, specialized in biotechnology,
2.4 Conclusion or the Aube-based company LDR Medical, spe-
cialized in spine surgery.
The various data analysed show us that the dis- In another register, we can mention IMASCAP
continuation of the marketing of a unicompart- created in 2009, by Jean CHAOUI, in the field of
mental prosthesis results from a combination of computer-assisted shoulder surgery at the labora-
several factors: tory of medical information (LaTIM-Telecom
Factors Related to the Implant Bretagne, Brest).
Its components His initiative was supported by surgeons
experts in shoulder surgery and by Telecom
–– Type of sterilization Bretagne. This work aroused the interest of sev-
–– Thickness of the polyethylene eral major players in the shoulder arthroplasty
–– Characteristics of certain components (e.g. market and in 2018 IMASCAP was acquired by a
rigidity of CrCo) world leader in shoulder prostheses: the American
company.
Its evolution in time Wright Medical (the latter having itself been
acquired more recently by Stryker).
–– Survival Most of these conglomerates are under
–– Mechanical loosening of femoral and tibial American leadership, and more than 2/3 of the
implants world market for orthopaedic implants is held by
–– Premature wear of polyethylene six companies.
28 S. Laurent et al.

This situation may lead to fears of a future and mental knee prosthesis. Acta Orthop Scand.
progressive sidelining of French and European 1990;61(6):578–9.
8. Lindstr A, Stenström A. Polyethylene wear of the
designers, which could lead to an impoverish- PCA unicompartmental knee: rospective 5 (4-8) year
ment of both the industrial and intellectual of the study of 120 arthrosis knees. Acta Orthop Scand.
design of prosthetic implants. 1992;63(3):260–2.
There is also a risk linked to a standardization 9. Johannes Plate MD, Ali Mofidi MB, Sandeep
Mannava MD, Cara Lorentzen MD, Beth Smith P,
of products and concepts by American ­companies, Thorsten Seyler MD, et al. Unicompartmental knee
sometimes focused on dogmatic ideas and arthroplasty: past, present, future. Reconstr Rev. [cité
concepts. 10 mai 2020];2(1). Disponible sur: http://reconstruc-
Historically, this does not reflect their cul- tiveview.org/ojs/index.php/rr/article/view/15
10. Grelsamer RP, Cartier P. A unicompartmental knee
ture but may have been dictated by judicial replacement is not “half a total knee”: five major dif-
influence, resulting in a certain chilliness, a ferences. Orthop Rev. 1992;21(11):1350–6.
lack of ingenuity and evolution both in terms 11. Marmor L. The Modular (Marmor) knee: case report
of surgical techniques but also in terms of with a minimum follow-up of 2 years. Clin Orthop
Relat Res. 1976;120:86–94.
solutions. 12. Marmor L. Unicompartmental knee arthroplasty. Ten-
All orthopaedic surgeons need to think deeply to 13-year follow-up study. Clin Orthop Relat Res.
in order to consider strong actions, necesasary to 1988;226:14–20.
reverse a potentially disastrous trend. 13. Cazenave A. Cazenave A, Cartier P- prothèse uni-
compartimentale marmor évolution genesis- In
« Arthroplastie du genou de 1°intention : expériences
cliniques ». Ed SAURAMPS médical 2008, pp. 155–
References 162. 7.
14. Marmor L. Unicompartmental arthroplasty of the
1. Mariani EM, Bourne MH, Jackson RT, Jackson ST, knee with a minimum ten-year follow-up period. Clin
Jones P. Early failure of unicompartmental knee Orthop Relat Res. 1988;228:171–7.
arthroplasty. J Arthroplasty. 2007;22(6):81–4. 15. Cartier P, Khefacha A. Fixed-bearing unicompart-
2. Koskinen E, Paavolainen P, Eskelinen A, Harilainen mental knee arthroplasty. In: Bellemans J, Ries
A, Sandelin J, Ylinen P, et al. Medial unicompartmen- MD, Victor JMK, éditeurs. Total knee arthroplasty
tal knee arthroplasty with Miller-Galante II prosthe- [Internet]. Berlin/Heidelberg: Springer-Verlag; 2005
sis: mid-term clinical and radiographic results. Arch [cité 10 avr 2019], pp. 317–21. Disponible sur: http://
Orthop Trauma Surg. 2009;129(5):617–24. springer.com/10.1007/3-­540-­276580_50
3. Berger RA, Meneghini RM, Jacobs JJ, Sheinkop 16. Cohen I, Zeev F, Hendel D, Blankstein A, Chechick A,
MB, Della Valle CJ, Rosenberg AG, et al. Results of Rzetelny V. Unikompartimenteller kniegelenkersatz
unicompartmental knee arthroplasty at a minimum mit zementierter MOD3TM-prothese: a prospective
of ten years of follow-up. J Bone Joint Surg Am. study. Eur J Trauma Emerg Surg. 1999;25:287–93.
2005;87(5):999–1006. 17. Chaigneau M. [Advantages and disadvantages of eth-
4. Lindstr A, Stenström A, Lewold S. Multicenter study ylene oxide sterilization of medicosurgical equipment
of unicompartmental knee revision: PCA, Marmor, and pharmaceutical products]. Bull Acad Natl Med.
and St Georg compared in 3,777 cases of arthrosis. 1983;167(6):627–31.
Acta Orthop Scand. 1992;63(3):256–9. 18. Mezache F, Mazouz H, Amrani H. Principes bioméca-
5. Gacon G. Résultats des prothèses unicomparti- nique de la prothèse du genou. 2013;5.
mentales PCA du genou. Orthopédie Traumatol. 19. Padgett DE, Stern SH, Insall JN. Revision total knee
1992;2(1):61–70. arthroplasty for failed unicompartmental replace-
6. Skyrme AD, Mencia MM, Skinner PW. Early failure ment. J Bone Joint Surg Am. 1991;73(2):186–90.
of the porous-coated anatomic cemented unicompart- 20. Spector BM, Ries MD, Bourne RB, Sauer WS, Long
mental knee arthroplasty: a 5- to 9-year follow-up M, Hunter G. Wear performance of ultra-high molec-
study. J Arthroplasty. 2002;17(2):201–5. ular weight polyethylene on oxidized zirconium total
7. Christensen OM, Christiansen TG, Johansen knee femoral components. J Bone Joint Surg Am.
T. Polyethylene failure in a PCA unicompart- 2001;83-A(Suppl 2 Pt 2):80–6.
 Anteromedial Osteoarthritis
and UKA 3
Samuel W. King, Bernard H. Van Duren,
and Hemant Pandit

3.1 Introduction UKA are increasing but the procedure still only
represents 8–12% of all knee arthroplasties [8–
Symptomatic knee osteoarthritis occurs in up to 10]. This chapter discusses the history of antero-
16.7% of people over 45 years of age [1]. Total medial osteoarthritis, the development of the
knee arthroplasty (TKA) is an extremely success- UKA, and its results, indications, and
ful and popular procedure for the treatment of contraindications.
end-stage knee osteoarthritis, and demand con-
tinues to increase significantly [2, 3]. It involves
the replacement of both tibiofemoral articular 3.2 Anteromedial Osteoarthritis
compartments but disease is often only present in
one. In approximately 60% of patients knee Knee osteoarthritis is often present only in the
osteoarthritis is restricted to the medial compart- anterior part of the medial tibiofemoral compart-
ment only [4–7]. A significant proportion of these ment. Ahlback et al. studied 370 knees with
patients may be treated with a unicompartmental osteoarthritis and found that 85% of these had
knee arthroplasty (UKA) providing they fulfil degeneration limited to only one compartment,
certain criteria. Its primary indication is antero- and that the medial compartment was 10 times
medial osteoarthritis of the knee, and require- more likely to be affected than the lateral [6].
ments include an intact anterior cruciate ligament Further radiographical study of 94 patients with
(ACL). Proponents of UKA cite its improved symptomatic knee osteoarthritis by Hernborg and
clinical outcomes and fewer side effects. Rates of Nilsson also demonstrated a predominance of
medial compartment disease. Ninety percent of
patients had disease isolated to this region with
S. W. King · B. H. Van Duren very little long-term progression laterally [11].
Leeds Institute of Rheumatic and Musculoskeletal Studies of the knee at this time generally used
Medicine, University of Leeds,
Leeds, West Yorkshire, UK anteroposterior plain knee radiographs and lat-
eral films were rarely utilised, often suggested to
H. Pandit (*)
Leeds Institute of Rheumatic and Musculoskeletal be of little clinical use [12].
Medicine, University of Leeds, Anteromedial osteoarthritis of the knee (also
Leeds, West Yorkshire, UK known as anteromedial gonarthrosis) was first
Nuffield Department of Orthopaedics, Rheumatology proposed as a distinct clinicopathological entity
and Musculoskeletal Sciences (NDORMS), by White et al. in their 1991 study [13]. They
University of Oxford, Oxford, UK studied resected tibial plateaus in 46 patients who
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2024 29

A. Clavé, F. Dubrana (eds.), Unicompartmental Knee Arthroplasty,
https://doi.org/10.1007/978-3-031-48332-5_3
30 S. W. King et al.

had undergone UKA for medial compartment ciency, wear areas were larger and had migrated
osteoarthritis. The ACL was intact in all speci- more posteriorly and progressed to the lateral
mens, degenerative lesions were centred anteri- compartment. As the authors hypothesised, this
orly on the medial tibial plateau, and posterior suggests that ACL rupture allows posterior femo-
cartilage was spared. These degenerative findings ral subluxation, posterior tibiofemoral contact,
were consistent with changes noted on lateral and posterior progression of medial compartment
knee radiographs. The authors hypothesised that wear. Radiographical correlation has also been
the anatomical findings explained their clinical noted. A study of 200 knees demonstrated a 95%
examination. In anteromedial osteoarthritis, genu correlation between preservation of posterior
varum is present on extension but is correctable medial tibial plateau on lateral radiograph and
with knee flexion. Intact cruciate ligaments and intra-operative findings of an intact ACL. The
lateral articular surfaces allow the medial femoral authors also found that 100% of knees with
condyle to roll posteriorly in flexion and articu- degenerative changes noted on lateral radiograph
late with the posterior medial tibial plateau. The had a deficient ACL [17].
articular cartilage, and therefore also tibial pla- The mechanism of chronic ACL rupture in
teau height, is preserved here. Preservation of knee osteoarthritis is believed to be both mechan-
tibial plateau height allows the varus deformity to ical and nutritional [14]. Direct physical damage
correct when the knee flexes. The correction ten- by osteophytes at condylar margins is one mech-
sions the medial collateral ligament (MCL) to anism [16]. Further, the ACL is intra-articular
full length, preventing its contracture over time. and so is at risk of devascularisation caused by
No soft tissue release is therefore required to cor- chronic synovitis. The removal of ACL synovium
rect MCL length intra-operatively. in rabbit models was found to cause very similar
The presence of an intact ACL in all knees changes to the ACL as those observed in human
undergoing UKA described by White et al. is of osteoarthritis [18].
key importance [13]. In chronic rupture due to Taken together the findings of these studies
osteoarthritis, the ACL first loses its synovial suggest that an intact ACL is necessary to confine
covering, then splits longitudinally. After this cartilage wear to the anteromedial tibial plateau.
collagen bundles begin to stretch and lose Chronic damage to and eventual rupture of the
strength, before the ACL finally ruptures [14]. ACL in the presence of osteoarthritis are likely
Following this the ligament may eventually be caused by direct physical and vascular damage,
absorbed and disappear. A later study investi- and in turn lead to spreading of wear areas poste-
gated the effects of damage to functionally intact riorly on the medial tibial plateau and to the lat-
ACL. Knees undergoing UKA with higher grades eral compartment.
of ACL damage had more full thickness loss of Anteromedial osteoarthritis has characteristic
cartilage in the anteromedial region of the tibial findings upon clinical assessment [14]. Pain is
plateau. Cartilage loss migrated laterally and pos- not necessarily localised to the medial compart-
teriorly with increasing ACL damage [15], ment, but is present on walking, worse on stand-
Harman et al. studied 143 tibial plateaus resected ing and reduced on sitting. There is a varus
during TKA for osteoarthritis and demonstrated deformity on knee extension of 5–15°, which is
the effects of functionally impaired ACLs [16]. passively correctable at 20 degrees of flexion,
They tested ACL integrity intra-operatively and and spontaneous correction at 90 degrees of flex-
used digital imaging to study plateau wear pat- ion. On intra-operative inspection, both cruciate
terns. In varus knees with intact ACL, their find- ligaments are intact. The cartilage of the antero-
ings were consistent with intra-operative reports medial tibial plateau and inferior medial femoral
of knees with intact ACL during UKA; wear was condyle are eroded with bone-on-bone contact,
present in the middle to anterior aspect of the while at the posterior aspect of both the cartilage
medial plateau. In varus knees with ACL defi- is preserved. The articular cartilage of the lateral
3 Anteromedial Osteoarthritis and UKA 31

compartment is at full thickness. The MCL length In 1976, Goodfellow et al. in Oxford first pro-
is preserved, while the posterior capsule is short- posed the use of meniscal bearing knee prosthe-
ened [14]. ses, initially for bicompartmental tibiofemoral
arthroplasty [29]. They later described its appli-
cation for UKA, suggesting adverse outcomes in
3.3 Unicompartmental Knee previous UKA to be caused by poor patient selec-
Arthroplasty tion, inadequate prosthesis design, and surgical
technique [30]. From 1982, this Oxford UKA
The aim of the UKA is to replace the diseased (OUKA) was mainly used in knees with isolated
articular compartment when knee osteoarthritis is medial osteoarthritis and intact ACL [30].
confined to a single compartment. The soft tis-
sues and opposite compartment are preserved,
allowing them to resume their physiological 3.3.2 Indications
function.
In an osteoarthritic knee where symptoms justify
arthroplasty, the Oxford group describe a series
3.3.1 History of indications for the use of UKA [14].
Intact ACL is necessary for UKA. Deschamps
The concept of UKA for the prevention of pain in et al. noted that the majority of knees with ACL
osteoarthritis was described by Campbell in 1940 laxity noted pre-operatively, UKA failed. Most of
when he used vitallium plates within the medial these failures required further surgery after a
compartment of arthritic knees [19]. McKeever mean time of 3.5 years [31]. These findings were
and MacIntosh then trialled the use of metal supported by those reported by Goodfellow et al.
inserts to replace the tibial surface of a single Their study of 301 patients up to 9 years follow-
compartment in valgus and varus deformities. ing UKA found a 95% survival rate at for knees
This provided pain relief but overall unsatisfac- with intact ACL, compared with 81% in knee
tory results due to prosthesis migration [20, 21]. with damaged or absent ACL [32].
McKeever later added a keel to his tibial plateau There must be full thickness preservation of
prosthesis to overcome this [22]. Gunston and the lateral tibial plateau articular cartilage to
polycentric UKA devices were introduced in the allow UKA. Wear within this compartment sug-
early 1970s [22]. The St Georg sled was devel- gests impending failure of the ACL and is an
oped in 1969, and good results were reported at 4 absolute contraindication for UKA. Fibrillation
year follow-up in a study of 294 patients [23]. and chondromalacia are often seen in the lateral
This was a cemented polycentric metal femoral compartment caused by chronic synovitis within
condyle articulating on flat polyethylene tibial the joint and are not of concern. Goodfellow et al.
components, as were the Mamor implants devel- describe their use of valgus-stressed radiographs
oped in 1972 [24]. These first-generation modern to assess lateral compartment cartilage thickness
implants were at high risk of deformation and [14]. They report little or no deterioration of the
early wear, and so were further developed to lateral compartment in follow-up of over 10 years
introduce a metal-backed component [25]. Good following UKA in patients pre-operatively screen
clinical results were reported for both implant with this method [33].
designs in single compartment disease [23, 26]. The Oxford group also require any varus
However, some groups reported poor outcomes deformity to be fully correctable in 20 degrees of
in UKA, often due to inadequate patient selection flexion, and for posterior cartilage to be intact
or material failures [27, 28]. In conjunction with within the medial tibial plateau [14]. These two
rapid developments and improvements in out- requirements are complementary; as previously
comes following TKA, this led to many surgeons discussed, intact posterior cartilage allows ten-
abandoning the use of UKA altogether [22]. sioning of the MCL on flexion of the knee,
32 S. W. King et al.

p­ reventing contracture, and allowing correction cedure is less. Siman et al. reported a study of
of varus deformity without soft tissue release. patients aged over 75 years, comparing UKA
Flexion deformity is often present in antero- with TKA. Patients in the UKA group had shorter
medial osteoarthritis of the knee. The posterior operative times, shorter length of stay, less intra-­
joint capsule shortens due to chronic synovitis operative blood loss, and post-operative transfu-
and voluntary reduction in extension caused by sion requirement. Their mobility also improved
pain. Osteophytes may also restrict posterior cap- better and more quickly post-operatively [8].
sular ligament movement, as well as directly Other studies agree, with further findings of
impinging on extension anteriorly. UKA is per- reduced blood loss [42, 43], shorter length of stay
missible in flexion deformity of up to 15°. This and lower readmission rate [44], and reduced
will usually correct spontaneously after surgery incidence of thromboembolism, infection, stroke,
as soft tissue release is not required and so the and myocardial infarction [45] for UKA com-
posterior capsule is not required to stretch beyond pared with TKA.
its physical constraints following the procedure. Functional outcomes are also better for
Flexion deformity beyond 15° is generally indic- UKA. Rougraff et al. compared 81 tricompart-
ative of ACL deficiency. mental knee arthroplasties with 120 UKA and
found improved range of motion and ambulatory
function in the UKA group. The gait is more
3.3.3 Other Indications physiological and biomechanics of the knee more
completely restored [46]. UKA also preforms
In addition to anteromedial osteoarthritis, UKA better than TKA when compared using outcome
has been proposed for other indications. Focal scores. A study of 390 knees comparing TKA
avascular necrosis of the medial femoral condyle with twin-peg OUKA by Lum et al. in 2016
or tibial plateau requires evaluation with found improved knee society scores at approxi-
MRI. Good results have been reported with both mately 5 years post-operatively [47]. Another
the Marmor knee and OUKA [34, 35]. UKA has study of 101 patients aged over 75 years of age
also been used for failed high tibial osteotomy found better Knee Society, Forgotten Joint and
with persistent symptoms. However, results have Knee Injury Osteoarthritis Outcome Scores at
generally been poor and inferior to TKA and so last follow-up [48]. Similarly, better Forgotten
this application is not recommended [36–39]. Joint Scores were seen for UKA in patients one-
Lateral compartment osteoarthritis represents and two-years post-operative.
approximately 10% of all unicompartmental Revision rates are often used as a comparative
osteoarthritis. Lateral UKA is challenging both measure of success for an implant and are
due to anatomical constraints and lesser preva- recorded at individual, local, regional, and
lence and so reduced surgical experience. Reports national levels. For UKA 10-year survival rates
vary with respect to suitability of lateral of more than 90% have been reported [49, 50],
UKA. The Oxford group report poorer outcomes with a centre which specialises in UKA reporting
for lateral OUKA compared with medial OUKA a 10-year survival rate of 96% in 1000 phase 3
due to high dislocation rate [40, 41]. However, a OUKA [51]. However, revision rate of UKA has
recent systematic review of the literature found been reported as higher than those for TKA by a
no difference between medial and lateral for all number of sources; according to registry data,
UKA [9]. UKA are 2.1–2.8 times more likely to be revised
[45, 52, 53]. Several factors must be considered
when interpreting these outcomes. Revision of
3.3.4 Outcomes in UKA UKA is technically easier. This is reflected in a
lower threshold for revision: Goodfellow et al.
UKA allows preservation of bone stock and soft reported that 63% of UKA implants with a post-­
tissues, and the physiological impact of the pro- operative Oxford knee score (OKS) of under 20
3 Anteromedial Osteoarthritis and UKA 33

(very poor) were revised but only 12% of TKA higher volume for better results [14]. Additionally,
with the same score [53]. UKA is also more patient selection is controversial. Kozinn and
likely to be used in younger, more active patients Scott have made recommendations for the char-
because of its better functional outcome, and acteristics of ideal fixed-bearing UKA candidates
these demographics are independent risk factors in their 1989 paper [62]. These are often applied
for revision. Additionally, there is a large differ- as eligibility criteria and are as follows: isolated
ence in revision rates for UKA between high and medial compartment disease; aged less than
low volume surgeons [54]. Some knee surgeons 60 years; low level of physical activity; weight
may only perform one or two UKA per year, less than 82 kg; a cumulative angular deformity
while best results are seen in those for whom of less than 15°; both cruciate ligaments intact; a
UKA makes up at least 20% of their operative pre-operative range of flexion of 90; a flexion
caseload [55]. Despite higher revision rates, contracture of <5°; minimal pain at rest; no radio-
UKA remains more cost-effective than TKA graphic or intra-operative evidence of chondro-
especially for younger patients [56]. calcinosis or patellofemoral osteoarthritis; no
Outcomes and mechanism of failure vary for inflammatory arthropathy. Applying these as
different types of UKA. A randomised study of strict eligibility criteria, some groups have found
56 knees comparing fixed and mobile bearing only 6% of patients to be eligible for UKA [63].
UKAs found improved kinematics and lower Others also apply additional exclusion criteria
incidences of radiolucencies in the mobile bear- such as presence of lateral compartment and
ing group, but with no difference in patient patellofemoral osteophytes and lateral compart-
reported outcome score [57]. Mobile bearing ment chondromalacia, and find even fewer to be
UKAs are more likely to fail with early bearing suitable [64]. The Oxford group applies less
dislocation, while late polyethylene wear occurs stringent criteria, removing restrictions including
more commonly in fixed-bearing implants [58]. those on patellofemoral arthritis, weight, and age
Despite this difference in failure mode, a system- [14]. They contend that they have applied these
atic review of the literature by Ko et al. found no criteria for a number of years with excellent clini-
significant difference in overall reoperation rate cal results, and the discrepancy may be because
between the two types [59]. A systematic review the original criteria were for fixed-bearing UKA
of 10 papers and 1199 knees comparing cemented while the OUKA is a mobile bearing implant.
UKAs with cementless found no difference with
respect to clinical outcome and revision and
reoperation rate, but that for cementless implants, 3.4 Conclusion
operative times were shorter and in post-­operative
radiographs, there was a lower incidence of Unicompartmental knee arthroplasty is a poten-
radiolucent lines [60]. tial alternative to total knee arthroplasty for
selected indications, primarily isolated anterome-
dial osteoarthritis. There is increasing demand
3.3.5 Limitations to UKA Use for knee arthroplasty, and indications are expand-
ing. Younger, more active patients with greater
Approximately 60% of patients with knee osteo- physical demands and higher expectations are
arthritis have disease isolated to the medial com- now undergoing joint arthroplasties. The rate of
partment [4–7], and approximately one-third of UKA is increasing, and the procedure represents
all patients eligible for TKA are suitable for UKA an excellent option for a significant proportion of
[14, 61]. However, only 8–12% of arthroplasties patients. The retention of other articular compart-
are unicompartmental [8, 9]. The reasons for this ments and soft tissues provide excellent clinical
are multifactorial. UKA is generally perceived as outcomes and a procedure with lower rates of
more challenging, and surgeons may be reluctant complications and better recovery. Revision rates
to perform the procedure given the need for for UKA remain higher than TKA, but these may
34 S. W. King et al.

be artificially inflated. Careful patient selection 10. National Joint Registry. National Joint Registry 15th
annual report. National Joint Registry; 2018.
and adequate volume of procedures are vital for 11. Hernborg JS, Nilsson BE. The natural course of
best outcomes. untreated osteoarthritis of the knee. Clin Orthop Relat
Professor Pandit is a National Institute for Res. 1977;123:130–7.
Health Research (NIHR) Senior Investigator. 12. Altman RD, Fries JF, Bloch DA, et al. Radiographic
assessment of progression in osteoarthritis. Arthritis
The views expressed in this article are those of Rheum. 1987;30(11):1214–25.
the author(s) and not necessarily those of the 13. White SH, Ludkowski PF, Goodfellow
NIHR, or the Department of Health and Social JW. Anteromedial osteoarthritis of the knee. J Bone
Care. Joint Surg Br. 1991;73(4):582–6.
14. Goodfellow JW, Pandit HG, Dodd CA, Murray
DW. Unicompartmental arthroplasty with the Oxford
knee. 2nd ed. Oxford: Goodfellow Publishers Limited.
References 15. Rout R, McDonnell S, Hulley P, et al. The pattern of
cartilage damage in antero-medial osteoarthritis of the
1. Lawrence RC, Felson DT, Helmick CG, et al. knee and its relationship to the anterior cruciate liga-
Estimates of the prevalence of arthritis and other rheu- ment. J Orthop Res. 2013;31(6):908–13. https://doi.
matic conditions in the United States. Part II. Arthritis org/10.1002/jor.22253.
Rheum. 2008;58(1):26–35. https://doi.org/10.1002/ 16. Harman MK, Markovich GD, Banks SA, Hodge
art.23176. WA. Wear patterns on tibial plateaus from varus and
2. Culliford D, Maskell J, Judge A, et al. Future pro- valgus osteoarthritic knees. Clin Orthop Relat Res.
jections of total hip and knee arthroplasty in the 1998;352:149–58.
UK: results from the UK clinical practice research 17. Keyes GW, Carr AJ, Miller RK, Goodfellow JW. The
Datalink. Osteoarthr Cartil. 2015;23(4):594–600. radiographic classification of medial gonarthrosis.
https://doi.org/10.1016/j.joca.2014.12.022. Correlation with operation methods in 200 knees.
3. Kurtz S, Ong K, Lau E, Mowat F, Halpern Acta Orthop Scand. 1992;63(5):497–501.
M. Projections of primary and revision hip and knee 18. Robinson D, Halperin N, Nevo Z. Devascularization
arthroplasty in the United States from 2005 to 2030. of the anterior cruciate ligament by synovial strip-
J Bone Joint Surg Am. 2007;89(4):780–5. https://doi. ping in rabbits. An experimental model. Acta Orthop
org/10.2106/JBJS.F.00222. Scand. 1992;63(5):502–6.
4. Ledingham J, Regan M, Jones A, Doherty 19. Campbell WC. Interposition of vitallium plates in
M. Radiographic patterns and associations of osteoar- arthroplasties of the knee. Preliminary report. By
thritis of the knee in patients referred to hospital. Ann Willis C. Campbell, 1940. Clin Orthop Relat Res.
Rheum Dis. 1993;52(7):520–6. 1988;226:3–5.
5. Davies AP, Vince AS, Shepstone L, Donell ST, 20. MacIntosh DL. Hemiarthroplasty of the knee using
Glasgow MM. The radiologic prevalence of patel- a space-occupying prosthesis for painful Varus and
lofemoral osteoarthritis. Clin Orthop Relat Res. valgus deformities. J Bone Jt Surg. 1958;40(A):1431.
2002;402:206–12. 21. MacIntosh DL, Hunter GA. The use of the hemiar-
6. Ahlbäck S. Osteoarthrosis of the knee. A radio- throplasty prosthesis for advanced osteoarthritis and
graphic investigation. Acta Radiol Diagn (Stockh). rheumatoid arthritis of the knee. J Bone Joint Surg Br.
1968;Suppl 277:7–72. 1972;54(2):244–55.
7. Rout R, McDonnell S, Benson R, et al. The histologi- 22. Bruni D, Iacono F, Akkawi I, Gagliardi M, Zaffagnini
cal features of anteromedial gonarthrosis—the com- S, Marcacci M. Unicompartmental knee replacement:
parison of two grading systems in a human phenotype a historical overview. Joints. 2013;1(2):45–7.
of osteoarthritis. Knee. 2011;18(3):172–6. https://doi. 23. Engelbrecht E, Siegel A, Rottger J, Buchholz
org/10.1016/j.knee.2010.04.010. HW. Statistics of total knee replacement: partial and
8. Siman H, Kamath AF, Carrillo N, Harmsen WS, total knee replacement, design St. Georg: a review
Pagnano MW, Sierra RJ. Unicompartmental knee of a 4-year observation. Clin Orthop Relat Res.
arthroplasty vs total knee arthroplasty for medial 1976;120:54–64.
compartment arthritis in patients older than 75 years: 24. Marmor L. The modular knee. Clin Orthop Relat Res.
comparable reoperation, revision, and complication 1973;94:242–8.
rates. J Arthroplast. 2017;32(6):1792–7. https://doi. 25. Palmer SH, Morrison PJ, Ross AC. Early catastrophic
org/10.1016/j.arth.2017.01.020. tibial component wear after unicompartmental knee
9. van der List JP, McDonald LS, Pearle AD. Systematic arthroplasty. Clin Orthop Relat Res. 1998;350:143–8.
review of medial versus lateral survivorship in unicom- 26. Marmor L. Marmor modular knee in unicompartmen-
partmental knee arthroplasty. Knee. 2015;22(6):454– tal disease. Minimum four-year follow-up. J Bone
60. https://doi.org/10.1016/j.knee.2015.09.011. Joint Surg Am. 1979;61(3):347–53.
3 Anteromedial Osteoarthritis and UKA 35

27. Lindstrand A, Ryd L, Stenström A. Polyethylene 41. Weston-Simons JS, Pandit H, Kendrick BJL, et al.
failure in two total knees. Wear of thin, metal-­ The mid-term outcomes of the Oxford Domed
backed PCA tibial components. Acta Orthop Scand. Lateral unicompartmental knee replacement.
1990;61(6):575–7. Bone Joint J. 2014;96-B(1):59–64. https://doi.
28. Insall J, Aglietti P. A five to seven-year follow-up org/10.1302/0301-­620X.96B1.31630.
of unicondylar arthroplasty. J Bone Joint Surg Am. 42. Schwab P-E, Lavand’homme P, Yombi JC, Thienpont
1980;62(8):1329–37. E. Lower blood loss after unicompartmental than
29. Campi S, Tibrewal S, Cuthbert R, Tibrewal total knee arthroplasty. Knee Surg Sports Traumatol
SB. Unicompartmental knee replacement - current Arthrosc. 2015;23(12):3494–500. https://doi.
perspectives. J Clin Orthop Trauma. 2018;9(1):17–23. org/10.1007/s00167-­014-­3188-­x.
https://doi.org/10.1016/j.jcot.2017.11.013. 43. Price AJ, Webb J, Topf H, et al. Rapid recovery after
30. Goodfellow JW, Tibrewal SB, Sherman KP, O’Connor oxford unicompartmental arthroplasty through a short
JJ. Unicompartmental Oxford meniscal knee arthro- incision. J Arthroplast. 2001;16(8):970–6.
plasty. J Arthroplast. 1987;2(1):1–9. 44. Drager J, Hart A, Khalil JA, Zukor DJ, Bergeron SG,
31. Deschamps G, Lapeyre B. [Rupture of the anterior Antoniou J. Shorter hospital stay and lower 30-day
cruciate ligament: a frequently unrecognized cause of readmission after unicondylar knee arthroplasty
failure of unicompartmental knee prostheses. Apropos compared to total knee arthroplasty. J Arthroplast.
of a series of 79 lotus prostheses with a follow-up 2016;31(2):356–61. https://doi.org/10.1016/j.
of more than 5 years]. Rev Chir Orthop Reparatrice arth.2015.09.014.
Appar Mot. 1987;73(7):544–51. 45. Liddle AD, Judge A, Pandit H, Murray DW. Adverse
32. Goodfellow J, O’Connor J. The anterior cruciate liga- outcomes after total and unicompartmental knee
ment in knee arthroplasty. A risk-factor with uncon- replacement in 101 330 matched patients: a study of
strained meniscal prostheses. Clin Orthop Relat Res. data from the National Joint Registry for England and
1992;(276):245–52. Wales. Lancet. 2014;384(9952):1437–45. https://doi.
33. Weale AE, Murray DW, Crawford R, et al. Does org/10.1016/S0140-­6736(14)60419-­0.
arthritis progress in the retained compartments 46. Laurencin CT, Zelicof SB, Scott RD, Ewald
after “Oxford” medial unicompartmental arthro- FC. Unicompartmental versus total knee arthroplasty
plasty? A clinical and radiological study with a in the same patient. A comparative study. Clin Orthop
minimum ten-year follow-up. J Bone Joint Surg Br. Relat Res. 1991;273:151–6.
1999;81(5):783–9. 47. Lum ZC, Lombardi AV, Hurst JM, Morris MJ,
34. Marmor L. Unicompartmental arthroplasty for osteo- Adams JB, Berend KR. Early outcomes of twin-peg
necrosis of the knee joint. Clin Orthop Relat Res. mobile-­bearing unicompartmental knee arthroplasty
1993;294:247–53. compared with primary total knee arthroplasty.
35. Langdown AJ, Pandit H, Price AJ, et al. Oxford Bone Joint J. 2016;98-B(10 Supple B):28–33.
medial unicompartmental arthroplasty for https://doi.org/10.1302/0301-­6 20X.98B10.BJJ-­
focal spontaneous osteonecrosis of the knee. 2016-­0414.R1.
Acta Orthop. 2005;76(5):688–92. https://doi. 48. Fabre-Aubrespy M, Ollivier M, Pesenti S, Parratte
org/10.1080/17453670510041772. S, Argenson J-N. Unicompartmental knee arthro-
36. Meding JB, Keating EM, Ritter MA, Faris PM. Total plasty in patients older than 75 results in better clini-
knee arthroplasty after high tibial osteotomy. A cal outcomes and similar survivorship compared to
comparison study in patients who had bilateral total knee arthroplasty. A matched controlled study.
total knee replacement. J Bone Joint Surg Am. J Arthroplasty. 2016;31(12):2668–71. https://doi.
2000;82(9):1252–9. org/10.1016/j.arth.2016.06.034.
37. Scott RD, Cobb AG, McQueary FG, Thornhill 49. Parratte S, Ollivier M, Lunebourg A, Abdel MP,
TS. Unicompartmental knee arthroplasty. Eight- to Argenson J-N. Long-term results of compart-
12-year follow-up evaluation with survivorship analy- mental arthroplasties of the knee: long term
sis. Clin Orthop Relat Res. 1991;271:96–100. results of partial knee arthroplasty. Bone Joint
38. Vorlat P, Verdonk R, Schauvlieghe H. The Oxford uni- J. 2015;97-B(10 Suppl A):9–15. https://doi.
compartmental knee prosthesis: a 5-year follow-up. org/10.1302/0301-­620X.97B10.36426.
Knee Surg Sports Traumatol Arthrosc. 2000;8(3):154– 50. Mohammad HR, Strickland L, Hamilton TW, Murray
8. https://doi.org/10.1007/s001670050206. DW. Long-term outcomes of over 8,000 medial
39. Rees JL, Price AJ, Lynskey TG, Svärd UC, Dodd CA, Oxford phase 3 unicompartmental knees—a system-
Murray DW. Medial unicompartmental arthroplasty atic review. Acta Orthop. 2018;89(1):101–7. https://
after failed high tibial osteotomy. J Bone Joint Surg doi.org/10.1080/17453674.2017.1367577.
Br. 2001;83(7):1034–6. 51. Pandit H, Jenkins C, Gill HS, Barker K, Dodd C, a.
40. Murray DW, Goodfellow JW, O’Connor JJ. The F, Murray DW. Minimally invasive Oxford phase 3
Oxford medial unicompartmental arthroplasty: unicompartmental knee replacement: results of 1000
a ten-year survival study. J Bone Joint Surg Br. cases. J Bone Joint Surg Br. 2011;93(2):198–204.
1998;80(6):983–9. https://doi.org/10.1302/0301-­620X.93B2.25767.
36 S. W. King et al.

52. Chawla H, van der List JP, Christ AB, Sobrero MR, 58. Cheng T, Chen D, Zhu C, et al. Fixed- versus mobile-­
Zuiderbaan HA, Pearle AD. Annual revision rates of bearing unicondylar knee arthroplasty: are fail-
partial versus total knee arthroplasty: a comparative ure modes different? Knee Surg Sports Traumatol
meta-analysis. Knee. 2017;24(2):179–90. https://doi. Arthrosc. 2013;21(11):2433–41. https://doi.
org/10.1016/j.knee.2016.11.006. org/10.1007/s00167-­012-­2208-­y.
53. Goodfellow JW, O’Connor JJ, Murray DW. A cri- 59. Ko Y-B, Gujarathi MR, Oh K-J. Outcome of unicom-
tique of revision rate as an outcome measure: partmental knee arthroplasty: a systematic review of
­re-­interpretation of knee joint registry data. J Bone comparative studies between fixed and mobile bear-
Joint Surg Br. 2010;92(12):1628–31. https://doi. ings focusing on complications. Knee Surg Relat
org/10.1302/0301-­620X.92B12.25193. Res. 2015;27(3):141–8. https://doi.org/10.5792/
54. Liddle AD, Pandit H, Judge A, Murray DW. Effect ksrr.2015.27.3.141.
of surgical caseload on revision rate following total 60. Campi S, Pandit HG, Dodd C, Murray DW. Cementless
and unicompartmental knee replacement. J Bone Joint fixation in medial unicompartmental knee arthro-
Surg Am. 2016;98(1):1–8. https://doi.org/10.2106/ plasty: a systematic review. Knee Surg Sports
JBJS.N.00487. Traumatol Arthrosc. 2017;25(3):736–45. https://doi.
55. Hamilton TW, Rizkalla JM, Kontochristos L, et al. org/10.1007/s00167-­016-­4244-­5.
The interaction of caseload and usage in determin- 61. Svärd UC, Price AJ. Oxford medial unicompartmental
ing outcomes of unicompartmental knee arthroplasty: knee arthroplasty. A survival analysis of an indepen-
a meta-analysis. J Arthroplast. 2017;32(10):3228– dent series. J Bone Joint Surg Br. 2001;83(2):191–4.
3237.e2. https://doi.org/10.1016/j.arth.2017.04.063. 62. Kozinn SC, Scott R. Unicondylar knee arthroplasty. J
56. Ghomrawi HM, Eggman AA, Pearle AD. Effect of Bone Joint Surg Am. 1989;71(1):145–50.
age on cost-effectiveness of unicompartmental knee 63. Stern SH, Becker MW, Insall JN. Unicondylar knee
arthroplasty compared with total knee arthroplasty in arthroplasty. An evaluation of selection criteria. Clin
the U.S. J Bone Joint Surg Am. 2015;97(5):396–402. Orthop Relat Res. 1993;286:143–8.
https://doi.org/10.2106/JBJS.N.00169. 64. Ritter MA, Faris PM, Thong AE, Davis KE, Meding
57. Li MG, Yao F, Joss B, Ioppolo J, Nivbrant B, Wood JB, Berend ME. Intra-operative findings in varus
D. Mobile vs. fixed bearing unicondylar knee arthro- osteoarthritis of the knee. An analysis of pre-­
plasty: a randomized study on short term clinical out- operative alignment in potential candidates for uni-
comes and knee kinematics. Knee. 2006;13(5):365–70. compartmental arthroplasty. J Bone Joint Surg Br.
https://doi.org/10.1016/j.knee.2006.05.003. 2004;86(1):43–7.
 Conventional Indications
for Unicompartmental Knee 4
Arthroplasty

Caroline Vincelot Chainard and Henri Robert

Surgical management of tibiofemoral osteoar-

thritis (OA) continues to be debated after more
than 50 years. Three solutions are possible: high
tibial or distal femoral osteotomy (HTO/DFO),
total knee arthroplasty (TKA) and unicompart-
mental knee arthroplasty (UKA) after failure of
well-conducted medical treatment (change to the
patient’s activities, nonsteroidal anti-­
inflammatory drugs or NSAIDs, chondroprotec-
tive medications, intra-articular injections,
orthotics). Osteotomies were long reserved for
young, active subjects, and arthroplasties tended
to be preferable in older subjects who were
largely inactive. Schematically, tibial osteoto-
mies restore good anatomical alignment of the
tibia when a deformity tends to be epiphyseal
(constitutional varus or Lévigne’s epiphysis), but
require slight overcorrection (Fig. 4.1). They are
indicated in unicompartmental OA, in young
subjects whose knee is stable and has complete
mobility. They have several limitations: nonim-
mediate weightbearing, long rehabilitation and
lengthy sick leave. They expose patients to spe-
cific complications: neurovascular, delayed con-
solidation or even disunion, infection and
cosmetically unpleasant result in overcorrection.

C. V. Chainard
Orthopaedic Surgery Department, Angers University Fig. 4.1 Anterior view of a knee with epiphyseal varus
Hospital, Angers, France deformity

H. Robert (*)
Orthopaedic Surgery Department, Haut Anjou
Hospital, Chateau-Gontier, France

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2024 37

A. Clavé, F. Dubrana (eds.), Unicompartmental Knee Arthroplasty,
https://doi.org/10.1007/978-3-031-48332-5_4
38 C. V. Chainard and H. Robert

In our experience, HTO continues to be indicated the choice of treatment [6]. The rate of UKA var-
in young, athletic patients with epiphyseal varus ies from 0% to 50% (mean 9%) in indications in
deformity and Ahlbäck stage ≤2. Osteotomies the United Kingdom [7]. These rates are 5% in
still have unconditional defenders who, by their the USA, 7.6% in Denmark in 2010, 7.9% in
technical expertise and volume of activity, mini- Australia in 2018, 8% in Sweden since 2014,
mise the risks [1]. According to these authors, 10% in New Zealand in 2009, 12% in France and
HTO remains possible for Ahlbäck stage OA ≥ 3 14% in Switzerland [8]. Survival rates at 10 and
and absence of the anterior cruciate ligament 20 years have improved with better patient selec-
(ACL). Currently, indications for HTO are tion, implants and placement techniques:
declining even though UKA is increasing [2]. In “Modern cemented uni-knee replacement pro-
France, 9500 UKA were implanted in 2011 and vided durable pain relief and long-term restora-
12,250 in 2019, i.e. a 36% rise in 8 years (www. tion of knee function” [9]. UKA offers many
atih.sante.fr). advantages: near-normal kinematic, preservation
UKA has progressively developed from the of bone stock and ligaments, less invasive sur-
1970s under the impulsion of Leonard Marmor in gery, simpler postoperative follow-up, possible
the USA and Phillipe Cartier in France. Initial outpatient surgery, lower morbidity (pain, infec-
disappointing results with UKA, often due to tion, stiffness), lower mortality (myocardial
technical errors or improper indications, led to infarction, pulmonary embolism, stroke) and bet-
restrictive indications and many contraindica- ter function [10]. Functional scores (patients’
tions. In a multicentre study, there was only 67% pain, mobility and satisfaction) are better with
survival of UKA in a series by Hernigou and UKA than with TKA, and current 10-year sur-
Deschamps [3]. These poor results have been vival rates are close to those of TKA [11–14].
confirmed by other authors or registers: 80% sur- The objective of UKA surgery is to perform
vival at 10 years for UKA in a Finnish register tibiofemoral resurfacing in order to correct
versus 91% to 94% for TKA, and 10% revision monocompartmental wear, without restoring a
for UKA versus 3–10% for TKA at 10 years in normal axis in a lower limb in anteroposterior
the Swedish or UK register [4]. Surgical revision presentation. In medial UKA, overall residual
rates are biased because revising a UKA is con- varus deformity of about 2° (HKA ≈ 178°) is
sidered easier than revising a TKA. Therefore, desirable to avoid decompensation of the contra-
the indication for UKA revision will be estab- lateral compartment [12] (Fig. 4.2). The persis-
lished more easily and the register rates reflect tence of this varus deformity does not expose the
this difference [5]. patient to early PE deterioration, particularly in
Considering these initial results, TKA with mobile-bearing UKA [15]. UKA cannot correct a
broad indications, a simpler technique and satis- diaphyseal or metaphyseal bone defect (disunion,
factory results have left little place for UKA. Yet sequelae of osteotomy, epiphyseal varus defor-
the functional results of TKA are highly variable mity). Anteromedial tibiofemoral OA-(AMOA)-
depending on the articles (up to 20% of dissatis- (often after meniscectomy) is the leading
fied patients despite the well-established indica- indication for UKA (> 90%), followed by ON
tion for TKA) and always better in publications (5%), the sequelae of fracture of the medial or
by an experienced team than in national registers. lateral tibial plateau [8].
However, we are witnessing a return to UKA in
4 Conventional Indications for Unicompartmental Knee Arthroplasty 39

Fig. 4.3 The “finger sign”, the patient’s finger indicates

the elective site of pain

varus is assessed with the knee in 20° flexion; it

can be partial or complete (Fig. 4.4 a, b). The
radiological assessment includes at least: an ante-
rior view with comparative weightbearing views
(a profile view in extension with weightbearing), a
patellofemoral view at 30° and the long axis with
weightbearing on one foot. Narrowing of the TF
joint space is graded on a weightbearing scale
according to the Ahlbäck classification (four
stages). Overall varus is the sum total of narrowing
of the joint space and epiphyseal varus. It can be
Fig. 4.2 Decompensation at 10 years postop. with valgus considered that the total loss of MTF cartilage (5
deformity of a medial UKA mm) results in varus of 5°, i.e. 1° per millimetre
[16]. Testing in valgus flexion will correct the
deformity in the absence of epiphyseal varus
4.1 Anteromedial Osteoarthritis because there is no retraction of the medial collat-
(AMOA) eral ligament (MCL), and this will make it possi-
ble to recognise the thickness of the cartilage in the
The clinical assessment should seek to identify the lateral compartment. In cases of epiphyseal varus,
specific site of pain (the “finger sign” pointing to a deformity of up to 7° can persist [13]. A preop-
the MTF joint space) (Fig. 4.3) and the mechanical erative deformity greater than 15° should lead dis-
characteristic (pain on walking, prolonged stand- tension of the convexity (lateral dislocation) to be
ing). Four points of the clinical examination are suspected, which contraindicates UKA. An X-ray
important: mobility in flexion–extension, sagittal assessment can be supplemented by a forced varus
stability, patellofemoral mobility and reducibility image in cases of moderate MTF (narrowing of the
of the varus deformity. AOMA is often accompa- joint space), which may be enhanced. Tibial sub-
nied by moderate flexion deformity of 10–15° and luxation in an anterior view with weightbearing,
loss of complete flexion of 10–20°. Sagittal stabil- often worsened by an X-ray view in valgus, is a
ity, evidencing competent ACL, should be sought contraindication to UKA (Fig. 4.5). A profile in
comparatively by the Lachman test or laximetry extension makes it possible to detect rupture of the
measurements (KT-1000, Telos, GNRB®). ACL by evidencing anterior tibial translation [17]
Patellar mobility is tested: the patella is mobile, (Fig. 4.6). MRI or CT scans has no place in the
painless and there is no clash. Reducibility of standard assessment.
40 C. V. Chainard and H. Robert

a b

Fig. 4.4 (a) Varus knee deformity under load. (b) Complete reducibility of varus, without hypercorrection

Fig. 4.6 Anterior subluxation in a sagittal view under

Fig. 4.5 Lateral subluxation of AMOA, which load, which contraindicates UKA
contraindicates UKA
4 Conventional Indications for Unicompartmental Knee Arthroplasty 41

4.2 Osteonecrosis (ON) Table 4.1 Absolute contraindications according to

Kozinn and Scott [20]

ON occurs following a localised vascular break Age < 60 years

in the subchondral bone with secondary chondral Weight > 82 kg
Manual worker
damage opposite and then sequestration. It can be
Patellofemoral osteoarthritis
primary and more rarely secondary (lengthy cor-
Frontal deformity >15°
ticosteroid therapy, trauma, transplantation, Chondrocalcinosis
chronic alcohol abuse, lupus, etc.). ON most
often affects the femoral condyle in females over
50 years of age. ON is manifest by unilateral Table 4.2 Absolute contraindications according to
sudden-­ onset pain with no X-ray changes. Deschamps [21]
Medical imaging makes it possible to confirm the Bi- or tricompartmental OA
diagnosis. X-rays underevaluate signs of aseptic Anterior tibial translation of more than 10 mm or a
soft stop in the Lachman test
osteonecrosis in the early stage. Later in progres-
Frontal laxity of the convexity
sion, one or more of the following signs are
BMI > 30 kg/m2
observed: a subchondral radiotransparent area Rheumatic or other inflammatory disorders.
with/without a sclerotic border, flattening of the
joint surface of the femoral condyle, free bodies
in the joint interspace and a periosteal reaction; Table 4.3 Relative contraindications according to
Deschamps [21]
subsequently, degenerative lesions can develop.
MRI detects early lesions, with sensitivity equiv- Osteoporosis of the tibial plateau, particularly in a
context of obesity
alent to that of scintigraphy but with better speci- Pseudarthrosis/disunion after fracture of the tibial
ficity. MRI is recommended in the preoperative plateau, after proximal tibial osteotomy.
assessment to evaluate the volume of necrotic
bone to be resected.
The indications for medial or lateral UKA in The Oxford team has published results on 1000
ON are rare, with 3.3% of UKA implanted in the mobile medial UKA by comparing survival rates
Mayo Clinic between 2002 and 2014 [18]. in the “Ideal indications” group (68%) and the
Transplantation of cancellous bone or cemented “Less than ideal indications” group (32%) accord-
filler for a condylar defect can be performed. The ing to Kozinn and Scott criteria [20]. Survival rates
results of UKA in ON are good with 93% sur- at 10 years were 93.6% and 97%, respectively,
vival at 10 years after UKA in primary ON, but (p > 0.05) [16]. Series of fixed medial UKA also
are less optimal in secondary ON [18]. The sur- have high rates of survival after 10 years’ follow-
vival rate at 12 years is 96.7% in 31 UKA accord- up: 93% for Lecuire et al. [22] and 98% for Lustig
ing to a report by Parratte et al. [19]. et al. [23]. In a meta-analysis of 44 articles on
9463 knees, the revision rates were comparable in
the short term between fixed versus mobile-bear-
4.3 Conventional Indications ing UKA [24]. Results from Parratte et al. [25]
confirm this study based on 156 UKA (fixed and
Whenever a patient who has electively symptom- mobile-bearing) with at least 15 years’ follow-up.
atic severe AOMA, as seen in X-ray weightbear- UKA use by surgeons is highly variable depend-
ing views, requires joint replacement, UKA can ing on their country, experience and, in particular,
be offered. Kozinn and Scott [20] in 1989 and trust in this implant. For surgeons who accept the
then Deschamps [21] in 1998 published a num- indication, UKA rates vary between 10% and over
ber of absolute and relative contraindications for 50%. The Oxford group using mobile-bearing
fixed medial UKA (Tables 4.1, 4.2, and 4.3). UKA surpasses 50% in indications for AOMA [7].
42 C. V. Chainard and H. Robert

4.4 Discussion functional score is poor. Consequently, age

of Contraindications should not be a contraindication to UKA accord-
ing to Kennedy et al.: “Earlier surgery may be
We are going to analyse certain contraindications preferable”, but HTO remains perfectly justified
in light of the literature and our experience. in hyperactive patients [33].

4.4.1 Age < 60 Years 4.4.2 Chondrocalcinosis

Clinical and radiological results for UKA in Chondrocalcinosis is characterised by deposits

patients >60 years of age are good [26]. In of calcium pyrophosphate crystals in the knee
patients under 60 years of age, HTO are recom- joint cartilage, meniscus or synovial membrane
mended for Ahlbäck stages ≤2, while UKA is (prevalence of 5%). It is well identified by X-rays
recommended in older patients [27]. The main and confirmed by histology. It is necessary to
risk of revision in the long-term follow-up of differentiate chondrocalcinosis “disease”, which
UKA in patients <60 years is wear on the PE is a real contraindication, from chondrocalcino-
[27]. This risk increases with younger subjects, sis “as an accompaniment”, which is asymptom-
activity and follow-up [26]. PE wear will be man- atic, nonprogressive and routinely observed
ifested by mechanical-type pain, acceleration of during progression of osteoarthritis disease
residual varus deformity and instability under (Fig. 4.7 a, b). According to Hernigou et al.,
weightbearing. Argenson et al. have reported there is no deterioration of the opposite compart-
11.4% PE wear requiring a simple change in a ment or reduction in survival rates at 10 or 15
series of 35 UKA (Miller–Galante arthroplasty) years compared to a series without chondrocalci-
implanted in patients 41–49 years of age [27]. In nosis [34].
this series, a single UKA patient underwent revi-
sion surgery for loosening of the implant (at
5 years’ postoperatively) and survival rate at 4.4.3 Patellofemoral Osteoarthritis
12 years was 80.6%. In a series by Pennington (PFOA)
et al. [28] (Miller–Galante arthroplasty), the rate
of revision surgery due to PE wear was 4.5% and Anterior pain in the knee should be analysed to
survival 92% at 12 years’ postoperatively. Oxford differentiate typically patellar pain from pain
UKA in patients >60 years of age show better caused by flexion with anterior tibial osteophytes,
survival rates at 10 years (96%) than Oxford ACL conflict in the femoral intercondylar notch
UKA in patients <60 years of age (91%). or posterior capsule retraction. These causes of
PE with fixed-bearing UKA should be highly pain identifiable in a profile X-ray view (anterior
reticulated and has a minimum thickness of tibial osteophyte or in the notch) are accessible
7–8 mm [29]. with a surgical release technique (resection of
Oxford UKA can have low PE thickness osteophytes, notchplasty) (Fig. 4.8 a, b). A patella
(meniscal bearing), but greater than 3 mm [30]. with little mobility and a painful flap, particularly
TKA in subjects <60 years of age can also with lateral PF narrowing, can be a contraindica-
yield high survival rates: 96% at 12 years accord- tion. PFOA will progress slowly and manifest
ing to Morgan et al. [31] and 95% for Duffy et al. more than 10 years’ postoperatively in 10% of
[32]. The decision to perform arthroplasty (TKA fixed-plateau UKA cases [35]. UKA with mobile-­
and UKA) in patients <60 years of age should be bearing plateau seem much more “patella
motivated by intensity of pain, functional impair- friendly” with PFOA. Only stage-four PFOA
ment and in full knowledge of specific risks (“bone against bone”) or fixed patellar sublux-
(early revision) with each implant. The benefit ation is a contraindication (Fig. 4.9); moderate
will be that much greater when the preoperative nonsymptomatic joint space narrowing present in
4 Conventional Indications for Unicompartmental Knee Arthroplasty 43

a b

Fig. 4.7 (a) Bimeniscal chondrocalcinosis with AMOA. (b) Medial UKA with good results at seven years’ follow-up

a b

Fig. 4.8 (a) Bone bloc in the anterior notch at the origin of flexion’s contracture. (b) Release of the anterior notch

54% of UKA cases in the Oxford group do not medial or lateral UKA. The condylar implant
penalise long-term results [16]. Medial or lateral should not be impinging forward to avoid a sec-
patellar and trochlear osteophytes in mirror ondary conflict [36]. Symptomatic PFOA (with-
image can be symptomatic (Fig. 4.10). out complete joint space narrowing) can be the
Regularisation of osteophytes after a parapatellar source of pain in cases of squatting but is not an
approach is a prerequisite to placement of a absolute contraindication to UKA [10].
44 C. V. Chainard and H. Robert

authors concluded: “The thresholds proposed by

Kozinn and Scott using weight, age, activity, the
state of the patellofemoral joint and chondrocal-
cinosis should not be considered to be contraindi-
cations for the use of the Oxford UKR” [16].
According to Bonutti et al., for severe obesity
(BMI > 35 kg/m2), the clinician should probably
remain cautious in UKA indications [40].

Fig. 4.9 Lateral patellofemoral arthritis and subluxation

is a contraindication to UKA
4.4.5 Absence of the Anterior
Cruciate Ligament (ACL)

It is necessary to differentiate previous post-­

traumatic rupture of the ACL, often accompa-
nied by meniscal lesions, from “trophic”
ruptures occurring progressively in a degenera-
tive process. Such secondary types of OA in
younger, active patients are characterised by
asymmetrical loss of tibiofemoral substance
related to anterior translation of the tibia (ante-
rior wear on the condyle and posterior wear on
Fig. 4.10 Patellofemoral osteophytes does not
the tibia). The anterior cartilage in the tibia
contraindicate UKA
remains intact for a long time, and an X-ray in
extension can appear falsely normal; only a
weightbearing view X-ray (Schuss) with 30° to
4.4.4 Overweight Patients 40° flexion will always be useful because the
cupule is posterior. According to Deschamps, if
Excess weight (BMI > 30 kg/m2 or weight > 82 kg) a profile view with weightbearing in extension
may expose the patient to premature wear of the shows 10-mm anterior translation of the tibia, a
polyethylene followed by loosening of a tibial UKA is strongly contraindicated [15]. In these
implant [13, 20]. cases, conduct of a UKA will be sanctioned
Many series have not shown a difference early by anterior loosening of the implant or
between patients with BMI < or > 30 kg/m2. posterior wear (Fig. 4.11) [17, 41]. It is possible
Cavaignac et al. compared 200 cases of UKA to combine UKA and simultaneous ACL sur-
(full poly-cemented) in patients whose BMI was gery with good results in these cases of second-
<30 kg/m2 with 80 UKA in patients whose BMI ary rupture. Perioperative discovery of
was >30 kg/m2. There was no difference in the degenerative damage to the ACL is not a contra-
Knee Society Score or survival rate at 12 years’ indication to UKA because there is no fixed
mean follow-up [37]. These results are confirmed anterior translation of the tibia due to capsule
by series by Tabor et al. [38] (80% survival at rigidity and posterior tibial osteophytes. Sagittal
20 years’ follow-up) and Xing et al. [39] (178 positioning of the tibial implant should not leave
UKA at 2 years’ follow-up). Results are also a sagittal slope ≥ 5° [41]. Combined one-stage
good in mobile-bearing-plateau UKA. Pandit surgery remains difficult with longer follow-up.
et al. [16] studied the results of 1000 mobile-­ Medium-term (5-year) results of UKA with a
bearing-­plateau UKA according to Kozinn and deteriorated ACL do not differ from those of
Scott criteria. With 10 years’ follow-up, these UKA with a competent ACL [42].
4 Conventional Indications for Unicompartmental Knee Arthroplasty 45

References
1. Lobenhoffer P. Indication for Uni knee replacement
versus osteotomy around the knee. J Knee Surg.
2017;30(8):769–73.
2. Rodriguez-Merchan CR. Unicompartmental knee
osteoarthritis (UKOA): unicaaaompartmental knee
arthroplasty (UKA) or high tibial osteotomy (HTO)?
Arch Bone J Surg. 2016;4(4):307–13.
3. Hernigou P, Deschamps G. Prothèses uni comparti-
mentales du genou. In: Symposium SOFCOT 1995.
Rev Chir Orthop. 1996:Supp 1, pp. 23–60.
4. Koskinen E, Eskelinen A, Paavolainen P, Pulkkinen
P, Rernes V. Comparison of survival and cost-­
effectiveness between unicondylar arthroplasty and
total knee arthroplasty in patients with primary osteo
arthritis. Acta Orthopedica. 2008;79-4:499–07.
5. Murray DW, Liddle AD, Dodd CAF, Pandit H. UKA:
is the glass half full or half empty? Bone Joint J.
2015;97-B(Suppl A):3–8.
6. Argenson JN, Flechner X. Minimally invasive
UKA. Knee. 2004;11:341–7.
7. Murray DW, Parkinson RW. Usage of UKA. Bone
Joint J. 2018;100-B(4):432–5.
8. Swiss Register. http://www.swissorthopaedics.ch/
images/content/SIRIS/KurzfassungF.pdf.
9. Argenson JN, Blanc G, Aubaniac JM, Parratte
S. Modern unicompartmental knee arthroplasty with
cement: a concise FU at a mean of 20 years of a previ-
ous report. J Bone Joint Surg Am. 2013;95-A:905–9.
10. Kim TK, Mittal A, Meshram P, Kim WH, Choi
Fig. 4.11 Early failure of UKA (radiolucent line under SM. Evidence-based surgical technique for medial
the tibial plateau) by anterior subluxation of the tibia unicompartmental knee arthroplasty. Knee Surg
(previous neglected rupture of the ACL) Related Res. 2021;33:2. Online 2021 Jan 7.
11. Liddle AD, Pandit H, Judge A, Murray DW. Patient
report outcomes following total and uni knee replace-
ment: a study of 14076 matched patients from NJR for
4.5 Conclusion England and Wales. Bone Joint J. 2015;97-B:793–801.
12. Newman J, Pydisetty RV, Ackroyd
C. Unicompartmental or total knee replacement. The
“Conventional indications” for UKA are too 15 years results of a prospective randomized con-
restrictive and should be revised. If they were trolled trial. J Bone Joint Surg Am. 2009;91(1):52–7.
applied, only 10% of AMOA would be eligible 13. Deschamps G, Chol C. Fixed bearing uni knee arthro-
[10]. It is possible to exceed reasonably the 20% plasty. Patient selection and operative technique.
Orthop Trauma Surg Res. 2011;97:648–61.
rate in AMOA. The risk of revision surgery (due 14. Deng M, Hu Y, Zhang Z, Zhang H, Qu Y, Shao
to tibial malposition) of UKA is much higher G. Unicondylar knee replacement versus total knee
when the annual number of UKA procedures replacement for the treatment of medial knee osteo-
performed by a surgeon is low. “With usage rates arthritis: a systematic review and meta analysis. Arch
Orthop Trauma Surg. 2021;141:1361–72.
of 20% or more, patients should have all the ben- 15. Gulati A, Pandit H, Jenkins C, Chau R, Dodd CA,
efits of UKA without the high revision rate” [5, Murray DW. The effect of leg alignment on the out-
10, 11]. come of uni knee replacement. J Bone Joint Surg Br.
2009;91-B:469–74.
46 C. V. Chainard and H. Robert

16. Pandit H, Jenkins C, Gill HS, Price AJ, Murray PE tibial components: the effects of knee alignment
DW. Unnecessary contraindications for mobile-­ and polyethylene thickness on creep and wear. Open
bearing UKA. J Bone Joint Surg Br. 2011;93-B:622–8. Orthop J. 2008;11(2):51–6.
17. Deschamps G, Lapeyre B. Rupture of the ACL is fre- 30. Pandit H, Kendtick B, Bottomley N, Price A, Murray
quently unrecognised cause of failure of UKA. Rev D, Dodd C. The implications of damage to the lateral
Chir Orthop Reparatrice Appar Mot. 1987;1:323–30. femoral condyle on medial unicompartmental knee
18. Chalmers BP, Mehrotra KG. Reliable outcomes and replacement. J Bone Joint Surg Br. 2010;92(3):374–9.
survivorship of unicompartmental knee arthroplasty 31. Morgan M, Brooks S, Nelson RA. Total knee
for isolated compartment osteonecrosis. Bone Joint J. arthroplasty in young active patients using a highly
2018;100-B(4):450–4. congruent fully mobile prosthesis. J Arthroplasty.
19. Parratte S, Argenson JN, Dumas J, Aubaniac 2007;22(4):525–30.
JM. Unicompartmental knee arthroplasty for 32. Duffy GP, Trousdale RT, Stuart MJ. Total knee arthro-
avascular osteonecrosis. Clin Orthop Relat Res. plasty in patients ≤ 55 years. Clin Orthop Relat Res.
2007;464:37–42. 1998;356:22–7.
20. Kozinn SC, Scott R. Unicondylar knee arthroplasty. J 33. Kennedy LG, Newman JH, Ackroyd CE, Dieppe
Bone Joint Surg Am. 1989;71-A:145–50. PA. When should we do knee replacements? Knee.
21. Deschamps G, Cartier P, Epinette JA, Hernigou 2003;10:161–6.
P. Indications et limites des prothèses 34. Hernigou P, Pascale W, Pascale V, Homma Y. La chon-
Unicompartimentales. Cahiers d’enseignement de la drocalcinose primaire ou secondaire influence-t-elle
SOFCOT. 1998;65:287–96. la survie à long terme d'une arthroplastie unicompar-
22. Lecuire F, Berard JB, Martres S. Minimum timentale ? Clin Orthop Relat Res. 2012;470:973–9.
10-year follow-up results of ALPINA cementless 35. Berger RA, Meneghini RM. The progression of patel-
hydroxyapatite-coated anatomic unicompartmen- lofemoral arthrosis after medial unicompartmental
tal knee arthroplasty. Eur J Orthop Surg Traumatol. replacement: results at 11 to 15 years. Clin Orthop
2014;24(3):385–94. Relat Res. 2004;428:92–9.
23. Lustig S, Elguindy A, Servien E. 5- to 16- year fol- 36. Hernigou P, Deschamps G. Patellar impinge-
low-­up of 54 consecutive lateral unicondylar knee ment following UKA. J Bone Joint Surg Am.
arthroplasties with a fixed-all polyethylene bearing. J 2002;84-A:1132–7.
Arthroplasty. 2011;26(8):1318–25. 37. Cavaignac E, Lafontan V, Reina N. Obesity has no
24. Peersman G, Stuyts B, Vanderlangenbergh T, Cartier adverse effect on the outcome of unicompartmental
P, Fennema P. Fixed- versus mobile-bearing UKA: knee replacement at a minimum follow-up of seven
a systematic review and meta-analysis. Knee Surg years. Bone Joint J. 2013;95-B(8):1064–8.
Sports Traumatol Arthrosc. 2015;23(11):3296–305. 38. Tabor OB Jr, Tabor OB, Bernard M, Wan
25. Parratte S, Pauly V, Aubaniac JM, Argenson JN. No JY. Unicompartmental knee arthroplasty: long-term
long-term difference between fixed and mobile success in middle-age and obese patients. J Surg
medial unicompartmental arthroplasty. Clin Orthop Orthop Adv. 2005;14(2):59–63.
Relat Res. 2012;470:61. 39. Xing Z, Katz J, Jiranek W. Unicompartmental knee
26. Berger RA, Meneghini RM, Jacobs JJ. Results of arthroplasty: factors influencing the outcome. J Knee
unicompartmental knee arthroplasty at a minimum Surg. 2012;25(5):369–73.
of ten years of follow-up. J Bone Joint Surg Am. 40. Bonutti PM, Goddard MS, Zywiel MG. Outcomes of
2005;87(5):999–1006. unicompartmental knee arthroplasty stratified by body
27. Argenson JN, Parratte S, Bertani A. The new arthritic mass index. J Arthroplasty. 2011;26(8):1149–53.
patient and arthroplasty treatment options. J Bone 41. Hernigou P, Deschamps G. Posterior slope of the
Joint Surg Am. 2009;91(Suppl 5):43–8. tibial implant and the outcome of UKA. J Bone Joint
28. Pennington DW, Swienchowski JJ, Camargo Surg Am. 2004;86-A:506–11.
M. Unicompartmental knee arthroplasty in patients 42. Boissonneault A, Pandit H. No difference in survivor-
sixty years of age or younger. J Bone Joint Surg Am. ship after unicompartmental knee arthroplasty with or
2003;85-A(10):1968–73. without an intact anterior cruciate ligament. Knee Surg
29. Hernigou P, Poignard A, Filippini P, Zilber Sports Traumatol Arthrosc. 2013;21(11):2480–6.
S. Retrieved unicompartmental implants with full
 The Modern Indications for Medial
UKA the “Oxford Philosophy” 5
Deciphered

T. Gicquel, J. C. Lambotte, F. X. Gunepin,

and Arnaud Clavé

Since the first unicompartmental knee arthroplas- stricter selection criteria than those initially pub-
ties (UKA) in the 1950s, the implants’ design, lished. According to regularly published criteria,
surgical technique and patient selection criteria only 2–6% of knee replacement surgery patients
have evolved with the early failures and suc- would be eligible for UKA.
cesses, improving satisfaction, functional results, In contrast, the Oxford team of John
and implant survival. Goodfellow and John O’Connor, designer of the
Progressively, the indications have been Oxford Unicompartmental Knee Arthroplasty
refined and the list of contraindications length- (OUKA), has in the last three decades validated
ened. In 1989, Kozinn and Scott published in the broader patient selection criteria [2, 3] that can
American JBJS a Current Concept Review refer- currently be considered “modern”. In particular,
ring to and proposing the characteristics of the they have demonstrated that the rate of revision
ideal candidate to undergo UKA (Table 5.1) [1]. surgery for UKA decreases when the number of
In the international literature, proposals have arthroplasties performed by the surgeon increases
since been revised and sometimes changed in (Table 5.2 and Figs. 5.1 and 5.2). This phenome-
order to be formulated into absolute selection cri- non seems to relate to significant differences in
teria under the term “Kozinn and Scott criteria” the level of technical control, patient selection
or “traditional criteria”. In reality, most publica- quality, and threshold for early revision surgery
tions using “Kozinn and Scott criteria” refer to between surgeons with high and low numbers of
arthroplasties performed [6].
The latter data has led to awareness that if
T. Gicquel · F. X. Gunepin
Clinique Mutualiste de la Porte de L’Orient, the technical benefit for patients is to be sus-
Lorient, France tained, it is important to increase the number
Rennes University Teaching Hospital. Orthopaedic of replacement surgeries. Yet this increase can
and Traumatology Surgery Department, only be made uniformly by redefining the
Rennes, France selection criteria, i.e. increasing the indica-
J. C. Lambotte tions and/or ignoring any unnecessary
Rennes University Teaching Hospital. Orthopaedic contraindications.
and Traumatologie Surgery Department,
The widening of the selection criteria, nev-
Rennes, France
ertheless, should be scientifically validated for
A. Clavé (*)
application with complete patient safety.
Orthopaedic and Traumatology Surgery Department,
Saint-George Private Hospital, Nice, France Therefore, this postulate raises questions
that we will attempt to answer:
LaTIM, UMR 1101 INSERM-UBO, Brest, France

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2024 47

A. Clavé, F. Dubrana (eds.), Unicompartmental Knee Arthroplasty,
https://doi.org/10.1007/978-3-031-48332-5_5
48 T. Gicquel et al.

Table 5.1 Profile of the ideal candidate to undergo medial unicompartmental arthroplasty (UKA) according to the
original Kozinn and Scott text published in 1989 [1]
Age Greater than 60 years and low functional demand.
But cementless implants can be indicated in younger patients who meet the other criteria.
Weight Less than 82 kg.
Level of activity The patient should not be physically extremely active or perform overly strenuous work.
Pain Should not be painful at rest because this may be the sign of an inflammatory component.
The procedure better relieves pain produced by weightbearing and walking.
Joint mobility Flexion arc greater than 90°.
Permanent flexion deformity less than 5°.
Deformity Varus deformity less than 10°.
The deformity should be suitable for perioperative correction after removal of the osteophytes.
Perioperative The final decision should be made after arthrotomy.
considerations Very small cartilaginous lesions in the non-weightbearing area of the lateral compartment are
not contraindications to UKA.
Patellofemoral pain is a relative contraindication.
Asymptomatic patellar chondromalacia is not a contraindication.
In cases of patellar subchondral exposure or in a weightbearing area of the lateral compartment,
total knee arthroplasty is recommended.
The two cruciate ligaments should be intact.
UKA is an effective treatment of avascular necrosis localised in a single compartment.
Patients with inflammatory disorders such as rheumatoid arthritis are not good candidates for
UKA.
Radiological chondrocalcinosis is a relative contraindication.

Table 5.2 Relationship between the number of Number of UKA implanted Annual revision rate
unicompartmental arthroplasties (UKA) performed annually (%)
each year by a surgeon and the revision rate based on Between 1 and 2 4
figures from the UK National Joint Registry,
10 2
according to Liddle et al. [4]
30 1

Fig. 5.1 From Liddle

et al. [5]
5 The Modern Indications for Medial UKA the “Oxford Philosophy” Deciphered 49

Fig. 5.2 From Liddle [5]

Is it possible to increase the annual number 5.1.1 Anteromedial Osteoarthritis

of UKA by widening the indications? (AMOA) [10]
Is it possible to increase the annual volume
of replacement procedures by ignoring certain Tibiofemoral anteromedial osteoarthritis
“traditional” contraindications? (AMOA) corresponds to a nosological entity
Particular case of ACL: What strategy resulting in relatively isolated wear on the inter-
should be adopted in patients with medial tib- nal tibiofemoral compartment in the anterome-
iofemoral OA and a nonfunctional or rup- dial aspect with a functional anterior cruciate
tured ACL? ligament. It is the main indication for UKA.
The clinical and radiological physical diagno-
sis is explained by several pathophysiological
5.1 Can the Annual Number characteristics:
of Arthroplasties
Be Increased by Widening 1. Tibial wear (on the cupula) is anterior because
the Indications? the anterior cruciate ligament is functional.
In complete extension, the knee has a varus
The main indication chosen by the authors and deformity. The posterior condylar shell is par-
generally in the literature is tibiofemoral antero- tially retracted or in conflict on osteophytes,
medial OA (AMOA) [7, 8]. This is a nosological preventing correction of the varus deformity
concept that has been proposed and then vali- (Fig. 5.3b) and contributing to permanent
dated by the Oxford team since the 1980s [7, 8] flexion deformity of the knee (Fig. 5.3a).
and corresponds to medial tibiofemoral knee OA Moreover, postoperatively, its progressive
with a competent anterior cruciate ligament and stretching will make it possible to see this
for which simple, reproducible diagnostic criteria flexion deformity correct itself until relatively
have been established [9]. late progression.
50 T. Gicquel et al.

2. Since the posterior parts of the tibial plateau 3. As a result of continuous flexion–extension
and femoral condyle are not worn out, when movements during the day, the medial collat-
the knee is bent 90° the point of tibiofemoral eral ligament (MCL) is stretched during each
contact is made in an area of preserved carti- flexion, which prevents its retraction.
lage. The height of the joint space is restored Therefore, the deformity is reducible during a
(Fig. 5.3c) and the varus deformity spontane- forced valgus movement at 20°, making it
ously reduced with the patient in the seated possible to relax the posterior condylar shell
position (Fig. 5.3d). (Fig. 5.3e and f).

Fig. 5.3 Pathophysiology of

anteromedial wear of the knee. (a and
a b
b) in maximum extension, the retracted
condylar shell limits extension and
prevents reduction of the varus
deformity. (c and d) in flexion at 90°,
the unworn part of the joint surface is
in contact, making it possible to reduce
varus wear and prevent retraction of
the medial collateral ligament. (e and f)
unlocked knee (20° flexion), the
relaxed condylar shell makes it
possible to reduce varus wear

c d

e f
5 The Modern Indications for Medial UKA the “Oxford Philosophy” Deciphered 51

a b c

Fig. 5.4 X-rays of the same knee: anterior view with weightbearing (a) showing incomplete wear (Ahlbäck 2) increas-
ing in weightbearing view (b) and in forced varus (c), which reveal Ahlbäck 3 bone-on-bone contact

5.1.1.1 Clinically dard X-ray film, a film in schuss (weightbear-

Assessment of functional impact is the main fac- ing) position (Rosenberg view) or a forced
tor making it possible to establish the indication varus view may be necessary (Fig. 5.4).
for prosthetic surgery (TKR or UKA). In an To propose UKA, complete wear with
AMOA presentation, in our view, the location of exposure of subchondral bone is essential. A
pain does not appear to be a decisive factor UKA procedure in cases of incomplete wear
between TKR and UKA. Therefore, although exposes the patient to a high risk of residual
patients describe medial pain (the “finger sign”) pain, a poor functional result with 25% of
in most cases, this is not essential because in a patients not being improved, and an implant
third of cases pain may not be solely medial and revision rate increased sixfold [14–16].
may also be anterior, posterior, and more rarely 2. A profile X-ray view (Fig. 5.5) makes it pos-
lateral, which does not affect the outcome [11]. sible to locate the tibial cup and ensure that
Ligament testing should find reducibility of the anterior cruciate ligament is indeed func-
varus (either by forced valgus at 20° flexion or in tional [9]. In fact, without a functional ACL,
seated position), confirming that the MCL is not the biomechanics of the knee are altered and
retracted and suggesting that we are indeed in the the internal compartment no longer plays its
AMOA setting. Assessment of the ACL by the role of a stable medial pivot during flexion–
Lachman–Trillat test or anterior “drawer test”, extension, which is the source of wear on the
which has poor sensitivity in this context, is not anterior cupula. The internal femoral condyle
considered sufficient to predict the condition of will have greater sagittal travel and move back
the ACL [12, 13]. into the tibia, ultimately resulting in wear of
the posterior part of the tibial plateau. For
5.1.1.2 Radiologically some authors, however, this assessment is not
A radiographic assessment is sufficient but cer- always easy to do [17] and therefore raises the
tain X-ray views are essential: potential perioperative risk of the procedure
being converted into TKR. MRI in this case
1. An anteroposterior view with weightbearing could make it possible to conclude with
should reveal complete wear of the tibiofemo- greater certainty [13].
ral compartment with bone-on-bone contact 3. An X-ray view in forced valgus should make
(Ahlbäck >3). If this is not visible on a stan- it possible to assess the status of the lateral
52 T. Gicquel et al.

Fig. 5.5 Anterior wear

in a profile X-ray view
and presentation of tibial
wear perioperatively

a b

Fig. 5.6 Tibiofemoral knee OA, weightbearing view (a) the cartilage height of the lateral tibiofemoral compart-
and in forced valgus (b), making it possible to confirm the ment despite the presence of lateral osteophytosis and
reducibility of varus wear deformity and conservation of meniscal calcinosis

tibiofemoral compartment and its conserva- the MCL is not retracted and retains its physi-
tion. The presence of lateral osteophytes is not ological characteristics (Fig. 5.6).
a contraindication to medial UKA [18], but a 4. A patellofemoral X-ray series (Skyline view)
loss of lateral cartilage height should exclude making it possible to assess this joint. See
medial UKA [19]. Moreover, it confirms that below.
5 The Modern Indications for Medial UKA the “Oxford Philosophy” Deciphered 53

Hamilton and Clavé [9] have validated a five knees, however, Chalmers et al. recommend
decision-making checklist: the Oxford vigilance in light of the enhanced risk of infec-
Radiological Decision Aid (https://www. tion and loosening of the prosthesis found in
oxfordpartialknee.net/content/dam/zb-­ cases of osteonecrosis secondary to numerous
minisites/oxford-­p artial-­k nee-­h cp/docu- local injections of corticosteroids [27].
ments/oxford-­decision-­aid-­flyer.pdf). Although this disorder is much rarer [28],
osteonecrosis of the medial tibial plateau is also
accessible by UKA [29, 30].
5.1.2 Aseptic Osteonecrosis In all cases, preoperative MRI is necessary to
confirm the diagnosis and assess the extent of
It is possible to propose UKA in the setting of necrosis. The use of a cemented implant may be
femoral osteonecrosis. This has been demon- necessary, and its availability in the operating
strated for many years by a number of authors room should be verified if the operator does not
[20, 21] and survival in this indication reaches usually perform a cemented procedure, particu-
92% at 15 years [22–26]. In a series limited to larly a femoral implant (Fig. 5.7).

a b

c d

Fig. 5.7 Radiographic (a and b) and MRI (c and d) presentation of aseptic osteonecrosis of the medial femoral
condyle
54 T. Gicquel et al.

5.1.3 Medial Tibiofemoral OA After 5.1.4 Bicompartmental UKA

Tibial Osteotomy (HTO)
Two circumstances can lead to the placement of
Prior HTO is not a systematic obstacle to the con- two unicompartmental arthroplasties (including
duct of UKA [31]. Nevertheless, this indication is one medial UKA) in the same knee:
complex and has been little studied. Limits
should be determined both in terms of ligament –– Either at the outset, in the same surgical stage
status and alignment of the lower limb or bone in cases of bicompartmental wear, particularly
structure (mechanical medial proximal tibial involving the medial tibiofemoral compart-
angle, mMPTA) [32, 33]. ment. This proposal is not promoted, mainly as
The ideal situation is a lower limb in residual a precautionary measure by the Oxford team
varus, without hypercorrection of tibial varus [10], but many experienced authors offer it.
(mMPTA ≤90°). Preoperative MRI makes it –– or at an interval after an initial UKA in cases
possible to ensure a satisfactory lateral of secondary deterioration of another knee
compartment. compartment. In this chapter, we will only dis-
Use of fixed-plateau UKA and navigation cuss the use of medial UKA to treat deteriora-
ensures that the operator has not produced tion of the medial tibiofemoral compartment
excessive intra-articular correction [34] sometime after a patellofemoral or lateral tib-
(Fig. 5.8). iofemoral UKA.

a b d f

c e g

Fig. 5.8 AMOA 22 years after an HTO (a). The preop- ral compartment in an X-ray view in forced valgus (c). (d
erative assessment revealed reducible residual varus of 4° and e) Immediate postoperative X-ray views. (f and g)
(b) and conservation of the height of the lateral tibiofemo- X-ray views with 7 years’ postoperative follow-up
5 The Modern Indications for Medial UKA the “Oxford Philosophy” Deciphered 55

5.1.4.1 Medial Unicompartmental 5.1.5 OA After Fracture

Arthroplasty and Simultaneous
Patellofemoral Arthroplasty Indications for medial UKA after a fracture are
The status of the patellofemoral compartment is very rare. There are no series in the literature.
seldom a contraindication to UKA (see the fol- Particular vigilance should be paid to possible
lowing chapter) and therefore indications for ligament lesions concomitant to fracture, particu-
combined medial UKA and patellofemoral UKA larly MCL, which can contraindicate placement
are rare [12]. of UKA.

5.1.4.2 Simultaneous Medial

and Lateral Unicompartmental 5.1.6 Joint Deterioration
Arthroplasty in Inflammatory Disorders
It is useful to remember that initially the OUKA
designed in 1976 was implanted solely as a Joint deteriorations related to active systemic
Bi-UKA until 1982, the year when the indication inflammatory disease of the knee joint are not
for medial UKA and the AMOA concept pre- indications for UKA. No recent publication has
vailed [10]. reported the use of unicompartmental implants in
A recent review of the literature [35] suggests this context.
that this combination can be proposed to patients
presenting with bicompartmental tibiofemoral
OA, with a nonsymptomatic patellofemoral joint, Take-Home Messages
an intact anterior cruciate ligament, a reducible 1. Tibiofemoral knee anteromedial OA
deformity, and conserved joint mobility. (AMOA) and aseptic osteonecrosis of
Few results are reported in the literature and it the femoral condyle are the main indi-
is not possible for the time being to recommend cations for unicompartmental
this strategy. Its utility may lie primarily in pres- arthroplasty.
ervation of the joint biomechanics and proprio- 2. It is possible to perform UKA in other
ception provided by preservation of the cruciate types of mechanical deterioration of the
ligaments [36]. knee, but there are few such indications
and results are unpredictable.
5.1.4.3 Medial Unicompartmental 3. Joint deteriorations in systemic inflam-
Arthroplasty at an Interval matory diseases are not indications for
After Another UKA.
Unicompartmental
Arthroplasty
In the setting of deterioration of the medial tibio-
femoral compartment after an initial procedure, 5.2 Can the Annual Number
UKA seems to be an attractive alternative to TKR of Procedures Be Increased
to treat deterioration of the medial compartment by Limiting the Current
after lateral UKA [37]. In this case, the first UKA Contraindications?
should have enabled significant relief during a
certain period of time and deterioration of the Although it is not possible to increase the number
medial compartment secondarily until bone-on-­ of procedures by widening the indications, it
bone contact in order to propose medial UKA. does appear important to examine the validity of
56 T. Gicquel et al.

the “traditional” contraindications. When Kozinn older subjects [48–52], data from registers often
and Scott criteria are strictly applied, less than show evidence of an unfavourable effect of
10% of surgical treatments of knee OA are eligi- younger age on implant survival [53, 54].
ble for UKA [38]. Yet under 10%, the annual In comparison with TKR [55], UKA in
revision rate is greater than 2% [39]. When the younger patients may enable better mobility
UKA usage rate increases and reaches a level of without significantly improving functional
between 40% and 60%, the revision rate decreases scores.
and is not then significantly different from that of
TKR (Fig. 5.2) [5, 39]. These rates can only be
reached by reducing the list of “traditional” con- Viewpoint
traindications. High body weight, young age, Treatment of knee OA in younger patients
major physical activity, radiological chondrocal- is complex. Although all of them suffer
cinosis, clinical and radiological evidence of intensely from knee symptoms, this popu-
patellofemoral damage, medial subluxation of lation is relatively heterogeneous in terms
the tibia, and osteophytes in the lateral plateau of physical activity, occupational activity,
should no longer be considered as formal contra- and intra-articular lesions.
indications to the conduct of UKA. For Pandit Surgeons faced with increasing demand
and Hamilton, patients presenting with these cri- from their patients should not propose
teria, deemed “contraindications” by Kozinn and UKA if the indication is not clearly estab-
Scott, have a UKA result that is at least as good lished. It is imperative that joint space nar-
in both function and survival as patients who rowing is complete. Although this was
present with “ideal” criteria according to Kozinn proposed for a time, perhaps wrongly,
and Scott [40, 41]. through the concept of transient UKA [56],
UKA should not be seen as a simpler alter-
native to TKR that is avoided because the
5.2.1 Body Weight patient is too young and/or the joint is not
worn out enough.
Obese patients have similar implant survival and In surgical registers, the overall number
similar or even better functional improvement to of UKA is relatively high but a large pro-
patients in other BMI classes [42–46]. Moreover, portion of the implants are inserted by sur-
survival seems independent of BMI [38]. geons who individually perform few
In comparison with TKR [47], UKA in obese procedures [4]. The hypothesis can reason-
patients may be useful with better postoperative ably be formulated that if such procedures
joint mobility, lower risk of infection (0% vs. involve younger patients, the indication is
0.5%), and decreased need for mobilisation dur- not or seldom mastered, explaining in part
ing general anaesthesia (3.7% versus 9.2%). the poor results in registers on these
Therefore, body weight is not a contraindica- patients.
tion to UKA which, if we use the Kozinn and
Scott restrictive criteria, would prevent its con-
duct in half of patients [40] (Fig. 5.9). In patients with a high level of activity preop-
eratively, implant survival and function are at
least as good as in patients with a lower level of
5.2.2 Age and Physical Activity activity. Therefore, the level of physical activity
should not be considered as a contraindication to
Although younger patients (≤ 50 years, ≤55 years UKA [57, 58].
or ≤ 60 years of age, according to studies) in On the other hand, patients over the age of
cohort studies have results (functional, quality of 75 years present results that are at least as good as
life, and implant survival) that are as good as in in younger patients or those who undergo TKR
5 The Modern Indications for Medial UKA the “Oxford Philosophy” Deciphered 57

Fig. 5.9 (a) Post, BMI

preoperative, and delta
Oxford knee score a OKS <25
according to BMI; (b)
survival over time 25-30
depending on 40 30-35
BMI. From Murray et al.
[42] 35-40

>40
30

20

10

0
Postop Preop Change

b
1.00 45

0.98

0.96
Cum Survival

0.94

0.92

0.90

0 1 2 3 4 5 6 7 8 9 10
Follow up period
58 T. Gicquel et al.

[59]. In older and more fragile patients, the 5.2.3 Flexion Deformity
choice of UKA over TKR seems logical to avoid
the more serious complications with TKR Up to 15°, flexion deformity is not a contraindi-
[60–62]. cation to UKA. Intra-articular procedures, par-
Because of the many external factors influ- ticularly release of the intercondylar notch and
encing the choice, the decision between TKR, removal of osteophytes (at the foot of the ACL,
HTO (high tibial osteotomy), and UKA in behind the medial condyle and medial tibial pla-
younger patients has not been fully deter- teau), make it possible to correct a major part of
mined, but age and physical activity should the flexion deformity. If all these procedures for
not be deemed a­bsolute contraindications at release have been correctly performed, a possible
the risk of needlessly excluding 24% and 10% postoperative minor flexion deformity will cor-
of patients from UKA, respectively [40] rect itself over the year after surgery by progres-
(Fig. 5.10). sive stretching of the periarticular structures.

Fig. 5.10 (a) survival

over time according to
a
age (> or < at 60 years);
(b) survival at 12 years
and Oxford knee score
based on Tegner activity
score [57] and Murray
et al. The London Knee
Meeting 2014

b
5 The Modern Indications for Medial UKA the “Oxford Philosophy” Deciphered 59

Fig. 5.11 Radiographic criteria for patellofemoral wear tialknee.net/content/dam/zb-­m inisites/oxford-­p artial-­
adapted from the Oxford Decision Aid [9]. If any criteria knee-­h cp/documents/oxford-­d ecision-­a id-­f lyer.pdf)
are “met”, the status of the patellofemoral compartment is (courtesy of Zimmer-Biomet)
not a contraindication to UKA (https://www.oxfordpar-

Beyond 15°, the status of the ACL and the in the absence of severe lateral wear, a defective
causes of flexion deformity must be correctly patellofemoral radiological alignment does not
assessed. In fact, flexion of more than 15° is fre- affect the functional result of UKA according to
quently related to a deficient or absent ACL and/ some authors [71].
or major deterioration of the other knee The presence of cartilage lesions, perhaps dis-
compartments. covered perioperatively, should not needlessly
worry the surgeon. Even when loss of substance
is total (Fig. 5.11), including from the bottom of
5.2.4 Patellofemoral Joint the trochlea, the result of UKA is unaffected if
this cartilage loss does not involve the lateral
The status of the patellofemoral joint has always facet of the patella [64, 68, 72–74]. Therefore,
been the source of tension and debate. Currently, there is no need to change the indication periop-
it continues to be an area of concern and uncer- eratively, but patients must be informed that ante-
tainty for many surgeons, often leading as a pre- rior pain can take time to disappear and that
cautionary measure to preferring TKR. residual discomfort on climbing stairs can persist
In most cases, however, the condition of the [75] (Fig. 5.12).
patellofemoral joint does not represent a contra-
indication to the conduct of medial UKA.
Clinically, the presence of anterior pain is not 5.2.5 Chondrocalcinosis
a factor for dissatisfaction or failure [63, 64]. The
existence of an authentic patellofemoral pain Although meniscal chondrocalcinosis (meniscal
syndrome [65] has never been studied in the lit- calcinosis) is not always associated with histo-
erature and would merit further investigation. logical chondrocalcinosis and reciprocally, it
Radiologically, whenever the patella remains seems that patients who have histological chon-
centred or the wear is solely medial, this does not drocalcinosis are at greater risk of having
have any consequences [63, 66–69]. In contrast, decreased implant survival [76].
it is necessary to be vigilant in cases of lateral The presence of radiological signs of chon-
subluxation or lateral wear [70] and to consider drocalcinosis (meniscal calcifications) is not a
this as a potential contraindication. Nevertheless, contraindication to the conduct of UKA [76, 77].
60 T. Gicquel et al.

Fig. 5.12 According to

Hamilton [64], survival
of medial UKA with and
without full-thickness
cartilage loss in the
patellofemoral joint

5.3 Specific Case: What Strategy

Should Be Adopted
in Patients with Medial
Tibiofemoral OA
and a Nonfunctional or
Ruptured ACL?

The question of the possibility of performing uni-

compartmental arthroplasty on a knee that pres-
ents with a nonfunctional anterior cruciate
Fig. 5.13 Survival over time according to compliance
with Kozinn and Scott criteria. From Pandit et al. [40] ligament is delicate. The absence of a functional
anterior cruciate ligament has long been consid-
ered as an absolute contraindication due to more
Take-Home Messages frequent tibial loosening of the implant and pre-
1. If the surgical indication is correctly mature polyethylene wear [2, 78, 79], probably
established (tibiofemoral knee antero- explained by biomechanical changes such as the
medial OA), the patient’s age and physi- pushing back of the tibiofemoral contact point
cal activity are not absolute [80] and sliding rather than rolling of the femoral
contraindications to medial UKA. condyle [81] (soaping phenomenon).
2. Analysis of the patellofemoral joint’s However, the good functional results of UKA
condition is difficult. Patellofemoral in younger patients with high functional demand
contraindications to UKA are marginal [57, 58] and the procedure’s low morbidity in
and mainly related to extreme wear of fragile elderly patients [60] have encouraged
the joint’s lateral facet. Anterior pain some teams to use these implants even in patients
does not predict an unfavourable result with a nonfunctional ACL.
of UKA. The population of patients with a nonfunc-
3. Unjustified contraindications are tional ACL in medial tibiofemoral OA is hetero-
responsible for limiting the use of UKA geneous and it is possible to differentiate the
by surgeons in the treatment of knee three different nosological entities described
anteromedial OA (AMOA) (Fig. 5.13). below.
5 The Modern Indications for Medial UKA the “Oxford Philosophy” Deciphered 61

5.3.1 Advanced Anteromedial OA Patients have not always complained of instabil-

(Advanced AMOA) ity or have changed their physical activities to
avoid it.
This occurs in generally elderly patients, over 70 This corresponds to Postero Medial Osteo
years of age, who initially have tibiofemoral knee Arthritis (PMOA) as opposed to Antero Medial
anteromedial OA, but progression of the disorder Osteo Arthritis (AMOA), although the term is not
and, in particular, of osteophytes of the intercon- used in the literature.
dylar notch has gradually led to loss of function- As an alternative to TKR, it appears possible
ing due to erosion of the ACL. to propose UKA together with ligamentoplasty
Patients rarely complain of knee instability of the ACL. Initial functional results are very sat-
because loss of ACL function is progressive and isfactory and likely meet these younger patients’
the knee is stabilised by the osteoarthritic expectations [88–93]. Nevertheless, long-term
ankylosis. survival remains little documented and requires
As an alternative to TKR, UKA without recon- more extensive studies [92, 94]. In addition, the
struction of the ACL can be considered, taking combination of these two procedures makes them
care to decrease the tibial slope in comparison to more complex and should be reserved for sur-
either the native slope of the medial tibial plateau geons who are experienced in both.
or the implant’s usual slope [82–85]. In return,
implant survival and functional result do not
seem to be different from that of UKA with a 5.3.3 Post-Ligamentoplasty
functional ACL in a population of the same age Tibiofemoral Knee
[84, 86, 87]. Nevertheless, it is necessary to be Anteromedial OA (Post-­
cautious during surgery because balancing such Ligamentoplasty AMOA)
implants is made difficult by the imperative to
reduce the slope, possible retraction of the ACL, This nosological entity is probably the most
and posterior progression of wear. complex of all because it necessitates assess-
ing the functional aspect of the ligamentoplasty
(Table 5.3). This form of OA affects younger
5.3.2 Posteromedial OA (PMOA) or even very young patients (sometimes under
50 years of age) who often have high to very
This affects younger patients, generally under 60 high functional demand and whose knee has
years of age. Rupture of the ACL is traumatic, already “benefitted” from several ligament
longstanding and sometimes ignored by patients, surgeries.
and wear on the tibial plateau is immediately pos- The conduct of TKR is clearly the simplest
terior by anterior tibial translation and wear on and most proven solution. Tibial valgus osteot-
the posterior segment of the medial meniscus. omy also retains all of its utility when wear is

Table 5.3 Factors for or against the functional aspect of a ligamentoplasty in a patient with medial tibiofemoral knee
OA and a history of reconstruction of the anterior cruciate ligament
Factors supporting nonfunctional
Factors supporting functional ligamentoplasty ligamentoplasty
Aetiology of Iatrogenic after meniscectomy in a pronounced Residual laxity responsible for meniscus/
deterioration varus knee deformity cartilage deterioration
History Only one reconstruction Repeated ligamentoplasty
Time aspect Recent ligamentoplasty Previous ligamentoplasty
Symptoms Stable Unstable
Laxitya No laxity Laxity
aLaxity is difficult to judge because ligamentoplasty can be relaxed by loss of cartilage height in the medial compart-
ment; the conduct of a UKA can sometimes retighten a ligamentoplasty that initially appears lax in testing
62 T. Gicquel et al.

moderate. Discussing the conduct of UKA is pos- The case of a Lemaire-type extra-articular
sible, but a precise analysis of the functional ligamentoplasty raises a different problem: the
aspect of a ligamentoplasty is essential even if it joint that is naive for prior ligamentoplasty
is complex. encourages the conduct of UKA + ligamento-
Therefore, in cases of satisfactory ligamento- plasty, but it is important to assess the lateral
plasty (Table 5.3), UKA may be considered. In compartment correctly and to be mindful of dete-
other cases, the combination of UKA with revi- rioration of the lateral tibiofemoral compartment,
sion of a ligamentoplasty, although theoretically particularly with older surgeries [95, 96].
feasible in our opinion, does not seem appropri-
ate due to the many technical difficulties in both
the choice of the type of graft and placement of 5.3.4 Take-Home Message
the tibial tunnel. (Table 5.4)

Table 5.4 Summary of various nosological categories in relation to medial tibiofemoral OA with a native nonfunc-
tional anterior cruciate ligament
1. Advanced AMOA 2. PMOA 3. Post-ligamentoplasty AMOA
Patient’s age “Elderly” “Young” “Young”
Pathophysiology Initial anteromedial Previous rupture To be determined in order to define the
OA of the ACL therapeutic project
Progressive erosion Initial posterior
of the ACL wear of the tibial
plateau
Stability Stable +/− stable +/− stable
Therapeutic UKA (or TKR) ACL + UKA (or UKA or TKR according to the condition
TKR) of the ligamentoplasty
5 The Modern Indications for Medial UKA the “Oxford Philosophy” Deciphered 63

5.4 Conclusion Furthermore, these cases are often complex

with an uncertain result.
Widening the selection criteria for UKA-­ Question 2: Is it possible to increase the
eligible patients is necessary in order to reach annual number of procedures by ignoring cer-
a number and rate of UKA which can limit the tain “traditional” contraindications?
rate of revision surgery. Widening the ­selection Although it is not possible to increase the
criteria essentially involves reviewing the limi- number of procedures by widening the indica-
tation of contraindications. However, it is not tions, it does appear important to examine the
appropriate to seek to widen the indication of validity of the “traditional” contraindications.
UKA at the risk of running into technical dif- When the Kozinn and Scott criteria are strictly
ficulties in its conduct and exposing the patient applied, less than 10% of surgical types of knee
to an uncertain result. OA are eligible for UKA [38]. Yet under 10%, the
Tibiofemoral knee anteromedial OA annual revision rate is greater than 2% [39].
(AMOA) is the prime indication for medial When the rate of UKA use increases and reaches
UKA and, if reasonable contraindications are a value of between 40% and 60%, the rate of revi-
followed, makes it possible to reach a UKA sion surgery decreases and is not then signifi-
rate of 30% to 40% in the clinical practice of cantly different from that of TKR (Fig. 5.2) [5,
this chapter’s authors. 39]. These rates can be reached by reducing the
list of “traditional” contraindications. High
weight, young age, major physical activity, radio-
5.5 Final Take-Home Messages logical chondrocalcinosis, clinical and radiologi-
cal patellofemoral damage, medial subluxation
Question 1: Is it possible to increase the annual of the tibia, and osteophytes in the lateral plateau
number of UKA by widening the indications? should not be seen as formal contraindications to
The main indication retained by the Oxford the conduct of UKA. Patients presenting with cri-
team is tibiofemoral anteromedial OA (AMOA) teria considered as “contraindications” under
[7, 8]. It is a nosological concept that they have Kozinn and Scott criteria have, according to
proposed and then validated since the 1980s [7, Pandit and Hamilton, a UKA result at least as
8], corresponding to medial tibiofemoral knee good in functional aspect and survival as patients
OA with a competent anterior cruciate ligament presenting with “ideal” Kozinn and Scott criteria
and for which simple, reproducible diagnostic [40, 41].
criteria have been established [9]. Avoiding unnecessary contraindications
Other indications of UKA exist, such as osteo- makes it possible to increase the rate of use
necrosis (femoral and tibial), bicompartmental and annual volume of UKA in a safe manner
OA (bi-UKA), failure of tibial valgus osteotomy and consequently to decrease the implant revi-
(TVO), post-traumatic OA, etc., but these are sion rate.
much rarer [7, 8]. Although widely validated in Question 3: Specific case: What strategy
femoral osteonecrosis, the use of UKA in other should be adopted in patients with medial tib-
indications remains marginal and the volume of iofemoral OA and a nonfunctional or rup-
publications is low. tured ACL?
AMOA is the main indication, representing The absence of an anterior cruciate ligament is
over 95% of UKA procedures for the Oxford a contraindication to UKA, whether according to
team. Good knowledge of it is therefore essen- “Kozinn and Scott” criteria or to the designers of
tial in the identification of patients potentially the OUKA system. Despite this, many surgeons
eligible for this surgery. continue to show interest in the topic and scien-
Widening the indications to marginal indi- tific publications are becoming more numerous.
cations (less than 5%) is not a good method of UKA is then conceived as an alternative to TKR
increasing the volume of such procedures. either to optimise the functional result in younger
64 T. Gicquel et al.

patients or as a result of its lower morbidity in the 10. Goodfellow J, O’Connor J, Pandit H, et al.
elderly. Unicompartmental arthroplasty with the Oxford knee.
2nd ed. Goodfellow Publishers Limited; 2015.
Nevertheless, a precise but difficult analysis 11. Liddle AD, Pandit H, Jenkins C, et al. Preoperative
of the nosological category is essential since pain location is a poor predictor of outcome after
the treatments proposed will not be the same Oxford unicompartmental knee arthroplasty
between elderly patients presenting with pro- at 1 and 5 years. Knee Surg Sports Traumatol
Arthrosc. 2013;21:2421–6. https://doi.org/10.1007/
gressive erosion of the ACL in advanced s00167-­012-­2211-­3.
AMOA, younger patients presenting with pos- 12. Scott CEH, Holland G, Krahelski O, et al. Patterns
teromedial OA or younger patients presenting of cartilage loss and anterior cruciate ligament status
with post-ligamentoplasty medial tibiofemoral in end-stage osteoarthritis of the knee. Bone Joint J.
2020;102-B:716–26. https://doi.org/10.1302/0301-­
OA. Not all of these patients will be eligible for 620X.102B6.BJJ-­2019-­1434.R1.
UKA, and for those who are the conduct of 13. Johnson AJ, Howell SM, Costa CR, Mont MA. The
combined ligamentoplasty must be discussed. ACL in the arthritic knee: how often is it pres-
ent and can preoperative tests predict its presence?
Clin Orthop Relat Res. 2013;471:181–8. https://doi.
org/10.1007/s11999-­012-­2505-­2.
References 14. Hamilton TW, Pandit HG, Inabathula A, et al.
Unsatisfactory outcomes following unicompartmen-
1. Kozinn SC, Scott R. Unicondylar knee arthroplasty. J tal knee arthroplasty in patients with partial thickness
Bone Joint Surg Am. 1989;71:145–50. cartilage loss: a medium-term follow-up. Bone Joint
2. Goodfellow JW, Kershaw CJ, Benson MK, O’Connor J. 2017;99-B:475–82. https://doi.org/10.1302/0301-­
JJ. The Oxford knee for unicompartmental osteo- 620X.99B4.BJJ-­2016-­1061.R1.
arthritis. The first 103 cases. J Bone Joint Surg Br. 15. Niinimäki TT, Murray DW, Partanen J, et al.
1988;70:692–701. Unicompartmental knee arthroplasties implanted for
3. Carr A, Keyes G, Miller R, et al. Medial unicom- osteoarthritis with partial loss of joint space have high
partmental arthroplasty. A survival study of the re-operation rates. Knee. 2011;18:432–5. https://doi.
Oxford meniscal knee. Clin Orthop Relat Res. org/10.1016/j.knee.2010.08.004.
1993;295:205–13. 16. Pandit H, Gulati A, Jenkins C, et al. Unicompartmental
4. Liddle AD, Pandit H, Judge A, Murray DW. Effect of knee replacement for patients with partial thick-
surgical caseload on revision rate following total and ness cartilage loss in the affected compartment.
unicompartmental knee replacement. J Bone Joint Knee. 2011;18:168–71. https://doi.org/10.1016/j.
Surg Am. 2016;98:1–8. https://doi.org/10.2106/ knee.2010.05.003.
JBJS.N.00487. 17. Heurtin T. Evaluation radiographique pré opératoire
5. Liddle AD, Pandit H, Judge A, Murray DW. Optimal de la gonarthrose femoro-tibiale médiale: reproduct-
usage of unicompartmental knee arthroplasty: a study ibilité, fiabilité et corrélation avec les atteintes liga-
of 41,986 cases from the National Joint Registry for mentaires: étude prospective. 2018.
England and Wales. Bone Joint J. 2015;97-B:1506– 18. Hamilton TW, Choudhary R, Jenkins C, et al.
11. https://doi.org/10.1302/0301-­620X.97B11.35551. Lateral osteophytes do not represent a contraindica-
6. Liddle AD. CORR insights®: no differences in out- tion to medial unicompartmental knee arthroplasty:
comes scores or survivorship of unicompartmental a 15-year follow-up. Knee Surg Sports Traumatol
knee arthroplasty between patients younger or older Arthrosc. 2017;25:652–9. https://doi.org/10.1007/
than 55 years of age at minimum 10-year followup. s00167-­016-­4313-­9.
Clin Orthop Relat Res. 2019;477:1447–9. https://doi. 19. Pandit H, Spiegelberg B, Clavé A, et al. Aetiology
org/10.1097/CORR.0000000000000797. of lateral progression of arthritis following Oxford
7. Goodfellow JW, Tibrewal SB, Sherman KP, medial unicompartmental knee replacement: a case-­
O’Connor JJ. Unicompartmental Oxford Meniscal control study. Musculoskelet Surg. 2016;100:97–102.
knee arthroplasty. J Arthroplast. 1987;2:1–9. https:// https://doi.org/10.1007/s12306-­015-­0394-­8.
doi.org/10.1016/s0883-­5403(87)80025-­6. 20. Cartier P, Gaggiotti G, Jully JL. [*primary osteonecro-
8. White SH, Ludkowski PF, Goodfellow sis of the medial femoral condyle. Unicompartmental
JW. Anteromedial osteoarthritis of the knee. J Bone or total replacement?]. Int Orthop. 1988;12:229–35.
Joint Surg Br. 1991;73:582–6. https://doi.org/10.1007/BF00547168.
9. Hamilton TW, Pandit HG, Lombardi AV, et al. 21. Marmor L. Unicompartmental arthroplasty for osteo-
Radiological Decision Aid to determine suitabil- necrosis of the knee joint. Clin Orthop Relat Res.
ity for medial unicompartmental knee arthroplasty: 1993;294:247–53.
development and preliminary validation. Bone Joint 22. Ollivier M, Jacquet C, Lucet A, et al. Long-term results
J. 2016;98-B:3–10. https://doi.org/10.1302/0301-­ of medial unicompartmental knee arthroplasty for
620X.98B10.BJJ-­2016-­0432.R1.
5 The Modern Indications for Medial UKA the “Oxford Philosophy” Deciphered 65

knee avascular necrosis. J Arthroplast. 2019;34:465– tal knee arthroplasty? Continuous series of fifty nine
8. https://doi.org/10.1016/j.arth.2018.11.010. procedures. Int Orthop. 2016;40:2511–8. https://doi.
23. Bruni D, Iacono F, Raspugli G, et al. Is unicompart- org/10.1007/s00264-­016-­3241-­0.
mental arthroplasty an acceptable option for sponta- 35. Wada K, Price A, Gromov K, et al. Clinical out-
neous osteonecrosis of the knee? Clin Orthop Relat come of bi-unicompartmental knee arthroplasty
Res. 2012;470:1442–51. https://doi.org/10.1007/ for both medial and lateral femorotibial arthritis: a
s11999-­012-­2246-­2. systematic review-is there proof of concept? Arch
24. Fukuoka S, Fukunaga K, Taniura K, et al. Medium-­ Orthop Trauma Surg. 2020. https://doi.org/10.1007/
term clinical results of unicompartmental knee s00402-­020-­03492-­6.
arthroplasty for the treatment for spontaneous osteo- 36. Fuchs S, Tibesku CO, Genkinger M, et al.
necrosis of the knee with four to 15 years of follow- Proprioception with bicondylar sledge prosthe-
­up. Knee. 2019;26:1111–6. https://doi.org/10.1016/j. ses retaining cruciate ligaments. Clin Orthop Relat
knee.2019.06.007. Res. 2003;406:148–54. https://doi.org/10.1097/01.
25. Langdown AJ, Pandit H, Price AJ, et al. Oxford medial blo.0000038053.29678.a5.
unicompartmental arthroplasty for focal spontaneous 37. Lustig S, Lording T, Frank F, et al. Progression
osteonecrosis of the knee. Acta Orthop. 2005;76:688– of medial osteoarthritis and long-term results of
92. https://doi.org/10.1080/17453670510041772. lateral unicompartmental arthroplasty: 10 to 18
26. Yoon C, Chang MJ, Chang CB, et al. Does unicom- year follow-up of 54 consecutive implants. Knee.
partmental knee arthroplasty have worse outcomes 2014;21(Suppl 1):S26–32. https://doi.org/10.1016/
in spontaneous osteonecrosis of the knee than in S0968-­0160(14)50006-­3.
medial compartment osteoarthritis? A systematic 38. Stern SH, Becker MW, Insall JN. Unicondylar knee
review and meta-analysis. Arch Orthop Trauma arthroplasty. An evaluation of selection criteria. Clin
Surg. 2019;139:393–403. https://doi.org/10.1007/ Orthop Relat Res. 1993;286:143–8.
s00402-­019-­03125-­7. 39. Murray DW, Liddle AD, Liddle A, et al.
27. Chalmers BP, Mehrotra KG, Sierra RJ, et al. Reliable Unicompartmental knee arthroplasty: is the glass
outcomes and survivorship of unicompartmental half full or half empty? Bone Joint J. 2015;97-B:3–8.
knee arthroplasty for isolated compartment osteone- https://doi.org/10.1302/0301-­620X.97B10.36542.
crosis. Bone Joint J. 2018;100-B:450–4. https://doi. 40. Pandit H, Jenkins C, Gill HS, et al. Unnecessary con-
org/10.1302/0301-­620X.100B4.BJJ-­2017-­1041.R2. traindications for mobile-bearing unicompartmental
28. Carpintero P, Leon F, Zafra M, et al. Spontaneous col- knee replacement. J Bone Joint Surg Br. 2011;93:622–
lapse of the tibial plateau: radiological staging. Skelet 8. https://doi.org/10.1302/0301-­620X.93B5.26214.
Radiol. 2005;34:399–404. https://doi.org/10.1007/ 41. Hamilton TW, Pandit HG, Jenkins C, et al. Evidence-­
s00256-­005-­0926-­7. based indications for mobile-bearing unicompart-
29. Kamenaga T, Hiranaka T, Hida Y, et al. mental knee arthroplasty in a consecutive cohort of
Unicompartmental knee arthroplasty for sponta- thousand knees. J Arthroplast. 2017;32:1779–85.
neous osteonecrosis of the medial tibial plateau. https://doi.org/10.1016/j.arth.2016.12.036.
Knee. 2018;25:715–21. https://doi.org/10.1016/j. 42. Murray DW, Pandit H, Weston-Simons JS, et al. Does
knee.2018.04.006. body mass index affect the outcome of unicompart-
30. Yang W-M, Zhao C-Q, Lu Z-Y, et al. Clinical char- mental knee replacement? Knee. 2013;20:461–5.
acteristics and treatment of spontaneous osteone- https://doi.org/10.1016/j.knee.2012.09.017.
crosis of medial tibial plateau: a retrospective case 43. Cavaignac E, Lafontan V, Reina N, et al. Obesity
study. Chin Med J. 2018;131:2544–50. https://doi. has no adverse effect on the outcome of unicompart-
org/10.4103/0366-­6999.244113. mental knee replacement at a minimum follow-up of
31. Valenzuela GA, Jacobson NA, Buzas D, et al. seven years. Bone Joint J. 2013;95-B:1064–8. https://
Unicompartmental knee replacement after high doi.org/10.1302/0301-­620X.95B8.31370.
tibial osteotomy: invalidating a contraindication. 44. Plate JF, Augart MA, Seyler TM, et al. Obesity has
Bone Joint J. 2013;95-B:1348–53. https://doi. no effect on outcomes following unicompartmen-
org/10.1302/0301-­620X.95B10.30541. tal knee arthroplasty. Knee Surg Sports Traumatol
32. Schlumberger M, Oremek D, Brielmaier M, et al. Arthrosc. 2017;25:645–51. https://doi.org/10.1007/
Prior high tibial osteotomy is not a contraindication s00167-­015-­3597-­5.
for medial unicompartmental knee arthroplasty. Knee 45. Affatato S, Caputo D, Bordini B. Does the body mass
Surg Sports Traumatol Arthrosc. 2020;29(10):3279. index influence the long-term survival of unicompart-
https://doi.org/10.1007/s00167-­020-­06149-­4. mental knee prostheses? A retrospective multi-centre
33. Jones GG, Clarke S, Jaere M, Cobb JP. Failed study. Int Orthop. 2019;43:1365–70. https://doi.
high tibial osteotomy: a joint preserving alterna- org/10.1007/s00264-­018-­4217-­z.
tive to total knee arthroplasty. Orthop Traumatol 46. Molloy J, Kennedy J, Jenkins C, et al. Obesity should
Surg Res. 2019;105:85–8. https://doi.org/10.1016/j. not be considered a contraindication to medial Oxford
otsr.2018.11.004. UKA: long-term patient-reported outcomes and
34. Gicquel T, Lambotte JC, Polard JL, et al. Is tibial cut implant survival in 1000 knees. Knee Surg Sports
navigation alone sufficient in medial unicompartmen-
66 T. Gicquel et al.

Traumatol Arthrosc. 2019;27:2259–65. https://doi. follow-up. J Arthroplast. 2019;34:1364–8. https://doi.

org/10.1007/s00167-­018-­5218-­6. org/10.1016/j.arth.2019.03.038.
47. Lum ZC, Crawford DA, Lombardi AV, et al. Early 59. Fabre-Aubrespy M, Ollivier M, Pesenti S, et al.
comparative outcomes of unicompartmental and Unicompartmental knee arthroplasty in patients
total knee arthroplasty in severely obese patients. older than 75 results in better clinical outcomes and
Knee. 2018;25:161–6. https://doi.org/10.1016/j. similar survivorship compared to total knee arthro-
knee.2017.10.006. plasty. A matched controlled study. J Arthroplasty.
48. Lee M, Chen J, Shi LC, et al. No differences in out- 2016;31:2668–71. https://doi.org/10.1016/j.
comes scores or survivorship of unicompartmental arth.2016.06.034.
knee arthroplasty between patients younger or older 60. Liddle AD, Judge A, Pandit H, Murray DW. Adverse
than 55 years of age at minimum 10-year followup. outcomes after total and unicompartmental knee
Clin Orthop Relat Res. 2019;477:1434–46. https:// replacement in 101330 matched patients: a study of
doi.org/10.1097/CORR.0000000000000737. data from the National Joint Registry for England
49. Felts E, Parratte S, Pauly V, et al. Function and qual- and Wales. Lancet. 2014. https://doi.org/10.1016/
ity of life following medial unicompartmental knee S0140-­6736(14)60419-­0.
arthroplasty in patients 60 years of age or younger. 61. Hunt LP, Ben-Shlomo Y, Clark EM, et al. 45-day
Orthop Traumatol Surg Res. 2010;96:861–7. https:// mortality after 467,779 knee replacements for
doi.org/10.1016/j.otsr.2010.05.012. osteoarthritis from the National Joint Registry
50. Greco NJ, Lombardi AV, Price AJ, et al. Medial for England and Wales: an observational study.
mobile-bearing unicompartmental knee arthro- Lancet. 2014;384:1429–36. https://doi.org/10.1016/
plasty in young patients aged less than or equal to S0140-­6736(14)60540-­7.
50 years. J Arthroplast. 2018;33:2435–9. https://doi. 62. Brown NM, Sheth NP, Davis K, et al. Total knee
org/10.1016/j.arth.2018.03.069. arthroplasty has higher postoperative morbidity than
51. Heyse TJ, Khefacha A, Peersman G, Cartier unicompartmental knee arthroplasty: a multicenter
P. Survivorship of UKA in the middle-aged. analysis. J Arthroplast. 2012;27:86–90. https://doi.
Knee. 2012;19:585–91. https://doi.org/10.1016/j. org/10.1016/j.arth.2012.03.022.
knee.2011.09.002. 63. Beard DJ, Pandit H, Ostlere S, et al. Pre-operative
52. Swienckowski JJ, Pennington DW. Unicompartmental clinical and radiological assessment of the patel-
knee arthroplasty in patients sixty years of age or younger. lofemoral joint in unicompartmental knee replace-
J Bone Joint Surg Am. 2004;86-A(Suppl 1):131–42. ment and its influence on outcome. J Bone
https://doi.org/10.2106/00004623-­200409001-­00004. Joint Surg Br. 2007;89:1602–7. https://doi.
53. Jeschke E, Gehrke T, Günster C, et al. Five-year sur- org/10.1302/0301-­620X.89B12.19260.
vival of 20,946 unicondylar knee replacements and 64. Hamilton TW, Pandit HG, Maurer DG, et al. Anterior
patient risk factors for failure: an analysis of German knee pain and evidence of osteoarthritis of the patel-
insurance data. J Bone Joint Surg Am. 2016;98:1691– lofemoral joint should not be considered contra-
8. https://doi.org/10.2106/JBJS.15.01060. indications to mobile-bearing unicompartmental
54. Liddle AD, Judge A, Pandit H, Murray knee arthroplasty: a 15-year follow-up. Bone Joint
DW. Determinants of revision and functional out- J. 2017;99-B:632–9. https://doi.org/10.1302/0301-­
come following unicompartmental knee replacement. 620X.99B5.BJJ-­2016-­0695.R2.
Osteoarthr Cartil. 2014;22:1241–50. https://doi. 65. Cook C, Mabry L, Reiman MP, Hegedus EJ. Best
org/10.1016/j.joca.2014.07.006. tests/clinical findings for screening and diagno-
55. Goh GS-H, Bin Abd Razak HR, Tay DK-J, et al. sis of patellofemoral pain syndrome: a systematic
Unicompartmental knee arthroplasty achieves greater review. Physiotherapy. 2012;98:93–100. https://doi.
flexion with no difference in functional outcome, qual- org/10.1016/j.physio.2011.09.001.
ity of life, and satisfaction vs total knee arthroplasty 66. Kang SN, Smith TO, Sprenger De Rover WB,
in patients younger than 55 years. A propensity score-­ Walton NP. Pre-operative patellofemoral degen-
matched cohort analysis. J Arthroplast. 2018;33:355– erative changes do not affect the outcome after
61. https://doi.org/10.1016/j.arth.2017.09.022. medial Oxford unicompartmental knee replace-
56. Engh GA. Orthopaedic crossfire—can we justify uni- ment: a report from an independent centre. J
condylar arthroplasty as a temporizing procedure? In Bone Joint Surg Br. 2011;93:476–8. https://doi.
the affirmative. J Arthroplast. 2002;17:54–5. https:// org/10.1302/0301-­620X.93B4.25562.
doi.org/10.1054/arth.2002.32448. 67. Lim JW-A, Chen JY, Chong HC, et al. Pre-existing
57. Ali AM, Pandit H, Liddle AD, et al. Does activity patellofemoral disease does not affect 10-year
affect the outcome of the Oxford unicompartmental survivorship in fixed bearing unicompartmen-
knee replacement? Knee. 2016;23:327–30. https:// tal knee arthroplasty. Knee Surg Sports Traumatol
doi.org/10.1016/j.knee.2015.08.001. Arthrosc. 2019;27:2030–6. https://doi.org/10.1007/
58. Crawford DA, Adams JB, Lombardi AV, Berend s00167-­018-­5169-­y.
KR. Activity level does not affect survivorship of 68. Deckard ER, Jansen K, Ziemba-Davis M, et al.
unicondylar knee arthroplasty at 5-year minimum Does patellofemoral disease affect outcomes in con-
5 The Modern Indications for Medial UKA the “Oxford Philosophy” Deciphered 67

temporary medial fixed-bearing unicompartmental of a series of 79 Lotus prostheses with a follow-up

knee arthroplasty? J Arthroplast. 2020. https://doi. of more than 5 years]. Rev Chir Orthop Reparatrice
org/10.1016/j.arth.2020.03.007. Appar Mot. 1987;73:544–51.
69. Berend KR, Lombardi AV, Morris MJ, et al. 80. Suggs JF, Li G, Park SE, et al. Knee biomechanics
Does preoperative patellofemoral joint state after UKA and its relation to the ACL—a robotic
affect medial unicompartmental arthroplasty sur- investigation. J Orthop Res. 2006;24:588–94. https://
vival? Orthopedics. 2011;34:e494–6. https://doi. doi.org/10.1002/jor.20082.
org/10.3928/01477447-­20110714-­39. 81. Blunn GW, Walker PS, Joshi A, Hardinge K. The
70. Munk S, Odgaard A, Madsen F, et al. Preoperative dominance of cyclic sliding in producing wear in
lateral subluxation of the patella is a predictor of total knee replacements. Clin Orthop Relat Res.
poor early outcome of Oxford phase-III medial 1991;273:253–60.
unicompartmental knee arthroplasty. Acta Orthop. 82. Adulkasem N, Rojanasthien S, Siripocaratana N,
2011;82:582–8. https://doi.org/10.3109/17453674.20 Limmahakhun S. Posterior tibial slope modification
11.618915. in osteoarthritis knees with different ACL conditions:
71. Burger JA, Kleeblad LJ, Laas N, Pearle AD. The cadaveric study of fixed-bearing UKA. J Orthop Surg
influence of preoperative radiographic patello- (Hong Kong). 2019;27:2309499019836286. https://
femoral degenerative changes and malalignment on doi.org/10.1177/2309499019836286.
patellofemoral-specific outcome scores following 83. Zumbrunn T, Schütz P, von Knoch F, et al. Medial
fixed-bearing medial unicompartmental knee arthro- unicompartmental knee arthroplasty in ACL-deficient
plasty. J Bone Joint Surg Am. 2019;101:1662–9. knees is a viable treatment option: in vivo kinematic
https://doi.org/10.2106/JBJS.18.01385. evaluation using a moving fluoroscope. Knee Surg
72. Beard DJ, Pandit H, Gill HS, et al. The influence Sports Traumatol Arthrosc. 2020;28:1765–73. https://
of the presence and severity of pre-existing patello- doi.org/10.1007/s00167-­019-­05594-­0.
femoral degenerative changes on the outcome of the 84. Hernigou P, Deschamps G. Posterior slope of the
Oxford medial unicompartmental knee replacement. tibial implant and the outcome of unicompart-
J Bone Joint Surg Br. 2007;89:1597–601. https://doi. mental knee arthroplasty. J Bone Joint Surg Am.
org/10.1302/0301-­620X.89B12.19259. 2004;86-A:506–11.
73. Adams AJ, Kazarian GS, Lonner JH. Preoperative 85. Suero EM, Citak M, Cross MB, et al. Effects of tibial
patellofemoral chondromalacia is not a contraindica- slope changes in the stability of fixed bearing medial
tion for fixed-bearing medial unicompartmental knee unicompartmental arthroplasty in anterior cruci-
arthroplasty. J Arthroplast. 2017;32:1786–91. https:// ate ligament deficient knees. Knee. 2012;19:365–9.
doi.org/10.1016/j.arth.2017.01.002. https://doi.org/10.1016/j.knee.2011.07.004.
74. Berger Y, Ftaita S, Thienpont E. Does medial patel- 86. Engh GA, Ammeen DJ. Unicondylar arthroplasty
lofemoral osteoarthritis influence outcome scores in knees with deficient anterior cruciate ligaments.
and risk of revision after fixed-bearing unicom- Clin Orthop Relat Res. 2014;472:73–7. https://doi.
partmental knee arthroplasty? Clin Orthop Relat org/10.1007/s11999-­013-­2982-­y.
Res. 2019;477:2041–7. https://doi.org/10.1097/ 87. Boissonneault A, Pandit H, Pegg E, et al. No dif-
CORR.0000000000000738. ference in survivorship after unicompartmental
75. Konan S, Haddad FS. Does location of patellofemo- knee arthroplasty with or without an intact anterior
ral chondral lesion influence outcome after Oxford cruciate ligament. Knee Surg Sports Traumatol
medial compartmental knee arthroplasty? Bone Joint Arthrosc. 2013;21:2480–6. https://doi.org/10.1007/
J. 2016;98-B:11–5. https://doi.org/10.1302/0301-­ s00167-­012-­2101-­8.
620X.98B10.BJJ-­2016-­0403.R1. 88. Pandit H, Beard DJ, Jenkins C, et al. Combined
76. Kumar V, Pandit HG, Liddle AD, et al. Comparison anterior cruciate reconstruction and Oxford
of outcomes after UKA in patients with and without unicompartmental knee arthroplasty. J Bone
chondrocalcinosis: a matched cohort study. Knee Joint Surg Br. 2006;88:887–92. https://doi.
Surg Sports Traumatol Arthrosc. 2017;25:319–24. org/10.1302/0301-­620X.88B7.17847.
https://doi.org/10.1007/s00167-­015-­3578-­8. 89. Tinius M, Hepp P, Becker R. Combined unicom-
77. Hernigou P, Pascale W, Pascale V, et al. Does primary partmental knee arthroplasty and anterior cruciate
or secondary chondrocalcinosis influence long-term ligament reconstruction. Knee Surg Sports Traumatol
survivorship of unicompartmental arthroplasty? Clin Arthrosc. 2012;20:81–7. https://doi.org/10.1007/
Orthop Relat Res. 2012;470:1973–9. https://doi. s00167-­011-­1528-­7.
org/10.1007/s11999-­011-­2211-­5. 90. Volpin A, Kini SG, Meuffels DE. Satisfactory out-
78. Goodfellow J, O’Connor J. The anterior cruciate liga- comes following combined unicompartmental knee
ment in knee arthroplasty. A risk-factor with uncon- replacement and anterior cruciate ligament recon-
strained meniscal prostheses. Clin Orthop Relat Res. struction. Knee Surg Sports Traumatol Arthrosc.
1992;276:245–52. 2018;26:2594–601. https://doi.org/10.1007/
79. Deschamps G, Lapeyre B. [Rupture of the anterior s00167-­017-­4536-­4.
cruciate ligament: a frequently unrecognized cause of 91. Tecame A, Savica R, Rosa MA, Adravanti P. Anterior
failure of unicompartmental knee prostheses. Apropos cruciate ligament reconstruction in association with
68 T. Gicquel et al.

medial unicompartmental knee replacement: a ret- 94. Iriberri I, Suau S, Payán L, Aragón JF. Long-term
rospective study comparing clinical and radiologi- deterioration after one-stage unicompartmental knee
cal outcomes of two different implant design. Int arthroplasty and anterior cruciate ligament recon-
Orthop. 2019;43:2731–7. https://doi.org/10.1007/ struction. Musculoskelet Surg. 2019;103:251–6.
s00264-­019-­04341-­x. https://doi.org/10.1007/s12306-­018-­0582-­4.
92. Ventura A, Legnani C, Terzaghi C, et al. 95. Castoldi M, Magnussen RA, Gunst S, et al. A
Unicompartmental knee replacement combined randomized controlled trial of bone-patellar
to anterior cruciate ligament reconstruction: mid- tendon-bone anterior cruciate ligament recon-
term results. J Knee Surg. 2019. https://doi. struction with and without lateral extra-articular
org/10.1055/s-­0039-­1692647. tenodesis: 19-year clinical and radiological follow-
93. Tian S, Wang B, Wang Y, et al. Combined unicom- ­up. Am J Sports Med. 2020;48:1665–72. https://doi.
partmental knee arthroplasty and anterior cruciate org/10.1177/0363546520914936.
ligament reconstruction in knees with osteoarthri- 96. Saithna A, Thaunat M, Delaloye JR, et al. Combined
tis and deficient anterior cruciate ligament. BMC ACL and anterolateral ligament reconstruction. JBJS
Musculoskelet Disord. 2016;17:327. https://doi. Essent Surg Tech. 2018;8:e2. https://doi.org/10.2106/
org/10.1186/s12891-­016-­1186-­5. JBJS.ST.17.00045.
 How to Deal with a Fixed-Bearing
Medial Unicompartmental Knee 6
Arthroplasty Implant?

Camille Steltzlen and Nicolas Pujol

The concept of unicompartmental knee arthro- arthroplasties [8]. The choice of a full-­
plasty was proposed by McKeever and MacIntosh polyethylene tibial implant only or metal-backed
in the USA in the 1950s [1]. It was then devel- tibial implant remains debated; results are diver-
oped by Marmor [2], who introduced the concept gent. Nevertheless, it seems that the latest clinical
of modular resurfacing arthroplasty. In France, results support metal-backed tibial implants [7, 9,
Philippe Cartier, Philippe Hernigou, and Gérard 10]. In contrast, studies on the fixation of
Deschamps were the first to popularise this pro- cemented or cementless implants have not shown
cedure. At the SOFCOT symposium in 1996, any significant difference [11, 12]. Based on
they established solid foundations for its indica- Gérard Deschamps’s arguments, we also prefer
tions and surgical technique [3, 4]. an implant based on cuts rather than a resurfacing
The surgical indication and preoperative plan- one. Since wear is mainly tibial, the use of resur-
ning are key factors to the procedure’s success. A facing arthroplasty may risk lowering the articu-
precise, reliable, and reproducible surgical tech- lation by an increase in femoral displacement
nique will make it possible to obtain a satisfac- [13].
tory functional result, as well as prolonged This chapter is divided into two parts. In the
lifespan of the implant. In order to achieve pre- first part, we will detail the preoperative planning
cise and reproducible placement quality, it based on analysis of the radiological assessment
appears that an orthopaedic surgeon would need and in the second part, we will discuss the surgi-
to perform about 40 knee arthroplasties per year cal technique.
[5–7]. An orthopaedic knee surgeon should know
how to establish the indication and perform uni-
compartmental arthroplasty correctly; to do so, s/ 6.1 Preoperative Planning
he will need to follow a learning curve and
acquire experience. In this chapter, we will The objective is to reproduce the articulation’s
describe the principles for placement of metal-­ orientation in both the frontal and sagittal planes.
backed fixed-bearing unicompartmental knee Changes have been made in the last few years
with the adoption of the concept of anatomical
implants whose technique makes it possible to
C. Steltzlen (*) · N. Pujol reproduce the initial knee deformity prior to wear
Service de Chirurgie Orthopédique et progression. Therefore, it must be tailored to
Traumatologique, Centre Hospitalier de Versailles, each patient’s anatomy.
Le Chesnay, France
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2024 69

A. Clavé, F. Dubrana (eds.), Unicompartmental Knee Arthroplasty,
https://doi.org/10.1007/978-3-031-48332-5_6
70 C. Steltzlen and N. Pujol

6.1.1 Tibial and Femoral Bone Cuts

The different bone cuts, particularly tibial cut,

must have been planned based on the imaging
assessment. The assessment should consist of an
AP, sagittal views and weightbearing X-rays. A
skyline view (at 30° flexion) is also necessary,
even though the presence of osteophyte is not a
contraindication to UKA if the joint is asymp-
tomatic [14, 15]. Full-length lower limb X-ray is
the key investigation in preoperative planning. It
will assess the overall deformity of the lower
limb. The objective is to determine the origin of
the deformity, constitutional or acquired, in order
to correct wear only. Some authors recommend
dynamic X-rays to verify reducibility of the
deformity. We do not use them.

6.1.1.1 Coronal Plane

One factor often emphasised is the notion of
postoperative undercorrection. In our opinion,
restoration of the anatomy is more relevant. The Fig. 6.1 Angle of the tibial cut, which in this example is
goal is not to take a fixed value as the postopera- at 88°
tive objective, but to restore the initial deformity
by correcting wear only [16–19]. Nevertheless, a lowed during the conduct of femoral distal and
postoperative deformity should remain limited. It posterior cuts. They should be parallel to the tib-
may be located between 7° and 10° of the overall ial cut to avoid positioning on the femoral
deformity [17]. The persistence of a major post- implant’s edges in flexion and extension.
operative deformity can carry a risk of increasing
the rate of early implant failure [20–22]. In order 6.1.1.2 Sagittal Plane
to assess the resection height, it is first necessary The tibial slope should also be assessed preoper-
to determine an objective for the final deformity. atively and restored at the end of the procedure
For example, if the initial deformity is 9° varus (Figs. 6.3 and 6.4). A decrease would have the
and the targeted final deformity is 3° varus, the effect of closing the space for flexion, limiting
axis would need to be corrected by 6°. In arthro- flexion movements, and increasing stress on the
plasties with dependent cuts, the axis is corrected posterior part of the tibial plateau with an
by a single tibial cut. In order to correct the axis increased risk of implant loosening. An excessive
by 6° using a tibial implant with a minimal thick- increase would multiply the tension on the ante-
ness of 8 mm, it is necessary to resect 2 mm of rior cruciate ligament at the risk of rupture and
bone on the tibial side. After assessing the resec- decreasing the space in extension, and therefore
tion height, it is necessary to assess the angle of limiting extension movements [24]. Then
inclination in the tibial section plane (Figs. 6.1 between 4° and 8° of postoperative posterior
and 6.2) [23]. This incline should also be fol- slope is recommended [25].
6 How to Deal with a Fixed-Bearing Medial Unicompartmental Knee Arthroplasty Implant? 71

Fig. 6.2 Postoperative restoration of the tibial coronal

plane inclination

Fig. 6.3 Preoperative assessment of the tibial slope

72 C. Steltzlen and N. Pujol

Fig. 6.5 Physica ZUK Lima Corporate® insert (courtesy

of LIMA)

Fig. 6.4 Postoperative restoration of the tibial slope

6.2 Surgical Technique

In this section, we will describe a surgical tech-

nique for dependent cuts arthroplasty derived
from the Miller–Galante arthroplasty technique
(Fig. 6.5).

6.2.1 Approach

A medial parapatellar approach is used. The inci-

sion starts at the upper border of the patella and
extends up to the ATT (anterior tibial tuberosity)
(Fig. 6.6). It is performed through the medial
patellar flange. It can sometimes be necessary to
increase the size of the approach by going through
fibres of the vastus medialis over about 1 cm in
order to dislocate the patella more easily. In a
second phase, it is necessary to expose the medial
tibial plateau without releasing the medial collat- Fig. 6.6 Parapatellar approach
6 How to Deal with a Fixed-Bearing Medial Unicompartmental Knee Arthroplasty Implant? 73

eral ligament, which could result in a tendency to

overcorrect the deformity by increasing the poly-
ethylene in response to the ligament laxity
induced by its release. For that, a double-bend
retractor is inserted just under the collar of the
tibial osteophytes, without prior release, and this
is amply sufficient to expose it. Osteophytes in
the tibia and femur are removed. This stage is
important to facilitate correction of the acquired
deformity and enable good implant positioning in
the frontal plane. The femoral implant should be
positioned close to the intercondylar notch so
that it is opposite the tibial implant. Excision of
osteophytes from the intercondylar notch may
also correct a slight preoperative flexion
deformity.

6.2.2 Tibial Resection

Cuts are dependent.

First, tibial resection is performed. It is neces-
sary to calculate preoperatively the targeted final
postoperative deformity. Thus, we should be able
to determine the height of tibia cut, which alone
prefigures the amount of correction when using
dependent cuts implants.

6.2.2.1 In the Coronal Plane Fig. 6.7 Extramedullary viewfinder aligned on the tibial
The surgeon starts by placing the extramedullary crest
cutting guide with the knee in 90° flexion. At this
stage, it is necessary that the resection plane ological slope. After assessing the level of
reproduces the physiological tibial coronal angle resection, a pin is introduced into the section
(Fig. 6.2). To reproduce this angle in moderate guide to verify its sagittal axis and check that it
deformities, the surgeon can align the ancillary follows the native tibial slope (Fig. 6.9).
material with the tibial crest by modulating Vertical cut is then performed, first using a sag-
adjustment at the level of the malleolar clamp ittal saw. It should leave the tibial spine intact,
(Fig. 6.7). A probe is then introduced into the sec- particularly insertion of the anterior cruciate
tion guide and the bottom of the bone cup, which ligament. The axis of this section should fol-
will enable bone resection of the height planned low the axis of the lateral wall of the medial
preoperatively (Fig. 6.8). condyle.
The tibial plateau is removed with the knee in
6.2.2.2 In the Sagittal Plane slight flexion. The quantity removed should not
Assessment of the tibial slope is an important be greater posteriorly. It reflects the future tibial
stage. The objective is to reproduce the physi- slope of the implant.
74 C. Steltzlen and N. Pujol

Fig. 6.9 Control of the cut height and tibial slope

Fig. 6.8 Stylus used to evaluate the cut level

6.2.3 Femoral Cuts

6.2.3.1 Distal Femoral Cut

In the frontal plane, both the distal femoral and
the tibial cut are dependent on each other. It is
performed with the knee in complete extension to
avoid creating a flexion or recurvatum deformity
in the cut (Fig. 6.10). Its height is defined by the
instrumentation; it corresponds to the thickness
of the femoral implant. The distal femoral cutting
guide is fixed on an adjustable height spacer
(Fig. 6.11). It starts at 8 mm, which is the mini-
mum implant thickness. At this stage, it is possi-
ble to assess the final deformity and residual
ligament laxity. If the space is deemed too tight,
it is necessary to cut the tibia again. If excessive
laxity is found, the thickness can be increased
using a 1-mm increments spacer. In cases of
major residual laxity with excess wear on the Fig. 6.10 Distal femoral cut with the knee in extension
6 How to Deal with a Fixed-Bearing Medial Unicompartmental Knee Arthroplasty Implant? 75

6.2.3.2 Posterior Femoral Cut

After removing the distal femoral cut, the size of
the femoral implant is assessed. The implant
should not overflow the anterior border of the dis-
tal cut (Fig. 6.12). After determining its size, the
section guide is placed with the knee in 90° flex-
ion. The implant should be positioned in the cor-
onal plane near the (intercondylar) notch so that
the femoral component will be in the middle of
the tibial component (Fig. 6.13). The posterior
cut should be parallel to the tibial cut to avoid
edge loading of the femoral component in
flexion.

6.2.4 Finalisation of the Tibia

and Testing

After evaluating the size of the tibia, anchoring

points are made in the tibial bone.
We recommend verifying the anteroposterior
position of the tibial implant. It should be posi-
tioned on the anterior cortex of the tibia to avoid
subsequent sinking of the implant. Such tibial
sinking is possible, particularly in cases of
osteoporotic bone, and this generally occurs in
its anterior part. Trial implants are then inserted
into the tibia and femur. It is necessary to place
the tibial insert on the tibial baseplate at a height
defined with the tibial spacer used during the
distal femoral cut. This height starts at 8 mm
and increases by 1-mm increments. After insert-
ing the implants, the existence of physiological
laxity must be verified. In fact, there has not
been any ligament release in this approach and
the implant only corrects wear (which is the
purpose of this type of implant and technique);
Fig. 6.11 Distal femoral cut spacer (courtesy of LIMA) there is no residual laxity and the knee has a
near-normal ligament kinematic presentation.
The positioning of the implants is also checked.
femur, it is possible to use 1- or 2-mm femoral The femoral implant should be parallel to the
blocks. Using these blocks will have the effect of tibial implant in extension and flexion and
decreasing femoral cut to distalize the femoral should be in the middle of the tibia (Figs. 6.14
implant without raising the joint line, which and 6.15).
would occur if the tibial spacer was increased.
76 C. Steltzlen and N. Pujol

Fig. 6.12 Control of the size of

femoral implant (courtesy of
LIMA)

Fig. 6.13 Positioning close to the notch in the femoral

cut guide Fig. 6.14 Trial in extension
6 How to Deal with a Fixed-Bearing Medial Unicompartmental Knee Arthroplasty Implant? 77

3. Cartier P, Cheaib S. Unicondylar knee arthroplasty.

2–10 years of follow-up evaluation. J Arthroplast.
1987;2(2):157–62.
4. Hernigou P, Deschamps G. Les prothéses unicom-
partimentales du genou. Symposium 70° Reunion
annuelle de la SOFCOT. Rev Chir Orthop Reparatrice
Appar Mot. 1996;82(1 Suppl):23–60.
5. Badawy M, Fenstad AM, Bartz-Johannessen CA,
Indrekvam K, Havelin LI, Robertsson O, et al.
Hospital volume and the risk of revision in Oxford
unicompartmental knee arthroplasty in the Nordic
countries –an observational study of 14,496 cases.
BMC Musculoskelet Disord. 2017;18(1):388.
6. Badawy M, Espehaug B, Indrekvam K, Havelin LI,
Furnes O. Higher revision risk for unicompartmen-
tal knee arthroplasty in low-volume hospitals. Acta
Orthop. 2014;85(4):342–7.
7. Zambianchi F, Digennaro V, Giorgini A, Grandi G,
Fiacchi F, Mugnai R, et al. Surgeon’s experience
influences UKA survivorship: a comparative study
between all-poly and metal back designs. Knee Surg
Sports Traumatol Arthrosc. 2015;23(7):2074–80.
8. Baur J, Zwicky L, Hirschmann MT, Ilchmann T,
Clauss M. Metal backed fixed-bearing unicondylar
knee arthroplasties using minimal invasive surgery:
a promising outcome analysis of 132 cases. BMC
Musculoskelet Disord. 2015;16(1):177.
Fig. 6.15 Trial in flexion
9. Hutt JRB, Farhadnia P, Massé V, LaVigne M,
Vendittoli P-A. A randomised trial of all-­polyethylene
and metal-backed tibial components in unicom-
partmental arthroplasty of the knee. Bone Joint J.
6.3 Conclusion 2015;97-B(6):786–92.
10. Koh IJ, Suhl KH, Kim MW, Kim MS, Choi KY, In
A good surgical indication, precise preoperative Y. Use of all-polyethylene tibial components in uni-
compartmental knee arthroplasty increases the risk of
planning, and reliable placement technique are early failure. J Knee Surg. 2017;30(8):807–15.
the three essential factors to obtain good repro- 11. Kerens B, Schotanus MGM, Boonen B, Boog P, Emans
ducible results in the medium and long terms PJ, Lacroix H, et al. Cementless versus cemented
[26–30]. Twenty percent of our indications for Oxford unicompartmental knee arthroplasty: early
results of a non-designer user group. Knee Surg
knee replacements are unicompartmental knee Sports Traumatol Arthrosc. 2017;25(3):703–9.
arthroplasties. It is also possible to perform this 12. Campi S, Pandit HG, Dodd C, Murray
procedure at the same time on both knees with DW. Cementless fixation in medial unicompartmen-
morbidity equivalent to total knee arthroplasty in tal knee arthroplasty: a systematic review. Knee Surg
Sports Traumatol Arthrosc. 2017;25(3):736–45.
a single knee [31]. 13. Deschamps G, Chol C. Fixed-bearing unicompart-
mental knee arthroplasty. Patients’ selection and
operative technique. Orthop Traumatol Surg Res.
References 2011;97(6):648–61.
14. Adams AJ, Kazarian GS, Lonner JH. Preoperative
patellofemoral chondromalacia is not a contraindica-
1. Johal S, Nakano N, Baxter M, Hujazi I, Pandit H, tion for fixed-bearing medial unicompartmental knee
Khanduja V. Unicompartmental knee arthroplasty: the arthroplasty. J Arthroplast. 2017;32(6):1786–91.
past, current controversies, and future perspectives. J 15. Lim JW-A, Chen JY, Chong HC, Pang HN, Tay DKJ,
Knee Surg. 2018;31(10):992–8. Chia S-L, et al. Pre-existing patellofemoral disease
2. Marmor L. Marmor modular knee in unicompartmen- does not affect 10-year survivorship in fixed bear-
tal disease. Minimum four-year follow-up. J Bone ing unicompartmental knee arthroplasty. Knee Surg
Joint Surg Am. 1979;61(3):347–53. Sports Traumatol Arthrosc. 2019;27(6):2030–6.
78 C. Steltzlen and N. Pujol

16. Zuiderbaan HA, van der List JP, Chawla H, Khamaisy 25. Weber P, Schröder C, Schwiesau J, Utzschneider S,
S, Thein R, Pearle AD. Predictors of subjective out- Steinbrück A, Pietschmann MF, et al. Increase in the
come after medial unicompartmental knee arthro- tibial slope reduces wear after medial unicompart-
plasty. J Arthroplast. 2016;31(7):1453–8. mental fixed-bearing arthroplasty of the knee. Biomed
17. Hernigou P, Deschamps G. Alignment influences Res Int. 2015;2015:736826.
wear in the knee after medial unicompartmental 26. Biswal S, Brighton RW. Results of unicompartmental
arthroplasty. Clin Orthop. 2004;423:161–5. knee arthroplasty with cemented, fixed-bearing pros-
18. Kim KT, Lee S, Kim TW, Lee JS, Boo KH. The influ- thesis using minimally invasive surgery. J Arthroplast.
ence of postoperative tibiofemoral alignment on the 2010;25(5):721–7.
clinical results of unicompartmental knee arthro- 27. Winnock de Grave P, Barbier J, Luyckx T, Ryckaert
plasty. Knee Surg Relat Res. 2012;24(2):85–90. A, Gunst P, Van den Daelen L. Outcomes of a fixed-­
19. Perkins TR, Gunckle W. Unicompartmental knee bearing, medial, cemented unicondylar knee arthro-
arthroplasty: 3- to 10-year results in a community plasty design: survival analysis and functional score
hospital setting. J Arthroplast. 2002;17(3):293–7. of 460 cases. J Arthroplasty. 2018;33(9):2792–9.
20. Vasso M, Del Regno C, D’Amelio A, Viggiano D, 28. Argenson J-NA, Blanc G, Aubaniac J-M, Parratte
Corona K, Schiavone PA. Minor varus alignment pro- S. Modern unicompartmental knee arthroplasty with
vides better results than neutral alignment in medial cement: a concise follow-up, at a mean of twenty
UKA. Knee. 2015;22(2):117–21. years, of a previous report. J Bone Joint Surg Am.
21. Gulati A, Pandit H, Jenkins C, Chau R, Dodd C, 2013;95(10):905–9.
Murray DW. The effect of leg alignment on the 29. Parratte S, Ollivier M, Lunebourg A, Abdel MP,
­outcome of unicompartmental knee replacement. J Argenson J-N. Long-term results of compartmental
Bone Joint Surg Br. 2009;91(4):469–74. arthroplasties of the knee: long term results of partial
22. Gulati A, Chau R, Simpson DJ, Dodd C, Gill HS, knee arthroplasty. Bone Joint J. 2015;97-B(10 Suppl
Murray DW. Influence of component alignment on A):9–15.
outcome for unicompartmental knee replacement. 30. Kim KT, Lee S, Lee JS, Kang MS, Koo KH. Long-­
Knee. 2009;16(3):196–9. term clinical results of unicompartmental knee
23. Asada S, Inoue S, Tsukamoto I, Mori S, Akagi arthroplasty in patients younger than 60 years of age:
M. Obliquity of tibial component after unicompart- minimum 10-year follow-up. Knee Surg Relat Res.
mental knee arthroplasty. Knee. 2019;26(2):410–5. 2018;30(1):28–33.
24. Takayama K, Matsumoto T, Muratsu H, Ishida K, 31. Siedlecki C, Beaufils P, Lemaire B, Pujol
Araki D, Matsushita T, et al. The influence of pos- N. Complications and cost of single-stage vs. two-­
terior tibial slope changes on joint gap and range of stage bilateral unicompartmental knee arthroplasty:
motion in unicompartmental knee arthroplasty. Knee. a case-control study. Orthop Traumatol Surg Res.
2016;23(3):517–22. 2018;104(7):949–53.
 Principles of the Oxford® (Zimmer
Biomet) Unicompartmental Knee 7
Arthroplasty (OUKA)

François Hardeman and Arnaud Clavé

7.1 Introduction The Oxford® knee arthroplasty is unique in

design and philosophy. It enables rolling/sliding
This chapter describes the principles relating to biomechanics via its mobile polyethylene (PE)
placement of the Oxford® (Zimmer Biomet) uni- insert, which has a dual articulation: its flat sur-
compartmental knee replacement. The technique face articulates with the flat tibial implant (metal
and philosophy described here are based on the back) on the one side and its concave upper part
instrumentation and fourth-generation implants articulates fully congruently with the spherical
known as Microplasty®. femoral component (Fig. 7.1) on the other. When

a b

Fig. 7.1 Oxford® unicompartmental knee arthroplasty with mobile bearing (a) 3/ view of an Oxford® unicompart-
mental knee in a bone model (b) frontal view of an Oxford® unicompartmental knee in a bone model

F. Hardeman
Department of Orthopaedic Surgery and
Traumatology, Jan Ypermanziekenhuis,
Ypres, Belgium
A. Clavé (*)
Department of Orthopaedic Surgery and
Traumatology, Saint George Private Hospital,
Nice, France
LaTIM, INSERM-UBO UMR 1101, Brest, France

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2024 79

A. Clavé, F. Dubrana (eds.), Unicompartmental Knee Arthroplasty,
https://doi.org/10.1007/978-3-031-48332-5_7
80 F. Hardeman and A. Clavé

the components are positioned correctly, there- unicompartmental replacement is completely dif-
fore, this concept limits wear of the polyethylene ferent because it essentially involves a procedure
mobile insert. designed to restore wear, which is based on bal-
The implant was designed and developed fol- ancing the ligaments in order to obtain the correct
lowing a meeting with Prof. John O’Connor position. Therefore, it should not be considered
(Engineer) and John Goodfellow (Orthopaedic as a bone procedure, but rather as a soft tissue
Surgeon) in Oxford in 1966. procedure. Its purpose is to restore normal liga-
The first knee replacements were performed ment tension. The technique is based on tension
in 1976, but as bi/unicompartmental (medial/lat- at rest on the medial collateral ligament (MCL).
eral) procedures in indications for three-­ Since the MCL is isometric in the entire range of
compartment knee OA as an alternative to knee joint amplitude, it constitutes a very reliable
TKR. The first Oxford knee replacement done as guide to reconstructing the joint space height in
a medial UKA was performed by John the medial compartment [1].
Goodfellow in 1982! The indication for Oxford knee replacement is
In its latest version (implant and instrumenta- based on anterointernal knee OA (osteoarthritis).
tion material) dating from 2011, called This is characterised by the presence of a func-
Microplasty®, the Oxford knee replacement tionally intact ACL (anterior cruciate ligament)
comes in cemented and cementless versions, and the existence of conserved cartilage thick-
each including five sizes of femoral components ness at the posterior part of the medial tibial pla-
and seven sizes of anatomical tibial components. teau and in the posterior medial femoral condyle.
The mobile PE has thicknesses ranging from 3 to Therefore, when the knee is placed in flexion, the
9 mm, and there is a range based on the size of MCL is taut, avoiding with the passage of time
the femoral component. retraction of the periarticular soft tissue including
Worldwide, the most common surgical treat- the MCL and posterior capsule (Fig. 7.2). Tension
ment for knee OA (independently of degenera- on the MCL will make it possible to determine
tion type) is total knee arthroplasty, which often the position of the components for the purpose of
requires so-called independent bone resections recreating physiological native alignment, mobil-
and ligament releases. The philosophy of Oxford ity, and stability.
7 Principles of the Oxford® (Zimmer Biomet) Unicompartmental Knee Arthroplasty (OUKA) 81

Fig. 7.2 Anteromedial

osteoarthritis (a and b)
a b
Slack MCL due to
cartilage wear and
AMOA; (c and d)
Tension of the MCL in
flexion due to the
physiological femoral
roll-back thanks to a
preserved ACL. The
joint line level is also
preserved as there is no
wear at the back of the
tibia and on the posterior
part of the femoral
condyle; (e and f)
Correction of varus
deformity caused by the
wear. Premorbid state
obtained due to the
physiological behaviour Flexion deformity Varus deformity
of the MCL and ACL c d

e f
82 F. Hardeman and A. Clavé

7.2 Role of the Collateral physiological tension of the MCL and cruciate
Ligaments ligaments, which are normal and efficient
(Fig. 7.3).
In a normal knee that has not undergone wear, Balancing the ligaments in independent TKR
biomechanical behaviour differs between the techniques with resection almost necessarily
medial and lateral compartments. The lateral involves releasing the medial (or lateral) struc-
compartment has a larger opening in flexion, evi- tures to recreate the rectangular spaces that are
dencing greater laxity of the LCL in flexion, balanced in flexion and extension. In the setting
while in the internal compartment, due to the iso- of Oxford UKA, ligament release should never
metric behaviour of the MCL, the 2-mm joint be performed. Integrity of the MCL is imperative
opening is constant throughout the arch of and makes it possible to conserve joint mobility,
movement. stability, and physiological hip-knee alignment
Permanent flexion, which among other things (HKA), as well as an optimal / physiological
is caused by retraction of the posterior capsule kinematic behaviour of the mobile polyethylene
on osteophytes, is a process frequently observed insert during the entire range of motion of the
in knee OA. In AMOA, this permanent flexion knee. Appropriate tension on the MCL is critical
generally does not exceed 15° so long as the throughout the procedure, and therefore it is nec-
ACL is functionally intact. Consequently, in a essary to be careful to protect it at all times and
knee suffering from anteromedial OA with more never release or damage it. The proper balance
than 20° flexion, the medial compartment can between flexion and extension is created by
open in such a way as to regain premorbid joint removing bone from the distal femur and not by
space height because this opening is dictated by performing ligament releases.
7 Principles of the Oxford® (Zimmer Biomet) Unicompartmental Knee Arthroplasty (OUKA) 83

a b

c d

Fig. 7.3 (a and b) Tension of the MCL in flexion due to femoral condyle; (c and d) Correction of varus deformity
the physiological femoral roll-back thanks to a preserved caused by the wear. Premorbid state obtained due to the
ACL. The joint line level is preserved as there is no wear physiological behaviour of the MCL and ACL
at the back of the tibia and on the posterior part of the
84 F. Hardeman and A. Clavé

7.3 Restoring the Joint Space In anteromedial knee OA, the cartilage
Height remains conserved in the posterior condyle.
Therefore, posterior femoral resection will aim to
In the medial condyle, the knee flexes and extends remove the same quantity of bone and cartilage
around a centre of rotation that coincides with the as will be replaced by the femoral component,
femoral insertion of the femoral MCL. As the reconstructing native joint offset in flexion
MCL has a constant length throughout the entire (Fig. 7.5). The thickness of the femoral compo-
joint amplitude, it is necessary to restore the nent varies by size, ranging from 5.5 mm for XS
­correct joint level height (Fig. 7.4). If the joint to 7.45 mm for XL components.
space height is changed, the resulting centre of In contrast, there is significant wear in the dis-
rotation will change, which can decrease MCL tal femur. This makes the distal femur unreliable
tension in midflexion and increase it in hyperflex- to restore the joint space in extension. In the
ion. Thus, causing pain or dislocation of the mobile Oxford concept, the joint space in extension is
polyethylene insert. Consequently, it is important reconstructed by copying the space in flexion to
to reconstruct the correct/native joint line. the space in extension. Therefore, the concept of

Fig. 7.4 Physiological

tension of the MCL
allows to restore the
joint line level and to
respect the centre of
rotation of the medial
femoral condyle

9 13

Radius of
Posterior femoral saw-cut implant
Anatomical joint level

Fig. 7.5 The posterior femoral cut aims to resect an amount of bone and cartilage equal to the thickness of the femoral
component. Thus it will reproduce the native joint offset in flexion
7 Principles of the Oxford® (Zimmer Biomet) Unicompartmental Knee Arthroplasty (OUKA) 85

a single radius implant used by Oxford and the 7.4 Execution of the Procedure
isometric characteristic of the MCL make this
recession logical and reliable. 7.4.1 Tibial Cut and Positioning
The height of tibial cut has little importance of the Tibial Implant
for restoration of the joint space. However, care
should be taken not to resect too much bone on The level of tibial cut height is ensured with the
the tibia in order to avoid damaging the distal aid of instrumentation parts called calibration
insertion of the MCL or placing the implant in a spoons and a G-clamp. The spoons are inserted
more fragile and smaller area of bone. Yet a mini- with the knee at about 90° flexion and fit the
mum quantity of bone must be removed in order femoral condyle (Fig. 7.6). They are available
to create a sufficient space to accommodate the in different sizes matching the sizes of the fem-
knee replacement components. The tibial oral components (from XS to XL). In most
metal-­
­ back insert has a constant thickness of cases, a medium size can be used. Once
3 mm irrespective of the implant size. The mini- inserted, the anterior part of the spoon must be
mum thickness of the polyethylene mobile insert located at the level of placement of the premor-
is 3 mm. In flexion, therefore, after performing bid cartilaginous surface of the distal condyle,
tibial resection and before performing femoral i.e. about 2 to 3 mm in front of the area of ebur-
posterior cut, the space should be at least 6 mm. nated bone in the distal femur. The size of the
If more bone has been resected on the tibia, this spoon in itself is unimportant because it will
can be corrected by increasing the thickness of not determine the choice of final implant.
the PE without theoretically affecting the joint However, it can provide relatively simply an
space height. approximation of the size of femoral compo-

a b

Fig. 7.6 G-Clamp and sizing spoon used to defined the stretch enough. (b) A thicker spoon will compensate more
height of the tibial cut. (a) The spoon has to be inserted in cartilage and bone loss. This will put more tension on the
the joint and then as a spacer will stretch the MCL. If you MCL. A thicker spoon will remove less bone on the tibial
can twist it (up to 90°) easily, it means that the MCL is not side
86 F. Hardeman and A. Clavé

nent to use. Nevertheless, an abacus exists patient, indicates the theoretical sizes of the
which, depending on the size and sex of the femoral and tibial component to use.

Female Male
Size (cm) Femur Tibia Size (cm) Femur Tibia
<153 X-small A or B <160 Small A, B or C
153–165 Small A, B or C 160–175 Medium C or D
165–175 Medium C or D 170–185 Large E or F
>175 Large E >185 X-large F

Spoons are available in three different thick- can be chosen in most cases. Therefore, by using
nesses: 1, 2, and 3 mm. The thickness of the a 4 G-clamp that aims for a 4-mm insert, this
spoon makes it possible to adjust the resection gives us the possibility of using a higher or lower
height. Using a 3-mm spoon will remove 1 mm insert thickness. It is only in short (and low-­
less in the tibia than a 2-mm spoon. In our experi- weight) patients that using in a 3-mm PE is speci-
ence, a spoon of 1 mm thickness is appropriate in fied. However, this has the disadvantage that it is
80% of cases. In cases of greater tibial bone loss impossible to choose a lower insert thickness if
and if excessive laxity of the MCL persists, how- necessary since the minimum thickness of poly-
ever, a 2- or 3-mm spoon can be indicated. If the ethylene inserts is 3 mm. The advantage in seek-
spoon can be pivoted to 90° when inserted ing to use a 3-mm insert in short patients is to
between the tibia and femur, this indicates a reduce the height of the tibial cut and therefore
defect in MCL tension; therefore, the joint space limit bone resection. Indeed, it is known that the
is not restored/refilled and it is necessary to repeat greater the tibial resection, the smaller the
the test with a thicker spoon (Fig. 7.6). weightbearing surface and the weaker the bone,
A G-clamp is a part that attaches between the increasing the risk of stress shielding and fracture
spoon and extramedullary guide/rod, making it of the tibial plateau.
possible to adjust its positioning height (Fig. 7.7). Physiological values of the tibial slope are tra-
It is available in two sizes: three or four depend- ditionally between 0° and 15°. The Oxford team
ing on the thickness of the preferred mobile does not recommend trying to reproduce the
insert. Similarly, in our experience, a G-clamp 4 patient’s natural physiological slope, with the

a b

3 4

Fig. 7.7 G-Clamp and sizing spoon (a) A 3 G-Clamp ally leads to a 7 mm height cut for a 4 mm mobile PE
usually leads to a 6 mm height cut for a 3 mm mobile PE (3 mm metal back and 4 mm PE)
(3 mm metal back and 3 mm PE); (b) A 4 G-Clamp usu-
7 Principles of the Oxford® (Zimmer Biomet) Unicompartmental Knee Arthroplasty (OUKA) 87

objective for them and in all cases being a tibial

slope of 7°. All studies reporting good results and
excellent survival have been conducted in a large
range of patients for whom the tibial slope had
been defined as 7° posterior [2].
The tibial extramedullary guide/rod makes it
possible to perform tibial sectioning perpendicu-
lar to the anatomical axis of the tibia and without
considering the orientation of the joint space. In
Oxford UKA and contrary to TKR, postoperative
alignment of the lower limb is not determined by
positioning the metal-back tibial implant in
varus/valgus. Besides this, the concept of a com-
pletely congruent mobile insert makes it possible
to tolerate a certain difference in angulation
between the tibial plateau and femoral compo-
nent, without loss of congruence. A 5° error in
positioning the tibial implant in varus or valgus is Fig. 7.8 Top view of the tibial biscuit
considered acceptable [3].
The sagittal tibial cut is done flush with the
tibial insertion of the ACL and aims to include a mobile insert against the vertical wall of the tibial
small part of the medial tibial spine. Certain component or between a metal-back medial over-
median fibres of the tibial ACL insertion must lap and the MCL.
occasionally be removed for better exposure of When the tibial biscuit is removed, it can be
the intercondylar notch and the direction of the inspected to confirm the diagnosis of anterome-
sagittal resection. Traditionally, the reciprocating dial knee OA. Standard wear shows complete
saw is placed just on the inside of the apex of the loss of cartilage in the anteromedial part, and
medial tibial spine and is aimed at the homolat- normal cartilage thickness remains in the poste-
eral anterosuperior iliac spine. The sagittal direc- rior part of the tibial slice (Fig. 7.8). At this stage,
tion of the cut is in the flexion/extension axis of the flexion space can be determined. Therefore, if
the tibia. This plane can be determined by mov- a 4 G-clamp has been used, the space obtained in
ing the tibia into flexion and extension. Another flexion should be 7 mm, corresponding to a 3-mm
aiming point is the ipsilateral anterosuperior iliac metal-back thickness (invariable irrespective of
spine. It is important to avoid excessive external the size of tibial implant) plus the 4 mm sought
rotation of the cut: posteriorly the cut must be for the mobile insert. From that point, the space
close to the insertion of the posterior cruciate in flexion is defined and will not change through-
ligament. The operator must be careful not to out the procedure. To determine whether enough
raise his/her hand (and the powerdrive, allowing space has been created the femoral resection
the blade to plunge) and cut the posterior part of guide set at 4 can be inserted. When the feeling is
the tibia too deeply to avoid damaging the poste- tight 1 mm of cartilage from the posterior femo-
rior tibial cortex. Damaging it drastically ral condyle should be removed, thus proximalis-
increases the risk of fracture. If the sagittal resec- ing the joint line with 1 mm. A recut on the tibia
tion is done in too medial a position, the result to remove an extra 2 mm of bone is not
will be a smaller metal-back tibial implant size, recommended.
resulting in less optimal distribution of The size of the tibial component can be deter-
­mechanical load and increasing the risk of frac- mined (seven different sizes exist) by seeking
ture or stress-shielding pain. This can also lead to optimal coverage between the upturned tibial
insufficient tibial coverage and conflict of the resection specimen placed against a contralateral
88 F. Hardeman and A. Clavé

test implant. Optimal tibial coverage is necessary 7°

to prevent the risk of complications such as pain,
loosening of the insert, subsidence, or fracture. Mechanical
Components should be balanced against the pos- Axis of axis of limb
terior and medial tibial cortices and not overlap femoral shaft
anteriorly. Medial overlap can be tolerated if less
than or equal to 2 mm [4]. In the event of overlap
of more than 2 mm, it is recommended to per-
form a sagittal cut in a more lateral position by
resecting a 2-mm slice flush or at the level of the
medial tibial spine. The metal-back tibial implant
must rest on the posterior cortex because the
mobile insert may, in deep flexion, slide beyond
the posterior border of the metal back and risk its
dislocation.

7.5 Femoral Sections,

Positioning of the Femoral
Implant and Balancing
of the Spaces

Given that the femoral component and polyethyl-

ene mobile insert are fully congruent, the mobile
insert will follow the femoral component all Fig. 7.9 The femoral component is placed parallel to the
along the area of flexion–extension with no femoral mechanical axis thanks to the intramedullary rod
change to the area of contact between the two links to a 7° instrumentation spider-link
parts.
A centromedullary rod will be introduced into increase the area of contact between the femoral
the femoral shaft in the direction of the homolat- component and the mobile insert and allow
eral anterosuperior iliac spine, with its point of hyperflexion. Adjustment to 7° valgus compared
entry located about 10 mm directly above the lat- to the diaphyseal axis will position the compo-
eral wall of the medial femoral condyle. This is nent along an axis parallel to the mechanical axis
slightly medial to the standard rod insertion for of the femur (Fig. 7.9).
TKA. It will be connected to a positioning guide The mediolateral position of the femoral posi-
available in five sizes (corresponding to the sizes tioning guide is not controlled automatically by
of five femoral components — XS to XL). This instrumentation and must be adjusted manually.
guide will be used to make the holes necessary The objective should be to position it at the centre
for proper positioning of the posterior femoral of the femoral condyle, avoiding putting it in the
section guide and drilling spigot for the distal medial half of the medial femoral condyle. The
femoral condyle. It is also used to receive the two two borders of the guide match the width of the
fixation contact points of the femoral component, corresponding femoral component, such that any
whether cemented or not. This part of the ancil- medial or lateral overlap can be identified and
lary material makes it possible to control flexion avoided.
and the components’ position in varus/valgus, Once this guide has been positioned, first the
ideally placing the femoral component in neutral 4- and then the 6-mm holes are drilled. They will
rotation at 10° flexion and 7° valgus. The femoral be used for positioning the posterior femoral
component is best positioned at 10° flexion to resection guide, which is the first femoral resec-
7 Principles of the Oxford® (Zimmer Biomet) Unicompartmental Knee Arthroplasty (OUKA) 89

Window
Spigot

Stop

Fig. 7.11 Milling of the distal femoral condyle and

spigot. The different spigot sizes allow a 1 mm incremen-
tal milling

are determined, the intramedullary rod and all

retractors should be removed because they exert
traction on the ligaments and other soft tissue,
affecting the ligament balance and opening the
Fig. 7.10 Sagittal view of the posterior cut. The amount internal compartment. Therefore, it is also
of bone and cartilage (+ saw blade thickness) is equal to important to check thoroughly for the absence
the thickness of the femoral component of osteophytes that can place the MCL under
tension at the level of the femoral condyle, par-
tion to be performed. This resection, generally in ticularly under the femoral insertion of the
a healthy bone and cartilage area, should remove MCL, and if applicable excise them. The osteo-
a quantity of bone equal to the thickness of the phytes on the tibial side should not be removed
femoral component minus the thickness of the since removal poses a high risk of damaging the
saw blade, thereby recreating a physiological MCL.
posterior offset condyle (Fig. 7.10). To evaluate the difference between the physi-
The distal femoral condyle is then resurfaced ological space in flexion and the space in exten-
by milling. In order to do this, spigots with col- sion, a trial tibia and trial femur are inserted. The
lars of variable thickness will make it possible to space in extension is evaluated in 20° flexion,
adjust the depth of drilling. They are inserted into relaxing the posterior capsule. Measuring the
the 6-mm hole and referenced in two points: the joint space in complete extension exerts tension
bottom of the 6-mm drill hole and the worn sur- on the capsule, limiting opening of the internal
face of the distal femur. The spigots are available compartment, and may also be misleading in
with collars of incremental 1-mm sizes ranging cases of permanent flexion. In anteromedial knee
from 0 to 7 mm. Length of the spigot is identical OA, fixed flexion deformity normally does not
on both sides of the collar, meaning that it can be exceed 15° and measurement of the space in
inserted in both directions. The first milling extension in 20° of flexion is therefore consid-
should be with a spigot of 0 because it will ered more reliable. The space in flexion is evalu-
remove the least amount of bone, enabling clean ated at 90° of the previous position, i.e. 110°
milling. The drill fitting into the spigot has a win- flexion. 110° flexion also corresponds to a posi-
dow through which the spigot can be seen, pre- tion that is perpendicular overall between the
venting milling of the distal condyle going further femoral and tibial components: with the femoral
than required (Fig. 7.11). Usually a first milling component placed at 10° flexion and the metal-­
with spigot 0 will not create a spherical surface of back tibial implant with 7° posterior slope. After
the distal femur yet. drilling with spigot 0, the flexion space evaluated
After this first milling, a zero point is estab- at 110° must be larger than the extension space
lished. This will be the reference for the follow- evaluated at 20° because in the flexion space is
ing measurements. The next stage is the first final (the posterior cut will not be altered), but the
measurement of the difference between the ref- extension space has not been finalised to match
erence space in flexion and the space in exten- the flexion space (more bone has to be removed
sion. When the spaces in flexion and extension from the distal femur).
90 F. Hardeman and A. Clavé

Tests are performed using trial spacers that lodged more deeply, in particular, as stated previ-
measure the residual space between the femoral ously, if it has been impacted by force with a
and tibial components. At this stage, the single-­ large mallet. This would result in excessive with-
pegged femoral trial and tibial trial without keel drawal of bone from the distal femur, risking cre-
are used. These trial spacers place the MCL under ating a too much extension space. Therefore, it is
tension and thus correct the joint space. In flex- important to perform successive millings step-­
ion, the medial compartment is generally the by-­step without skipping steps so as not to risk
same size as the G-clamp used at the start of the over-milling the distal condyle.
procedure. If a 4 G-clamp was used, usually the Precautions should also be taken when treat-
flexion space should be 4 mm with the femoral ing avascular necrosis of the distal femoral
and tibial trial in place. condyle because a major bone defect may exist
Key is to determine the difference between the in the distal femur. In this case, it is recom-
space in flexion and the space in extension. The mended to perform the first milling with a 0
space in extension should be subtracted from the spigot without milling up to the stop, leaving
space in flexion. If the flexion space is 4 (a num- about 2 mm of margin. If the first test shows a
ber 4 spacer is the most appropriate) and the difference in flexion/extension space greater
extension is 1 (a spacer of 1 is the most appropri- than 2 mm, a second milling with the same 0
ate), there is a difference of 3. The spigots deter- spigot can be performed advancing the drill up
mine the depth of milling and so the quantity of to the stop.
bone that will be removed from the distal femur. In rare cases of excessive milling, the space in
Therefore, if the difference between flexion and extension can again be distalised by inserting a
extension is 3, it is necessary to choose a spigot 3.5-mm cortical screw, parallel and next to the
of 3 for the second milling. This one will remove 6-mm hole. By using piling and adjusting its
3 mm of extra bone from the distal femur. It is depth, good balance of the spaces can be reached,
crucial to gently insert this spigot to prevent dam- particularly by retesting with the trial compo-
aging the 6 mm hole, since the basis of the 6 mm nents (Fig. 7.13). In this specific case, the final
hole is the reference for distal femoral bone implant always has to be cemented.
removal. Moreover, it is important to note that the
After this second milling, the sharp edges of sphero-spherical shape of the femoral component
bone on the side and the bone cuff around the enables a certain tolerance in its positioning and
6 mm hole are removed, creating a spherical dis- a difference in varus/valgus, flexion/extension
tal femoral condyle. It is important not to exert and rotation ranging up to 10° is considered
force on the spigots in order not to distort this acceptable (Fig. 7.14).
reference point, particularly from the time when Once a proper balance of the spaces has been
the cuff of bone has been removed. obtained, it is very important, using the appropri-
Next the flexion and extension spaces must be ate ancillary material, to excise posterior osteo-
measured again. If the spaces are balanced, the phytes and remove a possible anterior conflict
procedure can be finished by completing the fem- between the mobile insert and femur with an
oral and tibial preparation. If the space in exten- appropriate drill. Indeed, posterior conflicts on
sion is less than the space in flexion, a third osteophytes and anterior bone are the main causes
milling should be performed. If the difference is of dislocation or wear of mobile inserts. For this,
equal to 1 (e.g. flexion space of 4, extension the anti-impingement guide should be used.
space of 3), it is necessary to choose a larger size Therefore, this step should not be neglected
spigot. If the last milling has been performed under any circumstances.
with a spigot of 3, with the residual difference Finalisation of the tibia preparation comes
equal to 1 the spigot to use is 4 (3 + 1) (Fig. 7.12). last. With the aid of specific instrumentation for
Nevertheless, it is necessary to remember that preparation of the tibia, it can be verified that the
during successive millings, the spigot may be chosen size is fully suitable before the tibial
7 Principles of the Oxford® (Zimmer Biomet) Unicompartmental Knee Arthroplasty (OUKA) 91

Fig. 7.12 How to choose the spigot a 1 mm

to mill the distal femoral condyle
and to balance space in flexion and
extension. (a) Flexion space: 4 mm
and extension space: 1 mm. To
balance the space you have to
increase the space in extension by
4–1 = 3 mm. Thus, you have to mill
3 more mm on the femoral distal
condyle; (b) If a 0 spigot was used
in a, then we have to choose a spigot
4 mm
that will allowed us to mill 3 more
mm. Then we will use a 3–0 = 3
spigot!; (c) Perfect milling; (d)
Perfectly balanced flexion and
extension spaces

Size 3 spigot

4 mm
d

4 mm

insertion pin is made. This stage is important and slot that is too deep or too posterior, with result-
should be performed preferably with a recipro- ing damage to the posterior cortex, two errors
cating “toothbrush” saw blade to avoid making a that risk to fracture the tibial plateau.
92 F. Hardeman and A. Clavé

Fig. 7.13 The space in extension can be distalised by inserting a 3.5-mm cortical screw, parallel and next to the 6-mm
hole. By using adjusting its depth, good balance of the spaces can be reached

Once the final implants have been placed, it is

possible to perform a final test with trial mobile
inserts to define the right thickness/height and
verify the absence of conflict (anterior, posterior
or against the lateral wall of the metal back) or
dislocation. If doubt persists between two
­thicknesses/sizes, in our view, it is preferable to
prefer a smaller thickness PE to limit the risk of
excess MCL tension and overloading of the
external compartment, promoting the occurrence
of pain, PE dislocation, and OA progression. The
most commonly used PE thicknesses are 3, 4, and
5 mm. Use of PE of more than 6 mm is suspect
and should raise questions and, in particular, sus-
picions of damage to the MCL.

7.6 Take-Home Message

Due to the specific characteristics of anterome-

Fig. 7.14 A femoral malposition up to 10° in any dial knee osteoarthritis (including a functionally
direction is considered acceptable intact ACL), the main indication for internal
7 Principles of the Oxford® (Zimmer Biomet) Unicompartmental Knee Arthroplasty (OUKA) 93

UKA, cartilage is conserved in the posterior part References

of the tibia and posterior femoral condyle. The
MCL and joint capsule are retightened in flexion, 1. Victor J, Wong P, Witvrouw E, Sloten JV, Bellemans
preserving their physiological characteristics. J. Quelle est l’isométrie de la sphère fémoro-­
patellaire, de la sphère collatérale superficielle et
The Oxford knee replacement is a mobile des ligaments collatéraux latéraux du genou ? Am
polyethylene insert whose femoral joint surface J Sports Med. 2009;37(10):2028–36. https://doi.
is fully congruent with the sphero-spherical fem- org/10.1177/0363546509337407.
oral component. Consequently, it reproduces 2. Price AJ, Svard U. Une deuxième décennie
d’analyse de survie sur l’arthroplastie unicompar-
similar knee kinematics to physiological condi- timentale du genou à Oxford. Clin Orthop Relat
tion with almost no wear. Res. 2011;469(1):174–9. https://doi.org/10.1007/
Its principle is to copy the physiological space s11999-­010-­1506-­2.
existing in flexion to the space in extension 3. Gulati A, Chau R, Simpson DJ, Dodd CA, Gill HS,
Murray DW. Influence de l’alignement des com-
(where wear occurs). By using the isometric posants sur les résultats pour le remplacement du
characteristics of a healthy MCL and placing it genou unicompartimental. Knee. 2009;16(3):196–9.
under physiological tension, we are able to mea- https://doi.org/10.1016/j.knee.2008.11.001.
sure the difference in space and perform progres- 4. Chau R, Gulati A, Pandit H, Beard DJ, Price AJ, Dodd
CA, Gill HS, Murray DW. Porte-à-faux du composant
sive and controlled drilling of the distal femoral tibial après un remplacement du genou unicomparti-
condyle, giving a good ligament balance and rec- mental - est-ce important ? Genou. 2009;16(5):310–3.
reating a joint space and physiological joint https://doi.org/10.1016/j.knee.2008.12.017.
biomechanics.
 Lateral Unicompartmental Knee
Arthroplasty 8
Axel Schmidt, Matthieu Ollivier,
and Jean-Noël Argenson

8.1 Introduction gical option for external tibiofemoral osteoarthri-

tis treatment but is a more complicated procedure
Lateral unicompartmental knee arthroplasty with longer postoperative follow-up [7–9].
(UKA) accounts for a minority of all UKA per-
formed for osteoarthritis [1], about 10% [2]. Its
rarity is related to the low incidence of genu val- 8.2 Anatomy
gum deformity [3, 4] and better long-term toler-
ance of lateral osteoarthritis (OA) compared to The asymmetry between the lateral and medial
medial osteoarthritis [3, 4]. The specificity of the tibiofemoral space is explained by the anatomical
indications and the anatomical and kinematic characteristics specific to each compartment [10].
characteristics of the external knee compartment Concerning the lateral tibial plateau, the ana-
make lateral UKA surgery more challenging to tomical specificities are convexity of the cartilage
perform than medial UKA. surface, a lower anteroposterior and mediolateral
Alternatives to external UKA for the treat- size than the medial plateau, a long axis of 10–15°
ment of symptomatic, early-stage (Ahlbäck of internal rotation, and a lower external tibial
grades 1 and 2) isolated external unicompartmen- slope than in the internal compartment [11]. The
tal OA in young and active patients are distal lateral femoral condyle, often hypoplastic in val-
femoral and/or proximal tibial varus derotation gus deformity of the knee, is divergent with a lon-
osteotomies, which can correct a valgus defor- ger anteromedial axis oriented in a posterolateral
mity and therefore decrease the load exerted in direction.
the lateral compartment [5]. However, the results Congruence between the femoral condyle and
and survival rates for varus osteotomies are gen- tibial plateau closely depends on the presence,
erally less predictable, the surgical technique is shape, and mobility of the lateral meniscus. In
more difficult to perform than valgus osteotomy cases of meniscectomy, stability is altered,
and postoperative follow-up is longer than for quickly resulting in the osteoarthritic degenera-
UKA [6]. Total knee replacement is the other sur- tive alterations that explain the high number of
young subjects who develop lateral osteoarthritis
after meniscectomy [12].
A. Schmidt (*) · M. Ollivier · J.-N. Argenson All these anatomical specificities in the lateral
Institut du Mouvement et de l’Appareil Locomoteur tibiofemoral compartment will affect the surgical
(Institute of Movement and the Musculoskeletal technique, implants’ positioning, and even the
System), CHU (University Hospital Centre) Sud, choice of implant used [13, 14].
Marseille, France

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2024 95

A. Clavé, F. Dubrana (eds.), Unicompartmental Knee Arthroplasty,
https://doi.org/10.1007/978-3-031-48332-5_8
96 A. Schmidt et al.

Although cartilage wear develops primarily in the tibia is in a position of maximum stability
the anterior part of the medial compartment, with the femur. This process, called “the screw-­
osteoarthritis often starts in the posterior part of home mechanism”, is key to knee stability in
the lateral compartment [15]. This particularity is normal gait [18]. External rotation of the tibia
important to orient the paraclinical preoperative during extension should be understood by the
diagnosis, with Rosenberg X-ray views making it operator, who must position the femoral implant
possible to assess the severity of wear correctly. as laterally as possible. Excess internal transla-
tion of the femoral component may result in
conflict with the mass of spinous processes in
8.3 Kinematics of a Native Knee extension.
and with UKA

The biomechanics and kinematics differ between 8.4 Indications and Preoperative
the medial and lateral tibiofemoral compartment. Assessment
These variations will explain the technical par-
ticularities during lateral UKA placement. The indications for lateral UKA are based on
Kinematics of a native knee consists of pro- anatomical and radiological criteria grouped in
gressive external rotation of the femur on the Table 8.1 [19].
tibia during flexion in combination with posterior Age is no longer an absolute contraindication
recession of the femoral condyles, which is for UKA [20–22]. In particular, good results have
greater for the lateral femoral condyle (10 mm) been reported in cases of post-traumatic osteoar-
than for the medial condyle (2 mm). This involves thritis in patients under 60 years of age [23, 24].
the concept of “medial pivoting of the knee”: the
medial compartment is the area of knee stability
while the lateral compartment is the area of Table 8.1 Indications and contraindications of lateral
UKA
mobility [16].
The femur is in neutral position at 0° rotation Indications Contraindications
during extension and will turn progressively dur- – Primitive lateral – OA in other compartments:
osteoarthritis Internal tibiofemoral,
ing flexion up to 7° external rotation in the mid- (secondary to patellofemoral
dle of flexion. Translation of the medial femoral constitutional (particularly at the expense
condyle seems to be correlated with the integrity valgus knee). of the lateral facet).
of the ACL (anterior cruciate ligament) [17] – Avascular – Chronic anterior laxity.
osteonecrosis – Frontal laxity.
while mobility of the lateral femoral condyle in (femoral condyle or – Valgus deformity >15° or
the tibial plateau seems to be independent of tibial plateau). nonreducible [25].
osteoarthritic degenerative changes and ACL – Post-traumatic – Preoperative knee flexion
status. secondary OA deformity >15°.
(fracture mainly of – Knee stiffness (flexion
At the end of extension, between 0° and 20° the lateral tibial <100°).
flexion, external rotation of the tibia occurs after plateau). – History of femoral or tibial
tension is exerted by the two cruciate ligaments, – Lateral post-­ osteotomies.
making it possible to block the knee. Therefore, meniscectomy OA. – Inflammatory disorders.
8 Lateral Unicompartmental Knee Arthroplasty 97

8.5 Clinical Examination

The preoperative clinical examination should be

especially attentive to detail and look for several
points considered as contraindications to perfo-
roint amplitudes will be assessed as well as the
origin of pain, which should be localised pre-
cisely in the lateral tibiofemoral compartment
(“finger sign”) with no sign of associated patel-
lofemoral or internal tibiofemoral damage.
Alignment of the lower limb should be
assessed to look for a valgus knee deformity,
which will be quantified (absolute value and
reducibility) and compared to the opposite limb.
The assessment of sagittal and frontal ligament
laxities will make it possible to evaluate the
integrity of the central pivoting and collateral
ligaments. Special attention should be paid dur-
ing examination of the ACL because clinical
signs can be a little difficult to interpret in light of
pain associated with OA and intra-articular
effusion.
Fig. 8.1 X-ray of the right knee (anteroposterior, profile
and anteroposterior schuss views) revealing isolated
8.6 Imaging lateral tibiofemoral OA

The standard preoperative radiological assess- Preoperative, clinical, and X-ray assessment
ment should include anteroposterior and profile should seek to determine the origin of the valgus
views of the knee with weightbearing on one deformity in order to differentiate six situations:
foot, an axial view of the patella in 45° flexion, an
anteroposterior schuss view, pan-goniometry of –– Lateral femoral condyle dysplasia [26].
the lower limbs and weightbearing views in varus –– Post-traumatic valgus secondary to a fracture
and valgus. This preoperative assessment makes of the tibial plateau or lateral condyle [23].
it possible to evidence an isolated characteristic –– Lateral post-meniscectomy pain syndrome
of external tibiofemoral OA, with no internal tib- [27].
iofemoral or patellofemoral damage, and to –– Avascular osteonecrosis of the femoral con-
quantify the severity of OA according to the dyle or lateral tibial plateau.
Ahlbäck classification. X-ray views with weight- –– Valgus secondary to coxofemoral disorder in a
bearing will make it possible to assess reducibil- native or prosthetic hip [28].
ity of the deformity in the frontal plane and –– Valgus secondary to congenital tibial defor-
thickness of the cartilage in the internal compart- mity [29].
ment in X-ray views in varus position. Pan-­
goniometry will assess the overall deformity of Femoral condyle hypoplasia is the most com-
the lower limbs and look for the origin of a valgus mon cause [4]. In specific cases, the position of
deformity by calculating the femoral and tibial the femoral component should be adapted to the
mechanical angles (Figs. 8.1 and 8.2). severity of the dysplasia. In cases of severe hypo-
Specificity of preoperative planning for lateral plasia, the femoral implant should be positioned
UKA: more distally and more posteriorly to correct dys-
98 A. Schmidt et al.

population normally concerned. Placement of a

UKA makes it possible to correct the intra-­
articular deformity secondary to a fracture.
Lustig et al. [23] have reported improvement in
pain and function with post-traumatic lateral
UKA, as well as excellent survival at 5 and 10
years with 80% good results at 15 years. Even
though the number of indications is very limited
and it involves a rigorous procedure, lateral UKA
can be an effective option for treatment of post-­
traumatic OA. Survival of implants in this con-
text seems comparable to survival of lateral UKA
for primary OA.
The last two situations are less common and
are more often indications for treatment with
TKR or tibial osteotomies [29].

8.7 Surgical Technique

8.7.1 Approach

A lateral parapatellar approach is traditionally

used for this procedure even though some authors
have reported the possibility of performing it
with a medial parapatellar approach [30].
The skin incision extends from the lateral
upper pole of the patella and ends distally 2 cm
Fig. 8.2 Pan-goniometry of the lower limbs evidencing below the joint space on the lateral border of the
valgus disorder (HKA angle = 187°) ATT. Lateral arthrotomy is then performed, and
the patella will be pushed back medially.
plasia at its origin in both the sagittal plane and To improve intra-articular exposure, the lat-
frontal plane, in order to restore an intra-articular eral facet (external vertical patellectomy) and lat-
space that is closer to the knee’s normal eral patellar osteophytes can be resected in
anatomy. combination with excision of the lateral portion
In cases of post-traumatic [23] or post-­ of Hoffa’s fat pad tissue.
meniscectomy OA and osteonecrosis, there is no The entire joint is then explored, making it
bone dysplasia to be compensated. However, possible to confirm the isolated characteristic of
poor quality of subchondral bone as well as an OA and the absence of anterior laxity (integrity
architectural anomaly (of the tibial plateau or of the ACL).
femoral condyle) secondary to comminution and During exposure of the external tibiofemoral
depression should be anticipated. A precise pre- joint, it is important that release of the peripheral
operative assessment by CT scan will make it capsule around the tibial plateau and osteophytes
possible to plan potential bone graft procedures be as minimal as possible to leave the peripheral
or insertions of screws for reinforcement in order ligament structures intact. This is an essential
to obtain a satisfactory bone support for place- point during UKA placement, ensuring final
ment of the implant. Cases of post-traumatic OA undercorrection that is favourable to good liga-
affect patients who are often younger than the ment tension [31]. In fact, similarly to medial
8 Lateral Unicompartmental Knee Arthroplasty 99

UKA, the principles of ligament balance on both bility in flexion in cases of excessive slope, or
the lateral collateral ligament and fascia lata are stiffness in cases of an insufficient slope generat-
not applicable in external unicompartmental ing a tight lateral compartment [33, 34]. An
surgery. excessive slope will also have the effect of
In a first phase, osteophytes in the intercondy- increasing anterior tibial translation, which is the
lar notch are resected to prevent conflict with the source of secondary lesions of the ACL [35].
ACL, responsible for secondary rupture. To conclude, sagittal tibial resection should be
Regarding osteophytes in the lateral femoral performed as close as possible to the tibial spine
­condyles, it is important to conserve them ini- mass while conserving them. It is performed with
tially because they will aid subsequent position- the knee in flexion and should follow a line con-
ing of the femoral implant [32]. In fact, the latter necting two markers for identification:
should be positioned as laterally as possible when
the knee is in flexion, which implies that in some –– With the knee in flexion, it involves the most
cases it will be partially pressing on the condylar medial point of the middle of the anteroposte-
osteophytes. rior axis of the external tibial plateau, behind
Before any bone resection, it is useful to the ACL insertion.
note the contact point between the anterior part –– With the knee in extension, the most medial
of the femoral condyle and the anterior part of point of the anterior part of the external tibial
the tibial plateau when the knee is in extension. plateau, in front of the ACL insertion.
This point will serve as a marker subsequently
for the positioning of implants (size and Because of the natural orientation of the exter-
direction). nal tibial plateau in external rotation (“screw-­
home mechanism” [18]), this line will cross the
patellar ligament, which must be carefully
8.7.2 First Stage: Tibial Resection retracted to avoid injuring it during bone
resection.
Horizontal tibial resection is performed with an
extramedullary guide. The objective is to perform
bone resection that is the most economical and 8.7.3 Second Stage: Femoral
conservative possible, orthogonal to the mechan- Resections
ical tibial axis, in order to avoid having to increase
the thickness of the tibial insert to restore align- Distal femoral resection should be as minimal as
ment and stability [2]. possible to “distalise” the femoral implant, com-
In fact, in valgus knee deformity, where OA pensating on the one hand for possible hypopla-
generally affects the femoral condyle more [31], sia of the femoral condyle and on the other hand
minimal bone resection will ensure a larger area for wear, essentially femoral in valgus knee
of resection, enabling better weightbearing on the deformity. Since wear develops mainly in the
tibial cortex for the future implant. If the surgeon posterior part of the lateral compartment, carti-
wants to conserve the patient’s valgus deformity lage often persists in the distal femoral condyle,
partially, this should not be done during tibial even in cases of post-traumatic OA after a tibial
resection but preferably during femoral resection, plateau fracture. This cartilaginous remainder
where generally the origin of valgus knee defor- should be removed to resect the distal condyle
mity lies. There is very little metaphysis tibial [2].
deformity in valgus knee deformity. Each implant has specific characteristics, but
The tibial slope of the lateral compartment, the major principles and stages are similar. The
which is lower than that of the medial compart- distal femoral resection can be performed using
ment [11], should be reproduced to prevent insta- two different techniques:
100 A. Schmidt et al.

–– Independent resections: performed with the to avoid subsequent conflict between the implant
aid of an intramedullary resection guide. and patella due to excess anterior coverage.
–– The opening of an intramedullary femoral Once the posterior section and the bevel have
guide is centred above the apex of the inter- been performed, it is essential to look for and
condylar notch. Distal femoral resection is resect any possible posterior osteophytes in order
performed depending on the HKS angle (angle to obtain the best joint amplitude and prevent
between the mechanical femoral axis and the conflict between the polyethylene insert during
anatomical femoral axis, generally between 4° major flexion.
and 6°).
–– Dependent sections: the knee is placed in
extension after resection of the tibia. The dis- 8.7.4 Third Stage: Positioning
tal femoral resection guide is placed in the tib- the Implants
iofemoral space as done with a spacer.
The size of the tibial implant is chosen once all
The posterior femoral resection and bevelling bone resections have been performed. It involves
will then be performed using guides for appropri- the best compromise between maximum tibial
ate resection size. coverage without over dimensioning or overlap-
Posterior femoral resection should be as mini- ping the implant in the frontal and sagittal planes.
mal as possible to compensate for posterior con- The tibial implant should be as close as possible
dylar offset and therefore obtain a tibiofemoral to the tibial spine mass and have 15–20° internal
space in flexion similar to the tibiofemoral space rotation.
in extension. With this technique, the femoral The femoral implant should be placed, with
implant will not reproduce the patient’s initial the knee in flexion, in external rotation and as
anatomy but will be used to compensate for ­laterally as possible, sometimes meaning that it
hypoplasia of the femoral condyle. lies partially on osteophytes [32]. This technique
Rotation of the resection guide is essential makes it possible to obtain ideal contact with the
because it will determine the rotation of the tibia without entering into conflict with the tibial
future implant. Considering the natural divergent spines during a shift in position of the knee in
aspect of the lateral femoral condyle compared to extension related to the divergent anatomical
the medial condyle, it will be necessary to posi- shape of the lateral femoral condyle during
tion the resection guide in order to avoid having flexion.
excess internal rotation of the implant in flexion, The knee is then placed in maximum flexion
which would result in a conflict with the anterior and in internal rotation to facilitate final prepara-
tibial spines during a shift to extension. tion of the tibia, consisting of creating anchoring
Size of the resection guide is a compromise contact points for the keel pin of the final implant
between an anatomical position centred on the in the subchondral bone.
femoral condyle and the long axis of impact per- The knee’s stability is then tested with the test
pendicular to the tibial plateau. Special attention implants and a test insert. During movements of
should be paid so as not to “overdimension” the flexion/extension, the medial part of the femoral
femoral implant. There should instead be a ten- implant should remain opposite to the middle of
dency to “underdimension” the femoral compo- the tibial implant. It is important to look for a
nent. The anterior limit of impact should be conflict between the femur and the tibial spine
located at the level of the mark previously made mass in extension, which would be secondary to
before bone resection, at the anterior contact lack of external rotation in flexion of the femoral
point between the femur and tibia on the knee in implant. Testing the spaces in flexion and exten-
extension. This marker lies about 1–2 mm below sion will assess residual frontal laxity, patellar
the border between bone and cartilage created by travel, and the absence of conflict between the
distal femoral resection and will make it possible patella and femoral implant, leaving the surgeon
8 Lateral Unicompartmental Knee Arthroplasty 101

to choose the correct thickness of the PE insert,

often thicker in lateral UKA than for medial
UKA because of femoral dysplasia.
At the end of the surgical procedure, it is nec-
essary to make certain that slight residual lateral
laxity is still present when the knee is tested by
blocking it in varus with 15° flexion (unlocked
knee). In fact, it is essentially to undercorrect the
deformity in lateral UKA in order to avoid any
overconstraint of the medial compartment, which
is essential for long-term survival [24] (develop-
ment of OA in the medial compartment).
The strategy of lateral UKA corresponds to a
resurfacing procedure whose goal is to correct
only intra-articular wear while leaving the extra-­
articular deformity intact. In any event, it does
not involve a procedure whose purpose is correc-
tion of the lower limb deformity.
The final tibial implant is cemented and placed
first in the knee in complete flexion and in inter-
nal rotation to increase exposure of the lateral
compartment. The femoral implant is then
cemented, and the PE insert is placed after care-
ful cleaning of the metal-backed tibia (Fig. 8.3).

Fig. 8.3 Perioperative views of a Zimmer/Lima ZUK

metal-backed implant for lateral UKA (left panel: knee in
complete extension, middle panel: knee in mid-flexion,
right panel: knee in 90° flexion)
102 A. Schmidt et al.

8.8 Traditional Technical Errors on tibial “all-polyethylene” cemented fixed

and Perioperative Difficulties implants and on resurfacing on the femoral
­component, Lustig et al. [24] reported excellent
Overcorrection of a valgus deformity towards an long-­term survival with rates of 94.4% at 10
axis in varus will result in excess constraint in the years and 91.4% at 15 years, as well as very sat-
medial tibiofemoral compartment and early isfactory clinical scores (Fig. 8.4). The main fac-
development of medial OA. tor in failure resulting in repeat surgery in this
The divergent anatomical shape of the lateral study was OA progression in the medial tibio-
femoral condyle during flexion should be consid- femoral compartment. No case of patellofemoral
ered to avoid conflict between the implant and tib- OA was reported. These results were confirmed
ial spines during extension. Special attention by Deroche et al. [40], who found OA progres-
should be paid during tibial resection to avoid an sion in the medial compartment as the main cause
excess slope that would affect ligament alignment. of failure (87.5%) in a series of identical implants,
The tibial implant should be positioned with followed by aseptic loosening of the tibial
internal rotation of 15–20° and aligned with a implant (12.5%). With mean follow-up of 17.9
natural tibial slope. years, the revision surgery rate was 20.5%, and
for patients who did not undergo revision, the sat-
isfaction rate was excellent at 90.5% with good
8.9 Results and Survival clinical results. Survival was 82.1% at 15 years
of Lateral UKA and 79.4% at 20 years.
Excellent results have also been reported with
Concerning modern UKA, the literature reveals cemented implants consisting of a metal-backed
very good results with mean and long-term sur- tibial implant (Fig. 8.5) and using a so-called
vival greater than 90% [19, 36, 37]. In cases of resection technique in the short [42] and long
failure, surgical revision is easier with better terms [32]. Argenson et al. [32] found excellent
results and more satisfied patients with UKA revi- survival with these implants of 92% at 10 years
sion by TKR than TKR revision by TKR [38]. and 84% at 16 years for isolated involvement of
Long-term clinical and radiological results of the lateral compartment (essential OA, post-­
lateral UKA are good with functional scores, traumatic OA, osteonecrosis).
patient satisfaction, and survival at 10, 16, and 22 Concerning mobile-bearing plateau implants,
years comparable to the results of medial UKA Fornell et al. [41] found excellent survival of
[32]. Recent studies report lower rates of revision 97.5% at 5 years with a 2.4% revision rate at
than those recorded in the first studies on lateral 49 months’ follow-up. The main reason for fail-
UKA. In a 2002 series with 21-year follow-up, ure in series with the mobile-bearing plateau in
Ashraf et al. [39] found 10-year survival of 83% the lateral compartment was dislocation of the
and 74% for 15 years. Currently, studies report insert [43]. Long-term survival of the implant
survival greater than 90% in the medium term will also depend on postoperative alignment. In
and 80% in the long term [40, 41]. fact, a defect in alignment with postoperative
This improvement in results is related to better varus will have the consequence of worsening
preoperative selection of patients, surgical tech- the development of OA in the medial compart-
niques, and prosthetic implants. ment. A defect in postoperative alignment in the
upper frontal plane with 7° residual valgus
deformity has been reported as a risk factor for
8.10 Failure and Revision early failure with a risk 7 times higher for re-
revision [44].
The main reason for failure of lateral UKA is OA External OA can also affect young and active
progression in other compartments, particularly patients who wish to resume physical activity
the medial compartment. In a retrospective series similar to their preoperative level quickly. Canetti
8 Lateral Unicompartmental Knee Arthroplasty 103

Fig. 8.4 Postoperative repeat X-ray control for lateral UKA with a full-PE tibial implant from the company Corin
(image used with the permission of Dr. Guillaume Demey)
104 A. Schmidt et al.

8.11 Conclusion

With appropriate surgical technique (positioning,

implant size) and selected patients, lateral uni-
compartmental knee arthroplasty is an effective
procedure for isolated damage in the lateral com-
partment with good long-term results similar to
those with medial UKA [31, 32].
Given the anatomical and biomechanical dif-
ferences between the medial compartment and
lateral compartment, the technical specificities
should be known and mastered when performing
lateral UKA. Lateral UKA should be considered
as a resurfacing procedure without correction of
the deformity to prevent early failure related to
development of medial tibiofemoral OA. Implants
with a fixed plateau may be preferable consider-
ing the important mobility of the lateral tibio-
femoral compartment.

References
1. Parratte S, Ollivier M, Lunebourg A, Abdel MP,
Argenson J-N. Long-term results of compart-
mental arthroplasties of the knee: long term
results of partial knee arthroplasty. Bone Joint
J. 2015;97-B(10 Suppl A):9–15. https://doi.
org/10.1302/0301-­620X.97B10.36426.
2. Scott RD. Lateral unicompartmental replacement: a
road less traveled. Orthopedics. 2005;28(9):983–4.
3. Ranawat AS, Ranawat CS, Elkus M, Rasquinha
VJ, Rossi R, Babhulkar S. Total knee arthroplasty
for severe valgus deformity. J Bone Joint Surg
Am. 2005;87(Suppl 1(Pt 2)):271–84. https://doi.
Fig. 8.5 Postoperative repeat X-ray control of lateral
org/10.2106/JBJS.E.00308.
UKA with a Zimmer/Lima ZUK metal-backed tibial
4. Rossi R, Rosso F, Cottino U, Dettoni F, Bonasia DE,
implant
Bruzzone M. Total knee arthroplasty in the valgus
knee. Int Orthop. 2014;38(2):273–83. https://doi.
org/10.1007/s00264-­013-­2227-­4.
et al. [45] found a rate of return to contact sports 5. Madelaine A, Lording T, Villa V, Lustig S, Servien E,
with low or moderate impact of between 94% Neyret P. The effect of lateral opening wedge distal
and 100% after 4.2–10.5 months in a series of femoral osteotomy on leg length. Knee Surg Sports
Traumatol Arthrosc. 2016;24(3):847–54. https://doi.
athletic patients who underwent surgery either by
org/10.1007/s00167-­014-­3387-­5.
conventional or robot-assisted technique. In a 6. de Andrade MAP, Gomes DCFF, Portugal AL, de A
review of the literature, Witjes et al. [46], who e. Silva GM. Distal femoral varusing for osteoarthri-
combined 7 studies (including 2 on lateral UKA), tis of valgus knee: a long-term follow-up. Rev Bras
Ortop. 2015;44(4):346–50. https://doi.org/10.1016/
found a rate of return to high-impact sports in
S2255-­4971(15)30165-­8.
8%, 22% for intermediate impact, and 70% for 7. Arirachakaran A, Choowit P, Putananon C, Muangsiri
low-impact activities. S, Kongtharvonskul J. Is unicompartmental knee
8 Lateral Unicompartmental Knee Arthroplasty 105

arthroplasty (UKA) superior to total knee arthro- ral joint during normal gait: three-dimensional analy-
plasty (TKA)? A systematic review and meta-­ sis. Clin Orthop Surg. 2015;7(3):303–9. https://doi.
analysis of randomized controlled trial. Eur J Orthop org/10.4055/cios.2015.7.3.303.
Surg Traumatol. 2015;25(5):799–806. https://doi. 19. Walker T, Aldinger PR, Streit MR, Gotterbarm
org/10.1007/s00590-­015-­1610-­9. T. Lateral unicompartmental knee arthroplasty – a
8. Drager J, Hart A, Khalil JA, Zukor DJ, Bergeron SG, challenge. Oper Orthop Traumatol. 2017;29(1):17–
Antoniou J. Shorter hospital stay and lower 30-day 30. https://doi.org/10.1007/s00064-­016-­0476-­2.
readmission after unicondylar knee arthroplasty 20. Fabre-Aubrespy M, Ollivier M, Pesenti S, Parratte
compared to total knee arthroplasty. J Arthroplast. S, Argenson J-N. Unicompartmental knee arthro-
2016;31(2):356–61. https://doi.org/10.1016/j. plasty in patients older than 75 results in better clini-
arth.2015.09.014. cal outcomes and similar survivorship compared to
9. Zuiderbaan HA, van der List JP, Chawla H, Khamaisy Total knee Arthroplasty. A matched controlled study.
S, Thein R, Pearle AD. Predictors of subjective out- J Arthroplasty. 2016;31(12):2668–71. https://doi.
come after medial unicompartmental knee arthro- org/10.1016/j.arth.2016.06.034.
plasty. J Arthroplast. 2016;31(7):1453–8. https://doi. 21. Price AJ, Dodd C, Svard UGC, Murray DW. Oxford
org/10.1016/j.arth.2015.12.038. medial unicompartmental knee arthroplasty in
10. Miyatake N, Sugita T, Aizawa T, et al. Comparison patients younger and older than 60 years of age. J
of intraoperative anthropometric measurements of Bone Joint Surg Br. 2005;87(11):1488–92. https://doi.
the proximal tibia and tibial component in total knee org/10.1302/0301-­620X.87B11.16324.
arthroplasty. J Orthop Sci. 2016;21(5):635–9. https:// 22. Parratte S, Argenson J-NA, Pearce O, Pauly V,
doi.org/10.1016/j.jos.2016.06.003. Auquier P, Aubaniac J-M. Medial unicompart-
11. Weinberg DS, Williamson DFK, Gebhart JJ, mental knee replacement in the under-50s. J Bone
Knapik DM, Voos JE. Differences in medial and Joint Surg Br. 2009;91(3):351–6. https://doi.
lateral posterior tibial slope: an osteological review org/10.1302/0301-­620X.91B3.21588.
of 1090 tibiae comparing age, sex, and race. Am 23. Lustig S, Parratte S, Magnussen RA, Argenson J-N,
J Sports Med. 2017;45(1):106–13. https://doi. Neyret P. Lateral unicompartmental knee arthroplasty
org/10.1177/0363546516662449. relieves pain and improves function in posttraumatic
12. Longo UG, Ciuffreda M, Candela V, et al. Knee osteoarthritis. Clin Orthop Relat Res. 2012;470(1):69–
osteoarthritis after arthroscopic partial meniscec- 76. https://doi.org/10.1007/s11999-­011-­1963-­2.
tomy: prevalence and progression of radiographic 24. Lustig S, Lording T, Frank F, Debette C, Servien E,
changes after 5 to 12 years compared with contralat- Neyret P. Progression of medial osteoarthritis and
eral knee. J Knee Surg. 2019;32(5):407–13. https:// long term results of lateral unicompartmental arthro-
doi.org/10.1055/s-­0038-­1646926. plasty: 10 to 18 year follow-up of 54 consecutive
13. Demange MK, Von Keudell A, Probst C, Yoshioka implants. Knee. 2014;21(Suppl 1):S26–32. https://
H, Gomoll AH. Patient-specific implants for lateral doi.org/10.1016/S0968-­0160(14)50006-­3.
unicompartmental knee arthroplasty. Int Orthop. 25. Kozinn SC, Scott R. Unicondylar knee arthroplasty. J
2015;39(8):1519–26. https://doi.org/10.1007/ Bone Joint Surg Am. 1989;71(1):145–50.
s00264-­015-­2678-­x. 26. Feldman DS, Goldstein RY, Kurland AM, Sheikh
14. Greco NJ, Cook GJE, Lombardi AV, Adams JB, Taha AM. Intra-articular osteotomy for genu valgum
Berend KR. Lateral unicompartmental knee arthro- in the knee with a lateral compartment deficiency. J
plasty utilizing a modified surgical technique and spe- Bone Joint Surg Am. 2016;98(2):100–7. https://doi.
cifically adapted fixed-bearing implant. Surg Technol org/10.2106/JBJS.O.00308.
Int. 2019;34:371–8. 27. Pengas IP, Nash W, Khan W, Assiotis A, Banks J,
15. Gulati A, Chau R, Beard DJ, Price AJ, Gill HS, McNicholas MJ. Coronal knee alignment 40 years
Murray DW. Localization of the full- thickness car- after total meniscectomy in adolescents: a prospective
tilage lesions in medial and lateral unicompartmental cohort study. Open Orthop J. 2017;11:424–31. https://
knee osteoarthritis. J Orthop Res. 2009;27(10):1339– doi.org/10.2174/1874325001711010424.
46. https://doi.org/10.1002/jor.20880. 28. Barrios JA, Heitkamp CA, Smith BP, Sturgeon MM,
16. Argenson J-NA, Komistek RD, Aubaniac J-M, et al. Suckow DW, Sutton CR. Three- dimensional hip and
In vivo determination of knee kinematics for subjects knee kinematics during walking, running, and single-­
implanted with a unicompartmental arthroplasty. limb drop landing in females with and without genu
J Arthroplasty. 2002;17(8):1049–54. https://doi. valgum. Clin Biomech (Bristol, Avon). 2016;31:7–11.
org/10.1054/arth.2002.34527. https://doi.org/10.1016/j.clinbiomech.2015.10.008.
17. Du PZ, Markolf KL, Boguszewski DV, McAllister 29. van Lieshout WAM, van Ginneken BJT, Kerkhoffs
DR. Femoral contact forces in the anterior cruciate GMMJ, van Heerwaarden RJ. Medial closing wedge
ligament deficient knee: a robotic study. Arthroscopy. high tibial osteotomy for valgus tibial deformities:
2018;34(12):3226–33. https://doi.org/10.1016/j. good clinical results and survival with a mean 4.5
arthro.2018.06.051. years of follow-up in 113 patients. Knee Surg Sports
18. Kim HY, Kim KJ, Yang DS, Jeung SW, Choi HG, Traumatol Arthrosc. 2019. https://doi.org/10.1007/
Choy WS. Screw-home movement of the tibiofemo- s00167-­019-­05480-­9.
106 A. Schmidt et al.

30. Sah AP, Scott RD. Lateral unicompartmental knee 39. Ashraf T, Newman JH, Evans RL, Ackroyd
arthroplasty through a medial approach. Surgical CE. Lateral unicompartmental knee replacement
technique. J Bone Joint Surg Am. 2008;90(Suppl 2 Pt survivorship and clinical experience over 21 years. J
2):195–205. https://doi.org/10.2106/JBJS.H.00257. Bone Joint Surg Br. 2002;84(8):1126–30.
31. Ollivier M, Abdel MP, Parratte S, Argenson 40. Deroche E, Batailler C, Lording T, Neyret P, Servien
J-N. Lateral unicondylar knee arthroplasty (UKA): E, Lustig S. High survival rate and very low wear of
contemporary indications, surgical technique, and lateral unicompartmental arthroplasty at long term:
results. Int Orthop. 2014;38(2):449–55. https://doi. a case series of 54 cases at a mean follow-up of 17
org/10.1007/s00264-­013-­2222-­9. years. J Arthroplasty. 2019;34(6):1097–104. https://
32. Argenson J-NA, Parratte S, Bertani A, Flecher X, doi.org/10.1016/j.arth.2019.01.053.
Aubaniac J-M. Long-term results with a lateral 41. Fornell S, Prada E, Barrena P, García-Mendoza A,
unicondylar replacement. Clin Orthop Relat Res. Borrego E, Domecq G. Mid-term outcomes of mobile-­
2008;466(11):2686–93. https://doi.org/10.1007/ bearing lateral unicompartmental knee arthroplasty.
s11999-­008-­0351-­z. Knee. 2018;25(6):1206–13. https://doi.org/10.1016/j.
33. Karimi E, Norouzian M, Birjandinejad A, Zandi R, knee.2018.05.016.
Makhmalbaf H. Measurement of posterior tibial slope 42. Kim KT, Lee S, Kim J, Kim JW, Kang MS. Clinical
using magnetic resonance imaging. Arch Bone Jt results of lateral unicompartmental knee arthro-
Surg. 2017;5(6):435–9. plasty: minimum 2-year follow-up. Clin Orthop
34. Moreland JR, Bassett LW, Hanker GJ. Radiographic Surg. 2016;8(4):386–92. https://doi.org/10.4055/
analysis of the axial alignment of the lower extremity. cios.2016.8.4.386.
J Bone Joint Surg Am. 1987;69(5):745–9. 43. Pandit H, Jenkins C, Beard DJ, et al. Mobile bear-
35. Dejour H, Bonnin M. Tibial translation after anterior ing dislocation in lateral unicompartmental knee
cruciate ligament rupture. Two radiological tests com- replacement. Knee. 2010;17(6):392–7. https://doi.
pared. J Bone Joint Surg Br. 1994;76(5):745–9. org/10.1016/j.knee.2009.10.007.
36. Vasso M, Del Regno C, Perisano C, D’Amelio A, 44. Perkins TR, Gunckle W. Unicompartmental knee
Corona K, Schiavone Panni A. Unicompartmental knee arthroplasty: 3- to 10-year results in a community
arthroplasty is effective: ten year results. Int Orthop. hospital setting. J Arthroplast. 2002;17(3):293–7.
2015. https://doi.org/10.1007/s00264-­015-­2809-­4. 45. Canetti R, Batailler C, Bankhead C, Neyret P, Servien
37. Pandit H, Jenkins C, Gill HS, Barker K, Dodd C, E, Lustig S. Faster return to sport after robotic-­
Murray DW. Minimally invasive Oxford phase 3 assisted lateral unicompartmental knee arthroplasty:
unicompartmental knee replacement: results of 1000 a comparative study. Arch Orthop Trauma Surg.
cases. J Bone Joint Surg Br. 2011;93(2):198–204. 2018;138(12):1765–71. https://doi.org/10.1007/
https://doi.org/10.1302/0301-­620X.93B2.25767. s00402-­018-­3042-­6.
38. Lunebourg A, Parratte S, Ollivier M, Abdel MP, 46. Witjes S, Van Geenen RCI, Koenraadt
Argenson J-NA. Are revisions of unicompartmental KLM. Expectations of younger patients concerning
knee arthroplasties more like a primary or revision activities after knee arthroplasty: are we asking the
TKA? J Arthroplasty. 2015;30(11):1985–9. https:// right questions? Qual Life Res. 2017;26(2):403–17.
doi.org/10.1016/j.arth.2015.05.042. https://doi.org/10.1007/s11136-­016-­1380-­9.
 Kinematic Alignment Technique
for Medial Unicompartmental 9
Knee Arthroplasty

Charles C. J. Rivière, Philippe Cartier,

and Cédric Maillot

9.1 Summary 9.2 Definition

The kinematic alignment (KA) technique for uni- The traditional technique for partial or total knee
compartmental knee arthroplasty (UKA) has arthroplasty (TKA) consists of reproducing the
been performed successfully for decades, even systematised positioning of prosthetic compo-
though the terminology “kinematic” is of more nents, neglecting anatomical variations of the
recent introduction. This chapter helps the sur- knee between each subject. Therefore, the stan-
geon understand the theoretical bases and surgi- dard for decades has been to align the prosthetic
cal principles of the technique. The objective is to components perpendicularly on the femoral and
encourage surgeons to use the KA technique tibial mechanical axes in the frontal plane, and to
because it is a simple, safe, more anatomical, produce an identical tibial slope for all patients;
more physiological and probably clinically this is called the mechanical alignment (MA)
advantageous method compared to traditional technique for TKA [1] and UKA [2]. At the cost
mechanical alignment for UKA. More investiga- of altering the knee anatomy and physiological
tions are necessary to better define its clinical ligament balance, the MA technique was sup-
impact and if there are limits to follow for the posed to ensure surgical reproducibility and clin-
alignment of prosthetic components. ical results. Nevertheless, functional performance
and patient perception of MA knee replacements
were sometimes disappointing [3, 4], despite the
precise implantation of sophisticated implants [5,
6]. This has led to the development of more per-
sonalised and physiological implantation tech-
niques that better reflect the individual anatomy
of the knee and balance of soft tissue, known
C. C. J. Rivière (*) under the term kinematic alignment (KA)
Clinique du Sport Bordeaux-Mérignac,
(Fig. 9.1) [1, 7].
Mérignac, France
Based on the same principle as KA-TKA, the
The Lister Hospital, London, UK
aim is to co-align the UKA components with the
P. Cartier kinematic axes that dictate native movement of
Clinique Hartmann, Neuilly-sur-Seine, France
the tibia around the femur [8, 9], and to resect an
C. Maillot equivalent bone and cartilage thickness to that of
Hôpitaux Universitaire Beaujon – Bichat, APHP,
the implant. Therefore, the components are
Paris, France

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2024 107
A. Clavé, F. Dubrana (eds.), Unicompartmental Knee Arthroplasty,
https://doi.org/10.1007/978-3-031-48332-5_9
108 C. C. J. Rivière et al.

Fig. 9.1 This figure illustrates the kinematic alignment the pre-arthritic knee; these are de facto personalised and
(KA) technique for unicompartmental knee arthroplasty physiological methods of implanting prosthetic knee
(UKA) and total knee arthroplasty (TKA). The KA tech- components
nique aims to restore the anatomy and ligament balance of

aligned in parallel to the cylindrical axis and per- School, France) on the metaphyseal–epiphyseal
pendicular to the tibial longitudinal axis, respec- axis of the proximal tibia [12].
tively (Fig. 9.2). In simplified terms, the KA The terminology describing the alignment
technique for UKA aims to produce “true”’ joint technique is derived from the reference marker
surfacing by restoring the level and three-­ used to align UKA components: the KA and MA
dimensional orientation of the native joint space techniques align the UKA components on the
in the implanted knee compartment (Fig. 9.3). In kinematic axes of the knees and on the mechani-
cases of medial KA-UKA, the native medial and cal axis of the long bones, respectively (Figs. 9.4
posterior slopes of the medial tibial plateau, as and 9.5). Because of the different objective of
well as the frontal and axial constitutional orien- alignment, the KA and MA techniques differ in
tation of the medial femoral condyle, are restored almost each stage of the implantation procedure
(Fig. 9.3). The KA technique for UKA was popu- (see the following section on surgical technique)
larised decades ago under the name “Cartier’s [8]. In clinical assessment of UKA, therefore, it
Angle” [10, 11]. Philippe Cartier developed it in is important to consider the alignment technique
the 1970s under the influence of studies by by differentiating between KA and MA position-
Christophe Lévigne and Michel Bonnin (Lyon ing of the components [8] (Fig. 9.4).
9 Kinematic Alignment Technique for Medial Unicompartmental Knee Arthroplasty 109

Fig. 9.2 This figure illustrates the “academic” definition low line). These kinematic axes dictate movement of the
of the kinematic alignment (KA) technique for medial tibia around the femur during flexion–extension move-
UKA. Implants are parallel to the cylindrical axis (green ment of the knee. (courtesy of Medacta)
line) and perpendicular to the tibial longitudinal axis (yel-

7 mm

1 mm

6.5 mm

1 mm

8.5 mm
1 mm

Fig. 9.3 This figure illustrates the “simplified” definition bone section, the sawblade line of 1 mm and 2 mm of car-
of the kinematic alignment (KA) technique for medial tilage loss. Therefore, the physiological soft tissue bal-
UKA. Positioning of the implants results in “true resurfac- ance and kinematics of the knee are restored, probably
ing” of the medial compartment of the knee. The thick- facilitating optimal clinical results. (courtesy of Medacta)
ness of the implant is equal to the total thickness of the
110 C. C. J. Rivière et al.

Fig. 9.4 This figure illustrates the two principal tech- ment of the tibia around the femur. The mechanical align-
niques for medial UKA alignment. The kinematic align- ment (MA) technique positions the components by taking
ment (KA) technique positions the prosthetic components as reference the mechanical axis of the long bones
on the kinematic axis of the knee, which dictates move-

Fig. 9.5 This X-ray composite image illustrates the during MA-UKA, resulting in a tibial implant concentrat-
radiographic aspect of medial KA-UKA (left image) and ing stress on the medial cortex during weightbearing of
MA-UKA (right image). During the KA technique, the the limb. The bony section is represented by a blue dotted
UKA components are positioned along the mechanical line
axis of the tibia and femur. Note alteration of the anatomy
9 Kinematic Alignment Technique for Medial Unicompartmental Knee Arthroplasty 111

9.3 Utility of the KA Technique is repeated by many studies which have reported
the harmful effect of a change in the anatomy of
The KA technique enables personalised and rela- the medial tibial plateau during medial UKA
tively physiological implantation as a result of implantation [16–19].
conservation of the ligament balance, knee kine-
matics and transmission of stresses to the metaph-
yseal bone. The KA technique also optimises 9.4 Scientific Evidence
dynamic interaction (i.e., throughout the knee
movement arc) between the components, thereby The KA technique for UKA has many theoretical
reducing the risk of edge loading between the advantages compared to the MA technique, most
femoral component and the polyethylene insert of which remain to be scientifically demon-
(Figs. 9.4 and 9.5). strated. This paucity of scientific evidence is
By leaving the medial compartment anatomy explained by the fact that UKA alignment tech-
of the knee intact, i.e. height and orientation of niques to date have been seldom or poorly dis-
the joint surfaces, the KA technique is biome- cussed and that no study has been designed to
chanically healthy: compare the value of KA and MA techniques [8].
The KA technique conserves tibial bone stock
• The components are aligned on the native and exerts stress on the metaphyseal tibial bone
kinematic axes of the knee, which dictate in a more physiological manner [20–22] than the
physiological movement of the tibia around MA technique. By reducing stress on the tibial
the femur. metaphyseal cortex, the KA technique may
• The natural tension of the medial collateral reduce the risk of fracture of the tibial plateau
ligament (MCL) and physiological kinematics and of secondary residual pain in remodelling
of the knee are promoted. metaphyseal bone [20–22].
• The tibial component is perpendicular to the By restoring the posterior slope of the native
subchondral trabecula whose orientation is tibial plateau, the KA technique makes it possible
dictated by Wolff’s law, adapted to mechani- to conserve the physiological tension of the
cal stress [13], and is oriented parallel to the medial collateral ligament when the knee is
ground during walking [14]. This relatively flexed. This can potentially reduce the frequency
physiological weightbearing of the tibial bone, of complications such as pain and residual stiff-
which seems able to reduce shearing stress at ness in UKA [23].
the implant–bone fixation interface, may be The KA technique enables the prosthetic com-
beneficial to the implants’ lifespan. ponents to interact optimally throughout the
knee’s arc of motion. This optimal dynamic inter-
While the KA and MA techniques for medial action of the components can reduce complica-
UKA are designed to restore constitutional align- tions related to the “edge-loading effect”, i.e.
ment of the lower limb and knee in erect posture, dislocation of the mobile polyethylene insert or
the KA technique is likely to better reproduce accelerated wear on a fixed polyethylene insert.
natural frontal alignment of the limb when the It has been demonstrated that kinematically
knee is flexed. This is the result of conservation aligned Oxford® UKA components adapt better
of the posterior tibial slope solely with the KA to supporting bone with significant reduction in
technique. Consequently, KA enables more phys- the risks of overlap or prosthetic underdimen-
iological biomechanics of the knee by reduced sioning in comparison to MA positioning [9].
alteration of the anatomy and soft tissue balance. This may be clinically beneficial by reducing the
Similar to KA-TKA, compliance with the risk of residual pain and optimising the implants’
anatomy of the knee during implantation of the lifespan.
UKA components may make it possible to obtain Lastly, it is biomechanically healthy to per-
optimal clinical results [1, 8, 15]. This approach form KA-UKA that places the tibial component
112 C. C. J. Rivière et al.

perpendicular to the subchondral trabeculae [13] patient satisfaction at 1-year follow-up were 44
and parallel to the ground during walking [14]. (median value) and 98% (unpublished data),
By reducing the shearing stress at the tibial respectively.
implant fixation interface and constraining the
metaphyseal bone in a relatively physiological
manner, the KA technique could benefit the long-­ 9.5 Surgical Technique
term results of UKA.
Many studies have reported good long-term KA for UKA is a technique with independent
results of KA-UKA: acceptable lifespan of sectioning that can be performed simply and reli-
implants, high functional performance, often nat- ably with manual instrumentation. Measuring the
ural perception of the prosthetic joint and high thickness of bone resection with vernier callipers
patient satisfaction after KA-UKA [10, 12, makes it possible to control the quality of bone
24–27]. resection, and potentially decide on bone resec-
By simulating the Oxford® medial KA-UKA tion in case of an imperfect first cut. Sophisticated
in 40 models of an OA knee, Rivière et al. [9] technological assistance with three-dimensional
observed that frontal, sagittal and axial planning (e.g. personalised surgical [31] and
­orientations of Oxford® KA components were robotic instrumentation) can also be useful.
always within the alignment range recommended It is likely that most UKA implants currently
by the Oxford group. available on the market (fixed or mobile insert,
Three radiostereometric studies have demon- metal-backed or full-polyethylene, implants with
strated that fixation of KA components is reliable resurfacing or not) are suitable for the KA tech-
given their low migration during the first 2 years nique, provided that the instrumentation so
after implantation and that the limit of 6° varus permits.
orientation for the tibial component could be rec- As previously mentioned, KA and MA surgi-
ommended [28–30]. cal techniques differ at this stage of the implanta-
Many studies have reported the harmful effect tion procedure except for axial and sagittal
of a change to the anatomy of the medial tibial rotation of the tibial and femoral components,
plateau during implantation of a UKA [16–19]. respectively (Table 9.1). The first author
A systematic review concluded in the good implanted the Oxford® UKA (Zimmer Biomet)
safety and efficacy of KA-UKA in the medium following the KA technique and using so-called
and long terms [8]. No fracture of the tibial pla- Phase 3® manual instrumentation. This rudimen-
teau and low rates of unexplained proximal tibial tary instrumentation was preferable to more
pain (0.8%), tibial implant loosening (2%) and recent Microplasty® instrumentation because the
aseptic failure of the implant (5.6%) were latter requires the operator to perform
reported in an assessment of 593 KA-UKA with MA-UKA. Please refer to video 1 illustrating the
3.2 and 12 years’ follow-up [8]. Alignment of the KA technique for implantation of an Oxford®
lower limb and tibial component was slightly in medial UKA. In video 1, you will see the “tibia
the varus position (mean values of 3° to 5°), and first” technique, followed by measured resection
the tibial component remained parallel to the of the posterior condyle, and then distal femoral
ground with the patient in the erect position. reaming to balance the space in subextension
The author has manually performed 150 (more precisely at 10° knee flexion).
Oxford® medial KA in the last 2 years. Among Tibial resection for medial Oxford® UKA
these, one patient was suffering from residual (Fig. 9.6): The posterior slope is guided by a pin
anterior knee pain and underwent revision sur- inserted in the joint space and resting on the ante-
gery (TKA); no other complications or revisions rior and posterior borders of the medial tibial pla-
were evidenced (data from the National Joint teau. The medial slope is guided by the
Registry). The Oxford Knee Scores (OKS) and anteroposterior axis of the flexion facet of the
9 Kinematic Alignment Technique for Medial Unicompartmental Knee Arthroplasty 113

Table 9.1 Different recommendations for positioning medial UKA implants between mechanical alignment (MA) and
kinematic alignment (KA) techniques. The two surgical techniques differ significantly because only flexion of the femo-
ral component and axial rotation of the tibial component follow the same recommendations
MA technique KA technique
Femoral Flexion Identical
component Frontal Perpendicular to the mechanical axis Parallel to the tibial section
section
Posterior Parallel to the MA tibial section Parallel to the KA tibial section
section
Tibial Axial rotation Identical (parallel to the lateral wall of the medial femoral condyle)
component Frontal Perpendicular to the mechanical axis of Perpendicular to the anteroposterior
orientation the tibia axis of the flexion facet of the medial
condyle
Posterior Systematic slope whose value ranges Parallel to the slope of the medial
slope between 2° and 7° depending on the tibial plateau
implant used

Fig. 9.6 This composite image shows the anatomical frontal rotation (varus–valgus) and the posterior slope of
markers used to align the tibial component during medial the tibial component, respectively. The images on the
KA-UKA. Images on the left show the anteroposterior right show the resultant tibial resection, which aims to
axis of the flexion facet of the condyle (blue lines) and a restore the medial (upper image) and posterior (lower
pin inserted in the joint space in the anterior and posterior image) tibial slopes
borders of the medial plateau. They will be used to adjust

medial condyle [32]; the tibial section has to be the thickness of the Oxford® tibial implant (mini-
perpendicular to this axis. Once performed, tibial mum 6.5 mm) after considering cartilage loss
resection is verified visually to assess the poste- (2 mm), possible bone loss (rare during isolated
rior slope (it should follow the native slope) and anteromedial wear) and thickness of the saw-
the anteroposterior and mediolateral dimensions, blade (1 mm). The tibial section is generally ori-
which should not exceed those of the tibial trial ented in slight varus during medial UKA, but this
base. The thickness of the tibial cut is measured varies by subject. Following these recommenda-
with vernier callipers; the objective is to equalise tions, the tibial component will be aligned paral-
114 C. C. J. Rivière et al.

lel to the femoral cylindrical axis and which provides information on the frontal orien-
perpendicular to the tibial longitudinal axis, tation of the medial femoral condyle. The femo-
respectively. Considering the thickness of the ral component is generally oriented in slight
resected bone measured with vernier callipers, valgus position during medial UKA, with valgus
the surgeon can determine the thickness of the varying among patients depending on the origi-
polyethylene insert necessary to restore the pre-­ nal knee anatomy (Fig. 9.7). The thickness of the
arthritis height of the medial tibial plateau joint posterior femoral section is then verified with
surface. vernier callipers; its thickness has to be equal to
Femoral resection for medial Oxford® UKA that of the prosthetic posterior condyle after con-
(Fig. 9.7): Axial and frontal rotations of the fem- sidering the thickness of the sawblade (1 mm)
oral component are determined by following the and possible cartilage loss (rare in isolated
anteroposterior axis of the flexion facet and the anteromedial wear). The last stage, which is very
medial wall (after removing medial osteophytes) simple, consists of balancing the spaces at 90°
from the medial femoral condyle, respectively and 10° flexion with slight distal femoral ream-
(Fig. 9.7). Frontal alignment can be guided by a ing. Considering that the Oxford® femoral com-
pin inserted along the medial wall of the condyle ponent is an implant with resurfacing of
(after resectioning osteophytes) between the approximately 3 mm in thickness and that the
medial collateral ligament and the condylar bone, condylar cartilage measures about 2 mm in thick-

Fig. 9.7 These composite images show the anatomical tion of the femoral component. The image on the right
markers used to align the femoral component during exe- shows the femoral height aligned on the anteroposterior
cution of medial KA-UKA. The image on the left shows axis of the flexion facet of the condyle, and a pin posi-
the anteroposterior axis of the flexion facet of the condyle tioned along the medial wall of the condyle and indicating
which can be marked by electrocauterisation; this line is the frontal orientation of the condyle
used to adjust the axial rotation and mediolateral transla-
9 Kinematic Alignment Technique for Medial Unicompartmental Knee Arthroplasty 115

ness, reaming of the distal femur should not 5. Ollivier M, Parratte S, Lunebourg A, Viehweger
remove more than 1–2 mm of subchondral bone. E, Argenson J-N. The John Insall award: no
functional benefit after unicompartmental knee
By following these stages, the femoral compo- arthroplasty performed with patient-specific instru-
nent will be aligned perpendicular to the femoral mentation: a randomized trial. Clin Orthop Relat Res.
cylindrical axis. 2016;474(1):60–8.
Thanks to quality control of bone sections, 6. Fu J, Wang Y, Li X, et al. Robot-assisted vs. con-
ventional unicompartmental knee arthroplasty:
which is performed at each stage of surgery with systematic review and meta-analysis. Orthopade.
the aid of vernier callipers, the decision on the 2018;47(12):1009–17.
polyethylene insert’s thickness can be deduced. It 7. Rivière C, Vigdorchik JM, Vendittoli P-A. Mechanical
is essential to ensure that residual laxity of alignment: the end of an era! Orthop Traumatol Surg
Res. 2019;105(7):1223–6.
1–2 mm remains when stress is exerted on the 8. Rivière C, Sivaloganathan S, Cartier P, Villet L,
flexed knee in the valgus position. It is possible to Vendittoli PA, Cobb J. Kinematic Alignment Is A
underdimension the insert by 1– 2 mm if desired Reliable Technique For Implanting medial UKA: a
to protect the lateral compartment of the knee by systematic review. KSSTA. 2020;30(3):1082–94.
9. Rivière C, Harman C, Leong A, Cobb J, Maillot
undercorrecting alignment of the limb. C. Kinematic alignment technique for medial
OXFORD UKA: an in-silico study. Orthop Traumatol
Surg Res. 2019;105(1):63–70.
9.6 Conclusion 10. Deschamps G, Chol C. Fixed-bearing unicompart-
mental knee arthroplasty. Patients’ selection and
operative technique. Orthop Traumatol Surg Res.
The kinematic alignment (KA) technique for uni- 2011;97(6):648–61.
compartmental knee arthroplasty (UKA) has 11. Cartier P. Story of my passion. Knee.
been performed successfully for decades. This 2014;21(1):349–50.
12. Cartier P, Sanouiller JL, Grelsamer
chapter helps the surgeon understand the theo- RP. Unicompartmental knee arthroplasty surgery:
retical bases and surgical principles of the tech- 10-year minimum follow-up period. J Arthroplast.
nique. The objective is to encourage surgeons to 1996;11(7):782–8.
use the KA technique because it is a simple, safe, 13. Sampath SA, Lewis S, Fosco M, Tigani D. Trabecular
orientation in the human femur and tibia and
more anatomical, more physiological and proba- the relationship with lower-limb alignment for
bly clinically advantageous method compared to patients with osteoarthritis of the knee. J Biomech.
traditional mechanical alignment for UKA. Other 2015;48(6):1214–8.
investigations are necessary to better define its 14. Asada S, Inoue S, Tsukamoto I, Mori S, Akagi
M. Obliquity of tibial component after unicompart-
clinical impact and the limits of prosthetic com- mental knee arthroplasty. Knee. 2019;26(2):410–5.
ponent alignment. 15. Rivière C, Iranpour F, Auvinet E, et al. Mechanical
alignment technique for TKA: are there intrinsic
Acknowledgements We wish to thank Medacta for pro- technical limitations? Orthop Traumatol Surg Res.
viding the images illustrating kinematic alignment of the 2017;103(7):1057–67.
MOTO® implant. 16. Chatellard R, Sauleau V, Colmar M, Robert H,
Raynaud G, Brilhault J. Medial unicompartmental
knee arthroplasty: does tibial component position
influence clinical outcomes and arthroplasty survival?
Orthop Traumatol Surg Res. 2013;99(4):S219–25.
References 17. Lee SY, Bae JH, Kim JG, et al. The influence of
surgical factors on dislocation of the meniscal bear-
1. Rivière C, Iranpour F, Auvinet E, et al. Alignment ing after Oxford medial unicompartmental knee
options for total knee arthroplasty: a systematic review. replacement: a case–control study. Bone Joint J.
Orthop Traumatol Surg Res. 2017;103(7):1047–56. 2014;96-B(7):914–22.
2. Whiteside LA. Making your next unicompartmental 18. Lo Presti M, Raspugli GF, Reale D, et al. Early fail-
knee arthroplasty last. J Arthroplast. 2005;20:2–3. ure in medial unicondylar arthroplasty: radiographic
3. Baker PN, Petheram T, Avery PJ, Gregg PJ, Deehan analysis on the importance of joint line restoration. J
DJ. Revision for unexplained pain following unicom- Knee Surg. 2019;32(09):860–5.
partmental and total knee replacement. J Bone Joint 19. Zambianchi F, Digennaro V, Giorgini A, et al.
Surg Am. 2012;94(17):e126. Surgeon’s experience influences UKA survivorship:
4. NJR 15th Annual report 2018. a comparative study between all-poly and metal
116 C. C. J. Rivière et al.

back designs. Knee Surg Sports Traumatol Arthrosc. of posterior Tibial slope after UKA. J Knee Surg.
2015;23(7):2074–80. 2019;32(05):468–74.
20. Dai X, Fang J, Jiang L, Xiong Y, Zhang M, Zhu 27. Bruni D, Akkawi I, Iacono F, et al. Minimum thick-
S. How does the inclination of the tibial component ness of all-poly tibial component unicompartmen-
matter? A three-dimensional finite element analysis tal knee arthroplasty in patients younger than 60
of medial mobile-bearing unicompartmental arthro- years does not increase revision rate for aseptic
plasty. Knee. 2018;25(3):434–44. loosening. Knee Surg Sports Traumatol Arthrosc.
21. Inoue S, Akagi M, Asada S, Mori S, Zaima H, Hashida 2013;21(11):2462–7.
M. The valgus inclination of the Tibial component 28. Soavi R, Loreti I, Bragonzoni L, La Palombara PF,
increases the risk of medial tibial condylar fractures Visani A, Marcacci M. A roentgen stereophotogram-
in unicompartmental knee arthroplasty. J Arthroplast. metric analysis of unicompartmental knee arthro-
2016;31(9):2025–30. plasty. J Arthroplast. 2002;17(5):556–61.
22. Zhu G-D, Guo W-S, Zhang Q-D, Liu Z-H, Cheng 29. Ensini A, Barbadoro P, Leardini A, Catani F, Giannini
L-M. Finite element analysis of mobile-bearing uni- S. Early migration of the cemented tibial component
compartmental knee arthroplasty: the influence of of unicompartmental knee arthroplasty: a radiostere-
Tibial component coronal alignment. Chin Med J. ometry study. Knee Surg Sports Traumatol Arthrosc.
2015;128(21):2873–8. 2013;21(11):2474–9.
23. Wahal N, Gaba S, Malhotra R, Kumar V, Pegg 30. Barbadoro P, Ensini A, Leardini A, et al. Tibial com-
EC, Pandit H. Reduced bearing excursion after ponent alignment and risk of loosening in unicom-
Mobile-­ bearing unicompartmental knee arthro- partmental knee arthroplasty: a radiographic and
plasty is ­associated with poor functional outcomes. J radiostereometric study. Knee Surg Sports Traumatol
Arthroplast. 2018;33(2):366–71. Arthrosc. 2014;22(12):3157–62.
24. Heyse TJ, Khefacha A, Fuchs-Winkelmann S, Cartier 31. Jones GG, Clarke S, Harris S, et al. A novel patient-­
P. UKA after spontaneous osteonecrosis of the knee: specific instrument design can deliver robotic level
a retrospective analysis. Arc Orthop Trauma Surg. accuracy in unicompartmental knee arthroplasty.
2011;131(5):613–7. Knee. 2019;26(6):1421–8.
25. Heyse TJ, Khefacha A, Peersman G, Cartier 32. Freeman MA, Pinskerova V. The movement
P. Survivorship of UKA in the middle-aged. Knee. of the normal tibio-femoral joint. J Biomech.
2012;19(5):585–91. 2005;38(2):197–208.
26. Franz A, Boese C, Matthies A, Leffler J, Ries
C. Mid-term clinical outcome and reconstruction
 Computer-Assisted and Robotic
Unicompartmental Knee 10
Arthroplasties

Constant Foissey, Cécile Batailler, Elvire Servien,

and Sébastien Lustig

10.1 Introduction UKA is a rare surgical procedure in France. In

2011, 9500 UKA were performed versus 70,200
The development of ancillary instruments (instru- TKR [7]. The surgical technique is demanding
mentation) for the unicompartmental knee arthro- and the operator must follow strict rules to obtain
plasties (UKA) has made it possible to improve optimal function and survival. The different
positioning of the implant and its reproducibility. parameters to be considered are as follows:
Each stage in this progression has increased the
reliability of UKA: • The tibiofemoral mechanical axis [8]
• Ligament tension
• In the early 1970s, Marmor [1] developed a • Tibial slope [9]
rudimentary instrumentation device allowing • Positioning in the sagittal plane of the femur
near-‘freehand’ placement of the implant. [10]
• In the 1980s, Cartier [2] introduced a tibial • Rotation of the implants [11]
cutting guide and a test condyle with contact • Size of the implants [12]
points. • Contact point: centring the femoral implant on
• In the 1990s, an intra-medullary rod was con- the tibia across all degrees of mobility
sidered for a femoral implant, but that under- • Restoration of the joint space [13]
mined the mini-invasive aspect of the
procedure. The latest technological advances make it pos-
• Computer-assisted navigation was developed sible to control these parameters to different
towards the end of the 1990s and quickly pre- degrees. Navigation can be performed in two dif-
sented encouraging results [3–6]. ferent ways: by navigating tibial resection alone
• Lastly, robotic-assisted surgery developed or by navigating tibial and femoral resection [14].
progressively with the introduction of The first procedure makes it possible to control
Acrobot® in 2003, MAKO® in 2015 and the first three parameters whilst the second con-
Navio®. trols the first four. Robotic assistance makes it
possible to control all these parameters.

C. Foissey · C. Batailler (*) · E. Servien · S. Lustig

Service de chirurgie orthopédique, Hôpital de la
Croix Rousse, Hospices civils de Lyon, Lyon, France

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2024 117
A. Clavé, F. Dubrana (eds.), Unicompartmental Knee Arthroplasty,
https://doi.org/10.1007/978-3-031-48332-5_10
118 C. Foissey et al.

10.2 Navigation and Medial UKA Once all parameters have been validated, the
final implants are sealed. Another control of the
10.2.1 Installation of UKA tibiofemoral mechanical axis can then be per-
with Isolated Navigation formed with the final implants.
on the Tibial Plateau This system is independent of the type of
medial UKA installed.
The system described uses OrthoPilot®
(B. Braun–Aesculap), which does not require
preoperative imaging and functions with periop- 10.2.2 Placement of a UKA
erative dynamic image acquisition of the hip, with Navigation on the Tibial
knee, ankle and visible femoral and tibial points. Plateau and Femoral Condyle
The femoral and tibial sensors are inserted
percutaneously. The approach is standard. Preoperative preparation requires radiographic
Once the markers are acquired, the axis of the measurement of two elements:
limb appears and reducibility of the deformity
can be tested. • Frontal orientation of the femoral epiphysis:
The tibial sectioning guide is positioned, and the angle between the femoral mechanical
navigation can start by adjusting varus/valgus, axis and the tangent to the femoral condyles.
the tibial slope and height of the resection by • The presumed size of the femoral component
computer-acquired data (Fig. 10.1). Once the (seen in the profile X-ray view).
correct position has been found, the sectioning
guide is attached to the bone and tibial cutting of The acquisition system allows a mini-invasive
the bone is performed with the oscillating saw. approach without cutting into the tendon inser-
The remaining cutting takes place without tion of the vastus medialis [15]. The acquisitions
navigation with a femoral cut dependent on the are similar to the technique previously described.
tibial cut, taking care not to place the condyle in The same applies to the tibial cut.
recurvatum in order to avoid conflict with the The tibiofemoral space is then measured in
patella. The navigator is again used when tests flexion and in extension. Depending on the gap in
are performed to verify the axis (177°±2°) and flexion and extension, femoral cuts are planned:
safety margin. A safety margin >1° contraindi- frontal and sagittal orientation, height of distal
cates use of a mobile plateau due to the risk of its and posterior resection, thickness of the tibial
dislocation. component, residual laxity in flexion and
extension.
Once the data are validated, a semicircular
frame is fixed under navigation control directly to
the femoral reference screw (Fig. 10.2).
All cuts are made by adjusting the different
section guides on this frame: with a saw for the
posterior tibial cut and the bevel and milling of
the femur for the distal section (Fig. 10.3).
The procedure ends as in the previous tech-
nique. The navigator is again used at the time
testing is conducted to verify the axis (177°±2°)
and safety margin. Once all parameters have been
validated, the final implants are sealed. Another
control of the tibiofemoral mechanical axis can
Fig. 10.1 Computer showing the choice of tibial section:
0° varus, 3° slope and 5-mm resection
then be performed with the final implants.
10 Computer-Assisted and Robotic Unicompartmental Knee Arthroplasties 119

Fig. 10.2 Mini-invasive implantation of a UKA:

preoperative and control views under navigation of the
femoral section guide

Various navigation systems can be used with

different systems to make the cuts. However, the
acquisition of preoperative parameters and plan-
ning are similar between each one.
Fig. 10.3 Femoral cuts

10.2.3 Results of Medial UKA These results were confirmed in 2012 [16]
Implanted with Navigation with 93.9% accuracy for HKA, 84.8% for the
mechanical tibial axis, and 100% for the poste-
In the context of medial UKA with isolated nav- rior tibial slope in 33 medial UKA surgically
igation on the tibial plateau, a 2009 study [3] implanted with assistance from the OrthoPilot®
comparing 20 navigated UKA versus 20 stan- system.
dard UKA on the accuracy of the postoperative In 2011 [17], a review of 81 medial UKA with
HKA (hip–knee–ankle) angle compared to the tibial and femoral navigation found 94% accu-
target objective (178°) found 85% accuracy racy regarding HKA, with implantation consid-
with navigation vs. 60% without navigation ered as radiologically perfect in 77% of cases and
(p < 0.05). 97% 2-year survival.
120 C. Foissey et al.

10.3 Robotic-Assisted Medial UKA

Various robotic systems can be used for

implanting a medial UKA. This involves mainly
the NAVIO® system from Smith & Nephew and
the MAKO® system from Stryker. The essential
difference between the two lies in the need to
perform a preoperative CT scan or not.

10.3.1 Surgical Technique—Navio®

System (Smith & Nephew)

The NAVIO® system (Smith & Nephew) does not

require preoperative imaging. It is based on kine-
matic preoperative image acquisitions of the hip,
knee and ankle via bone morphing and on acqui-
sition of points of interest.

10.3.1.1 Positioning of the Patient

The patient is placed in the supine position, with
a side block and distal block to maintain the knee
at 90°.
The NAVIO® PFS console consists of three
components (Fig. 10.4):

• An infrared camera placed 1 m from the area

of interest.
• A touchscreen. Fig. 10.4 The Navio® system
• A console that controls the robotic-assisted
drill connected to it by a cable.

No cutting ancillary instrument is necessary.

The only preoperative imaging is a standard
radiographic assessment.
The first stage is positioning the femoral and
tibial sensors on the skin (Fig. 10.5). These sen-
sors should be visible throughout the procedure
and for extreme amplitudes of the knee.
A medial parapatellar incision is made conven-
tionally from the upper pole of the patella to about
1 cm below the joint space over a length of about
10 cm. Arthrotomy is performed at the mid-vastus
medialis. It is important that osteophytes be
Fig. 10.5 Positioning of the patient and sensors
removed before any image acquisition in order to
have an appropriate ligament balance.
10 Computer-Assisted and Robotic Unicompartmental Knee Arthroplasties 121

10.3.1.2 Acquisition of Points recorded to finish tibial acquisition with a bone

of Interest morphing phase.
In order to ensure that the sensors are stable
throughout the procedure, a reference point is 10.3.1.3 Planning
identified in the tibia and femur, making it possi- This enables effective dynamic planning, reflect-
ble to verify at any time with a probe that the sen- ing the reducibility of the deformity.
sors have not moved. The first stage consists of choosing the femo-
The centre of the hip, ankle and knee axis of ral implant’s size, which can be modified at any
flexion are acquired by complete flexion–exten- time during planning (Fig. 10.7).
sion movement without stress in the varus/valgus
position. The same flexion–extension movement Key Points
is then performed with stress in valgus to record –– Obtain ideal bone coverage without overdi-
the reducibility of the deformity. This dynamic mensioning to avoid having an implant that
acquisition is essential because it enables the sys- overlaps, entering into conflict with the
tem to consider ligament laxity during the plan- patella, tibial spine mass or soft tissue, or on
ning stage. It is also essential that the deformity the contrary an underdimensioned implant
be reduced with moderation. that risks sinking in a secondary phase.
The points of interest are then acquired on the –– Moderately increasing flexion of the femoral
femur: the centre of the knee, the most distal implant can make it possible to obtain an
point of the medial condyle, the most posterior appropriate size with optimal bone coverage.
point and the most anterior point of the medial
condyle. Femoral acquisition continues with a The desired position for the femoral compo-
phase of bone morphing from the area of interest nent is then determined in the three spatial planes.
using the probe (Fig. 10.6). Exact positioning of the implant compared to the
The same sequence is then repeated in the shape of the femoral condyle is depicted with the
tibia: the centre of the tibia, the most distal point angular values (varus/valgus, flexion, rotation)
of the tibial cup, the most posterior point, the (Fig. 10.8). Usually, it is necessary to distalise
most medial point and the most anterior point. somewhat the femoral implant in order to com-
The anteroposterior axis of the tibia is also pensate for distal wear and to follow the bony

Fig. 10.6 Bone morphing of the femoral condyle

122 C. Foissey et al.

Fig. 10.7 Over- and underdimensioned femoral implants

–– Avoid any conflict with the patella (hence the

importance of marking the most anterior
point).

Usually, the femoral component should be as

close as possible to the intercondylar notch in
order to improve the contact points between the
femoral and tibial implants.
The same stages are then performed for the
tibial component. At the outset, the size of the
implant and thickness of the polyethylene are
decided. The positioning in varus/valgus, the tib-
ial slope, rotation and mediolateral positioning of
the implant are then chosen in line with the tibial
Fig. 10.8 Planning the positioning of the femoral
spines.
implant: 0° varus, 0° rotation, and 45° flexion

Key Points
contour posteriorly to avoid increasing posterior –– Tibial bone cutting should be minimal (usu-
femoral offset and to tighten the knee in flexion. ally 4–5 mm).
–– The tibial slope is adapted to the patient’s
Key Points anatomy; it is higher in the medial positioning
The objective is to: (equal to or less than 5°) than in the lateral
position and determines flexion, stability and
–– Preserve the height of the joint space. survival.
–– Prevent a conflict anteriorly with the tibial –– The tibial cut is usually orthogonal in the tib-
spine mass. ial mechanical axis. In cases of a major varus
10 Computer-Assisted and Robotic Unicompartmental Knee Arthroplasties 123

Fig. 10.9 Planning the overall balancing of the knee according to the positioning of the femoral and tibial implants

deformity, it is preferable to maintain a few conversely rapid failure of the implant

degrees of varus in the tibial section in order (undercorrection).
to improve the contact points between the fem- –– Importance of the ligament balance in the
oral and tibial implants. frontal plane, for which we strive to maintain
a residual safety margin balanced in flexion
The next step makes it possible to visualise the and in extension.
results of planning in terms of angular correction
(preoperative versus postoperative) between 0° The last stage in planning consists of visualis-
and 120° of flexion, as well as ligament balance ing the contact points between the two implants
(Fig. 10.9). At this stage, the positioning of the during flexion, which, if necessary, makes it pos-
tibial (varus/valgus, slope, rotation, section sible to lateralise or medialise one or both of the
height) and femoral (varus/valgus, flexion, rota- implants to better centre this contact point
tion, cut height) implants can be changed to (Fig. 10.10).
­visualise directly the effects on final angular cor-
rection and ligament balance. These parameters Key Point
consider not only static acquisition but also initial Avoid any risk of impingement responsible for
dynamic acquisition and therefore reducibility of premature wear of the polyethylene.
the deformity with each degree of flexion.
10.3.1.4 Preparation of Bone
Key Points Surfaces
–– Obtain slight undercorrection to avoid contra- Generally, we start with the femur, which is the most
lateral decompensation (overcorrection) or readily accessible. An automatic feedback system
124 C. Foissey et al.

Fig. 10.10 Adjustment of the mediolateral positioning of the implants to centre the tibiofemoral contact point

Once the femur is prepared, we move onto the

tibia, following the same procedure. It is possible
to use the most anterior part of the bone cut to
press with a saw and saw the most posterior part
of the tibia.
A rasp (grater) makes it possible to flatten the
bone sections once milling is complete. The
meniscus, readily accessible at this stage, is then
removed.
The last stage consists of milling the anchor-
ing points of the femoral implant (Fig. 10.13).

Fig. 10.11 Milling of the femoral condyle. The mill 10.3.1.5 Tests and Final Implants
automatically retracts when the operator strays outside the It is then possible to insert the test implants
planned area and to visualise onscreen the angular correc-
tion obtained, as well as the balancing on all
makes it possible to mill only the planned area. If the amplitudes of flexion (Fig. 10.14). Cementing
operator strays outside this area, the milling cutter and fixation of the final implants are done
retracts, making erroneous bone resection impossi- according to the operator’s usual practice. It is
ble in an unwanted area (Fig. 10.11). The depth of then possible to recheck the angular correction
bone to be removed is continuously visualised by and balancing of the knee with the final
changing colour (Fig. 10.12). implants.
10 Computer-Assisted and Robotic Unicompartmental Knee Arthroplasties 125

a b

Fig. 10.12 Onscreen visualisation of milled areas of bone yet to be removed (a) distal femur, (b) tibia)

Fig. 10.13 Milling of the femoral points

a b

Fig. 10.14 Positioning of the test implants (11a) and control of balancing on all joint amplitudes (11b)
126 C. Foissey et al.

10.3.2 Surgical Technique—MAKO®

System (Stryker)

The MAKO® system bases its modelling on a

preoperative CT scan. This makes it possible to
plan the positioning of the implant, as well as the
desired sizes, upstream of the procedure.

10.3.2.1 Positioning of the Patient

The patient is placed in the supine position, with
an uncumbersome lateral block (in order not to
hinder the robot’s progression) and a distal block
to maintain the knee at 90°. Fig. 10.15 Modelling of the knee by coupling with the
The MAKO® system consists of four preoperative CT scan: acquisition of specific points
components:
ative scan by precise modelling perioperatively
–– An infrared camera placed 1 m from the area the knee that has undergone surgery (Fig. 10.15).
of interest. Acquisition of ligament balance is relatively
–– A screen for the surgeon. similar to the NAVIO® system. The lower limb
–– A control console for the engineer. deformity is acquired extension, with 90° flexion
–– A robotic arm, which is positioned on the side and maximum flexion, by simulating weightbear-
to be treated with surgery, close to where the ing. Reduction of the deformity is acquired in
surgeon stands. complete extension and then every 20° to obtain
the ligament balance throughout amplitude of the
No cutting ancillary instrument is necessary. joint.
A CT scan is performed preoperatively
according to a precise protocol in order to plan 10.3.2.3 Planning
the implants’ positioning. Planning can be under- Preoperative planning makes it possible to choose
taken a few days prior to the procedure or just the most appropriate implant size in order to have
before. It will be adjusted during the procedure optimal bone coverage without overlap, and then
following acquisition of the ligament balance. to position precisely the femoral and tibial
Positioning of the femoral and tibial sensors implants in the 3D reconstruction obtained with
percutaneously is similar to that of the Navio® the CT scan (Fig. 10.16). The positioning can be
system. The incision and approach are also modified in the three spatial planes (coronal, sag-
unremarkable. Osteophytes must be removed
­ ittal and axial) for the two implants by also
after acquisition of the reference points (osteo- adjusting the heights of the bone cuts.
phytes potentially comprise areas of acquisition) Perioperatively, planning can be adjusted to
and before the ligament balance. reflect the ligament balance acquired at the start
of the procedure, as well as contact points
10.3.2.2 Acquisition of Points between the femur and tibia on the entire joint
of Interest amplitude.
Acquisition of the centre of the hip and the centre
of the knee is approximately identical to that of 10.3.2.4 Preparation of Bony Areas
the NAVIO® system. For modelling the knee, Lastly, the bone cuts will be made with a robotic
however, the CT scan makes it possible to avoid arm controlled by the surgeon with haptic
bone morphing. Therefore, the acquisition points ­feedback allowing a certain rapidity (Fig. 10.17).
performed on the femur and tibia make it possi- The robotic arm can use either a narrow saw or
ble to couple the patient’s knee with the preoper- milling cutter, as chosen by the surgeon. Once the
10 Computer-Assisted and Robotic Unicompartmental Knee Arthroplasties 127

Fig. 10.16 Planning the positioning of the implant with CT scan acquisition (a) and control of the femur–tibia contact
points perioperatively (b)
128 C. Foissey et al.

reduction in excessive tibial cuts lowers the risk

of complications in the tibial implant (postopera-
tive pain, early loosening, secondary
displacement).
Other robotic systems requiring a preoperative
scan (Acrobot® and Mako®) found similar results
[3, 11, 12]. The literature did not report any dif-
ference in terms of precision of the cuts, clinical
result or survival of implants between the
MAKO® and NAVIO® technologies [20, 21].
However, an advantage was reported in favour of
the MAKO® system in the dimensioning of the
implants. The NAVIO® system is dependent on
the acquisition of points of interest, which can
prove difficult behind the femur and tibia. Use of
the MAKO® system, coupled with CT scans,
Fig. 10.17 Robotic arm with haptic feedback enables better restitution of the posterior femoral
offset and less underdimensioning with the pos-
bone cuts have been made, tests may be con- terior part of the tibia [22]. Lastly, the MAKO®
ducted in a conventional manner. The lower limb technology has enabled faster procedure times
axis and residual laxity will be determined by the thanks to preoperative planning and the robotic
robotic system during the tests. arm [20, 21].
Robotisation significantly decreases improper
positioning and alignment errors. Consequently,
10.3.3 Results it is possible to expand certain indications and
regularly perform procedures that are known to
In a retrospective study, 80 robotic UKA versus be demanding and difficult, such as lateral UKA,
80 standard UKA with 1.5 years’ follow-up were bicompartmental arthroplasties and reconstruc-
analysed in our department [18]. We found a sig- tion of the anterior cruciate ligament (ACL) asso-
nificant decrease in aberrant values regarding the ciated with UKA.
HKA angle, tibial slope and orientation of the
tibial section, which falls within the logic of
results reproducibility. Analysis of revisions References
revealed the absence of revision due to malposi-
tion of the implants. The last advantage found 1. Marmor L. The Marmor knee replacement. Orthop
was the absence of a learning curve; none of the Clin North Am. 1982;13(1):55–64.
first implants placed by younger surgeons using 2. Cartier P, Cheaib S. Unicondylar knee arthroplasty:
2–10 years of follow-up evaluation. J Arthroplast.
robotic control presented any aberrant values. 1987;2(2):157–62.
Robotic planning readily enables acquisition of 3. Ayach A, Plaweski S, Saragaglia D. Computer-­
precise techniques prior to validation, which is assisted uni knee arthroplasty for genu varum defor-
highly valuable in a university department. mity. Results of axial correction in a case-control
study of 40 cases. In: 9th annual meeting of CAOS-­
Restitution of the joint space in the resurfac- International proceedings; 2009. p. 4–7.
ing implants was also studied [19] and a signifi- 4. Cossey AJ, Spriggins AJ. The use of computer-­
cant difference was found in favour of a robotic assisted surgical navigation to prevent malalignment
procedure with a mean decrease of 1.4 mm vs. in unicompartmental knee arthroplasty. J Arthroplast.
2005;20(1):29–34.
4.7 mm. Therefore, the femoral implant was less 5. Jenny J-Y, Boeri C. Unicompartmental knee pros-
distalised, with good restitution of the joint space thesis implantation with a non-image-based navi-
and decreased thickness of the tibial cut. This gation system: rationale, technique, case-control
10 Computer-Assisted and Robotic Unicompartmental Knee Arthroplasties 129

comparative study with a conventional instrumented 14. Jenny J-Y, Saragaglia D. Navigation informatisée
implantation. Knee Surg Sports Traumatol Arthrosc. des prothèses unicompartimentales du genou. In:
2003;11(1):40–5. Prothèses partielles de genou. ELSEVIER MASSON.
6. Jung KA, Kim SJ, Lee SC, Hwang SH, Ahn (Cahiers d’enseignement de la SOFCOT); 2012.
NK. Accuracy of implantation during computer-­ 15. Jenny J-Y. Navigated unicompartmental knee replace-
assisted minimally invasive Oxford unicompart- ment. Sports Med Arthrosc Rev. 2008;16(2):103–7.
mental knee arthroplasty: a comparison with 16. Saragaglia D, Picard F, Refaie R. Navigation of
a conventional instrumented technique. Knee. the tibial plateau alone appears to be sufficient in
2010;17(6):387–91. computer-­ assisted unicompartmental knee arthro-
7. Haute Autorité de Santé. Rapport d’évaluation - plasty. Int Orthop. 2012;36(12):2479–83.
Implants articulaires du genou. 2012. 17. Jenny J-Y, Saussac F, Louis P. Navigated, minimal
8. Lustig S, Lording T, Frank F, Debette C, Servien E, invasive, mobile bearing unicompartmental knee
Neyret P. Progression of medial osteoarthritis and prosthesis. A 2-year follow-up study. Orthopaedic
long term results of lateral unicompartmental arthro- Proc. 2012;94-B(SUPP_XXXVII):271.
plasty: 10 to 18 year follow-up of 54 consecutive 18. Batailler C, White N, Ranaldi FM, Neyret P, Servien
implants. Knee. 2014;21(Suppl 1):S26–32. E, Lustig S. Improved implant position and lower
9. Nunley RM, Nam D, Johnson SR, Barnes CL. Extreme revision rate with robotic-assisted unicompartmen-
variability in posterior slope of the proximal tibia: tal knee arthroplasty. Knee Surg Sports Traumatol
measurements on 2395 CT scans of patients undergo- Arthrosc. 2018;27:1232. https://doi.org/10.1007/
ing UKA? J Arthroplast. 2014;29(8):1677–80. s00167-­018-­5081-­5.
10. Kaya Bicer E, Servien E, Lustig S, Demey G, Ait Si 19. Herry Y, Batailler C, Lording T, Servien E, Neyret P,
Selmi T, Neyret P. Sagittal flexion angle of the femoral Lustig S. Improved joint-line restitution in unicom-
component in unicompartmental knee arthroplasty: is partmental knee arthroplasty using a robotic-assisted
it same for both medial and lateral UKAs? Knee Surg surgical technique. Int Orthop. 2017;41(11):2265–71.
Sports Traumatol Arthrosc. 2010;18(7):928–33. 20. Leelasestaporn C, Tarnpichprasert T, Arirachakaran
11. Servien E, Fary C, Lustig S, Demey G, Saffarini M, A, Kongtharvonskul J. Comparison of 1-year out-
Chomel S, et al. Tibial component rotation assessment comes between MAKO versus NAVIO robot-assisted
using CT scan in medial and lateral unicompartmen- medial UKA: nonrandomized, prospective, compara-
tal knee arthroplasty. Orthop Traumatol Surg Res. tive study. Knee Surg Relat Res. 2020;32(1):13.
2011;97(3):272–5; https://www.em-­consulte.com/en/ 21. Porcelli P, Marmotti A, Bellato E, Colombero D,
article/288184. Ferrero G, Agati G, et al. Comparing different
12. Servien E, Saffarini M, Lustig S, Chomel S, Neyret approaches in robotic-assisted surgery for unicom-
P. Lateral versus medial tibial plateau: morphometric partmental knee arthroplasty: outcomes at a short-­
analysis and adaptability with current tibial compo- term follow-up of MAKO versus NAVIO system.
nent design. Knee Surg Sports Traumatol Arthrosc. J Biol Regul Homeost Agents. 2020;34(4 Suppl.
2008;16(12):1141–5. 3):393–404. Congress of the Italian Orthopaedic
13. Weber P, Schröder C, Laubender RP, Baur-Melnyk Research Society
A, von Schulze PC, Jansson V, et al. Joint line recon- 22. Batailler C, Bordes M, Lording T, Nigues A, Servien
struction in medial unicompartmental knee arthro- E, Calliess T, et al. Improved sizing with image-based
plasty: development and validation of a measurement robotic-assisted system compared to image-free and
method. Knee Surg Sports Traumatol Arthrosc. conventional techniques in medial unicompartmental
2013;21(11):2468–73. knee arthroplasty. Bone Joint J. 2021;103-B(4):610–8.
 Full Polyethylene or Metal Back?
11
Hubert Lanternier and Arnaud Clavé

11.1 Introduction tence of the two concepts is prompting reflection

on their mechanical, biological, and surgical
Although total knee arthroplasty is still the “gold specificities and respective indications.
standard” with medium- and long-term survival
rates of 92–100% [1, 2], the good results of uni-
compartmental knee arthroplasty (UKA), partic- 11.2 The Different Types of Tibial
ularly functional, have made them an alternative Implants
choice for younger patients with high functional
demand who have unicompartmental knee osteo- The “full-polyethylene” implant (Fig. 11.1) is a
arthritis (OA) [3]. UHMWPE single block, generally with a flat
Nevertheless, implant failure on the tibial side joint surface, which will be cemented directly
remains a major cause of failure [4]. Over time, onto the tibial section. Its minimum thickness to
this has led to the development of different types limit complications due to wear and creep was
of tibial components. Currently, we can identify validated by the SOFCOT consensus in 1995 and
two major groups: was set at 9 mm.
Metal back refers to the implantation on the
–– Full-polyethylene (FPE) implants. tibial section of a metal baseplate, generally
–– Metal-back implants for which a polyethylene made of Stellite (chromium–cobalt) and more
(PE) insert can be either fixed or mobile. rarely titanium, for which bone anchoring can be
achieved with or without cementing. Similarly,
The progressive introduction of metal-back there are different manufacturer designs, includ-
implants has come to supplement the traditional, ing stemmed implant and variable number of
older “full-polyethylene” offering, and coexis- contact points (or a screw-in type), making it pos-
sible to reinforce the metal-back implant’s fixa-
H. Lanternier tion to the tibial bone. This metal-back implant is
Department of Orthopaedic Surgery and designed to receive a polyethylene insert, either
Traumatology, Polyclinique de l’Europe, fixed or mobile. For fixed plateau, the polyethyl-
Saint-Nazaire, France
ene is generally flat and noncongruent (Fig. 11.2)
A. Clavé (*) with the minimum thickness validated at the
Department of Orthopaedic Surgery and
1995 SOFCOT symposium of 6 mm and 9 mm
Traumatology, Saint George Private Hospital,
Nice, France for the screw-in format. This thickness was deter-
mined to limit the risk of wear. Using a ­metal-­back
LaTIM UMR 1101 INSERM-UBO, Brest, France

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2024 131
A. Clavé, F. Dubrana (eds.), Unicompartmental Knee Arthroplasty,
https://doi.org/10.1007/978-3-031-48332-5_11
132 H. Lanternier and A. Clavé

implant involves the thinnest possible PE to limit

the depth/height of tibial cut. To reduce stress and
therefore creep and wear on the PE, Goodfellow
had the idea in the 1970s of developing a mobile
plateau concept. The PE has become concave and
congruent with the femoral implant and has two
surfaces for sliding, one with the femoral implant
and the other with the metal-­back tibial implant
(Fig. 11.3). As the femoral implant has a spheri-
cal shape, the contact surface is optimised, theo-

Fig. 11.1 P Uni Full-Polyethylene HLS Univ-­ Fig. 11.2 P Uni Fixed Metal-Back Uni-­ highFlex™,
evolution™, Tornier (Courtesy of Tornier/Stryker) Zimmer (Courtesy of Zimmer-Biomet)

a b

Fig. 11.3 P Uni Mobile Metal-Back Oxford™, Zimmer Biomet (a) 3/4 view from the medial side (b) Frontal view
11 Full Polyethylene or Metal Back? 133

retically decreasing and better distributing economical resection while gaining in surface
stresses and shearing forces. The advantage is the area by going as close as possible to the central
possibility of using thinner PE. pivotal point.
It is then important to choose the most appro-
priate implant and fixation for this specific con-
11.3 The Problem text. The design of the tibial implant, which is
available in several sizes, should cover maximum
Tibial replacement requires a highly specific uni- surface area without entering into conflict with
compartmental implant procedure since it is nec- the capsular ligaments.
essary to resect a sufficient thickness of bone to The type of implant (full PE or metal back) in
leave room for a suitable thickness of polyethyl- contact with bone will govern the transfer of
ene (PE). To do that, we have to sacrifice the weightbearing, which we will describe.
underlying subchondral bone layer, thereby los-
ing a solid and stable support, as the underlying
cancellous bone is less dense. Nothing very dif- 11.4 Biomechanical: Stress, Wear
ferent from total knee replacement you might and Creep
think, if it were not for three major differences.
The first one lies in the surface area of support 11.4.1 Finite Element Model Parts/
provided. In the case of UKA, it is much less than Transition of Rigidity/
half of what is available for TKR since the latter Elasticity/Young’s Modulus
has the entire surface of tibial spine mass shifted
from stability to transfer of weightbearing. It is important to know the finite element model
The second specificity lies in the lesser quality parts and their mechanical characteristics [5–7].
of the bone segment offered to support the Young’s modulus (elasticity modulus) is the
UKA. In fact, a TKR can be supported by dense constant that connects the compression or trac-
bone close to cortical bone throughout the periph- tion stress and start of deformation before the
ery, while UKA is deprived of this benefit near to limit of elasticity.
the tibial spine mass since the resection here
exposes low-density cancellous bone that does –– PE, a constant element in UKA, has a Young’s
not have the primary purpose of stress transfer modulus evaluated between 0.4 and 0.7 GPa;
during compression. In summary, two thirds of it is homogeneous [7].
the peripheral bone are dense and much more –– The alloy (Cr–Co) used for metal-back
suitable for weightbearing than in the central implants is also homogenous with a much
third. The bony base of a UKA, therefore, is higher Young’s modulus of 193 GPa [5] to
much less homogeneous. 225 GPa [7].
The third difference lies in the level of liga- –– The polymethylmethacrylate cement is around
ment balance needed to maintain residual varus: 4 GPa [7].
this point can only be acquired by sufficient –– Lastly, the mechanical characteristics of the
bone resection, often deeper than in a TKR. And tibial bone chosen to receive the implant are
the problem worsens because we know that the extremely variable according to its cortical or
deeper we descend into bone, the more the sur- cancellous structure and depending on the
face area decreases and bone density subjects. Furthermore, resistance in compres-
diminishes. sion, the one that interests us, is more impor-
Therefore, tibial implantation is demanding: a tant than in distraction. The literature reports
small surface area, little homogeneity and low-­ very different figures with maximum values of
density bone. The surgeon, aware of these deli- 30 GPa for cortical bone and minimum
cate parameters, will strive to perform a precise, 0.3 GPa for cancellous bone [6, 7].
134 H. Lanternier and A. Clavé

Therefore, the metal alloy has a very high cally higher stresses in the four weightbearing
modulus while the PE, cement and cancellous positions with, depending on the area, signifi-
bone have much lower and “relatively” similar cant differences ranging from +57% to +223%.
moduli. Implant fixation must incorporate these They concluded, therefore, that mobile metal-
elements because much demand is placed on the back implants enable better distribution of
sealing interface between a rigid and less rigid stresses on the surface area of the tibia and up to
body (Table 11.1). 3 cm below the tibial implant (Fig. 11.4). The
areas of excess stress were more common and
more pronounced in the absence of metal-back
11.4.2 Stress and Strain implants, which induced both higher testing of
the bone–implant interface (cemented or not)
In 2010, Small et al. [8] studied in vitro in a which transmits the stresses and greater risk of
validated biomechanical model the effect of collapse of the underlying bone structure. These
metal backing (mobile metal-back vs. full PE localised excess stresses and their less good dis-
study) on the stress transmitted and found on the tribution to the underlying tibia, moreover, were
tibia and its subchondral bone depending on responsible for the poor results obtained by full
flexion of the knee (extension, 45° and 90° flex- PE in a clinical study by Aleto [4]. In fact, 87%
ion). For full-PE implants, they found statisti- of the revised full-PE internal UKA versus 53%

Table 11.1 Material properties assigned lo finite element model part.20, 21 Cortical and cancellous bone properties
apply to loading in compression
Model Part Elastic modulus (GPa) Poisson's ratio Elements
AP Cortical bone 16.7 0.3 105,375
Cancellous bone 0.155 0.3 93,880
PMMA cement 2.4 0.3 19,691
AP tibia 0.69 0.46 23,950
MB Cortical bone 16.7 0.3 105,375
Cancellous bone 0.155 0.3 96,340
PMMA cement 2.4 0.3 6,371
MB tibial tray (CoCr) 210 0.3 16,594
Polyethylene insert 0.69 0.46 22,313
AP all-polyethyne, MB metal-backed, GPa gigapascal, PMMA polyethylmethacrylate, CoCr cobalt chrome
Mechanical properties assigned to finite element model parts (Scott CEH et al.) [7].

a b

Fig. 11.4 Distribution and intensity of stress on the tibia depending on the type of tibial implant: (a) Full PE: stress
hotspot in posteromedial view, (b) mobile metal back: gradual distribution of intensity and distribution of stress [8]
11 Full Polyethylene or Metal Back? 135

for MB (p = 0.04) were due to collapse of the ­ etal-­back implants (2.7 MPa). This was due to
m
fixed tibial plateau [4]. more uniform stress distribution over a wider sur-
Several teams have studied the stress exerted face area of the PE with mobile metal-back
on PE and its distribution based on the type of implants (Fig. 11.5). Therefore, only these PE
tibial plateau. Simpson et al. [9] developed an had peaks of contact forces below the PE level of
experimental model enabling analysis of the resistance (17 MPa); the other models were sub-
peaks of stress and strain exerted on the four ject to stresses three times higher (Fig. 11.6).
types of tibial plateaus: mobile metal back, par- Similarly, in a study of stresses by von Mises, the
tially congruent fixed metal back, flat fixed metal peaks of stress were lower than the limit of PE
back and full PE. Their results showed that the fatigue only for mobile metal-back implants
intensity of the contact stress, identical for the (Fig. 11.7). Moreover, these peaks of stresses
fixed metal back and full PE (44.3, 48.6 and varied inversely with the thickness of the PE, and
45.9 MPa), was much lower on the mobile the authors observed that for fixed metal-back

a b c d

Fig. 11.5 Peak of intensity of contact stress (in MPa) depending on the type of plateau: (a) mobile metal back, (b)
semi-congruent fixed metal back, (c) fixed metal back, (d) full PE [9]

MBF
60
FPE
MBM
MBF partiellement congruent
50
Contact Stress (MPa)

40

Polyethylene fatigue failure stress upper limit

30

20
Polyethylene fatigue failure stress lower limit

10

0
20 25 30 35 40 45 50 55 60
Knee Flexion Angle (degrees)

Fig. 11.6 Comparison of contact stress peaks depending on the type of plateau (on y-axis flexion of the knee in
degrees, on x-axis pressure in MPa) [9]
136 H. Lanternier and A. Clavé

30

25
von Mises Stress (MPa)

20

Polyethylene fatigue failure stress lower limit

15
MBF
FPE
MBM
10
MBF partiellement congruent

0
20 25 30 35 40 45 50 55 60
Knee Flexion Angle (degrees)

Fig. 11.7 Comparison of stress peaks from von Misses depending on the type of plateau (on y-axis flexion of the knee
in degrees, on x-axis pressure in MPa) [9]

implants with PE of 8.5 mm, those in the von O’Connor [13], as well as Psychoyiosis et al. [14]
Mises study were greater than the limit of PE or Price et al. [15], reported much lower linear
fatigue, an observation that was identical with wear rates ranging from 0.01 to 0.08 mm/year for
full PE of 15 mm. These results, therefore, call mobile inserts (Oxford™, Biomet Warsaw,
into question the minimum thickness of 6 mm USA).
and 9 mm for fixed metal-back and full-PE It is interesting to note that in these two stud-
implants. On the contrary, peaks of PE stresses of ies, the difference in wear was observed less in
mobile metal-back implants of 3.5 mm and terms of volumetric wear: 17.3 mm3/year for full-
2.5 mm were less than the limit of fatigue. ­PE and fixed metal-back implants [12] and 6 to
47 mm3/year for mobile metal-back implants [13,
14]. A possible explanation is that in congruent
11.4.3 Volume and Reasons for Wear fixed plateaus, wear is distributed mainly in the
“femoral-meniscal” area of excess stress while
The main industrial and mechanical factors that for mobile plateaus, wear occurred over a larger
affect wear are type of PE, its method of sterilisa- section of the plateau and in both the “femoral-­
tion and thickness, congruence of parts and con- meniscal” and “tibial-meniscal” areas.
tact surface area [10, 11]. We note that these values are not very differ-
Several studies have examined PE wear and ent to what Wroblewski found in 1985 for
creep in vivo on specimens explanted to replace Charnley total hip replacements (16 mm3/year)
an implant. Ashraf et al. [12] reported a mean lin- [16].
ear wear rate of 0.15 mm/year for fixed plateau In vitro studies that examine PE wear in UKA
implants (full PE and fixed metal back) in 2004. are rare in the literature and find more wear for
The total mean wear rate was highly related to mobile metal-back implants than for fixed pla-
the duration of implantation. Argenson and teaus (fixed metal back or full PE) [17]. Kretzer
11 Full Polyethylene or Metal Back? 137

et al. found a mean wear rate in vitro of 10.7 mg

and 5.38 mg/106 cycles for medial and lateral
mobile metal-back implants (Univation
Mobile™, Aesculap, Germany) versus 7.51 mg
and 3.04 mg/106 cycles for the fixed metal back.
Moreover, higher generation of PE debris
occurred with mobile inserts. These statistically
significant differences led the author to conclude
in the existence of a greater risk of aseptic loos-
ening of the implant with mobile inserts [17].
Other studies have examined the characteris-
tics of PE degradation: Manson et al. in 2010
analysed the degree (subjective score) and type
of wear for three types of uni replacements, Fig. 11.9 Burnishing-­type lesions on the upper aspect of
PE in full PE, which are dominant in full-PE and fixed
Oxford™ (mobile metal back), Miller-Galante™ metal-back implants [17]
(fixed metal back) and Repicci™ (full PE),
removed during revision surgery [18]. Seven
types of wear were identified: scratching
(Fig. 11.8), pitting, burnishing (Fig. 11.9),
embedded debris, abrasion (Fig. 11.10), delami-
nation and creep. They found wear scores that
were statistically higher (and similar) for the full-
­PE (33.6) and fixed metal-back implants (33.7)
than for mobile metal-back implants (22.6). The
latter, however, presented wear on the metal-back
side of the PE with a mean score of 16.3 reinforc-
ing the notion of wear on both sides of the mobile
plateaus. Concerning the type of wear, the same
quantity of pitting and embedded debris lesion
types was found for the three groups. The full-PE Fig. 11.10 Abrasion-­type lesions on the lower aspect of
and fixed metal-back implants showed evidence the PE of mobile metal-back implants [17]
of wear induced by shearing forces (creep,
delamination and burnishing). The mobile metal-­ back implants mainly presented lesions such as
abrasion and scratching. The authors concluded
that the mobile metal-back implants showed evi-
dence of wear similar to that found in PE inserts
of THR (total hip replacement) and related to
abrasion and adhesion forces, while the reasons
for the wear of fixed metal-back and full PE are
similar to those of TKR in relation to forces and
stress of shearing and in fatigue.
In their in vitro study, Kretzer et al. systemati-
cally found on the upper sides of the mobile
metal-back and fixed metal-back implants, the
burnishing processes considered as the least
harmful wear [17]. The lower sides of the fixed
meta-back implant in contact with the metal back
Fig. 11.8 Scratching and creep-type lesions [17] presented evidence of wear manifesting as abra-
138 H. Lanternier and A. Clavé

sion and creep/scratching; these lesions were 11.5.1 Full PE

probably due to micromovements in the PE–
metal back interface (backside wear lesion). The full-PE tibial implant is cemented and fixed,
Significant lesions such as abrasion, on the and there are two interfaces, bone to cement and
­contrary, were systematically found on the lower cement to PE.
side of the mobile metal-back PE inserts [17, 19]. An experimental study by Scott [7] examined
The results of these studies, nevertheless, are the theoretical causes of mechanical disappoint-
to be interpreted with caution due to: ment with full PE to reiterate that 40% of revision
surgeries may be performed because of unex-
–– The low number of cases studied each time. plained pain, a figure that can appear disconcert-
–– PE of unequal manufacture, some old and ing for a surgeon. The study’s author, in particular,
sterilised with processes that are now impli- examined transfer of stresses to the tibial epiphy-
cated in their degradation. sis to observe that a full-PE implant generates
–– Performances specific to each implant and more abnormal stress (possibly causing pain)
probably different, irrespective of the model. than metal-back implants, the increase in PE
–– Studies conducted on specimens explanted for thickness from 6 to 10 mm decreasing but not
revision surgery and therefore in a “pathologi- cancelling out anomalies. These stresses are
cal” context. located in the anteromedial area and descend dis-
–– Or in vitro studies that cannot perfectly repro- tally while the metal base that logically ensures
duce the complexity of the knee joint their more harmonious distribution is more
kinematics. extensive in surface and less deep.
Hernigou [20] studied a series of removed
full-PE implants to confirm that although wear
11.5 The Strategy, the Options was relatively constant, creep was more pro-
nounced on parts with a lower thickness and par-
The type of implant (full PE or metal back) in ticularly in heavy patients, suggesting the relative
contact with bone, therefore, will govern the incompetence of thin and unrigid PE to transfer
transfer of stresses. The bone implantation sur- load harmoniously.
face is irregular in its mechanical characteristics, This biomechanical reflection is echoed in
has a relatively low elasticity modulus and needs clinical studies that report lower survival over
to be made weightbearing. The issue is delicate, time, particularly in heavy and active patients. In
and the complexity of the situation is confirmed a series of 1746 implants, Stefano Bini [21] found
by the frequency of bone–cement borders; two a significantly higher risk of revision surgery for
strategies can contradict each other. full PE in comparison to metal-back implants.
If the harmonious transition of elasticity is KR Berend [22] also found more cases of implant
preferred, the full-PE implant is attractive: bone, loosening with full PE in heavy patients.
cement and PE elasticity moduli are relatively
similar. However, PE, if thin [7, 17–19], is sus-
ceptible to creep deformation or creep and will 11.5.2 Metal Back
struggle to transfer the load equally to the entire
bone section. The metal-back implant is attractive in principle:
If homogenous transfer of stress to all con- it provides a rigid base that can distribute load
sidered surfaces is a priority, then very thick and over a bone surface inhomogeneously. Moreover,
less deformable PE or a metal-back implant it offers the possibility of uncemented fixation
should be prioritised. That solution is reassuring and opens the option of meniscal replacement.
but comes at the expense of a brutal transition of Nevertheless, it should be remembered that this
elasticity. strategy will multiply the interfaces and insert a
11 Full Polyethylene or Metal Back? 139

metal part that is very rigid between two ele- rearwards and makes it possible to ensure good
ments (bone and PE) which are less so. adaption before adding the PE. The introduction
The study by Scott [7] showed us that the of a full implant with integrated or full PE imme-
transmission of stresses was optimised by the diately encumbers the surgical field and hinders
metal plate. Michael Berend [23], who conducted good visibility. And it is all the more difficult
a photoelasticity study on the transmission of when the tibial element is installed with proce-
stresses, found pronounced internal stresses (par- dures directed “below and rearwards”, encourag-
ticularly posterointernal) for a full-PE implant, ing the implant to slide rearwards during its final
but he also demonstrated relatively significant impaction. With cumbersome full PE, it is diffi-
internal stresses that are more anterior with metal cult to perform unhindered vertical implantation,
backs [23]. The stresses were not cancelled out, a procedure that is somewhat easier with a metal
they were shifted: rearwards, more homogeneous back fin. This subtlety does not pose a problem
and less deep. He suggested that these stresses with TKR because implantation on a dislocated
could explain internal pain during the first 6 or subluxated tibia occurs mainly from the upper
months after the implantation of a metal-back to lower aspect.
implant, as if it was necessary to allow a stress
fracture to heal [23].
The metal-back implant with fixed plateau 11.5.3 Clinical Results
introduced a new interface, that of industrial fixa-
tion of PE on the base. A number of studies in the There are clinical studies comparing full-PE
1990s [24–26] mentioned the responsibility of implants to metal-back implants, and they pres-
micromovements that are likely to release parti- ent discrepant results. Gleeson et al. compared 47
cles that can promote the occurrence of backside metal-back mobile implants (Oxford™) and 57
wear [17–19, 27]. These reasons for wear are not full-PE implants (St Georg™ Sled) at 2 years’
the same as those resulting from contact of the follow-up and found a better functional result and
upper aspect of the PE with the femoral compo- lower number of complications (particularly
nent and may be more hazardous, particularly pain) [3]. Bhattacharya et al. in 2012 with slightly
because the trapped concentration of PE residue longer follow-up (5.6 and 3.7 years on average,
near to or in contact with bone is higher. The respectively) found the opposite results after ana-
methods of metal/PE assembly have since lysing 49 metal-back mobile (Oxford™) and 91
improved but the quality of this fixation should full-PE implants (Preservation™, DePuy, USA)
be analysed when selecting an implant. [28].
Mobile implants fall within a different logic; A meta-analysis by Smith T.O. in 2009 [29]
they are marked by PE solidity with the aim of compared mobile plateaus with fixed plateaus
decreasing wear: translation movements are (MBF and full PE) and did not find any differ-
shifted to the flat lower aspect while sliding in ence in clinical results, patient satisfaction or
flexion is allocated to the concave upper aspect, complications, simply reporting a lower fre-
which can be completely congruent with the fem- quency of tibial radiological borders with mobile
oral component. There are two technical speci- metal-back implants [29]. In a series of 144 full-
ficities to this type of implant. The first, the more PE cases (HLS Uni-Evolution™, Tornier,
­
difficult, is the need for perfect ligament balance France), however, Lustig found 26.5% of such
to ensure good implant kinematics and stability, borders starting with year one, not progressive,
which will obviously contribute to a good clinical with only a 3.5% revision rate because of loosen-
result. This is a demanding type of implant. The ing of the implant [30].
second, easier one is the quality of exposure dur- The conclusion of the lower number of clini-
ing tibial sealing: introduction of a single, rela- cal studies that have compared full-PE to metal-­
tively thin metal plateau offers better visibility back implants (and to a greater extent fixed
140 H. Lanternier and A. Clavé

plateaus to mobile plateaus) is that, currently, Full PE are the type of implants most affected
there is no definitive clinical evidence supporting by aseptic loosening, which points in the direc-
one type of plateau over another [31], although tion of studies by Simpson, Small and Aleto
one randomised study in 2015 favoured metal-­ among others on the existence of excess stresses
back implants [32]. A large number of recent with this type of plateau [4, 8, 24, 25].
studies report retrospective clinical studies. This may also account for the higher number
The vast majority of them evaluated results by of cases of unexplained pain and tibial collapse
establishing an implant survival curve and clini- found in the tibia with full PE [4, 8, 28, 37].
cal score (KSS, HSS, GIUM, Oxford Score, etc.). PE dislocations are an almost-exclusive com-
Few studies detail the different components of plication of mobile metal-back implants [3, 38].
the clinical and functional evaluation, particu- No consensus is found in the literature on the
larly the joint amplitudes. The survival rates earlier occurrence of failure depending on the
ranged from 82% to 98% at 10 years for different type of plateau. For some authors such as
types of plateaus, and no specific type seems to Bhattacharya, full PE is subject to earlier failure
provide better results, although over the very [28], but Gleeson found the opposite results in his
long-term mobile metal-back implants appear study with a higher early failure rate for mobile
superior with 91% survival at 20 years [33] ver- metal-back implants [3].
sus 85% for full-PE implants [34, 35] and 86% at The complexity of UKA revision UKA for
15 years for fixed metal-back [36] implants. The totalisation is assessed differently by the authors
functional results given by clinical scores do not [4, 18, 37, 39]. The technical difficulty, corre-
show any differences between the families. lated with the cause of failure, nevertheless
Restoration of more anatomical kinematics of the remains more often in the tibial component [4,
knee for mobile metal-back implants does not 39]. Yet one of the advantages of the full-PE
translate into a marked improvement in clinical implant lies in conservation of tibial bone stock
performance. [40]. For some authors, the higher number of
As in comparative studies, analysis of retro- tibial plateau collapses with this type of implant
spective series shows very similar results in terms requires more frequent bone grafts and reconsti-
of functional score or survival, even though they tutions, ultimately making revision surgery more
can vary within a given family. complicated [4, 37].
However, although these survival rates appear Several authors have shown a correlation
similar irrespective of the type of plateau, it between the rate of revision surgery and number
should be underlined that the authors did not all of UKA performed each year, suggesting that
use the same causes of failure in their calcula- experience tallies with results, which is espe-
tions, making a strict comparison impossible. cially true for mobile metal-back implants [41,
Traditional UKA complications include com- 42]. Mobile metal-back UKA is commonly con-
plications that are more pathognomonic, such as sidered as more technically demanding and
type of tibial plateau, and can be correlated with requires a steeper learning curve.
biomechanical processes.
OA progression in the opposite compartment
and infection are not affected by the type of tibial 11.6 Conclusion
implant [29], although this first point is some-
times debated [19]. After clarification of these factors, how can we
On the contrary, PE wear and creep are found choose?
mainly in fixed plateau implants and tended ini- The literature remains difficult to use: many
tially to involve the old-generation full-PE very large series analyse numerous criteria but
implants. This trend is declining with new full- overlook the model used [43]; others simply
­PE models for which less revision surgery due to compare uncemented implants (and therefore
wear has been found. metal-back implants) to cemented implants that
11 Full Polyethylene or Metal Back? 141

probably combine full PE and metal back but • Mobile metal-back implants enable better res-
with no precise details [21]. Moreover, full-PE toration of the knee joint’s kinematic
implants are the oldest and series on them may be presentation.
tainted by implant design, PE quality or manu-
facturing issues, which can penalise results. In terms of wear:
Most clinical series that can be used support
metal-back implants with a lower rate of revision • Mobile metal-back implants are less subject to
surgery due to loosening of the implant, particu- creep and wear than full-PE and fixed metal-­
larly in heavy and active patients [44]. Yet tran- back implants, even though volumetric wear
sient upper metaphyseal pain has been reported in studies seems to question that notion.
with both full-PE [7] and metal-back implants • The reasons for wear of full-PE and fixed
[23], perhaps because of conflicts in elasticity or metal-back implants correspond to stresses
bone stress. This point, which is sometimes and strains in shearing forces and fatigue
observed in clinical practice, is considered tem- which are similar to those of TKR, contrary to
porary and should not lead to hasty revision sur- mobile metal-back implants, where the abra-
gery for unexplained pain during year one sion events found are similar to those of THR
postoperatively [23, 41]. (total hip replacement).
The metal-back implant has a good reputation,
the survival rate seems better, it opens the door to Clinically:
cementless fixation and enables the use of a
mobile insert without cement. It should certainly • No difference was found in the functional
be preferred in heavy and active subjects. results and patients’ satisfaction rate. The
The full-PE implant also does not lack merit: mid- to long-term survival rates are similar to
its favourable elasticity modulus makes it possi- recent series.
ble to offer appropriate surgery to older, less • More complications manifesting as aseptic
active patients in the hope of minimising the loosening of the implant and collapse of the
painful events related to possible conflict in tibial plateau were found for full PE.
elasticity. • Dislocation of a PE insert represents an exclu-
It emerges from the literature that UKA is a sive complication of mobile-insert metal-back
good procedure whose conduct should be metic- implants.
ulous and will yield good, lasting results pro- • Difficulties in revision surgery for totalisation
vided that the right indication based on a clinical, are correlated with the reason for failure.
radiographic and psychological analysis has been Among these, tibial plateau collapses are a
made. The implant model chosen is clearly cause of complexity.
important, but ranks behind quality of the indica-
tion and surgical procedure.
References
Take-Home Messages
1. McCalden RW, Robert CE, Howard JL, Naudie DD,
Biomechanical level:
McAuley JP, MacDonald SJ. Comparison of outcomes
and survivorship between patients of different age
• Metal-back implants enable better distribution groups following TKA. J Arthroplast. 2013;28:83–6.
of loads and stresses in the implant/bone inter- https://doi.org/10.1016/j.arth.2013.03.034.
face and underlying tibial bone tissue. 2. Gioe TJ, Killeen KK, Hoeffel DP, Bert JM, Comfort
TK, Scheltema K, et al. Analysis of unicompartmental
• The stresses and strains exerted on PE in knee arthroplasty in a community-based implant reg-
mobile metal-back implants remain below the istry. Clin Orthop Relat Res. 2003;416:111–9. https://
level of its mechanical resistance. doi.org/10.1097/01.blo.0000093004.90435.d1.
142 H. Lanternier and A. Clavé

3. Gleeson RE, EVANS R, Ackroyd CE, WEBB J, 16. Wroblewski BM. Direction and rate of socket wear in
Newman JH. Fixed or mobile bearing unicompart- Charnley low-friction arthroplasty. J Bone Joint Surg
mental knee replacement? A comparative cohort study. (Br). 1985;67:757–61.
Knee. 2004;11:379–84. https://doi.org/10.1016/j. 17. Kretzer JP, Jakubowitz E, Reinders J, Lietz E, Moradi
knee.2004.06.006. B, Hofmann K, et al. Wear analysis of unicondylar
4. Aleto TJ, Berend ME, Ritter MA, Faris PM, Meneghini mobile bearing and fixed bearing knee systems: a
RM. Early failure of Unicompartmental knee arthro- knee simulator study. Acta Biomater. 2011;7:710–5.
plasty leading to revision. J Arthroplast. 2008;23:159– https://doi.org/10.1016/j.actbio.2010.09.031.
63. https://doi.org/10.1016/j.arth.2007.03.020. 18. Manson TT, Kelly NH, Lipman JD, Wright TM,
5. Silver FH. Biomaterials medical devices & tissue Westrich GH. Unicondylar knee retrieval analy-
engineering: an integrated approach. New York: sis. J Arthroplast. 2010;25:108–11. https://doi.
Chapmann & Hall; 1993. org/10.1016/j.arth.2010.05.004.
6. Baïetto S. Modèles visco-élastiques de remodelage 19. Kwon O-R, Kang K-T, Son J, Kwon S-K, Jo S-B, Suh
osseux. Approche théorique, numérique et expéri- D-S, et al. Biomechanical comparison of fixed- and
mentale. 2004. mobile-bearing for unicomparmental knee arthro-
7. Scott CEH, Eaton MJ, Nutton RW, Wade FA, Evans plasty using finite element analysis. J Orthop Res.
SL, Pankaj P. Metal-backed versus all-polyethylene 2014;32:338–45. https://doi.org/10.1002/jor.22499.
unicompartmental knee arthroplasty: proximal tibial 20. Hernigou P, Poignard A, Filippini P, Zilber
strain in an experimentally validated finite element S. Retrieved Unicompartmental implants with full
model. Bone Joint Res. 2017;6:22–30. https://doi. PE Tibial components: the effects of knee alignment
org/10.1302/2046-­3758.61.BJR-­2016-­0142.R1. and polyethylene thickness on creep and Wear. Open
8. Small SR, Berend ME, Ritter MA, Buckley CA, Orthop J. 2008;2:51–6. https://doi.org/10.2174/18743
Rogge RD. Metal backing significantly decreases 25000802010051.
Tibial strains in a medial Unicompartmental knee 21. Bini S, Khatod M, Cafri G, Chen Y, Paxton
arthroplasty model. J Arthroplast. 2010;26:1–6. EW. Surgeon, implant, and patient variables may
https://doi.org/10.1016/j.arth.2010.07.021. explain variability in early revision rates reported
9. Simpson DJ, Gray H, Lima DD, Murray DW, Gill for unicompartmental arthroplasty. J Bone Joint
HS. The effect of bearing congruency, thickness and Surg. 2013;95:2195–202. https://doi.org/10.2106/
alignment on the stresses in unicompartmental knee JBJS.L.01006.
replacements. Clin Biomech. 2008;23:1148–57. 22. Berend KR, Lombardi AV, Mallory TH, Adams JB,
https://doi.org/10.1016/j.clinbiomech.2008.06.001. Groseth KL. Early failure of minimally invasive
10. Engh GA, Dwyer KA, Hanes CK. Polyethylene wear unicompartmental knee arthroplasty is associated
of metal-backed tibial components in total and uni- with obesity. Clin Orthop Relat Res. 2005;440:60–6.
compartmental knee prostheses. J Bone Joint Surg https://doi.org/10.1097/01.blo.0000187062.65691.e3.
(Br). 1992;74:9–17. 23. Berend ME. State of the art in partial knee arthro-
11. Kendrick BJL, Simpson DJ, Kaptein BL, Valstar ER, plasty Mobile bearing. San Francisoc: AAOS; 2012.
Gill HS, Murray DW, et al. Polyethylene wear of 24. Parks NL, Engh GA, Topoleski LD, Emperado J. The
mobile-bearing unicompartmental knee replacement Coventry Award. Modular tibial insert micromotion.
at 20 years. J Bone Joint Surg (Br). 2011;93:470–5. A concern with contemporary knee implants. Clin
https://doi.org/10.1302/0301-­620X.93B4.25605. Orthop Relat Res. 1998;356:10–5.
12. Ashraf T, Newman JH, Desai VV, Beard D, Nevelos 25. Conditt MA, Thompson MT, Usrey MM, Ismaily
JE. Polyethylene wear in a non-congruous unicom- SK, Noble PC. Backside wear of polyethylene tibial
partmental knee replacement: a retrieval analysis. inserts: mechanism and magnitude of material loss.
Knee. 2004;11:177–81. https://doi.org/10.1016/j. J Bone Joint Surg Am. 2005;87:326–31. https://doi.
knee.2004.03.004. org/10.2106/JBJS.C.01308.
13. Argenson J-N, O'Connor JJ. Polyethylene wear 26. Steklov N, Chao N, Srivastav S. Patient-specific uni-
in meniscal knee replacement. A one to nine-year compartmental knee resurfacing arthroplasty: use
retrieval analysis of the Oxford knee. J Bone Joint of a novel interference lock to reduce tibial insert
Surg (Br). 1992;74:228–32. micromotion and backside wear. Open Biomed Eng
14. Psychoyios V, Crawford RW, O'Connor JJ, Murray J. 2010;4:156–61. https://doi.org/10.2174/187412070
DW. Wear of congruent meniscal bearings in unicom- 1004010156.
partmental knee arthroplasty: a retrieval study of 16 27. Wasielewski RC, Parks N, Williams I, Surprenant H,
specimens. J Bone Joint Surg (Br). 1998;80:976–82. Collier JP, Engh G. Tibial insert undersurface as a
15. Price AJ, Short A, Kellett C, Beard D, Gill H, Pandit contributing source of polyethylene wear debris. Clin
H, et al. Ten-year in vivo wear measurement of a fully Orthop Relat Res. 1997;345:53–9.
congruent mobile bearing unicompartmental knee 28. Bhattacharya R, Scott CEH, Morris HE, Wade F,
arthroplasty. J Bone Joint Surg (Br). 2005;87:1493–7. Nutton RW. Survivorship and patient satisfaction of
https://doi.org/10.1302/0301-­620X.87B11.16325. a fixed bearing unicompartmental knee arthroplasty
11 Full Polyethylene or Metal Back? 143

incorporating an all-polyethylene tibial component. 37. Saenz CL, McGrath MS, Marker DR, Seyler TM,
Knee. 2012;19:348–51. https://doi.org/10.1016/j. Mont MA, Bonutti PM. Early failure of a unicom-
knee.2011.04.009. partmental knee arthroplasty design with an all-­
29. Smith TO, Hing CB, Davies L, Donell ST. Fixed ver- polyethylene tibial component. Knee. 2010;17:53–6.
sus mobile bearing unicompartmental knee replace- https://doi.org/10.1016/j.knee.2009.05.007.
ment: a meta-analysis. Orthop Traumatol Surg 38. Pandit H, Hamilton TW, Jenkins C, Mellon SJ, Dodd
Res. 2009;95:599–605. https://doi.org/10.1016/j. CAF, Murray DW. The clinical outcome of minimally
otsr.2009.10.006. invasive phase 3 Oxford unicompartmental knee
30. Lustig S, Paillot J-L, Servien E, Henry J, Ait Si arthroplasty: a 15-year follow-up of 1000 UKAs.
Selmi T, Neyret P. Cemented all polyethylene tib- Bone Joint J. 2015;97-B:1493–500. https://doi.
ial insert unicompartimental knee arthroplasty: a org/10.1302/0301-­620X.97B11.35634.
long term follow-up study. Orthop Traumatol Surg 39. Whittaker J-P, Naudie DDR, McAuley JP, McCalden
Res. 2009;95:12–21. https://doi.org/10.1016/j. RW, MacDonald SJ, Bourne RB. Does bearing design
otsr.2008.04.001. influence midterm survivorship of unicompartmental
31. Confalonieri N, Manzotti A, Pullen C. Comparison of arthroplasty? Clin Orthop Relat Res. 2010;468:73–
a mobile with a fixed tibial bearing unicompartimental 81. https://doi.org/10.1007/s11999-­009-­0975-­7.
knee prosthesis: a prospective randomized trial using 40. Rouanet T, Combes A, Migaud H, Pasquier G. Do
a dedicated outcome score. Knee. 2004;11:357–62. bone loss and reconstruction procedures differ at
https://doi.org/10.1016/j.knee.2004.01.003. revision of cemented unicompartmental knee pros-
32. Hutt JRB, Farhadnia P, Massé V, LaVigne M, theses according to the use of metal-back or all-­
Vendittoli P-A. A randomised trial of all-poly- polyethylene tibial component? Orthop Traumatol
ethylene and metal-backed tibial components Surg Res. 2013;99:687–92. https://doi.org/10.1016/j.
in unicompartmental arthroplasty of the knee. otsr.2013.03.018.
Bone Joint J. 2015;97-B:786–92. https://doi. 41. Pandit H, Jenkins C, Gill HS, Barker K, Dodd CAF,
org/10.1302/0301-­620X.97B6.35433. Murray DW. Minimally invasive Oxford phase 3 uni-
33. Price AJ, Svard U. A second decade lifetable sur- compartmental knee replacement: results of 1000
vival analysis of the Oxford unicompartmental knee cases. J Bone Joint Surg (Br). 2011;93:198–204.
arthroplasty. Clin Orthop Relat Res. 2011;469:174–9. https://doi.org/10.1302/0301-­620X.93B2.25767.
https://doi.org/10.1007/s11999-­010-­1506-­2. 42. Zambianchi F, Digennaro V, Giorgini A, Grandi G,
34. Steele RG, Hutabarat S, Evans RL, Ackroyd CE, Fiacchi F, Mugnai R, et al. Surgeon's experience
Newman JH. Survivorship of the St Georg sled influences UKA survivorship: a comparative study
medial unicompartmental knee replacement beyond between all-poly and metal back designs. Knee Surg
ten years. J Bone Joint Surg (Br). 2006;88:1164–8. Sports Traumatol Arthrosc. 2015;23:2074–80. https://
https://doi.org/10.1302/0301-­620X.88B9.18044. doi.org/10.1007/s00167-­014-­2958-­9.
35. O'Rourke MR, Gardner JJ, Callaghan JJ, Liu SS, 43. Jeschke E, Gehrke T, Günster C, Hassenpflug J,
Goetz DD, Vittetoe DA, et al. The John Insall Award: Malzahn J, Niethard FU, et al. Five-year survival of
unicompartmental knee replacement: a minimum 20,946 Unicondylar knee replacements and patient
twenty-one-year followup, end-result study. Clin risk factors for failure: an analysis of German insur-
Orthop Relat Res. 2005;440:27–37. https://doi. ance data. J Bone Joint Surg. 2016;98:1691–8. https://
org/10.1097/01.blo.0000185451.96987.aa. doi.org/10.2106/JBJS.15.01060.
36. Naudie D, Guerin J, Parker DA, Bourne RB, Rorabeck 44. Argenson J-NA, Chevrol-Benkeddache Y, Aubaniac
CH. Medial unicompartmental knee arthroplasty with J-M. Modern unicompartmental knee arthroplasty
the miller-Galante prosthesis. J Bone Joint Surg Am. with cement: a three to ten-year follow-up study. J
2004;86–A:1931–5. Bone Joint Surg Am. 2002;84:2235–9.
 Recovery After Partial Knee
Arthroplasty and Daycare Surgery 12
A. Sharma, H. A. Wilson, C. O’Neill, A. Alvand,
N. Bottomley, A. J. Price, and W. F. M. Jackson

Historically, knee arthroplasty was considered a Beard et al. [4] published a pilot study in
surgical procedure that required prolonged post-­ which all patients were discharged within
operative hospitalisation. In more recent years, 24 hours from the time of surgery with no signifi-
the development of ambulatory surgery and cant complications being noted. They commented
enhanced recovery pathways across a broad that convalescence at home removes the patient
range of surgical specialties has gained consider- from the threat of hospital acquired infections,
able interest. Reported benefits include improved permits a more functional rehabilitation and the
patient satisfaction, reduced perioperative com- cost for the institution is reduced, other published
plication rates and greater cost-effectiveness. series have followed suit reporting similar
Unlike total knee arthroplasty (TKA) which findings.
involves quite an extensive surgical dissection The average reported length of stay in
and a typical 2–3 day inpatient stay, unicom- enhanced recovery programmes for UKA has
partmental knee arthroplasty (UKA) can be per- already decreased to 1 day with good results, and
formed via a more minimally invasive approach in the United States health care system UKA has
with considerably less soft tissue trauma. For been performed safely with rates of discharge of
this reason, it is a procedure ideally suited to up to 100% on the same day as surgery. [5, 6]
early discharge home after surgery and for many All published studies surrounding outpatient
patients this procedure can be done in an outpa- total joint arthroplasty from Europe have a well-­
tient setting. In addition, the literature demon- established enhanced recovery protocol in place.
strates that it allows for safe, efficient care with As a result of their investment in time and
fewer perioperative complications, which in resources, they have seen their length of stay
turn leads to higher patient satisfaction [1, 2]. gradually decrease to a point where day case
Importantly, when compared to inpatient stay arthroplasty has become feasible. The philosophy
there is no increased risk of complications or of marginal gains has been shown to provide suc-
changes in patient outcomes with similar levels cess in the field of elite sport. Similarly, over the
of anxiety and pain being experienced as the last decade, our philosophy has been to enhance
enhanced recovery group requiring inpatient patient recovery and patient satisfaction by care-
stay [ 3]. fully examining all the processes and individual
components of the patient pathway with the aim
of introducing incremental improvements to each
A. Sharma · H. A. Wilson · C. O’Neill · A. Alvand facet. We are now in a position where we have
N. Bottomley · A. J. Price · W. F. M. Jackson (*) created well defined pathways and standard
Nuffield Orthopaedic Centre, Oxford, UK

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2024 145
A. Clavé, F. Dubrana (eds.), Unicompartmental Knee Arthroplasty,
https://doi.org/10.1007/978-3-031-48332-5_12
146 A. Sharma et al.

o­ perating procedures that guide correct patient In the pre-operative phase, sedatives are
selection, reproducible anaesthetic and surgical avoided as they may impair post-operative mobil-
techniques. We can select the correct patients that isation and contribute to delays in discharge. Any
meet the criteria for safe same day discharge and premedication is intravenous as the absorption of
importantly manage the medical consequences oral analgesia may be unpredictable.
that we have created from surgery. The patient should be placed early on the the-
All patients with symptoms and radiographic atre list, aiming to complete surgery before mid-
features of end-stage single compartment tibio- day in order to maximise the likelihood of same
femoral osteoarthritis (as described by the Oxford day discharge.
group) considered suitable candidates for UKA Our anaesthetic of choice is a general anaes-
are assessed for suitability for daycase surgery. thetic (GA). If there is a strong patient preference
The inclusion criteria for discharge on the day of for spinal anaesthesia or contraindications to GA,
surgery were patients with co-morbidities that then spinal anaesthesia can be performed with or
were considered stable and their home situation without sedation. Spinal opioids are avoided.
allowed safe discharge with appropriate care. Patients undergoing GA are often supplemented
There were no arbitrary limitations such as with an adductor canal block under ultrasound
American Association of Anaesthesiologist guidance in the anaesthetic room with 20mL of
(ASA) grade, age or body mass index. 0.25% levobupivacaine. Femoral and sciatic
Within our institute, all patients listed for sur- nerve blocks are discouraged in order to avoid
gery attend a pre-operative optimisation clinic. delays in post-operative mobilisation. Our stan-
They are assessed by members of the orthopae- dard pre-operative antibiotic regime of co-­
dic, medical, anaesthetic, nursing, physiotherapy amoxiclav is administered before induction.
and occupational therapy teams allowing optimi- Further doses of antibiotics are not given as there
sation of co-existing medical conditions. is good evidence to suggest that one dose of pro-
Following medical optimisation, any patients still phylactic antibiotics is sufficient [7]. Other rou-
deemed to be at risk of unstable medical condi- tine intraoperative medications include
tions who may require more intensive post-­ intravenous tranexamic acid at induction unless
operative monitoring are excluded from the day contraindicated. We avoid topical tranexamic
of surgery discharge pathway. Furthermore, if a acid during UKA due to concerns regarding
screening questionnaire reveals that a patient potential chondrotoxicity affecting the remaining
lives alone or has any concerns about discharge compartments [8], intravenous dexamethasone,
arrangements, they can be assessed by an occu- intravenous ondansetron, intravenous
pational therapist. paracetamol and intravenous diclofenac (omitted
All patients from their initial outpatient clinic if any contraindications to NSAIDs).
and subsequent pre-operative clinic are informed Surgery is performed in the supine position
about the perioperative plan in order to manage with a thigh support and a high thigh tourniquet
expectations and reinforce the idea of day of sur- using a standard minimally invasive approach.
gery discharge. Every patient is given consistent We use the Oxford (Biomet, Warsaw, Indiana)
advice by all members of the multidisciplinary microplasty instrumentation for all patients.
team. Written information in the form of a patient Following introduction of the final trial implants,
information leaflet on daycase surgery is pro- local anaesthetic (40mL of Ropivacaine 7.5 mg/
vided to all patients, with instructions on how to mL + 0.5 mL 1:1000 adrenaline, made up to a
make preparations for same day discharge. This total volume of 100 mL with 0.9% NaCl) is infil-
principle is similarly applied on the day of sur- trated methodically into the posterior capsule,
gery when the patient is reviewed by the operat- periosteum, synovium, skin margins and quadri-
ing surgeon, anaesthetist, nursing staff and ceps using a 19-gauge spinal needle. The skin is
physiotherapist. infiltrated up to 3 cm from the margins of the
12 Recovery After Partial Knee Arthroplasty and Daycare Surgery 147

wound. The tourniquet is deflated, haemostasis post-operative day. If for social, medical or geo-
achieved, and subsequently the wound is closed graphic reasons patients are unable to go home
in layers with application of a wool and crepe on day 0, they are admitted overnight. If the
compression bandage. Drains are not used in our patient is subsequently suitable for discharge on
routine practice. day 1, they receive the same post-operative
Post-operatively, all patients are initially instructions as those patients discharged on day
recovered in main theatre recovery and 0. If the patient remains an invariant on day 2, the
promptly moved to the dedicated day case sur- compression bandage is removed, knee flexion is
gical unit. All patients are allowed to eat and commenced and they can be discharged when
drink freely. In the recovery ward, rescue anal- medically fit and mobilising safely. Outpatient
gesia is provided in the form of paracetamol, physiotherapy can be arranged to improve range
oxycodone, or an intravenous infusion of fen- of motion or mobility in selected cases but is not
tanyl if required. Morphine is avoided in order our routine practice.
to reduce sedation, nausea and vomiting. All patients discharged on day 0 or day 1
Additional doses of tranexamic acid are given return to the UKA clinic on the closest weekday
either as an inpatient or outpatient depending on the fifth post-operative day. Nurses redress the
on the time of discharge. wound, and patients have a single physiotherapy
Shortly following return to the day surgery session comprising additional gentle knee flexion
unit, an assessment of sensory and motor func- and extension exercises along with a comprehen-
tion is performed by the physiotherapy team. If sive booklet of exercises and advice. Surgical
these are adequate, static quadriceps and active team review or referral to outpatient physiother-
foot and ankle exercises are commenced. Active apy is available on the same day if any concerns
or passive knee flexion exercises are initially dis- are noted. All patients have access to physiother-
couraged. We encourage our patients to keep apy led drop-in sessions if required.
their compression bandaging intact and maintain All patients are routinely reviewed either by a
knee extension until they have returned to a des- surgeon or specialist physiotherapist at 6 weeks.
ignated UKA clinic on the fifth post-operative Following the formal introduction of the day-
day for further review. All patients are mobilised case pathway, we have continued to monitor our
fully weight bearing with crutches under the practice. In our unit we found that of all patients
supervision of the physiotherapy team. Patients presenting to clinic 73% were suitable for same
must also demonstrate the ability to safely nego- day discharge, and that we achieved same day
tiate steps or stairs prior to discharge. Post-­ discharge in 72% of these patients (118/164).
operative X-rays are obtained prior to discharge. Within the first 30 post-operative days, a total
The X-ray department prioritise daycase patients of 12 (9%) of patients who went home on the
in order to avoid delays to discharge. Patients are same day required additional assessment, with
actively educated and encouraged to rest the leg only five (4%) required readmission. Of the five
in elevation to reduce post-operative swelling and readmitted to hospital one was admitted with a
when at home to walk with the assistance of significant pulmonary embolus (PE) despite
crutches. receiving venous emboli prevention. Two patients
All patients are provided with detailed infor- were readmitted with leg swelling which were
mation on the importance of taking the prescribed investigated with ultrasound scans, which
post-operative analgesics, laxatives and anti-­ excluded deep vein thromboses, the fourth patient
emetics. In addition, patients have direct access had pain management issues following discharge,
to a 24-hour telephone helpline service for both and went on to require a manipulation under
medical and orthopaedic advice if they experi- anaesthetic (MUA) 9 weeks post-operatively, and
ence any concerns during the time period between lastly one patient required revision of the surgical
discharge and first planned review on the fifth wound with additional debridement with implant
148 A. Sharma et al.

retention (DAIR) with bearing exchange at Within our institute, introduction of a daycase
3 weeks post-operatively. Of those that were UKA pathway has provided a marked decrease in
reassessed as an outpatient one required addi- the average length of stay. The financial savings
tional physiotherapy for stiffness, one required from a safe and effective day of surgery discharge
oral antibiotics for a superficial wound infection, pathway are considerable. It has been safe, effec-
the remaining five all had minor issues that tive, and the patient satisfaction has been high.
required dressing changes or reassurance.
There was no significant difference between
the number of additional general practitioner References
(GP) appointments and Accident and Emergency
(A + E) hospital admissions, between those 1. Cleary PD, Greenfield S, Mulley AG, et al. Variations
in length of stay and outcomes for six medical and
patients that went home on the same day and surgical conditions in Massachusetts and California.
those that did not. JAMA. 1991;266:73–9.
2. Kim S, Losina E, Solomon DH, Wright J, Katz
JN. Effectiveness of clinical pathways for total
knee and total hip arthroplasty: literature review. J
12.1 Conclusions Arthroplast. 2003;18:69–74.
3. Hoorntje A, Koenraadt KLM, Boeve MG, Van Geenen
Introducing an effective outpatient arthroplasty RCI. Outpatient unicompartmental knee arthro-
protocol requires a multidisciplinary approach. It plasty: who is afraid of outpatient surgery? KSSTA.
2017;25:759–66.
is essential for all members of the team to be 4. Beard DJ, Murray DW, Rees JL, Price AJ, Dodd
engaged with the process. The whole care path- CA. Accelerated recovery for unicompartmental knee
way needs to be examined and optimised, with replacement—a feasibility study. Knee. 2002;9:221–4.
the principle of multiple small improvements to 5. Gondusky JS, Choi L, Khalaf N, et al. Day of surgery
discharge after unicompartmental knee arthroplasty:
processes leading to substantial gains in quality an effective perioperative pathway. J Arthroplast.
for the whole care episode. 2014;29:516–9.
This is a constantly evolving pathway with 6. Cross MB, Berger R. Feasibility and safety of per-
continuous audit and improvements where neces- forming outpatient unicompartmental knee arthro-
plasty. Int Orthop. 2014;38:443–7.
sary. From our experience to date, we feel the 7. Tan TL, Shohat N, Rondon AJ, et al. Perioperative
most important factors to achieve successful antibiotic prophylaxsis in Total joint arthroplasty:
same day discharge are a well-described pathway a single dose is as effective as multiple doses. JBJS
with precise “standard operating procedures”, a (AM). 2019;101(5):429–37.
8. Tuttle JR, Feltman PR, Ritterman SA, Ehrlich
consistent team message and good patient MG. Effects of tranexamic acid cytotoxicity on
education. in vitro chondrocytes. Am J Orthop (Belle Mead NJ).
As an example, our delayed knee flexion pro- 2015;44(12):E497–502.
gram [9], allows early safe mobilisation, reduces 9. Jenkins C, Jackson W, Bottomely N, Price A, Murray
D, Barker K. Introduction of an innovative day sur-
pain and swelling at 24-48rs, and has been shown gery pathway for unicompartmental knee replace-
to have no detrimental consequences at 6 weeks ment: no need for early knee flexion. Physiotherapy.
regarding range of motion. 2019;105(1):46–52.
 Utility of Bilateral Single-Stage
Unicompartmental Knee 13
Arthroplasty

Quentin Nicolas, Arnaud Clavé, Fabien Ros,

and Frédéric Dubrana

13.1 Introduction plasty has since emerged, although its appropri-

ateness remains widely debated [6–12].
Knee osteoarthritis is commonly a bilateral dis- Advocates of simultaneous procedures [9, 13–
ease [1]. Sayeed et al. [2] reported that in 26% of 17] mention the following benefits: shorter total
cases with minimal disease, patients who under- hospital stay, single anaesthesia, patient conve-
went knee replacement surgery were reoperated nience and satisfaction, and decreased cost for
on the opposite side within 5 years. the healthcare system with no increase in mor-
The expected increase in its incidence [3] bidity or mortality alongside comparable func-
combined with the legitimate desire for rapid tional results. Opponents of this approach [10,
patient recovery (particularly since the develop- 18–20] criticise its increase in perioperative com-
ment of ERAS, enhanced recovery after surgery) plications including cardio-circulatory events
and growing socioeconomic pressure have and pulmonary embolism, as well as the higher
encouraged some authors to investigate bilateral rates of blood transfusion, confusion and death.
single-stage knee arthroplasty. Chan et al. in 2009 [21] were the first to investi-
Interest in bilateral single-stage total knee gate bilateral single-stage unicompartmental
arthroplasty (SS-TKA) began in the late 1970s, (SS-UKA) knee arthroplasty. Several studies have
primarily for patients with inflammatory arthritis reported that a unicompartmental procedure enables
[4, 5]. Abundant literature on total knee arthro- faster functional recovery with a shorter hospital stay,
similar failure rate, lesser surgical trauma and lower
blood loss than TKA [22–25]. Therefore, the uni-
Q. Nicolas · F. Ros · F. Dubrana
Service de Chirurgie Orthopédique et compartmental procedure is considered as having a
Traumatologique (Department of Orthopaedic lower risk than TKA [26] and theoretically is more
Surgery and Traumatology), University Regional suitable for bilateral management. However, the lit-
Hospital Centre, Brest, France erature on this subject is scarcer [18, 21, 27–35].
A. Clavé (*) In this chapter, we will attempt to review the
Service de Chirurgie Orthopédique et utility of bilateral single-stage unicompartmental
Traumatologique (Department of Orthopaedic
Surgery and Traumatology), Clinique Saint-George, knee arthroplasty (SS-UKA) by studying:
Nice, France
Laboratoire d’Analyse et de Traitement de
• The procedure’s safety.
l’Information Médicale (LaTIM) (Laboratory of • Pain and the patient perioperative experience.
Medical Information Processing), UMR1101 • Recovery of function.
INSERM-UBO (INSERM: French National Institute • Cost.
of Health and Medical Research), Brest, France

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2024 149
A. Clavé, F. Dubrana (eds.), Unicompartmental Knee Arthroplasty,
https://doi.org/10.1007/978-3-031-48332-5_13
150 Q. Nicolas et al.

13.2 The Procedure’s Safety Since the adage “first do no harm” must pre-
vail, the evaluation of excess risk related to the
Most studies on bilateral single-stage knee procedure’s bilateral nature is the subject that
arthroplasty have involved TKR (total knee most interested the authors. We reviewed the var-
replacement), with contradictory data in the lit- ious terms associated with its evaluation in the
erature on the procedure’s safety [36, 37]. literature (see Table 13.1).
In their retrospective case–control study on 52
patients who underwent SS-UKA (104 knees) vs.
52 (unilateral) u-TKR patients, Ahn et al. [38] 13.2.1 Estimation of Blood Losses
reported fewer perioperative complications, less and Transfusion Rates
blood loss, lower transfusion rates and faster clin-
ical recovery in the SS-UKA group. In their case-control study on SS-UKA (70 knees)
Therefore, given its less invasive characteris- vs. (two stage) TS-UKA (282 knees), Berend
tics, shorter surgery, anaesthesia and hospital et al. [28] did not report any blood transfusions in
stay, faster functional recovery and lower mortal- either group.
ity rate [26, 39, 40], unicompartmental arthro- In a case-control study on SS-UKA (102
plasty seems more appropriate for single-stage knees) vs. TS-UKA (102 knees), Biazzo et al.
bilateral procedures than TKR. [18] found a significant difference in postopera-

Table 13.1 Summary table of various systemic and local intrahospital complications or at 6 months
Systemic Local
In-hospital
Major  • Death  • Mobilisation under general
 • AF anaesthesia (GA)
 • Myocardial infarction  • Slow healing
 • Diabetes  • Proximal phlebitis
 • Pulmonary embolism  • Fracture
 • Pancreatitis
Minor  • Minor cardiac dysrhythmia  • Distal phlebitis
 • Hypertension  • Sciatica
 • Dyspnoea  • Oedema
 • Asthma  • Wound bleeding
 • Intestinal ileus  • Disunion
 • Alteration of liver enzymes  • Haematoma requiring its
 • Hypotension evacuation
 • Dysuria  • Algodystrophy
At 6 months
Major  • Death  • Infection of the prosthesis
 • Pulmonary embolism  • Revision of the implant
 • Atrial fibrillation  • Mobilisation under GA
 • Jaundice  • Proximal phlebitis
 • Heart disease
Minor  • Oedema and haematoma
 • Sciatica
 • Unequal lower limb length
 • Superficial infection of the wound
scar
 • Distal phlebitis
 • Algodystrophy
13 Utility of Bilateral Single-Stage Unicompartmental Knee Arthroplasty 151

tive decrease in haemoglobin at D3 (3.1 g/dL vs. bin (2 g/dL vs. 1.3 g/dL, no indication on the day
2.4 g/dL) and transfusion rate (4 vs. 0, a transfu- of postoperative control) or the transfusion rate
sion was performed if the haemoglobin was <8 g/ (1 vs. 3, transfusion if anaemia <7 g/dL or < 10 g/
dL and in one patient with clinical signs of anae- dL in patients with heart disease).
mia). Statistical analysis of the decrease in hae- In a retrospective case-control study on
moglobin correlated with duration of surgery SS-UKA (72 knees) vs. TS-UKA (90 knees),
supports a greater decrease if surgery surpasses Tong Ma et al. [32] did not find any significant
90 min. difference in postoperative haemoglobin (10.5 g/
In their case-control prospective study on dL vs. 11.1 g/dL, no indication on the day of
SS-UKA (248 knees) vs. TS-UKA (94 knees), postoperative control) or the transfusion rate (nil
Chen et al. [29] did not find a significant differ- in both groups).
ence in fall in haemoglobin postoperatively or In their retrospective cohort study on 38
transfusion rate (−1.45 g/dL vs. -1.30 g/dL, 1 vs. SS-UKA procedures (76 knees), Akhtar et al.
0 transfusions in the control group). [27] found an average postoperative fall in hae-
In their case-control study on SS-UKA (100 moglobin of 1.8 g/dL (no indication on the day of
knees) vs. u-UKA (100 knees), Clavé et al. [30] postoperative control) and absence of postopera-
did not find any significant difference in real tive transfusion.
blood losses (465 mL vs. 396 mL), lower haemo- The current literature seems to support a
globin at D3 (10.8 g/dL vs. 11.2 g/dL) or transfu- greater decrease in haemoglobin in the simulta-
sion rate (3 vs. 7) between the 2 groups. neous procedure, without however being associ-
In a retrospective case-control study on ated with an increased transfusion rate. Duration
SS-UKA (78 knees) vs. TS-UKA (108 knees), of surgery <90 min and use of tranexamic acid
Feng et al. [31] concluded in a significant differ- more or less in combination with a preoperative
ence for decrease in haemoglobin at D3 postop- blood donation or perioperative cell-salvage pro-
eratively (2.9 g/dL vs. 0.6 g/dL) but no significant tocol for patients at risk of bleeding appear to be
difference for transfusion rate (1 vs. 0). protective factors and therefore can be advised.
For Romagnoli et al. [33], who conducted a
retrospective case-control study on SS-UKA
(382 knees) vs. 299 u-UKA, their protocol 13.2.2 Duration of Anaesthesia
included preoperative donation of packed red and Tourniquet Placement
blood cells (RBC). The authors found a signifi-
cant difference in decrease in haemoglobin (−4 g/ For Feng et al. [31], the duration of anaesthesia in
dL vs. -2.8 g/dL, measured at hospital discharge), the SS-UKA group was 120.2 mins vs. 141.6
but also in transfusions (24 vs. 13). There was no mins for the TS-UKA group (significant differ-
significant difference in autologous blood trans- ence). Chen and Tong Ma(ref) found similar
fusion in patients in the “perioperative blood results in their studies.
donation” group. It should be noted that the Akhtar et al. [27] found a mean duration of
anaesthesia protocol did not include tranexamic tourniquet placement of 83 min for a simultane-
acid in a systemic or local injection and that the ous procedure, which is similar to that in the
lower limit for transfusion was haemoglobin study by Siedlecki et al. [34].
<8 g/dL or clinical signs of anaemia. For Chan et al. [21], the duration of tourniquet
In their case-control study on SS-UKA (88 placement was 109.1 min for a simultaneous pro-
knees) vs. TS-UKA (52 knees), Siedlecki et al. cedure versus 114.86 min for a 2-stage proce-
[34] concluded that there is no significant differ- dure. The authors concluded in a nonsignificant
ence in the postoperative decrease in haemoglo- difference.
152 Q. Nicolas et al.

13.2.3 Perioperative Complications no significant difference in terms of complica-

tions was evidenced with complication rates of
Berend et al. [28] retrospectively compared the 10% and 7%, respectively.
placement of unicompartmental knee replace- Siedlecki et al. [34] found major complication
ment in 141 patients (282 knees) in sequential rates of 9.1% and 15.4%, respectively, and minor
management and 35 patients (70 knees) in simul- complication rates of 4.5% and 3.8%, respec-
taneous management. They did not find a signifi- tively (no difference between the 2 cases).
cant difference in complications and concluded Feng et al. [31], who compared the proce-
in the absence of a major complication. dure’s safety in a retrospective case-control study
Chen et al. [29] compared 124 patients (248 on 39 SS-UKA (78 knees) vs. 54 TS-UKA (108
knees) managed simultaneously and 47 (94 knees) with 42 months’ follow-up, evidenced a
knees) managed sequentially. They found five complication rate in the SS-UKA group of 10.3%
minor complications and zero major complica- versus 9.3% in the TS-UKA group; however,
tions in each group. there was no significant difference.
Romagnoli et al. [33] compared 191 SS-UKA Only the study by Chan et al. [21] found con-
(382 knees) and 299 u-UKA. The mortality rate tradictory results versus the rest of the literature.
at 2 years was similar (four deaths) with one case In their retrospective case-control study with 159
probably related to surgery in the u-UKA group. SS-UKA (318 knees) and 80 TS-UKA (160
No significant difference was found in complica- knees), there is a significant difference between
tions or revisions of the implants with at least the major complication rate in the SS-UKA group
2 years’ follow-up. (8.2%, 13 patients) and the TS-UKA group (0%,
For Tong Ma et al. [32], who compared 36 0 patients). Nine patients with VTED and one
SS-UKA (72 knees) and 45 TS-UKA (90 knees) death subsequent to a massive pulmonary embo-
with mean follow-up of 50 months, found no lism were reported. It should be specified that the
complication such as death, pulmonary embo- surgical follow-up was marked by the absence of
lism or infection of the prosthesis. Furthermore, preventive anticoagulant treatment. Furthermore,
they reported three complications in the SS-UKA they did not find any significant difference in the
group and five in the TS-UKA group with no sig- rate of minor complications (2.5% for the
nificant difference. SS-UKA group versus 3.15% for the TS-UKA
In a retrospective case-control study on 51 group) (Table 13.2).
patients with SS-UKA (102 knees) and 51
patients with TS-UKA (102 knees) by Biazzo Take-Home Message
et al. [18], no deaths, episode of confusion, The current literature, although narrow and diver-
VTED (venous thromboembolic disease) or hos- gent, suggests that simultaneous bilateral proce-
pitalisation in the ICU during the first 30 days dures are safe and do not seem to increase the
were reported. One internal tibial plateau frac- postoperative rate of complications. However, it
ture, one TIA and one slight renal impairment should be specified that all the studies mentioned
were found in the SS-UKA group and two report non-randomisation limitations and a pos-
patients with algodystrophy in the TS-UKA sible selection bias with younger patients, fewer
group. The authors concluded the absence of a comorbidities and volunteers/motivated partici-
significant difference in complications between pants in the simultaneous procedure groups.
the two groups. Therefore, it appears prudent to reiterate the util-
Clavé et al. [30] retrospectively compared 50 ity of patient selection for a simultaneous proce-
SS-UKA (100 knees) and 100 u-UKA patients; dure, at least early in experience.
 13

Table 13.2 Summary table

Number of cases (solely
Authors medial UKA) Type of arthroplasty Study design Complications Follow-up
Berend et al. [28] 70 SS-UKA vs. 282 Oxford Retrospective Rate of cardiac, pulmonary problems and 3 months
TS-UKA comparative of similar superficial infections in the two
groups
Romagnoli et al. [33] 382 SS-UKA vs. 299 Not specified Retrospective No significant difference in major or 6 months
u-UKA comparative minor complications
Tong Ma et al. [32] 72 SS-UKA vs. 90 Oxford Retrospective No significant difference 50 months
TS-UKA comparative
Chen et al. [29] 248 SS-UKA vs. 94 Not specified Prospective case-control No significant difference 2 years
TS-UKA
Clavé et al. [30] 100 SS-UKA vs. 100 Oxford Retrospective No significant difference Min. 2 years
u-UKA case-control
Siedlecki et al. [34] 84 SS-UKA vs. 52 ZUK Retrospective No significant difference in major or 17.6 months
Utility of Bilateral Single-Stage Unicompartmental Knee Arthroplasty

TS-UKA comparative minor complications

Chan et al. [21] 318 SS-UKA vs. 160 Oxford Retrospective 8.2% major complication in the SS-UKA 30 days
TS-UKA comparative group. No major complication in the
TS-UKA group.
No anticoagulant treatment.
153
154 Q. Nicolas et al.

13.3 Perioperative Pain who have investigated bilateral single-stage

and Personal Experience arthroplasty have focused on risk analysis.
(Tables 13.3, 13.4, 13.5 In a retrospective cohort study, Powell et al.
and 13.6) [15] investigated early pain in bilateral single-­
stage total knee arthroplasty and did not find any
Knee arthroplasty is associated with pain that is significant difference in the use of opioids.
considered moderate to severe, which can delay However, the VAS (visual analogue scale) scores
ambulation and the patient’s return home. Several were statistically different in the first 24 h after
studies published in the last 2 years have involved surgery. The procedure took place under general
improvement of postoperative pain management anaesthesia, with a pneumatic tourniquet and no
and ambulation after total knee arthroplasty [23, periarticular or deep nerve block local injection.
25, 37, 41, 42]. A thesis researched at Brest University
In this chapter, we will answer a practical Regional Hospital Centre involved conducting a
question, often raised by our patients: is bilateral prospective case-control study including 74
single-stage unicompartmental knee arthroplasty patients in each group (SS-UKA vs. u-UKA)
(SS-UKA) more painful than TS-UKA? (Clavé and Ros, 48).
There is a paucity of data in the literature on Their primary assessment endpoint was cumu-
early postoperative pain as the majority of authors lative analgesic use in opioid equivalents during

Table 13.3 Narcotic use mean dose equivalents (1DE = 10 mg IM morphine)

Intraoperative 0–24 24–48h 48–72h Cumulative for
postoperative postoperative postoperative first 72h
Unilateral DE 2,1±1,67 6,08±3,26 3,40±2,60 1,67±1,29 13,43±6,92
TKR (mean
±SD)
Bilateral DE 2,01±1,47 7,14±3,66 4,44±2,85 1,69±1,61 15,09±5,84
TKR (mean
±SD)
P 0.91 0.16 0.17 0.67 0.86

Table 13.4 Analog pain scor (0 = No pain, 10 = Maximun Pain)

Intraoperative 0–24 24–48h 48–72h Cumulative for
postoperative postoperative postoperative first 72h
Unilateral Pain score 4,53±1,92 4,26±1,29 4,53±1,12 3,78±1,85 4,04±1,87
TKR (maen
±SD)
Bilateral Pain score 5,83±2,21 5,49±1,69 4,12±1,45 3,93±1,79 4,11±1,40
TKR (maen
±SD)
P 0.03 0.001 0.19 0.62 0.91

Table 13.5 Mean VAS score based on postoperative time and groups
Mean VAS Control group (n=74) Case group (n=74) P
0–6 h 1,68(±0,35) 2,00(±0,37) 0.2
6–12 h 2,14(±0,37) 2,18(±0,34) 0.87
12–24 h 2,27(±0,35) 3,04(±0,41) 0.31
24–48 h 2,18(0,35) 2,57(±0,38) 0.13
48–72 1,15(±0,26) 1,42(±0,29) 0.17
Cumulative: 0–72 h 9,90(±0,99) 11,24(±1,11) 0.07
13 Utility of Bilateral Single-Stage Unicompartmental Knee Arthroplasty 155

Table 13.6 Distribution of VAS score based on postop- patients who underwent a bilateral surgical
erative time and groups procedure.
Control group No significant differences existed between the
VAS Periods (n=74) Case group (n=74) two groups concerning VAS scores in the five
H0–H6 n % n %
periods of interest or cumulative VAS scores
0–3 67 90.5 65 87.8
(H0–H72) (Table 13.5); categorical analysis of
4–6 7 9.5 9 12.2
sup 7 0 0 0 0
postoperative VAS confirmed the more painful
H6–H12 n % n % trend of the H12–H24 period but without signifi-
0–3 63 85.1 65 87.8 cance (VAS <3: 73% of u-UKA patients vs.
4–6 11 14.9 8 10.8 62.2% of SS-UKA patients) (Table 13.6).
sup7 0 0 0 0
H12–H24 n % n % Take-Home Message
0–3 54 73 46 62.2 The single-stage bilateral procedure does not
4–6 19 25.7 26 35.1 appear to be more painful or less well experi-
sup7 1 1.3 2 2.7
enced by patients than a traditional unilateral
H24–H48 n % n %
procedure.
0–3 60 81.1 54 73
4–6 13 17.6 18 24.3
sup7 1 1.3 2 2.7
H48–H72 n % n % 13.4 Functional Recovery
0–3 72 97.3 69 93.3
4–6 2 2.7 5 6.7 Most studies presenting the functional results of
sup7 0 0 0 0 single-stage bilateral knee arthroplasty also
involve TKA and find good clinical results [7, 9,
the first three postoperative days (Table 13.3). 44], even in patients over 70 years of age [45,
Surgery was performed under general anaesthe- 46].
sia supplemented by peripheral nerve block The literature contains less information on
anaesthesia of the adductors with 50 cc of functional recovery for the single-stage bilateral
ROPIVACAINE 2% and a periarticular local strategy (see Table 13.7):
injection (ROPIVACAINE 2% 100 mg, In their retrospective case-control study on
KETOPROFEN 50 mg, ADRENALINE 0.5 mg) SS-UKA (70 knees) vs. TS-UKA (282 knees)
administered in each operated knee. with mean final follow-up of 19.4 months and
The sum total of analgesic use (H0–H72) cal- 13.9 months, respectively (significant differ-
culated in opioid equivalents found in the ence), Berend et al. [28] found a significant dif-
SS-UKA group was 21.61 mg (±3.70) versus ference for the Knee Society Function Score and
19.11 mg (±3.12) in the control group. The dif- Lower Extremity Activity Score in favour of the
ference was not significant. Moreover, outside SS-UKA group (87.9 and 72.9, 11.3 and 10.2,
the H12–H24 period, use of analgesics did not respectively); the Knee Society Pain Scores and
differ between the two groups (Table 13.3). These Knee Society Clinical Scores are similar in the
results on analgesic use, moreover, are consistent two groups. However, the groups were not homo-
with the literature. Essving et al. [43] in a similar geneous with younger patients, a lower BMI
setting (unilateral medial infiltration analgesia (body mass index) and more favourable Knee
UKA) found analgesic use of 20 mg (±30 mg) in Society Clinical Score for the SS-UKA group,
opioid equivalents. which could result in a selection bias.
The authors explained the difference in anal- In their prospective case-control study on
gesic use for the H12–H24 period by the progres- SS-UKA (248 knees) vs. TS-UKA (94 knees)
sive regression of the combined local anaesthetics with final follow-up of 2 years, Chen et al. [29]
and deep nerve block, potentially more painful in did not find any significant difference for OKS
156 Q. Nicolas et al.

Table 13.7 Functional recovery/score for single-stage bilateral UKA

Control group:
Case group: two-stage strategy
one-stage (TS-UKA) or
Study Functional simultaneous unilateral surgery
name score strategy (u-UKA) p value
Preop score Preop score
Postop score Postop score
(mean duration (mean duration
of follow-up in of follow-up in
months) months)
Berend Knee society 46 38 <0.0001
et al. [28] clinical score
91.4 (19.4) 90.1 (13.9) NS
(0.0013)
Knee society 11.6 9.5 NS
pain score
44.6 (19.4) 46.8 (13.9) NS
(0.0013)
Knee society 58.9 55.6 NS
function
scores
87.9 (19.4) 72.9 (13.9) <0.0001
(0.0013)
Lower 11.3 (19.4) 10.2 (13.9) <0.0001
extremity (0.0013)
activity score
Chen Oxford knee 34 29 0.001
et al. [29] score
18 (6) 18 (6) NS (NS)
17 (24) 16 (24) NS (NS)
Knee society 44.5 47 NS
knee score
88 (6) 88 (6) NS (NS)
90 (24) 90 (24) NS (NS)
Clavé OKS 27.5 25.2 NS
et al. [30]
41.8 (6) 40.5 (6) NS (NS)
44.5 (44.4) 42.2 (61.2) NS (?)
KOOS 56.7 52.9
85.28 (6) 84.1 (6) NS (NS)
91.8 (44.4) 87.9 (61.2) NS (NS)
Feng KSS scores 115 115 NS
et al. [31]
170 (12) 167.5 (12) NS (NS)
Tong Ma OKS 40.8 40.5 NS
et al. [32]
25.1 (1) 23.2 (1) NS (NS)
20 (3) 19.7 (3) NS (NS)
19.2 (6) 18.8 (6) NS (NS)
18.3 (50) 18 (50) NS (NS)
Ros and OKS 36.59 38.33 0.04
Clavé
et al. [35]
13 Utility of Bilateral Single-Stage Unicompartmental Knee Arthroplasty 157

Table 13.7 (continued)

Control group:
Case group: two-stage strategy
one-stage (TS-UKA) or
Study Functional simultaneous unilateral surgery
name score strategy (u-UKA) p value
Preop score Preop score
Postop score Postop score
(mean duration (mean duration
of follow-up in of follow-up in
months) months)
46.91 (6) 47.31 (6) NS (NS)
49.47 (12) 48.89 (12) NS (NS)
Delta OKS 10.5 12.9 0.03
preop-M12

functional scores, Knee Society Function Scores with final follow-up of 1 year, Feng et al. [31] did
or Knee Society Knee Scores. Several biases not find any significant difference for the KSS
existed in this study. The first is a selection bias score. It is possible that a selection bias exists
since patients in the SS-UKA group have higher since patients in the SS-UKA group were younger
preoperative functional scores than those in the and in better health than in most of these studies.
TS-UKA group. The second is a type-1 error bias In a prospective case-control study on
since there was no randomisation, this being SS-UKA (74 patients, 148 knees) vs. u-UKA (74
potentially reinforced by the inclusion of “more” patients) with final follow-up at 1 year, Ros et al.
motivated patients in the SS-UKA group. [35] (article in press) found a significant differ-
In their case-control study with a control ence in favour of the case group in the analysis of
group based on a prospective matched series of OKS gain between preoperative status and M12.
SS-UKA (100 knees) vs. u-UKA (100 knees) Given the absence of randomisation, a type-1
with mean follow-up of 3.7 years and 5.1 years, error remains possible, however.
respectively, Clavé et al. [30] did not find any sig-
nificant difference for the OKS, KOOS and IKS
functional scores. The two groups, however, do 13.4.1 Patient Satisfaction
not have the same inclusion periods (maximum
difference of 10 years); therefore, a bias in expe- For Clavé et al. [30], 96% (48/50) of patients in
rience remains possible. The absence of randomi- the SS-UKA group recommended this procedure
sation, moreover, can also introduce a selection with an excellent satisfaction rate of 74% at last
bias. follow-up (3.7 years on average) versus 94% and
In their retrospective case-control study on 36 79% for the control group (unilateral procedure
SS-UKA (72 knees) and 45 TS-UKA (90 knees) with 5.1 years of follow-up on average). They did
with mean follow-up of 50 months, Tong Ma not find any significant difference between the
et al. [32] did not find any significant difference groups. Results were similar in the Clavé and Ros
in OKS score. A selection bias was noted with, in et al. study.
both groups, patients who were younger (mean
age 65 years) and in “better” health (80% of the
population < ASA II, mean BMI 25) than can be 13.4.2 Early Functional Recovery
found in comparable series.
In a 2019 retrospective case-control study on In their case group (SS-UKA), Clavé and Ros
SS-UKA (78 knees) vs. TS-UKA (108 knees) et al. [35] found 62.2% of patients who made
158 Q. Nicolas et al.

their first round trip at 24 h post-surgery and however, duration of rehabilitation was longer for
100% at 3 days versus 68.9% and 100% for the the SS-UKA group (9.2 vs. 7.8 days).
control group (nonsignificant difference). Feng et al. [31] in 2019 also found a longer
Concerning ascent and descent of 8 steps, total duration of hospital stay in the TS-UKA
Clavé and Ros et al. [35] did not find any signifi- group (7.5 vs. 4.2 days); these figures are consis-
cant difference between the 2 groups, with 33.8% tent with those of Siedlecki.
of patients able to ascend a stairway at 24 h For Clavé and Ros et al. [35], the medical
postop in the control group versus 24.3% in the duration of stay was evaluated based on criteria
case group and 100% in both groups at D3. for hospital discharge so as not to be biased by
These data on early postoperative resumption delay in discharge for administrative or nonmedi-
of function could not be compared to those in the cal reasons. Therefore, based on their study,
literature because, as far as we know, no other 36.5% of patients were discharged at 24 h post-
study has been conducted on this subject. operatively, 92% at 48 h and 100% at 72 h in the
control group compared to 27% at 24 h, 89.2% at
Take-Home Message 48 h and 100% at 72 h in the case group. The
Here too, the literature on single-stage bilateral authors concluded in the absence of a significant
UKA provides less information compared to difference.
TKA. However, studies all converge in the same
direction, with clinical scores and patient satis-
faction as good as or even better than for unilat- 13.5.2 Cost of Hospitalisation
eral UKA or two-stage strategies (TS-UKA).
Immediate postoperative physical rehabilitation For Siedlecki et al. [34], cost of hospitalisation
does not seem to be impacted by the bilateral was significantly higher in the TS-UKA group
nature of the procedure. These results neverthe- (€11,766.7 vs. €5626.4).
less remain subject to criticism since the studies For Chen et al. [29], the hospital stay savings
published have several biases, particularly were 8892 USD for the SS-UKA group.
younger, more motivated and healthier cohorts. For Feng et al. [31] in 2019, the cost of hospi-
tal stay in the SS-UKA group was 11,294.2 USD
versus 12,846 USD for the TS-UKA group; the
13.5 Cost difference was lower but nevertheless
significant.
13.5.1 Duration of Stay It should be noted that none of these studies
assessed the financial impact on outpatient care
For Siedlecki et al. [34], total mean hospital stay (nursing, rehabilitation, cost of medical treat-
was 6.7 days for the SS-UKA group vs. 13.9 days ments at home) or societal costs, particularly
for the TS-UKA group (significant difference). regarding early return to work. There is also no
For Chen et al. [29], total hospital stay was study on the cost of possible increased morbidity
significantly shorter by 3 days in the SS-UKA even though studies on the risks of these proce-
group (5 vs. 8 days). dures are reassuring in nature.
Akhtar et al. [27] found a duration of hospital
stay of 3.5 days for the simultaneous procedure Take-Home Message
versus 2 days for a unilateral procedure. They In the current literature, bilateral procedures
concluded in a reduced hospital stay with two-­ appear economically preferable. Therefore, it is
stage bilateral procedures. very important that health authorities take that
For Romagnoli et al. [33], no significant dif- into account and remove the current financial
ference was found for duration of hospital stay regulatory restraints in order to facilitate use of
between the SS-UKA group and u-UKA group; simultaneous bilateral procedures.
13 Utility of Bilateral Single-Stage Unicompartmental Knee Arthroplasty 159

13.6 Conclusion—Take-Home 7. Forster MC, Bauze AJ, Bailie AG, Falworth MS,
Oakeshott RD. A retrospective comparative study of
Message bilateral total knee replacement staged at a one-week
interval. J Bone Joint Surg Br. 2006;88-B(8):1006–10.
Bilateral single-stage unicompartmental knee 8. Husted H, Troelsen A, Otte KS, Kristensen BB, Holm
arthroplasty (SS-UKA), in light of the current lit- G, Kehlet H. Fast-track surgery for bilateral total knee
replacement. J Bone Joint Surg. 2011;93(3):6.
erature, seems to be a safe procedure for which 9. Kim Y-H, Choi Y-W, Kim J-S. Simultaneous bilateral
perioperative patient experience is good, particu- sequential total knee replacement is as safe as uni-
larly thanks to good management of analgesia lateral total knee replacement. J Bone Joint Surg Br.
and pain that is ultimately well tolerated. 2009;91-B(1):64–8.
10. Oakes DA, Hanssen AD. Bilateral total knee replace-
Furthermore, immediate postoperative rehabilita- ment using the same anesthetic is not justified by
tion does not seem to be impacted by the bilateral assessment of the risks. Clin Orthopa Relat Res.
aspect of the procedure. Functional scores, more- 2004;428(87):91.
over, are identical in the medium and long terms. 11. Parvizi J, Sullivan TA, Trousdale RT, Lewallen
DG. Thirty-day mortality after total knee arthroplasty.
For all studies that examined the subject, costs J Bone Joint Surg Am. 2001;83(8):1157–61.
for the healthcare system are decreased. However, 12. Ritter MA, Meding JB. Bilateral simultaneous total
current fee schedules imposed by France’s health knee arthroplasty. J Arthroplast. 1987;2(3):185–9.
authorities are such that the procedures incur a 13. Hutchinson JRM, Parish EN, Cross MJ. A compari-
son of bilateral uncemented total knee arthroplasty:
loss for healthcare facilities and surgeons, pre- simultaneous or staged? J Bone Joint Surg Br.
venting the study and promotion of these proce- 2006;88-B(1):40–3.
dures on a larger scale. 14. Leonard L, Williamson DM, Ivory JP, Jennison C. An
Nevertheless, it is appropriate while awaiting evaluation of the safety and efficacy of simultane-
ous bilateral total knee arthroplasty. J Arthroplast.
larger randomised studies to remain reasonable 2003;18(8):972–8.
in terms of the indications and to perform rigor- 15. Powell RS, Pulido P, Tuason MS, Colwell CW,
ous patient screening and selection. Ezzet KA. Bilateral vs unilateral Total knee arthro-
plasty: a patient-based comparison of pain levels
and recovery of ambulatory skills. J Arthroplast.
2006;21(5):642–9.
References 16. Shetty GM, Mullaji A, Bhayde S, Chandra Vadapalli
R, Desai D. Simultaneous bilateral versus unilateral
1. Andersson G, Academy A, of Orthopaedic Surgeons. computer-assisted total knee arthroplasty: a prospec-
The burden of musculoskeletal diseases in the United tive comparison of early postoperative pain and func-
States: prevalence, societal and economic cost. tional recovery. Knee. 2010;17(3):191–5.
Rosemont, IL: American Academy of Orthopaedic 17. Yoon H-S, Han C-D, Yang I-H. Comparison of simul-
Surgeons; 2008. taneous bilateral and staged bilateral Total knee
2. Sayeed SA, Sayeed YA, Barnes SA, Pagnano MW, arthroplasty in terms of perioperative complications.
Trousdale RT. The risk of subsequent joint arthro- J Arthroplast. 2010;25(2):179–85.
plasty after primary unilateral Total knee arthroplasty, 18. Biazzo A, Masia F, Verde F. Bilateral unicompart-
a 10-year study. The Journal of Arthroplasty Sept. mental knee arthroplasty: one stage or two stages?
2011;26(6):842–6. Musculoskelet Surg. 2019;103(3):231–6.
3. Kurtz S, Ong K, Lau E, Mowat F, Halpern 19. Lombardi AV, Mallory TH, Fada RA, Hartman JF,
M. Projections of primary and revision hip and knee Capps SG, Kefauver CA, et al. Simultaneous bilateral
arthroplasty in the United States from 2005 to 2030. J Total knee arthroplasties: who decides? Clin Orthop
Bone Joint Surg. 2007;89(4):780–5. Relat Res. 2001;392(319):29.
4. Gradillas EL, Volz RG. Bilateral Total knee replace- 20. Memtsoudis SG, González Della Valle A, Besculides
ment under one anesthetic. Clin Orthop Relat Res. MC, Gaber L, Sculco TP. In-hospital complications
1979;(140):153–8. and mortality of unilateral, bilateral, and revision
5. Head WC, Paradies LH. Ipsilateral hip and knee TKA: based on an estimate of 4,159,661 discharges.
replacements as a single surgical procedure. J Bone Clin Orthop Relat Res. 2008;466(11):2617–27.
Joint Surg Am. 1977;59(3):352–4. 21. Chan WCW, Musonda P, Cooper AS, Glasgow MMS,
6. Bullock DP, Sporer SM, Shirreffs TG. Comparison Donell ST, Walton NP. One-stage versus two-stage
of simultaneous bilateral with unilateral total knee bilateral unicompartmental knee replacement: a com-
arthroplasty in terms of perioperative complications. parison of immediate post-operative complications. J
J Bone Joint Surg Am. 2003;85(10):1981–6. Bone Joint Surg Br. 2009;(10):91-B, 1305–9.
160 Q. Nicolas et al.

22. Beard DJ, Davies LJ, Cook JA, MacLennan G, Price 34. Siedlecki C, Beaufils P, Lemaire B, Pujol
A, Kent S, et al. The clinical and cost-effectiveness of N. Complications and cost of single-stage vs. two-­
total versus partial knee replacement in patients with stage bilateral unicompartmental knee arthroplasty:
medial compartment osteoarthritis (TOPKAT): 5-year a case-control study. Orthop Traumatol Surg Res.
outcomes of a randomised controlled trial. Lancet. 2018;104(7):949–53.
2019;394(10200):746–56. 35. Ros Fabien, These pour le diplôme d’etat de docteur
23. Lombardi AV, Berend KR, Walter CA, Aziz-Jacobo en medecine, des de chirurgie generale, universite de
J, Cheney NA. Is recovery faster for mobile-bearing medecine de brest-bretagne occidentale, Prothèse uni-­
unicompartmental than total knee arthroplasty? Clin compartimente de genou unilatérale vs bilatérale en
Orthop Relat Res. 2009;467(6):1450–7. un temps opératoire: étude cas-témoin du vécu post-­
24. Newman J, Pydisetty RV, Ackroyd opératoire et des résultats fonctionnels à court terme,
C. Unicompartmental or total knee replacement. J 2018.
Bone Joint Surg. 2009;91(1):6. 36. Liu L, Liu H, Zhang H, Song J, Zhang L. Bilateral
25. Price AJ, Waite JC, Svard U. Long-term clinical total knee arthroplasty: simultaneous or staged?
results of the medial Oxford Unicompartmental knee A systematic review and meta-analysis. Medicine.
arthroplasty. Clin Orthop. 2005;435:10. 2019;98(22):e15931.
26. Liddle AD, Judge A, Pandit H, Murray DW. Adverse 37. Morrey BF. Safety of simultaneous bilateral Total
outcomes after total and unicompartmental knee knee arthroplasty: a meta-analysis. Yearbook of
replacement in 101 330 matched patients: a study of Orthopedics. 2008;2008(105):6.
data from the National Joint Registry for England and 38. Ahn JH, Kang DM, Choi KJ. Bilateral simultaneous
Wales. Lancet. 2014;384(9952):1437–45. unicompartmental knee arthroplasty versus unilateral
27. Akhtar KSN, Somashekar N, Willis-Owen CA, total knee arthroplasty: a comparison of the amount
Houlihan-Burne DG. Clinical outcomes of bilateral of blood loss and transfusion, perioperative complica-
single-stage unicompartmental knee arthroplasty. tions, hospital stay, and functional recovery. Orthop
Knee Jan. 2014;21(1):310–314. Traumatol Surg Res. 2017;103(7):1041–5.
28. Berend KR, Morris MJ, Skeels MD, Lombardi 39. Arirachakaran A, Choowit P, Putananon C, Muangsiri
AV, Adams JB. Perioperative complications of S, Kongtharvonskul J. Is unicompartmental knee
simultaneous versus staged Unicompartmental arthroplasty (UKA) superior to total knee arthro-
knee arthroplasty. Clin Orthop Relat Res. plasty (TKA)? A systematic review and meta-­
2011;469(1):168–73. analysis of randomized controlled trial. Eur J Orthop
29. Chen JY, Lo NN, Jiang L, Chong HC, Tay DKJ, Surg Traumatol. 2015;25(5):799–806. https://doi.
Chin PL, et al. Simultaneous versus staged bilateral org/10.1007/s00590-­015-­1610-­9.
unicompartmental knee replacement. Bone Joint J. 40. Dalury DF, Fisher DA, Adams MJ, Gonzales
2013;95-B(6):788–92. RA. Unicompartmental knee arthroplasty compares
30. Clavé A, Gauthier E, Nagra NS, Fazilleau F, Le Sant favorably to total knee arthroplasty in the same
A, Dubrana F. Single-stage bilateral medial Oxford patient. Orthopedics. 2009;32(4):1213–26. orthosu-
Unicompartmental knee arthroplasty: a case-control persite.com/view.asp?rID=38057
study of perioperative blood loss, complications 41. Carlsson LV, Albrektsson BEJ, Regnér LR. Minimally
and functional results. Orthop Traumatol Surg Res. invasive surgery vs conventional exposure using the
2018;104(7):943–7. miller-Galante Unicompartmental knee arthroplasty. J
31. Feng S, Yang Z, Sun J-N, Zhu L, Wang S, Guo K-J, Arthroplast. 2006;21(2):151–6.
et al. Comparison of the therapeutic effect between 42. Reilly KA, Beard DJ, Barker KL, Dodd CAF, Price
the simultaneous and staged unicompartmental knee AJ, Murray DW. Efficacy of an accelerated recov-
arthroplasty (UKA) for bilateral knee medial com- ery protocol for Oxford unicompartmental knee
partment arthritis. BMC Musculoskelet Disord. arthroplasty—a randomised controlled trial. Knee.
2019;20(1):340. 2005;12(5):351–7.
32. Ma T, Tu Y-H, Xue H-M, Wen T, Cai M-W. Clinical 43. Essving P, Axelsson K, Otterborg L, Spännar H, Gupta
outcomes and risks of single-stage bilateral A, Magnuson A, et al. Minimally invasive surgery
Unicompartmental knee arthroplasty via Oxford did not improve outcome compared to conventional
phase III. Chin Med J. 2015;128(21):2861–5. surgery following unicompartmental knee arthro-
33. Romagnoli S, Zacchetti S, Perazzo P, Verde F, Banfi plasty using local infiltration analgesia: a random-
G, Viganò M. Onsets of complications and revi- ized controlled trial with 40 patients. Acta Orthop.
sions are not increased after simultaneous bilateral 2012;83(6):634–41.
unicompartmental knee arthroplasty in comparison 44. Hooper GJ, Hooper NM, Rothwell AG, Hobbs
with unilateral procedures. Int Orthop (SICOT). T. Bilateral total joint arthroplasty. J Arthroplast.
2015;39(5):871–7. 2009;24(8):1174–7.
13 Utility of Bilateral Single-Stage Unicompartmental Knee Arthroplasty 161

45. Adili A, Bhandari M, Petruccelli D, de Beer 46. Severson EP, Mariani EM, Bourne MH. Bilateral
J. Sequential bilateral total knee arthroplasty under total knee arthroplasty in patients 70 years and
1 anesthetic in patients ≥75 years old. J Arthroplast. older. Orthopedics. 2009;32(5):316. https://doi.
2001;16(3):271–8. org/10.3928/01477447-­20090501-­13.
 Sports and Functional Activities
Following Unicondylar Knee 14
Arthroplasty

David A. Crawford and Keith R. Berend

14.1 Introduction In patients with end-stage knee arthritis, the

two main surgical treatment options are total
The goal of most knee arthroplasty procedures is knee arthroplasty (TKA) or unicondylar knee
to decrease pain and increase function. Those arthroplasty (UKA). While TKA is a more com-
patients with knee osteoarthritis (OA) have a pro- monly performed procedure for knee OA, many
gressive decline in their daily functioning, work patients with end-stage knee arthritis are candi-
and sports related activities [1]. This decline in dates for unicondylar knee arthroplasty (UKA)
function is important to the patients’ overall [9].
health as regular exercise has been shown to The indications and benefits of UKA are
reduce mortality, stimulate weight loss, reduce addressed in other chapters of this book; how-
anxiety/depression, and improve bone density ever, some benefits of UKA related to sports and
[2–4]. Studies have shown that undergoing a joint activity should be noted. Compared to TKA,
arthroplasty can reduce a patient’s risk of major UKA has been shown to have improved knee
cardiovascular events compared to those with kinematics, knee range of motion, and functional
arthritis that do not have surgery [5]. The demands outcome scores [10–12]. The preservation of the
on knee arthroplasty implants continue to cruciate ligaments and better knee kinematics
increase as life expectancy is higher than decades likely contribute to why a UKA tends to “feel
before [6], and patients want to stay active and more normal” than a TKA [13]. This more nor-
engaged in their working activities up to and after mal feeling may ultimately contribute why
retiring [7]. More younger patients are also seek- patients are able to return to a higher level of
ing knee arthroplasty as a treatment for arthritis, sport after UKA compared to TKA [8].
and these younger patients are expecting to return This chapter will review the published litera-
to a high level of activity [8]. ture on patient activity and participation in sports
following UKA.

D. A. Crawford (*) 14.2 Assessing Patient’s

Joint Implant Surgeons, Inc., New Albany, OH, USA
e-mail: [email protected]
Functional Activity
K. R. Berend
There are many functional and clinical outcome
Joint Implant Surgeons, Inc., New Albany, OH, USA
scores used to assess patients following lower
Mount Carmel Health System,
extremity surgery. Each of these scoring systems
New Albany, OH, USA

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2024 163
A. Clavé, F. Dubrana (eds.), Unicompartmental Knee Arthroplasty,
https://doi.org/10.1007/978-3-031-48332-5_14
164 D. A. Crawford and K. R. Berend

has varying levels of so-called ceiling effects (Table 14.1) and UCLA activity score [17]
[14]. For example, the Knee Society Functional (Table 14.2). These scoring metrics greater
score that is commonly used to assess functional separate activity levels and give examples of
outcome after knee arthroplasty is composed of specific sporting activities. When interpreting
only 3 questions: 1) walking distance, 2) stair functional outcomes following arthroplasty, it
climbing, and 3) walking aids used [15]. While is also important to know whether the compar-
this is a validated outcome score, this score would ative time frame is the patient’s pre-arthritic
not be able to differentiate between patients that functional level or their functional level just
can walk an unlimited distance from those that prior to arthroplasty. This information helps
can also golf, run, or ski. answer the question of whether patients can
More sports specific scoring systems that return to their pre-symptomatic level of activ-
are commonly reported in arthroplasty litera- ity or just improve from their symptomatic
ture include the Tegner activity level [16] arthritic level.

Table 14.1 Tegner activity level [16]

Level Description
0 Sick leave or disability pension because of knee problems
1 Work—sedentary (secretarial, etc.)
2 Work—light labor; walking on uneven ground possible, but impossible to backpack or hike
3 Work—light labor (nursing, etc.)
4 Work—moderately heavy labor (e.g., truck driving, etc.)
5 Work—heavy labor (construction, etc.); competitive sports—cycling, cross-country skiing; recreational
sports—jogging on uneven ground at least twice weekly
6 Recreational sports—tennis and badminton, handball, racquetball, down-hill skiing, jogging at least
five times per week
7 Competitive sports—tennis, running, motorcars speedway, handball; recreational sports- soccer,
football, rugby, bandy, ice hockey, basketball, squash, racquetball, running
8 Competitive sports—racquetball or bandy, squash or badminton, track and field athletics (jumping,
etc.), down-hill skiing
9 Competitive sports—soccer, football, rugby (lower divisions), ice hockey, wrestling, gymnastics,
basketball
10 Competitive sports—soccer, football, rugby (national elite)

Table 14.2 UCLA activity scale [17]

Level Description
1 Wholly inactive, dependent on others, and cannot leave residence
2 Mostly inactive or restricted to minimum activities of daily living
3 Sometimes participates in mild activities, such as walking, limited housework, and limited shopping
4 Regularly participates in mild activities
5 Sometimes participates in moderate activities such as swimming or could do unlimited housework or
shopping
6 Regularly participates in moderate activities
7 Regularly participates in active events such as bicycling
8 Regularly participates in active events, such as golf or bowling
9 Sometimes participates in impact sports such as jogging, tennis, skiing, acrobatics, ballet, heavy labor,
or backpacking
10 Regularly participates in impact sports
14 Sports and Functional Activities Following Unicondylar Knee Arthroplasty 165

14.3 Defining the Level of Activity 14.4 Activity Level After UKA

Terms such as “low impact,” “high impact,” “low Once patients become symptomatic with knee
activity,” and “high activity” are often used in the arthritis, their activity level decreases. Fisher
literature, but the definition and consensus on et al. who found that only 25% of patients who
these terms are vague. Patients often hear that underwent UKA were still participating in activi-
they may return to “low-impact” activities after ties that they did before symptom onset [1].
knee arthroplasty, but what exactly does that Correspondingly, pre-surgery UCLA activity lev-
mean? Running is often considered a “high els in patients undergoing UKA is low, ranging
impact” activity, while biking and swimming are from 3.3 to 5.3. [1, 21–23]. The goal of UKA is
more “low impact” activities. Some surgeons to improve patient’s pain and increase their activ-
have defined low activity as a Tegner level 4 or ity from prior to surgery. However, we should
less [18]. Robertson et al. defined high activity also aim to return patients to as close to pre-­
patients as those who completed >three million symptom level as possible.
gate cycles/year or 1 h of activity/day [19]. Work A few studies have compared post-operative
out of the Scripps Clinic has helped quantify the UKA activity to pre-symptom activity. Walker
impact of certain common activities on knee et al. compared patients sporting activity after
arthroplasty. D’Lima et al. implanted sensors in UKA to their activity level prior to any restricting
tibial components in vivo and measured forces symptoms of osteoarthritis. They found that 93%
from inside the prosthetic knee during various of patients were involved in at least one physical
activities and reported contact stresses for the fol- activity prior to the onset of symptoms and 92%
lowing activities (Table 14.3). Interesting golf, of patients participated in at least one sporting
which is often touted as a “low impact” sporting activity after surgery. This 1% decline repre-
activity, produced some of the highest joint sented 6 patients who had quit their pre-­symptoms
forces. Rowing was the only activity to have less sports, but 5 patients who began new sports after
than bodyweight force on the knee [20]. surgery [22]. Fisher et al. found that 93% of
patients after UKA were able to return to their
same level of activity as before knee arthritis
Table 14.3 Tibial forces after TKA during specific activ-
ities [20] symptoms [1], while others have found slightly
Multiple of body
less patients returning to their pre-symptom level
Activity weight of activity at 80.1% [24]. Ho et al. compared
Cycling 1.3 UCLA activity score prior to knee pain and after
Treadmill 2.05 UKA. They found that there was a significant
Walking on ground 2.6 decline in UCLA from 8.1 pre-knee pain to 7.4
Rowing 0.85 after surgery [25]. However, a score 7.4 is still
Tennis—Forehand 3.6 quite high correlating to an activity level between
Tennis—Backhand 3.1
“regularly participating in active events such as
Jogging 4.3
bicycling” and “regularly participating in active
Golf driving swing—Leading 4.5
leg events such as golf or bowling” [17].
Golf driving swing—Opposite 3.2 Most studies have compared post-operative
leg activity level to the patients’ immediate pre-­
166 D. A. Crawford and K. R. Berend

Table 14.4 Published results reporting rate of return to erated polyethylene wear and aseptic loosening
activity following UKA [32, 33]. A person with an average activity level
Number of Return to activity rate produces approximately 1.0 million knee cycles/
Study subjects (%) year where highly active individuals about 3.2
Fisher et al. 66 93
million knee cycles/year [34]. The concern over
[1]
Naal et al. 83 95
polyethylene wear in arthroplasty has slowed
[21] since polyethylene manufacturing has improved
Walker et al. 45 98 with decreased wear and oxidation [35].
[22] Furthermore, polyethylene wear in UKA is not as
Walker et al. 93 93 much of an issue as in TKA. In the mobile bear-
[23]
ing Oxford knee (Zimmer Biomet, Warsaw, IN),
Ho et al. [25] 36 87
for example, the 20-year wear was shown to be
only 0.4 mm [36]. Polyethylene wear is also a
operative activity level (Table 14.4). Overall relatively infrequent failure mode in UKA, repre-
activity level does tend to increase after UKA senting only 4% of UKA revisions [37].
with significant improvements in UCLA to mean In TKA literature, there are conflicting studies
scores of 6.3 to 7.1. [1, 22, 23, 26] and Tegner on activity level and the relationship to implant
activity levels of 2.6 to 4.0 [26–29]. failure. Some studies have shown a positive cor-
Patients may want to know what specific relation between activity level and arthroplasty
sports they can expect to be able to return to after failure [38–40], while others have not demon-
UKA. Common sports activities that increase strated any correlation [41, 42]. There has been
after UKA are swimming, hiking, aerobics, golf, limited research on the specific question of activ-
and dancing. However, patients may also expect a ity level and failure in UKA. One of the few stud-
decreased participation in certain sports such as ies was published by Al et al., who evaluated the
skiing, jogging, tennis, and soccer [1, 22, 26]. effect of activity level on survivorship of the
Time to return to sporting activities varies Oxford knee. They separated patients by post-­
between patients. Walker et al. found that in those operative Tenger score, with 4 or less being low
patient that returned to activity, 27% did so by the activity and 5 or greater being high activity. They
first month, 56% within 3 months, and 77% by found that the high activity group had 40% less
6 months after surgery [22]. revisions than the low activity group. Each 1
point increase in Tenger score was associated
with around 30% fewer revisions. They further
14.5 Effect of Activity Level found, which may be intuitive, that the more
on UKA Implant Survivorship active patients were younger [18]. Greco et al.
reported on 340 patients under the age of 50 years
Historically, the recommended activity level fol- old that underwent a medial UKA with the
lowing knee arthroplasty has been guided by Oxford mobile bearing implant. At a mean of
physician gestalt and consensus statements rather 6.1 years, only 2% of patients had a revision for
than objective publications of deleterious effects aseptic loosening and there were no revision for
from specific activities [30]. In a consensus state- polyethylene wear [43].
ment by the Knee Society in 2001, the recom-
mended activities following knee arthroplasty
were bowling, golfing, walking, swimming, and 14.6 Summary
dancing [31].
The concern from surgeons about high patient Patients who undergo UKA can expect to return
activity level is that certain activities may shorten to most activities they participated prior to knee
the survivorship of the arthroplasty due to accel- symptoms. Certain activities such as skiing and
14 Sports and Functional Activities Following Unicondylar Knee Arthroplasty 167

jogging may decline, while others such as swim- in the same patient: a comparative study. Clin Orthop
ming may increase. There does not appear to be a Relat Res. 1991;273:151–6.
14. Steinhoff AK, Bugbee WD. Knee Injury and
negative effect of increased activity on the survi- Osteoarthritis Outcome Score has higher respon-
vorship of UKA. siveness and lower ceiling effect than Knee Society
Function Score after total knee arthroplasty. Knee
Surg Sports Traumatol Arthrosc. 2016;24(8):2627–33.
15. Insall JN, Dorr LD, Scott RD, Scott WN. Rationale of
References the knee society clinical rating system. Clin Orthop
Relat Res. 1989;248:13–4.
1. Fisher N, Agarwal M, Reuben SF, Johnson DS, Turner 16. Tegner Y, Lysholm J. Rating systems in the evalua-
PG. Sporting and physical activity following Oxford tion of knee ligament injuries. Clin Orthop Relat Res.
medial unicompartmental knee arthroplasty. Knee. 1985;198:43–9.
2006;13(4):296–300. 17. Amstutz HC, Thomas BJ, Jinnah R, et al. Treatment
2. American College of Sports Medicine Physicians of primary osteoarthritis of the hip: a comparison of
Statement. The recommended quantity and quality of total joint and surface replacement arthroplasty. J
exercise for developing and maintaining cardiovas- Bone Joint Surg Am. 1984;66:228.
cular and muscular fitness in healthy adults. Med Sci 18. Ali AM, Pandit H, Liddle AD, Jenkins C, Mellon S,
Sports Exerc. 1990;22:265. Dodd CA, Murray DW. Does activity affect the out-
3. Macnicol MF, McHardy R, Chalmers J. Exercise test- come of the Oxford unicompartmental knee replace-
ing before and after hip arthroplasty. J Bone Joint ment? Knee. 2016;23(2):327–30.
Surg. 1980;62:326. 19. Robertson NB, Battenberg AK, Kertzner M,
4. Ries MD, Philbin EF, Groff GD, et al. Improvement Schmalzried TP. Defining high activity in arthroplasty
in cardiovascular fitness after total knee arthroplasty. patients. Bone Joint J. 2016;98-B(1 Suppl A):95–7.
J Bone Joint Surg Am. 1996;78:1696. 20. D'Lima DD, Steklov N, Patil S, Colwell CW Jr. The
5. Ravi B, Croxford R, Austin P, Lipscombe L, Bierman Mark Coventry Award: in vivo knee forces during
A, Harvey P, Hawker G. The relation between total recreation and exercise after knee arthroplasty. Clin
joint arthroplasty and risk for serious cardiovascular Orthop Relat Res. 2008;466(11):2605–11.
events in patients with moderate-severe osteoarthritis: 21. Naal FD, Fischer M, Preuss A, et al. Return to sports
propensity score. BMJ. 2013;347:f6187. and recreational activity after unicompartmental knee
6. Arias E, Heron M, Xu J. United States Life Tables, arthroplasty. Am J Sports Med. 2007;35:1688–95.
2013. Natl Vital Stat Rep. 2017;66(3):1–64. 22. Walker T, Gotterbarm T, Bruckner T, Merle C, Streit
7. Maxwell JL, Keysor JJ, Niu J, et al. Participation fol- MR. Return to sports, recreational activity and
lowing knee replacement: the MOST cohort study. patient-reported outcomes after lateral unicompart-
Phys Ther. 2013;93(11):1467–74. mental knee arthroplasty. Knee Surg Sports Traumatol
8. Witjes S, van Geenen RC, Koenraadt KL, van der Arthrosc. 2015;23(11):3281–7.
Hart CP, Blankevoort L, Kerkhoffs GM, Kuijer 23. Walker T, Streit J, Gotterbarm T, Bruckner T, Merle
PP. Expectations of younger patients concerning C, Streit MR. Sports, physical activity and patient-­
activities after knee arthroplasty: are we asking the reported outcomes after medial unicompartmental
right questions? Qual Life Res. 2017;26(2):403–17. knee arthroplasty in young patients. J Arthroplast.
9. Berend KR, Berend ME, Dalury DF, Argenson JN, 2015;30(11):1911–6.
Dodd CA, Scott RD. Consensus statement on indi- 24. Pietschmann MF, Wohlleb L, Weber P, Schmidutz
cations and contraindications for medial unicom- F, Ficklscherer A, Gülecyüz MF, et al. Sports activi-
partmental knee arthroplasty. J Surg Orthop Adv. ties after medial unicompartmental knee arthro-
2015;24(4):252–6. plasty Oxford III—what can we expect? Int Orthop.
10. Banks SA, Fregly BJ, Boniforti F, Reinschmidt C, 2013;37(1):31–7.
Romagnoli S. Comparing in vivo kinematics of uni- 25. Ho JC, Stitzlein RN, Green CJ, Stoner T, Froimson
condylar and bi-unicondylar knee replacements. Knee MI. Return to sports activity following UKA and
Surg Sports Traumatol Arthrosc. 2005;13:551–6. TKA. J Knee Surg. 2016;29(3):254–9.
11. Hollinghurst D, Stoney J, Ward T. No deteriora- 26. Jahnke A, Mende JK, Maier GS, Ahmed GA,
tion of kinematics and cruciate function 10 years Ishaque BA, Schmitt H, Rickert M, Clarius M,
after medial unicompartmental arthroplasty. Knee. Seeger JB. Sports activities before and after medial
2006;13:440–4. unicompartmental knee arthroplasty using the
12. Li MG, Yao F, Joss B, et al. Mobile vs. fixed bearing new Heidelberg sports activity score. Int Orthop.
unicondylar knee arthroplasty: a randomized study 2015;39(3):449–54.
on short term clinical outcomes and knee kinematics. 27. Pandit H, Jenkins C, Gill HS, Barker K, Dodd CA,
Knee. 2006;13:365–70. Murray DW. Minimally invasive Oxford phase 3
13. Laurencin CT, Zelicof SB, Scott RD, Ewald unicompartmental knee replacement: results of 1000
FC. Unicompartmental versus total knee arthroplasty cases. J Bone Joint Surg Br. 2011;93(2):198–204.
168 D. A. Crawford and K. R. Berend

28. Weston-Simons JS, Pandit H, Kendrick BJ, Jenkins C, mobile-bearing unicompartmental knee replacement
Barker K, Dodd CA, Murray DW. The mid-term out- at 20 years. J Bone Joint Surg Br. 2011;93(4):470–5.
comes of the Oxford domed lateral unicompartmental 37. Van der List JP, McDonald LS, Pearle AD. Systematic
knee replacement. Bone Joint J. 2014;96-B(1):59–64. review of medial versus lateral survivorship
29. Yim JH, Song EK, Seo HY, Kim MS, Seon in unicompartmental knee arthroplasty. Knee.
JK. Comparison of high tibial osteotomy and unicom- 2015;22(6):454–60.
partmental knee arthroplasty at a minimum follow-up 38. Lavernia CJ, Sierra RJ, Hungerford DS, et al.
of 3 years. J Arthroplast. 2013;28(2):243–7. Activity level and wear in total knee arthroplasty: a
30. Swanson EA, Schmalzried TP, Dorey FJ. Activity rec- study of autopsy retrieved specimens. J Arthroplast.
ommendations after total hip and knee arthroplasty: a 2001;16(4):446–53.
survey of the American Association for hip and Knee 39. Odland AN, Callaghan JJ, Liu SS, Wells CW. Wear
Surgeons. J Arthroplast. 2009;24(6):120–6. and lysis is the problem in modular TKA in the young
31. Healy WL, Iorio R, Lemos MJ. Athletic activ- OA patient at 10 years. Clin Orthop Relat Res. 2011
ity after joint replacement. Am J Sports Med. Jan;469(1):41–7.
2001;29(3):377–88. 40. Rohrbach M, Lüem M, Ochsner PE. Patient and
32. Harris WH. Wear and periprosthetic osteolysis: the surgery related factors associated with fatigue type
problem. Clin Orthop Relat Res. 2001;393:66–70. polyethylene wear on 49 PCA and DURACON
33. Horikoshi M, Macaulay W, Booth RE, Crossett retrievals at autopsy and revision. J Orthop Surg Res.
LS, Rubash HE. Comparison of interface mem- 2008;22(3):8.
branes obtained from failed cemented and cement- 41. Bisschop R, Brouwer RW, Van Raay JJ. Total knee
less hip and knee prostheses. Clin Orthop Relat Res. arthroplasty in younger patients: a 13-year follow-up
1994;309:69–87. study. Orthopedics. 2010;33(12):876.
34. Seedhom B, Dowson D, Wright V. Wear of solid phase 42. Diduch DR, Insall JN, Scott WN, et al. Total knee
formed high density polyethylene in relation to the replacement in young, active patients: long-term
life of artificial hips and knees. Wear. 1973;24:35–51. follow-up and functional outcome. J Bone Joint Surg
35. Grupp TM, Fritz B, Kutzner I, Schilling C, Bergmann Am. 1997;79(4):575–82.
G, Schwiesau J. Vitamin E stabilised polyethylene 43. Greco NJ, Lombardi AV Jr, Price AJ, Berend
for total knee arthroplasty evaluated under highly ME, Berend KR. Medial Mobile-Bearing
demanding activities wear simulation. Acta Biomater. Unicompartmental Knee Arthroplasty in Young
2017;48:415–22. Patients Aged Less Than or Equal to 50 Years. J
36. Kendrick BJ, Simpson DJ, Kaptein BL, Valstar ER, Arthroplast. 2018;33(8):2435–9.
Gill HS, Murray DW, et al. Polyethylene wear of
 Complications
of Unicompartmental Knee 15
Replacement

Stefano Campi

15.1 Introduction 15.2 Progression of Osteoarthritis

in the Retained
Complications after unicompartmental knee Compartments
replacement occur with a similar incidence than
after TKA. However, their management can be Progression of osteoarthritis in the retained com-
difficult in the hands of low volume surgeons, partments represents the most common cause of
who unfortunately tend to experience more than failure of UKA in the majority of clinical series,
expert surgeons. The higher susceptibility to revi- with an incidence between 0.9 and 7% [1–3] [4–
sion and the scepticism towards UKR of many 6]. It represent the third cause of failure in the
orthopaedics surgeons are a threat for patients National Joint Registry of England and Wales,
undergoing UKR, especially if they encounter a with a revision rate of 2.27 for 100 component
complication. It is therefore really important to years [7].
recognise and manage the most common compli- The causes of OA progression in the lateral
cations of UKR and protect patients from unnec- compartment are still controversial.
essary revisions and/or overtreatment. A second Overcorrection of the mechanical axis of the
opinion from an experienced UKR surgeon can lower limb is considered as a relevant cause by
be really helpful in dealing with complex many authors, causing overload of the lateral
situations. compartment of the knee. This complication is
The most common causes of failure and com- due to an error in the surgical technique implying
plications of UKR will be discussed in this the release or damage of the medial collateral
chapter. ligament, inadequate bone resections, or the use
of thick inserts with LCM stretching. However,
OA progression is frequently observed in knees
that are not overcorrected [1]. Overcorrection is
probably one of the possible causes of OA pro-
gression, but not the only one [8].
It has been suggested that OA progression is
time dependent. However, long-term studies did
S. Campi (*) not show an increase of this complication over
Department of Medicine and Surgery, Università time [9].
Campus Bio-Medico di Roma, Fondazione Other factors, such as BMI or chondrocalcino-
Policlinico Universitario Campus Bio-Medico, Rome,
Italy
sis, have been suggested as possible causes of OA

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2024 169
A. Clavé, F. Dubrana (eds.), Unicompartmental Knee Arthroplasty,
https://doi.org/10.1007/978-3-031-48332-5_15
170 S. Campi

progression, but this hypothesis has not been sup- Clinically, the main symptom is pain, which
ported by clinical studies [10, 11]. can correspond to the affected compartment or be
Finally, it has been suggested that lateral OA referred elsewhere. It is important to notice that
is mainly related to the conditions of the lateral the radiographic evidence of lateral OA is not
compartment at the moment of the operation, and always associated with pain.
that the missed diagnosis of chondral damage can There are two treatments for symptomatic
cause the subsequent failure of the implant [8]. progression of OA. The first is revision to
Consequently, in the author’s opinion, it is man- TKA. This can be performed extending the old
datory to obtain stress-views or Rosenberg views incision through a medial parapatellar approach.
besides standard, weight-bearing radiographs. The second option is the addition of a lateral
The use of MRI is controversial and still debated. UKR. Thus technically more demanding, the lat-
The progression of osteoarthritis is usually ter is an effective procedure which should be pre-
diagnosed on standard weight-bearing X-rays. In ferred in patients that have been happy with their
some cases, stress X-rays or Rosenberg views are medial UKR for years [13]. This operation can be
needed to highlight this complication (Fig. 15.1). performed through a lateral approach or extend-
Joint space narrowing or its disappearance are the ing the medial skin incision and performing a lat-
most relevant radiographic finding. In contrast, eral parapatellar approach, or using a TKA
the presence of marginal osteophytes is not spe- approach.
cific and it is common also in the presence of pre-
served cartilage [12].

Fig. 15.1 Osteoarthritis progression in the lateral compartment shown on Rosenberg views, with lateral joint space
narrowing, sclerosis, and marginal osteophytes
15 Complications of Unicompartmental Knee Replacement 171

15.3 Infection

Periprosthetic joint infections are relatively

uncommon in partial knee replacement, with an
incidence around 1% [14]. The diagnostic algo-
rithm is the same as in total knee replacement
and should be based on the most recent
guidelines.
Acute infections can be treated with a DAIR
procedure, while late infections usually require a
revision to total knee replacement. This can be
performed either as a single-stage or a two-stage
procedure. The author’s preference is to perform
a two-stage procedure. In the first-stage, beside
implant removal, it is important to perform also
the total knee resections to remove the retained
cartilage, as it can be contaminated with bacteria
[15]. A standard spacer, either fixed or articulat-
ing, is implanted and followed by i.v. and/or oral
antibiotics. Once the infection is cleared, the sec-
ond stage can be performed. Fig. 15.2 Physiological radiolucent lines around the
tibial component of a well-functioning mobile-bearing
UKR. Physiological RL are usually less than 2 mm in
thickness, non-progressive and with a sclerotic margin
15.4 Aseptic Loosening

Aseptic loosening is a frequent cause of failure in of fixation or infection [16]. It has to be noticed
the joint registries. However, it is really uncom- that the presence of RL is common in all cemented
mon in case series from high volume centres. implants; however, it is more evident in UKR
This discrepancy has different possible explana- than TKR because of the absence of a central
tions. First, revisions for unexplained pain are keel. Furthermore, the presence of RLs is influ-
frequently categorised as loosening even in the enced by the X-ray alignment, so that even few
absence of clear evidence of such complication. degrees of inclination of the beam can hide or
Second, the wrong interpretation of peripros- show them.
thetic radiolucent lines (RL) is frequent among While failure, the diagnosis of tibial loosening
surgeons that are not familiar with is usually evident on standard X-ray, the loosen-
UKR. Radiolucent lines are frequently encoun- ing of the femoral component is less frequent but
tered in the X-rays of well-functioning UKRs also less easy to diagnose. If there is the suspect
(Fig. 15.2). It is paramount to distinguish physi- of femoral loosening, a lateral X-ray with the
ological and pathological radiolucencies. A phys- knee in flexion and then in extension can high-
iological RL is usually less than 2 mm in light this problem showing position changes of
thickness, non-progressive, with a sclerotic mar- the component.
gin. Physiological RL are not correlated with In case of aseptic loosening is generally revi-
loosening and do not affect the survival of the sion to TKR. The revision of UKR to UKR is
implant. In contrast, a progressive, poorly defined option in early loosening. However, this indica-
radiolucent line that is thicker than 2 mm is con- tion is controversial and should be performed is
sidered “pathological” and correlated with failure selected cases by experienced surgeons.
172 S. Campi

15.5 Fracture

Perioperative medial condyle fracture is a rare

but known complication of UKRs [17]. The aeti-
ology of fracture is likely to be multifactorial.
Technical errors such as a deep tibial resec-
tion, a medialised positioning of the tibial com-
ponent, the damage of the posterior cortical bone,
and the use of a heavy hammer are known risk
factors [18]. In addition, patient characteristic
such as small sizes and poor bone quality can
increase the risk of fracture. The risk seems
higher in specific subset of patients such as the
Asian population, in which the size and morphol-
ogy of the proximal tibia can predispose to frac-
turing of the medial condyle.
A careful surgical technique can significantly
reduce the risk of this complication.
Most of the fractures are diagnosed intraoper-
atively or in the first 4–8 weeks after surgery,
when patients increase mobilisation and
weight-bearing.
A standard X-ray is usually enough for the Fig. 15.3 X-ray of a periprosthetic fracture of the tibial
diagnosis (Fig. 15.3). condyle after medial unicompartmental knee replacement,
Some surgeons have experienced and diagnosed 4 weeks after the operation
increased number of fracture using cementless,
mobile-bearing UKRs. This is probably related to
the interference generated around the keel of the 15.6 Bearing Dislocation
tibial component during implantation. Even in Mobile Bearing Designs
though this complication is rare, the suspect
should be higher when cementless UKRs are Bearing dislocation is a relatively uncommon
used, especially at the beginning of the learning complication of mobile-bearing designs. With
curve. the recent design and surgical technique improve-
The conservative treatment with restricted ments, the dislocation rate is around one on 200
weight-bearing is usually effective for undis- cases (0.5%) [2]. Bearing dislocations tend to
placed or minimally displaced fractures diag- occur early. Primary dislocations are usually
nosed within 3 months after surgery. In case of caused by a surgical error: inadequate osteophyte
displaced fractures (with stable tibial component removal, poor gap balancing, MCL damage,
over the fragment), open reduction and internal retained cement, or a femoral component sited
fixation with screws and/or buttress plate are the too far from the sagittal wall allowing the bearing
preferred treatment. In case of late diagnosis or to spin. A “loose” bearing is often a reason of
loose components, revision to TKR with stemmed concern for surgeons that are not familiar with
components is usually required [18]. the procedure. However, this is rarely a cause of
15 Complications of Unicompartmental Knee Replacement 173

Anteromedial pain is not uncommon in the

first 6–12 months after the operation, and it
resolves spontaneously in the vast majority of
cases. One possible cause is related to bone over-
load and remodelling in the proximal tibia after
the operation [20]. These cases must be treated
conservatively with active monitoring and reas-
surance of the patient. A second opinion from an
expert surgeon can be helpful for managing com-
plex patients. Some studies report benefits from
the off-label use of high dosages of clodronate
[21].
A relevant pain in the anteromedial region in
the first 2–3 months after the surgery can also be
caused by a tibial condyle fracture. This should
be always suspected and ruled out with appropri-
ate imaging.
Possible causes of persistent pain are compo-
nent malpositioning with significant overhang
(causing soft tissue irritation), retained cement
Fig. 15.4 Lateral view showing a dislocated bearing in
the supra-­patellar pouch
fragments or loose bodies, impingement, tendi-
nopathies, meniscal tears, progression of OA in
dislocation. In contrast, a bearing that is too tight the retained compartment. However, the most
increases the risk of dislocation [18]. A bearing frequent cause of unexplained pain is a wrong
dislocation can also result from a trauma or twist- indication to surgery. The results of UKR in par-
ing injury of the knee. tial thickness cartilage loss are not predictable
A bearing dislocation is usually characterised and correlated with a higher incidence of persis-
by acute pain and functional impairment. tent pain [22]. The surgical treatment should
However, it can be relatively silent in some cases. always be reserved to patients with end-stage
Bearing dislocations are usually noticeable on a OA, with no exceptions. The bone-on-bone con-
standard X-ray (Fig. 15.4). tact is not always revealed by standard, weight-­
In most cases, the treatment consists in the bearing X-rays and has to be confirmed with
arthrotomy, joint examination to assess and further imaging (stress X-rays, Rosenberg views).
address possible causes of dislocation, and bear- If a patient with unexplained pain has been oper-
ing exchange. In case of recurrent dislocation or ated elsewhere, it is fundamental to review the
severe imbalancing of the gaps, revision to TKA preoperative imaging. In case the operation was
can be necessary. performed on a partial thickness disease, a revi-
sion is likely to be ineffective. In these cases of
“overtreatment”, further operations usually lead
15.7 Unexplained Pain to disappointing results and should be avoided.
Better results are achieved with pain therapy.
Unexplained pain is the second most common Adequate imaging is important to assess com-
cause of revision of UKR according to the NJR, ponent positioning. An overhang greater than
while it is relatively uncommon in case series 2 mm can hypothetically cause soft tissue irrita-
from high volume centres (0.6% of revisions) tion and pain. However, this is not always the
[19]. case. The nature and features of pain should be
174 S. Campi

Arthroscopy can be useful also when a meniscal

tear is suspected, or in case of proper “unex-
plained pain” when other possible causes of fail-
ure have been investigated.
When a revision is needed, conversion to TKR
is usually the most reliable option. However,
selected cases in the hands of expert surgeons can
be treated with partial revision.
As a general rule, in case of unexplained pain, a
UKR should be treated with the same approach
and threshold to revision of a TKR: early revisions
(less than 2 years post-op) should be avoided,
unless there is an obvious cause of failure.

References
1. Emerson RH Jr, Higgins LL. Unicompartmental knee
arthroplasty with the oxford prosthesis in patients
with medial compartment arthritis. J Bone Joint Surg
Am. 2008;90(1):118–22.
Fig. 15.5 X-ray showing a retained cement fragment 2. Pandit H, Jenkins C, Gill HS, Barker K, Dodd
medial to the bearing. This patient initially experienced CA, Murray DW. Minimally invasive Oxford
mechanical symptoms, which settled in few weeks’ time phase 3 unicompartmental knee replacement:
after the fragment moved in the posterior recess results of 1000 cases. J Bone Joint Surg Br.
2011;93(2):198–204.
3. Khan OH, Davies H, Newman JH, Weale
evaluated meticulously to confirm this hypothe- AE. Radiological changes ten years after St. Georg
sis. An injection with local anaesthetics over the sled unicompartmental knee replacement. Knee.
tender spot can be useful to confirm the 2004;11(5):403–7.
diagnosis. 4. Kim KT, Lee S, Kim JH, Hong SW, Jung WS,
Shin WS. The survivorship and clinical results of
Malalignment of the components is frequently minimally invasive Unicompartmental knee arthro-
observed in X-rays of well working implants. plasty at 10-year follow-up. Clin Orthop Surg.
Therefore, it is really important to rule out other 2015;7(2):199–206.
possible causes of failure (including the indica- 5. Pandit H, Hamilton TW, Jenkins C, Mellon SJ, Dodd
CA, Murray DW. The clinical outcome of minimally
tion to surgery) before considering revision invasive phase 3 Oxford unicompartmental knee
surgery. arthroplasty: a 15-year follow-up of 1000 UKAs.
In contrast, overcorrection of limb alignment, Bone Joint J. 2015;97-B(11):1493–500.
overstuffing of the medial compartment, and 6. Campi S, Pandit HG, Dodd CAF, Murray
DW. Cementless fixation in medial unicompartmen-
MCL stretching are less tolerated and often cause tal knee arthroplasty: a systematic review. Knee Surg
problems on the short- or long-term follow-up. Sports Traumatol Arthrosc. 2017;25(3):736–45.
Retained cement fragments or loose bodies 7. National Joint Registry for England, Wales and
are not uncommon, especially when using a min- Northern Ireland. 13th annual report., 2016.
8. Pandit H, Spiegelberg B, Clave A, McGrath C, Liddle
imally invasive approach (Fig. 15.5). The patients AD, Murray DW. Aetiology of lateral progression of
usually refer mechanical symptoms while walk- arthritis following Oxford medial unicompartmental
ing, with sharp, occasional pain. In some cases, knee replacement: a case-control study. Musculoskelet
the symptoms resolve spontaneously over time. Surg. 2016;100(2):97–102.
9. Price AJ, Waite JC, Svard U. Long-term clini-
However, if there is evidence of a loose fragment cal results of the medial Oxford unicompart-
and the patients is symptomatic, arthroscopic mental knee arthroplasty. Clin Orthop Relat Res.
removal is a viable and effective option. 2005;435:171–80.
15 Complications of Unicompartmental Knee Replacement 175

10. Murray DW, Pandit H, Weston-Simons JS, Jenkins C, cal radiolucency following Oxford unicompartmental
Gill HS, Lombardi AV, Dodd CA, Berend KR. Does knee replacement and its relationship to outcome. J
body mass index affect the outcome of unicompart- Bone Joint Surg Br. 2009;91(7):896–902.
mental knee replacement? Knee. 2013;20(6):461–5. 17. Ji JH, Park SE, Song IS, Kang H, Ha JY, Jeong
11. Hernigou P, Pascale W, Pascale V, Homma Y, JJ. Complications of medial unicompartmental knee
Poignard A. Does primary or secondary chondro- arthroplasty. Clin Orthop Surg. 2014;6(4):365–72.
calcinosis influence long-term survivorship of uni- 18. Goodfellow J, O'Connor J, Pandit H, Dodd C, Murray
compartmental arthroplasty? Clin Orthop Relat Res. D. Unicompartmental arthroplasty with the Oxford
2012;470(7):1973–9. knee, vol. 1. 2nd ed. Goodfellow Publishers Limited;
12. Waldstein W, Kasparek MF, Faschingbauer M, 2015.
Windhager R, Boettner F. Lateral-compartment osteo- 19. Mohammad HR, Strickland L, Hamilton TW, Murray
phytes are not associated with lateral-compartment DW. Long-term outcomes of over 8,000 medial
cartilage degeneration in arthritic Varus knees. Clin Oxford phase 3 Unicompartmental knees-a system-
Orthop Relat Res. 2017;475(5):1386–92. atic review. Acta Orthop. 2018;89(1):101–7.
13. Pandit H, Mancuso F, Jenkins C, Jackson WFM, 20. Pegg EC, Walter J, Mellon SJ, Pandit HG, Murray DW,
Price AJ, Dodd CAF, Murray DW. Lateral unicom- D'Lima DD, Fregly BJ, Gill HS. Evaluation of factors
partmental knee replacement for the treatment of affecting tibial bone strain after unicompartmental
arthritis progression after medial unicompartmental knee replacement. J Orthop Res. 2013;31(5):821–8.
replacement. Knee Surg Sports Traumatol Arthrosc. 21. Frediani B, Giusti A, Bianchi G, Dalle Carbonare
2017;25(3):669–74. L, Malavolta N, Cantarini L, Saviola G, Molfetta
14. Liddle AD, Judge A, Pandit H, Murray L. Clodronate in the management of differ-
DW. Determinants of revision and functional out- ent musculoskeletal conditions. Minerva Med.
come following unicompartmental knee replacement. 2018;109(4):300–25.
Osteoarthr Cartil. 2014;22(9):1241–50. 22. Pandit H, Gulati A, Jenkins C, Barker K, Price AJ,
15. Goodfellow JOCJ, Dodd CAF, Murray Dodd CA, Murray DW. Unicompartmental knee
DW. Unicompartmental arthroplasty with the Oxford replacement for patients with partial thickness car-
knee. New York: Oxford University Press; 2006. tilage loss in the affected compartment. Knee.
16. Gulati A, Chau R, Pandit HG, Gray H, Price AJ, 2011;18(3):168–71.
Dodd CA, Murray DW. The incidence of physiologi-
 What to Do If a Medial
Unicompartmental Knee 16
Arthroplasty Fails

F. -X. Gunepin, L. Tristan, G. Le Henaff, O. Cantin,

and T. Gicquel

16.1 Introduction involve implants of different design and inser-

tion. North European and Australian registers
The rate of revision surgery for unicompartmen- have evidenced this [3–6].
tal knee arthroplasty (UKA) can be very different When a medial unicompartmental knee arthro-
in the literature and vary by a factor of 1 to 2 plasty fails, it is sometimes difficult to diagnose
depending on the publication: from 91% survival loosening of the implant. Certainly, unexplained
at 20 years in a series on the Oxford implant to pain can be an early manifestation of loosening
21% failure at 15 years for the Australian register of the implant, excess stress or micro-mobility
[1]. The UK register lies between the two with processes. The combination of unexplained pain
88% survival at 10 years. However, all authors and loosening, all series combined, represents
agree that survival increases with the expertise of 2/3 to 3/4 of the causes of revision surgery. The
the centres and practitioners [2]. Oxford teams make a distinction between techni-
Early failures tend to result from improper cal error and improper indication [4, 7].
indication or incorrect surgical technique with a Dislocation of the polyethylene (PE) insert is
few complex regional pain syndromes. In the specific to implants with a mobile insert (Table 16.1).
long term, causes of failure are progression of In all cases, it is necessary to seek to identify
osteoarthritis (OA) in the other knee compart- this failure in order to offer the patient the most
ments or wear of the polyethylene insert. appropriate therapeutic solution for his/her
Complications such as fracture or infection occur situation.
more randomly. The reason for consultation is the onset of
Analysing the causes of medial UKA failure is pain more or less associated with deterioration of
relatively difficult because retrospective studies a functional result that was previously satisfac-
tory. The context can be sudden (trauma, disloca-
F. -X. Gunepin (*) · L. Tristan
tion of the PE) or slowly progressive (wear). The
G. L. Henaff · O. Cantin combination with inflammatory signs should
Clinique Mutualiste de la porte de L’Orient, suggest sepsis, which will require prompt aggres-
Lorient, France sive management to save the implant.
T. Gicquel
Clinique Mutualiste de la porte de L’Orient,
Lorient, France
Unexplained pain should suggest the
Department of Orthopaedic Surgery and
Traumatology, Rennes University Teaching Hospital, hypothesis of an unstable implant.
Rennes, France

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2024 177
A. Clavé, F. Dubrana (eds.), Unicompartmental Knee Arthroplasty,
https://doi.org/10.1007/978-3-031-48332-5_16
178 F. -X. Gunepin et al.

Table 16.1 Aetiology of UKA failures

Oxford (%) NJR UK (%) Epinette/SFHGa (%) Australian register (%)
Unexplained pain 65 25 5.5 12
OA progression 10 20 15 14
Loosening of the implant (T/F/ 10 (7 and 32 44 (25/6/13) 54
T + F) 3)
Dislocation/wear of the PE insert 7 6 12 4.8 (2, 8/2)
Infection 6 5 2 4
Technical error 2 6 11.5 3
Fracture 2 2 4 2
a
SFHG: French Society of Hip and Knee Surgery

16.2 Specific Case of Allergy Before a joint implant is inserted, comple-

tion of an allergy screening questionnaire
To date, there has been no published report on is advisable. In cases of unexplained loos-
UKA failure due to allergy, but many situations ening of the implant or pain, the hypothesis
of pain and inflammation remain unexplained. of an allergic origin should be considered.
The processes of an allergic complication in knee
arthroplasties are regularly described. The diag-
nosis should be considered in a consultation by
questioning the patient on their history of atopic 16.3 Measures to Be Taken
dermatitis, asthma or metal intolerance (belt on Wear Progression
buckle, expansion band of a wristwatch, costume
jewellery, earrings). Wear progression can occur in patellofemoral or
The operator may want to use the question- lateral tibiofemoral OA. The therapeutic solu-
naire developed by the SFHG [8] (https://www. tions range from total knee arthroplasty or
sfhg.fr/accueil/fiches-­d-­information/allergies-­et-­ replacement of the deteriorated compartment. In
prothèses/) to aid management. In case of doubt, the case of lateral or patellofemoral unicompart-
an assessment should be performed. It is based mental implants, the surgeon should refer to con-
on conducting skin patch tests [8–9]. The sensi- ventional indications for these procedures. The
tivity of these tests is an imperfect reflection of only factor that should be considered is the degree
the biological reality, and a skin allergy is not of wear on a medial UKA. Any alteration in the
necessarily correlated with joint allergy. internal implant should suggest use of total knee
Whatever the reason, the preoperative hypothesis arthroplasty.
of possible allergy should lead the surgeon to For a lateral UKA, the preoperative axis of the
consider the choice of implant; the unavailability limb should be in valgus position, and it is logical
of a titanium or surface-treated UKA should lead to maintain slight postoperative valgus in order
him/her to consider using a total arthroplasty or not to overload the medial UKA, which is by
to continue with further investigations by con- definition an older implant and therefore poten-
ducting a MEmory Lymphocyte Immuno tially already involved. Other criteria, of course,
Stimulation Assay (MELISA), the only specific should be followed: reducibility of the valgus
test for metal allergy but of limited access (no deformity, conserved joint amplitude and effi-
laboratory in France) [10]. cient ligament system [11].
16 What to Do If a Medial Unicompartmental Knee Arthroplasty Fails 179

For patellofemoral arthroplasty, indications Other investigations (needle puncture, scintig-

are rare. They can be considered in light of UKA raphy, CT scan, etc.) will be used as needed and
deterioration with the onset of patellofemoral oriented by the possible diagnoses: loosening of
pain in combination with Iwano radiological the implant, sepsis, allergy.
stage 3 or 4. The central pivotal point should be Choice of the TKR should also consider pres-
intact, and joint amplitude should be conserved ent or potential loss of bone substance (fixation
or with limited and reducible stiffness (<10°). method). The choice can involve:
The axis of the lower limb should be close to nor-
mal [12]. –– First-line implant more or less under stress.
Once the indication has been established, the –– First-line implant with a stem or augment.
approach will be chosen based on the one per- –– Specific case of a first-line implant with an
formed for medial UKA. autologous bone graft augment.
–– An implant for revision surgery more or less
under weightbearing.
Wear progression in the lateral compart-
ment can benefit from revision by external Medial UKA implants should be kept as long
UKA subject to a reducible genu valgum. as possible during revision surgery to give the
surgeon the most precise view of the lower limb
axis and joint space height [14].
Lewis et al. studied the impact of different
16.4 TKR After UKA solutions on TKR survival for UKA revision
through data from the Australian register [15].
In cases of UKA failure, revision surgery with They did not find any difference in survival
TKR is the most common approach. In the SFHG between posterior-stabilised or non-­
series combining 425 failures of UKA alone, 36 weightbearing implants. On the contrary, they
underwent revision by UKA, i.e. 8.5% [13], and showed with a statistically significant difference
in a series by Lewis, this figure rose to 11.2% that use of a stem (with or without augments/
(follow-up of 45,615 UKA in the Australian reg- blocks) increased survival at 10 years (87% ver-
ister). The implant should be chosen following an sus 81% without a stem). They also noted, irre-
attentive clinical and paraclinical assessment. spective of the TKR used (posterior-stabilised or
Clinically, it is necessary to assess the mor- non-weightbearing, with or without a stem, with
photype, joint amplitudes, muscle capital and or without augments/blocks), that the survival of
competence of the different ligament structures. cemented implants is systematically better than
The surgeon should note the position and size of cementless implants.
scars and the existence and extent of any joint Scott et al. showed that the use of augments
effusion (which can be punctured and drained). and stems is more common in UKA revision sur-
Every effort should be made to obtain the previ- gery with metal backs than for full polyethylene
ous surgery report. tibial implants, which are more economical in
The paraclinical assessment consists of stan- terms of bone tissue resection [16].
dard X-ray views (anteroposterior and profile During placement of a TKR for medial UKA
with weightbearing, patellofemoral series). failure, the lateral compartment is often the
Long-leg radiographs are advisable to provide healthy one. The reference for the mechanical
details of the limb axis, but also the presentation sectioning guide is therefore the lateral tibial gle-
of the tibial and femoral diaphysis in order to use noid surface with a conventional section of
a stem approach. In case of doubt on the intact- 10 mm in height below the reference level. At this
ness of the collateral ligament planes, X-rays stage, the operator can assess the difference
with stress on the joint should be performed. between the section and the healthy area of bone
180 F. -X. Gunepin et al.

under the medial tibial implant. It is this ­difference must be available in the operating room at the
that should guide the operator in the choice of slightest doubt.
revision technique. This assessment is made pre- Perioperatively, it is conventional to start with
operatively based on anteroposterior X-ray views tibial resection with the medial implant in posi-
with one magnification or a CT scan. tion. The probe marker is placed in the healthy
Perioperatively, mechanical guides, but also nav- compartment, which is most often in the lateral
igation, can be used. Crawford et al. recommend compartment. It is then possible to evidence the
the systematic use of sealed implants. They use level of resection under the lateral compartment
the fixation of stems and augments when the (at 10 mm for most implants), and to verify at
medial tibial section is more than 5 mm above the what level resection will be performed in the
lateral section [17]. A study by Marinier et al., medial level. At that stage, it is necessary to refer
SOFCOT 2017 congress, confirmed that, beyond to the table. Below a 5-mm defect, the cut can
cutting of 14 mm, a first-line implant risks being descend and be compensated by thicker polyeth-
insufficient [18]. ylene. This is all the more relevant when the tibia
Deficiency of the medial collateral ligament is large in size (Diagram 16.1). This can some-
should be analysed. If it involves incompetency times be problematic in small sizes because low-
related to UKA failure, revision of the UKA ering of the section can result in a decrease in the
should enable restoration of the MCL tension, weightbearing area of the tibial implant. It is in
and Table 16.2 can be used. If not, it will be nec- this situation that an autologous bone graft aug-
essary to plan the use of stress implants, which ment may be useful.
If the difference between resection height lat-
erally and in the medial healthy area is greater
Table 16.2 Choice of TKR type for UKA revision than 5 mm, it would be necessary to choose a
Bone defect Bone quality stem more or less in combination with an aug-
compared to ment (which can sometimes also be performed at
the lateral
the expense of the lateral resection).
section Good Mediocre or poora
≤5 mm Cemented primary Cemented
implant primary implant
(Fig. 16.1a) with a short (or
long) stem The unpredictable nature of UKA revision,
(Fig. 16.1b) related to perioperative discovery, makes it
>5 mm Cemented primary Cemented necessary to have a TKR with stem and
implant with a primary implant
short (or long) with a long stem augment. At the slightest doubt over the
stem and augmentb and augment quality of the MCL, it would also be neces-
(Fig. 16.1c) (Fig. 16.1d) sary to have a weightbearing revision TKR.
a
Bone quality and BMI (body mass index)
b
Possibility of using an autologous bone graft
16 What to Do If a Medial Unicompartmental Knee Arthroplasty Fails 181

a b

c d

Fig. 16.1 Choice of TKR type for UKA revision: (a) primary, (b) primary with short (or long) stem, (c) primary with
short (or long stem) and augment, (d) primary with long stem and augment

Diagram 16.1 If the tibial cut is to big the loss of width can result in a harmful decrease in the tibial implant size and
inadequacy with the femoral implant

uled height of the lateral cut and the medial cut in

16.5 Principle of the Autologous
the healthy area must be assessed. This difference
Bone Graft Augment/Block
provides the thickness of the defect to be filled. It
is necessary to start with subchondral resection
Whenever use of an augment is necessary, and if
laterally and then to perform sectioning at the
bone quality is satisfactory, the lateral tibial
height of the defect to be filled medially and then
resection can be used as an autologous bone graft
a final resection to reach the height of the final
augment. This makes it possible for the surgeon
lateral resection (Diagrams 16.2a, 16.2b and
to work in the conditions of a first-line total
16.2c).
arthroplasty. The difference between the sched-
182 F. -X. Gunepin et al.

Diagram 16.2a How to calculte the thickness of the bone graft augment

Diagram 16.2b Preparation of the bone graft augment

An autologous bone graft augment is an alternative to metal block/augment, which enables

bone saving provided that the bone is of good quality. The delicate phase is preparation of the
tibial baseplate.

Diagram 16.2c Positionning of the bone graft augment

The bone graft must be sized and positioned in is recommended to use a sealed tibial baseplate
the medial compartment. Temporary stabilisation [15–19].
with a pin is performed. The guide for prepara-
tion of the tibial baseplate is positioned. It is rec-
ommended to prepare tibial stamping for
An autologous bone graft augment is an
baseplates with a different wing by making a saw
alternative to metal block/augment, which
cut so as not to risk splitting the graft. After this
enables bone saving provided that the bone
preparation, temporary pins can be replaced with
is of good quality. The delicate phase is
compression screws, taking care not to enter into
preparation of the tibial baseplate.
contact with the imprint of the tibial baseplate
(Fig. 16.2). Even for satisfactory bone quality, it
16 What to Do If a Medial Unicompartmental Knee Arthroplasty Fails 183

Fig. 16.2 First-line TKR revision with a medial autologous bone graft augment affixed by screws and use of
navigation

16.6 UKA and Sepsis –– Chronic in all cases when symptoms progress
for more than four months.
The occurrence of sepsis in UKA is governed by In cases of acute infection, imaging has
the principles of infection management in a knee little utility [21]. Surgical management should
implant. The strategy will be based on the condi- be as early as possible with lavage synovec-
tions in which the infection occurred. The diag- tomy and change of the polyethylene implant
nosis of infection is suggested based on clinical if possible (difficulty in cases of a solid or
findings and a warm swollen knee that is painful fully sealed polyethylene implant). Antibiotic
on mobilisation, weightbearing and may progress therapy should be the subject of a multidisci-
until total functional disability. General signs are plinary discussion with the infectious disease
not always present, and the clinical presentation specialist and bacteriologist. Either the micro-
may be asymptomatic, particularly for chronic organism is known upstream of surgery (nee-
forms. Conversely, the existence of a discharge or dle puncture, blood cultures) and antibiotic
fistula is almost always a point of certainty. therapy will be immediately targeted, or the
Needle puncture of the knee and a laboratory microorganism is unknown and antibiotic
assessment should be performed systematically. therapy will be probabilistic, broad spectrum
The speed of erythrocyte sedimentation and and by intravenous route, until the results of
C-reactive protein elevation can remain low, but perioperative samples. It should be remem-
various information should be crosschecked bered that survival at 1 year after conservative
(Table 16.3, [20]). treatment is 76% in a series by Chalmers [22].
The infection can be: In cases of chronic infection, imaging has a
more important place but its specificity is tricky
–– Acute: early postoperative (up to one month to assess (simple scintigraphy or radiolabelled
postoperatively) or haematogenic of early WBC, CT scan) and surgery consists of remov-
diagnosis (less than one month between the ing the implants. The strategy for surgery in
start of symptoms and diagnosis). one or two stages remains debated. TKR results
184 F. -X. Gunepin et al.

Table 16.3 From The Journal of Arthroplasty Vol. 27 No. 8 Suppl. 12,012
n Cutoff % Sensitivity % Specificity
ESR 172 21 mm/h 79,2 [73–85] 73,0 [66–80]
CRP 158 14 mg/L 82,6 [77–89] 80,7 [75–87]
Synovial WBC 96 6200/μL 90,0 [84–96] 96,5 [93–100]
PMN 91 60% 90,9 [85–97] 93,8 [89–99]
ESR: erythrocyte sedimentation rate = VS.
Synovial WBC: number of leucocytes per microliter of needle puncture fluid
PMN: polymorphonuclear neutrophils = altered white blood cells found in histological bone samples

after sepsis in UKA or TKR are identical irre-

spective of the strategy (77% survival at 2 years
for Bauer [23]).
The existence of risk factors increases the
probability of infection during initial surgery,
but also for revision surgery [21–23].

The occurrence of acute or chronic infec-

tion is always a serious event. Even with
optimal management, the risk of failure in
revision surgery is 25% at 2 years.

16.7 Specific Case of Dislocation

of the Polyethylene Implant

This complication is the result of mobile implants Fig. 16.3 Anteroposterior view of the knee with
and therefore requires a dedicated chapter. Its weightbearing and PE dislocation
incidence is low, assessed at 1.2 per 1,000
patients/year in the UK national register of causes is responsible in the case analysed.
implants. Fig. 16.4 can guide this analysis.
It should be considered based on sudden dete- Apart from defects in the implant’s design, the
rioration of the clinical result; this requires dislocation depends on two major categories:
prompt consultation with the patient. Diagnosis defects in mechanical stress of the implant com-
is confirmed by a clinical examination and X-ray partment and obstacles to proper movement of
assessment: anteroposterior view of the knee the insert.
(Fig. 16.3) profile and long-leg radiography. It is In these two major groups, we find causes
essential to have the surgical report with details related to the indication for UKA, a perioperative
of the implants inserted. technical issue and secondary causes, occurring
Once these details have been collected, it is most often sometime after implant surgery.
necessary to determine which of the possible
16 What to Do If a Medial Unicompartmental Knee Arthroplasty Fails 185

Dislocation of the mobile insert Design insert issue

Mechanical constraint issue Insert displacement issue

MCL insufficiency Imbalance Cam effect Abnormal mobility

Asymmetrical space Symmetrical space

between flexion and between flexion and
extension extension

Indication Voluntary
issue History of serious underdimensioning of
MCL sprains PE to limit stress on the
external compartment

Posterior cam effect Femoral malposition

– Posterior osteophyte 27% Mobility of the insert
– Meniscal remainder in rotation and loss of
Technical issue Improper evaluation of – Persistent fragment efficacy of the
Peroperative MCL Improper preparation for of cement anterio and posterior walls
the ligament tension
lesion 3% bone implants 30%
3% Tibial malposition
Anterior cam effect 24%
– Improper preparation Insert wall constraint
of the anterior bevel resulting in dislocation

Issue after an Chronic laxity after Wear of the insert Post-traumatic

Loosening 3%
interval UKA 3% 3% 4%

Fig. 16.4 Decision Tree in case of PE dislocation

Analysis of the causes of dislocation in an (Fig. 16.5), adjusting the stress on the implant
Asian series found 87% dislocation attributable depending on the cases.
to a perioperative technical issue. Once the cause
was determined, the latter guides the technique to In cases of asymmetry of the joint spaces in
use to restore optimal knee function. flexion and extension, revision by TKR makes it
possible to correct these differences during bone
Technical Conduct resection. Cases of malposition of an implant
The Oxford team first proposes, depending on the also require a change of implant.
cases, reduction by external procedure, which Lastly, in cases where no other anomaly is
makes it possible to resolve the issue under found, and the space in flexion and extension is
anaesthesia. In most cases, however, arthrotomy identical, it is necessary to determine if the insert
is the rule [24]. is too thin or too thick.
Different situations are to be differentiated for In most cases, the surgeon will try with an
revision on dislocation of an UKA insert: insert increased by 1 mm. Here too, testing of
complete mobility with the trial insert makes it
–– In cases with a CAM effect, which may be possible to verify the absence of dislocation by a
observed perioperatively in some cases, the “nutcracker” effect in flexion or extension.
procedure consists of removing the bone frag- In all cases where the UKA is kept, the crucial
ments or cement responsible for a conflict. stage lies in tests of the insert size in flexion and
Another insert can then be reintroduced and extension. It is necessary to make certain that the
control of good stability should be performed sensation of retention of the ancillary material
on complete mobility of the knee to ensure the size in flexion and extension is the same. It is
absence of any other cause. important to be wary of an insert that has to be
–– In cases of MCL insufficiency, the anatomical greatly increased in size, which can suggest rup-
conditions are no longer met to enable good ture of the MCL. In our experience, another
mechanical motion of the implant, therefore obstacle consists of first testing with the trial
revision should be planned with TKR mobile inserts. It can be very difficult to remove
186 F. -X. Gunepin et al.

Fig. 16.5 Constraint TKA with a metal augment in the medial compartment

them in some cases. They are used after testing in –– Technical defect (sagittal cut is too large or
flexion and extension with the size of the trial too high, impaction of the tibial implant).
insert of the ancillary material. –– Postoperative trauma.

In cases of osteosynthesis, use of a plate pro-

16.8 Fracture and Sinking duces better results according to Seeger et al. [25]
(Fig. 16.6) (Fig. 16.7).
A perioperative sagittal fracture during impac-
Fractures and sinking of the implant are rare tion of a tibial baseplate is not a systematic indica-
complications (0% to 10%) depending on the tion for a switch to TKR. If satisfactory
series and almost exclusively affect the tibia [17]. osteosynthesis is possible, UKA can be maintained.
The causes are multifactorial and often combine: Mobilisation and postoperative relief of weight-
bearing will be assessed on a case-by-case basis.
–– Defect in indication (osteoporosis or major If the indication for TKR is chosen, the selec-
osteopenia, BMI, axis). tion criteria will be those mentioned in Table 16.2.
16 What to Do If a Medial Unicompartmental Knee Arthroplasty Fails 187

Fig. 16.6 Fracture and sinking. It is necessary to analyse with/without mobilisation. In cases of fracture with dis-
the stability of the tibial implant. If stability of bone/ placement, osteosynthesis can be considered if the implant
implant is maintained and the fracture is not displaced, has remained solid with the bone. If not, it will be neces-
treatment can be conservative with relief of weightbearing sary to plan revision of the UKA

Fig. 16.7 Postoperative CT scan for unexplained pain and post-osteosynthesis control

detect the cause is vital because revision surger-

16.9 Conclusion
ies for unexplained pain are those which produce
less optimal results [26].
The number of UKA is constantly increasing due
Recourse to total arthroplasty is the most
to this surgical procedure’s excellent functional
widely used solution. In the tibia, it is necessary
results. Failures are also increasing too, although
to use implants sealed with a stem and augment
the curve is not parallel because of relevant indi-
whenever necessary.
cations and reliable techniques.
Failure occurs most often due to loosening of
the tibial implant, but attentive screening to
188 F. -X. Gunepin et al.

References and combined uni plus patellofemoral replace-

ment: indications and surgical technique. Joints.
2015;3(1):42–8.
1. Johal S, Nakano N, Baxter M, Hujazi I, Pandit
13. Epinette JA, Leyder M, Saragaglia D, Pasquier G,
H. Unicompartmental knee arthroplasty: the
Deschamps G, Société Française de la Hanche et du
past, current controversies, and future perspec-
Genou. Is unicompartmental-to-­ unicompartmental
tives. J Knee Surg. 2018;31(10):992–8. https://doi.
revision knee arthroplasty a reliable option ?
org/10.1055/s-­0038-­16255961.
Case-control study. Orthop Traumatol Surg Res.
2. Hansen EN, Ong KL, Lau E, Kurtz S, Lonner
2014;100(1):141–5. https://doi.org/10.1016/j.
JH. Unicondylar knee arthroplasty has fewer com-
otsr.2013.10.013.
plications but higher revision rates than total knee
14. Weißenberger M, Petersen N, Bölch S, et al. Revision
arthroplasty in a study of large united states data-
of unicompartmental knee arthroplasty using the in
bases. J Arthroplast. 2019;34(8):1617–25. https://doi.
situ referencing technique. Oper Orthop Traumatol.
org/10.1016/j.arth.2019.04.004.
2020;32(4):273–83. https://doi.org/10.1007/
3. Australian Joint Registry: AOANJRR. https://aoanjrr.
s00064-­020-­00656-­w.
sahmri.com/documents/10180/677574/Table+KP8.
15. Lewis PL, Davidson DC, Graves SE, De Steiger
png.
RN, Donnelly W, Cuthbert A. Unicompartmental
4. Kennedy JA, Palan J, Mellon SJ, Esler C, Dodd CA,
knee arthroplasty revision to TKA: are tibial stems
Pandit HG, Murray DW. Most unicompartmental
and augments associated with improved survivor-
knee replacement revisions could be avoided: a radio-
ship? J Arthroplast. 2019;34(8):1617–25. https://doi.
graphic evaluation of revised Oxford knees in the
org/10.1016/j.arth.2019.04.004.
National Joint Registry. Knee Surg Sports Traumatol
16. Scott CE, Powell-Bowns MF, MacDonald DJ,
Arthrosc. 2020;28:3926–34. https://doi.org/10.1007/
Simpson P, Wade F. Revision of unicompartmen-
s00167-­020-­05861-­5.
tal to total knee arthroplasty: does the unicompart-
5. Epinette JA, Brunschweiler B, Mertl P, Mole D,
mental implant (metal-backed vs. all-polyethylene)
Cazenave A. French Society for Hip and Knee: uni-
impact the total knee arthroplasty? J Arthroplast.
compartmental knee arthroplasty modes of failure:
2018;33(7):2203–9. https://doi.org/10.1016/j.
wear is not the main reason for failure: a multicentre
arth.2018.02.003.
study of 418 failed knees. Orthop Traumatol Surg Res.
17. Crawford DA, Berend KR, Lombardi AV. Management
2012;98(6 Suppl):S124–30. https://doi.org/10.1016/j.
of the failed medial unicompartmental knee arthro-
otsr.2012.07.002. Epub 2012 Aug 24
plasty. J Am Acad Orthop Surg. 2018;26(20):e426–
6. Dahl W, Robertsson O, Lidgren L, Miller L, Davidson
33. https://doi.org/10.5435/jaaos-­d-­17-­00107.
D, Graves S. Unicompartmental knee arthroplasty in
18. Marinier ES, Peltier A, Gaillard R, Cheze L, Servien
patients aged less than 65. Acta Orthop.
E, Neyret P, Lustig S. Conséquence de la hauteur de
2010;81(1):90–4. https://doi.org/10.3109/174536710
la coupe osseuse tibiale sur la laxité du genou dans le
03587150.
plan frontal : étude biomécanique cadavérique. Revue
7. Goodfellow J, Goodfellow J, O’Connor J, Pandit HG,
de Chirurgie Orthopédique et Traumatologique.
Dodd C, Murray DW. Unicompartmental arthroplasty
2017;103(7):S27. https://doi.org/10.1016/j.
with the oxford knee. Oxford: Goodfellow Publishers
rcot.2017.09.015.
Lim; 2015.
19. Leta TH, Lygre SH, Skredderstuen A, Hallan G,
8. Mertl P, Cazenave A. https://www.sfhg.fr/accueil/
Gjertsen J-E, Rokne B, Furnes O. Outcomes of
fiches-­d-­information/allergies-­et-­prothèses/.
unicompartmental knee arthroplasty after asep-
9. Dietrich KA, Mazoochian F, Summer B, Reinert
tic revision to total knee arthroplasty. J Bone Joint
M, Ruzicka T, Thomas P. Intolerance reactions to
Surg. 2016;98(6):431–40. https://doi.org/10.2106/
knee arthroplasty in patients with nickel/cobalt
jbjs.o.00499.
allergy and disappearance of symptoms after revi-
20. Society of Unicondylar Research and Continuing
sion surgery with titanium-based endoprostheses. J
Education. Diagnosis of periprosthetic joint infec-
Dtsch Dermatol Ges. 2009;7(5):410–3. https://doi.
tion after unicompartmental knee arthroplasty.
org/10.1111/j.1610-­0387.2008.06987.x. Epub 2009
J Arthroplast. 2012;27(8):46–50. https://doi.
Jan 15
org/10.1016/j.arth.2012.03.033.
10. Stejskal V, Hudecek R, Stejskal J, Sterzl I. Diagnosis
21. de Santé HA. Recommandations de la Haute Autorité
and treatment of metal induced side effects. Neuro
de Santé. Prothèse de hanche ou de genou: diagnostic
Endocrinol Lett. 2006;27(Suppl 1):7–16.
et prise en charge de l’infection dans le mois suiv-
11. Servien E, Merini A, Lustig S, Neyret P. Lateral
ant l’implantation : Méthode Recommandation pour
uni-compartmental knee replacement: current
la pratique clinique. 2014. https://www.has-­sante.fr/
concepts and future directions. Knee Surg Sports
upload/docs/application/pdf/2014-­03/rbp_argumen-
Traumatol Arthrosc. 2013;21(11):2501–8. https://doi.
taire_prothese_infectees_vd_.pdf.
org/10.1007/s00167-­013-­2585-­x.
22. Chalmers B, Kapadia M, Chiu Y, Henry M, Miller
12. Romagnoli S, Marullo M, Massaro M, Rusteni E,
A, Carli A. Treatment and outcome of periprosthetic
D’Amario F. Corbella M : Bi-unicompartmental
joint infection in unicompartmental knee arthro-
16 What to Do If a Medial Unicompartmental Knee Arthroplasty Fails 189

plasty. J Arthroplast. 2020;35(7):1917–23. https://doi. tions. J Arthroplast. 2020;35(1):265–71. https://doi.

org/10.1016/j.arth.2020.02.036. org/10.1016/j.arth.2019.08.004.
23. Bauer T, Piriou P, Lhotellier L, Leclerc P, Mamoudy 25. Seeger JB, Jaeger S, Röhner E, Dierkes H, Wassilew
P, Lortat-Jacob A. Résultats des changements de pro- G, Clarius M. Treatment of periprosthetic tibial pla-
thèse de genou pour infection multicentrique portant teau fractures in unicompartmental knee arthroplasty:
sur 107 cas d'infections sur prothèse totale de genou. Plates versus cannulated screws. Arch Orthop Trauma
Revue de Chirurgie Orthopédique et Traumatologique. Surg. 2013;133:253–7.
2006;92(7) 26. Kerens B, Boonen B, Schotanus MG, Lacroix H,
24. Bae JH, Kim JG, Lee SY, Lim HC, Yong I, MUKA Emans PJ, Kort NP. Revision from unicompart-
Study Group. Epidemiology of bearing dislocations mental to total knee replacement: the clinical out-
after mobile-bearing unicompartmental knee arthro- come depends on reason for revision. Bone Joint J.
plasty: multicenter analysis of 67 bearing disloca- 2013;95–B:1204e8.
 Results and Registry Data
for Unicompartmental Knee 17
Replacements

A. Rahman, A. D. Liddle, and D. W. Murray

The results of unicompartmental knee replace- the clinical results of UKR in general and focus
ment can be gathered from three main sources: on the results of the Oxford UKR (OUKR) in
reports from national registries, observational more detail, as there are many more publications
studies, and randomised controlled studies. These on this implant than any other in current use.
all have advantages and disadvantages.
National registries have very large numbers,
but tend only to track a single outcome measure 17.1 Registry-Based Comparisons
rate of revision. As the numbers are very large, of UKR and TKR:
statistically significant associations are often Interpretations
found, but these do not imply causation. There and Limitations
are also large numbers of observational studies.
Although most are short term, we will focus on All national registries have found that the revi-
those reporting 10-year outcomes or more. There sion rate of UKR is about three times that of
are very few randomised studies available. While TKR. As a result, they tend to conclude that UKR
their results are very reliable, they tend to have has a poorer outcome than TKR, discouraging
highly selected populations, and hence are not surgeons from using UKR. This conclusion is not
necessarily generalisable to all patients and justified; the main reason the revision of UKR is
surgeons. higher than that of TKR is that the threshold for
We will first compare the results of unicom- revision is much lower. Therefore, higher revi-
partmental (UKR) and total (TKR) knee replace- sion rate of UKR does not necessarily suggest
ments, as this is critical in determining whether that UKR has a worse outcome than TKR.
UKR should be done at all. We will then review The New Zealand Joint Registry (NZJR) col-
lects data about revision rates and post-operative
Oxford Knee Score (OKS) 6 months after the
A. Rahman (*) · D. W. Murray operation. The OKS assesses overall knee joint
Oxford Orthopaedic Engineering Centre, NDORMS, pain and function and is categorised into ‘Poor’,
University of Oxford, Oxford, UK
‘Fair’, ‘Good’ and ‘Excellent’ [1]. Data from the
Nuffield Orthopaedic Centre, Oxford, UK NZJR demonstrates that UKR not only have
e-mail: [email protected]
more Excellent results, but also fewer Poor
A. D. Liddle results than TKR (Fig. 17.1). Therefore, the
MSk Lab, Imperial College London, London, UK
higher revision of UKR cannot be because UKR
Department of Trauma and Orthopaedics, Imperial has a poorer outcome [2].
College Healthcare NHS Trust, London, UK

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2024 191
A. Clavé, F. Dubrana (eds.), Unicompartmental Knee Arthroplasty,
https://doi.org/10.1007/978-3-031-48332-5_17
192 A. Rahman et al.

Fig. 17.1 TKR revision

rate classified by Oxford
knee score categories,
based on data from the
New Zealand Joint
Registry. Graph adapted
from Goodfellow et al.
(2010) [2]

Fig. 17.2 Revision

rates for UKA and TKA
having different PROMS
results at 6 months
post-surgery [2]

The NZJR also compares patients’ 6-month This very large difference is not surprising
OKS with their subsequent revision rate because a revision of UKR is usually a simple
(Fig. 17.2). It found that whatever the outcome conversion to a primary TKR and the outcome
score, the subsequent revision rate of UKR is would generally be expected to be good, whereas
about five times higher than that of TKR. This a revision of a TKR is usually complex, requiring
suggests that factors independent of outcome stems, wedges, and stabilised implants, and the
scores increase the UKR revision rate. These outcome is unpredictable. We therefore conclude
factors would be closely related to the threshold that the higher revision rate of UKR is not
for revision. The most striking difference in because they have poorer outcomes (Fig. 17.1
revision rate occurs in patients who have a demonstrates UKRs have better outcomes at least
worse score post-operatively than pre-opera- in the short and medium term), but because they
tively (i.e. those who have a post-operative OKS have a lower threshold for revision.
less than about 20). These patients have a 10% Most surgeons would agree that the relative
chance of being revised if they have a TKR, and ease of revision (if there was a problem) is an
a 60% chance of being revised if they have a advantage of the UKR over the TKR. The conse-
UKR. quence of it being easy to revise is that the thresh-
17 Results and Registry Data for Unicompartmental Knee Replacements 193

old for revision is lower, and therefore the and the transfusion rate (OR 0.25, CI 0.17–
revision rate is higher. The higher revision rate of
0.37) were all less. Complications also occurred
UKR should thus not be considered to be a seri- less frequently with UKR: for example, the
ous problem, as it is a manifestation of an odds ratio of thromboembolism was 0.49 (CI
advantage. 0.39–0.62), infection was 0.5 (CI 0.38–0.66),
Patients may have a poor result after UKR or stroke was 0.37 (CI 0.16–0.86), and myocardial
after a TKR. If a patient has a poor result after a
infarct was 0.53 (CI 0.30–0.90).
TKR, the knee will probably not be revised. The mortality following UKR was also sig-
National registries will classify this to be a suc-
nificantly lower than following TKR. In the first
cess, whereas the patient will consider this to be
30 days, the hazard ratio was 0.23 (CI 0.11–0. 50,
a failure. Conversely, if a patient has a poor result
p < 0.001); in the first 90 days, it was 0.46 (CI
following a UKR, it will probably be revised and0.31–0.69, p < 0.001). This difference in mortal-
have a successful outcome. National registries ity was not just observed in the short term. The
will classify this as a failure, whereas the patient
survival curves progressively separated for
will consider this to be a success. A patient-­ 4 years and thereafter remained parallel, suggest-
centred approach to knee replacements should ing the effect of surgery on mortality lasted for
place a greater emphasis on patient beliefs rather
4 years. At 8 years, the mortality following UKR
than registry conclusions. was 0.87 (CI 0.80–0.94 p < 0.001) that of TKR
(Fig. 17.3a, b).
In this matched comparison, it was found that
17.2 Matched Comparisons the revision rate in UKR was 2.1× that of TKR,
with UKR and TKR and the overall reoperation rate was 1.3× higher.
Registry Data To put the adverse outcomes in perspective, it
was concluded that if 100 patients receiving TKR
When national registries compare different received a UKR instead, the results would be
implants, they usually analyse unmatched data. around one less death and three more revisions in
However, UKR is generally implanted in the first 4 years after surgery.
younger and fitter patients than TKR [3, 4]. Liddle et al. in a separate matched study com-
Younger and fitter patients are more likely to pared the patient-reported outcomes of about
have higher revision rates and lower complica- 15,000 UKR and TKR [7]. The primary outcome
tion rates. Hence, a fair comparison between measure was the post-operative Oxford knee
UKR and TKR requires matched patients. score at 6 months after the operation. The OKS
Liddle et al. (2014) compared adverse events in was significantly better for the UKR (UKR 38 vs.
matched UKR and TKR [5]. The data was TKR 36, p < 0.0001). The difference in OKS is
obtained from the National Joint Registry of relatively small; however, many more patients
England, Wales, Northern Ireland and the Isle achieved excellent OKS (>41) with UKR (odds
of Man (NJR) and other national databases. ratio 1.59, CI 1.47–1.73, p < 0.001). EQ-5D data
Over 100,000 UKR and TKR were matched at a was also collected, and a significantly better
1:3 ratio using propensity score analysis on 20 overall score was achieved with UKR (p < 0.001).
outcome measures. It was found that there were The four subscales relating to mobility, pain,
many advantages of UKR compared to function, and self-care were significantly better,
TKR. For example, the length of stay was 1.38 and no statistical difference was found in the anx-
(CI 1.33–1.43) days shorter, and the re-­ iety subscale. The level of patient satisfaction
admission rate within the first year (incidence was also assessed, and patients were 1.3× more
rate ratio 0.65, CI 0.58–0.72), the intraopera- likely to be report excellent satisfaction with
tive complication rate (OR 0.73 CI 0.58–0.91), UKR than TKR.
194 A. Rahman et al.

Fig. 17.3 Post-­ a

operative mortality for
matched UKA and TKA
(a) for the first year and
(b) for 8 years, from UK
registry data [6].
(Courtesy of AD Liddle)

Nearly all other outcome measures favoured

17.3 Randomised Control Trials the UKR, and several were statistically signifi-
Comparing UKR and TKR cant: at 5 years, EQ-5D VAS was 75.4 vs. 71.1
(p = 0.004), self-reported knee improvement was
The 5-year results of the Total Or Partial Knee 95.2% vs. 90.1% (p = 0.016), and self-reported
Arthroplasty Trial (TOPKAT) were recently pub- likelihood of willing to have operation was 91.2%
lished [8]. In this randomised study of 528 vs. 84.3% (p = 0.010).
patients, the primary outcome measure was the The most remarkable finding of the study was
OKS. UKR was significantly better than TKR at that the revision rate of UKR was not higher than
1 year after surgery, and better at 5 years (though that of TKR at 5 years. When analysed based on
the difference was not statistically significant). intention to treat, it was actually lower following
UKR was also significantly better when the UKR, although not significantly (3% n = 8 v 5%
whole 5 year period was taken together. The most n = 12), and when analysed based on the treat-
marked difference was a higher proportion of ment it was the same (4% n = 10 v 4% n = 10).
patients achieving ‘Excellent’ outcomes Various other endpoints were studied, including a
(OKS > 41) [9]. composite failure measure, defined as ­reoperation
17 Results and Registry Data for Unicompartmental Knee Replacements 195

or revision surgery or no appreciable improve- 2 weeks, UKR had significantly better NPRS (3.7
ment in OKS. At 5 years, there were 26 (9.9%) vs. 7.8, p < 0.001), KSS (86.5 vs. 81.4, p < 0.001),
failures in the UKR group and 37 (14.0%) fail- and FJS (90.5 vs. 79.5, p < 0.001). These signifi-
ures in the TKR group (p = 0.118). The death rate cant differences persisted at 6 weeks.
at 5 years was also lower following UKR (2.3%, Considerably more literature is available com-
n = 6) than following TKR (4.2%, n = 11), paring non-matched cohorts. However, readers
although the difference was not significant. should take greater caution in interpreting find-
A detailed health economic analysis was also ings. A 2019 meta-analysis analysed 36 cohort
undertaken, and this showed that UKR was both studies covering a wide range of outcome mea-
more effective (0.24 additional quality adjusted sures [13]. It found UKR procedures were 23.8
life years, 95% CI 0.046 to 0.434), and had lower (CI 9.8–37.8) minutes shorter than TKR, required
healthcare costs for surgery and aftercare (−£910, 1.7 (CI 1.2–2.3) days shorter hospital stays,
95% CI −1503 to −317) than TKR, during the enabled 8.7 (CI 5.6–11.8) degrees greater range
5 years of follow-up. of movement, and had better pain and function
A longer but small 15-year randomised con- scores.
trolled trial comparing 52 fixed bearing UKR and
50 TKR found similar outcomes: patients with
UKR reported higher Excellent Bristol knee 17.5 Decreasing the Revision Rate
scores (71.4% UKR vs. 52.6% TKR) and higher in UKR
survivorship based on revision or failure (89.8%
vs. 78.7%). However, this study was underpow- A striking finding is that in registry studies, UKR
ered to test for statistical significances [10]. has substantially higher revision rates than TKR,
whereas in randomised studies and matched
cohort studies, there are no marked differences.
17.4 Comparative Cohort Studies There can be various reasons for this, but proba-
Between UKR and TKR bly the most important are surgeon-related
factors.
We were only able to identify a few matched In national registries, most surgeons are
cohort studies. found to be doing very small numbers of UKR,
Burn et al. (2018) matched 590 UKR to 590 whereas in published series, surgeons tend to do
TKR from prospective cohorts and assessed OKS large numbers of UKR. The data from the NJR
and EQ-5D outcomes over 10 years after surgery would suggest that about half the surgeons
[11]. At 1 year, UKR patients reported signifi- doing knee replacement do some UKR [6]. For
cantly better OKS [40.3 (CI 39.5–41.0) vs. 35.9 those doing UKR, the most common number
(CI 35.0–37.6)] and EQ-5D [0.82 (CI 0.80–0.83) implanted per year is one, the second is two, and
vs. 0.74 (CI 0.72–0.76)] scores. When OKS was the third is three; the average number is five
divided into pain and function sub-scores, both (Fig. 17.4).
remained significantly better for UKR [OKS When the number of UKR performed per sur-
pain: 23.8 (CI 23.3–24.2) vs. 22.0 (21.5–22.5), geon per year was compared to revision rate, it
OKS function: 16.5 (CI 16.2–16.7) vs. 14.1 was found that surgeons doing small numbers
(13.8–14.5)]. These differences persisted and had a high revision rate. Surgeons doing one or
remained statistically significant throughout the two UKR per year have a 4% failure rate per year,
10 years. which would equate to about 60% survival at
A recent study compared the early post-­ 10 years. The revision rate dramatically decreases
operative outcomes of matched 150 UKR and with increasing numbers. Surgeons doing about
150 TKR, assessing the Numeric Pain Rating 10 UKR per year have a revision rate of 2% per
Scale (NPRS), American Knee Society Score year, whereas those doing ≥30 UKR per year
(KSS), and Forgotten Joint Score (FJS) [12]. At have a revision rate of 1% per year (Fig. 17.5).
196 A. Rahman et al.

Fig. 17.4 Histogram

demonstrating the
distribution of UKR
caseload among
surgeons performing
UKR in England and
Wales [6]. (Courtesy of
AD Liddle)

Fig. 17.5 LOWESS

curve demonstrating the
effect of increasing
caseload on revision
rates following UKR (up
to 50 cases), based on
NJR data [6]. (Courtesy
of AD Liddle)

Surgeons cannot easily increase the size of [4]. Surgeons doing less than this have a very
their knee replacement practice. Hence, the only high revision rate, and they should either con-
way they can increase the numbers of UKR they sider stopping doing UKR, or doing more. A
do is by increasing the proportion of UKR in minimum acceptable usage is about 20% [16].
their knee replacement practice, which we have With the Oxford UKR (OUKR), the revision rate
defined as ‘usage’ of UKR. Figure 17.6 shows decreases until surgeons are doing about 50% of
the relationship between revision rate and usage their knees as UKR (Fig. 17.7). At this rate, a
of UKR for the Oxford knee based on NJR data. matched study of UKR and TKR shows that the
Fixed bearing UKR has a relatively similar curve revision and reoperation rates of UKR and TKR
that drops to a minimum at 20%, and then steadily are similar [14].
increases rather than dropping further [14]. Hamilton, in a meta-analysis of the outcome
In Fig. 17.6, the shape of the graph is not what of published studies of the OUKR compared the
would be expected. As Kozinn and Scott’s (1989) influence of caseload and usage on revision rate
ideal indications for UKR are satisfied in perhaps [16]. He found that high-usage (≥20%) surgeons
5% of patients [15], one would expect the revi- had low revision rates, whether their caseload
sion rate to increase above 5% usage, but it does was high or low, and conversely low-usage
not. The average usage in the NJR is about 10% (<20%) surgeons had high revision rates whether
17 Results and Registry Data for Unicompartmental Knee Replacements 197

Fig. 17.6 Annual

revision rate plotted
against the proportion of
a surgeon’s knee
replacement practice
that are UKR [14].
(Courtesy of AD Liddle)

their caseload was high or low.1 This suggests contact, UKR is also unreliable and should not be
that usage is more important than caseload in performed [20].
decreasing revision rate. A 2018 health economics study compared the
Usage is a manifestation of indications. lifetime health and financial benefits of UKR to
Surgeons following very narrow indications will TKR and how they vary with surgeon usage [8].
have a very low usage and will probably have Overall, UKR dominated TKR for both males
poor results, whereas surgeons following broad and females in all age groups as it provided a
indications will have high usage and probably greater lifetime health gain at a lower cost.
better results. The indications for fixed bearing However, surgeon usage had a marked effect on
UKR are relatively poorly defined. However with the findings. If the usage was <10%, then there
the mobile-bearing OUKR, the indications are was no lifetime health gain of UKR over TKR
well defined, evidence based, and are satisfied in [∆QALY: −0.04 (CI −0.32 to 0.21)] even though
about 50% of cases needing knee replacements the costs were less [∆Costs £ − 127 (CI −429 to
[17, 18]. Therefore, to obtain optimal results with 127)]. If the usage was >10%, there were sub-
the OUKR, surgeons should adhere to the recom- stantial benefits of UKR with both greater life-
mended indications and have a usage of over time health gain [∆QALY: 0.26 (CI 0.12 to 0.40)]
20% and ideally somewhere in the region of 50%. and lower lifetime costs [∆Costs (£ − 758 (−939
Surgeons with low usage are likely to be using to −579)].
UKR in inappropriate patients. There is a com-
mon but incorrect perception that UKR should be
used for early arthritis, when the surgeon feels 17.6 Long-Term Results of UKR
disease in not severe enough for TKR, and where (≥20 Years)
the surgeon expects a TKR may not perform well
[19]. In early arthritis, without bone-on-bone The longest series of UKR we are aware of is a
series of 125 medial mobile-bearing Phase 1
1
In this meta-analysis, high caseload was defined as >12 Oxford UKR implanted in 104 patients by Dr.
UKR per year, and low caseload was defined as ≤12 UKR U. Svard and his colleagues in Sweden begin-
per year.
198 A. Rahman et al.

ning in 1983 (Table 17.1). No patients were lost 10-year survival of 88% in 461 implants [25],
to follow-up. In their final review, all but two and from those reaching 10 years, Steele
patients were revised or deceased. The two that reported that 86% survived to 20 years in 203
were alive were followed up at over 30 years. In implants [26]. The overall survival at 20 years
this study, failure was defined as either a revision can hence be calculated to be approximately
or a ‘poor’ result (based on the HSS knee score) 75%.
at last follow-up or death. Overall, the Oxford
knee was a successful and definitive knee
replacement in 84% of patients over their life- 17.7 Mid-Term Results for Fixed
time [21]. As far as we know, no other knee Bearing UKR (~10 Years)
replacement has had such complete follow-up,
and such good lifetime results. Prior to this pub- Most of the studies reporting 10-year results of
lication, a standard follow-up had been done of fixed bearing UKR are on the Miller-Galante,
his 683 UKR and these achieved a 20-year sur- Marmor, and St Georg UKRs, which are no lon-
vival of 91% [22]. The most common mode of ger commonly used (Table 17.2). A wide range
failure was progression of disease laterally, but of 10-year survivals, from 70% to 94%, have
remarkably, at 20 years this had only occurred in been reported in various cohorts. The only fixed
2.3% of patients. This demonstrates that if the bearing UKR implants in common use, as
operation is done appropriately in appropriate reported by the NJR, with published 10-year
patients, progression of arthritis is not inevitable. survival is the ZUK (now sometimes called
This comes as important evidence for surgeons Physica ZUK). There are many other devices in
who do not use UKR, as most are concerned use for which 10-year data is not currently
about revision for progression of arthritis. available.
There are publications of 20-year results of The ZUK is an evolution of the Miller-­
three fixed bearing devices (Table 17.2). The Galante. Three publications report the 10-year
20-year survival of the Miller-Galante UKR was survival rates which range between 94 and 98%
reported by Argenson to be 74% (CI 67–71) in [27–29] (Table 17.4). Unfortunately, none of the
an independent series of 160 implants [23]. papers reports the number of cases at risk at
Common causes of failure in the second decade 10 years, so it is impossible to assess the reliabil-
were patella femoral problems and wear. A ity of these estimates. The numbers at risk may
designer series of 68 Miller-Galante UKR by have been relatively low as the average follow-up
Foran had a higher 90% survival [24]. A single was relatively short. Only one paper quoted the
surgeon series of 103 Marmor UKR was fol- ‘loss to follow up’ rate and one appeared to over-
lowed by O’Rourke for a minimum of 21 years, estimate the survivorship by incorrectly assum-
with a survival of 84% (CI 76–92%) at 20 years, ing it was the same the percentage of cases not
and 72% (CI 58–95%) at 25 years. A series of St revised. The indications and contraindications
Georg Sled UKR from Bristol has been reported used for UKR by the different groups are also
at various time intervals. Ansari reported a different.
Table 17.1 Survival of the Phase 1 and Phase 2 Oxford UKR (10-year results)
Principal surgeon or No. of
Implant author Date References No. of knees Age Time (years) Survival (%) revisions Reasons for revision
Phase 1–2 U Svard 2006 [22] 683 70 20 92
Phase 1–2 Svard 2001 [44] 124 70 10 95
Phase 2 Emerson 2010 54 64 20 84 9
Phase 1–2 Emerson 2008 [45] 54 64 10 85 5 Disease progression (7),
loosening (1), impingement (1)
Phase 1–2 Murray 1998 [46] 143 71 10 98 1 Disease progression (2),
infection (1), pain (1),
loosening (1)
Phase 2 Rajesekhar 2004 [47] 135 70 10 94
Phase 1–2 Kumar 1999 [48] 100 71 10 85 7 Patient selection (see text, 4),
disease progression (2), fracture
(1)
Phase 1–2 Price 2005 [49] 52 56 10 91
Phase 1–2 Price 2005 [49] 512 71 10 96
17 Results and Registry Data for Unicompartmental Knee Replacements

Phase 2 Vorlat 2006 [50] 149 66 10 84 24

Phase 1–2 Koskinen 2007 [51] 1113 64 10 85
Phase 1–2 Lidgren 2010 [52] 749 10 86
Phase 1–3 Price 2011 [53] 683 70 15 92 29 Disease progression (10),
loosening (9), infection (5),
pain (3), bearing dislocation (2)
199
200 A. Rahman et al.

17.8 Mid-Term Results for Mobile-­ implant (1.25%), dislocation of the mobile bear-
Bearing OUKR (~10 Years) ing (0.58%), and pain (0.57%). Complications
were rare at 0.83%. The studies reported an aver-
The Phase 3 Oxford UKR is the version of the age weighted 10-year OKS score of 40 (out of
OUKR that is currently used. It is available with 48) and an average weighted KSS-objective score
both cemented fixation, introduced in 1998, and of 86 (out of 100).
cementless fixation, introduced in 2004 Before the cementless Oxford Phase 3 UKR
(Table 17.3). A review of Oxford Phase 3 UKR in was generally released, 2 small randomised trials
2017 found 15 studies which had a follow-up of were undertaken [31, 32]. These showed similar
10 years or longer. It assessed a total of 8658 second year migration, as assessed by radio-­
implants and found a 10-year survival of 93% stereometric analysis, with both cemented and
and a 15-year survival of 89% [30]. Revisions cementless components. In addition, they found
were due to progression of the arthritis to the lat- markedly decreased incidence of tibial radiolu-
eral component (1.42%), aseptic loosening of the cent lines with the cementless components, sug-
Table 17.2 Survival and outcomes for fixed-bearing UKR (10-year results)
Principal Number of
surgeon or knees Follow-up Survival No. of Outcome
Implant author Date References (Lateral) Age (years) (%) revisions Reason for revision measure Latest score
Genesis Heyse 2012 [54] 261 (78) 54 10 94 15 Wear/loosening (6), disease KSS-­ 92.0–97.2
progression (4), pain (4), instability (2) Fcn
Marmor O’Rourke 2005 [55] 136 71 21 84 19 Disease progression (9), loosening (8), KSS-­ 53
(minimum) pain (2) Fcn
Marmor Squire 1999 [56] 140 (15) 68 15 87 14 Disease progression (7), tibial KSS-­ 71
subsidence (6), pain (1) Obj
Marmor Tabor 1998 [57] 67 61 10 84 11 Subsidence (6), disease progression KSS-­ 77 (5–100)
(2), inflammatory disease (2), not Fcn
stated (1)
Marmor Cartier 1996 [58] 207 65 10 93 7 Not given KSS 75%
(both) ‘excellent’
Marmor, Heck 1993 [59] 294 (39) 68 10 91 16 Loosening (11), disease progression HSS 50%
C1,C2 (4), infection (1) ‘excellent’
Marmor Marmor 1988 [60] 60 (7) 63 10 70 21 Loosening (11), disease progression Various -
(8), other (2)
Miller-­ Argenson 2013 [23] 160 66 20 74 19 Disease progression (12), loosening KSS-­ 88
Galante (2), wear (5), infection (1 Fcn (45–100)
Miller-­ Rachha 2013 [61] 74 64 10 93 5 Disease progression (2), pain (2), KSS-­ 75.5
Galante infection (2) Fcn (45–90)
Miller Foran 2012 [24] 62 (3) 68 15 93 5 Disease progression (2), pain (1), HSS 80%
17 Results and Registry Data for Unicompartmental Knee Replacements

Galante component dissociation (1), not stated ‘excellent’

(1)
Miller-­ John 2011 [62] 94 (9) 67 10 94 7 Disease progression (5), loosening (2) BKS 43.6
Galante (28–50)
Miller-­ Naudie 2004 [63] 113 68 10 90 11 Disease progression (4), wear (3), KSS-­ 80
Galante loosening (2), pain (1), traumatic Fcn (20–100)
ligament rupture (1)
St Georg Ansari 1997 [25] 461 70 10 88 19 Not given Not
given
St Georg Steele 2006 [26] 203 67 10–20 86 18 Disease progression (7), bearing wear Not
(second (3), aseptic loosening (4), component given
decade) fracture (2), infection (2)
UNIX Hall 2013 [64] 85 (20) 65 10 92 Not given OKS 38
201
202 A. Rahman et al.

gesting improved fixation. A multi-centre 10-year cemented OUKR was modified to include the same
study of the first 1000 cementless UKR found design. The first 100 cases of Dual Peg cemented
97% survival (CI 92–100%), and no significant OUKR had good outcomes similar to the Phase 3
differences in survival or clinical outcome [38]. A 5-year matched registry study of 2834
between designer and independent centres [33]. Single Peg and 2834 Dual Peg cemented Oxfords
In 2019, a single-centre case series of 1000 found a 26% decrease in revision (Single Peg 5.2%,
cementless OUKR found a 10-year survival of Dual Peg 3.8%), with significant >50% reductions
98% (CI 96–99%), with a mean 10-year OKS of in revisions for aseptic loosening (Single Peg 0.4%,
41, and KSS-objective of 89 [34] (Table 17.3). Dual Peg 0.1%, p = 0.03) and pain (Single Peg
Multiple studies have directly compared the 0.8%, Dual Peg 0.3%, p = 0.01) [39].
cemented and cementless OUKR. A 10-year In 2012, new microplasty instrumentation was
matched registry study of 14,814 OUKR found introduced, facilitating improved positioning of
significantly greater survival [cemented 90% (CI the femoral component and preventing impinge-
88–92%), cementless 93% (CI 90–96%)]. Of the ment. Comparative cohort studies found the
causes of revision, the greatest improvement was microplasty instrumentation reduce average sur-
in implant loosening (58% reduction, from 1.00% gery time by 15%, reduce the time range of pro-
in cemented to 0.42% in cementless)[35]. A fol- cedures [40, 41], reduce malalignment [41, 42],
low-­up study of these 14,814 implants was done and reduce the rate of bearing dislocation [42].
to assess the effect of surgeon caseload on out- The broader clinical benefits were established by
comes. It found that cementless Oxford had lower a five-year matched registry study of 7953 micro-
risk of revision across all surgical caseload plasty and 7953 non-microplasty procedures,
groups (Hazard Ratios: low-volume 0.74, which found a significant 31% reduction in revi-
medium-volume 0.79, high-volume 0.80)2 [36]. sion rate with microplasty (97% vs. 95%, Hazard
A separate study comparing detailed patient Ratio 0.77, p = 0.008) [43].
outcomes for 267 cemented and 278 cementless
OUKR at 5 years found superior outcome scores
for the cementless with OKS (43 vs. 41, p = 0.008) 17.9 Registry Reports on Fixed
and EQ-5D-5L index (0.87 vs. 0.81, p = 0.0001, and Mobile-Bearing UKR
higher is better). The most remarkable difference,
however, was in pain. Four independent pain mea- In 2019, the UK National Joint Registry found
sures recorded significantly less pain with the that the Oxford UKR comprises more than half all
cementless: ICOAP (5 vs. 11, p < 0.0001, lower is UKR performed in the UK, at 68098. This is fol-
better), OKS pain (18.2 vs. 16, p < 0.0001, higher lowed by the ZUK UKR at 14973, and Sigma HP
is better), AKSS pain (46.2 vs. 43.1, higher is bet- UKR at 10445, both of which have a fixed bear-
ter), and EQ-5D (0.492 vs. 0.789, p < 0.0001, ing. Other implants had a total implant number
lower is better). Across all patients, 61% of those below 10,000 [3]. The general conclusion from
with the cementless OUKR had no pain, compared registry data is that both good mobile and fixed
to 43% of those with the cemented [37]. bearing devices perform very well, on a national
Following the success of the cementless OUKR basis, at least up to 10 years. With the mobile-­
(which had a dual peg femoral component), the bearing device, the indications are well defined
and are satisfied in about 50% of knees requiring
2
Low volume was defined as <10 cases/year, medium vol-
replacement. With fixed bearing devices, the indi-
ume 10–29 cases/year, and high volume ≥ 30 cases/year. cations and contraindications are not well defined.
Table 17.3 Survival and outcomes for Phase 3 and cementless Oxford UKR (10-year results)
Principal 10 year
surgeon or Number Follow-up Survival No. of Outcome Latest
Implant author Date References of knees Age (years) (%) revisions Reason for revision measure score
Phase 3 Kristensen 2012 [65] 794 64 10 95 49 Disease progression (16), loosening (11), pain
(10), infection (4), fracture (2), other (6)
Phase 3 Jones 2012 [66] 1000 67 10 94
Phase 3 Lim 2012 [67] 400 69 10 94 14 Bearing dislocation (12), disease progression (1), OKS 37.8
infection (1) KSS-O 85.1
KSS-F 86.9
Phase 3 Davidson 2012 [68] 124 90
Phase 3 Keys 2013 [69] 107 97
Phase 3 Briant-Evans 2013 [70] 827 10 91 41
Phase 3 Faour-Martin 2013 [71] 416 59 10 95 29 Infection (15), bearing dislocation (2), persistent KSS-O 90.2
pain (8), Aseptic loosening (4) KSS-F 88.6
Phase 3 Yoshida 2013 [72] 1279 77 10 96 25 Aseptic loosening (12), bearing dislocation (10), OKS 40.8
periprosthetic fracture (2), lateral progression (1)
Phase 3 Kristensen 2013 [73] 695 64 10 85 51 Progression (16), aseptic loosening (11), pain
(10), infection (4), periprosthetic fracture (2),
malposition (2), instability (4), other (2)
Phase 3 Nagy 2013 107 10 97
Phase 3 Kim 2015 [74] 166 10 91 16 Bearing dislocation (8), bearing fracture (1), KSS-O 85.4
aseptic loosening (5), periprosthetic fracture (1), KSS-F 80.5
infection (1)
17 Results and Registry Data for Unicompartmental Knee Replacements

Phase 3 Emerson 2016 [75] 213 67 10 88 20 Lateral progression (9), aseptic loosening (4), KSS-O 93
chronic haemarthrosis (3), bearing dislocation KSS-F 78
(1), bearing fracture (1), other (2)
Phase 3 Lisowski 2016 [76] 138 72 10 92 11 Lateral progression (6), PFJ progression (2), pain OKS 41.9
(2), bearing dislocation (1) KSS 81
Phase 3 Bottomley 2016 [77] 1084 67 10 93 46 Lateral progression (13), aseptic loosening (12),
bearing dislocation (7), infection (7), pain (5),
periprosthetic fracture (1), other (1)
Cementless Campi 2018 [33] 1000 66 10 97 25 Lateral progression (9), bearing dislocation (6), OKS 41.7
(9 surgeons) periprosthetic fracture (2), pain (4), aseptic
loosening (1)
Cementless Mohammad 2020 [78] 1000 66 10 98 15 Lateral progression (4), bearing dislocation (7), OKS 41.2
(2 surgeons) periprosthetic fracture (1), pain (2), aseptic KSS-O 89.1
203

loosening (1) KSS-F 80.4

 204

Table 17.4 Survival and outcomes for the Physica ZUK UKR (10-year results)
Principal surgeon Number of Follow-up Survival No. of Outcome
Implant or author Date References knees (lateral) Age (years) (%) revisions Reason for revision measure Latest score
ZUK Nicolai 2019 [27] 452 (14) 67 10 98 6 Disease progression (3), MCL Rupture KSS-­ 93.4
(1), arthrofibrosis (1), subsidence (1) Obj 91.0
KSS-Fn
ZUK Grave 2018 [28] 460 66 10 94 11 Infection (4), pain (3), disease OKS 77%
progression (2), synovitis (1) ‘excellent’
ZUK Vasso 2015 [29] 136 67 10 97.1 4 Infection (2), lateral progression (1), pain IKS 87.2
(1) (71–100)
A. Rahman et al.
17 Results and Registry Data for Unicompartmental Knee Replacements 205

References 16. Hamilton TW, et al. The interaction of caseload and

usage in determining outcomes of Unicompartmental
knee arthroplasty: a meta-analysis. J Arthroplast.
1. Kalairajah Y, et al. Health outcome measures in the
2017;32(10):3228–3237.e2.
evaluation of Total hip arthroplasties—a compari-
17. Hamilton TW, et al. Evidence-based indications for
son between the Harris hip score and the Oxford hip
Mobile-bearing Unicompartmental knee arthro-
score. J Arthroplast. 2005;20(8):1037–41.
plasty in a consecutive cohort of thousand knees. J
2. Goodfellow JW, O'Connor JJ, Murray DW. A cri-
Arthroplast. 2017;32(6):1779–85.
tique of revision rate as an outcome measure: re-­
18. Willis-Owen CA, et al. Unicondylar knee arthroplasty
interpretation of knee joint registry data. J Bone Joint
in the UK National Health Service: an analysis of can-
Surg Br. 2010;92(12):1628–31.
didacy, outcome and cost efficacy. 2009;16:473–8.
3. The New Zealand Joint Registry. Twenty Year Report:
19. Kennedy JA, et al. Most unicompartmental knee
January 1999 to December 2018. 2019.
replacement revisions could be avoided: a radio-
4. National Joint Registry. 16th Annual report 2019—
graphic evaluation of revised Oxford knees in the
National Joint Registry for England. Northern Ireland
National Joint Registry. Sports Traumatology,
and the Isle of Man: Wales; 2019.
Arthroscopy: Knee Surgery; 2020.
5. Liddle AD, et al. Adverse outcomes after total and
20. Hamilton T, et al. Radiological decision aid
unicompartmental knee replacement in 101,330
to determine suitability for medial unicom-
matched patients: a study of data from the National
partmental knee arthroplasty. Bone Joint J.
Joint Registry for England and Wales. Lancet.
2016;98-B(10_Supple_B)):3–10.
2014;384(9952):1437–45.
21. Price AJ, Svard U. 30-year survival of the Oxford
6. Liddle AD. Failure of unicompartmental knee replace-
Mobile bearing Unicompartmental knee arthroplasty.
ment [DPhil]. Oxford: University of Oxford; 2014.
Arthroscopy. 2017;33(10):e115–6.
7. Liddle A, et al. Patient-reported outcomes after
22. Price AJ, Svard U. 20-year survival and 10-year clini-
total and unicompartmental knee arthroplasty: a
cal results of the Oxford medial UKA. In 73rd Annual
study of 14,076 matched patients from the National
Meeting of the AAOS. Chicago; 2006, p. Il.
Joint Registry for England and Wales. Bone Joint J.
23. Argenson JN, et al. Modern unicompartmental knee
2015;97-B(6):793–801.
arthroplasty with cement: a concise follow-up, at a
8. Beard DJ, et al. The clinical and cost-effectiveness of
mean of twenty years, of a previous report. J Bone
total versus partial knee replacement in patients with
Joint Surg Am. 2013;95(10):905–9.
medial compartment osteoarthritis (TOPKAT): 5-year
24. Foran JR, et al. Long-term survivorship and failure
outcomes of a randomised controlled trial. Lancet.
modes of unicompartmental knee arthroplasty. Clin
2019;394(10200):746–56.
Orthop Relat Res. 2013;471(1):102–8.
9. Beard DJ, et al. Total versus partial knee replacement
25. Ansari S, Newman JH, Ackroyd CE, St. Georg
in patients with medial compartment knee osteoarthri-
sledge for medial compartment knee replacement.
tis: the TOPKAT RCT. 2020;24:20.
461 arthroplasties followed for 4 (1-17) years. Acta
10. Newman J, Pydisetty RV, Ackroyd
Orthop Scand. 1997;68(5):430–4.
C. Unicompartmental or total knee replacement: the
26. Steele RG, et al. Survivorship of the St Georg sled
15-year results of a prospective randomised con-
medial unicompartmental knee replacement beyond
trolled trial. J Bone Joint Surg Br. 2009;91(1):52–7.
ten years. J Bone Joint Surg Br. 2006;88(9):1164–8.
11. Burn E, et al. Ten-year patient-reported outcomes
27. Gill JR, Nicolai P. Clinical results and 12-year survi-
following total and minimally invasive unicom-
vorship of the Physica ZUK unicompartmental knee
partmental knee arthroplasty: a propensity score-­
replacement. Knee. 2019;26(3):750–8.
matched cohort analysis. Knee Surg Sports Traumatol
28. Winnock de Grave P, et al. Outcomes of a fixed-­
Arthrosc. 2018;26(5):1455–64.
bearing, medial, cemented Unicondylar knee arthro-
12. Blevins JL, et al. Postoperative outcomes of total
plasty design: survival analysis and functional score
knee arthroplasty compared to unicompartmental
of 460 cases. J Arthroplast. 2018;33(9):2792–9.
knee arthroplasty: a matched comparison. Knee.
29. Vasso M, et al. Unicompartmental knee arthro-
2020;27(2):565–71.
plasty is effective: ten year results. Int Orthop.
13. Wilson HA, et al. Patient relevant outcomes of uni-
2015;39(12):2341–6.
compartmental versus total knee replacement: system-
30. Mohammad HR, et al. Long-term outcomes of over
atic review and meta-analysis. BMJ. 2019;364:l352.
8,000 medial Oxford phase 3 Unicompartmental
14. Liddle AD, et al. Narrow indications predict poor
knees-a systematic review. Acta Orthop.
outcomes in unicompartmental knee replacement,
2018;89(1):101–7.
in International Society of Arthroplasty Registers
31. Kendrick BJL, et al. Cemented versus cement-
(ISAR) annual meeting. UK: Stratford-upon-Avon;
less Oxford unicompartmental knee arthroplasty
2013.
using radiostereometric analysis. Bone Joint J.
15. Stern SH, Becker MW, Insall JN. Unicondylar knee
2015;97-B(2):185–91.
arthroplasty: an evaluation of selection criteria. Clin
32. Pandit H, et al. Improved fixation in cementless uni-
Orthop Relat Res. 1993;286:143–8.
compartmental knee replacement: five-year results of
206 A. Rahman et al.

a randomized controlled trial. J Bone Joint Surg Am. 46. Dawson J, et al. Questionnaire on the perceptions of
2013;95(15):1365–72. patients about total knee replacement. J Bone Joint
33. Campi S, et al. Ten-year survival and seven-year func- Surg Br. 1998;80:63–9.
tional results of cementless Oxford unicompartmental 47. Rajasekhar C, Das S, Smith A. Unicompartmental
knee replacement: a prospective consecutive series of knee arthroplasty. 2- to 12-year results in a commu-
our first 1000 cases. Knee. 2018;25(6):1231–7. nity hospital. J Bone Joint Surg Br. 2004;86(7):983–5.
34. Mohammad HR, et al. Ten-year clinical and radio- 48. Kumar A, Fiddian NJ. Medial unicompartmental
graphic results of 1000 cementless Oxford unicom- arthroplasty of the knee. Knee. 1999;6:21–3.
partmental knee replacements. Knee Surg Sports 49. Price AJ, et al. Oxford medial unicompartmental knee
Traumatolo Arthrosc. 2020;28(5):1479–87. arthroplasty in patients younger and older than 60 years
35. Mohammad HR, et al. Comparison of the 10-year out- of age. J Bone Joint Surg Br. 2005;87(11):1488–92.
comes of cemented and cementless unicompartmen- 50. Vorlat P, Verdonk R, Schauvlieghe H. The Oxford
tal knee replacements: data from the National Joint unicompartmental knee prosthesis: a 5-year fol-
Registry for England, Wales, Northern Ireland and the low-­ up. Knee Surg Sports Traumatol Arthrosc.
Isle of Man. Acta Orthop. 2019:1–6. 2000;8(3):154–8.
36. Mohammad HR, et al. The effect of surgeon casel- 51. Koskinen E, et al. Unicondylar knee replacement for
oad on the relative revision rate of cemented and primary osteoarthritis: a prospective follow-up study
Cementless Unicompartmental knee replacements: an of 1,819 patients from the Finnish arthroplasty regis-
analysis from the National Joint Registry for England, ter. Acta Orthop. 2007;78(1):128–35.
Wales, Northern Ireland and the Isle of Man. J Bone 52. Robertsson O, et al. Knee arthroplasty in Denmark,
Joint Surg Am. 2020;102(8):644–53. Norway and Sweden. A pilot study from the Nordic
37. Rahman A, et al. Pain and function following arthroplasty register association. Acta Orthop.
Cementless and cemented Unicompartmental knee 2010;81(1):82–9.
replacement: a 5 year comparison. In: Virtual EFORT 53. Price AJ, Svard U. A second decade lifetable survival
congress. Vienna, Austria; 2020. analysis of the Oxford unicompartmental knee arthro-
38. White SH, Roberts S, Jones PW. The twin peg Oxford plasty. Clin Orthop Relat Res. 2011;469(1):174–9.
partial knee replacement: the first 100 cases. Knee. 54. Heyse TJ, et al. Survivorship of UKA in the middle-­
2012;19(1):36–40. aged. Knee. 2012;19(5):585–91.
39. Mohammad HR, et al. A matched comparison of 55. O'Rourke MR, et al. The John Insall award: unicom-
revision rates of cemented Oxford unicompartmental partmental knee replacement: a minimum twenty-­
knee replacements with single and dual peg femoral one-­year followup, end-result study. Clin Orthop
components, based on data from the National Joint Relat Res. 2005;440:27–37.
Registry for England, Wales, Northern Ireland and the 56. Squire MW, et al. Unicompartmental knee replace-
Isle of Man. Acta Orthop. 2020;91 ment. A minimum 15 year followup study. Clin
40. Berend K, et al. New instrumentation reduces opera- Orthop Relat Res. 1999;367:61–72.
tive time in medial unicompartmental knee arthro- 57. Tabor OB Jr, Tabor OB. Unicompartmental arthro-
plasty using the Oxford Mobile bearing design. plasty: a long-term follow-up study. J Arthroplast.
Reconstruct Rev. 2015;5(4) 1998;13(4):373–9.
41. Tu Y, et al. Superior femoral component alignment 58. Cartier P, Sanouiller JL, Grelsamer
can be achieved with Oxford microplasty instru- RP. Unicompartmental knee arthroplasty surgery.
mentation after minimally invasive unicompartmen- 10-year minimum follow-up period. J Arthroplast.
tal knee arthroplasty. Knee Surg Sports Traumatol 1996;11(7):782–8.
Arthrosc. 2017;25(3):729–35. 59. Heck DA, et al. Unicompartmental knee arthroplasty.
42. Gaba S, et al. Early results of Oxford Mobile bearing A multicenter investigation with long-term follow-up
medial Unicompartmental knee replacement (UKR) evaluation. Clin Orthop Relat Res. 1993;286:154–9.
with the Microplasty instrumentation: an Indian expe- 60. Marmor L. Unicompartmental knee arthroplasty. Ten-
rience. Arch Bone Joint Surg. 2018;6(4):301–11. to 13-year follow-up study. Clin Orthop Relat Res.
43. Mohammad HR, et al. New surgical instrumentation 1988;226:14–20.
reduces the revision rate of unicompartmental knee 61. Rachha R, Veravalli K, Sood M. Medium term results
replacement: a propensity score matched compari- of the miller-Galante knee arthroplasty with 10 year
son of 15,906 knees from the National Joint Registry. survivorship. Acta Orthop Belg. 2013;79(2):197–204.
Knee. 2020;27(3):993–1002. In Press 62. John J, Mauffrey C, May P. Unicompartmental knee
44. Svard UC, Price AJ. Oxford medial unicompartmental replacements with miller-Galante prosthesis: two
knee arthroplasty. A survival analysis of an indepen- to 16-year follow-up of a single surgeon series. Int
dent series. J Bone Joint Surg Br. 2001;83(2):191–4. Orthop. 2011;35(4):507–13.
45. Emerson RH Jr, Higgins LL. Unicompartmental knee 63. Naudie D, et al. Medial unicompartmental knee
arthroplasty with the oxford prosthesis in patients arthroplasty with the miller-Galante prosthesis. J
with medial compartment arthritis. J Bone Joint Surg Bone Joint Surg Am. 2004;86-A(9):1931–5.
Am. 2008;90(1):118–22.
17 Results and Registry Data for Unicompartmental Knee Replacements 207

64. Hall MJ, Connell DA, Morris HG. Medium to long-­ 72. Yoshida K, et al. Oxford phase 3 unicompartmental
term results of the UNIX uncemented unicompart- knee arthroplasty in Japan—clinical results in greater
mental knee replacement. Knee. 2013;20(5):328–31. than one thousand cases over ten years. J Arthroplast.
65. Wagner-Kristensen P. Follow up on 800 patients 2013;28(9 Suppl):168–71.
having a medial Oxford prosthesis at Vejle hos- 73. Kristensen PW, Holm HA, Varnum C. Up to 10-year
pital, Denmark. Oxford: Oxford Global Masters follow-up of the Oxford medial partial knee arthro-
Symposium; 2012. plasty—695 cases from a single institution. J
66. Jones L, et al. 10 year survivorship of the medial Arthroplast. 2013;28(9 Suppl):195–8.
Oxford unicompartmental knee arthroplasty. A 74. Kim KT, et al. The survivorship and clinical results
1000 patient non-designer series - the effect of sur- of minimally invasive unicompartmental knee
gical grade and supervision. Osteoarthr Cartil. arthroplasty at 10-year follow-up. Clin Orthop Surg.
2012;20:S90–1. 2015;7(2):199–206.
67. Lim HC, et al. Oxford phase 3 unicompartmental 75. Emerson RH, et al. The results of Oxford unicompart-
knee replacement in Korean patients. J Bone Joint mental knee arthroplasty in the United States: a mean
Surg Br. 2012;94(8):1071–6. ten-year survival analysis. Bone Joint J. 2016;98-­
68. Davidson JA, et al. A district general hospital experi- B(10 Supple B):34–40.
ence of Oxford partial knee replacement in the young 76. Lisowski LA, et al. Ten- to 15-year results of the
patient. Oxford: Oxford Global Masters Symposium; Oxford phase III mobile unicompartmental knee
2012. arthroplasty: a prospective study from a non-designer
69. Nagy M, Keys GW. Long-term outcome of unicom- group. Bone Joint J. 2016;98 B(10 Supple B):41–7.
partmental knee replacement in a district general hos- 77. Bottomley N, et al. A survival analysis of 1084 knees
pital (paper 234). Chicago: AAOS; 2013. of the Oxford unicompartmental knee arthroplasty: a
70. Briant-Evans T, et al. The Oxford phase 3 medial comparison between consultant and trainee surgeons.
unicompartmental knee replacement. 10 year results Bone Joint J. 2016;98-B(10 Supple B):22–7.
from an independent Centre: survival, function and 78. Mohammad HR, et al. Ten-year clinical and radio-
risk factors for revision. In: British Association for graphic results of 1000 cementless Oxford unicom-
Surgery of the Knee (BASK) Annual Meeting. Derby, partmental knee replacements. Knee Surg Sports
UK; 2013. Traumatol Arthrosc. 2020;28(5):1479–87.
71. Faour-Martin O, et al. Oxford phase 3 unicondylar knee
arthroplasty through a minimally invasive approach:
long-term results. Int Orthop. 2013;37(5):833–8.