SCIENTIFIC ARTICLE

Biomechanical Comparison of Parallel Versus 90-90

Plating of Bicolumn Distal Humerus Fractures With
Intra-Articular Comminution
Christopher Got, MD, John Shuck, BS, Alison Biercevicz, BS, Dave Paller, MS, Mary Mulcahey, MD,
Matthew Zimmermann, MD, Theodore Blaine, MD, Andrew Green, MD

Purpose To compare the biomechanical properties of 90-90 versus mediolateral parallel

plating of C-3 bicolumn distal humerus factures.
Methods We created intra-articular AO/Orthopaedic Trauma Association C-3 bicolumn frac-
tures in 10 fresh-frozen matched pairs of cadaveric elbows. We determined bone mineral
density of the metaphyseal region with dual-energy x-ray absorptiometry. The matched pairs
of elbows were randomly assigned to either 90-90 or parallel plate fixation. We tested
anteroposterior displacement at a rate of 0.5 mm/s to a maximum load of ⫾100 N for both
the articular and entire distal humerus segments. We tested torsional stability at a displace-
ment rate of 0.1 Hz to a maximum torque of ⫾2.5 Nm. After cyclical testing, we loaded the
specimens in torsion to failure.
Results There was no significant difference in the bone density of the paired specimens.
Compared with parallel fixation, 90-90 plate fixation had significantly greater torque to
failure load. Both plating constructs were equally sensitive to bone density. Both techniques
had the same mode of failure in torsion, a spiral fracture extending from the medial plate at
the metaphyseal-diaphyseal junction. There was no significant difference in the stiffness of
fixation of the articular fragment or the entire distal segment in anteroposterior loading.
Conclusions This study demonstrated that 90-90 and parallel plating had comparable biome-
chanical properties for fixation of comminuted intra-articular distal humerus fractures, and that
90-90 plating had greater resistance to torsional loading. (J Hand Surg 2012;37A:2512–2518.
Copyright © 2012 by the American Society for Surgery of the Hand. All rights reserved.)
Key words Comminuted intra-articular distal humerus, biomechanical testing, 90-90 versus
parallel fixation, torque to failure.

distal humerus frac- reduction and internal fixation is to achieve stable and

C
OMMINUTED AND DISPLACED
tures present a tremendous treatment challenge. accurate bony and articular reconstruction that permits
These injuries are usually the result of high- early range of motion. This is an important component
energy trauma in younger patients or low-energy falls of the initial rehabilitation process that leads to both an
in elderly patients with osteoporosis. The goal of open anatomical and functionally successful outcome.1 Ar-

From the Department of Orthopaedic Surgery, Warren Alpert Medical School, Brown University, Provi- Corresponding author: Christopher Got, MD, attn: Suzanne Swanson, Coop 1, Department of Or-
dence, RI. thopaedic Surgery, Warren Alpert Medical School Brown University, 593 Eddy Street, Providence, RI
Received for publication February 29, 2012; accepted in revised form August 30, 2012. 02903; e-mail: [email protected].

Smtih & Nephew supplied the hardware through an unrestricted research grant. T.B. has financial 0363-5023/12/37A12-0011$36.00/0
http://dx.doi.org/10.1016/j.jhsa.2012.08.042
involvement with Zimmer. A.G. has financial involvement with Tornier, Illuminos, Arthrex, Linvatec,
Smith & Nephew, and Synthes.
HardwarewasdonatedforthisprojectbySmithandNephewthroughanunrestrictededucationgrant.

2512 䉬 ©  ASSH 䉬 Published by Elsevier, Inc. All rights reserved.

 TESTING INTRA-ARTICULAR DISTAL HUMERUS 2513

beitsgemeinschaft für Osteosynthesefragen (AO)/Or-

thopaedic Trauma Association (OTA) C-3 bicolumn
intra-articular fractures present the greatest complexity
and potential for poor outcomes.
Despite years of experience and advances in surgical
technique and implants, the optimal technique and
method of internal fixation remain controversial. A
number of techniques that employ plate and screw
internal fixation have been developed. The 90-90 plat-
ing technique for distal humerus fractures was devel-
oped in response to failure of parallel posterior plating,
and was found to be superior to both Y-plate constructs
and screw fixation.2 The 90-90 fixation technique em-
ploys a plate in the sagittal plane, which provides re-
sistance to the forces generated in the flexion arc of the
elbow, and in the coronal plane. Dual plating has be-
come the method of choice for open reduction and
internal fixation of intra-articular distal humerus frac-
tures; the 90-90 system is the standard configuration
employed by most surgeons.3– 8 Recently, parallel (me-
FIGURE 1: Anteroposterior radiograph of parallel fixation.
diolateral) plating has been promoted as an alternative Arrows demonstrate osteotomy sites.
to 90-90 plating. Current techniques have evolved to
use either 90-90 or parallel plating systems that are
specifically designed with precontoured plates and
locking screws. Favorable clinical outcomes have been The specimens were maintained in a freezer at
reported with dual plating that offers adequate fixation ⫺20°C until approximately 24 hours before testing. We
to allow for early rehabilitation.9 –14 thawed the specimens to room temperature and re-
Although a number of recent studies report the re- moved all residual soft tissue with careful dissection to
isolate the distal humerus. Throughout preparation and
sults of biomechanical testing for internal fixation of
testing, we kept the specimens moist with a wrapping of
distal humerus fractures, most have failed to mimic
saline-soaked gauze and minimized freeze-thaw cycles.
current clinical standards of fracture fixation. The pur-
We assigned the matched pairs of elbows in a right–
pose of this study was to compare the fixation between
left alternating fashion into each of 2 fixation tech-
90-90 plating and parallel-mediolateral plating. It was
niques (90-90-medioposterolateral [group 1] or parallel-
our hypothesis that both forms of fixation would be
mediolateral plate systems [group 2]) (Figs. 1, 2). We
equivalent when tested in anteroposterior (AP) transla-
used a 0.8-mm sagittal handsaw to simulate a C-3
tion, torque stiffness, and torque to failure regardless of
fracture (AO/OTA classification) of the distal humerus.
bone mineral density.
The articular osteotomy was performed at the junction of
the trochlea and capitellum laterally and within the troch-
MATERIALS AND METHODS lear groove medially. We performed medial and lateral
We used 10 fresh-frozen matched cadaveric humeri supracondylar osteotomies based on precontoured plate
(n ⫽ 20; 4 males and 6 females) for this study. We locations (PERI-LOC; Smith and Nephew, Memphis,
screened the specimens radiographically for gross ana- TN). The lateral column osteotomy was located between
tomical defects and arthritis and excluded them from the third and fourth holes (counting proximal to distal) and
the study if we found any abnormalities. The mean age the medial column osteotomy within the fourth hole
of donors was 59 years (range, 45– 69 y). (counting proximal to distal) of the precontoured appro-
We determined bone mineral density (g/cm2) of the priate plate. These column osteotomies extended obliquely
specimens with dual-energy x-ray absorptiometry anal- into the coronoid and olecranon fossae, similar to com-
ysis (g/cm2) focusing on the metaphyseal region of the monly observed fracture patterns. We chose the location of
distal humerus. We performed side-to-side comparisons these osteotomies specifically to recreate the location of
and correlations between bone mineral density and me- metaphyseal fracture extension as defined by the AO/OTA
chanical testing data. classification.

JHS 䉬 Vol A, December 

2514 TESTING INTRA-ARTICULAR DISTAL HUMERUS

cm of humeral shaft exposed above the proximal ends

of the plates.
Mechanical testing
We performed mechanical testing of the specimens on
an Instron (Norwood, MA) servohydraulic test frame.
The proximal shaft of the humerus potted in a polyvi-
nylchloride pipe was rigidly attached to the load cell.
Testing occurred in 2 phases: The first phase focused
solely on the intra-articular fragment, and the second on
the entire articular segment in relation to the shaft. For
the AP testing, we applied a custom jig to the intra-
articular fracture fragment and attached it to the servo-
hydraulic actuator (Fig. 3). We placed a customized
piece of copper only around the comminuted trochlear
fragment to focus testing on the specific articular sur-
face fragment (Fig. 3A). This piece of copper allowed
for placement of screws from the jig to gain adequate
compression to stabilize the interface of the copper cuff
and the articular segment for testing. This minimized
any effect that cartilage and bone deformity might im-
FIGURE 2: Anteroposterior radiograph of 90-90 fixation.
Arrows demonstrate osteotomy sites.
part on the results. We performed cyclical loading in the
AP direction at a displacement rate of 0.5 mm/s to a
maximum load of ⫾100 N for a total of 125 cycles, in
an attempt to recreate stresses at the articular level seen
We performed internal fixation immediately after with early elbow range of motion.16
creating the osteotomies in both groups. All specimens After testing the intra-articular fragment, we potted
had each plate fixed proximally to the diaphysis with 3- the entire humerus distal to the transepicondylar level in
to 3.5-mm nonlocking screws in the 3 most proximal 2-part urethane compound (SmoothOn, Easton, PA), to
holes of the dual plates. focus testing on the entire articular segment of the
humerus in relation to the shaft. We tested the distal
Mediolateral/parallel plating humerus articular segment in the AP direction under the
Following the AO technique, with provisional Kirsch- same parameters used for the intra-articular segment
ner wire fixation in place, we affixed medial and lateral described above.
plates. Distally, both plates had all 3 locking screw (2.7 To test torsional stiffness, we rigidly attached the potted
mm) options filled. We carefully placed the screws to ends of the construct to both sides of the servohydraulic
avoid both the olecranon fossa and the articular surface, test frame. We tested torsional stiffness at a displacement
which always resulted in varying amounts of screw rate of 0.1 Hz to a maximum torque of ⫾2.5 Nm for a total
interdigitation (Fig. 1). of 20 cycles.16 After the cyclical testing protocol was
completed, we monotonically loaded the specimens in
90-90/perpendicular plating torsion to failure at a displacement rate of 1.0°/s (Fig. 4). In
As described by AO technique for perpendicular plating, each fixation group, we tested 5 specimens in internal
we achieved distal fixation of the articular components of rotation and 5 in external rotation.
the fracture with a single 3.5-mm fully threaded cortical We collected data digitally at a test frequency of 50
screw.15 This screw was placed in static mode, outside the Hz for all mechanical testing. Digital photographs of
plate, and centrally in the spool of the trochlea. We used all the test setup as well as modes of failure were obtained.
3 distal locking screw (2.7 mm) options in both the pos- We calculated stiffness using the slope of the initial
terolateral and medial plates (Fig. 2). linear region of the load or torque versus the actuator
After fixation of the distal humerus fractures, we position plot.
obtained anteroposterior and lateral radiographs to con-
firm satisfactory fracture reduction and internal fixation Statistical analysis
(Figs. 1, 2). We then potted the proximal aspect of the We used paired Student’s t-test to compare the bone
humerus potted in a polyvinylchloride pipe, leaving 1.5 density of the matched pairs of distal humeri. We also

JHS 䉬 Vol A, December 

 TESTING INTRA-ARTICULAR DISTAL HUMERUS 2515

FIGURE 3: A Testing apparatus for the intra-articular fracture segment demonstrated on a polyurethane (Sawbones Vashon, WA)
model. B Copper flange location is shown on the articular segment on a polyurethane (Sawbones, Vashon, WA) model.

To determine whether 90-90 or parallel plate fixation

was more sensitive to bone mineral density in ultimate
strength to failure, we performed a linear regression
analysis. We conducted an F-test to compare the slopes
of the 2 linear regression analyses.

RESULTS
The bone mineral densities ranged from 0.702 to 3.142
g/cm2. There were no significant differences in bone
mineral densities between groups—90-90 (0.924 ⫾
0.171) and parallel (0.928 ⫾ 0.138) (P ⫽ .893)— or
between sides within each matched pair (P ⫽ .153).
The torque to failure load of the 90-90 plate fixation
was significantly greater than the torque to failure load
of parallel plating (P ⬍ .05). In torque to failure testing,
we found no difference between 90-90 and parallel
fixation methods and their dependence on bone mineral
density (r2 ⫽ 0.482 vs r2 ⫽ 0.21 and F ⫽ 0.044; P ⫽
.838) (Fig. 5). Both fixation methods demonstrated the
same mode of failure, a spiral fracture beginning just
anterior to the medial plate in the proximity of the
second or third most proximal screw, regardless of the
direction of torque.
There were no significant differences in the AP stiff-
ness of the intra-articular fragment (P ⫽ .581), the AP
stiffness (P ⫽ .312), and torsional stiffness of the entire
distal articular segment (P ⫽ .40) (Table 1).
FIGURE 4: Testing configuration for testing torque to failure.
DISCUSSION
The 90-90 plating and medial-lateral parallel plating
used paired t-tests to evaluate the differences between techniques are 2 current, predominant methods for in-
AP plane intercondylar fragment and distal articular ternal fixation of comminuted intra-articular distal hu-
fragment stiffness, torque stiffness, and torque to fail- merus fractures. The results of this study suggest that
ure. A value of P ⬍ .05 was used a priori to determine the stiffness of these 2 fixation constructs in the modes
statistical significance. and directions tested is equivalent. This equivalence

JHS 䉬 Vol A, December 

2516 TESTING INTRA-ARTICULAR DISTAL HUMERUS

Torque to Failure vs BMD

80.00 y = 47.972x - 4.2588
R2 = 0.4821
70.00
Torque to Failure (N-m)

60.00 90-90

50.00
PARALLEL
40.00
Linear (90-90)
30.00

20.00 Linear (PARALLEL)

10.00 y = 57.054x - 20.874

R = 0.21026
0.00
0.6 0.7 0.8 0.9 1 1.1 1.2
BMD (g/cm2)

FIGURE 5: Torque to failure versus bone mineral density.

TABLE 1. Results of Testing Modalities

90/90 Parallel

Mean SD Mean SD P Value

Maximum deflection (fragment) (mm) 1.29 0.42 1.20 0.27 .581

Maximum deflection (entire segment) (mm) 0.67 0.08 0.70 0.08 .312
Delta deflection (mm) 0.62 0.39 0.50 0.24 .439
Stiffness fragment cycle 120 (N/mm) 150.61 47.69 156.74 32.58 .727
Stiffness entire segment cycle 120 (N/mm) 278.81 37.41 269.32 34.16 .400
Torque to failure (Nm) 44.07 16.87 31.92 16.23 .047
Maximum deflection at maximum torque (°) 22.14 10.58 14.92 10.84 .179

was maintained even when accounting for bone mineral might be a more important parameter than axial
density. torsion. Based on our experience and findings, we
Several other recent biomechanical studies re- believe that axial load is likely less important
ported comparisons of 90-90 and parallel plate when considering the early postoperative course.
internal fixation.16 –18 Schwartz et al18 tested ep- Early range of motion is not likely to subject the
oxy resin humeri and a fracture pattern similar to fixation construct to substantial axial load. In con-
that in our study with a comminuted intra-articular trast, the elbow is subjected to torsional loads even
segment. However, the intra-articular segment was with minimal motion.6 However, torsional testing
not loaded in isolation. They did not find a statis- focuses on the fixation strength between the shaft
tically significant difference in stiffness between and the articular segment as a whole, not individ-
the 2 constructs, whereas the parallel construct ual fracture fragments. It is nearly impossible to
demonstrated lower transverse strains in axial tor- place torsional loads on small articular fracture
sion and 90-90 experienced less strain in axial fragments. However, testing perpendicular to the
compression. They noted that given the mechanics axis of fixation (AP plane) of the intra-articular
and anatomy of the elbow joint, axial compression component accentuates any weaknesses in the ar-

JHS 䉬 Vol A, December 

 TESTING INTRA-ARTICULAR DISTAL HUMERUS 2517

ticular construct that may result in failure of fixa- a gravitational varus torque stress applied upon the
tion with early motion. forearm during daily minimal-use activities. They cited
Arnander et al17 tested a simple transverse distal the example of reaching for a glass of water and bring-
humeral osteotomy without articular involvement in ing it to the mouth.6 The possible consequence of
epoxy resin humeri. They examined sagittal plane torsional load is distraction of the lateral column from
bending forces and the parallel system had superior the posterior plate. An argument against the posterior
strength and stiffness. Stoffel et al16 compared the 2 plate and in favor of placing plating on the lateral
techniques in cadaveric osteoporotic bone. The fracture column is laid out. Our data do not support this hypoth-
pattern used included a transverse osteotomy proximal esis; rather, they suggest that 90-90 plating achieves
to the olecranon fossa with a simple sagittal fracture greater ultimate strength in torque. Also, in torque to
intra-articularly through the distal segment, equivalent failure, neither 90-90 nor parallel fixation was related to
to a C-2 fracture. They reported that parallel fixation bone mineral density. This finding is not consistent with
had superior stability in compression and external rota- those of Stoffel et al,16 who found that the 90-90 plating
tion and greater resistance to plastic deformation. Sev- system underperformed compared with the parallel con-
eral prior studies used a model with a segment of bone struct in fixation of osteoporotic bone. However, we did
loss in the supracondylar region; however, it is our not find either plating construct to be related to bone
experience that our model more closely mimics a typ- mineral density through our represented range of 0.702
ical clinical presentation. Our model represents a more to 3.142 g/cm2.
stable fracture pattern than those with segmental bone Finally, Shin et al12 recently reported a comparison
loss, but its clinical implications are still relevant. of clinical outcomes between the 2 constructs. The
Although Schwartz et al18 used a fracture model authors found no significant difference in outcomes
similar to ours, they did not specifically evaluate the between the parallel and 90-90 systems, as assessed by
fixation of the comminuted segment. Relevant limita- postoperative arc of flexion. They concluded that both
tions of previous studies include the use of resin bone constructs could provide adequate fixation and stability.
models,18 the assumption that stainless-steel plating Their study was likely underpowered and not random-
systems are equivalent to titanium plating systems,16 ized; the practical considerations of this type of study
failure to use a clinically relevant model,16,17 and vari- render a definitive answer about the best fixation diffi-
able fixation techniques and materials.16,18 cult to determine.
This study had a number of strengths in both design The limitations of this study included a relatively
and testing. The use of fresh-frozen cadaveric tissue small sample size. In addition, because there are no
most accurately reflects a clinical situation and enabled standard data for distal humerus bone densities, we
us to evaluate the effect of bone mineral density on the simply assessed for equal distribution with dual-energy
strength of the fixation constructs. Although fracture x-ray absorptiometry scanning. We scanned no other
fixation models that use composite materials can per- standardized areas such as the spine or hip to determine
form similarly to bone, they may not mimic modes of whether the World Health Organization criteria for os-
failure that occur in human bone. In addition, our use of teoporosis were met. However, we were not much
commercially available precontoured distal humerus concerned about any osteoporotic testing medium af-
plates for fixation is consistent with current clinical fecting outcome, given matching specimen assignments
practice. Prior studies have been inconsistent in both the to the 2 testing groups. Furthermore, the 90-90 system
mode of fixation and the materials. They tested stain- seemed to be equally affected by osteoporotic bone, as
less-steel versus titanium as equals, as well as precon- determined by linear regression analysis. It is therefore
toured versus reconstruction plates. a reasonable assumption that bone quality did not affect
The results of the current study offer evidence sup- the comparative results of the mechanical testing. Our
porting the clinical utility of the specific 90-90 and study, like many prior cadaveric biomechanical studies,
parallel plating constructs employed in this study for was not adequately powered (power ⫽ 0.05). When
open reduction and internal fixation of comminuted comparing the differences in mean bone mineral den-
intra-articular distal humerus fractures. The 2 constructs sity between groups, a post hoc analysis resulted in N ⫽
had equivalent stiffness when subjected to AP and 10,012 to reach a power of 0.8. To our knowledge, only
torsional loads. In the current study, 90-90 plate fixation Stoffel et al16 performed a study in a cadaveric model
required a significantly higher load for torque to failure that exceeded our number of specimens tested.
(P ⬍ .05). This finding has noteworthy clinical impli- Based on the findings of this study, we conclude that
cations. O’Driscoll et al6 highlighted the importance of in most clinical situations, both the 90-90 and parallel

JHS 䉬 Vol A, December 

2518 TESTING INTRA-ARTICULAR DISTAL HUMERUS

plating configurations are acceptable methods of fixa- 9. Celli A, Donini MT, Minervini C. The use of pre-contoured plates in
the treatment of C2-C3 fractures of the distal humerus: clinical
tion for comminuted intra-articular distal humerus frac- experience. Chir Organi Mov 2008;91:57– 64.
tures. Factors related to specific fracture patterns, sur- 10. Huang TL, Chiu FY, Chuang TY, Chen TH. The results of open
gical exposure, and implant placement can be reduction and internal fixation in elderly patients with severe frac-
tures of the distal humerus: a critical analysis of the results. J Trauma
considered on a case-by-case basis. To date, neither
2005;58:62– 69.
clinical outcome nor biomechanical testing has demon- 11. Sanchez-Sotelo J, Torchia ME, O’Driscoll SW. Complex distal
strated either system to be unequivocally superior. humeral fractures: internal fixation with a principle-based parallel-
plate technique. J Bone Joint Surg 2007;89A:961–969.
12. Shin SJ, Sohn HS, Do NH. A clinical comparison of two different
REFERENCES double plating methods for intraarticular distal humerus fractures. J
1. Green A. Open reduction and internal fixation with 90-90 plating of Shoulder Elbow Surg 2010;19:2–9.
bicolumn distal humeral fractures. Instr Course Lect 2009;58:515– 13. Soon JL, Chan BK, Low CO. Surgical fixation of intra-articular
199. fractures of the distal humerus in adults. Injury 2004;35:44 –54.
2. Helfet DL, Hotchkiss RN. Internal fixation of the distal humerus: a 14. Wildburger R, Mahring M, Hofer HP. Supraintercondylar fractures
biomechanical comparison of methods. J Orthop Trauma 1990;4: of the distal humerus: results of internal fixation. J Orthop Trauma
260 –264. 1991;5:301–307.
3. Anglen J. Distal humerus fractures. J Am Acad Orthop Surg 2005; 15. Robinson CM. Fractures of the distal humerus. In: Bucholz RW,
13:291–297. Heckman JD, Court-Brown CM, Tornetta P, eds. Rockwood and
4. Hausman M, Panozzo A. Treatment of distal humerus fractures in Green’s fractures in adults. Philadelphia: Lippincott Williams and
the elderly. Clin Orthop Relat Res 2004;425:55– 63. Wilkins, 2006:1083–1088.
5. Jacobson SR, Glisson RR, Urbaniak JR. Comparison of distal hu- 16. Stoffel K, Cunneen S, Morgan R, Nicholls R, Stachowiak G. Com-
merus fracture fixation: a biomechanical study. J South Orthop parative stability of perpendicular versus parallel double-locking
Assoc 1997;6:241–249. plating systems in osteoporotic comminuted distal humerus frac-
6. O’Driscoll SW. Optimizing stability in distal humeral fracture fixa- tures. J Orthop Res 2008;26:778 –784.
tion. J Shoulder Elbow Surg 2005;14:186S–194S. 17. Arnander MW, Reeves A, MacLeod IA, Pinto TM, Khaleel A. A
7. Ring D, Jupiter JB. Fractures of the distal humerus. Orthop Clin biomechanical comparison of plate configuration in distal humerus
North Am 2000;31:103–113. fractures. J Orthop Trauma 2008;22:332–336.
8. Self J, Viegas SF, Buford WL Jr, Patterson RM. A comparison of 18. Schwartz A, Oka R, Odell T, Mahar A. Biomechanical comparison
double-plate fixation methods for complex distal humerus fractures. of two different periarticular plating systems for stabilization of
J Shoulder Elbow Surg 1995;4:10 –16. complex distal humerus fractures. Clin Biomech 2006;21:950 –955.

JHS 䉬 Vol A, December 