INTRODUCTION

Interpretation of postoperative orthopedic radiographs comprises a significant portion of the practice

of not only subspecialized musculoskeletal radiologists but also general radiologists. A good
foundation and understanding of the most common performed orthopedic procedures is essential for
accurate interpretation of postoperative radiographs. This chapter reviews the basic concepts of joint
replacement, spinal fusion, and fracture fixation, which are some of the most common procedures
performed by orthopedic surgeons. In addition, the postoperative evaluation of various orthopedic
hardware including the imaging findings for common complications will be discussed.
JOINT REPLACEMENT
Joint replacement is one of the most common orthopedic procedures performed. Generalized
indications for joint replacement include severe osteoarthritis, avascular necrosis, trauma, and
inflammatory arthropathies such as rheumatoid arthritis. Absolute contraindications for joint
replacement include active local or systemic infection. Relative contraindications include obesity,
remote infection, unrepaired ligamentous injuries, and neurologic impairment. Prior to the advent of
joint replacement, surgical management of a painful or nonfunctional joint included joint arthrodesis
(e.g., joint fusion), osteotomy, nerve division, and joint debridement. Patients were afforded significant
improvement in quality of life with the development of joint replacement techniques; however, older
joint replacement components often suffered from premature wear. Recent advances in bio-materials
and joint replacement technology have led to marked improvements in the longevity of joint
prostheses. Orthopedic surgeons can now choose between a vast array of prosthetic devices, many
based on preference and familiarity. Though it is impossible for the radiologist to become familiar with
all the different devices in the market, the structural material and complications are shared among the
variety of different prostheses.
The main components of any modern joint arthroplasty include a metal alloy and a plastic
polyethylene liner. The low coefficient of friction between the metal alloy component and the
polyethylene component simulates movements of normal joints. Alloys represent the metallic
component of the prostheses. They are combinations of different metals such as chromium–cobalt,
chromium–cobalt–titanium, or chromium–cobalt–molybdenum. 1 These different alloys have individual
biomechanical properties based on their metal composition, and differ in terms of their resistance to
stress, strain, and tension. Polyethylene is the radiolucent liner of the prostheses. In other words, it is
not seen on the radiograph. In order to secure the prosthesis, the prosthesis may either be press fit
into the bone or cemented to the bone. Polymethylmethacrylate is the most commonly used cement to
secure the prosthesis into the medullary cavity of the bone. Cement is seen as a radiopaque lining
surrounding the prosthesis. Alternatively, porous-coated press-fit cementless prostheses demonstrate
an irregular surface coated with lucent bone growth-stimulating material to ensure adherence to the
surface.2 Another concept to be familiar with is the resistance of a prosthetic implant to motion,
whether in the anteroposterior (AP) direction or the axial direction. A constrained prosthesis has two
components that are directly linked together. As a result, there is full range of motion in only one
direction. In a nonconstrained prosthesis, it is the muscles, ligaments, and tendons that provide
stability with no connection between the two prosthetic parts. This not only provides the greatest
possible range of motion but is also most prone to joint subluxation or dislocation. Semiconstrained
prostheses allow intermediate motion in a given direction. 3
The postoperative radiograph evaluation includes at least two radiographic views of the prosthesis
at right angles to one another (e.g., orthogonal views) in addition to any specific views particular for
the joint imaged. For example, a complete examination of a total knee arthroplasty would include an
AP view, a lateral view, and possibly a sunrise view for adequate visualization of the patella. The
entire prosthesis and surrounding bone need to be imaged on the examination. The technical factors
of the radiograph must allow the examiner to distinguish between metal–bone, metal–cement, and
cement–bone interfaces. Other radiologic examinations such as arthrography, ultrasonography,
computed tomography (CT), magnetic resonance imaging, and nuclear scintigraphy also have specific
roles in evaluating joint replacement.
HIP PROSTHESES
Three types of hip prostheses exist: unipolar hemiarthroplasty, bipolar hemiarthroplasty, and total hip
arthroplasty (Figures 10-1 and 10-2). A unipolar hemiarthroplasty involves replacement of the femoral
head and neck without alteration to the native acetabulum. The femoral component includes either a
noncemented or cemented femoral stem with a femoral head that articulates directly with the native
acetabulum. This is the least common hip prosthesis and is typically performed in patients with
femoral head or femoral neck fractures and decreased life expectancy. A bipolar hip arthroplasty
includes a femoral stem with a small diameter femoral head and a separate acetabular cup (Figure
10-1). The outer portion of the acetabular cap articulates with the native acetabulum, while the inner
portion articulates with the femoral head as one unit. Again, the native acetabular surface is unaltered.
This design is most prone to dislocation as motion can occur between both the femoral head and
acetabular component and external surface of the acetabular component and the native acetabulum.
The total hip arthroplasty is the most common type of performed hip arthroplasty (Figure 10-2). In a
total hip arthroplasty, the articular surface of both the femur and the acetabulum is replaced. These
components may either be cemented or noncemented. The metallic acetabular cup includes a
radiolucent polyethylene liner that articulates directly with the metallic femoral head.

Figure 10-1. Bipolar prosthesis. AP view of the bipolar right hemiarthroplasty, with separate
acetabular cup. Note the radiolucent native articular cartilage surface.
Figure 10-2. Total hip replacement. AP view of the total hip arthroplasty, consisting of both the
femoral and acetabular components. The polyethylene liner separating the acetabular cup from the
femoral head is radiolucent. The AP view best illustrates the angle of inclination (normal between 30
and 55°).
The postoperative radiograph of the hip includes AP and lateral view including the entire femoral
stem and acetabular component. The AP film is used to measure the angle of inclination that is
optimal at 30–55° (Figure 10-2), and the lateral film is used to measure the angle of anteversion that
is optimal around 15°.1,2 The femoral component should be either parallel to the femoral shaft or in
slight valgus. Varus alignment increases the risk of stem migration, which can result in periprosthetic
fractures (Figure 10-3). With varus alignment, the lateral femoral cortex is most often injured.
Regardless of the location of the periprosthetic fracture, revision to a longer-stemmed revision
prosthesis is often needed. The femoral component should also be symmetric in the center of the
acetabular component. Smooth 2 mm or less radiolucent lines at the bone–cement interface can be
normal if not progressive. Subsidence (sinking in of the prosthesis) of less than 2 mm is also within
normal limits.4
Figure 10-3. Loose stem total hip prosthesis. (A) AP radiograph shows significant femoral stem
loosening (arrows) and varus alignment of the femoral stem tip. (B) Arthrogram of the hip reveals
contrast accumulation in between the bone and cement interface indicating loosening of the
prosthesis stem (arrows).
KNEE PROSTHESES
Most knee replacements are total knee replacements involving resurfacing of the femoral condyle and
the tibial plateau (Figure 10-4). The patella may either be simply resurfaced, or a patellar prosthesis
(e.g., a button) may be attached (Figure 10-4C). The patellar and tibial components may be cemented
or cementless. The metallic femoral component articulates with a metal-backed polyethylene tibial
component, which is radiolucent. Tricompartmental knee prostheses can further be subdivided into
posterior cruciate ligament (PCL) sparing or sacrificing prostheses. PCL sparing prostheses are most
commonly performed and have slightly improved gait. To differentiate between the two types of
prostheses, a large box is seen in the femoral component on the lateral film that articulates with the
polyethylene in the tibial tray that provides posterior stability. 1,2
Figure 10-4. Total knee arthroplasty (TKA). AP (A), lateral (B), and sunrise (C) views of TKAs show
cemented tibial and uncemented femoral components. In all images, the polyethylene component is
radiolucent and cannot be seen readily on radiograph. (A) The femorotibial component should be
aligned in 4–7)° of valgus, and the articular surface of the tibial component should be aligned parallel
to the ground. TKAs may involve simple patellar resurfacing (B) or placement of a patellar button (C).
Note in (C), patellar resurfacing and fracture of the patella are seen.
Standing AP (Figure 10-4A), lateral (Figure 10-4B), and patellar views (Figure 10-4C) are obtained
when evaluating the postoperative knee. The optimal alignment for the femorotibial component is 4–
7° valgus in the AP projection (Figure 10-4A) and neutral to minimal flexion on the lateral radiograph
(Figure 10-4B).2 In addition, the articular surface of the tibial component of the prosthesis should be
parallel to the ground on the standing views (Figure 10-4A). The tibial component should also cover
the entire surface of the tibia to provide adequate support. The femoral component should be 90° to
the long axis of the femoral shaft on the lateral view. 1
Unicompartmental knee prostheses have been used in younger patients with isolated medial or
lateral compartment arthritis (Figure 10-5). In these cases, a single femoral condyle and its tibial
articulating surfaces are resurfaced. Unicompartmental patellar prostheses have been shown to result
in suboptimal outcomes and are not routinely used.
Figure 10-5. Unicompartmental knee arthroplasty (UKA). AP (A) and lateral (B) radiographs of
UKA. The radiopaque line between the femoral and tibial components seen on the AP view (A)
corresponds to a metallic marker within the polyethylene component. Additionally, on the AP view (A),
note the periprosthetic lucency (arrows) that represents hardware loosening. There is linear soft tissue
calcification incidentally noted near the medial tibial condyle.
ANKLE PROSTHESES
The ankle is a complex joint, and success rate for joint replacement has been suboptimal. The lack of
success is likely due to inability to duplicate the normal mechanics of the ankle joint and inability to
restore the stabilizing effect of the ligaments. Although second-generation ankle prostheses have had
better outcomes than first-generation prostheses, ankle arthrodesis remains the treatment of choice in
managing the painful ankle joint.
SHOULDER PROSTHESES
Three types of surgeries exist for shoulder replacement: hemiarthroplasty (Figure 10-6), total shoulder
arthroplasty (Figure 10-7), and reverse shoulder arthroplasty (Figure 10-8). A shoulder
hemiarthroplasty is used in cases such as severe proximal humeral fractures and severe rotator cuff
tear where the patient still possesses a normal glenoid. The humeral component may be cemented or
noncemented, and articulate with the native glenoid. A total shoulder arthroplasty, usually performed
in severe glenohumeral osteoarthritis, has a metal or polyethylene-backed glenoid component (Figure
10-7).3 The reverse shoulder arthroplasty is performed in patients with a nonfunctioning rotator cuff
due to massive rotator cuff tear (Figure 10-8). In this case, the ball-shaped glenoid component aligns
with the cup of the humeral component. The cup of the humeral component is connected to the stem
portion of the prosthesis. Because these designs are held in place by the surrounding rotator cuff,
they are either semiconstrained or unconstrained and are more prone to dislocation. 1
Figure 10-6. Shoulder hemiarthroplasty. AP radiograph shows shoulder hemiarthroplasty. Note the
absence of any glenoid components. Also note that the superior aspect of the prosthetic head lies
above the greater tuberosity; this positioning helps prevent subacromial impingement.

Figure 10-7. Total shoulder arthroplasty. AP radiograph shows total shoulder arthroplasty. Note the
glenoid component contains radiopaque and radiolucent parts. Also note minimal lucency surrounding
the radiopaque glenoid component suggestive of loosening.
Figure 10-8. Reverse total shoulder arthroplasty. AP view shows reverse total shoulder
arthroplasty. Note the medialized center of rotation, which allows the deltoid muscle to substitute for
the deficient rotator cuff musculature to facilitate shoulder abduction.
Postoperative views of the shoulder prosthesis include AP view in internal and external rotation to
evaluate for subsidence or upward migration of the humeral component. In a patient with an intact
rotator cuff, impingement occurs if the most superior aspect of the prosthesis lies below the level of
the superior tip of the greater tuberosity. Trans-scapular Y or axillary views are also obtained to
assess for dislocation.
ELBOW PROSTHESES
Total elbow prostheses consist of both the humeral and ulnar components. Elbow prostheses can be
categorized by design, either as linked or nonlinked. Linked elbow prostheses can be likened to
constrained prostheses, whereas nonlinked elbow prostheses can be likened to nonconstrained
prostheses. The linked portions have a rigid hinge that connects the humeral component to the ulnar
component (Figure 10-9). Loosening, especially at the humeral component, is a major problem. The
unlinked prostheses have stemmed ulnar and humeral components that articulate via an interposed
polyethylene liner. In this case, stability is provided by the adjacent muscles, and intact tendons and
ligaments. Finally, radial head prostheses may be performed in cases of comminuted radial head
fractures (Figure 10-10).
Figure 10-9. Constrained left total elbow prosthesis. AP (A) and lateral (B) radiographs of
constrained left total elbow prosthesis. Hinged prostheses often suffer from loosening, as exhibited by
the periprosthetic lucency surrounding the humeral component that has led to periprosthetic fracture
of the distal humeral shaft (arrow).
Figure 10-10. Elbow radial head prosthesis. AP view of the right elbow illustrating radial head
prosthesis.
WRIST AND HAND PROSTHESES
Wrist arthroplasty is usually performed in patients with rheumatoid arthritis or severe osteoarthritis.
For replacement of individual carpal bones due to avascular necrosis or trauma, Silastic prostheses
have been used. The metacarpophalangeal and interphalangeal joints are commonly performed
arthroplasties in patients with severe rheumatoid arthritis. There are no clear indications in
management, and in most cases management often trends toward partial or total arthrodesis of the
wrist and the hand.
JOINT PROSTHETIC COMPLICATIONS
Bone fractures typically occur within the early postoperative period in patients with poor bone stock
such as osteoporotic patients. In the hip, excessive varus alignment of the femoral stem will
eventually predispose to early periprosthetic fracture (Figure 10-3A), requiring a long-stem revision
procedure. Fractures of the prosthesis or cement are usually delayed complications secondary to
long-term repetitive stress. Loosening is a common delayed complication shared by all prostheses
(Figures 10-3A,B, 10-5A, and 10-11). Repetitive mechanical stresses can cause loosening at the
cement–bone, prosthesis–bone and cement–prosthesis interfaces. Lucency that is less than 2 mm in
width and nonprogressive on follow-up radiographs is considered normal. Progression of lucency
greater than 2 mm or development of new, irregular areas of lucency is likely secondary to loosening
(Figures 10-3A,B, 10-5A, and 10-11).4 It is always important to have prior films available in addition to
short-term follow-up films to assess progression of loosening. In the hip, subsidence of the femoral
portion of the prosthesis that is greater than 5 mm is also indicative of loosening. Subsidence of the
acetabular component will also result in protrusio acetabuli, or migration of the prosthesis into the
pelvic cavity. Other signs of loosening in the hip prosthesis include cement fracture and sclerosis
(pedestal formation) at the tip of the prosthesis. 4

Figure 10-11. Loose femoral component of total knee arthroplasty (TKA). Lateral view of the TKA
illustrating loosening of the anterior aspect of femoral component at the site of the bone–metal
interface (arrow), as evidenced by the lucency between the femoral cortex and prosthesis.
Infection is a serious delayed complication of any joint replacement. There is considerable overlap
in differentiating infection from loosening. Additional clinical information, including laboratory analysis,
is needed to assess the likelihood of infection. Radiographically, the presence of irregular
periprosthetic lucency, periosteal reaction, and bone destruction is suggestive of infection rather than
loosening (Figure 10-12A,B). Focal areas of lucency are more suggestive of loosening than the
generalized lucency seen in infection. Additional signs of infection include soft tissue swelling, large
joint effusion (Figure 10-12B), and abscess formation. Joint aspiration is the most definitive technique
to diagnose septic arthritis. Arthrography can also be used to diagnose both loosening and infection.
Initially, the joint is aspirated for laboratory analysis. Next, iodinated contrast is injected into the joint.
Contrast accumulation around in the region of periprosthetic lucency is suggestive of loosening
(Figure 10-3B) or infection. Antibiotic-laced cement may be used after removal of infected prosthesis
(Figure 10-13). Other methods to diagnose prosthetic infection include ultrasound-guided joint fluid
aspiration and nuclear scintigraphy.5

Figure 10-12. Infected total knee arthroplasty (TKA). AP (A) and lateral (B) views of the TKAs. (A)
Both the femoral and tibial components of the TKA exhibit irregular periprosthetic lucency (arrows),
suggestive of infection. (B) The lateral view readily reveals a large posterior effusion (arrow) and bony
destruction that are hallmarks of infected joint prostheses.
Figure 10-13. Infected total knee arthroplasty (TKA) with antibiotic cement spacer. AP view of
the infected TKA with antibiotic cement spacer. Infected TKAs are often revised in a staged fashion:
first, the infected TKA is removed and an antibiotic spacer is placed as illustrated, and subsequently
once the infection has been eradicated with irrigation, debridement and antibiotics, the revision
surgery takes place.
Another relatively common complication of joint replacement is particle disease that is a host
inflammatory osteolytic response, which occurs after shedding of portions of the prosthesis (Figure
10-14). It is usually a response to the radiolucent polyethylene liner or methylmethacrylate. Although
they occur more commonly in hip prostheses, particle disease can also occur in any other prostheses.
Particle disease usually manifests as multiple well-defined lucencies that do not conform to the shape
of the prosthesis (Figure 10-14). Additional foci of endosteal scalloping may also be seen. Unlike
infection, a periosteal reaction is not seen in cases of particle disease. Along the same lines,
polyethylene wear is a common entity seen in both the hip and knee prosthesis (Figure 10-15).
Figure 10-14. Aggressive granulomatosis (particle disease) in total hip arthroplasty (THA). AP
view of the left hip arthroplasty with particle disease, as evidenced by lucencies around the prosthesis
components and multiple metallic particles in the joint space.
Figure 10-15. Polyethylene liner wear and displacement. AP view of the right hip. The femoral
head of the prosthesis is not centered in the acetabular cup due to wear and displacement of the
polyethylene liner.
Dislocation or subluxation may occur in either the early or late postoperative period. This is a
greater problem in semiconstrained or nonconstrained arthroplasties such as the shoulder or the
elbow if the surrounding muscles, tendons, and ligaments do not have the adequate strength to
prevent subluxation and dislocation. Another complication seen in various joint replacements is
heterotopic ossification seen around the periprosthetic region. Heterotopic ossification can also be
seen with other types of hardware as well (Figure 10-16). Patients at higher risk of heterotopic
ossification include patients with a history of ankylosing spondylitis, diffuse idiopathic skeletal
hyperostosis (DISH), and hypertrophic osteoarthritis.1 In advanced cases, heterotopic ossification can
limit mobility of the joint and may eventually cause joint fusion.

Figure 10-16. Heterotopic ossification around the hip joint. AP view of the right hip demonstrating
cannulated screw fixation of femoral neck fracture, with extensive heterotopic ossification (arrows)
about the greater trochanter. Heterotopic ossification in this region predictably results in significant
deficits of hip flexion and abduction.
SPINAL FUSION
Spinal fixation procedures are commonly encountered in today’s radiologic practice. The most
common indication for spinal surgery today is degenerative disk disease. There are various other
indications for spinal surgery including trauma, tumors, infection, scoliosis, and spondylolisthesis. The
goal of spinal fixation devices is to restore anatomic alignment; stabilize the bone during fusion; and
replace bone defects in cases of trauma, tumor, or infection. The same principles that apply to other
joints also apply to the spine. Fusion of a diseased joint will eliminate pain by eliminating the motion
between the painful joint, such as severely diseased disks within the lumbar spine. 6 It usually takes
6–9 months for solid fusion to be seen radiographically. The other important concept to realize is that
the spinal hardware is used to provide temporary fixation and stability by immobilizing the bone. The
function of the hardware is complete when osseous fusion occurs. Most intact implants are generally
left in place after bony fusion due to the morbidity involved in recurrent spinal surgery. This section will
discuss the procedures and range of hardware devices used in spinal fixation. The postoperative
complications will then be discussed.
SPINAL INSTRUMENTATION
Although many spinal fusion instrumentation systems exist, the basic components of each system can
be classified into a few general categories. Interpedicular screws are connected either by rods or
plates that span single or multiple vertebral body segments (Figures 10-17 to 10-22). Plates are also
commonly used in conjunction with cortical screws in anterior fusion of the cervical spine (Figure 10-
17A,B). There are various sizes of plates that can be used for both the anterior and posterior fusion
procedures. Rods are used to provide stability over short or long segments (Figures 10-17 to 10-22).
A common example is the Harrington rod used for scoliosis of the spine. Harrington rods help provide
distraction along the concavity and compression forces across the convexity in the treatment of
scoliosis. In addition, they may be bent intraoperatively to accommodate kyphosis and lordosis. Rods
can be attached to the spine by pedicle screws, wires, or cables. Disk spacers are inserted into the
intervertebral disk space after the diseased disk is removed (Figures 10-17A,B and 10-19A,B). They
are made of titanium or radiolucent material such as polyether ether ketone (PEEK). Surrounding
bone graft material is also used surrounding the disk spacer to provide additional stability. Bone graft
material is also used within the posterior elements in posterior spinal fusion to provide additional
stability. Finally, corpectomy (vertebral body replacement) may be necessary after major trauma or
destruction of the vertebral body by tumor or infection. Usually, the vertebral body is replaced by an
expandable hollow cylinder packed with bone graft material or cement. 6,7