Hip Replacement Imaging 

Arthroplasty of the hip may be categorized as a total hip arthroplasty or a hemiarthroplasty. In a total hip arthroplasty, the articular surfaces of both the acetabulum and femur are replaced. This involves either replacement of the femoral head and neck or replacement of the surface of the femoral head (resurfacing total hip arthroplasty) (see the images below); both procedures also replace the acetabulum.

In contrast to a total hip arthroplasty, a hemiarthroplasty involves replacement of the articular surface of the femoral head without surgical alteration to the acetabular articular surface. This may involve replacement of the femoral head and neck (unipolar hemiarthroplasty), replacement of the femoral head and neck with an additional acetabular cup that is not attached to the pelvis (bipolar hemiarthroplasty), or replacement of the surface of the femoral head (resurfacing hemiarthroplasty) (see the images below).

The hip joint may be replaced with a variety of materials, including metal, polyethylene, and ceramic. There are also various methods of arthroplasty fixation, such as polymethylmethacrylate (PMMA) cement and screw fixation, although cementless press fit and porous ingrowth arthroplasties may also be used.

Radiography is the primary imaging method for the evaluation of hip arthroplasties,^[1] and imaging of a hip arthroplasty and its complications primarily relies on the information that is obtained from routine radiography; however, there are specific roles for other imaging techniques, such as arthrography, computed tomography (CT) scanning, magnetic resonance imaging (MRI), ultrasonography, and nuclear medicine.

For patient education information, see the Arthritis Center and Foot, Ankle, Knee, and Hip Center, as well as Total Hip Replacement.

Postoperative assessment

Radiographs are essential for the evaluation of hip arthroplasties. It is important that the entire prosthesis is included on 2 orthogonal radiographs of acceptable technique. When evaluating the acetabular component on the frontal view, there is normally 30-50° of lateral inclination (see the image below).^[2]

On a cross-table (Manfredi) or true lateral radiograph, there is normally 5-25° of anteversion (see the images below).^[2]

The femoral component should be assessed for symmetry relative to the contralateral normal hip, when present. In the vertical direction, the center of the femoral head is assessed relative to the ischial tuberosities and the greater trochanter (see the first image below).^[2] In the horizontal direction, the center of the femoral head is assessed relative to the lateral margin of the acetabular tear drop (see the second image below).^[2]

Normal radiographic findings

With a cemented prosthesis, normal radiographic findings include a lucency that is less than 2 mm thick at the bone-cement interface; the lucency represents fibrous tissue and is outlined by a thin, sclerotic demarcation line (see the image below).^[3]

A lucency at the metal-cement interface of a cemented arthroplasty and one at the metal-bone interface of a cementless arthroplasty are typically related to surgical technique. These lucencies are normal if they are stable over time, but they generally should be less than 2 mm thick (see the image below). Lucencies should be followed up on radiographs because progression can indicate loosening.

Another normal finding at porous ingrowth surfaces is the presence of focal sclerosis or spot welds (see the first image below).^[3] Bone resorption may also be seen beneath the femoral flange with a cementless femoral component and with a cemented femoral component, in which a lucency may be up to 4 mm thick (see the second image below).

Focal osteopenia of the trochanteric regions due to stress shielding (diverted stress causes bone resorption) is considered a normal finding when the femoral component is secure (see the images below).

As an isolated finding, sclerosis at the tip of a cementless femoral component, or pedestal formation, is of unclear significance. A radiopaque cement restrictor or centralizer may be used with cemented femoral components, and cables or wires may be used after a trochanteric osteotomy or after a total hip arthroplasty revision. (See the images below.)

Wire fractures occur in up to 33% of hips and are usually insignificant without greater trochanteric displacement.^[4] However, fractured wires may cause an adjacent soft-tissue abnormality such as bursitis. A patient with more than 2 cm of trochanteric displacement may need a repeat operation. The normal radiographic findings described above can also be applied to other imaging modalities, such as CT scanning and MRI.

Early complications

Early complications include improper component placement, dislocation, and cement extrusion. Increased acetabular cup inclination and abnormal version—as well as a femoral component that is too long (causing muscle spasm)—may predispose the hip to a dislocation between the acetabulum and the femoral head (see the image below).

Most such dislocations occur in the immediate postoperative period. Varus angulation between the femoral stem and femoral diaphysis predisposes to femoral fracture at the femoral stem tip (see the first image below). Cement extrusion is typically asymptomatic (see the second image below).

Late complications

Late complications include dislocation, hardware failure, fracture, heterotopic ossification, prosthetic loosening, infection, and particle disease.

Hardware failure may consist of metal, ceramic, or polyethylene component fracture and displacement. Osseous fractures may involve the greater trochanter, femoral neck, acetabulum, and femoral diaphysis; these fractures may be related to trauma, stress shielding, or component loosening.^[5] (For depictions of these types of hardware failure, see the images below.)

Heterotopic ossification is usually asymptomatic, but it is seen in up to 39% of total hip arthroplasties^[4] and may begin by 2-3 weeks after surgery, with possible ankylosis by 12 weeks (see the images below).

Brooker and Bowerman classified heterotopic ossification as the following^[4] :

Class 1 – Islands of bone in soft tissues

Class 2 – >1 cm gap in heterotopic ossification between the femur and pelvis

Class 3 – < 1 cm gap

Class 4 – Bony ankylosis

Loosening

With a cemented component, loosening is suggested by the presence of component migration or tilt or a new cement fracture. (See the related images below.)

With a cementless component, a femoral component subsidence that is >10 mm or an increased number of metal beads that are displaced from the surface of a bone ingrowth prosthesis over time (bead shedding) also indicates abnormal component motion and loosening (see the image below).

A finding that is common to all types of component fixation is a >2 mm or progressive periprosthetic lucency (see the images below), although this may be due to coexisting infection or particle disease.

The location of periprosthetic lucencies can be described by their zones, which are based on anteroposterior and lateral hip radiographs. On an anteroposterior radiograph, the femoral zones are numbered 1-7, and the acetabular zones are referred to as I, II, and III (see the first image below). On a lateral hip radiograph, additional femoral zones are numbered from 8-14 (see the second images below).

Infection and particle disease

The identification of a lucency around a prosthesis (>2 mm or increasing lucencies) raises the clinical concern for an infection or particle disease (also called aggressive granulomatosis), with possible coexisting component loosening.^{[2, 4]} Infection, particle disease, and isolated mechanical loosening may appear similarly on radiographs. (See the related images below.)

However, a diffuse lucency suggests mechanical loosening or infection; multifocal lucencies can suggest particle disease or infection. With mechanical loosening, a diffuse lucency around the femoral component can be seen with a pistoning effect (see the first image below), or focal lucencies may be seen at the proximal and/or distal aspects from a toggling effect (see the second image below).

Evidence for polyethylene wear, which appears as an asymmetric position of the femoral head within the acetabular cup, is an important finding that also suggests particle disease. None of the above radiographic findings are specific for infection, and a normal-appearing radiograph does not exclude infection^[6] ; therefore, hip aspiration is indicated when excluding infection from the differential diagnosis.

Arthrography

Arthrography is primarily used to document intra-articular needle placement during fluoroscopic arthrocentesis to exclude infection. Dedicated arthrography can also be performed to evaluate prosthetic loosening.

Normally, intra-articular contrast medium extends from the rim of the acetabular cup to the intertrochanteric line (see the first image below); thus, intra-articular contrast extension at the bone-cement interface can indicate component loosening (see the second image below).^[3] However, lack of abnormal contrast extension does not exclude component loosening. Arthrography has also been shown to be unreliable in the evaluation of a noncemented hip arthroplasty.

When filling of the bursae or cavities around the hip occurs during arthrography, irregularity of the margins may indicate infection (see the images below).^[7] Although the iliopsoas bursa may communicate with the hip joint in approximately 10% of patients, in a series of painful hip arthroplasties, Berquist et al reported that contrast filling of the trochanteric bursa was the most common finding.^[7]

Ultrasonography should be considered before the performance of a fluoroscopic aspiration of a hip joint in order to exclude infection and to screen for any adjacent and overlying soft-tissue abscesses; needle placement at fluoroscopy could theoretically be passed through an unsuspected soft-tissue abscess, thus contaminating a sterile joint.

Degree of confidence

Radiography is reliable in the diagnosis of a dislocation, an osseous fracture, and hardware failure. Because abnormal lucencies around a prosthesis that are caused by an infection may appear similar to that which is seen with prosthetic loosening or particle disease, arthrocentesis is typically used to exclude a diagnosis of infection.

False positives/negatives

Both radiography and arthrography would not detect a soft-tissue infection; thus, a normal-appearing radiograph does not exclude the presence of an infection. Arthrography is not reliable in the evaluation of a noncemented hip arthroplasty. In addition, a normal arthrogram does not exclude the possibility of a loosening prosthesis.

Although the initial evaluation of a hip arthroplasty should begin with radiography, there is a definite role for CT evaluation in several situations. When there is concern about infection, CT scanning is complementary to radiography in that CT scans can show soft-tissue abscesses (see the image below).

A significant role for CT scanning is in the evaluation of osteolysis that is related to particle disease (see the image below). Although radiography is effective in identifying large, abnormal periprosthetic lucencies, CT scans better characterize and show the extent of the osteolysis.^[8] With multiplanar reformation in multiple imaging planes, CT scans also display the location of the osteolysis and assess the status of the adjacent normal bone before surgery.

Lastly, CT scans can show the location of fragmented or failed arthroplasty components and periprosthetic fractures (see the image below) as well as assess the acetabular component version.

To reduce metal artifacts when imaging a hip prosthesis with CT scanning, it is important to optimize various technical parameters (see the image below). The milliamperes (mAs) are increased (350-450 mAs in adults; up to 600 mAs if there are bilateral hip arthroplasties), but one must also take the radiation dose into consideration, especially if one is imaging children.

Additional methods to reduce artifacts include the use of lower pitch settings (to reduce cone beam artifacts with multichannel scanners), narrow detector element collimation, increased peak killivoltage (kVp) (140 kVp), and a smoother image reconstruction algorithm (eg, use of a standard soft-tissue filter vs a bone filter).^[9] The original data are reconstructed using 1.0-1.5–mm thick slices with a 50% overlap, and then multiplanar reformations are created using 1.5-2–mm thick slices in the coronal and sagittal planes.

Degree of confidence

As a complementary imaging study to radiography, CT scanning has a role in the evaluation of soft-tissue abnormalities and osteolysis. Artifacts can be reduced with optimized technical parameters.

False positives/negatives

Soft-tissue abnormalities immediately adjacent to a metal prosthesis may not be seen on CT scans due to the presence of artifacts, which can potentially cause a false-negative examination result; however, this depends on the quality of the image and the technician's success in reducing such artifacts.

MRI has some limitations due to the artifacts produced by the prosthesis that may obscure the adjacent soft tissue and any bone abnormalities. However, the role of MRI in the evaluation of hip arthroplasty has been investigated.

Modifications to standard MRI sequences can improve imaging of a hip arthroplasty, including using thin sections or reducing the voxel size, increasing the frequency-encoding gradient strength, using spin-echo or fast spin-echo (FSE) sequences rather than gradient echo, using short TI inversion recovery (STIR) rather than spectral flat saturation, using broader band width, and using lower magnetic field strength, as well as imaging of the prosthesis with its long axis longitudinal to the static magnetic field.^{[10, 11]}

In addition, because artifacts are most pronounced in the frequency-encoding direction, it is important to place the frequency-encoding gradient in a direction that is away from suspected pathology. Similar to CT scanning, MRI may show soft-tissue abnormalities, such as abscesses and bursae (see the images below), as well as osseous abnormalities, such as osteolysis from particle disease. Component loosening appears as periprosthetic low T1 and high T2 signals, whereas particle disease will show low T1 and low to intermediate T2 signals.^[10]

Degree of confidence

Similar to CT scanning, artifacts that are produced by a prosthesis on MRI may obscure any adjacent soft-tissue and bone abnormalities. However, these artifacts can be reduced by optimizing the MRI's technical parameters. Radiography remains an important imaging method to evaluate hip arthroplasty.

False positives/negatives

Artifacts that are produced by a prosthesis may obscure the immediate adjacent soft tissue and any bone abnormalities, resulting in a false-negative MRI examination.

Unlike CT scanning or MRI, with ultrasonography, the artifacts produced by metal occur deep to the prosthesis; therefore, periprosthetic fluid collections can be visualized with this modality. The plane of the ultrasound beam or the long axis of the transducer is positioned along the long axis of the femoral neck of the prosthesis. Often, a lower frequency transducer is needed to optimize the image resolution (< 10 MHz); a curvilinear transducer or a linear transducer with a trapezoidal function is helpful to increase the field of view.

The superficial contours of the arthroplasty and the adjacent acetabulum and femur allow identification, and the metal components will appear hyperechoic with posterior reverberation artifacts (see the image below). The native bone of the acetabulum and femur will also appear hyperechoic but with posterior acoustic shadowing.

A normal hip arthroplasty may show minimal hypoechoic tissue along the femoral neck component or no tissue at all.^[12] Abnormal fluid will appear anechoic or hypoechoic over the femoral neck component (see the images below).

Synovitis may also appear hypoechoic (see the image below), but this condition is more variable in echotexture, with possible flow on color or power Doppler imaging.^[12]

It is important to scan around the entire hip region to adequately assess for the possible presence of a soft-tissue fluid collection or bursae (see the images below); in addition, imaging deep to the skin incision is very important as fluid collections often occur here.

The patient may indicate a focal area of symptoms to guide scanning. Although the findings of infection are often nonspecific, extensive fluid collection with a pseudocapsule-to-bone distance of >3.2 mm, especially if it extends beyond the femoral neck area and there is hyperemia, is often a clue that infection may exist (see the image below).^[13] In a patient who has a large body habitus, the ultrasonographic resolution decreases, and anechoic fluid may appear hypoechoic.^[12]

Ultrasonography may be used in conjunction with fluoroscopy in the setting of infection to evaluate for soft-tissue abscesses and other extra-articular fluid collections.^[13] It is important to exclude an extra-articular fluid collection before fluoroscopic arthrocentesis because there is a theoretical risk of seeding a sterile joint by passing a needle through an overlying soft-tissue abscess.

Ultrasonography can also be used to guide percutaneous needle aspiration of a joint or soft-tissue abscess, as well as guide injection or aspiration of a bursa. Ultrasound-guided injection of anesthetic agents and steroids deep to the iliopsoas tendon can be completed in the case of an iliopsoas tendon impingement from an acetabular cup (see the image below).^[14]

Another advantage of ultrasonography is the ability to dynamically assess the hip joint and the adjacent structures; this modality can evaluate any snapping or symptomatic condition that requires joint movement or unusual positioning.

Degree of confidence

Significant joint effusions or extra-articular fluid collections can be identified easily with ultrasonography. In a patient with a large body habitus, it may be difficult to visualize or exclude a small joint effusion. In this setting, percutaneous joint aspiration is needed if the clinical suspicion for infection is high, preferably with the use of fluoroscopic guidance so that intra-articular needle placement can be confirmed with an iodinated contrast medium.

False positives/negatives

Infected and sterile joint effusions or soft-tissue fluid collections may appear similar on ultrasonograms. Aspiration is required in this setting. In a patient with a large body habitus, it may be difficult to visualize a small joint effusion and differentiate it from normal postoperative changes. Consider performing an arthrocentesis if the index of suspicion for infection is high.

False-negative and false-positive arthrocentesis results for infection have been described. One study found false-negative aspirations in 11 of 19 infected arthroplasties.^[3] Similarly, false-positive aspirations have been described in up to 21% of arthroplasties. Both clinical and radiographic evidence for infection can help to differentiate true-positive from false-positive cases; the type and quantity of an organism found in aspirates has not been determined to be helpful.^[3]

Nuclear medicine studies have also been used to diagnose prosthetic loosening or infection of hip arthroplasties. A cemented component may normally demonstrate radionuclide uptake on a bone scan within 1-2 years (see the images below).

Increased tracer uptake after this period can indicate infection, prosthetic loosening (see the first image below), or fracture (see the second image below); the sensitivities range from 50-100%.^[3] With a cementless component, increased radionuclide uptake on a bone scan may persist secondary to bone ingrowth.^[15] In general, a negative bone scan suggests infection or component loosening is unlikely.

Tracer uptake on a gallium scan that correlates with bone scan findings indicates infection with few false-positive results. Labeled white blood cell scans are more specific than gallium scans, but false-negative results are possible in these cases in the presence of chronic infection. To exclude cellulitis, the radionuclide uptake on labeled white blood cell scans should correspond to the bone scan findings; to exclude normal bone marrow, white blood cell scan uptake should not correspond to sulfur colloid uptake.

Fluorodeoxyglucose positron emission tomography (FDG-PET) scanning has a variable accuracy in the diagnosis of infection, ranging from 43-78%.^{[15, 16]}

Degree of confidence

Tracer uptake on a bone scan indicates loosening of a prosthesis or infection with a 50-100% sensitivity. Concordant uptake on a bone scan and gallium or labeled white blood cell scan localizes the abnormality to the bone. Correlation with radiographs is important. Percutaneous aspiration is often required to confirm the presence of an infection.

False positives/negatives

The combination of gallium with a bone scan reduces false-positive results. Chronic infection, however, may produce false-negative results with labeled white blood cells.