ACR Appropriateness Criteria®
Clinical Condition: Imaging After Total Hip Arthroplasty (THA)
Variant 1: Follow-up of the asymptomatic patient with a THA.
| Radiologic Exam Procedure |
Appropriateness Rating |
Comments |
| X-ray, hip |
9 |
|
| Radiostereogrammetry, hip |
1 |
|
| NUC, bone scan |
1 |
|
| NUC, bone/gallium scan |
1 |
|
| DXA, hip |
1 |
|
| CT, hip |
1 |
|
| QCT, hip |
1 |
|
| MRI, hip |
1 |
|
Appropriateness Criteria Scale
1 2 3 4 5 6 7 8 9
1 = Least appropriate 9 = Most appropriate
|
Note: Abbreviations used in the tables are listed at the end of the "Major Recommendations" field.
Variant 2: Evaluating patients with a painful primary THA -- suspected loosening and/or wear.
| Radiologic Exam Procedure |
Appropriateness Rating |
Comments |
| X-ray, hip |
9 |
|
| CT, hip |
5 |
For planning, if osteolysis is suspected or assessment of bone stock is needed. May be useful to see subtle wear. With multidetector, thin section CT, this study is providing more information. |
| Aspiration, hips, image-guided |
5 |
Arthrography may be done to evaluate loosening. Injection of anesthetic may be helpful to localize the source of pain. |
| NUC, bone/gallium scan |
2 |
|
| DXA, hip |
1 |
|
| Radiostereogrammetry, hip |
1 |
|
| QCT, hip |
1 |
|
| MRI, hip |
1 |
Recent data suggests it may be helpful in certain cases. |
| NUC, bone scan |
No consensus |
May help determine if one or both components are loose or if other abnormalities are present. Varies by institution. |
Appropriateness Criteria Scale
1 2 3 4 5 6 7 8 9
1 = Least appropriate 9 = Most appropriate
|
Note: Abbreviations used in the tables are listed at the end of the "Major Recommendations" field.
Variant 3: Evaluating patients with a painful primary THA -- suspected infection.
| Radiologic Exam Procedure |
Appropriateness Rating |
Comments |
| X-ray, hips |
9 |
|
| Aspiration/arthrography, hips |
9 |
Should be first exam following radiographs. Other studies may be helpful but not indicated routinely. Arthrography may document collections. |
| NUC, bone scan |
3 |
Sensitive but not specific enough to be used as the sole procedure. |
| NUC, bone, Gallium scan |
3 |
If bone scan is abnormal, gallium can follow to suggest infection. |
| NUC, WBC scan, Indium-111 |
3 |
WBC scan can be done without bone scan to diagnose infection, or can follow bone scan. |
| Ciprofloxacin (infection) imaging |
3 |
|
| CT, hip |
1 |
Usually not indicated. Only if soft tissue abscesses suspected. |
| MRI, hip |
1 |
Currently being investigated. |
Appropriateness Criteria Scale
1 2 3 4 5 6 7 8 9
1 = Least appropriate 9 = Most appropriate
|
Note: Abbreviations used in the tables are listed at the end of the "Major Recommendations" field.
Symptoms
Ninety percent of patients have good or excellent results after THA. Those with persistent postoperative pain after cemented prostheses may have had the incorrect diagnosis made originally, have infection, or, less commonly, have heterotopic bone formation. Studies of patients with uncemented prostheses indicate that pain occurring after an asymptomatic period suggests possible loosening. While patients with loosening or infection usually (but not always) have pain, those with particle disease and resulting osteolysis may be asymptomatic.
Imaging Techniques
Radiographs shortly after surgery are recommended as a baseline for future evaluation and are particularly important after revision surgery. The schedule for obtaining radiographs is uncertain and depends on the presence or absence of symptoms. The NIH consensus panel indicates that "appropriate x-ray examination" is required throughout life with periodic examination, perhaps at 5-year intervals, after the first 5 years. A review of 18,486 primary total hip arthroplasties in patients with osteoarthritis found radiological follow-up in order to monitor component loosening to be unnecessary in asymptomatic patients in the first five postoperative years. In asymptomatic patients with bone ingrowth prostheses, radiography may also be useful to gauge fixation early in the postoperative course. They are also helpful in identifying bone loss from aggressive granulomatous disease, although the sensitivity of radiographs in detecting osteolysis is limited. Since this complication occurs most often several years postoperatively, one study recommends radiographs in asymptomatic patients beginning 3 years after surgery.
All symptomatic patients should undergo radiography. Availability of old radiographs to compare to new ones facilitates the diagnosis of subtle changes such as may occur in loosening or in particle disease, or of rapid changes such as may occur with infection.
Roentgen Stereophotogrammetric Analysis
Roentgen stereophotogrammetric analysis (radiostereometry) is a technique that utilizes three components to achieve highly accurate measurements of component position. First, opaque reference markers are placed in the bone near the prosthesis (usually at the time of surgery). Second, two radiographs are obtained simultaneously using two radiographic tubes that are angled to each other. Third, a calibration cage is placed beneath the table. Sophisticated mathematical analysis is then performed. One study investigated the accuracy of this technique by performing duplicate measurements. The 99% confidence limits were 0.16 mm for proximal to distal translation, 0.3 degree for anterior posterior rotation, 0.8 degree or rotation about the longitudinal axis (anteversion-retroversion), and 0.34 degree for varus/valgus tilt. Radiostereometry has been used for the early detection of component migration and prosthetic wear. Because of its complexity, however, this test remains largely a research method.
Dynamic Computed Tomography
Dynamic computed tomography is performed by obtaining CT slices through the collar of the prosthesis and the femoral condyles with the leg in maximum internal and external rotation. Change in relative position has been shown in one series to correlate with loosening at surgery.
Bone Scintigraphy
Bone scintigraphic appearances after total hip arthroplasty are varied, reflecting the stress on the adjacent bone as well as any complications that occur. One study of asymptomatic cemented total hip prostheses indicated that persistent increased uptake could be seen at the tip of the femoral stem in about 10% of patients at 1 to 3 years after surgery; at the greater trochanter in 20%; and at the acetabulum in 12% at 2 years. The femoral shaft uptake decreased by 9 months after surgery. These authors recommend that a baseline bone scan be obtained between 9 and 12 months after surgery.
Normal bone scan appearances after uncemented total hip replacements depend on the type of prosthetic components used. One group of investigators examined 25 asymptomatic prostheses serially for two years after THA and found that bone phase images showed uptake at the tip of each prosthesis at some time during the study. Increased blood flow or focal blood pool activity should suggest a complication. Uptake at the area distal to the tip or along the lateral tip increased in some cases in the 12 to 24 month follow-up. Indium-111 WBC scanning also showed increased activity at the tip of the prosthesis in 80% of uncomplicated uncemented prostheses. These authors noted that baseline three-phase bone scans and In-111 WBC scans are of value.
Evaluation of Patient with Suspected Infection
Infection occurs in 1-2% of primary total hip arthroplasties and is even more frequent after revision procedures. Approximately one third of infections develop in the first 3 months, another one third within the first year, and another one third more than 1 year after surgery. Loosening occurring within the first 2 years after surgery suggests infection. Infection of failed hip prostheses may be underestimated since organisms may reside in a biofilm. As summarized in one report, no test is perfectly sensitive and specific for the diagnosis. This includes radiography. One study evaluated radiographs of 20 infected THAs and found half to be normal.
Computed Tomography
Prospective evaluation of patients with painful total hip arthroplasties using helical CT has shown that periosteal new bone formation was always associated with infection (100% specificity) but had only 16% sensitivity. Soft tissue findings were more accurate. Fluid collections in muscles and perimuscular fat had a 100% positive predictive value and the absence of joint distension a 96% negative predictive value.
Ultrasonography
Ultrasonographic findings may suggest infection after THA. One study found that a thick capsule (3.2 mm bone capsule distance) was 100% sensitive for the diagnosis but not entirely specific (74%). The combination of intraarticular effusion with extraarticular extension was indicative of infection.
Joint Aspiration
Joint aspiration, although not perfect, is "perhaps the most useful investigative tool for definitive confirmation of the presence or absence of infection." The sensitivity of preoperative aspiration ranges from 50-93% and the specificity from 82-97%. Thus, both false positive and false negative studies occur. Debate remains regarding the indications for arthrography. Some investigators indicate that aspiration may be particularly useful when there is no clinical suspicion of infection, while others suggest aspiration when there is an elevated erythrocyte sedimentation rate or C-reactive protein or clinical suspicion remains.
Magnetic Resonance Imaging
Magnetic resonance imaging (MRI) has demonstrated a peripherally enhancing juxta-articular fluid collection connecting to the implant in a patient with periprosthetic infection.
Technetium-99m Bone Scan
Technetium-99m bone scans are thought to be sensitive but not specific for postoperative infection. One study noted that a positive bone scan strongly suggests loosening or infection, while a negative scan does not exclude these abnormalities. Another study found that a negative bone scan makes infection very unlikely. The pattern of bone scan uptake is probably unreliable in differentiating infection from loosening.
Gallium-67 Citrate
Gallium-67 citrate accumulates in areas of infection but also in areas of new bone formation. Mechanisms of gallium uptake include granulocyte uptake, direct bacterial uptake, and lactoferrin binding and binding to bacterial siderophores. False negative scans may occur in patients treated with antibiotics. A positive gallium scan is very likely to indicate infection, but a normal scan does not exclude infection (sensitivity 37%, specificity 100%).
Indium-111-Labeled White Blood Cells
Indium-111-labeled WBCs may accumulate at the site of a number of inflammatory processes, including acute osteomyelitis, acute exacerbations of chronic osteomyelitis, septic arthritis, abscesses, and rheumatoid arthritis, but also in heterotopic bone, Paget disease, and other disorders. The sensitivity of the indium WBC scans for infection decreases as the chronicity of the infection increases. Sensitivities for the diagnosis of infection after total joint replacement range from 50-100% and specificities from 45-100%.
In addition to Indium-111, leukocytes may be labeled with Tc-99m stannous colloid. Tc-99m-labeled leukocyte scanning is less expensive than Indium-111-labeled leukocyte scintigraphy and has better imaging characteristics and greater availability. Tc99-m stannous colloid labels both granulocytes and monocytes, which should offer improved detection of chronic infections such as those associated with total joint replacement. The addition of marrow imaging to leukocyte scanning has improved accuracy to 90%. Delayed Tc99-m leukocyte imaging may be helpful.
One study noted that until additional studies are done, the routine use of Indium-111-labeled immunoglobulin-G cannot be recommended.
Tc 99-m-labeled antibiotic (ciprofloxacin) has been used to evaluate orthopedic infections. This agent should be specific for infection as opposed to other inflammatory processes, although mild uptake in noninfective inflammation and foreign body reaction does occur. Antibiotic treatment may interfere with uptake of this agent. One study found a sensitivity of 85%, specificity of 92%, and accuracy of 88% in orthopedic infections (including 18 total hip replacements). This was compared to 78%, 100%, and 90%, respectively, for the bone gallium combination. The ciprofloxacin scanning procedure enables studies to be performed at 1 and 4 hours and with improved resolution in contrast to the delay for gallium and Indium-111 labeled WBC scanning. Ciprofloxacin is not taken up in normal bone marrow, and its uptake is independent of the white blood cell count.
Evaluation of Patient with Suspected Loosening and/or Wear
Late aseptic loosening is the most common reason for implant failure. One study evaluated 18,486 primary cemented, uncemented, and combined total hip arthroplasties performed for osteoarthritis followed at least 10 years. The prevalence of acetabular loosing ranged from 0.6-13.9% and of femoral loosening from 1.7-12.5%.
Tc-99m-labeled antigranulocyte antibody Fab' fragments imaging is a more readily available, simpler examination than Indium-111-labeled autologous leukocyte scanning and has replaced that examination in Europe. A sensitivity of 91% and specificity of 47% have been found for the diagnosis of infected total hip arthroplasties.
Fluorine 18 FDG PET has been utilized in diagnosing of infection after joint replacement, but has performed similarly to three-phase bone scintigraphy.
Radiographs
Loosening is usually evaluated on radiographs. However, despite the presence of radiographic features indicating loosening, symptoms may be absent. As pointed out in one study, there is discrepancy in the identification and quantification of lucent zones, which are important indicators of radiographic loosening.
Roentgen stereogrammetry allows the detection of small changes in component position. One study found that the magnitude of subsidence in the first 2 years is a risk factor for subsequent component revision. Thus, the risk of femoral revision at 5-7 years postoperative was greater than 50% if subsidence was 1.2 mm or more during the first two years; the relative risk of revision increased 5.2 times for each mm of subsidence. One study defined loosening as migration (change in component position over time) demonstrated at roentgen stereogrammetry. Patients with migration showed signs of loosening on arthrography, scintigraphy, or at revision surgery, whereas those without migration detected with this technique did not demonstrate any evidence of loosening.
Dual Energy X-ray Absorptiometry
Dual energy x-ray absorptiometry (DXA) scanning of patients with loose prostheses indicated a decrease in bone mineral density (BMD) in every Gruen zone in comparison to the opposite unoperated hip. While comparison to the opposite side may be suboptimal, the pattern of bone loss differed from that associated with normal remodeling around successful prostheses.
Arthrography
The role of arthrography in documenting loosening of cemented components has been extensively studied. One group of investigators used refined criteria, high injection pressure, and subtraction technique and found a sensitivity of 96% and a specificity of 92% for demonstrating femoral component loosening and a sensitivity of 97% and a specificity of 68% for acetabular component loosening. Optimal arthrographic technique is important to demonstrate loosening.
The efficacy of arthrography in defining loosening of uncemented components is less well studied and less certain. One study analyzed contrast arthrography in 12 uncemented femoral components and found a sensitivity of 50% and a specificity of 100% for loosening evaluated at surgery. Contrast arthrography in 31 uncemented femoral components in another study showed a sensitivity of 59% and a specificity of 64%, lower than the results for cemented femoral components (sensitivity 76%, specificity 70%).
Anesthetic
Intra-articular injection of anesthetic that results in pain relief indicates an intra-articular cause for the symptoms. One study notes that pain relief suggests an intra-articular process, while the absence of pain relief is not helpful.
Bone Scan
The absence of increased uptake on bone scan is thought to be strong evidence against prosthetic complication. One study, however, found that even the combination of bone scan and radiography was only about 84% sensitive and 92% specific for loosening in patients without obvious radiographic loosening.
Intra-articular Radionuclide
Injection of intra-articular radionuclide was first used for evaluation of femoral component loosening, but later procedural changes allowed both acetabular and femoral components to be evaluated. The procedure is usually coupled with a bone scan, and the isotope injection is performed at the time of contrast arthrography. In one series of uncemented components, this study was 70% sensitive and 100% specific for femoral loosening with increased sensitivity over the contrast examination. In another study, the nuclear arthrogram performed better than or equal to the contrast arthrogram for evaluation of cemented and uncemented components. The combination of radionuclide and arthrographic procedures is advantageous. In a small series involving uncemented femoral stems, the sensitivity of combined exams was 90%, and the specificity was 100% when both studies were negative.
Magnetic Resonance Imaging
Initial investigations have suggested that MRI may also have a role in the evaluation of prosthetic complications. In one series, MRI documented femoral component loosening as low signal fluid collections parallel to the component on fast-spin-echo T1-weighted images.
Evaluation of Patients with Suspected Stress Shielding
Bone loss associated with decreased stress occurs around both cemented and uncemented prostheses. Although these changes may be frequent, they may not affect the long-term results. Cortical thinning, increased porosity, and decreased bone density are indicators of stress shielding on radiographs. Radiographs, however, are insensitive indicators of bone loss. One study evaluated the ability of three orthopedic surgeons to agree on the presence of bone loss around prostheses under optimal radiographic conditions. "Excellent" agreement was not achieved until bone loss averaged 70%.
Dual-Energy X-ray Absorptiometry
Dual-energy x-ray absorptiometry has been used to assess periprosthetic bone density. Reproducibility is severely compromised if identical hip rotation is not achieved on follow up examinations. One study found precision of the DXA exam to average 3.8% with a standard deviation of 1.2%. Another study noted that bone loss was obvious on radiographs when about 35% of bone loss was present on DXA scanning.
Quantitative Computed Tomography
Quantitative computed tomography (QCT) has demonstrated changes in acetabular density with time after total hip replacement.
Particle Disease (Aggressive Granulomatous Disease)
Localized areas of bone resorption occur around total hip arthroplasties in response to the release of small particles. The condition may occur in response to cement and/or polyethylene or metal particles. It is more frequent after uncemented total hip arthroplasties than cemented ones and is a more frequent complication than infection, dislocation, or extensive heterotopic bone formation. Loosening may or may not accompany the condition.
Radiographs
Radiographs are typically the first method of identifying these areas of bone resorption. One study found the area of the lesions to double in 2.2 years and the number of lesions to increase with time. However, particularly in the acetabulum, considerable bone loss is necessary before lesions are identified with certainty.
Computed Tomography
Improved CT scanning techniques enable better demonstration of bone adjacent to prostheses and may find a role for identifying granulomas. CT scans have helped determine the extent and location of clinically silent osteolysis. One study suggests a baseline CT scan seven to ten years after THA; additional follow-up CTs may then be done.
Magnetic Resonance Imaging
On MRI, focal periprosthetic intraosseous masses of low T1 and heterogeneous, largely low to intermediate T2 signal have been described in cases of aggressive granulomatous disease. Peripheral and some internal enhancement of these granulomas have been noted after intravenous gadolinium injection.
Summary
A large number of techniques are available for evaluating total hip arthroplasties. Radiographs, however, remain the standard imaging modality. Joint aspiration is the best available test for evaluation of joint infection. CT and MRI hold potential for assessing granulomatous disease.
Abbreviations
- CT, computed tomography
- DXA, dual energy x-ray absorptiometry
- MRI, magnetic resonance imaging
- NUC, nuclear medicine
- QCT, quantitative computed tomography
- WBC, white blood cell
