Prosthetic Joint Infection

Javier Cobo; Jose Luis Del Pozo


Expert Rev Anti Infect Ther. 2011;9(9):787-802. 

In This Article


Diagnosis and optimal treatment of PJI remains a challenge. Early postoperative infection and acute hematogenous are usually easily suspected, but sometimes it is difficult to distinguish superfical wound infections or cellulitis from true implant deep infections. However, late chronic infections are challenging to predict. Clinical signs and symptoms, laboratory tests, radiography and joint aspiration are insensitive, nonspecific or both.[26] In addition, artifacts produced by the prosthetic devices themselves hamper cross-sectional imaging modalities such as CT and MRI. Late chronic PJI has been defined as the presence of at least one of the following four criteria: isolation of the same microorganism from two cultures of joint aspirates or intraoperative periprosthetic tissue specimens; presence of acute inflammation on histopathological examination of periprosthetic tissue; sinus tract communicating with the prosthesis; or purulence in a joint space (as determined by the surgeon).[4,27,28]

Aseptic failure (i.e., aseptic loosening) is usually defined as failure of a prosthesis in the absence of any of these criteria.[29] Distinguishing infection from aseptic loosening is extremely important because their treatments are completely different. It has been suggested that some cases of aseptic failure are missed cases of PJI.[30] We believe that all patients diagnosed with aseptic failure should be evaluated via erythrocyte sedimentation rate and C-reactive protein testing. Furthermore, preoperative joint aspiration, with differential cell counts and culture of the synovial fluid, and histopathological studies – besides imaging – should also be performed. Guidelines for the diagnosis of periprosthetic joint infections at the hip and knee are recently available from the Academy of Orthopaedic Surgeons.[31] These guidelines emphasize the role of joint aspirate (sending it for white blood cell count) and differential white blood cell count.

Laboratory Studies

Increased peripheral blood leukocytes, erythrocyte sedimentation rate and C-reactive protein levels are neither sensitive or specific for PJI. After surgery, the erythrocyte sedimentation rate and the C-reactive protein level may remain elevated for weeks. Therefore, serial postoperative measurements are more informative than a single value. In addition, C-reactive protein and erythrocyte sedimentation rate may be elevated as a result of other inflammatory conditions or, conversely, may be falsely negative in the context of suppressive antimicrobial therapy or low-virulence organisms. However, a normal erythrocyte sedimentation rate along with a normal C-reactive protein level is suggestive of a very low probability of infection.[24] The role of new markers, including IL-6, procalcitonin and TNF-α, remains to be defined. Consequently, we recommend erythrocyte sedimentation rate and C-reactive protein testing for patients assessed for periprosthetic joint infection.

Histopathological Studies

Intraoperative frozen sections of periprosthetic tissues have excellent accuracy in predicting a diagnosis of PJI but have moderate accuracy in ruling out the diagnosis. The definition of acute inflammation in the periprosthetic tissue varies in studies from an average of one to ten neutrophils per high-power field at a magnification of 400.[32] There is insufficient information to distinguish five from ten neutrophils per high-power field as the best threshold needed for diagnosis. In addition, insufficient information is available to determine the efficacy of frozen sections in patients with an underlying inflammatory arthropathy. A meta-analysis of histopathological studies indicated that frozen section is a very good 'rule-in' test (i.e., a positive result has a high likelihood of infection; likelihood ratio [LR]+: 23), but is a relatively low-value 'rule-out' test (i.e., a negative result does not have a high likelihood of absent infection; LR-: 0.23).[31] Previous treatment with antibiotics may modify the nature of the inflammatory response, with more chronic inflammatory cells (i.e., plasma cells) and fewer neutrophils present. However, the nature and degree of infiltration with inflammatory cells may vary markedly among specimens from the same patient and even within individual tissue sections. Accordingly, careful sampling of the tissue surrounding a replacement is necessary.

Preoperative Joint Aspiration With Differential Cell Counts

Preoperative joint aspiration with differential cell counts is a valuable diagnostic tool for differentiating a septic from an aseptic process. We recommend joint aspiration of patients being assessed for late periprosthetic knee infections who have an abnormal erythrocyte sedimentation rate and/or C-reactive protein results. A synovial-fluid leukocyte count of more than 1700 per cubic millimeter or more than 65% neutrophils had sensitivities for prosthetic knee late infections of 94 and 97%, respectively, and specificities of 88 and 98%, respectively, in patients without underlying inflammatory joint diseases.[10] In a study of 201 painful hip arthroplasties, other investigators found that a synovial fluid leukocyte count of >4200/ml was 84% sensitive and 93% specific, and that a leukocyte differential of 80% neutrophils was 84% sensitive and 82% specific.[33]

Microbiologic Studies

Cultures are important in the diagnosis of infections associated with indwelling devices. They are critical to direct antimicrobial therapy. Commensal organisms on the skin most commonly infect implantable biomedical devices.[28] In this situation it is usually impossible to decide whether it is clinically significant or a contaminant from the skin of the patient, the medical staff obtaining the sample or the laboratory staff processing it.

Culture of the Synovial Fluid Aspirated fluid should be sent for microbiologic culture. Fluid Gram staining is very specific (more than 97%) but has a low sensitivity (less than 26%). It should be performed in cases of early and hematogenous infections but have no role in late chronic infections.[31] The most accurate specimens for detecting the infecting microorganism(s) are sampled from periprosthetic tissue. Intraoperative swab cultures have a low sensitivity, so swabs should be avoided. Positive cultures are detected in 65–94% of the cases.[28,32,34,35] At least six intraoperative tissue specimens should be sampled for culture.[28,34,35] The isolation of an indistinguishable microorganism from three or more independent specimens is highly predictive of infection (sensitivity: 65%; specificity: 99.6%; LR: 168.6).[28] The sensitivity of tissue culture increases as the number of specimens collected increases. Cultures of sinus tract exudates are often positive due to microbial skin colonization, making the interpretation very difficult, and accordingly should be avoided.

Cultures may be negative due to prior antimicrobial exposure, a low number of organisms, an inappropriate culture medium, fastidious organisms or prolonged transport time to the microbiology laboratory. To increase detection of microorganisms, enrichment broth media may be used; however, this carries the risk of amplifying any contaminant present in the specimen, potentially resulting in suboptimal specificity. When infection by an unusual microorganisms is present, routine cultures may be negative and special culture techniques (e.g., fungal culture, mycobacterial culture) may be necessary. In the same way, it is recommended to discontinue any antimicrobial therapy at least 2 weeks prior to tissue sampling for culture.[34] Perioperative prophylaxis at revision surgery should be replaced by empirical treatment started after tissue specimens have been collected for culture.[36] It has been demonstrated that there is a significant risk that PJIs will not be detected if a prolonged (14 day) microbiological culture period is not applied and that organisms isolated >1 week after the start of culture do not overproportionally reflect contaminating strains.[37]

Culture After Prosthesis Sonication Organisms associated with the infected joint may not be concentrated in the periprosthetic tissue but may actually be attached to the prosthesis as biofilm microorganisms. Consequently, obtaining a sample from the prosthesis might improve the diagnosis of PJI.[30,38,39] Vortexing and sonication of the implant is simple and can be performed in most microbiology laboratories. Culture of samples obtained by sonication of prostheses is more sensitive (78.5%) than conventional periprosthetic-tissue culture (60.8%) or synovial-fluid culture (56.3%) for the microbiologic diagnosis of prosthetic hip and knee infection, especially in patients who had received antimicrobial therapy within 14 days before surgery.[39] The specificities of sonicate-fluid culture, tissue culture and synovial-fluid culture were 98.8, 99.2 and 98.1%, respectively. The sensitivity and specificity of sonicate-fluid Gram staining were 44.7 and 100.0%, respectively. However, this technique is not yet widely available, and there is a considerable risk of contamination during processing.[38]

Molecular Tools Molecular tools are attractive for the diagnosis of device-related infection, especially in the case of culture-negative infection or the presence of fastidious microorganisms. Broad-range PCR amplification of conserved bacterial deoxyribonucleic acid sequences (e.g., 16S ribosomal RNA gene) followed by direct sequencing of the amplified product (i.e., for species determination) is a powerful technique that permits rapid detection and identification of bacteria.[40] This approach may, however, be associated with false-positive results. Also, this approach requires sequencing (which is not a rapid approach), and does not provide antimicrobial susceptibility results.

Imaging Studies

Plain Radiographs Plain radiographs are neither sensitive or specific; however, they may be helpful to detect infection when they are studied serially over time after implantation. Findings such as radiolucency, osteolysis and migration are observed in both infection and aseptic loosening.[41]

Computed Tomography Computed tomography provides better contrast between normal and abnormal tissue than plain radiography, and it is useful in detecting joint effusion, sinus tracts, soft tissue abscesses, bone erosion and periprosthetic lucency; however, imaging artifacts caused by metal implants limit its use.

Magnetic Resonance Imaging MRI can only be performed in patients with implants that are safe for MRI (i.e., nonferrimagnetic implants), such as those composed of titanium or tantalum, and provides greater resolution for soft-tissue abnormalities.

Radionuclide Imaging Radionuclide imaging is not affected by metallic hardware and is the current imaging modality of choice for evaluation of suspected joint replacement infection.[1] The primary role of nuclear medicine in the evaluation of a painful joint replacement is to differentiate aseptic loosening from infection. Combined leukocyte-marrow scintigraphy is currently regarded as the imaging modality of choice for diagnosing PJI.[42]

18F-fluoro-2-deoxyglucose-PET18F-fluoro-2-deoxyglucose (FDG)-PET enables visualization of hyperglycolytic inflammatory cells (i.e., leukocytes, macrophages and other immunologically active cells). Applied criteria for positivity can grossly be divided into four groups: FDG uptake in the periprosthetic soft tissue; increased FDG uptake at the bone–prosthetic interface; increased FDG uptake at the bone–prosthetic interface, while emphasizing that FDG uptake limited to the soft tissues adjacent to the neck of the prosthesis is not considered suggestive of infection (for hip prostheses only); and other criteria. Increased FDG uptake around the femoral head and neck or around the distal tip of the hip prosthesis (possibly due to foreign body reaction) may persist for years following hip arthroplasty in asymptomatic patients; consequently, it should not be interpreted as periprosthetic infection. However, FDG uptake along the interface between bone and hip prosthesis is virtually never seen in asymptomatic patients or in those with aseptic loosening and is, therefore, highly suggestive of infection. FDG uptake limited to the soft tissues or adjacent to the neck of the prosthesis is not considered suggestive of infection. Published results to date, however, are inconclusive. In a meta-analysis that reviewed 11 studies (including 635 prosthesis), the pooled sensitivity and specificity of FDG-PET for the detection of prosthetic hip or knee joint infection were 82.1% (95% CI: 68.0–90.8%) and 86.6% (95% CI: 79.7–91.4%), respectively.[43] The lower specificity of FDG-PET in knee prostheses may be related to the relatively limited knowledge about the incidence and pattern of nonspecific FDG uptake around knee prostheses. Meta-analytically, FDG-PET achieved moderate-to-high sensitivity and specificity in detecting prosthetic hip or knee joint infection. However, this result should be interpreted cautiously because significant heterogeneity was identified among the results of individual studies. Combined leukocyte–marrow scintigraphy is currently regarded as the imaging modality of choice for diagnosing PJI.[42] Two studies compared FDG-PET and combined leukocyte–marrow scintigraphy in the diagnosis of prosthetic hip and knee infection.[44,45] For prosthetic hip infection, FDG-PET and combined leukocyte–marrow scintigraphy demonstrated comparable specificities (93 and 95.1%, respectively); however, FDG-PET exhibited a substantially higher sensitivity (95.2 and 50%, respectively).[44] For prosthetic knee infection diagnosis, sensitivity and specificity of FDG-PET were 100 and 73%, respectively, and sensitivity and specificity of combined leukocyte–marrow scintigraphy were 100 and 93%, respectively.[45]

Antigranulocyte Scintigraphy Antigranulocyte scintigraphy with monoclonal antibodies or antibody fragments may be another attractive approach to detect PJI. A recent meta-analysis on the diagnostic performance of antigranulocyte scintigraphy included 13 studies with a total sample size of 522 prostheses and reported estimates of sensitivity and specificity of 83 and 80%, respectively.[46]