Introduction
While total joint arthroplasty (TJA) is considered a relatively safe procedure, complications may arise. Periprosthetic joint infection (PJI) continues to be the leading cause of failure for primary total knee and revision total hip arthroplasty. , The number of TJAs performed in the United States is anticipated to increase significantly by the year 2030 and exceed 3.8 million annually. Consequently, associated PJIs are projected to increase from 17,000 to 266,000 during this time. , Diagnosing PJI remains a major challenge and is a critical first step in appropriate management of this condition. The lack of a single test with absolute accuracy has led organizations and institutions to propose diagnostic criteria for PJI. This chapter will focus on some of the most commonly used markers for diagnosis and describe the evolution, differences, and evidence behind various definitions for PJI.
Diagnostic Tests
In the absence of an absolute test for PJI diagnosis, a combination of metrics is utilized that includes surgical information, physical findings, results of various tests on serum and synovial fluid, and histologic evaluation of periprosthetic tissue. The individual diagnostic performance of each marker, and its availability and relative cost, should guide clinicians in selecting the most suitable diagnostic approach. It should also be noted that test performance may differ according to the joint assessed (hip vs. knee), the time interval from surgery, and the virulence of the infecting organism.
Clinical Manifestations
Clinical findings can be very helpful in the assessment of a painful joint. Frequently reported signs and symptoms of PJI are pain, joint effusion, erythema, fever, drainage, and the existence of a sinus tract communicating with the prosthesis. While all official PJI definitions consider a sinus tract as definitive evidence of PJI, , , , other clinical manifestations are less specific and compel physicians to seek further evaluation.
The International Consensus Meeting (ICM) held in 2018 raised the importance of clinical findings in evaluation of PJI. Pain is considered the most common symptom in PJI. , , In one study evaluating patients undergoing revision surgery for failed total knee arthroplasty (TKA) due to PJI and aseptic causes, over 90% of patients in both groups complained of pain, making pain alone a very nonspecific symptom. Effusion and erythema following TKA are present in a significantly higher percentage of patients with PJI compared with patients experiencing failure due to aseptic reasons. , Systemic signs such as fever and tachycardia, though less common, are more specific findings related to PJI. However, they are also the least common.
From a practical standpoint, clinical findings can be used to establish a pretest probability of a diagnosis as well as aid in the subsequent interpretation of tests ordered as applied widely across the medical profession. These differences in pretest probability may also significantly alter the post-test probability of patients who present with similar laboratory findings. , A Fagan nomogram has been proposed as a practical easy-to-use tool to aid in the interpretation of clinical findings, taking into account serum and synovial fluid marker results ( Fig. 26.1 ).

Serum Markers
An inflammatory reaction due to an infection may result in an elevation of serum biomarkers. Recognition of the importance of serum testing in the initial evaluation of a painful artificial joint has resulted in the evaluation and testing of many serum biomarkers throughout the years.
White blood cell (WBC) count is typically ordered as part of routine blood work at the time of initial evaluation of an infection. However, unlike its utility in other fields of medicine, it has no role in the diagnosis of PJI due to very low sensitivity—around 20%. , Erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) remain the most commonly used markers in the workup of patients with suspected PJI; the vast majority of definitions for PJI include these two tests. These tests have earned their spot as screening tools in the initial workup of PJI because they are readily available and inexpensive. Several studies investigated the performance of these markers—while they are relatively sensitive, they lack specificity. In a meta-analysis of 23 studies (>3000 patients), the pooled sensitivity and specificity for ESR were 75% and 70%, respectively. For CRP, the pooled sensitivity and specificity were 88% and 74%, respectively. If elevated levels of one of these tests was considered as abnormal, then the sensitivity rose to 96%. , Notably, the cutoff used in the evaluation of both markers ranged considerably, with the most frequently used threshold being 30 mm/h for ESR and 1 mg/dL for CRP. These are also the cutoffs recommended by the ICM for diagnosis of chronic PJI. , Certain underlying inflammatory conditions, such as rheumatoid arthritis and other crystalline deposition diseases, exhibit higher baseline CRP and ESR, raising concern regarding the utility of commonly used thresholds in this subset of patients. However, the evidence suggests that these markers can be used effectively and without the need to adjust threshold rates. , Low-virulence infections and culture-negative cases have also been pointed out as instances in which commonly used cutoffs may be inaccurate. Kheir et al. performed a retrospective study on over 1000 revision arthroplasty cases (549 for PJI, 653 for aseptic causes) and compared commonly used serum marker cutoffs stratified based on the underlying organism-causing infection. Interestingly, they reported that culture-negative PJI and PJI caused by slow-growing organisms, such as coagulase-negative Staphylococcus , resulted in lower levels of ESR and CRP, making them prone to being misdiagnosed as aseptic failure.
Interleukin-6 (IL-6) produced by lymphoid and nonlymphoid cells as part of the inflammatory cascade is another marker that has been investigated extensively. Serum IL-6 levels increase rapidly in cases of infection and precede the release of CRP, making them a potentially more sensitive indicator for early PJI. Another possible benefit of IL-6 is its rapid return to normal levels following arthroplasty (15 hours) compared with CRP (62 hours), further suggesting a theoretical advantage for diagnosis of acute PJI. Albeit there are theoretical advantages, a meta-analysis by Xie et al. found the pooled sensitivities and specificities of serum IL-6 to be 72% and 89%, respectively. It should be noted that the studies included in the meta-analysis differed significantly in cutoffs used, which may have affected results, limiting a definite conclusion. However, a lack of clear advantage and lack of easy availability, together with associated expense, has made IL-6 a less appealing screening test than CRP and ESR.
Another serum marker that was recently explored is D-dimer. D-dimer has been used in the workup of venous thromboembolism (VTE) for many years. The rationale behind its use in the workup of PJI was based on several studies showing that systemic and local infections result in fibrinolytic activities. , Shahi et al. were the first to describe the association between elevated D-dimer (>850 ng/mL) and PJI. In their study, they were able to show higher sensitivity (89%) and higher specificity (93%) for D-dimer compared with ESR (73% and 78%) and CRP (79% and 80%), respectively. Notably, they excluded 11 patients due to multiple reasons that could have caused a rise in the level of D-dimer. Since the initial publication, there has been a growing interest in this serum marker, with conflicting results. , Zhang et al. examined the diagnostic accuracy of D-dimer through a meta-analysis of 9 studies that included 1592 patients total. The pooled sensitivity and specificity of D-dimer for PJI diagnosis were 82% and 73%, respectively. These results suggest a lower performance of D-dimer compared with the original report by Shahi et al. Nonetheless, it seems that it may have a similar performance to CRP, and their combined performance shows promising outcomes. As evidence suggests that D-dimer does have a role in PJI diagnosis, it was evaluated and included in the 2018 ICM definition for PJI. It should be noted that the discrepancies seen with D-dimer could be the result of variability in testing of D-dimer between labs. There is no standard method to measure D-dimer; thus, the level of this marker in serum versus whole blood and the method used to measure the marker vary greatly. Most publications in orthopaedics seem to use the HemosIL D-Dimer 500 kit on the ACL TOP instrument, which measures D-dimer in patient plasma and reports it in fibrinogen-equivalent units (ng/mL), It should be noted that D-dimer testing, assays, and reported outcomes are more complex and divergent compared with other commonly used markers.
Finally, while serum markers should indeed be the first step in the workup of patients suspected of PJI, these screening tests carry a certain false-negative rate. , Therefore, clinical judgment is critical in investigation of a painful joint to ensure that diagnosis of PJI is not overlooked. The presence of erythema, reduced range of motion, tachycardia, proximity to index surgery, multiple prior surgeries in the index joint, and elevated serum polymorphonucleate percentage (PMN%) were found to be important parameters that should raise suspicion for PJI and prompt aspiration of the joint even when serum testing may be normal. ,
Synovial Fluid Markers
Any elevated serum marker or a high clinical suspicion should prompt a joint aspiration based on Infectious Diseases Society of America (IDSA), American Academy of Orthopaedic Surgeons, and ICM recommendations. , , In general, synovial fluid provides more accurate information regarding the probability of infection compared with serum markers. However, it should be noted that measuring a biomarker in a close compartment is useful only if it is produced in a defined space with a barrier against the bloodstream.
Synovial fluid analysis for WBC count and PMN% are the most commonly used tests in the workup of PJI, with a high reported sensitivity and specificity ranging between 90% to 97% and 85% to 95%, respectively. , , A wide range of diagnostic thresholds has been proposed, limiting the interpretation and standardization of results. A variety of threshold values has been proposed in recent years, ranging from 730 [38] to 4450 cells/μL [36] for synovial fluid WBC count and 65% to 90% for PMN%. , A meta-analysis to evaluate the performance of synovial WBC count and PMN% demonstrated a sensitivity and specificity greater than 90%. Importantly, the specificity of synovial fluid WBC count was significantly increased by using the threshold value ≥3000 cells/μL compared with threshold values <1700 cells/μL ( P = .006), indicating that by using a threshold value above 3000 cells/μL, the WBC test resulted, on average, in 20 less false-positive diagnoses for every 100 cases. Thus, 3000 cells/μL was chosen as the threshold for chronic PJI by the 2018 ICM.
Anatomic site also appears to influence the level of synovial fluid WBC count and PMN%. In the abovementioned meta-analysis sensitivity of synovial fluid, PMN% was significantly higher in detecting PJI after TKA than THA ( P = .034). These findings support those who believe that thresholds should be different for TKA and THA. The optimal thresholds appear to be higher in THA compared with TKA. However, in a multicenter study by Higuera et al., optimal thresholds of 3966 cells/μL for WBC count and 80% for PMN% were suggested for the diagnosis of PJI in THA patients. These thresholds seem to be closer to values for TKA than was previously believed. Also, the differences between TKA and THA could be biased by the differences in the type and virulence of the underlying organism. Further research should help clarify the need for an individual optimal threshold for PJI based on anatomic site.
Leukocyte esterase (LE) is an enzyme released from active neutrophils that can be measured using a colorimetric strip at point of care either preoperatively or intraoperatively. In a meta-analysis comparing different synovial markers, LE showed the highest specificity (95%). However, it also had the lowest sensitivity (77%) compared with other synovial markers, making it a poor rule-out test. Tarabichi et al. showed that initial serologic screening can immensely help in the interpretation of LE results. When LE results are concordant with ESR and CRP levels, PJI was accurately diagnosed or refuted with >95% certainty. When LE was discordant from serology, only the stricter LE thresholds (2+ or negative) were adequately diagnostic to potentially suggest a change in clinical decision-making, whereas the lesser thresholds (1+ or trace) should serve to prompt further workup. LE has been incorporated in the 2018 ICM definition for PJI to receive similar weight as synovial WBC count.
Alpha-defensin is an antimicrobial peptide released from activated neutrophils that acts to destroy the cell membrane of infective organisms. The alpha-defensin immunoassay provides a qualitative result at a relatively high cost per test. While preliminary reports showed outstanding performance reaching nearly 100% sensitivity and specificity, making this test almost an absolute test for PJI, further studies have reported a more realistic picture, with a recent meta-analysis showing a pooled sensitivity and specificity of 87% and 97%, respectively. Another recent meta-analysis showed a similar performance when examining only prospective studies. It also reported no differences in performance based on the method used (enzyme-linked immunosorbent assay [ELISA] vs. lateral flow) for testing. Shohat et al. examined the performance of an evidence-based algorithm for detecting PJI using the 2018 definition with and without taking the alpha-defensin results into consideration. They showed similar diagnostic performance of the algorithm and definition with and without alpha-defensin. Given the high cost and that there was no additional benefit in using alpha-defensin compared with commonly used and readily available tests, they recommend not using this test routinely.
Other synovial fluid markers that have been investigated are CRP and IL-6, with a pooled sensitivity of 85% and 81% and pooled specificity of 88% and 94%, respectively, based on a recent meta-analysis. These results do not support their routine use, as they do not appear to add benefit over currently used markers. Also, the additional benefit of synovial CRP over serum CRP has been called into question since this protein may not be synthesized in the joint; rather it is synthesized exclusively in the liver. Synovial CRP, however, appears to provide additional an datapoint in some cases in which the results of available tests do not allow diagnosis or exclusion of PJI.
Intraoperative Findings
Periprosthetic tissue examination can be extremely valuable in the diagnosis of PJI, especially in cases in which a definite diagnosis could not be reached preoperatively. Shohat et al. developed an evidence-based algorithm for diagnosing PJI and showed that out of 73 patients in whom a diagnosis was not reached preoperatively (17% of the entire cohort), 58 (79%) were correctly diagnosed based on intraoperative findings.
Charosky et al. was the first to describe the use of frozen-section histology to show an inflammatory response that could be suggestive of infection. In a meta-analysis comparing frozen-section histology with culture, the conclusion was that frozen sections taken intraoperatively accurately predicted PJI in culture-positive cases but were less reliable in ruling out the diagnosis. The threshold of PMN infiltrate that is necessary to make a diagnosis is controversial. Nevertheless, there are enough studies to support diagnostic thresholds of either 5 or 10 PMN in each of 5 high-power fields (HPFs). The Musculoskeletal Infection Society (MSIS) and ICM currently support 5 PMN per HPF as the threshold for PJI. , There are several limitations associated with histopathologic analysis that are worth noting and could significantly impact its utility in practice: (1) tissue sample selection is subjective and prone to error, (2) samples can be compromised during transfer and preparation, and (3) sample examination is highly user dependent. Other challenges include presence of foreign body macrophages that mimic neutrophils and neutrophils entrapped in superficial fibrin or adherent to endothelium or small veins, as those should be disregarded.
Intraoperative tissue samples should be sent for microbiologic examination. Positive cultures obtained from periprosthetic tissue at the time of surgery are not only important for diagnosis, they are also extremely valuable for treatment. The MSIS recommends that tissue samples be obtained from representative periprosthetic tissue or (synovial) fluid. At least three and no more than five samples should be sent for culture; in theory, Gram staining can rapidly aid in the diagnosis of PJI while in the operating room. However, while very specific, its low sensitivity, ranging between 5% to 30%, precludes its use in practice. , In their meta-analysis, Lee et al. showed that compared with other commonly used intraarticular markers, culture had the least sensitivity, with a pooled result of 62%. These findings are in accordance with previously published literature and support the practice of obtaining at least 5 tissue samples at the time of revision surgery. While 2 positive cultures of the same organism are highly suggestive of infection and agreed upon by the majority of official bodies as definite infection, a single positive culture result may generate confusion and difficulties in interpretation, especially in cases in which a low-virulence organism is isolated. While many would consider a single growth as a contaminant, evidence from the Danish registry shows that first-time clinically aseptic revisions with one positive culture had a higher risk of re-revision for PJI (relative risk, 2.63; P = .020). It was concluded that unexpected bacterial growth with common bacteria may be clinically important, even if only one of five biopsy cultures is positive. These findings are accounted for in the most recent diagnostic criteria.
Formal Definitions Related to Periprosthetic Joint Infection
Diagnostic Criteria
Prior to the last decade, research in the field of PJI was limited by lack of official standardized criteria. The earliest PJI definition was suggested in 2011 by a group of MSIS experts and was followed by many other bodies, including the Infectious Diseases Society of North America (IDSA) in 2013 and the ICM in 2013 and later in 2018. These definitions have become globally accepted among surgeons and have radically improved diagnostic assurance and improved both treatment and research in the field of PJI. Their use in research allowed for consistency in definition between studies and enhanced the potential for collaboration and the overall quality of literature.
The MSIS 2011 criteria ( Table 26.1 ) relies on major and minor criteria similar to the Dukes criteria for endocarditis. A single major criterion or four out of six minor criteria are thought to be suggestive of infection. The IDSA definition uses the same major criteria as MSIS but also includes purulence without another known etiology as definitive evidence of PJI. Unlike the MSIS, the IDSA does not use biomarkers for diagnosis. Rather, it uses factors such as growth of a virulent organism from a single culture or positive histopathology results as supporting evidence for PJI.
Major Criteria |
---|
1. There is a sinus tract communicating with the prosthesis. |
2. A pathogen is isolated by culture from at least two separate tissue or fluid samples obtained from the affected prosthetic joint. |
Minor Criteria |
1. Elevated serum erythrocyte sedimentation rate (ESR) and serum C-reactive protein (CRP) concentration |
2. Elevated synovial leukocyte count |
3. Elevated synovial neutrophil percentage (PMN%) |
4. Presence of purulence in the affected joint |
5. Isolation of a microorganism in one culture of periprosthetic tissue or fluid |
6. Greater than 5 neutrophils per high-power field in 5 high-power fields observed from histologic analysis of periprosthetic tissue at ×400 magnification |
a Diagnosis is made with one of the major criteria or four of the six minor criteria.
In August 2013, hundreds of international specialists assembled from around the globe for the first ICM. The experts endorsed the MSIS definition of PJI and modified it slightly. The ICM added the leukocyte esterase test as a minor criterion equivalent to elevated WBC count in synovial fluid and excluded the purulence surrounding the prosthesis from the minor criteria. Thus, out of the five minor criteria left, three should be present in order to diagnose PJI ( Table 26.2 ). Acceptable thresholds for minor criteria were also determined in this meeting based on time from index arthroplasty ( Table 26.3 ).
