Key points

Introduction

Periprosthetic joint infection (PJI) is a devastating complication seen in total joint arthroplasty (TJA) patients. Traditionally, the serologic diagnosis of PJI was performed by measuring inflammatory factors of white blood cell (WBC) levels, erythrocyte sedimentation rate (ESR), and C-reactive protein (CRP). Microbiologic analysis of synovial fluid and periprosthetic tissue has also been performed using histology and synovial fluid culture, which may not be highly sensitive for detecting PJI. Modern use of novel molecular methods of diagnosis, along with the use of serum and synovial fluid biomarkers, may improve the diagnosis of PJI and provide markers that may be used to monitor the resolution of joint infection.

Introduction

Periprosthetic joint infection (PJI) is a devastating complication seen in total joint arthroplasty (TJA) patients. Traditionally, the serologic diagnosis of PJI was performed by measuring inflammatory factors of white blood cell (WBC) levels, erythrocyte sedimentation rate (ESR), and C-reactive protein (CRP). Microbiologic analysis of synovial fluid and periprosthetic tissue has also been performed using histology and synovial fluid culture, which may not be highly sensitive for detecting PJI. Modern use of novel molecular methods of diagnosis, along with the use of serum and synovial fluid biomarkers, may improve the diagnosis of PJI and provide markers that may be used to monitor the resolution of joint infection.

History and physical examination

The first step for evaluating a patient with a possible PJI is to perform a thorough history and physical examination. The patient should be asked about a prior PJI because this has been shown to increase the risk of repeat PJI. The postoperative course should be questioned for postoperative drainage, need for superficial washouts, or prolonged postoperative antibiotics. At presentation, patients with PJI may report a history of fevers, chills, pain, and loss of function, including loss of range of motion and pain with ambulation.

On physical examination, the patient should be observed for signs of illness. The vital signs should be measured, including temperature, pulse, blood pressure, and respiration rate. Inspection of the joint is critical because erythema or drainage at the incision site, and swelling and warmth of the affected joint, may indicate a PJI. According to the Musculoskeletal Infection Society PJI criteria, the presence of a sinus tract from the surface of the skin to the implant is diagnostic of a PJI. Thus, the diagnosis of PJI may be made on physical examination alone.

Serology: erythrocyte sedimentation rate and C-reactive protein

The diagnosis of PJI still poses a significant challenge and there is no consensus on the most appropriate gold standard tests to diagnose PJI. Due to the ease and low-risk nature of blood draw, serum markers are an attractive diagnostic tool for PJI. However, serum markers are all subject to confounding comorbidities, such as systemic inflammation or other infections. When studies demonstrating the success of serum markers are carefully read, it is almost invariably found that the study has excluded subjects with inflammatory comorbidities and subjects on antibiotics. When this confounded population is included, the utility of serum biomarkers is found to decline. Inflammatory markers may be elevated in obesity. The timing of infection must also be taken into account because ESR and CRP are routinely elevated in the postoperative period. ESR is elevated up to 6 weeks after surgery and CRP is elevated up to 2 weeks after surgery. If these laboratory test values are elevated postoperatively, they may not be reliable criteria for diagnosing infection. ESR and CRP levels should be measured in joint arthroplasty patients who present with pain; preoperative screening can point to the presence of infection. The clinician must carefully consider whether the use of serum biomarkers will provide a reliable diagnosis in the patient being tested. To date, none of the serum biomarker tests have been developed and optimized specifically to detect PJI.

An ESR greater than 30 mm/hr or a CRP greater than 10 mg/L should raise the suspicion of PJI. Both the American Academy of Orthopaedic Surgeons (AAOS) and the recent International Consensus on Periprosthetic Joint Infection recommend performing a joint aspirate for cell count, differential, and culture in the setting of elevated serology.

Of note, serum WBC may not be a reliable test for the diagnosis of infection because it has a low sensitivity (55%) and specificity (66%) for diagnosing PJI. Also, the test may not add to the synovial WBC test.

If the ESR and CRP are not elevated, and the clinician has no suspicion of PJI, then a joint aspirate may be unnecessary. However, it must be kept in mind that PJI can exist in the setting of normal serology, especially with organisms such as Propionibacterium acnes , coagulase-negative Staphylococcus , Candida , Corynebacterium , Mycobacterium , and Actinomyces ( Box 1 ). Fig. 1 provides an algorithm for diagnosing PJI using history, physical examination, serologic testing, and synovial fluid analysis.

Algorithm for diagnosing PJI using history, physical examination, serologic testing of ESR and C-reactive protein, and synovial fluid analysis. AFB, acid-fast bacilli.

( From Parvizi CJ, Gehrke T. Proceedings of the International Consensus Meeting on Periprosthetic Joint Infection. Brooklandville (MD): Data Trace Publishing Company; 2013. p. 160; with permission.)

Serology: beyond erythrocyte sedimentation rate and C-reactive protein

There has been recent research evaluating other serum markers that can be used in the diagnosis of PJI. Procalcitonin (PCT) is a serum marker that is elevated in the presence of stimuli, such as bacteria, that are proinflammatory. Bottner and colleagues measured serum levels of interleukin (IL)-6, PCT, tumor necrosis factor (TNF)-α, CRP, and ESR in 78 subjects undergoing revision total knee or hip replacement for PJI. CRP and IL-6 had the highest sensitivity (95%) for detecting PJI when the levels were higher than 3.2 mg/dL and 12 pg/mL, respectively, and the investigators recommended combining CRP and IL-6 as a screening test. PCT levels (>0.3 ng/mL) were very specific (98%) but had a low sensitivity (33%). On the other hand, Hügle and colleagues reported that PCT had a higher sensitivity and specificity for diagnosing septic arthritis than CRP, with a sensitivity of 93% and specificity of 75% at a PCT cutoff of 0.25 ng/mL. Theoretically, this is possible because PCT is secreted by the mononuclear phagocyte system when stimulated by lipopolysaccharide. Based on these studies, PCT may be useful for distinguishing between bacterial infections of the joint and other causes of inflammation. Using this marker may also help determine whether an antimicrobial therapy might be effective that could reduce the duration of medication and minimize antimicrobial resistance.

Glehr and colleagues compared PCT, IL-6, and interferon (IFN)-α as serum biomarkers to WBC and CRP levels for diagnosing PJI in revision arthroplasty patients. Blood samples were taken preoperatively and on the first, third, and seventh postoperative days. The results demonstrated that PCT, IL-6, CRP, and WBC correlated with the diagnosis of PJI, although IFN-α did not. IFN-α has an important role in antiviral immunity but not in antimicrobial immunity and may not be detected in bacterial infections. In serum measurements, PCT greater than 0.35 ng/mL had a sensitivity of 80% and specificity of 37%, whereas the IL-6 greater than 2.55 pg/mL had a sensitivity of 92% and specificity of 59%. Other studies found similar results in the serum of subjects diagnosed with PJI. On the other hand, Worthington and colleagues and Drago and colleagues found that PCT was not elevated in the serum of PJI subjects. However, ESR, CRP, WBC, IL-6, soluble intercellular adhesion molecule-1, and serum IgG to short-chain exocellular lipoteichoic acid were all elevated in subjects with septic loosening. IL-6 is secreted by different immune cells, such as monocytes, macrophages, fibroblasts, and T2 lymphocytes after trauma. Because IL-6 triggers the release of CRP in liver cells, it can react much faster to infection than CRP and has been reported to be a sensitive marker for bacterial infection after TJA. Wirtz and colleagues demonstrated that increased IL-6 correlates to increased inflammatory activity and suggested that IL-6 is a better indicator of postoperative inflammatory response than CRP measurements after TJA. This finding is confounded because monocytes respond to polyethylene particles by secreting IL-6, and high concentrations of IL-6 have been found in the interface membrane surrounding loose implants. Even if IL-6 levels were increased in the peripheral blood after TJA, there have been no clinical studies conducted that show a correlation between failure of an aseptic implant and increased levels of IL-6. Therefore, IL-6 maybe useful for early detection of a septic process and for monitoring success of antibiotic therapy.

Shah and colleagues measured the serum levels of 25 different cytokines before and after TJA and identified cytokines associated with surgical trauma. Three of the 25 cytokines, including IL-6, monocyte chemoattractant protein (MCP)-1, and IL-2R, were associated with postsurgical trauma, which included 1 deep infection. The changes in IL-6 and MCP-1 seem to reflect increased inflammation in the subject with deep infection, and the levels of IL-2R in the same subject were lower than average but not markedly decreased. The investigators suggested that the combination of increased IL-6 at 6 hours and reduced levels of MCP-1 at 48 hours may be associated with infection.

Synovial fluid markers

In addition to serum biomarkers, synovial fluid biomarkers may aid in the diagnosis of PJI. To date, only the α-defensin test has been specifically optimized and made commercially available for the diagnosis of PJI. Using these biomarkers may be beneficial for the diagnosis of PJI because they are measured directly from the fluid in the suspected joint. This direct measurement may prove more reliable in the setting of patients with comorbidities, such as systemic inflammation or antibiotic treatment. However, obtaining synovial fluid is an invasive procedure and synovial fluid may not always be drawn from the joint.

Synovial biomarkers can be divided into 2 categories: cytokines and biomarkers with antimicrobial functions. At the site of infection, cytokines such as IL-1β, IL-6, IL-8, and IL-17 are released from macrophages and are increased in the synovial fluid of patients with diagnosed PJI. Similar to serum biomarkers, TNF-α is elevated in synovial fluid. IFN-δ is another cytokine that is elevated in PJI because it is a glycoprotein that is released in the presence of pathogens. Vascular endothelial growth factor is also increased in the synovial fluid of patients diagnosed with PJI because it is a marker of angiogenesis. These markers are all elevated in synovial fluid but are also elevated in other inflammatory conditions, such as rheumatoid arthritis and vasculitis.

More specific synovial fluid biomarkers for detecting PJI have been evaluated, including synovial CRP, α-defensin, human β-defensin (HBD)-2, HBD-3, leukocyte esterase (LE), and cathelicidin LL-37. CRP, which is elevated in the serum and synovial fluid of PJI patients, is a liver protein that is synthesized during acute inflammation when there are increased macrophages. Synovial fluid CRP greater than 9.5 mg/L in septic revision cases was found to have a sensitivity of 85% and a specificity of 95%, with an area under the curve (AUC) of 0.92. Although this may be a valuable diagnostic test, some hospital laboratories are unwilling to measure CRP levels in synovial fluid because machines may only be calibrated for serum CRP.

α-Defensin is another synovial fluid biomarker for diagnosing PJI that has higher sensitivity and specificity than synovial fluid CRP. An α-defensin test has recently been developed and commercialized specifically for the purpose of diagnosing PJI. α-Defensins are released from neutrophils in the presence of bacteria. It has been shown that an α-defensin level greater than 5.2 μg/ml has been found to have a sensitivity of 97% and a specificity of 96%. The diagnostic accuracy of α-defensin was demonstrated in a subject population that included those with systemic inflammatory diseases and antibiotic treatment. HBD-2 and HBD-3 are similar to α-defensin because they are secreted by neutrophils in inflammatory conditions and are active against gram-negative organisms and Candida . Synovial fluid HBD-3 was elevated in aspirates of PJI subjects with PJI with an AUC of 0.745.

LE is another biomarker that is elevated in the urine of patients with urinary tract infections and has been diagnosed by the dipstick technique. Although the LE test was developed as a leukocyte count estimation for use in urinalysis, some investigators have reported its off-label use on synovial fluid. LE is specifically found in neutrophils and is measured in synovial fluid by lysis of neutrophils and measuring all intracellular and extracellular esterase activity, which could provide an estimation of the synovial fluid WBC count. This inexpensive and rapid test has 93.3% sensitivity and 77.0% specificity for diagnosing PJI when compared with microbiology culture. This test must only be conducted on nonbloody synovial fluid because the presence of blood can interfere with the colorimetric change on the dipstick seen in this test.

Human host defense peptide LL-37 is a member of the cathelicidin family and is an antimicrobial protein peptide that induces immune mediators such as IL-8, prevents the formation of biofilm, and regulates the inflammatory response. Gollwitzer and colleagues determined that LL-37 was elevated in the synovial fluid of PJI subjects and had a sensitivity of 80% and specificity of 85% for the diagnosis of PJI, with an AUC of 0.875.