Difficult-to-Detect Low-Grade Infections Responsible for Poor Outcomes in Total Knee Arthroplasty
In this article, we present the case of a patient with a stiff, painful knee after total knee arthroplasty, and with erythrocyte sedimentation rate, C-reactive protein level, and synovial fluid analysis within normal limits. Open biopsy for cultures showed Propionibacterium acnes prosthetic joint infection, which was successfully treated with 2-stage revision and intravenous antibiotics.
Case Report
A 69-year-old man with a past medical history of hypertension underwent left primary TKA in 2012. In 2014, he presented to our office complaining of chronic left knee pain and stiffness that had developed insidiously over the first 3 months after surgery and never improved, despite rigorous physical therapy (Table). 
Despite not meeting MSIS diagnostic criteria, the patient elected to undergo open biopsy for synovial culture as a last resort. During surgery, there was no purulence in the joint, and frozen section showed <5 neutrophils per high-power field. All cultures from 5 separate synovial tissue samples grew P acnes, confirming the PJI diagnosis. Cultures turned positive after being incubated an average of 12.2 days (range, 10-14 days). Sensitivities showed the organism was responsive to oxacillin. The risks and benefits of 2-stage revision surgery were discussed with the patient at the next office visit, and he decided on 2-stage revision. On November 4, 2014, he underwent open synovectomy, irrigation and débridement with iodine and Dakin solution, hardware removal, and cement antibiotic spacer placement without complication (Figures 3A, 3B). 
Just before stage 2 revision on January 6, 2015, preoperative inflammatory markers were within normal limits. During surgery, additional cultures were taken from synovial tissue. At 15 days, these cultures showed no growth, confirming eradication of the infection. The patient underwent reimplantation without complication and had an uneventful postoperative course with no wound-healing issues (Figures 4A, 4B). 
Discussion
Because PJIs with low-virulence organisms can present with normal levels of inflammatory markers and negative fluid analysis and culture from joint aspirations, they pose a diagnostic challenge for arthroplasty surgeons. In this case report, there was a low index of suspicion for PJI based on radiographic, physical examination, and laboratory findings. Our patient did not meet MSIS diagnostic criteria for PJI before undergoing open biopsy. Initial cultures from joint aspiration of synovial fluid were negative, and inflammatory markers were within normal limits. However, all 5 synovial tissue biopsy specimens that were cultured confirmed a low-grade periprosthetic infection with P acnes—likely the reason for the poor outcome. This case supports Zappe and colleagues13 and Millett and colleagues,14 who found that a subset of patients with a low-grade organism PJI had normal to mildly elevated inflammatory markers and negative fluid analysis and cultures from joint aspirations.
Hardware-involved orthopedic infections are often caused by bacteria that form a biofilm, which can be difficult to culture. Biofilm matrix binds cells into aggregates, which grow only a single colony on culture media, decreasing positive yield. Therefore, synovial fluid cultures are often negative, because of the low number of planktonic cells removed by aspirate. Using FISH and PCR, Stoodley and colleagues16 found biofilm on hardware removed for “culture-negative aseptic loosening.” This is especially important for low-grade organism infections that lack a strong inflammatory response in the joint and that may be missed with traditional screening. This may be one reason our patient’s synovial fluid cultures and inflammatory markers were negative.
Another reason these low-grade infections can be missed is that P acnes is notoriously difficult to culture—it may take up to 15 days to grow in a special medium.20 Intraoperative cultures may be read as false-negative if not incubated the right amount of time. In many hospitals, aerobic and anaerobic cultures are discarded if there is no growth after 3 to 5 days. In our patient’s case, the earliest that cultures turned positive was on day 10—which is consistent with other reports, including one by Butler-Wu and colleagues,15 who suggested a minimum incubation of 13 days for optimal recovery of organisms. Our case highlights the importance of lengthening incubation to allow for growth of low-virulent organisms. Given the different types of management used for PJI and aseptic loosening, it is imperative that surgeons take cultures during revision TJA and that cultures are held up to 14 days to allow enough time for low-virulence organisms to grow.
Fortunately, PJI with low-virulence organisms can be treated successfully. Treating P acnes PJI with exchange arthroplasty and IV antibiotics has documented success rates as high as 92%.21 Again, we emphasize the importance of obtaining intraoperative cultures to determine antibiotic sensitivities, which can guide treatment. Our patient’s infection was eradicated with 2-stage revision and IV antibiotics, and his symptoms, ROM, and function improved significantly.
Diagnosing PJI after TJA can be challenging, as there is no definitive test that is sensitive, specific, rapid, and minimally invasive. Researchers have looked for novel serum or synovial fluid biomarkers that may be elevated in PJI. Synovial interleukin 6 (IL-6) and synovial α-defensin show great promise. In 2 separate studies, elevated IL-6 levels strongly correlated with infection.22,23 Jacovides and colleagues23 found that a synovial IL-6 level higher than 4270 pg/mL had a 100% positive predictive value and a 91% negative predictive value for diagnosing PJI. In some trials, synovial α-defensin has shown up to 100% sensitivity and specificity for PJI diagnosis. Most notably, in a trial by Frangiamore and colleagues,24 α-defensin levels were elevated to statistically significant levels in P acnes PJI, indicating this test may help in diagnosing PJI with low-virulence organisms. Finally, PCR has also shown promise in detecting low-grade joint infections. PCR uses 16 primers that allow not only for the identification of pan-genomic bacterial markers, specific bacterial organisms, and Candida, but also for the presence of antibiotic resistance markers. Use of pan-genomic PCR also allows for detection of a wider variety of pathogens, including organisms commonly missed by conventional culture methods.25Early intervention can significantly improve outcomes in PJI. Therefore, we recommend maintaining a high index of suspicion for low-virulence PJI in patients with chronic pain and decreased functionality after TJA with well-placed implants, despite their not meeting current MSIS diagnostic criteria for PJI. As new microbiological tools for detecting PJI with low-grade organisms are developed, use of these technologies can be incorporated into the diagnosis algorithm. Screening tools more sensitive in detecting low-grade organisms can help avoid the morbidity associated with interoperative synovial biopsies for culture and can allow for more efficient surgical planning. These tools, along with increased clinical awareness of potential PJIs, ultimately will lead to earlier detection, accurate diagnosis, and optimal treatment.
Am J Orthop. 2017;46(3):E148-E153. Copyright Frontline Medical Communications Inc. 2017. All rights reserved.
