Should all hip and knee prosthetic joints be aspirated prior to revision surgery?

Total joint arthroplasty is one of the most commonly performed surgical procedures in orthopedics. Unfortunately, a prosthetic joint eventually requires revision, either due to mechanical failure (i.e. instability, malalignment or malpositioning) or loosening of the prosthesis [1, 2]. Published records describe that around 10–20% of patients undergoing revision surgery have unexpected positive cultures during revision [3‐6], and these patients have a worse outcome compared with aseptic cases [7‐10]. Therefore, it is critical to rule out a periprosthetic joint infection (PJI) prior to revision surgery because the revision of an infected prosthesis requires other surgical and antibiotic interventions than a revision for a non-infected prosthesis. Currently, there is no single test to confirm or exclude a PJI. Low serum inflammatory parameters such as C-reactive protein (CRP) and erythrocyte sedimentation rates (ESR) are false negative in a proportion of patients with a low-grade infection [11‐14]. For this reason, it is advised not to rely on serum inflammatory parameters to rule out an infection [15]. Consequently, many experts recommend routine synovial fluid analysis prior to revision surgery. In particular, a synovial leucocyte count below 1500 cells/μL and/or negative biomarkers (e.g. alpha defensin or calprotectin) have a very high negative predictive value and may be used in this regard [16‐19]. However, it is known that specific clinical and implant characteristics increase the risk of having a PJI, such as revision prostheses and early prosthetic loosenings [20, 21]. On the other hand, prostheses being revised for mechanical reasons may have a lower odd of having a PJI. Along these lines, identifying a subgroup of patients undergoing revision surgery in whom a synovial fluid aspiration is not necessary would be favorable. Therefore, the aim of our study was to investigate if specific clinical and implant characteristics could be identified to rule out a PJI and in whom a synovial fluid aspiration would be redundant.

In this retrospective observational cohort study, we analyzed whether specific clinical or implant characteristics could be identified to rule out a PJI prior to revision surgery without the need for synovial fluid aspiration. Our results indicate that patients who undergo revision surgery due to mechanical failure of the prosthesis probably do not need to be aspirated if accompanied with low-inflammatory parameters. We could not identify a subgroup of patients with loosening of the prosthesis in whom not aspirating the joint would be justifiable.

To our knowledge, we are the first who explicitly report on the incidence of PJI in prostheses with mechanical failure. A study by Khalid et al. did determine the incidence of unrecognized PJI in 72 patients undergoing revision for aseptic failure without any clinical suspicion of infection. These so-called aseptic failures were defined as the patients with implant instability and/or polyethylene wear but also included cases with implant loosening. Within this population, five patients turned out to have an unrecognized PJI (6.9%), but it was not reported if the infections were only observed in the patients with prosthetic loosening [23]. The absence of infection in our patient population with mechanical failure should be interpreted with some caution. It should be noted that the gross majority of patients with mechanical failure in our analysis had low serum inflammatory markers. For this reason, we cannot conclude whether a synovial fluid aspiration is still needed in patients with an elevated CRP. In addition, our results need to be confirmed in a larger cohort of patients with merely mechanical failure, and a subanalysis on hips and knees should be performed. In our study, the majority of patients had mechanical failure of knees. As the aspiration of a hip joint is more complicated compared with knees, this group of patients is of particular interest.

Regarding patients with prosthetic loosening, the incidence of PJI in this particular group was high. The PJI incidence of 19% in patients with prosthetic loosening in our study is comparable with other reports [3‐6]. We could not identify a subgroup of patients with prosthetic loosening in whom a PJI could be ruled out preoperatively without synovial fluid aspiration. Commonly described patient-related risk factors for a PJI such as male sex, diabetes, rheumatoid arthritis and obesity did not appear to be risk factors in our study [24‐26]. A possible explanation might be that the importance of host-related factors are more pronounced when analyzing the risk of acute PJIs, rather than chronic ones. We did found that certain implant characteristics were associated with a higher odds of having a PJI. In accordance with previous studies [20, 21, 27], early loosening occurred more often in the presence of a PJI. Portillo et al. demonstrated that in the first 2 years after implantation, revisions were performed more often for PJI (69%) than for aseptic loosening (16%). Indeed, in most studies, early loosening is defined as loosening within 2 years after implantation [20, 21]. In our study, the most prominent difference was observed using a cut-off of 4 years instead of 2 years. In addition, we identified that a previously treated infection of the affected joint and/or having a revision prosthesis are both associated with a higher odds of having a PJI. Despite the identification of these implant-related risk factors, PJIs were still detected in our population far after the 4-year implantation period. Moreover, even combined with a low serum CRP, a PJI could not be ruled out, not even when using a lower threshold of 5 mg/L instead of the accepted 10 mg/L used in the diagnostic criteria for PJI [22]. This finding again underlines that we cannot solely rely on serum CRP and ESR, as bacteria are able to remain in a ‘dormant’ and asymptomatic state within a biofilm without inducing any systemic inflammatory response [28, 29].

An important limitation of our study is the limited number of patients with a PJI (n = 20), which hampers the possibility to perform detailed subanalyses on high and low-risk groups for infection. Second, the preoperative diagnosis of loosening is challenging. In our cohort, next to conventional X-rays to detect loosening, a bone scan was performed. This type of imaging may not be a routine practice in all hospitals and may result in false-positive scans when performed within 2 years after implantation for hips and within 5 years for knees. However, because the negative predictive value is high, we considered it as a valuable diagnostic tool to rule out infection [30]. Finally, we only studied CRP and ESR as serum inflammatory markers. There are other sensitive serum markers available, such as D-dimer and interleukin 6 (IL-6) that we have not incorporated in our daily clinical care [31, 32]. Although these markers do not exhibit a diagnostic accuracy of 100% either, applying a combination of serum inflammatory markers may be beneficial [33, 34].

In conclusion, in case of prosthetic loosening, all joints need to be aspirated before revision surgery as no specific characteristic could be identified to 100% rule out an infection. A prosthetic implant that needs to be revised for a mechanical reason without any signs of loosening probably does not need a synovial fluid aspiration, in particular, not when combined with low serum inflammatory markers.