Invited reply to the letter to the editor by McNally et al., 2021

Abstract

Thank you for the opportunity to respond to the comments on our paper Morgenstern et al. 2020 [1]. The detection of microbial heat production (microcalorimetry) has been studied for several biological and medical applications and has shown superior sensitivity and/or shorter time to microbial detection compared to conventional culture [2–4]. In our study we aimed to evaluate the performance of microcalorimetric analysis of synovial fluid in comparison to synovial fluid culture for the preoperative diagnosis of periprosthetic joint infection (PJI). In their Letter to the Editor (McNally et al. 2021), the authors raised their concern over the use of unvalidated definition criteria of PJI, which could potentially lead to overdiagnosing of infection. It is correct that the definition criteria we used for this study are highly sensitive and could be prone to over-diagnosing rather than underdiagnosing infection. Still, this classification system is presently being used in the clinical routine in many institutions (including ours) and has also been employed in several other studies [5–16]. In the last years, various PJI diagnostic criteria have been proposed and used by different authors. In the time our prospective study was conducted (2014 to 2015), the International Consensus Meeting (ICM) criteria [17], the Musculoskeletal Infection Society (MSIS) criteria [18] and the Infectious Disease Society of America (IDSA) criteria [19] were the most popular criteria available for PJI definition. However, we believed that these criteria were not appropriate for our study as they present crucial drawbacks. For instance, IDSA criteria do not consider an important criterion for PJI, namely the synovial fluid leukocyte count, which is inappropriate for a study that focuses on the preoperative investigation of synovial fluid. The 2013 ICM criteria were mainly based on expert opinion and studies with a low quality of evidence [20]. Furthermore, sonication of removed implants, which represents the most sensitive intraoperative method to detect low-virulent pathogens [21], was not considered neither in ICM nor in IDSA criteria, which consequently made them miss a considerable number of PJI [9, 10]. In fact, in a cohort of 136 patients with a painful prosthetic joint, Huard et al. [10] reported 41 PJI applying MSIS criteria, 50 with IDSA and 68 with proposed European Bone and Joint Infection Society (EBJIS) (our institutional) criteria, confirming the results published by Renz et al. [9]. ICM criteria predominantly missed those infections caused by low-virulent pathogens [9]. Another study showed that presence of a single minor criterion (but not fulfilling the 2013 ICM criteria for PJI) was associated with worse outcome of total joint revision [22]. These observations resulted in the modification of the ICM PJI definition criteria in 2018 [23]. In regard to the other comments, we are not ignoring the value of preoperative synovial biopsies as claimed by McNally et al. Still, we believe that this invasive procedure (requiring arthroscopy or open biopsy) should be performed only exceptionally, i.e. in case of a dry tap or inconclusive arthrocentesis results, as it poses intervention-related risk of additional infection. In addition, arthroscopically collected samples are probably not most representative for diagnosis of low-grade infection, as the biofilm location at the prosthesis-bone interface cannot be easily reached. The additional risk of performing a biopsy needs to be weighed against the moderate gain of additional information compared to only performing arthrocentesis. Some institutions still perform open biopsy as part of diagnostic routine, which, in our opinion, should be done only in exceptional cases with a good rational indication.