Diagnosing periprosthetic hip joint low-grade infection via arthroscopic neo synovium biopsies

1. Background

Periprosthetic joint infection (PJI) is one of the postoperative complications associated with total hip arthroplasty [1]. Particularly in cases of an assumed low-grade infection, it is crucial to determine its presence with high accuracy, as treatments like a two-stage exchange total hip arthroplasty often results in future limitations in function. A low-grade PJI is characterized by a subacute or chronic process with frequently absent clinical signs of an infection and normal conventional diagnostic parameters [2]. There is presently a lack of consensus on the recommended diagnosis and treatment of PJI. Proposed parameters to verify a PJI include diagnostic joint aspiration, serum inflammatory markers, intraoperative cultures, and pathological specimens [3, 4]. However, the relevance of serum inflammatory parameters and diagnostic joint aspiration has been challenged recently [1, 4, 5, 6, 7, 8, 9].

Over the last few years, the analysis of synovial fluid has gained increasing importance. Determination of the leucocyte count within synovial fluid can help to differentiate between low-grade infection and arthritic changes, even if the presence of bacteria has not been verified. However, diagnostic joint aspiration of the hip is often not as effective or successful as for the knee joint [12]. It is often impossible to gain a large enough sample to count the number of leucocytes within the synovial fluid. Sometimes no synovial fluid can be obtained at all [1, 13].

A histopathological classification of the etiology of implant failure was published by Krenn and Morawietz et al. and has been revised over the last few years [14, 15, 16, 17, 18]. As the joint cavity communicates with the periprosthetic area, the histological pattern of the neosynovium mirrors that of the periprosthetic membrane. The diagnosis of a PJI is based on reactive vascular proliferation, chronic edema, granulation tissue of fibroblasts, and an infiltration of neutrophils, plasma cells, and lymphocytes [16, 17].

2. Objective

Despite an extensive literature review, very few publications were found to report the diagnostic value of neosynovium biopsies obtained via a diagnostic arthroscopy of the hip. As we performed a similar study on the value of arthroscopic biopsies of the knee joint to predict PJI [19], the aim of this study was to evaluate the reliability of neosynovium biopsies to predict or rule out the existence of PJI after THA. In addition, we sought to determine whether the value of these biopsies in predicting low-grade infection was any different, compared to its use in the knee.

3. Methods

The institutional Ethical Committee approved the present study. During a retrospective study, 20 cases that had undergone a diagnostic hip arthroscopy with sampling of neosynovium biopsies between January 2012 and December 2016 were analyzed. The individual data and history of each patient are presented in Table 1. The mean age of the study group at the time of arthroscopy was 65.5 ± 12.2 years (mean ± standard deviation). A supposed PJI after THA or a differential diagnosis that included PJI in cases of early loosening of the prosthesis were the indications for diagnostic arthroplasty, during which neosynovial samples were obtained.

We analyzed the data regarding previous surgical interventions and serum parameters, such as the white blood cell count (WBC) and C-reactive protein (CRP) levels, prior to arthroscopy. The cut-off value for CRP was 0.5 mg/dL; and for WBC, 10.0 k/ μ L and 9.1 k/ μ L for female and male patients, respectively. In 10 cases, a revision arthroplasty was performed following the arthroscopy. During arthroscopy and revision arthroplasty, neosynovium biopsies were collected from five identical and well-defined locations close to the implant, for microbiological and histological analysis. Three samples were taken around the cup and two samples were taken around the stem. An infection was defined by the presence of an infectious specimen in at least two of five biopsies, or the histological proof of a type II or III membrane according to the classification of Krenn and Morawietz et al. [14, 15, 17]. According to our Institute for Medical Microbiology, the presence of an infecting specimen in only one of five samples is considered contamination. This procedure had also been applied to patients who had undergone total knee arthroplasty, as previously published [19].

Statistical analysis was conducted using the Graph Prism 5 package (GraphPad Software, Inc., La Jolla, CA, USA). Diagnostic parameters were described as true positive, true negative, false positive, and false negative in the determination of sensitivity, specificity, and positive and negative predictive values (PPV and NPV, respectively). The accuracy was determined by the percentage of the sum of true positive and true negative results among all results. Values were expressed as means with 95% confidence intervals. In keeping with the standards of the Biometric Institute of our organization, diagnostic parameters were compared using chi-squared analysis. A p -value 0.05 was considered statistically significant. A post hoc power analysis was performed.

5. Discussion

The most important finding of the present study was the high reliability of arthroscopic neosynovium biopsies in predicting a PJI. In suspected cases of a low-grade PJI, a proper diagnosis is essential to distinguish between low-grade PJI and other reasons for the persistence of pain. Furthermore, in cases of early loosening, a low-grade PJI has to be ruled out, as confirmation of its presence affects subsequent procedures. After verification of a PJI, the majority of the literature recommends a two-stage revision arthroplasty rather than a one-stage revision arthroplasty, because of the increased risk of a persistent PJI [1, 5, 20, 21, 22, 23, 24, 25] . Thus, the presumptive diagnosis has major consequences on the therapy that follows. The importance of preoperative diagnostic tools, like joint aspiration and the determination of serum inflammatory markers, has been challenged in the literature [8, 26]. Over the last few years, the analysis of synovial fluid with the determination of leucocyte counts represents a new approach and possible benefit [13]. However, for such analysis, more than 5 mL of synovial fluid needs to be obtained. A tool with high specificity and sensitivity is particularly needed in the detection of low-grade PJI. Thus, the aim of the current study was to evaluate the value of biopsies collected via diagnostic arthroscopy to confirm the presence of a PJI after THA.

For several decades, hip arthroscopy has been a standard procedure, employed particularly for the resection of femoroacetabular impingement and intra-articular lesions. The rate of complications associated with hip arthroplasty varies below 2% [1]. In comparison to diagnostic alternatives like joint

aspiration and assessment of serum inflammatory markers, the most common complications include hemarthrosis, infection, thromboembolic diseases, and anesthetic complications [27]. Although no complications occurred in the current study and no revisions were necessary, diagnostic arthroscopy is more invasive than diagnostic joint aspiration and serum sampling to determine inflammatory markers.

The histopathologic results were graded according to the published classification of periprosthetic membranes outlined by Krenn and Morawietz and colleagues [14, 15]. A high correlation between histopathological and microbiological diagnoses ranging between 89.0% and 89.7% was detected. In addition, high interobserver reproducibility between 85% and 95% was observed [14, 15]. The incidence of type I periprosthetic membranes was 54.3%; for type II, 19.7%; for type III, 5.4%; and for type IV, 15.4% [14]. A high correlation has been observed between the interpretation of periprosthetic membrane biopsies and that of neosynovium biopsies [16, 17]. In the current study, the incidence of type IV neosynovium biopsies was evidently higher than that reported by Krenn et al. This can be due to a difference in the composition of the patient cohorts examined.

It was previously possible to attribute the lifetime of different prostheses to different types of periprosthetic membranes, such that the longest lifetime was associated with type I membranes and the shortest, with type II [16]. These facts highlight the clinical relevance of the classification of periprosthetic membranes. Based on the available data, we believe that arthroscopic neosynovium biopsies, categorized according to the classification of Krenn and Morawietz et al. [14, 15], present a helpful tool in determining low-grade infections for cases in which conventional diagnostic tools have failed.

The identification of PJI is one of the most difficult issues encountered in arthroplasty. However, PJI is one of the most frequent reasons for the revision of arthroplasty in the knee and hip [28]. Several diagnostic algorithms have been published in recent research [4]. The diagnostic tools include pre- and intraoperative examinations; however, only the preoperative diagnostics can provide the necessary information to select between one- and two-stage revision arthroplasty. The results of the intraoperative histopathological evaluation become available only about 14 days postoperatively. The existing algorithms are often based on preoperative joint aspiration and the determination of serum inflammatory markers, as the main diagnostic tools [4, 29, 30, 31, 32].

Some studies have questioned the relevance of serum inflammatory markers [6, 8, 26]. For CRP, a sensitivity of 0.17 and a specificity of 0.81 have been published for a cut-off value 1.0 mg/dL, and 0.48 and 0.61, respectively for a cut-off value 0.5 mg/dL, which are consistent with the values recorded in the present study [8, 26]. Alternative serum markers might enhance the accuracy of preoperative testing [32, 33, 34]. For example, Ettinger et al. [32] investigated whether circulating biomarkers can preoperatively help to distinguish between aseptic prosthesis loosening and low-grade joint infection, and which biomarker combinations are the most accurate. They determined that CRP, in addition to interleukin 6 (IL-6), seems to be the most effective combination for preoperative discrimination of aseptic loosening vs. low-grade joint infection [32]. In addition, Glehr et al. [33] investigated the sensitivity and specificity of the biomarkers procalcitonin, IL-6, and interferon α (IFN- α ), as compared to conventional biomarkers (CRP and leukocyte counts) for PJI associated with revision arthroplasties. They concluded that procalcitonin and IL-6 were useful parameters in the detection of PJIs in revision arthroplasties, although CRP was generally superior. Thus, procalcitonin and IL-6 could be considered adjuvant tests when the diagnosis of a PJI is uncertain. Unfortunately, the diagnosis of assumed low-grade infection is even more complicated. Conventional clinical and laboratory tests assume a particular physiological range [2, 35, 36]. A low bacterial load and a protective biofilm might yield a negative joint aspiration [6]. This emphasizes the need for additional diagnostic tools. Synovial fluid CRP, as a diagnostic marker for PJI, has been proposed as an enhanced diagnostic tool [37, 38]. Wang et al. [39] performed a relevant systematic review and meta-analysis. They concluded that synovial fluid CRP is a good biomarker for the diagnosis of PJI, with high sensitivity and specificity [39]. However, Tetreault et al. [40] did not agree with this finding, as they could not demonstrate any benefits of CRP in synovial fluid for the detection of PJIs.

A greater volume of tissue and a lower influence of the biofilm on the results underline the relevance of neosynovium biopsies and histological evaluation. Only a few studies have been published that present data on the potential role of diagnostic arthroscopy in obtaining biopsies of the periprosthetic membranes or neosynovium of the hip joint. When considering diagnostics for PJI, a high sensitivity and NPV are crucial, as they help to determine whether a one- or two-stage revision arthroplasty should follow [41].

The current study demonstrated the high diagnostic accuracy of neosynovium biopsies obtained via arthroscopy. This accuracy tended to be higher than that of conventional serum inflammatory markers like CRP. Furthermore, the sensitivity of neosynovium biopsies was higher than that of serum CRP. However, because of the relatively small number of cases, the differences observed were not statistically significant.

Pohlig et al. [42] determined the diagnostic accuracy of arthroscopic biopsy in periprosthetic infections of the hip. Twenty consecutive patients were included, all of whom underwent percutaneous aspiration of synovial fluid and arthroscopic biopsy because of an assumed PJI of the hip, and subsequent one- or two-stage revision surgery. The best overall diagnostic value was identified for arthroscopic biopsy and a combination of bacteriological and histological analysis, with a sensitivity of 87.5%, specificity of 100% and accuracy of 95%. Bacteriological assessment of the synovial aspirate yielded a sensitivity of 50.0%, specificity of 91.7%, and accuracy of 75%. Those researchers concluded that arthroscopic biopsy is a superior diagnostic tool in comparison to the erythrocyte sedimentation rate, CRP determination, joint aspiration, and their respective combinations [42]. A similar study was conducted by Fink et al. [43], who also concluded that the biopsy technique has greater value than joint aspiration and CRP in the diagnosis of PJI of the hip. They demonstrated that biopsy should be preferred over the repetition of joint aspiration in cases of a negative aspirate, with increased CRP or other clinical signs of infection [43]. Williams et al. [41] reported a sensitivity of 0.83, a specificity of 0.90, a PPV of 0.74, and an NPV of 0.94 for diagnostic arthroscopy in cases of an assumed PJI of the hip joint. However, the accuracy reported was lower than that of joint aspiration [41].

Comparing the aforementioned data of studies that have analyzed the value of arthroscopic biopsies of the hip to those of the knee researchers arrived at similar conclusions. Fuerst and colleagues [45] reported a high sensitivity and NPV of 1.0 in their analysis of 15 cases of diagnostic arthroscopy of the knee, whereas the specificity and PPV were reportedly 0.95 and 0.87, respectively. These values are slightly higher than those reported in our previous study that analyzed the value of arthroscopic biopsies in the diagnosis of periprosthetic knee joint infections [19].

The limitations of this study include its retrospective design. Because a diagnostic arthroscopy is only indicated in special cases, the number of patients analyzed was relatively low. A post hoc power analysis indicated an insufficient number of patients for most comparisons. However, the lack of statistically significant differences in the present study does not necessarily preclude their presence in a larger sample of patients. In addition, the heterogeneity of the study cohort, in terms of the differences in age, sex, and body mass index, should be considered in the interpretation of the results. Nevertheless, our results supplement the current literature and highlight the need for further investigations. Another limitation of the present study is the fact that arthroscopy cannot achieve the periprosthetic membrane itself. However, the neo-synovium is achieved and their changes correlate with changes of the periprosthetic membrane. Thus, the samples obtained were sufficient for histopathologic evaluation to determine low-grade PJI.

We believe that diagnostic arthroscopy is indicated in cases of associated persistent disorders after THA or revision THA. This is particularly applicable for patients with early loosening of the prosthesis within two years, who show no signs of mechanical failure, or those with a history of previous infections in the affected hip, for whom the analysis of joint aspiration is either negative or impossible.

In cases of assumed PJI, a safe tool is required to determine the operative steps. In this study, diagnostic hip arthroscopy with neosynovium biopsies showed high accuracy and seemed to be helpful in predicting a low-grade PJI after THA. However, non-invasive diagnostic tools, such as serum biochemical parameters and joint aspiration with analysis of the synovial fluid should also be performed. If these diagnostic values do not confirm or rule out a PJI, then arthroscopic biopsy represents an additional, helpful tool in the diagnosis of a PJI.