Epidemiology of Fungal Periprosthetic Joint Infection: A Systematic Review of the Literature

Fungal prosthetic joint infection (fPJI) is a rare complication; nonetheless, it represents a significant diagnostic and therapeutic challenge. There are no official guidelines on the most effective approach to identify and treat fPJIs. This systematic review aims to review the current literature on fPJI management and provide a comprehensive overview of this topic, especially from an epidemiologic point of view. Studies eligible for this systematic review were identified through an electronic systematic search of PubMed, Scopus, and Web of Science until 30 September 2022. Further references were obtained by cross-referencing. Sixty-three studies met the inclusion criteria, reporting on 372 cases of fPJI; such cases were described mostly in case reports and small case series with only a few larger cohort studies. Diagnosis of fPJI is challenging because of its chronic and indolent clinical course; it is further complicated by the technical difficulty of harvesting fungal cultures. A two-stage revision was the primary procedure in 239 (64.2%) patients whereas DAIR and one-stage approaches were reported in 30 (8.0%) and 18 (4.8 %) cases. In conclusion, our study highlights the heterogeneity of the reported treatments of fPJI, particularly in terms of medical management. With concern to a surgical approach, a two-stage revision arthroplasty is generally suggested, considering fPJI a delayed or late infection. The need for multicenter, prospective studies to provide standardized protocols and improve the treatment of fungal PJI clearly emerges.


Introduction
Prosthetic joint infection (PJI) is one of the most common complications following arthroplasty, occurring in as many as 0.7% of total hip arthroplasties (THA) and 2% of total knee arthroplasties [1,2]. PJIs are mostly caused by bacteria, especially the Staphylococcus species, while fungal PJIs (fPJIs) are much rarer, accounting for about 1% of total PJIs [3,4]. Although several species of fungi may cause PJI, the most common is the Candida species [3,5]. In contrast to bacterial PJIs, the clinical presentation of fPJIs tends to be more subtle; acute onset of symptoms is unlikely. Thus, diagnosis is often late, [6] resulting in delayed initiation of appropriate therapy.
Fungi are notoriously difficult to isolate in culture due to the need for specialized culture mediums and longer incubation periods [7]. In some culture-negative cases, uncultivable organisms must be taken into account and alternative identification techniques must be considered [6]. Furthermore, the isolation of a fungus does not exclude the presence of bacteria: concomitant bacterial infection is shown to occur in 15% and 20% of fPJI cases [3]. Bacteria and fungi are in fact thought to act synergistically within the prosthetic biofilm to produce more virulent infections [8,9].
Among the fungal infections included, 161 were total hip arthroplasties (THA), 207 total knee arthroplasties (TKA), one total elbow arthroplasty, and three total shoulder arthroplasties. Three patients had a bilateral TKA infection [5,76]. The mean age across all studies was 64.8 ± 14.2 years. The mean follow-up period was 27 months, ranging between 3 and 136 months. However, not all the included studies reported on the duration of follow-up.
The use of many different antifungal agents has been reported. The choice was dependent on case-specific pathogens and medical comorbidities. The duration of intravenous therapy was also variable between studies with an average of 4 months, even if not reported in all studies. Intravenous regimens were typically followed by prolonged courses of oral antifungals. Fluconazole was the most used, both as a monotherapy and in combination. Intravenous amphotericin B was also used, often in combination with either fluconazole or flucytosine. The use of echinocandins (micafungin, anidulafungin, and caspofungin) was also described.
A two-stage exchange was the primary surgical procedure in 239 patients. A one-stage approach and DAIR were reported in 30 (8.0%) and 18 (4.8%) cases, respectively, mainly as case reports. In a small case series reporting on 10 cases treated with a single-stage approach, Klatte et al. [71] observed only one case with a recurrence of infection out of 10 cases. On the other hand, Ji et al. [80] observed a 36% fPJI recurrence rate. All other data on DAIR and one-stage can only be deduced by summarizing sparse and heterogeneous data reported in the literature. Success rates of the two-stage approach also vary widely. Azzam et al. [3] reported a success rate of 47.4% among 31 cases. Phelan et al. [68] described an eradication rate of 80%. In addition, Anagnostakos et al. [15] observed no recurrence among seven cases of fPJI treated by the two-stage exchange protocol, and Ueng et al. [11] reported 16 patients with a success rate of 50%.
A three-stage approach with planned spacer exchange was reported by Baecker et al. [72] in 18 cases with a 72% rate of reimplantation.
The diagnosis of fPJI is challenging because of its chronic and indolent clinical course. Upon physical examination, patients frequently only manifest mild symptoms (pain, swelling, and/or reduced range of motion in the absence of acute local inflammatory signs) that are similar to those of low-grade PJI [84]. A sinus tract represents a major diagnostic criterium according to the 2018 Musculoskeletal Infection Society (MSIS) criteria [85]; however, it was only reported in 0% [3] to 16% of cases [74]. Fungal PJI should therefore be suspected in patients with one or more specific risk factors [8,63], such as immunosuppression, obesity, diabetes, many previous revision surgeries as well as long-term antibiotic treatments [3,11,17,18,30,63,74,75,[77][78][79]. Moreover, chronic medical conditions such as prolonged steroid use, renal disease, malignancy, rheumatoid arthritis, chemotherapy, and hepatic diseases, which can contribute to poor host immunity, can increase the risk of fungal infections [65]. Riaz et al. [75] reported that antimicrobial therapy within three months before the diagnosis of PJI and the presence of wound drainage lasting longer than five days prior to the diagnosis of PJI is significantly associated with increased odds of fPJI when compared with bacterial PJI. Nonetheless, a recent study [30] reported that 32% of fPJI patients have no risk factors.
Systemic inflammatory markers, including erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), and D-dimer, are the first screening tool for PJI, but they are generally minimally elevated or even normal [75,86,87]. In the presence of high clinical suspicion, synovial fluid should be sampled and sent for testing and culture [88].
Additionally, an underestimation of the prevalence of fPJI can be due to the difficulty in culturing fungi. To optimize the diagnostic process, the use of selective fungal media and an adequate incubation time of 5 to 14 days is recommended [89]. In some culture-negative cases, uncultivable organisms must be taken into account and alternative identification techniques must be considered, such as polymerase chain reaction (PCR) and next-generation sequencing (NGS) [6]. Despite no uniform diagnostic standards existing for fPJI, according to the European Bone and Joint Infection Society's (EBJIS) definition of PJI [90], the isolation of an uncommon or highly virulent organism on only one sample should be considered a probable infection.
The significant variability of pathogens associated with fPJI further complicates their management. The most identified fungus is Candida albicans (in approximately 49% of cases) followed by Candida parapsilosis (27%) [75,91]. Candida albicans forms complex biofilms compared to other Candida spp. [91,92], thus contributing to its pathogenicity. Only 14 cases of fPJI caused by Aspergillus spp. have been reported in the literature [93]. Moreover, a high rate of mixed bacterial and fPJI have been reported (ranging between 16% [74] and 33% [17] of fPJIs), which increases difficulties in their diagnosis and treatment. Mixed infections are associated with significantly worse outcomes. Sidhu et al. [8] described a revision-free survival rate of 38% in mixed infections. Bacteria and fungi establish a mutually beneficial mixed biofilm that protects them from antimicrobial treatment [73]. For example, Staphylococci can affect the activity of antifungal drugs while staphylococcal proteinase can enhance the adhesion capability of C. albicans [94], thus favoring the survival of the yeast. On the other hand, the presence of Candida species may increase the growth of anaerobic bacteria by generating a hypoxic microenvironment [95], stimulate biofilm development in Enterococcus faecalis strains (which are normally unable to form biofilm on their own [96]), and lead to an enhanced tolerance of Staphylococcus aureus towards vancomycin [97].
Implant removal is of paramount importance for the successful treatment of fPJI [98]. A total of 18 cases treated with DAIR have been reported, often resulting in persistent infection requiring multiple revisions [5,12,32,33,46,61,74,80]. However, most of these data can be only deduced by extrapolating sparse data reported in the literature. Moreover, most of these cases are reported together with other treatments in heterogeneous series.
A one-stage approach was reported in 30 cases with discordant results. In a small case series reporting on 10 cases treated with a single-stage approach, Klatte et al. [71] observed only one case with a recurrence of infection out of 10 cases. On the other hand, Ji et al. [80] observed a 36% fPJI recurrence rate. One-stage exchange might be considered in unfit patients who are less likely to tolerate multiple procedures; however, similarly to bacteria PJI, the pathogen involved must be well studied, and the response to antifungal treatment must be predictable [71]. In the case of rare or resistant pathogens, a single-stage exchange is often ineffective. However, the interpretation of treatment success is difficult because of the small number and heterogeneity of the described cases [52,71,80].
Two-stage exchange is the most commonly used surgical approach for fPJI with rates of infection eradication up to above 90% [99]. However, success rates vary widely. Kuiper et al. [17] reviewed 164 cases of fPJI and reported a success rate for the two-stage exchange protocol of 84.8%. In the series by Azzam et al. [3], the success rate was 47.4%. Phelan et al. [68] reported a success rate of 80%. Anagnostakos et al. [15] described seven cases of fPJI treated by the two-stage exchange protocol without any recurrence. A therapy protocol with three-stage revision has also been described [72]. Repeated surgical debridement combined with a scheduled spacer exchange could facilitate continuous delivery of the highest local drug concentrations with optimized release kinetics [72,100].
The use of cement spacers loaded with antifungal agents has been suggested to be effective in eradicating local infections and reducing the duration of antifungal treatment. However, no definitive evidence exists on which type and dosage of drug need to be incorporated to achieve the best results [101]. Amphotericin is the most used, due to its relative heat stability in polymethylmethacrylate (PMMA) cement. However, Cunningham et al. [102] reported a decrease in compressive strength of the spacer over time with the addition of amphotericin. Voriconazole has shown some promise and is becoming increasingly used in cement spacers: it has been shown to have predictable elution in vitro and determines only a mild reduction in compressive strength [103].
In addition, there is no agreement on the choice of optimal medical treatment. The choice of antifungal agent should be driven by local patterns of resistance and patient factors [3,30]. Fluconazole and amphotericin B are agents characterized by good bone penetration and are effective against most fungi [17,78]. Unfortunately, they have several side effects, especially in patients with renal and hepatic impairment. Echinocandins, such as micafungin, caspofungin, and anidulafungin, have a more tolerable side effect profile [101]. There is also no single consensus regarding the duration of antifungal administration. In consideration of the indolent nature of fPJIs, the ideal interval between implant removal and reimplantation are unknown. For patients who are not candidates for surgery, lifelong suppressive therapy with oral antifungals, such as fluconazole, is mandatory [3,104]. For Candida infections, the European Society for Clinical Microbiology and Infectious Disease recommends implant removal with at least 14 days of parenteral antifungals followed by a subsequent minimum of 4 to 6 weeks of oral agents [105,106]. In the case of two-stage exchange, the International Consensus Meeting (ICM) recommends a minimum of 6 weeks treatment after prosthesis removal [107]. The Infectious Diseases Society of America (IDSA) guidelines recommend between 6 and 12 months [10]. A meta-analysis by Ueng et al. [11]. identified improved eradication of infection with prolonged systemic therapy from three to six months. However, Anagnostakos et al. [15] suggested that 6 weeks of administration of an antifungal agent can be sufficient. Reduced susceptibility of fungal organisms to antifungal agents has become an area of interest, with resistance being described in fPJI [52,108], particularly azole resistance in Candida PJI [67,109].
There are several limitations to this study. Many of the included studies are case reports or small case series. There is a real lack of long-term data on fungal PJI, with many series not reporting the outcome [17,110,111]. Additionally, many series used different outcome measurements. Moreover, the heterogeneity of reported pathogens does not allow for the determination of the correlation between pathogen type and treatment outcomes.
In conclusion, fungal PJI is a rare complication compared to bacteria PJI. The present literature review highlights the extreme heterogeneity in the reported treatments of fungal prosthetic joint infections. In particular, the most conspicuous discrepancies emerge in medical management both in terms of drug of choice and length of treatments. A two-stage revision arthroplasty is generally suggested for fPJI, considering it usually presents as delayed or late infection. Nonetheless, success rates were less favorable than in bacterial PJI, and patients affected by fPJI treated with a two-stage exchange protocol were more likely to relapse compared with patients with bacterial PJI that underwent the same treatment. The need for multicenter, prospective studies to provide standardized protocols and to improve the treatment of fungal PJI clearly emerges.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.

Conflicts of Interest:
The authors declare no conflict of interest.