Non-Candida Fungal Prosthetic Joint Infections

Background: Fungal prosthetic joint infections (PJIs) are rare, especially those caused by non-Candida species. Treatment has not been fully elucidated, since a plethora of antifungal and surgical interventions have been proposed. Τhis study represents an effort to clarify the optimal management of non-Candida fungal PJIs, by reviewing all relevant published cases. Methods: A thorough review of all existing non-Candida fungal PJIs in the literature was conducted. Data regarding demographics, responsible organisms, antifungal treatment (AFT), surgical intervention, time between initial arthroplasty and onset of symptoms, and time between onset of symptoms and firm diagnosis, as well as the infection’s outcome, were evaluated. Results: Forty-two PJIs, in patients with mean age of 66.2 years, were found and reviewed. Aspergillus spp. were isolated in most cases (10; 23.8%), followed by Coccidioides spp. (7; 16.7%) and Pichia anomala (5; 11.9%). Fluconazole was the preferred antifungal regimen (20 cases; 47.6%), followed by amphotericin B (18 cases; 42.9%), while the mean AFT duration was 9.4 months (SD = 7.06). Two-stage revision arthroplasty (TSRA) was performed in 22 cases (52.4%), with the mean time between stages being 5.2 months (SD = 2.9). The mean time between initial joint implantation and onset of symptoms was 42.1 months (SD = 50.7), while the mean time between onset of symptoms and diagnosis was 5.8 months (SD = 14.3). Conclusions: Non-Candida fungal PJIs pose a clinical challenge, demanding a multidisciplinary approach. The present review has shown that combination of TSRA separated by a 3–6-month interval and prolonged AFT has been the standard of care in the studied cases.


Introduction
Joint arthroplasty represents a life-enhancing procedure, providing pain relief and restoration of function, and thereby improving patients' quality of life. Hip and knee arthroplasty rates are projected to reach 572,000 and 3.48 million, respectively, in the USA by 2030 [1]. Joint reconstruction surgery has evolved over time, encompassing minimally invasive surgical approaches, perioperative pain management and blood transfusion reduction protocols, and navigation or robotic systems, as well as new prosthetic materials [2][3][4]. Nevertheless, complications have not yet been eliminated. Prosthetic joint infections (PJIs) have serious implications on the patient's quality of life, and in some cases may prove fatal [5].
Fungal organisms are responsible for~1-2% of PJIs-even less in cases where non-Candida species are the cause [6]. Due to the rarity of these infections, no clear guidelines exist regarding management [7]. Currently, on the basis of limited data, a two-stage revision arthroplasty (TSRA) combined with prolonged antifungal treatment (AFT) is suggested [7,8].
The present study represents an effort, by reviewing all published cases of non-Candida fungal PJIs, to clarify the medical and surgical treatment options and their success-namely, the eradication of the infection, as well as the maintenance of the viability and functionality of the prosthetic joint, offering the patient the best possible quality of life. The present review takes into account the fact that it covers vast and epidemiologically diverse geographical areas, and that over the course of the long time period studied, medical therapeutic management has changed dramatically.

Materials and Methods
A meticulous electronic search of the PubMed and MEDLINE databases was conducted to identify all existing articles related to the treatment of non-Candida PJIs through to March 2021. Alone and/or in combination, the terms "fungal joint infection", "fungal prosthetic joint infection", "fungal knee arthroplasty infection", "fungal knee infection", "fungal shoulder arthroplasty infection", "fungal shoulder infection", "fungal elbow arthroplasty infection", "fungal elbow infection", "fungal hip arthroplasty infection", and "fungal hip infection" were searched. In addition, terms including each fungal species (e.g., "Aspergillus joint infection", "Coccidioides joint infection", etc.) were also searched.
The review was limited to papers published in English and in peer-reviewed journals. The data extracted from these studies included age, gender, affected joint, responsible non-Candida fungal organisms, duration and type of AFT, type of surgical intervention, use of antifungal-agent-loaded cement, time between initial arthroplasty and onset of symptoms, and definitive diagnosis (microscopy/culture/histopathology). Furthermore, the results of medical and surgical treatment, along with the follow-up of each case, were studied and evaluated.
In order to reach a conclusion of the success rates of each surgical treatment, all interventions that failed in each report were thoroughly evaluated. Cases not including surgical interventions were excluded from the assessment of the surgical success rate.
Treatment was considered successful if all signs and symptoms of the infection disappeared and no recurrence was observed during the follow-up period.
Finally, the patients' Charlson Comorbidity Index was calculated based on the information provided by each study.
Data were recorded and analyzed using Microsoft Excel 2019 (Microsoft Corporation, Redmond, WA, USA).
Patients' mean Charlson Comorbidity Index was 3.2 (SD = 1.6). Furthermore, 26 cases (62%) referred to primary joint reconstruction, while the remaining 16 (38%) were revisions, with the mean number of revisions being 1.5 (SD = 0.7). A total of 10 patients (23.8%) were immunocompromised according to the available information from each report. The mean C-reactive protein (CRP) and erythrocyte sedimentation rate (ESR) at initial presentation were 58.8 mg/L (SD = 68.9) and 57 mm/h (SD = 32.4), respectively. The mean time interval between the initial joint implantation surgery and the symptomatology of the onset of the infection was 42.1 months (SD = 50.7), while the mean time interval between onset of symptoms and firm diagnosis was 5.8 months (SD = 14.3). The mean follow-up was 44.2 months (SD = 24.5).
The total success rate of any surgical intervention was 61%. TSRA showed a 66.7% success rate; ORSA, arthrodesis, and debridement had success rates of 66.7% each; and RA had a 50% success rate. In the amputation case, the infection was eradicated. All eight cases (100%) that did not receive surgical intervention were considered, by the authors, as successfully treated only by medical means. It is of note that all of these eight patients were put on lifelong antifungal suppressive treatment ( Table 2).
More specifically regarding TSRA, the mean time interval between the initial removal of the implants and the final re-implantation was 5.2 months (SD = 2.9). Regarding the cement spacer used, in nine cases an antifungal regimen was impregnated in the cement; amphotericin B was used in five of these cases, voriconazole in three, and itraconazole in one.

Case No
Year

Discussion
Fungal PJIs are uncommon and extremely challenging regarding their management [5,6]. Most such infections are caused by Candida species [5,35]. Hence, limited data and information exist regarding PJIs caused by non-Candida organisms [8]. The present study aims, by reviewing published data, to clarify the characteristics, treatment options, and outcomes of non-Candida fungal PJIs.
PJIs, along with other invasive fungal infections, represent a major cause of morbidity and mortality in current medical practice. Optimal treatment of fungal PJIs remains unclear, since no certain guidelines exist regarding the antifungal regimen and the indicated surgical intervention [7,8]. TSRA and long-term AFT are proposed due to lack of data [5,7,8]. Information about the kind of AFT, its duration, and its success rate, as well as type of surgical intervention, the use of antifungal agents in cement, and the time intervals between the two stages of TSRA, are of utmost importance for the clarification of the best medical treatment and the improvement of the surgical management of these cases.
The present study has reviewed 42 cases of non-Candida fungal PJIs over 37 years, with a mean follow-up of 44.2 months. The incidence of fungal PJIs incidence is expected to rise, due to the increasing number of prosthetic joint reconstruction surgeries worldwide [1,8].
Immunosuppression and systemic disease have been widely acknowledged as risk factors for invasive fungal infections, while revision joint surgeries increase the risk of infection [5]. In the present study, a total of 10 patients (23.8%) were immunocompromised, while in 38% of the cases a revision reconstruction surgery had been performed.
Fungal PJI is frequently of hematogenous origin [8,14,16]. However, intraoperative contamination by fungal skin pathogens may also occur [36]. If the contamination originates in the skin, symptoms usually appear early in the postoperative period. In the studied cases, the mean time interval between initial joint reconstruction surgery and onset of symptoms was 42.1 months (ranging from 0.7 to 120), supporting the theory of hematogenous spread. The mean time from onset of symptoms to firm diagnosis was 5.8 months. Fungal PJIs usually present with indolent symptoms and, therefore, diagnosis may be delayed.
In the present review, the predominant fungi involved were Aspergillus spp., identified in 10 cases (23.8%), followed by Coccidioides spp. in 7 cases (16.7%) and Pichia anomala in 5 (11.9%). In 12 cases (28.6%), co-infection with bacteria was diagnosed. Concomitant bacterial infection has already been reported in 15-20% of fungal PJI cases [37]. Coccidioides spp. are geographically restricted dimorphic fungi represented rather heavily in the present review. However, it must be noted that most of the published cases originate in the United States, where these fungi are endemic.
Invasive Aspergillus infections are typically seen in patients with significant underlying immunosuppression [13,14,17]. However, in the present review, most of the reported Aspergillus PJIs occurred in immunocompetent individuals.
Several surgical intervention options have been described for the treatment of fungal PJIs [7,8]. In the reviewed population, TSRA was preferred in most cases (52.4%), with a success rate of 66.7%. The mean time between the two surgical stages was 5.2 months, and in nine cases cement impregnated with an antifungal agent was used. ORSA, arthrodesis, and debridement each exhibited a 66.7% success rate, while for RA it was 50%. Eradication of the infection was the result in one case of amputation.
It is of note that arthrodesis, RA, and amputation, although proven to be successful treatment options-since the infection was finally eradicated-are associated with a high negative impact on the patient's quality of life, and may possibly lead to loss of independence. Furthermore, in the eight cases that did not receive any surgical intervention (cases 15, 17, 19, 21, 22, 24, 27, and 40, shown in Tables 1 and 2), although the treatment's outcome was considered by authors to be successful, all patients subsequently commenced lifelong fungal suppression; therefore, it is understood that definitive eradication of the infection was not achieved. These reports raise the question of whether lifelong fungal suppression by AFT is a feasible option in cases where further surgery may lead to failure, either due to technical difficulties (e.g., extensive bone loss, challenging reconstructive options) or to patient comorbidities.
Guidelines for the treatment of fungal osteoarticular infections exist; however, no clear recommendations are available for the treatment of such PJIs [7]. Therefore, the duration of treatment is mainly based on the clinical and laboratory findings of each case and the physicians' experience with such infections. It is, therefore, of utmost importance to carry out susceptibility testing to obtain accurate MIC values following the isolation of the fungus, taking into account that different species of fungi (e.g., yeast, molds, etc.) are characterized by intrinsic resistance to certain antifungal compounds [8]. Additionally, it must be noted that for a number of molds, laboratory methods indicating MICs are not standardized and unanimously accepted, while the immune status of the patient plays a major role [38].
In all of the reviewed cases, the causative fungus was identified. However, the microbiological procedures used were not described in most of them. Moreover, since the present review spans almost four decades, it is understandable that identification techniques have evolved over time, ranging from simple microscopy, pathology, and culture to modern molecular methods [38]. Hence, although the information about the fungal cause of cases is sufficient, information about specific identification procedures has been inadequate.
Regarding AFT, fluconazole was the preferred antifungal agent (47.6%), followed by amphotericin B (42.9%). Fluconazole was extensively used in the reviewed cases, although this agent is ineffective against molds. However, it must be taken into account that fluconazole and amphotericin B deoxycholate were the only available agents in the early years of the reviewed cases. Fluconazole has on rare occasions been associated with severe hepatotoxicity [39]. Hence, liver function tests should be performed regularly during prolonged fluconazole therapy, while amphotericin B, although an effective broad spectrum regimen, is relatively toxic, and its side effects-including renal dysfunction-may restrict its long-term use, which is essential for PJI cases [40]. The liposomal compounds of amphotericin B have reduced the drug's nephrotoxicity considerably, but long-term use of these agents may be still problematic [40]. Voriconazole, which was introduced in 2003, has proven to be the drug of choice against Aspergillus spp. This agent has changed the management of Aspergillus infections dramatically over the past several decades; with all the characteristics of azole compounds, it is moderately hepatotoxic and much less nephrotoxic than all amphotericin compounds [41].
The mean AFT duration was found to be 9.4 months. The final outcome was successful in 33 cases (78.6%). It is of note, however, that the success rate drops to 66.7% in cases of bacterial co-infection. Concomitant bacterial infections occur in between 15 and 20% of fungal cases, while the poor prognosis for co-infective PJIs has already been underlined [42].
The present review has some limitations. There is heterogeneity between different joint reconstruction surgeries (hip, knee, and elbow), while not all information was available from each case, such as specific signs and symptoms of the infection, antifungal dosages, mode of administration, monitoring of serum levels, MICs, and adverse events. Another point not clarified is the specific microbiological techniques used for fungal identification (especially modern ones, such as PCR and/or β-D-glucan). Nevertheless, this study reviews all of the non-Candida fungal PJIs in a systematic way, offering valuable insights regarding epidemiology, severity, surgical management, and medical treatments that changed dramatically over the long study period, as well as outcomes.
It is important that over the past several decades a number of new antifungal agents (e.g., voriconazole, posaconazole, echinocandins, and isavuconazole) were introduced, providing more medical treatment options, hopefully associated with better results [41].
The present review shows that non-Candida fungal PJIs represent a very challenging clinical entity. A combination of the proper medical AFT, based on susceptibility testing (when feasible) and surgical intervention, seems to represent the current standard management. There have been reports of successful treatment of such cases with OSRA and debridement. However, TSRA should be strongly recommended. The combination of TSRA separated by 3-6 months and a prolonged period of AFT is suggested on the basis of limited data. Lifelong fungal suppression with a proper agent (based on the type of the fungal species) is suggested in cases where surgery is not desirable (either due to patient comorbidities or technical difficulties predicting uncertain results). More data and research are needed, focusing on proper treatment-since the results of therapeutic procedures and policies, such as the antifungal-loaded cement spacers and AFT duration, remain unclear-in order to conclude the optimal management approach.
Funding: This research received no external funding.

Institutional Review Board Statement:
Not applicable, the present study is a literature review.

Informed Consent Statement:
Not applicable, the present study is a literature review.
Data Availability Statement: Not applicable.