Infections are microbially induced local tissue or systemic inflammatory responses that cause complex diagnostic and therapeutic problems [71]. In contrast to bacterial PJI, the clinical symptoms of fPJI are usually chronic and less likely to present with acute manifestations [15]. A previous study reported that fungal infections are estimated to cause 1% of all PJI [72]. Candida is the most common pathogen, accounting for approximately 80% of fPJI [19]. Candida infections accounted for 76.2% of all cases in the results of this paper, which is generally consistent with previous statistics. Nine different Candida species were recorded, and the most common Candida types were Candida parapsilosis (22/63) and Candida albicans (9/63). Globally, the overall incidence of fPJI infections has remained stable over the years, but we found that different geographic locations had a significant impact on the incidence, with the prevalence of fPJI in the United States being almost eight times higher than in Europe (2.3% vs. 0.3%) [73]. With the current increasing number of immunosuppression and human life expectancy, the incidence of fungal PJI is also on the rise, ranging from 1–2% [54, 74, 75]. Therefore, accurate diagnosis and treatment of this catastrophic complication are crucial.
As mentioned above, fungal infections are associated with several risk factors or chronic diseases and can lead to a decrease in host immunity. fPJI develops mainly due to hematogenous spread and contamination during surgery. Blood-borne fungal infections, especially caused by Candida, are less associated with an immunocompromised state and more strongly associated with alterations in the commensal microbiota caused by antibiotic and steroid use or with epithelial barrier dysfunction secondary to surgical intervention [76]. However, one study reported that 32% of patients with fungal PJI had no risk factors [77]. Similarly, our study found no risk factors in 38% (24/63) of patients. Additionally, as the number of procedures or intra-articular punctures increases, the resulting chance of direct inoculation is expected to be higher; hence, worthy of greater attention.
In patients with fPJI after TKA, there are no specific clinical manifestations in the early stages of the disease, and concomitant systemic symptoms are even less common. Later in the course of the disease, there is often joint pain, swelling, exudation, sinus tract formation, reduced joint mobility, and in severe cases, joint stiffness. Kuiper et al. [19] found that common clinical symptoms of fPJI after arthroplasty were joint pain and swelling, and a small number of patients had increased joint skin temperature and even joint sinus tract formation. In our study, joint pain (47/63) and swelling (40/63) were likewise the most common symptoms. Because the clinical presentation of fPJI is similar to that of bacterial PJI, early diagnosis of fPJI is more difficult. Therefore, fungal infection should be suspected even in healthy patients who present with bacterial PJI typical symptoms and fail to respond to standard antimicrobial therapy.
In terms of diagnosis, imaging can only provide limited value in that we can observe prosthesis loosening, fracture, or dislocation, the presence or absence of intra-articular effusion, and changes in the extent of the effusion but cannot provide a definitive diagnosis [78]. ESR and CRP levels should be determined according to the diagnostic guidelines of the Infectious Diseases Society of America (IDSA) and the Musculoskeletal Infection Society [79, 80]. However, one study suggested that systemic inflammatory markers (ESR and CRP) may be only slightly elevated or even normal in fPJI [81]. Azzam et al. [82] reviewed routine laboratory indicators of inflammation in patients diagnosed with PJI from 1999 to 2006 and found no value for the diagnosis of fPJI. Anagnostakos et al. [43] also agree that CRP and ESR are not significant for the diagnosis of fPJI. Therefore, clinicians should be aware that the diagnosis may be delayed or missed if only such laboratory markers are used.
Although imaging and serologic testing are important tools for orthopedic surgeons, isolation of infectious microorganisms from tissue within the infected joint remains central to the diagnosis of fPJI [83, 84]. Currently, culture-based methods are the "standard" for identifying pathogenic microorganisms. This allows for targeted antimicrobial therapy and increases the chances of successful treatment [85]. The 2018 International Consensus Conference recommended that microbiological cultures of the same organism in two or more synovial fluid or tissue specimens should be used as the primary diagnostic criteria for PJI [8]. If an infection is suspected, immediate arthrocentesis aspiration or incision and drainage may be performed, and a specimen of joint fluid may be sent for culture. At least three samples of periprosthetic tissue (ideally five) should be obtained using different instruments to ensure that there is no contamination between the specimens. However, fungal cultures need to be incubated in specific media for 4 weeks or more in sterile conditions at 24–25°C [86]. Despite this, approximately 20–50% of patients have negative cultures, which may be due to previous antibiotic treatments, the pathogen's own factors, culture techniques, and laboratory contamination [87, 88].
The most common reason for a negative fungal culture is preoperative antibiotic use [87, 89]. Guidelines from the American College of Orthopaedic Surgeons [90] and the Infectious Diseases Society of America [80] recommend stopping antibiotic therapy at least 2 weeks before aspiration or intraoperative specimen collection. Another important pathogen's factor is the presence of biofilms, which can make pathogens viable but non-culturable (VBNC) [91, 92]. Specifically, the limited nutrient supply within the biofilm, low oxygen concentration, and high level of toxic metabolites can therefore cause the metabolic activity of the pathogen to significantly decrease and enter a dormant phase, and VBNC cells temporarily lose their ability to grow on conventional bacterial media [93, 94], resulting in a false negative culture. Negative cultures cause patients to miss the best choice of antimicrobial agents and therefore lead to higher failure and mortality rates.
Because fungal cultures have a relatively low positive rate in clinical practice, which makes clinical diagnosis and treatment difficult, the technique of using ultrasound to degrade the biofilm on the surface of artificial prostheses followed by routine fungal cultures has now been demonstrated in clinical work. In 14 patients with fPJI studied by Trampuz et al. [95] routine tests and cultures were negative, and after ultrasonic degradation of the artificial prosthesis surface biofilm, the cultures were positive again. Huang et al. [96] found that the ultrasonic degradation method of culture has the advantage of detecting multiple microbial or fungal infections. However, some limitations of ultrasonication were also found, such as the absence of a standardized procedure [97], and water contamination in an ultrasonicator may lead to false positive results [98].
Given the shortcomings of fungal culture techniques, alternative techniques capable of detecting fungi, such as molecular techniques, can be used as complementary methods. Polymerase chain reaction (PCR) technology has the advantages of fast detection, high sensitivity, and specificity. PCR is a method of designing universal primers according to the highly conserved region of fungal rDNA to amplify the target rDNA fragment and sequence the product. Rapid detection of pathogens is the key to effective treatment and good recovery. PCR technology compensates for the shortcomings of traditional methods. Shin et al. [99] used PCR to correctly identify all species of Candida in 73 blood culture bottles, including culture bottles containing Candida and bacteria and three kinds of mixed Candida culture bottles. Morace et al. [100] used PCR and culture methods to detect Candida albicans in the blood. In that study, the sensitivity of the two probes using PCR technology was 100%, and the specificity of each probe was 97% and 72%. The results showed that PCR technology had higher sensitivity and specificity. However, the detection of fungi by PCR also has its shortcomings; for example, sequencing requires specific primers, making it difficult to detect rare fungi [101, 102]. The detection of fungi by PCR technology can only determine whether there are fungi in the samples but cannot determine whether they are pathogenic fungi or live or dead bacteria. Additionally, it cannot detect the sensitivity and drug resistance to antifungal drugs. Due to the false positive results of PCR and the lack of standardized operation and criteria, the United States Food and Drug Administration has not approved the use of PCR in the clinical diagnosis of fungal infections [103]. Therefore, PCR detection should be combined with other methods, clinical symptoms, and imaging, which is conducive to the accurate treatment of fPJI patients.
The difficulty of fPJI treatment can be attributed to the rarity of fungal-induced periprosthetic infections. Systematic application of antifungal agents is essential to control or eradicate fungal infections. Regarding antifungal treatment, 24 patients (38.1%) were treated with a single antifungal regimen, 23 patients (36.5%) were treated with two antifungal agents simultaneously or consecutively, and 16 patients (25.4%) were treated with more than two antifungal agents. Amphotericin B and fluconazole were widely used in the articles included in this study. The importance of selecting drug therapy based on antifungal susceptibility test results and patient factors has been highlighted in the literature [77, 79, 104]. In the absence of antifungal susceptibility test results, fluconazole and amphotericin B are the drugs of choice when administered systemically for fPJI [105]. There is a high incidence of amphotericin B adverse reactions [106]. Moreover, fluconazole is associated with severe hepatotoxicity; therefore, liver function should be monitored regularly during extended fluconazole therapy. Although liposomal compounds of amphotericin B significantly reduce its nephrotoxicity [107], the duration of antifungal treatment is currently uncertain. The IDSA recommends that most patients with fPJI should be treated with amphotericin B or fluconazole for at least 6 weeks after the second stage arthroplasty [80]. However, the inclusion of cases showed that the duration of treatment usually lasted until the infection disappeared, as determined by systemic inflammatory markers (CRP and ESR) or knee fungal culture results.
The situation poses great challenges when fungi and bacteria are present together. It is worth noting that in the case of bacterial co-infection, the cure success rate decreased from 71.4–50%. This is because Candida and specific bacteria establish a mutually beneficial biofilm that protects them from being killed by drugs. For example, Staphylococcus can affect the activity of antifungal drugs, and staphylococcal proteases enhance the adherence of Candida albicans, thus, favoring the survival of Candida [108]. However, Kong et al. [109] found that a barrier of polysaccharides secreted by Candida, which covers the bacterial surface and physically prevents the interaction between bacteria and antibiotics, led to an increased tolerance of S. aureus to vancomycin. The use of caspofungin reduces the ability of Candida to form biofilms and can penetrate them [110, 111]. In mixed PJI, an abatement of the infection occurs only after caspofungin treatment [112]. Therefore, the biofilm penetration effect is particularly important for the successful treatment of mixed PJI.
Currently, there are no guidelines for the surgical treatment of peri-fungal prosthetic infections. Surgical treatment options include debridement with retention of the prosthesis, OTRA, and STRA. According to the literature, in most cases, prosthesis-preserving debridement does not control fungal infections [82, 113]. However, it has also been suggested that periprosthetic fungal infections can be successfully treated by debridement alone in immunocompetent patients and those with early fungal infections [28, 32]. Our own case was diagnosed with fPJI, promptly treated with effective medications, debridement with a preserved prosthesis, and replacement of polyethylene liner, and no recurrence of infection was seen at five years of follow-up. Nevertheless, in the literature we included, there were very few cases of simple debridement, showing the higher risk of recurrence and how narrow the indications are.
OSRA may be an option for patients who are unlikely to tolerate multiple surgeries. However, the relatively small number of cases has led to less convincing evidence. In addition to case reports, only two single-center studies on OSRA of fPJI were identified [20, 114], in which Klatte et al. has reported the outcome of four cases of single-stage revision of fPJI after TKA, with only one recurrent infection. In another study, Ji et al. reported seven patients with fPJI treated with OSRA and observed one recurrent infection at a mean 5-year follow-up. They thought that OSRA could be successful in treating fPJI. Nine of 63 included cases were treated with OSRA, and one case had infection recurrence during follow-up, which is a success rate not significantly different from the above-mentioned studies, suggesting that there is hope regarding the efficacy of OSRA. However, in the case of drug-resistant fungi, OSRA is unlikely to be effective.
TSRA has received a strong recommendation in a guideline for fPJI treatment [7, 19, 20]. TSRA involves removal of the prosthesis, thorough debridement, placement of cement spacers, and treatment with sensitive antifungal medications. The second stage of prosthesis replantation is performed after the symptoms are disappeared, the joint fluid culture is negative, and the ESR and CRP are normal. Wang et al. [47] treated five patients with fPJI after TKA with TSRA and had a mean follow-up of 41.6 months (range 24–65 months) after prosthesis reimplantation, during which no patient had reinfection or revision for any reason. Additionally, in a single-center retrospective study of seven fPJI patients undergoing TSRA, there were no cases of recurrent infection after a mean follow-up of 28 months [43]. Thirty-two of our included cases underwent TSRA, the median time between the two stages was 16 weeks (range 1–48 weeks), and all the infection was controlled during the follow-up period. Antibiotic-impregnated bone cement spacers are considered a therapeutic option when performing TSRA on fPJI; however, the effectiveness of this approach remains controversial [77, 105, 115]. In the present study, antibiotic-impregnated bone cement was used in 35 patients, single agent in 13 cases (37.1%) and combined agent in 22 cases (62.9%). Vancomycin was the preferred regimen in 21 cases ((60%), in two as a single regimen), followed by gentamicin in 15 cases ((42.8%), in two as a single regimen), amphotericin B in 13 cases ((37.1%), in five as a single regimen), clindamycin in 5 cases ((14.3%), not as single regimen), tobramycin and voriconazole in 2 cases each ((5.7%), in one as a single regimen each) and 1 case each of teicoplanin and azithromycin ((2.9%), both as a single regimen). Several studies have noted the low elution properties of antifungal agents from cement [116, 117]. The elution characteristics and mechanical properties of antifungal agents in cement are two important considerations in TSRA. There are few studies on antifungal agents and cement formulations, with amphotericin B being the most used. Amphotericin B has the advantage of thermal stability and powder form and can successfully treat fPJI [41]. However, another study claimed that although considerable amounts of amphotericin B could be detected, the level was not optimal or sufficient to control deep infections [117]. In recent years, voriconazole, which has been proven to have good elution and mechanical properties, has been increasingly used for bone cement pads with a very optimistic prospect [116, 118]. Although still controversial, TSRA is the most used surgical method for fPJI after TKA, which has a higher probability of successful infection eradication.
There were a few limitations to this systematic review. First, inconsistencies in data reporting across the literature limited the value of the statistics. Second, the quality assessment tools utilized in this article only included the assessment components, and it did not describe how to classify the quality of the articles. Third, the quality assessment might have been affected by inter-user variability, especially under the influence of the second limitation.