Global Epidemiology of Invasive Infections by Uncommon Candida Species: A Systematic Review

Abstract

Emerging and uncommon Candida species have been reported as an increasing cause of invasive Candida infections (ICI). We aim to systematize the global epidemiology associated with emergent uncommon Candida species responsible for invasive infections in adult patients. A systematic review (from 1 January 2001 to 28 February 2023) regarding epidemiological, clinical, and microbiological data associated to invasive Candida infections by uncommon Candida spp. were collected. In total, 1567 publications were identified, and 36 were selected according to inclusion criteria (45 cases). The chosen studies covered: C. auris (n = 21), C. haemulonii (n = 6), C. fermentati (n = 4), C. kefyr (n = 4), C. norvegensis (n = 3), C. nivariensis (n = 3), C. bracarensis (n = 1), C. duobushaemulonii (n = 1), C. blankii (n = 1), and C. khanbhai (n = 1). Over the recent years, there has been an increase in the number of invasive infections caused by uncommon Candida spp. Asia and Europe are the continents with the most reported cases. The challenges in strain identification and antifungal susceptibility interpretation were significant. The absence of clinical breakpoints for the susceptibility profile determination for uncommon Candida spp. makes interpretation and treatment options a clinical challenge. It is crucial that we focus on new and accessible microbiology techniques to make fast and accurate diagnostics and treatments.

1. Introduction

Over the years, the enormous advances in science, including medicine, have allowed a remarkable increase in average life expectancy. However, the rise in longevity has been partly at the expense of better therapeutic options [1]. Intensive care medicine, oncology, and intensive care units are some areas in which prophylaxis against bacteria and fungi has played a lead in preventing opportunistic infections in immunosuppressed individuals [2,3]. In contrast, the increased use of broad-spectrum antibiotics has contributed to increased opportunistic infections, particularly (but not exclusively) in immunocompromised patients, especially fungal infections [4,5,6,7].

The clinical manifestations of fungal infections include a broad spectrum ranging from superficial lesions to life-threatening invasive infections. Since the early 1980s, fungi have been leading in the etiology of invasive infections, with high mortality and morbidity rates. Over the last decade, there has been a dramatic increase in scientific studies associating sepsis with prior fungal infection [3,4,8,9]. According to Bassetti and his colleagues, invasive infections in intensive care units are increasing, and 80% of these infections are caused by Candida species [9]. Candida spp. are commonly distributed in the environment and are part of our normal microflora [10]. A transition from commensal to a pathogenic agent and evolution for invasive infections occur primarily in immunodepression situations or when barrier leakage occurs [11]. Candida infections can be classified as cutaneous, mucosal, or invasive when the infection reaches the bloodstream or other typically sterile sites and deep tissues [12]. Candida albicans is the most frequent species that causes invasive infections with severe repercussions on patient prognosis, morbidity, and mortality and the economic health system [10]. However, in recent decades, we have seen an increase in non-albicans invasive Candida infection (ICI) reports worldwide [13,14,15]. A recent ten-year analysis of the distribution of these species indicated that C. glabrata (new nomenclature: Nakaseomyces glabrata) is the most common, followed by C. parapsilosis, C. tropicalis, and C. krusei (new nomenclature: Pichia kudriavzevii) [12,16,17].

Uncommon Candida species play a significant role in the landscape of candidemia due to their emerging resistance to antifungal treatments, diagnostic challenges, varied clinical manifestations, epidemiological shifts, impact on patient outcomes, and influence on treatment guidelines. In 1991, the first case of C. haemulonii was reported in the USA and associated with a poor prognosis [18]. In 2009, C. auris was identified for the first time [19,20]. Since then, numerous studies have reported cases of invasive infections associated with high mortality and morbidity rates. These emerging pathogens have a low susceptibility to antifungal drugs and affect mainly immunocompromised patients [21]. C. nivariensis and C. bracarensis have been associated to distinct antifungal susceptibility pattern, and, together with C. glabrata (sensu stricto), form the C. glabrata species complex [22]. The unambiguous identification of C. nivariensis and C. bracarensis through traditional methods has been reported [22]. C. fermentati belong to the C. guilliermondii species complex, a heterogeneous taxonomic group with several morphologically indistinguishable species [23,24]. Consequently, ICI by uncommon Candida spp. has led to a public health concern worldwide such as C. auris [25,26,27].

This work aimed to systematically review the worldwide literature regarding the epidemiology, diagnostics, susceptibility profile, and clinical outcome of uncommon Candida species responsible for invasive infection.

2. Materials and Methods

The systematic review was carried out following the Cochrane Handbook for Systematic Reviews of Interventions and reported following the Preferred Reporting Items for Systematic Review and Meta-Analyses (PRISMA) guidelines [28,29].

2.1. Defining the Problem and Establishing the Guiding Question

The PICO anagram (population, intervention, comparison, and outcome) was defined and represented in Table S1 (Supplementary Materials) to perform the guiding question. According to the guiding question, this work did not use the comparison element [30]. The guiding question of this research was: “What is the clinical information, diagnostics, susceptibility profile, and clinical outcome of patients with invasive infections by uncommon Candida species?”

2.2. Data Source and Search Strategy

An electronic literature search comprising terms such as [(Candida OR “Candida species” OR “Candida spp”) AND (“invasive” OR “deep-seated” OR “bloodstream infection”) AND (emergent OR emergence OR emerging) AND (resistance OR resistant OR “multiple drug resistance” OR “multiple drug resistant” OR Pan OR “decrease susceptibility” OR “treatment failure” OR “poor outcome”)] was carried out in databases Pubmed, Web of Science, and SCOPUS to identify case reports of patients with invasive infection caused by uncommon Candida species. The study was conducted from 1 January 2001 to 28 February 2023. The search included all publication types except reviews or systematic reviews, and no language restrictions were applied.

2.3. Eligibility Criteria

Studies were included in this review if they met the following criteria: (1) information reporting invasive infection by uncommon Candida spp in patients (≥18 years old), (2) predisposing factors/underlying medical conditions, (3) method(s) of diagnosis, and susceptibility testing results, (4) information regarding treatment and other patient management strategies, and (5) patient outcome. Uncommon Candida species were defined as all Candida spp. except C. albicans, C. glabrata, C. parapsilosis, C. tropicalis, C. lusitaniae (new nomenclature: Clavispora lusitaniae), C. guilliermondii (new nomenclature: Meyerozyma guilliermondii), C. krusei, and C. dubliniensis. Invasive Candida infection, also known as invasive candidiasis or systematic candidiasis, involves bloodstream and/or deep-seated infections such as those in the brain, kidney, heart, liver, and lungs [12].

The following exclusion criteria were applied: (1) studies regarding noninvasive fungal infection, (2) studies regarding infections by C. albicans, C. glabrata, C. parapsilosis, C. tropicalis, C. lusitaniae, C. guilliermondii, C. krusei, and C. dubliniensis, and (3) studies with no patient clinical information and outcome, and studies with no laboratory data.

2.4. Data Extraction and Synthesis

The screening of publications was carried out by two independent reviewers (SP and SCO) based on the eligibility criteria. The literature screening was based on a 2-step process. First, the extracted studies were uploaded to EndNote 20.4.1 and Rayyan software for duplicate removal and further selection according to the title and abstracts [31]. The second phase was the review of the full text of selected studies. Disagreements were resolved by consensus. The PRISMA flowchart (Figure S1—Supplementary Materials) summarizes the literature search and screening strategies. A protocol was defined to independently synthesize and compare the extracted data from the selected studies. Data extracted included information about the publication date, country, age, gender, clinical history/underlying medical conditions, type of sample, Candida identification methods, susceptibility testing method and results, clinical outcome, microbiology identification method, and antifungal treatment or other management strategies.

2.5. Risk of Bias (ROB) Assessment

To evaluate the risk of bias in the studies included in this review, we used the National Institute of Health (NIH) Quality Assessment Tool for Case Series Studies. A 3-point scale was used to grade nine questions to evaluate the potential source of bias as good, fair, or poor. No studies were excluded based on quality. ROB assessment was performed independently by SP and IMM.

3. Results

3.1. Publication Characteristics

Detailed data on scientific papers included in this review are provided in Table S2 (Supplementary Materials). We identified 1567 publications, including 440 from PubMed, 550 from Web of Science, 577 from SCOPUS, and 47 publications, which were later manually added from the reference lists of identified papers. After checking for duplicates, 916 studies were enrolled for the title and abstract screening; 825 were excluded, and 88 were enrolled for a full-text screening and assessment.

Thirty-six publications reporting on 45 patients were included. Cases from Asia were the most common (n = 22, 47.8%), followed by Europe (n = 15, 33.3%) and America (n = 8, 17.8%). No data were obtained from African countries and New Zealand (Figure 1). The number of included cases per year is represented in Figure S2 (Supplementary Materials).

• Asia

Asia was the continent with the highest number of included cases in the present work, with a total of 22 cases. C. auris was the most isolated agent (n = 10), all related to candidemia (Table S2—Supplementary Materials). A case of C. auris candidemia was reported in Taiwan, related to a 64-year-old male Taiwanese patient who has lived in Vietnam for the past five years [32]. The isolates were closely related to the South Asian Clade (I). In the Middle East, two isolates of C. auris related to candidemia reported in Oman in 2017 were nested into two distinct clusters: the Indian and UK clusters [33]. Both patients never traveled outside their country.

C. fermentati was the second most reported agent associated with invasive infection, with four cases, including one case in which C. famata was also identified in the same isolate [34]. Two cases of C. haemulonii candidemia were also reported in Taiwan and South Korea, respectively, in 2010 and 2011 (Figure 2A and Table S2—Supplementary Materials) [32,35]. C. kefyr, C. nivariensis, C. norvegensis, and C. duobushaemulonii were documented in one case each, all related to candidemia (Table S2—Supplementary Materials).

In 2022, a new Candida spp. associated with an invasive infection case was reported for the first time in 2022: C. khanbhai [36]. A 55-year-old man with no known comorbidities was hospitalized in 2014 for the treatment of new-onset hospital-acquired pneumonia and ended up succumbing to C. khanbhai candidemia.

• Europe

It is worth noting the incidence of emerging infections in Europe in the last decade. This review included 15 cases of invasive infection in Europe reported from distinct European countries: six cases in Spain [37,38,39], three in Greece [40,41], two in France [42,43], and one in Belgium [44], Netherlands [45], Switzerland [46], and Italy [47] (Table S2—Supplementary Materials and Figure 2B). Candidemia was the most documented invasive infection, although hepatic infection and intraabdominal abscesses were also reported [47,48] (Table S2—Supplementary Materials).

C. auris was the most reported agent (n = 9), and all cases were associated with candidemia, with the exception of one related to a hepatic abscess in Greece [48]. Three cases of C. auris invasive infection had Clade information [41,44,45]. Two isolates of. C. auris related to candidemia cases were reported in Belgium and the Netherlands and were associated with Clade I. Both patients had a recent history of hospitalization in India and Kuwait, respectively. Another case of C. auris invasive infection was documented in 2023 in Greece [41]. The isolate was also related to Clade I. However, in these cases, the patient had no history of traveling outside of their country. The authors suggested a possible horizontal transmission.

C. norvegensis and C. nivariensis were documented in Spain and Italy [38,47]. Both were transplanted liver patients, and C. norvegensis was isolated from blood and bile specimens, respectively.

• America

A total of eight cases reported in the American continent were included: three cases of C. haemulonii [49,50,51], two cases of C. auris candidemia, and one case of C. kefyr, C. blankii [52], and C. bracarensis (Table S2—Supplementary Materials and Figure 2B) [53]. C. haemulonii was reported in South America; in contrast, North America had distinct uncommon Candida spp. in the USA and Canada (Figure 2C). Most cases were associated with candidemia (Table S2—Supplementary Materials). A case of liver abscess associated with C. haemulonii was reported in Peru in 2021 [51]: a 72-year-old woman with congenital polycystic kidney liver disease in end-stage renal disease was treated with caspofungin with a full recovery.

A recent C. blankii case was reported in the USA (2021): candidemia with possible endocarditis in a 63-year-old man who was immunocompetent but with a recent hospitalization history due to sepsis [52]. Another uncommon invasive infection was documented, also in the USA, related to a 64-year-old man with a malignancy pathology and a history of recent use of broad-spectrum antibiotics, in which C. kefyr was isolated in pleural fluid.

3.2. Risk of Bias—Quality Assessment

Three studies were rated as Fair due to a lack of information regarding the study question [54], study population [55], and the type of intervention [56] (Supplementary Materials, Table S3).

3.3. Patient Characteristics

Among the 45 patients, 12 (median 26.7%) were female, and 33 cases (median 73.3%) were male patients. In Europe, the male proportion was significantly higher in comparison with females (male: 80%; female: 20%). In contrast, the Middle East had a major proportion of female patients (female: 80%; male: 20%). The age range was from 26 to 88 years, with a mean and standard deviation of 59.1 ± 16.6 years.

3.4. Types of Infection

The most common type of infection was candidemia, with 40 cases in a total of 45 cases (median: 88.9%). In general, C. auris was the most frequently reported yeast to cause a bloodstream infection (n = 20, median 52.5%). C. haemulonii was presented in five cases of candidemia (12.5%); C. fermentati was isolated from four fungemia cases (10%); and C. famata was also isolated in one of these patients [34]. C. norvegensis and C. nivariensis were responsible for each one of the three cases of candidemia. One case had an isolation of uncommon Candida in blood, but no symptoms were associated with candidemia [45].

C. blankii, C. duobushaemulonii, C. bracarensis, and C. khanbhai invasive infections were less common, with only one case of each one of these [36,52,53,56]. However, atypical invasive infections were most common with these agents: endocarditis was associated with a C. blankii invasive infection and C. kefyr was reported in a case of pyelonephritis [40,52].

Eighteen patients (40%) were reported to have more than one Candida agent infection (bacterial and fungal). These cases were reported to have gastrointestinal, urinary, and upper respiratory tract infections or other unspecified infections at the same time or recently.

3.5. Underlying Conditions, Hospitalization Time, Treatment, and Clinical Outcome

Malignancy diseases were the most common underlying condition, present in 16 cases among a total of 45 cases. Diabetes mellitus was noted in 8 patients [32,33,41,52,56,57,58], and transplantation was documented in 10 cases [34,38,46,47,48,53,58,59]. Hospitalization time information was absent in 6 cases [45,46,50,52,55,57,58,60]; however, the remaining 42 cases had a minimum hospitalization time of 8 days and a maximum of more than 100 days. Only 21 cases had information related to mechanical ventilation: 17 patients needed intubation/ventilation. In 34 cases, prior exposure to antibiotics belonging to the broth spectrum was recorded; in 4 cases, this information was not provided. The antifungal agents used for the treatment of invasive candidemia were also documented in Table S5—Supplementary Materials. In a total of 45 patients, 5 cases had no information related to antifungal treatment [49,50,56,61,62].

A patient with C. khanbhai isolated in blood culture died before starting treatment [36]. Fluconazole and amphotericin were the most common antifungals used, documented in 14 cases. Caspofungin was the most documented echinocandin (n = 11), followed by anidulafungin and micafungin, 9 and 8 cases, respectively (Table S2—Supplementary Materials). Fluconazole was also the antifungal agent most used in prophylaxis and empirical treatment (Table S2—Supplementary Materials).

3.6. Mortality

In the 45 cases of uncommon Candida spp. invasive infection in our review, 19 patients died (42.2%). However, one death was not attributed to emergent Candida spp. invasive infection [53].

The average age of the population who ended up dying was 59.4 ± 16.6 years (range of 26–88 years). According to the gender category, 8 of 33 male patients and 7 of 12 women died (24.2% and 58.3%, respectively). The mortality specified for each uncommon Candida spp. is represented in Figure S3—Supplementary Materials.

In a total of 21 cases, 11 patients with C. auris infection died (52.3%). Two of the three C. fermentati cases and half of the C. norvegensis patients succumbed (66.7% and 50%, respectively).

There is only one case of C. khanbhai, in which the patient had an unfavorable clinical outcome [36]. In the case of C. duobushaemulonii candidemia, with unfavorable clinical outcomes, the treatment for Candida invasive infection was withdrawn due to poor neurological recovery [56]. No death was documented in the four patients with C. kefyr, and the three cases of C. nivariensis invasive infection were included in this study.

3.7. Laboratorial Diagnosis

This review documented 45 cases of uncommon Candida species as causative agents of invasive infection. Most of the authors reported that the identification of uncommon species was not initially correct. One case mentioned that the samples went to an external laboratory without further information [52]. Table S4—Supplementary Materials represents the misidentification isolates presented in our included cases. Among the 44 cases with information regardless of identification methods, only 11 cases had a corrected Candida identification in the first attempt (25%) through the VITEK^®, API^® or MicroScan^® system. In 24 cases (54.5%), the infectious agent was misidentified for the first time or not clearly identified with conventional methods.

Confirmation was performed through MALDI-TOF and, in some cases also, PCR sequencing (ITS and D1/D2 domain). It is important to mention that misidentification was also performed with MALDI-TOF: one case of C. auris candidemia was initially identified as C. haemulonii, and the identification of C. kefyr and C. duoshaemulonii candidemia (one case of each) failed [46,57].

The VITEK^® system was one of the most used identification routine techniques in 19 of 44 cases (43.2%). However, C. auris misidentification was common, with 11 in a total of 21 isolates with the VITEK^® system. C. haemulonii was the most frequent agent identified as C. auris (n = 4).

Fourteen cases were initially identified through the API^® system, and five cases with the MicroScan system. However, misidentification was presented in 11 and 4 cases with the API^® system and MicroScan^® system, respectively. Thirteen cases used identification agents with only one method, and DNA sequencing was the most used in these cases (n = 5).

The susceptibility tests of the reported cases included in the present work are summarized in Table S5 and Figure S4—Supplementary Materials. Information related to the susceptibility determination method(s) was absent in four articles [52,59,63,64]. To evaluate the susceptibility of the isolated agents, a variety of techniques were used, such as the Clinical and Laboratory Standards Institute (CLSI) and the European Committee on Antimicrobial Susceptibility Testing (EUCAST) broth microdilution, CLSI microdilution, Etest^®, Sensititre YeastOne^®, and automated VITEK 2^® yeast susceptibility system. CLSI broth microdilution was the most used method (n = 16). EUCAST broth microdilution method was used exclusively in European reported cases (n = 6), including in three isolates whose susceptibility profiles were also analyzed with another technique, the broth microdilution EUCAST method compared with the Etest^® technique in C. nivariensis and C. kefyr isolates and also compared with the Sensititre YeastOne^® technique in the C. auris isolate [40,42,44].

In one case related to a C. kefyr invasive infection in 2004, CLSI broth microdilution was performed [40]. The automated VITEK2^® yeast susceptibility system was applied only in on case reported in 2014, in Italy [47]. Another distinct colorimetric microdilution antifungal susceptibility test, MICRONAUT, was recently performed in one fatal case of C. auris [58].

Different mutations in FKS1 and ERG11 genes, associated with antifungal drug resistance, were analyzed in four and two studies, respectively (Table S5—Supplementary Materials) [32,58,59,65]. FKS1 was presented in two C. auris invasive infection cases, and one C. fermentation candidemia and ERG11 was detected in one immunocompromised patient with a C. auris fatal infection. A case reported by Tsai, M. H. et al. related to C. auris candidemia also evaluated FKS1 and ERG11 mutations, but none of them were detected [32].

3.7.1. C. auris

Regarding the susceptibility profile, C. auris had, in general, high values for fluconazole with a range between 2 and ≥256 mg/L; 14 of 21 C. auris isolates had a high MIC to fluconazole, according to the Center for Disease Control (CDC) breakpoints [66]. One case did not have information regarding fluconazole [64]. All the nine isolates of C. auris reported in Europe had high MICs of fluconazole (Table S5—Supplementary Materials).

Susceptibility testing for amphotericin B was performed in all cases, and the results revealed that 4 of 21 C. auris isolates had high MICs to this antifungal, according to the CDC breakpoints [66]. Curiously, three of the four C. auris cases with a high MIC to amphotericin B also showed high values for fluconazole [33,41,67].

The susceptibility testing results for other azoles showed 10 C. auris isolates with a high MIC to voriconazole, 2 cases of posaconazole, and 1 case of itraconazole (Table S5—Supplementary Materials). One isolate documented in Russia in 2019 exhibited a high MIC to caspofungin [55]. Most of the studies reported low MIC values to anidulafungin and micafungin.

A particular C. auris isolate reported in the USA presented an FKS1 mutation and revealed, according to the authors, high MICs to echinocandins [65] (Table S2—Supplementary Materials). Another isolate of C. auris associated with invasive infection exhibited FKS1 and ERG11 mutations and also high MICs to echinocandins and fluconazole [58]. All FKS1 mutations occurred in hotspot 1.

3.7.2. C. haemulonii

One reported case of C. haemulonii had no information related to the susceptibility testing method used [63]. In this case, the authors mention that the isolate is susceptible to echinocandins such as caspofungin or micafungin, and it was not resistant to new triazoles (e.g., voriconazole).

The cutoff of CLSI was used in four of five C. haemulonii isolate cases that presented with the susceptibility profile testing method (Table S5—Supplementary Materials). In these cases, the interpretation results were performed according to the susceptibility breakpoints established by the CLSI: ≤8 mg/L for fluconazole, ≤0.125 mg/L for itraconazole and ketoconazole, ≤4 mg/L for flucytosine, ≤1 mg/L for voriconazole, posaconazole, and amphotericin B, and ≤2 mg/L for caspofungin, anidulafungin, and micafungin [49,51,60]. According to these breakpoints, a total of four C. haemulonii isolates had high MICs to amphotericin B, fluconazole, and itraconazole (Table S5—Supplementary Materials). An isolate from a candidemia case reported in Brazil, with a previous history, showed a high MIC to 5-flucytosine [49]. No C. haemulonii isolate showed high MICs to echinocandins.

3.7.3. C. fermentati (New Nomenclature: Pichia fermentans)

The susceptibility profile results of C. fermentati (Table S5—Supplementary Materials) do not present significant discrepancies. No isolate showed high MICs to azoles such as fluconazole and voriconazole. Itraconazole was the exception since three isolates showed a high MIC to this azole. Only one isolate revealed a high MIC of amphotericin.

Konuma, T. et al. reported one case of C. fermentati candidemia in which an FKS1 mutation was identified, and, according to the authors, the susceptibility testing results showed a high MIC to echinocandins [59].

A case report of C. fermentati published in 2018 demonstrated a discrepancy between the in vitro and in vivo susceptibility to antifungal drugs, particularly in echinocandins [34]. A patient whose isolate revealed low MICs for amphotericin B and micafungin was initially treated with amphotericin and micafungin; however, blood cultures were persistently positive, and, ultimately, the patient died [34].

3.7.4. C. nivariensis (New Nomenclature: Nakaseomyces nivariensis)

Of the three cases of C. nivariensis invasive infection reported in the present work, none of them used CLSI broth microdilution. Cartier N. et al. performed two different instances of susceptibility testing in the same C. nivariensis isolate: Etest and broth microdilution according to EUCAST [42]. The results presented in Table S3 showed distinct MICs for azoles (fluconazole, voriconazole, and posaconazole), amphotericin B, and micafungin.

3.7.5. C. norvegensis (New Nomenclature: Pichia norvegensis)

A few cases have been reported of C. norvegensis associated with invasive infection, and most of the cases are associated with acute comorbidities such as solid and liquid malignancy or immunosuppression [25]. In fact, C. norvegensis was reported only in three included cases, and all of the patients were immunosuppressed and had leukemia [38,47,68]. The C. norvengensis susceptibility profile indicates high MICs for fluconazole in two included cases [38,68]. The echinocandins susceptibility profile was performed in only one case, through caspofungin MICs, which showed low values (0.047 mg/L) [38].

3.7.6. C. kefyr (New Nomenclature: Kluyveromyces marxianus or Candida pseudotropicalis)

In recent years, the number of C. kefyr invasive infection reports increased [38]. In our work, four cases were included, and all showed a low MIC to fluconazole and itraconazole [40,46,57,67]. Two cases reported in Europe showed a high MIC to amphotericin [46]. One included case did not present numerical MICs. However, the authors concluded that the isolates were susceptible to amphotericin and azoles [57].

3.7.7. C. famata (New Nomenclature: Debaryomyces hansenii), C. bracarensis (New Nomenclature: Nakaseomyces bracarensis), C. khanbhai, C. blankii, and C. duobushaemulonii

The susceptibility testing results of C. duobushaemulonii and C. khanbhai revealed a high MIC for fluconazole [57,58]. The C. famata susceptibility profile was documented in four cases (Table S5—Supplementary Materials), and the results suggested a susceptibility to both the polyene and azole antifungal classes [34,59]. On the contrary, the C. blankii isolate susceptibility profile showed high MICs for both amphotericin and azoles (Table S5—Supplementary Materials).

The susceptibility profile of the only case of C. bracarensis candidemia was performed with the YeastOne method, and results showed low MICs to all antifungals tested (Table S5—Supplementary Materials) except for itraconazole, whose susceptibility was classified as “susceptible dose-dependent” [53].

4. Discussion

Invasive infections by uncommon Candida species represent a significant and complex challenge in clinical microbiology and infectious disease management. These uncommon species are of particular concern due to their potential for resistance to standard antifungal treatments and their varying virulence profiles, being a great challenge for patients and clinicians.

4.1. Population and Comorbidities

According to numerous studies, the male gender has been identified as a risk factor for ICI, with a prevalence of between 52% and 60% [69,70]. A recent systematic review developed by Egger M. et al. has shown that gender is a risk factor for ICI, where females represent 51.2% of ICI [71]. According to the authors, immune responses can vary between females and males due to a variety of factors, including genetic and hormonal influences. For instance, the presence of two X chromosomes and the hormonal effects of estrogen and progesterone may contribute to different immune response patterns. However, it is important to recognize that immune responses are highly dynamic and context-dependent, and further investigation is mandatory in order to unveil the link between biological gender and ICI [71]. This fact was not in concordance with the studies enrolled in this systematic review. Moreover, there are few case reports of invasive infections by uncommon Candida species when compared with the most prevalent Candida species.

Other risk factors for ICI beyond demographic factors and gender also include comorbidities and medical interventions [11]. Immunosuppression conditions such as transplantation or HIV, solid and hematological malignancies, corticoid therapy, kidney diseases, and neutropenia are known major risk factors for ICI [11]. The mechanisms involved in each condition have been studied, noticing that the destabilization of the immune system is the theory basis [72,73]. In this review, it is worth noting the high proportion of transplanted and oncology patients. Diabetes was also a common comorbidity in the studies included. It is known that Candida spp. are capable of growing in biofilm forms, exhibiting enhanced resistance against most antifungal agents [74]. High levels of glucose in the blood (common in diabetes) promote an increase in Candida biofilm formation and, consequently, develop pathogenicity mechanisms [71]. Other important predisposing factors leading to ICI include recent antifungal treatments, previous broad-spectrum antibiotic therapy, and mechanical ventilation [75]. All these risk factors were presented in several patients in this review associated with unfavorable clinical outcomes.

4.2. Uncommon Candida spp. Identification Challenging

It is known that difficulties with Candida species identification, especially the emergence of new Candida species and the lack of awareness, have resulted in transmission and several outbreaks which have remained unnoticed [76]. The development of rapid and precise methods for identifying uncommon Candida invasive agents is a challenge, but it is imperative for patient care and to certify the appropriate implementation of measures to ensure the prevention of infection [22,77,78,79].

Among the conventional methods that traditionally identify pathogenic yeast based on biochemical and morphological characteristics at the species level that are commercially available, the VITEK^® 2 and API^® AUX systems are the most routinely used systems in clinical microbiology [80]. Unfortunately, the databases of these methods are limited, time-consuming, require high experience in this area, and generally identify only the most common pathogens’ yeasts [80]. In the last few years, evolution in molecular biology has provided the increasing use of DNA techniques as tools for yeast identification and characterization with a high accuracy [81]. Recently, MALDI-TOF and DNA sequencing (ITS and D1/D2 regions) have been discussed as accurate identification methods for C. auris [82].

An included study in this systematic review developed by Ruan, S. Y. et al. reported that the VITEK 2 system, in comparison with the VITEK^® 1 and API^® 32C system, is a better solution to C. haemulonii identification [60]. The VITEK^® 2 system results were highly consistent with molecular methods for the sequence analysis of the rRNA [60]. In fact, these conclusions are in concordance with the other five included cases in which C. haemulonii was correctly identified with the VITEK^® 2 system (Table S4—Supplementary Materials).

The identification of C. fermentati in one of three cases was successfully performed with the VITEK^® 2 system. DNA sequencing was performed, identifying C. fermentati in all the three cases. Konuma, T. and colleagues documented the VITEK^® 2 system misidentification of C. fermentati in a fungemia case [59]. The authors confirmed that this result is consistent with recent publications: C. fermentati is commonly misidentified as C. famata by the VITEK^® 2 system [83,84,85]. Since it is a cryptic species of C. guilliermondii, C. fermentati misidentification as C. guilliermondii consequently leads to underreported infections due to this pathogenic yeast [34].

C. nivariensis identification is another challenge in mycology laboratories since it is routinely misidentified as C. glabrata [86]. Poor identification with conventional methods is common. In the present work, all the isolates identified with the VITEK^® or API^® AUX systems were misidentified, or the identification was not conclusive [39,42]. Fortunately, in order to distinguish C. nivariensis from C. glabrata, several molecular methods, such as multiplex PCR, a fast, cost-effective, and reliable tool, can be used [87]. Currently, five different genetic Clades have been found related to C. auris species [88]. The South-Asian (Clade I), East-Asia (Clade II), Africa (Clade III), and South America (Clade IV) clades, and, more recently, a fifth clade (Clade V), separated from the others by more than 200,000 SNPs, were confirmed to have appeared in patients from Iran [89]. In this review, only four cases had results from a genetic analysis related to clade identification. Clade I was found in three cases reported in Europe (Netherlands, Greece, and Belgium) [41,44,45]. Two patients had a history of hospitalization in Asia, and, in another case (isolated in Greece), they had no travel information, but the authors suggested a horizontal transmission, since C. auris with identical sequences was detected from environmental screening.

4.3. Susceptibility Interpretation and Treatment Challenges

Antifungal susceptibility testing is an increasingly vital tool for understanding local and worldwide disease epidemiology, considered by clinicians to guide their treatment decision-making process, the treatment of fungal illnesses, and identifying antifungal resistance [90]. Distinct methods are currently available for evaluating the susceptibility of fungi to different antifungals. Broth microdilution is one of the most common susceptibility methods used in clinical and research microbiology laboratories.

Over the years, several commercial susceptibility methods have been available with the advantage of easier interpretation and effective MICs such as Etest and YeasOne [91]. Broth microdilution from CLSI and EUCAST, Etest^®, and Sensititre^® YeastOne were the most used methods in our cases.

Regarding susceptibility testing, there are two different approved reference procedures based on microdilution techniques for antifungal susceptibility testing: CLSI and EUCAST [92,93]. CLSI and EUCAST define breakpoints for C. albicans and other common Candida non-albicans, such as C. glabrata, C. parapsilosis, and C. tropicalis [94]. In 2020, the CDC suggested tentative MIC breakpoints for C. auris, and it was applied in the most recent published cases included in this systematic review [41,58,95]. Recently, CLSI also suggested breakpoints for some uncommon Candida spp, including species included in the present study, such as C. haemulonii, C. duobushaemulonii, and C. kefyr [96]. Although these new breakpoints are for uncommon species, most of the studies included in the present work emphasize the challenges regarding the interpretation of susceptibility results. Several papers revealed that, due to the absence of clinical breakpoints for the majority of uncommon Candida species, MICs obtained have been compared to the susceptibility cutoffs determined for common Candida species by CLSI or EUCAST [34,39,47,67]. Nonetheless, it has been reported that susceptibility varies significantly among Candida species [59]. Morita k. et al. used CLSI cutoff values of C. guilliermondii for the interpretation of susceptibility results of C. fermentati isolated from a case of candidemia [34]. Regarding C. guilliermondii and C. fermentati susceptibility profiles, there seems to be not much difference, with echinocandins having good antifungal activity against both species [34,97]. On the other hand, C. fermentati micafungin resistance has been documented and is associated with FKS1 mutation [59].

Multidrug-resistant Candida spp species are increasingly reported [97]. It was also observed in the present work that several isolates of distinct Candida spp. had high MICs for azoles, polyenes, and echinocandins, mainly fluconazole. Candida spp. antifungal resistance mechanisms can be conferred through gain-of-function mutations in different target pathway genes or in their transcriptional regulators [98]. The present work included four C. auris and C. fermentati cases, which analyzed the presence of the most known mutations associated with a reduced susceptibility to fluconazole and echinocandins: ergosterol biosynthesis pathways mutations (ERG11) and cell wall biosynthesis mutation (FKS1), respectively [32,58,59,65].

4.4. Global Epidemiology

Asia was the epicenter of several uncommon Candida species [20,99]. It was noted that there was a high incidence of reported cases documenting high MICs for distinct classes of antifungal agents (Table S5—Supplementary Materials). Due to this emergence of uncommon Candida species antifungal resistance, susceptibility testing is crucial to guide the management of ICI in Asian countries [35]. C. khanbhai associated with invasive infection was reported for the first time in Malaysia recently with a high MIC to amphotericin and azoles in a patient with no comorbidities, who died after hospitalization due to new-onset hospital-acquired pneumonia and posterior diagnosis of candidemia associated with this new Candida spp.

Asia is one of the epicenters of antimicrobial drug resistance and a growing concern related to the dissemination of multidrug-resistant pathogens [100]. A meta-analysis conducted by Habibzadeh, A. et al. compared fungal drug resistance between Europe and Asia [99]. The results demonstrated that Asia had a high prevalence, which was justified due to the poor global health infrastructure in most Asian countries [99]. These findings may be one of the reasons why Asia was the continent with the highest number of reported cases in our work.

Europe presented the highest number of C. auris isolates with a high MIC to fluconazole and voriconazole. Although the reasons for the changing epidemiology of invasive fungal infections in Europe are not entirely known, Lass-Flörl and colleagues suggested that it probably results from different factors, such as the increased use of fluconazole prophylaxis in cancer and transplant patients and changes in treatment strategies for various at-risk populations [101].

In Europe, mortality associated with C. auris invasive infections was reported [37,41,48]. One of the most puzzling characteristics of C. auris is the recent simultaneous and independent emergence of five genetically distinct clades on three continents [102]. It is known that the thermotolerance of C. auris compared with other Candida species is phylogenetically similar. Indeed, it has been suggested that global warming is a contributing factor to the emergence of C. auris [102]. Although the association between human and animal health and the shared environment has been documented in recent years, the current understanding of the effects of climate change and risk for yeast diseases is still incompletely understood [103]. However, Bajpai, Vivetek K. et al. documented the change in climate as a major reason for the growth of a number of these invasive infections, suggesting that global warming and moisture effects on pathogen sporulation and dispersal could benefit certain pathogens and the introduction of new potential vectors [104].

The European Centre for Disease Prevention and Control conducted surveys related to epidemiology, laboratory capacity, and preparedness for C. auris in Europe in two distinct periods: 2013 to 2017, and January 2018 to May 2019 [105]. However, since the information was not updated after the COVID-19 pandemic, a C. auris survey was conducted in April 2022 in order to understand the epidemiological situation and control measures implemented regarding C. auris infections in Europe [106]. According to the obtained data, C. auris was assessed as endemic in Spain, which was one of the first countries to report outbreaks, and five countries reported outbreaks in the period of 2019 to 2021: France, Greece, Germany, Denmark, and Italy [106]. Only two C. auris cases were not reported from outbreak zones: Belgium and the Netherlands [44,45].

In European countries, the distribution of invasive infections by uncommon Candida species differs from one country to another. C. norvegensis was reported in two cases, both transplanted patients [38,47]. This pathogen was isolated for the first time in asthmatic patients in 1954, in Norway [38]. In 1990, the first case of C. norvegensis invasive infection was reported in Denmark [107]. According to Pfaller M. et al., C. norvegensis is more prominent in Eastern European [108]. This fact concurs with the results of the present work since there are only three cases of C. norvegensis, two of which were reported in Europe [38]. In both cases, the patients underwent liver transplantation, a risk factor associated with C. norvegensis infection [25].

C. nivariensis was first described in Spain in 2005, collected from one institution from bronchoalveolar lavage, blood culture, and urine samples [109]. Only two cases of C. nivariensis invasive infections in Europe were included [39]. In contrast, no C. bracarensis cases were identified. In these cases, the importance of optimizing the identification techniques available is highlighted, focusing on speed, simplicity, and reliable results, suggesting MALDI-TOF MS as a suitable option [39]. The reported C. nivariensis candidemia in Spain was associated with nosocomial infection [39]. The authors documented the gardens and flowers presented in the hospital as possible sources of contamination [39].

Candida nivariensis and C. bracarensis are two species related to C. glabrata, being phenotypically indistinguishable [109,110,111].

Candida kefyr was reported in Europe in two case reports [40,46]. This species has gained recent importance as an emerging opportunistic yeast predominantly in immunocompromised patients [40]. Candida kefyr pyelonephritis cases are limited [40]. The authors highlighted the possibility of bloodstream dissemination from the urinary system or hematogenous spread to the kidneys.

Evidence from the included cases from American countries suggests that, although uncommon species are rare, they are becoming pathogens, namely, in immunosuppressed patients with high mortality rates [49,112]. In several studies, the challenges associated with the lack of economic, rapid, and accurate identification and susceptibility methods were also reported [50,65]. Although, in this systematic review, the number of cases in South America is limited, Sifuentes-Osornio, J. and colleagues documented that the incidence of candidemia is higher in this region than in Europe and the United States [113]. The authors also reported that the Candida spp. distribution in Latin America is distinct from other regions, which was associated with huge variations in the quality of health care, namely, in high-risk patients. In fact, C. haemulonii was the pathogen most reported in this region, in contrast to all the other locations.

With the present work, it was possible to understand the concordance in the challenges related to the emergence of uncommon invasive Candida species. It was also possible to understand that some species are most related to specific comorbidities; for example, in the included cases, C. norvegensis formed the majority of the species present in transplanted patients [38,47]. It was also observed that there was resistance to fluconazole of distinct uncommon Candida species independently of geographical localization [56].

In 2016, Calvo B., et al. reported the first outbreak of C. auris in America [114]. In 2018, an epidemiological survey in the USA reported that all the isolates were genetically related to the South American, African, East Asian, or South Asian clades, suggesting that C. auris was introduced in the USA through travel-related cases [115].

Global change, encompassing climate shifts, increased international travel, and altered land use, has promoted shifts in the natural balance between humans and micro-organisms, favoring the emergence of novel, resistant strains that challenge current medical therapies and public health initiatives.

4.5. Limitations

There are several studies reporting ICI due to uncommon Candida species; however, the exclusion of studies with no information regarding the clinical outcome or susceptibility results may have led to a bias in the selection and results.

Our findings are limited, nevertheless, by the quality and breadth of the low number of cases and data in the reports, which were not uniform (for example, distinct identification and susceptibility methods; and different comorbidities). The low number of studies related to some of the uncommon species may limit us from reaching conclusions.

5. Conclusions

It is undeniable that, in recent years, uncommon Candida species have emerged worldwide. Over the last few years, numerous hospitals have reported a significant and progressive shift in the etiology of invasive candidiasis in diverse patient populations and distinct hospital settings [116]. In our work, diagnostic delays and aggravating related to difficulty in fast and accurate identification were reported in the vast majority of the included studies [37,42]. Early and accurate identification and tailored antifungal therapy are essential for effective management and improving patient outcomes. Ongoing research into the pathogenesis, diagnosis, and treatment of these infections is critical to address the challenges they present in the clinical setting worldwide.