In this study, we collected S. cerevisiae isolates from 75 clinical patients and calculated the MICs. To our knowledge, this is the first clinical S. cerevisiae study conducted in East Asia. These S. cerevisiae isolates were isolated from vaginal secretions, cervical secretions, ALF, sputum, feces, and intestinal mucosa, representing the reproductive system, respiratory system, and digestive system, similar to previous reports [7, 8, 12]. Given the sources of isolate transmission, S. cerevisiae, a yeast widely used in the food industry, is normally detected in the human digestive tract. Young children are prone to coughing while eating, with a small lung capacity, and/or iatrogenic procedures such as tracheal intubation, which may also lead to foodborne yeast aspiration into the lungs. The pathogens responsible for VVC are usually transmitted through endogenous infection, which may originate from the intestine or through colonization of the vagina. Although S. cerevisiae may cause invasive infections, as in previous reports [5, 8, 13], our institution has not found any evidence of invasive cases, and S. cerevisiae has not been detected in the bloodstream.

Few studies have reported the antifungal susceptibility of clinical strains of S. cerevisiae. The distributions of the MICs of amphotericin B, fluconazole, itraconazole, voriconazole and flucytosine against the clinical isolates of S. cerevisiae in this study (Supplementary Table 1) are consistent with those reported by Borman et al. [14]. However, compared with the study by Górzyńska et al. [9], which was conducted in Poland with 55 cases, the mean MICs of amphotericin B, fluconazole, micafungin, caspofungin and flucytosine differed by 2-fold, whereas the mean MICs of itraconazole and voriconazole differed by approximately 4-fold. Possible reasons for this overall difference in antifungal susceptibility include the following: (1) These studies were conducted in different regions, Eastern Europe and East Asia, and there may be large differences in the strains of S. cerevisiae between the two groups. (2) The race, age and sex of the subjects included in the two studies were different; this study mainly included female VVC patients, and the antifungal susceptibilities of S. cerevisiae isolated from different sites of the body were different.

Among the 75 clinical isolates of S. cerevisiae in this study, 44 were the causative pathogen of VVC. Given the heterogeneity in the susceptibility of S. cerevisiae isolates from different isolation sites to antifungal agents, we grouped the isolates for analysis. Another difficulty in interpreting the antifungal susceptibility of S. cerevisiae as a pathogen is the lack of clinical breakpoints established by scientific or industrial societies. As a medical laboratory accredited by both ISO 15,189 and the CAP, we use the CLSI guidelines to interpret the resistance of clinical isolates, including breakpoints and ECVs, but these standards lack data on S. cerevisiae. Therefore, to evaluate the clinical therapeutic effect of antifungal agents as realistically as possible with in vitro experiments, we used VVC patients with pathogenic C. albicans as a control group via stratified sampling; C. albicans is the primary pathogen in more than half of VVC cases.

Treatment with azoles, including topical and oral therapies, is the preferred recommended regimen for VVC according to the Sexually Transmitted Infections Treatment Guidelines (2021) [15]. In this study, the mean MICs of azole agents against S. cerevisiae were significantly lower than those against C. albicans, and the proportion of voriconazole-resistant isolates of S. cerevisiae was significantly lower than that of C. albicans with respect to the breakpoint of C. albicans, approximately 4.5%. In addition, after antifungal treatment, S. cerevisiae caused a lower rate of RVVC than did C. albicans, suggesting that VVC caused by S. cerevisiae is easier to treat and has a better prognosis.

As the resistance of RVVC pathogens to azole agents becomes increasingly common [16], treatment options other than azoles are gaining attention. In our institution, we usually treat these patients with amphotericin B vaginal effervescent tablets, to which all 88 VVC pathogen isolates in this study were sensitive according to the breakpoint of C. albicans. Terbinafine is an allylamine antifungal drug that inhibits squalene epoxidase in the synthesis of ergosterol in fungal cells and causes the accumulation of squalene in the cells, killing the fungus. Terbinafine is usually suitable for treating skin and nail infections caused by superficial fungi [17]. In this study, the MIC of terbinafine against S. cerevisiae was significantly lower than that against C. albicans, suggesting that it has the potential to be an adjuvant treatment for VVC caused by S. cerevisiae. Echinocandins (including micafungin and caspofungin) are water-soluble lipopeptides that inhibit glucan synthase and are almost exclusively available as intravenous preparations, usually for the treatment of invasive candidiasis [18]. Although there were no cases of invasive S. cerevisiae in this study, the MICs of echinocandins against S. cerevisiae were significantly greater than those against C. albicans, suggesting that echinocandins should be used with caution in the treatment of invasive S. cerevisiae infection. Flucytosine is often used in combination with other antifungal agents because resistance to this agent is common among pathogenic fungi. In this study, although the mean MIC of flucytosine against C. albicans was nearly 50 times greater than that against S. cerevisiae, there was no significant difference between the two groups, suggesting that the efficacy of flucytosine alone is highly heterogeneous.

Compared with the study by Borman et al. [14] which also used the breakpoints of C. albicans to determine antifungal susceptibility, the overall resistance rates of S. cerevisiae to amphotericin B (0.0% vs. 0.4%), fluconazole (31.8% vs. 43.1%) and voriconazole (4.5% vs. 4.0%) in this study are similar. However, we were unable to conduct a horizontal comparison of the antifungal susceptibility of clinical S. cerevisiae isolates with more studies because the sample sizes, disease distributions, experimental methods, and interpretation criteria of each study varied greatly [19, 20].

In this study, we also isolated S. cerevisiae from children with pneumonia and diarrhea, but there was no evidence that these strains were directly related to the corresponding diseases. Studies have shown that for immunosuppressed patients, S. cerevisiae can cause systemic infections [6, 8, 13]. Therefore, pediatric leukemia wards in children’s hospitals need to be aware of the potential pathogenicity of S. cerevisiae, especially in patients with pneumonia symptoms and S. cerevisiae isolated from the respiratory system. Moreover, S. cerevisiae is often made into probiotic preparations and is used orally as an adjuvant treatment for patients. Although the literature [21] indicates that S. cerevisiae strains are transient fungi in the intestine and do not cause colonization, some studies have shown that oral preparations may also be a channel through which S. cerevisiae can cause systemic infection [22, 23].

Overall, this is the first study in East Asia to report the detection and distribution of S. cerevisiae in medical institutions and to compare the antifungal susceptibility of S. cerevisiae and C. albicans as causative pathogens of VVC. As a newly defined rare conditional pathogenic fungus, the clinical characteristics and antifungal susceptibility of S. cerevisiae deserve more attention. In this study, we not only conducted an overall antifungal susceptibility analysis of clinical S. cerevisiae isolates but also systematically reviewed the clinical information of patients and conducted an in-depth analysis of isolates from patients with various diseases. However, this study also has several limitations. First, patient information about lifestyle and occupation was lacking (such as bread baking, frequency of body cleaning, and oral or topical use of probiotics), so we cannot determine whether the source of S. cerevisiae was related to lifestyle and/or occupational exposure. Second, virulence factor and molecular biological testing of the isolates was lacking, so we cannot confirm whether there are genetic differences between the pathogenic isolates and the wild-type isolates, which needs to be further explored.