Determining Susceptibility in Candida Vaginal Isolates

ABSTRACT

Antifungal drug susceptibility tests (AST) for Candida albicans are increasingly demanded for women with refractory or recurrent Candida vaginitis due to fluconazole resistance. Given reduced activity of azole drugs at pH levels found in women with Candida vaginitis, it is proposed that AST be performed at pH 4.5, since testing at only the recommended pH 7.0 is likely to miss a significant number of clinically relevant azole-resistant C. albicans vaginal isolates.

TEXT

Optimal management of vaginitis by Candida species not infrequently requires performance of antifungal drug susceptibility tests (AST). Although not required on a routine basis, MICs are increasingly requested because of increased numbers of women infected with clinical resistance (1–6). C. albicans remains the dominant cause of yeast vaginitis, most frequently treated with fluconazole and topical azoles (5, 7). Utilizing CLSI M27-A4 methodology, organisms with MICs of ≤2 μg/mL are defined as susceptible to azoles, and MICs of ≥8 μg/mL are designated resistant (8–11). This breakpoint definition was based on clinical experience and drug pharmacokinetics; however, focus was based on consideration of invasive candidiasis. Little consideration was given to vulvovaginal candidiasis (VVC), with failure to recognize differences in drug pharmacokinetics and vaginal pH. Vaginal pH during episodes of VVC is 4 to 4.5. Both CLSI and EUCAST standards measure antifungal drug and yeast interaction only at pH 7.0 (8–11). Danby et al. reported the effect of growth medium pH in performing AST involving Candida species by evaluating a variety of antifungal drugs (2, 12). Both the azole drug class and polyenes showed reduced antifungal activity at pH 4.5, in contrast to flucytosine and echinocandin use. The reduced pH-dependent effect of all azole drugs had previously been reported by Marr et al. (13). Danby et al. suggested that only testing at pH 7 has clinical consequences, missing clinically relevant azole resistance in pathogenic isolates (12). We report our experience at the Wayne State University Vaginitis clinic, where MIC determination has been performed at pH 7 and 4.5 over the last 10 years.

AST are infrequently obtained and only at the request of the treating clinician suspecting clinical azole resistance, such as when a patient with proven acute VVC fails to respond clinically (symptoms and microscopy) to appropriate azole therapy, either oral or topical, or when a compliant patient receiving maintenance fluconazole therapy experiences breakthrough culture-confirmed symptomatic VVC but not simply on the basis of vaginal colonization, i.e., positive yeast cultures. Candida species identification was obtained using reinoculation of CHROMagar Candida plates (CHROMagar Microbiology Inc.), germ tube formation, and chlamydospore production. Broth microdilution tests based on the CLSI M27-A4 reference method but including both pH 7 and 4.5 were performed at MIC ranges of 0.125 to 64 mg/mL fluconazole (Pfizer), and pH was adjusted utilizing HCl and a MOPS (morpholinepropanesulfonic acid; Sigma-Aldrich) buffer solution. The MICs were read as the lowest antifungal concentration with substantially lower visual turbidity (50% growth reduction) compared to growth in the antifungal-free tubes. C. albicans isolates tested at pH 7 were defined as sensitive to fluconazole at ≤2 μg/mL, those with MIC of 4 μg/mL as sensitive dose dependent, and resistant with MIC of ≥8 μg/mL (8–11). ATCC strains of C. parapsilosis (22,019) and C. krusei (6,258) were used as a quality control. As background we randomly tested 100 C. albicans vaginal isolates for which MIC testing had not been requested. Results revealed fluconazole MIC₉₀ of ≤0.5 μg/mL at pH 7 and ≤2.0 μg/mL at pH 4.5.

From January 2012 to December 2021, the Vaginitis clinic each year experienced approximately 1,100 patient visits, including 200 to 300 new patients. Since 2015, AST were performed on 13% of C. albicans isolates. C. albicans strains dominated Candida isolates (74.8%).

We analyzed the results of 125 MIC determinations performed between 2015 and 2021 at pH 7 and pH 4.5, prompted by suspicion of clinical resistance. At pH 7, 38 C. albicans isolates (38/125, 30%) were determined to have fluconazole resistance, with MICs of ≥8 μg/mL. All 38 resistant isolates had replication of fluconazole resistance at pH 4.5, although the latter median MIC was 32 μg/mL. Accordingly, no additional benefit in measuring MIC at pH 4.5 was apparent. At pH 7, 33 C. albicans isolates (33/125, 26%) were determined to have fluconazole dose-dependent susceptibility (SDD; 4 μg/mL). Clinicians might be expected to treat these patients with a higher dose of fluconazole. However, when tested at pH 4.5, all 33 isolates had MICs that demonstrated resistance to fluconazole using a cutoff of ≥8 μg/mL. The mean fluconazole MIC was 16 μg/mL. Finally, at pH 7, 54 (42.3%) of C. albicans isolates were initially considered sensitive to fluconazole (MIC of ≤2 μg/mL); however, measurement of sensitivity at pH 4.5 revealed that 17/54 (31%) isolates were considered resistant (MIC of ≥8 μg/mL). Clearly these patients with C. albicans isolates with median MIC of 8 μg/mL at pH 4.5 should not have been offered fluconazole therapy. Furthermore, 9 C. albicans isolates had MICs of 1 μg/mL at pH 7, only 2 of 9 remained sensitive (≤2 μg/mL) at pH 4.5, and 3 were reclassified as SDD. Hence, even C. albicans isolates with MIC determinations of 1 μg/mL at pH 7 are problematic. Thirty C. albicans isolates had MIC values of ≤0.5 μg/mL at pH 7, none of which had MICs of ≥8 μg/mL at pH 4.5, i.e., none met the resistance definition.

When the pH of the medium used in azole susceptibility testing is reduced, a predictable reduction in drug activity is evident (2, 3, 12, 13). The clinical relevance of the increased MICs obtained at vaginal pH 4.5 is not widely appreciated and is not validated in the present study because of the absence of a treatment outcome. Among the C. albicans isolates tested susceptible or intermediate (DDS) at pH 7, a large number were resistant at pH 4.5. On the basis of the aforementioned results, we propose that determination of vaginal Candida organism fluconazole susceptibility is best achieved by measurement at pH 4.5 only, potentially resulting in improved recognition of fluconazole resistance and enhanced therapeutic outcome.