Abstract
Abstract
Cryptococcosis is a severe systemic mycosis, and relapse may be influenced by fungal subpopulations that survive fluconazole exposure. This retrospective cohort study investigated whether fluconazole minimum inhibitory concentration (MIC) and the HetR phenotype of the first available diagnostic cerebrospinal-fluid isolate were associated with recorded relapse within six months in 70 patients with neurocryptococcosis from Amazonas, Brazil. One clinical isolate per patient, comprising 64 Cryptococcus neoformans VNI and six Cryptococcus gattii VGII isolates, was evaluated by broth microdilution and endpoint-based HFL testing. Using an operational criterion of HFL/MIC ≥8, corresponding to visible growth at least three twofold dilutions above the MIC, 43 of 70 isolates (61.4%) were classified as displaying a HetR phenotype. Six-month follow-up was available for 69 patients. Recorded relapse occurred in 12 of 42 patients infected with HetR isolates (28.6%) and in 4 of 27 patients infected with non-HetR isolates (14.8%). Although relapse was numerically more frequent in the HetR group, the association was not statistically significant (conditional odds ratio, 2.27; exact 95% confidence interval, 0.59–10.96; two-sided Fisher’s exact test, p=0.248). The HFL/MIC ratio also did not differ significantly according to recorded relapse status (p=0.165). The HetR phenotype of the first available diagnostic isolate, defined operationally in this study as HFL/MIC ≥8, was frequent but was not significantly associated with recorded relapse within six months in this cohort.