Purpose: To study the epidemiology of oral candidiasis and the effect of treatment of thrush in human immunodeficiency virus (HIV)-infected patients.
Patients and methods: We conducted a prospective observational study of 92 patients over 1 year, including a nonblinded, randomized treatment trial of thrush with clotrimazole troches or oral fluconazole. Oral sites were cultured monthly and when thrush occurred. Candida albicans strains were typed by contour-clamped homogeneous electric field (CHEF) electrophoresis. Changes in strains were evaluated over time and in regard to their associations with particular sites, episodes of thrush, relapse after treatment, and colonization of sexual partners. Susceptibility to fluconazole was tested and CHEF analysis was done on these strains to determine the epidemiology of fluconazole resistance.
Results: Yeasts colonized 84% of patients. C albicans accounted for 81% of all isolates and was separated into 34 distinct strains. Most patients had persistent carriage of 1 or 2 dominant strains of C albicans. Three couples shared strains. Nineteen different C albicans strains caused 82 episodes of thrush in 45 patients. CD4 < 200/microL was associated with development of thrush. Clinical cure rates were similar with fluconazole (96%) and clotrimazole (91%), but mycologic cure was better with fluconazole (49%) than clotrimazole (27%). Following mycologic cure, colonization recurred with the same strain 74% of the time. Colonization with Torulopsis glabrata and Saccharomyces cerevisiae increased after treatment with either drug, but these organisms were never a sole cause of thrush. In a subset of 35 patients followed for over 3 months in whom fluconazole susceptibilities were performed, minimum inhibitory concentrations (MICs) to fluconazole increased only in those on fluconazole prophylaxis. Clinical failure of fluconazole was associated with an MIC > or = 64 micrograms/mL in 3 patients, and with an MIC of 8 micrograms/mL in 1 patient. In 2 of these 4 patients, the prior colonizing strain developed fluconazole resistance. In the other 2, new resistant strains were acquired.
Conclusions: Many different strains of C albicans colonize and cause thrush in patients infected with HIV. Patients are usually persistently colonized with a single strain, and recurrences following treatment are usually due to the same strain. Transmission of strains may occur between couples. Fluconazole and clotrimazole are equally effective in treating thrush, but mycologic cure occurs more often with fluconazole. Fluconazole resistance in C albicans occurs most often in patients who have low CD4 counts and are taking fluconazole prophylactically for recurrent thrush. Fluconazole resistance may occur through acquisition of a new resistant strain or by development of resistance in a previously susceptible strain.