OBJECTIVES

To investigate (1) the frequency of clinical resistance to oral polyenes or azole treatment for oral candidiasis, (2) the frequency of resistant in vitro Candida strains, (3) the relationship between the susceptibilities of in vitro Candida species and in vivo status in HIV patients.

DESIGN

Prospective cross-sectional study.

SETTING

Tertiary care clinic at Bocage Hospital, Dijon, France.

PATIENTS

HIV-infected patients with and without oral candidiasis.

INTERVENTIONS

Clinical examination, oral swab for mycologic investigations.

MAIN OUTCOME MEASURES

Clinical diagnosis of oral candidiasis, identification of the antifungal treatment given within the previous month, identification of Candida species, antimycogramm and determination of the minimal inhibitory concentration (MIC) for fluconazole, and measurement of T-helper cell count.

RESULTS

Within a 2-month period, 154 HIV-infected patients were studied: 46 heterosexuals, 51 intravenous drug users (IVDU), 52 homosexuals and five blood transfusion recipients. The percentages of patients with oral candidiasis were: 41, 80, 44 and 20%, respectively (P < 0.05); the mean T-helper cell counts were 200, 135, 210 and 238 x 10(6)/l cells, respectively (P < 0.05). Twenty-two patients (14.3%) had received recent azole treatment and 29 (18.8%) recent oral polyene treatment. Among the 84 patients with and the 70 patients without oral candidiasis, 78 and 28 Candida strains were isolated, respectively. Although Candida albicans represented the majority of Candida species (88 strains, 83%), the non-albicans strains were isolated more frequently in patients who had received recent antifungal treatment. No strains were resistant to ketoconazole, miconazole or econazole; however, six (5.6%), 16 (15%) and 10 (9.5%) were intermediately susceptible to the three drugs, respectively. Twelve (13.6%) of the 88 C. albicans, five of the six C. (Torulopsis) glabrata, one of the five C. tropicalis and all three C. krusei strains were resistant to fluconazole. These resistant strains were separated as follows: 41.1% of C. albicans strains resistant to fluconazole were isolated from patients who had received recent azole therapy, 17.6% from patients who received recent oral polyene, and 3.7% from patients who had not received any recent antifungal treatment (P = 0.004). The mean MIC of these three categories of isolates were 3.6, 1.6 and 0.6 mg/l, respectively (P = 0.06).

CONCLUSIONS

Oral candidiasis and fluconazole-resistant Candida isolates are more frequently found in IVDU. Treatments using azoles select non-albicans strains and are associated with decreased susceptibilities of C. albicans strains to fluconazole in particular. These findings show that prolonged azole treatment in severely immunocompromised patients should be avoided.