[The incidence, etiology and clinical significance of visceral mycoses in patients with AIDS].

The incidence, aetiology and clinical significance of visceral mycoses in HIV-infected subjects were evaluated by a retrospective survey of the clinical and microbiological records of 237 consecutive AIDS patients followed-up since 1984. Seventy-four patients out of 237 (31.2%) (56 males, 18 females; 55 IV drug abusers, 7 heterosexuals, 6 homobisexuals, 3 blood recipients and 3 children with congenitally-acquired HIV infection) presented 77 different episodes of visceral fungal infection as a whole, represented by candidiasis in 56 cases (oesophageal 45, pulmonary 5, sepsis 2, eye involvement 2, endocarditis and invasive oropharyngeal infection in the remaining 2 patients), cryptococcosis in 17 cases (meningoencephalitis in all subjects, with disseminated infection in 11 of them), and aspergillosis in 4 cases (pulmonary 2, cerebral and cranio-facial in the remaining 2 patients). In 57 out of 74 patients (77%), visceral mycoses were diagnostic or concurrent with the diagnosis of AIDS. Fungal diseases, as a whole, showed a significantly higher incidence (p < 0.03) among drug abusers, whereas homobisexual men presented a significantly lower frequency (p < 0.001, chi-square test) than AIDS patients with other risk factors for HIV infection. The onset of cryptococcosis was significantly associated with the male sex (p < 0.005, Fisher exact test). All subjects suffering from a visceral mycosis were severely immunosuppressed, with a higher rate of neutropenia in patients developing Candida and Aspergillus spp. infection (23 out of 56 patients with visceral candidiasis and 3 out of 4 cases of aspergillosis had an absolute neutrophil count lower than 1500 cells/mm3), while a severe reduction in CD4+ lymphocyte count was more evident among patients with cryptococcosis (13 out of 17 patients had a CD4+ cell count lower than 50/mm3). After remission of the primary episode of fungal infection (obtained in 80.5% of cases), the incidence of relapse observed in a long follow-up period (mean time 57.6 +/- 39.2 weeks) was elevated both for patients with cryptococcosis (7 cases out of 17) and subjects with candidiasis (19 cases out of 53), with no significant difference among patients receiving a secondary prophylaxis or not (22 relapses observed in 53 patients treated with maintenance antifungals versus 4 episodes in 8 patients followed for a comparable mean time with no antimycotic treatment). Fifty-two out of 74 patients (70.3%) have died up to now; in 21 of them death was due to or associated with the visceral mycosis (cryptococcosis in 11 cases, candidiasis in 8, aspergillosis in 2).(ABSTRACT TRUNCATED AT 400 WORDS)