killing: 25% of the strains were killed by 0.4 ,ug or less of amphotericin B alone per ml, and 75% of the strains were killed with the addition of 6.25 ,ug of rifampin per ml. In vivo studies are needed for confinnation of the usefulness of combined amphotericin B and rifampin therapy.Despite an interval of 21 years since its introduction, amphotericin B remains the most efficacious agent in the treatment of systemic mycoses (4). Numerous attempts to reduce the toxicity of amphotericin B have met with limited success. Currently, the most satisfactory ways to limit the undesirable side effects of the drug are to reduce the total quantity administered or to give amphotericin B over a prolonged time period (every-other-day dosage). One method to reduce the amount without compromising efficacy is to combine the drug with a compound which increases killing of organisms by combined action without adding toxicity. Tetracycline, 5-fluorocytosine, rifampin, actinomycin D, and mycophenolic acid glucuronide have all shown a propensity to increase the potency of amphotericin against various fungi and yeasts (8,9).Although Candida species are generally susceptible to relatively small quantities of amphotericin B, Candida endophthalmitis, endocarditis, meningitis, osteomyelitis, peritonitis, pancreatitis, arthritis, and systemic candidiasis are difficult to treat. Reports of all these Candida infections have increased impressively. This study was designed to determine the potential for rifampin to increase the potency of amphotericin B against multiple clinical isolates of Candida in vitro at clinically achievable levels of both drugs and to evaluate the microtiter system for testing synergism with these antifungal components. MATERIALS AND METHODSForty strains of Candida were tested. Twenty-eight strains were isolated from the laboratory of Harbor General Hospital; the remaining strains were from the American Type Culture Collection (Rockville, Md.).These were the same strains used in a previous study of the combined activity of 5-fluorocytosine and amphotericin B and were identified and maintained as previously described (11

The induction of neutropenia and immunosuppression by the administration of nitrogen mustard (HN2) decreased the frequency and altered the morphology of clinically detectable hematogenous Candida endophthalmitis in the rabbit model of disseminated candidiasis. Whereas 95% of eyes in rabbits infected with Candida albicans without pretreatment with HN2 developed typical lesions of hematogenous Candida endophthalmitis, only 6.2% of eyes in rabbits that had been given 3.0 mg of HN2 per kg developed clinically detectable endophthalmitis. Lesions that developed in the severely immunocompromised and neutropenic rabbits were small and atypical in appearance. From these data, we conclude that ophthalmoscopic examination may not be a sensitive diagnostic modality for disseminated candidiasis in severely immunocompromised, neutropenic patients.

Anterior chamber aspiration and vitreous aspiration were evaluated as diagnostic procedures for establishing a specific microbiological diagnosis in the rabbit model of hematogenous Candida endophthalmitis. Vitreous aspiration was the most successful procedure, confirming the diagnosis of hematogenous Candida endophthalmitis in 62% of eyes with documented intraocular infection. When animals with only the most severe clinical endophthalmitis were considered, vitreous aspiration confirmed the diagnosis in 89% of eyes evaluated. Vitreous aspiration correlated well with the extent of clinical endophthalmitis, as well as with postmortem, whole-vitreous cultures. Gram staining of the aspirate was additive to culture results in confirming the diagnosis. Anterior chamber aspiration was positive in only 1 of 58 eyes evaluated (1.7%). Additionally, muscle biospy was evaluated in this study as a tool for establishing a microbiological diagnosis in the rabbit model of disseminated candidiasis. Only 2 of 131 biospy specimens contained detectable Candida. Although vitreous aspiration may be associated with ocular complications, in certain clinical settings this procedure may be valuable in establishing the definitive microbiological diagnosis of hematogenous Candida endophthalmitis.

No abstract