To the Editor: In the valuable study by Powderly et al. (March 19 issue)1 comparing fluconazole with amphotericin B as suppressive therapy for cryptococcal meningitis in patients with the acquired immunodeficiency syndrome (AIDS), the authors state that relapse was not predicted by routine sampling of the cerebrospinal fluid in any of the patients with meningitis. It would be of great interest to know whether their frequent monitoring of serum cryptococcal-antigen titers helped predict relapses of cryptococcal disease during suppressive therapy for cryptococcal meningitis. Persistence of Cryptococcus neoformans in the prostate gland, a source of recurrent infection, is common after amphotericin… © 1992, Massachusetts Medical Society. All rights reserved.