Fungemias were reviewed in 110 immunocompromised patients hospitalized between November 1, 1974, and December 31, 1977, a Memorial Sloan-Kettering
Cancer Center (MSKCC). The incidence of Candida tropicalis
fungemia increased each year. Seventy-six percent of the patients with C. tropicalis
fungemia and 32.5 percent of those with C. albicans
fungemia had either
leukemia or
lymphoma. Seventy-seven percent of the C. parapsilosis fungemias were related to
total parenteral nutrition. Thirty-seven percent of the patients with C. albicans
fungemia were receiving oral prophylactic
nystatin therapy. The source of
fungemia was often difficult to determine: in 60 percent of the patients, only blood cultures were positive for C. tropicalis or Torulopsis glabrata; no cultures were positive for the fungus from any other site before the episode occurred. Serologic tests, including a highly sensitive passive hemagglutination test, showed fourfold increases in titer only inconsistently. A passive hemagglutination-inhibition test for circulating
antigen was positive in 50.9 percent of 57 patients with
fungemia who were tested and may be a valid indication for treatment.
Fungemia usually represented a severe and often fatal disease. The over-all mortality of the 110 patients with
fungemia was 79 percent whereas only 23 percent of the patients with C. parapsilosis
fungemia died. Among the patients who received more than 200 mg of
amphotericin B, 71 percent died despite treatment.