In our multicenter trial [1] involving 257 patients, we did not observe a significant increase in the incidence of colonization or infection by either C. krusei or T. glabrata in patients with acute leukemia who were receiving prophylactic fluconazole (400 mg/d). Similar results were found in the randomized, placebo-controlled trial of prophylactic fluconazole in 357 patients undergoing bone marrow transplantation [2]. The single case report and two small clinical trials of low-dose prophylactic fluconazole (50 or 200 mg/d) referred to by Drs. Akiyama, Sakamaki, and Onozawa also did not prove that prophylactic fluconazole enhances the risk for infection or colonization by these organisms [3-5]. Nevertheless, in vitro susceptibility tests indicate that C. krusei and T. glabrata isolates are frequently resistant to fluconazole. Thus, we agree that prophylactic fluconazole cannot be relied on for preventing infection by either C. krusei or T. glabrata. Single cases or uncontrolled, retrospective studies of patients receiving low-dose prophylactic fluconazole (200 mg/d) such as that of Akiyama and coworkers cannot tell us to what degree fluconazole prophylaxis increases the risk for colonization or infection. Well-controlled clinical trials of large numbers of patients given adequate doses of prophylactic fluconazole are necessary.

We disagree with the analysis by Drs. Pavia and Riley. We neither expressed enthusiasm nor made recommendations regarding the use of prophylactic fluconazole or empiric amphotericin B. Instead, we reported certain benefits (decreased colonization and infection by Candida species other than C. krusei, good safety profile, and no observed increase in infection or colonization by resistant isolates) and limitations (no reduction in amphotericin use or overall mortality) of prophylactic fluconazole in patients with acute leukemia. We also found a lower probability of invasive fungal infection in fluconazole recipients compared with placebo recipients among patients not receiving empiric amphotericin B (P = 0.03; see our Figure 3 RF 1 *). We encourage physicians to carefully assess all our results in making decisions about whether to use prophylactic fluconazole.