Nystatin

Despite intensive infection prevention programs, the clinical course of patients with granulocytopenia continues to be complicated by fever and infection. The infections are most often caused by bacteria, especially aerobic gram-negative bacilli, normally dwelling in the alimentary tract; anaerobic bacteria rarely cause infections in the immunocompromised leukemic patients. Decontamination of the alimentary tract with agents that are selectively directed against aerobic gramnegative bacilli and do not affect the anaerobic flora, therefore, may prevent infections in these patients. Several studies have shown the efficacy of trimethoprim-sulfamethoxazole, gentamicin, nalidixic acid and colistin for the prevention of infections Abbreviations used in this paper: PPA, pipemidic acid; CLM, colistin sodium methanesulfonate; NYS, nystatin; ara-C, cytarabine; HDARA-C, high dose ara-C; DNR, daunorubicin; DHAD, mitoxantrone; ACR, aclarubicin; ANLL, acute nonlymphocytic leukemia; ALL, acute lymphocytic leukemia; GVN, gentamicin, vancomycin and nystatin.

Adequate selection of the eligible study population is challenging since these high-risk patients are difficult to identify, and the primary end-poin of prophylaxis should be carefully determined. Since the prevalence of candidemia and severe fungal infections in ICU patients approximates 1% and 2%, respectively, a large sample size would be required to demonstrate the efficacy of antifungal prophylaxis to reduce these rates. Such trial would therefore be impractical to conduct from both economic and clinical points of view and, more importantly, would not be clinically relevant. In addition, far more patients become colonized than infected during the ICU stay, and colonization by Candida spp. Constitutes an independent risk factor for invasive candidiasis. Finally, preemptive antifungal therapy has been recently proposed in heavily colonized, nonimmunocompromised ICU patients with well established risk factors for invasive candidiasis when their general condition abruptly deteriorates. In this subset of unstable, critically ill patients, the risk of invasive candidiasis is so high that the potential benefit of immediate antifungal treatment initiated before culture results outweighs its potential side effects. Such therapeutic strategy has recently been shown to be cost-effective in a highly selected target population that may represent approx. 7% of all ICU patients. Accordingly, we believe that future trials aimed at evaluating the usefulness of nystatin prophylaxis should include the use of preemptive therapy with predefined indications as a relevant end-point in double-blind placebo-controlled studies. We hypothesize that early nystatin prophylaxis administered in nonimmunocompromised ventilated ICU patients at high risk of invasive candidiasis would reduce the progression of colonization by Candida spp.