Systemic antifungal prophylaxis after hematopoietic stem cell transplantation: A meta-analysis

Background Hematopoietic stem transplant recipients are subject to increased risk for invasive fungal infections. Objective This meta-analysis was undertaken to explore the comparative effectiveness of systemic antifungal prophylaxis in hematopoietic stem cell transplant recipients. Methods We searched PubMed and The Cochrane Register of Randomized Controlled Trials up to March 2013 for randomized studies on systemic antifungal prophylaxis after hematopoietic stem cell transplantation. We performed a meta-analysis on the relative effectiveness of systemic antifungal prophylaxis on proven or probable invasive fungal infections using direct and indirect effects. Relative effectiveness was reported as odds ratio (OR) for invasive fungal infections, causative agent, empirical antifungal therapy, and withdrawals due to drug adverse events. Results Twenty evaluable studies provided data on 4823 patients. The risk for invasive fungal infections while on prophylaxis was 5.1% (95% CI, 3.6-6.8%). In patients receiving fluconazole, risks of proven or probable invasive fungal infections (OR = 0.24; 95% CI, 0.11-0.50; number needed to treat [NNT] = 8), systemic candidiasis (OR = 0.11; 95% CI, 0.05-0.24; NNT = 7), and overall need for empiric antifungal treatment (OR = 0.60; 95% CI, 0.44-0.82; NNT = 8) were reduced compared with patients receiving placebo. Itraconazole was more effective than fluconazole for the prevention of aspergillosis (OR = 0.40; 95% CI, 0.19-0.83; NNT = 23) at the expense of more frequent withdrawals (OR = 3.01; 95% CI, 1.77-5.13; number needed to harm = 6). Micafungin was marginally more effective than fluconazole for the prevention of all mold infections (OR = 0.35; 95% CI, 0.10-1.18; NNT = 79) and invasive aspergillosis (OR = 0.19; 95% CI, 0.03-1.11; NNT = 78) and reducing the need for empiric antifungal treatment (OR = 0.40; 95% CI, 0.13-1.21; NNT = 8). There was a relative lack of comparisons between different antifungal prophylactic strategies, including the newer azoles, voriconazole and posaconazole, in this population. Direct effects derived from single studies showed marginally significant effects for voriconazole compared with fluconazole regarding invasive aspergillosis (OR = 0.50; 95% CI, 0.20-1.20; NNT = 35) and the need for empiric treatment (OR = 0.72; 95% CI, 0.50-1.06; NNT = 15). Voriconazole compared with itraconazole (OR = 0.59; 95% CI, 0.40-0.88; NNT = 8) and posaconazole compared with amphotericin B (OR = 0.28; 95% CI, 0.06-1.24, marginal significance; NNT = 3) were better regarding empirical antifungal treatment. Conclusions Even when on antifungal therapy, invasive fungal infection will develop in 1 of 20 patients undergoing hematopoietic stem cell transplantation. There is evidence for the comparable effectiveness of different antifungal drugs used for prophylaxis. Fluconazole is the most widely studied agent, but micafungin might prove to be more effective. There is a relative paucity of studies for the newer azoles, although both voriconazole and posaconazole give proof of their comparative or higher effectiveness to fluconazole in single randomized studies.