Antifungal prophylaxis in patients with haematological malignancies receiving chemotherapy or haematopoietic stem cell transplantation: a systematic review and network meta-analysis.
To compare systemic antifungal prophylaxis for proven/probable invasive fungal disease (IFD) in high-risk haematology populations using randomised controlled trials, with separate mortality, discontinuation, and formulation analyses.
Six databases and conference sources were searched to 2 September 2026. Bayesian random-effects network meta-analysis included RCTs with traceable proven/probable IFD definitions. Mixed-policy, setting, formulation, mortality, and discontinuation analyses were reported separately.
CRD420261490010 (retrospective).
Twenty-one RCTs (42 arms; 6,338 participants; 363 events) informed the primary network. Versus inactive control, voriconazole (OR 0.042, 95% CrI 0.001-0.793) and liposomal amphotericin B (OR 0.335, 95% CrI 0.127-0.819) had credible intervals excluding 1. Posaconazole suspension showed favourable but imprecise estimates. No mortality comparison demonstrated clear benefit (17 RCTs; 582 deaths). Exploratory peri-HSCT analyses suggested fewer IFDs but more discontinuations with itraconazole. In AML/MDS chemotherapy, posaconazole effects were comparator-dependent. Direct evidence did not distinguish delayed-release tablets from suspension.
No regimen was consistently superior across settings and outcomes. Voriconazole and liposomal amphotericin B had the clearest evidence versus inactive control. Posaconazole effects were comparator-dependent; peri-HSCT itraconazole showed fewer IFDs but more discontinuations. No mortality benefit was established. Prophylaxis should be risk-adapted to clinical setting, baseline mould risk, formulation, tolerability, and feasibility.
Six databases and conference sources were searched to 2 September 2026. Bayesian random-effects network meta-analysis included RCTs with traceable proven/probable IFD definitions. Mixed-policy, setting, formulation, mortality, and discontinuation analyses were reported separately.
CRD420261490010 (retrospective).
Twenty-one RCTs (42 arms; 6,338 participants; 363 events) informed the primary network. Versus inactive control, voriconazole (OR 0.042, 95% CrI 0.001-0.793) and liposomal amphotericin B (OR 0.335, 95% CrI 0.127-0.819) had credible intervals excluding 1. Posaconazole suspension showed favourable but imprecise estimates. No mortality comparison demonstrated clear benefit (17 RCTs; 582 deaths). Exploratory peri-HSCT analyses suggested fewer IFDs but more discontinuations with itraconazole. In AML/MDS chemotherapy, posaconazole effects were comparator-dependent. Direct evidence did not distinguish delayed-release tablets from suspension.
No regimen was consistently superior across settings and outcomes. Voriconazole and liposomal amphotericin B had the clearest evidence versus inactive control. Posaconazole effects were comparator-dependent; peri-HSCT itraconazole showed fewer IFDs but more discontinuations. No mortality benefit was established. Prophylaxis should be risk-adapted to clinical setting, baseline mould risk, formulation, tolerability, and feasibility.
Authors
Shao Shao, Li Li, Cui Cui, Feng Feng, Chen Chen, Huang Huang, Fan Fan, Pei Pei, Chen Chen, Feng Feng
View on Pubmed