Systematic review and mixed treatment comparison of randomized evidence for empirical, pre-emptive and directed treatment strategies for invasive mould disease.

Randomized controlled trials (RCTs) provide the most reliable estimates of the effects of treatments. However, not all treatments are compared in available RCTs, making comparison of treatments problematic. Mixed treatment comparisons (MTCs) can provide estimates of the comparative effects of treatments across a range of available therapeutic options. MTCs use networks of available direct comparisons to estimate differences in treatments that have not been estimated in trials via a common comparator. We conducted a systematic review and MTCs of comparative RCTs in haematological patients of anti-mould active agents used for the empirical treatment of febrile neutropenia (Analysis 1), and pre-emptive therapy (Analysis 2) of invasive mould diseases. In addition, we summarized the evidence available associated with the use of directed treatment strategies (Analysis 3). For empirical therapy, caspofungin proved superior to amphotericin B, liposomal amphotericin B, amphotericin B lipid complex and voriconazole in the outcome of survival, but no agents showed superiority for treatment response. There was no evidence of a difference between pre-emptive and empirical strategies on mortality outcomes. For directed therapy, voriconazole was superior to amphotericin B for overall survival, and both voriconazole and liposomal amphotericin B were superior to amphotericin B and amphotericin B colloidal dispersion on the outcome of response. While limited to some degree by the availability of RCTs, the MTCs reported here provide the best available evidence of relative therapeutic success for different available treatment strategies.