Abstract
In PANTHER-IPF (Prednisone, Azathioprine, and N-Acetylcysteine: A Study That Evaluates Response in Idiopathic Pulmonary Fibrosis) (May 24 issue),1 Raghu et al. report increases in the risks of death and hospitalization for patients with idiopathic pulmonary fibrosis who were treated with a widely used triple-therapy regimen of prednisone, azathioprine, and N-acetylcysteine, as compared with placebo. However, the mean follow-up of only 32 weeks and the relatively low number of deaths may limit the robustness of the findings. By way of comparison, the Japanese study of pirfenidone by Azuma et al.2 was stopped early on the basis of five acute exacerbations in the placebo group, as compared with no exacerbations in the pirfenidone group, an important finding that was not reproduced in subsequent large clinical trials.3,4 In PANTHER-IPF, the extrapolated annual rate of death of 2.0% in the placebo group and 19.8% in the combination-therapy group are at the absolute extremes of the ranges observed so far.5,6 There was no significant between-group difference in the time to death or disease progression, which suggests that disease progression was more frequent in the placebo group. This appears to indicate an inconsistency of findings, potentially due to the handling of missing data, differences in rates of death, or differential dropout rates. Can the authors explain this apparent paradox?

This publication has 10 references indexed in Scilit: