Reply to Y. Yu et al

Abstract

We appreciate the comments from Yu et al 1 regarding the statistical analysis and methodology used in our study. 2 Summarily, although they agree with our approach of using meta-pooled prevalence of acquired genomic alterations (Acq-GAs) as historical controls for later-line anti-EGFR-antibody (Ab) therapy for comparison with first-line anti-EGFR-Ab therapy data generated from CALGB/SWOG-80405 trial, they feel that a Bayesian approach would be a better metaanalytical approach, since it can stabilize the variance of the study-specific prevalence and likely to perform better in setting of sparse data.