442 pts underwent 4 cycles of induction chemo w/ gemcitabine +/- erlotinib f/b continued gem or RT+cape 800 mg bid. RT involved 54 Gy/30 fx w/o prophylactic nodal irradiation. Erlotinib had trend to poorer survival (HR 1.19, p = 0.09). Only ∼1/2 of pts underwent 2nd randomization. MS did not differ b/t the CRT (15.2 mos) vs. chemo (16.5 mos) groups (NSS). CRT associated with decreased with LRF (32% vs. 46%, p = 0.03) but increased DM (60% vs. 44%, p = 0.04). Toxicities were similar (except worse nausea with CRT), and there was increased time off therapy for the CRT group (6.1 vs. 3.7 mos, p = 0.02) (HammelP et al., JAMA 2016). Criticism: use of single-agent chemo in the induction arm. Of note, QA revealed 68% RT plans had minor or major deviations, and 9% of pts randomized to CRT never rcvd RT.