Tivantinib plus erlotinib versus placebo plus erlotinib in Asian patients with previously treated nonsquamous NSCLC with wild-type EGFR: First report of a phase III ATTENTION trial.

8044 Background: A randomized phase II study suggested that adding a c-Met inhibitor tivantinib (T) to an EGFR-TKI erlotinib (E) may prolong PFS in non-squamous NSCLC patients (pts) with 1-2 prior chemotherapy. Furthermore, the subset analysis suggested that survival benefit was more prominent in pts with wild-type EGFR (WT-EGFR) than in those with EGFR mutation. Therefore we conducted the phase III ATTENTION trial to compare overall survival (OS) between T+E vs. placebo (P)+E, in Asian pts with pretreated NSCLC with WT-EGFR. Methods: Four-hundred-and sixty Asian (Japan, Korea, and Taiwan) pts who had been treated with 1-2 chemotherapy for non-squamous NSCLC with WT-EGFR were planned to be randomized between T+E and P+E. The doses of T were 360mg BID and 240mg BID, for CYP2C19 extensive metabolizers (EM) and poor metabolizers (PM), respectively, whereas that of E was 150 mg QD for both EM and PM. The primary endpoint was OS, and secondary endpoints included PFS, ORR, and safety. Exploratory analysis tested biomarkers including c-Met expression and KRAS mutation. Results: New enrollment was stopped when 307 pts had been randomized, following the Safety Review Committee’s recommendation based on an observed imbalance in interstitial lung disease (ILD) between the groups. Accordingly,153 and 154 pts were treated with T+E and P+E, respectively. Patient backgrounds including CYP2C19 genotype were well balanced. Results are summarized below. Typical ≥Gr.3 adverse events in T+E were anemia (13.2 %), neutropenia (10.5 %), and leukopenia (6.6 %). Fourteen pts (including 3 deaths) and 6 pts (0 deaths) developed ILD in T+E and P+E, respectively. Conclusions: Although this trial lacked a statistical power due to the prematurely termination owing to toxicity concern, our results indicated some sign for benefit for adding T to E in this population. OS was numerically prolonged but this did not reach statistical significance. Exploratory biomarker analysis may identify subsets in which benefit and risk from this treatment is well balanced. Clinical trial information: NCT01377376. Median OS Median PFS ORR T+E 12.9 mo 2.9 mo 8.4% P+E 11.2 mo 2.0 mo 6.5% HR 0.891 0.719 P value 0.427 0.019