Randomized Phase 2 Study of Tivantinib (ARQ 197) Plus Erlotinib Versus Single-Agent Chemotherapy in Previously Treated KRAS Mutant Advanced Non-Small Cell Lung Cancer (NSCLC): Metastatic Non-Small Cell Lung Cancer

Abstract

genotype, patient demographics, and survival from diagnosis were evaluated by multivariate Cox regression. Results: Amongst brain metastases, KRAS mutations were identified in 43% (49/113, 95% CI 35%-53%) adenocarcinomas (ADC), 15% (2/13, 95% CI 4%-42%) large cell carcinomas (LCC) and 0/13 squamous cell carcinomas (SQC)/large cell neuroendocrine carcinomas (LNEC). EGFR mutations were identified in 11% ADC (12/114, 95% CI 6%-18%) and 0/26 LCC/SQC/LNEC. MET amplification was identified in 16% (30/191, 95% CI 11%-22%), occurring in 19% ADC (23/120), 20% LCC (4/20), 8% SQC (2/25) and 4% LNEC (1/26). Co-occurrence of KRAS mutations and MET amplification was observed in 12% (12/103, 95% CI 7%-19%) of ADC tested for both. In 17 paired primary lung and brain metastases tested for KRAS, 6 paired samples had mutations in both sites; 2 had KRASmutations in brain metastasis only. In 38 paired primary lung and brain metastases tested for MET, amplification was identified in 3 brain metastases not identified in the primary lung tumor; 35 paired samples did not have MET amplification in either site. In ADCs, KRAS mutations in brain metastases were more common than in liver (43% vs 9%, mid-PZ.04), and EGFR mutations were more common in liver than in brain (45% vs 11%, midPZ.004).MET amplification across histologies was more common in brain compared with liver (16% vs 0, mid-PZ.02). No correlation was found between genotype and gender, age, or smoking status in brain metastases patients. Controlling for extracranial disease, smoking, and age at diagnosis, patients with ADC/LCC brain metastases with co-occurrence of KRAS mutations and MET amplification had inferior OS compared with patients without co-occurrence of these alterations (HR 3.6, PZ.01). Conclusions: The high rate of MET amplification and co-occurrence with KRAS mutations in NSCLC brain metastases implicate MET as a therapeutic target in NSCLC brain metastases. Author Disclosure: L.C. Villaruz: None. M.A. Socinski: None. V. Derrick: None. S.A. Bartoluzzi: None. B. Kurland: None. M. Nikiforova: None. T.F. Burns: None. J. Siegfried: None. R. Hamilton: None. S. Dacic: None.