BARCELONA – Mutations in the tumor suppressor KEAP1 or STK11 genes in patients with metastatic non–small cell lung cancer (mNSCLC) in the randomized, phase 3 MYSTIC trial experienced poorer outcomes than did patients without the mutations, according to an exploratory analysis of trial data.
Mutations in ARID1a, however, were associated with improved overall survival (OS) among patients in the trial who were treated with durvalumab + tremelimumab.
Profiling of circulating tumor DNA from 943 evaluable baseline plasma specimens showed median OS of 7.4 vs. 12.9 months among 170 patients with KEAP1 mutations (m) vs. in 773 with KEAP1 wild type (hazard ratio, 1.64), and 6.8 vs. 12.6 months in patients with STK11m vs. 796 with STK11wt (HR, 1.52),, the Price Family Professor of Medicine, director of Thoracic Oncology, and codirector of Cancer Immunotherapy at Columbia University Irving Medical Center, New York, reported at the World Conference on Lung Cancer.
Objective response rates (ORR) in the groups, respectively, were 17.6% vs. 27.7% for KEAP1m vs. KEAP1wt, and 16.3% vs. 27.6% for STK11m vs. STK11wt, Dr. Rizvi said at the conference, sponsored by the International Association for the Study of Lung Cancer.
This was regardless oftreatment arm; the open-label, multicenter, global trial compared durvalumab monotherapy or durvalumab plus tremelimumab with platinum-based chemotherapy for the first-line treatment of patients with epidermal growth factor receptor and anaplastic lymphoma kinase wild-type, locally-advanced or metastatic NSCLC.
Mutations in the ARID1a gene, however, had no impact on OS (12.6 vs. 11.4 months in 114 vs. 829 patients with ARID1am vs. ARID1awt; HR, 0.94), and ORRs, respectively, were 35.1% vs. 24.6%.
When comparing outcomes by treatment arm, the ORRs with chemotherapy were 15.1% vs. 34% for KEAP1m vs. KEAP1wt, 12.2% vs. 33.6% for STK11m vs. STK1wt, and 28.1% vs. 31% for ARID1am vs. ARID1awt.
The ORRs in the durvalumab arm were 16.7% vs. 25.2% for KEAP1m vs. KEAP1st, 14.5% vs. 25.7% for STK11m vs. STK11wt and 25.6% vs. 23.4%, respectively, and in the durvalumab + tremelimumab arm they were 20.6% vs. 23.9% for KEAP1m vs. KEAP1wt, 21.6% vs. 23.6% for STK11m vs. STK11wt.
“The key finding here is really the ARID1a response,” Dr. Rizvi said, noting the “pretty impressive response rates” of 51.3% with ARID1am vs. 19.4% for ARID1awt.
The relationship between gene alterations and response to anti-programmed death-1 (PD-1) therapy with and without anti-CTLA-4 therapy is not well characterized. These findings, which suggest that KEAP1 and STK11 mutations are prognostic for OS in mNSCLC, and that ARID1am may be predictive of OS benefit in patients receiving durvalumab + tremelimumab, provide insights to the potential impact of specific mutations on response to immunotherapy, Dr. Rizvi said.
“STK11 and KEAP1 mutations ... are relatively common mutations – they are actually the third and fourth most common mutations in lung cancer after p53 and KRAS,” he said, adding that they influence outcomes and need to be factored in to analyses of outcomes in lung cancer. “ARID1am patients were about 10% of the population and they did particularly well with durvalumab and tremelimumab, and I think these exploratory analyses can help us think about how we use [tumor mutational burden] and outcomes among cancer patients in future trials.”
The MYSTIC trial was sponsored by AstraZeneca. Dr. Rizvi disclosed royalties related to intellectual property/patents filed by MSKCC and Personal Genome Diagnostics.
SOURCE: Rizvi N et al. WCLC 2019: Abstract