Afatinib (AFA) plus bevacizumab (BEV) combination after osimertinib (OSIME) failure for aDvanced EGFR-mutant non-small cell lung cancer (NSCLC): A multicenter prospective single arm phase II study (ABCD-study)

Abstract

Background: 3rd-generation EGFR-TKI, OSIME showed a remarkable efficacy for T790M+ NSCLC. Moreover, OSIME has proven superior progression-free survival (PFS) over 1st-generation EGFR-TKIs in front-line setting, and it will be widely used in front-line and later settings. Various resistant mechanisms were reported: C797S (cis/ trans); EGFR uncommon mutations such as E709K and L844V; T790M- conversion; HER2-amp; MET-amp; PIK3CA/BRAF mutations; and histological transformation such as small cell and EMT. AFA is an irreversible EGFR-TKI with a potency as pan- HER inhibitor. Preclinical and clinical studies showed high sensitivity to EGFR uncommon mutations. Preclinical studies also showed a synergistic effect of AFA and BEV after acquired resistance (AR) to EGFR-TKIs. Our previously-conducted phase II study found a promising efficacy of AFA+BEV after AR: response rate (RR), 18.8%; disease control rate (DCR), 90.7%; and median PFS (mPFS), 6.3 months. Notably, preclinical data has suggested EGFR-TKI+BEV combination could overcome MET-associated resistance. Based on the above background, we hypothesize several resistant mechanisms (e.g., C797S (trans), uncommon EGFR mutations, and HER2) to OSIME are sensitive to AFA, and synergism of additional BEV maximizes AFA sensitivity. Trial design: This ABCD-study is a phase II study conducted by HANSHIN Oncology Group. ECOG PS 0-1 patients with EGFR-mutant NSCLC are enrolled after failure of OSIME. Liquid/histological rebiopsies before enrollment after OSIME failure are essential to examine the resistant mechanisms. Liquid /histological samples are analyzed to confirm gene mutations/fusions and copy-number gain using next-generation sequencers. AFA is prescribed at 30-40 mg QD, and BEV is administered at 15 mg/kg tri-weekly until progression. The primary endpoint is PFS. Sample size of 26 is based on a null mPFS of 4.0 months, an alternative mPFS of 6.0 months, power of 80% and onesided 15% of type I error. We plan an exploratory analysis regarding differential efficacies according to each resistant mechanism.