A comprehensive model of genetic-features predicts outcome of personalized adjuvant treatment in resected EGFR-mutant stage II-IIIA NSCLC: Results from a phase III trial (CTONG 1104-ADJUVANT)

Abstract

Background: Results of the ADJUVANT trial established adjuvant gefitinib as an optimal choice for EGFR-mutated stage II-IIIA NSCLC patients. However, clinical benefit varied among patients. To investigate this heterogeneity, we performed comprehensive tumor genomic analyses on these patients. Here, we report the predictive MEDUSA model (Multiple-biomarker Evaluation to Determine the Utilization of Specific Adjuvant therapy) that can guide clinical decision of adjuvant therapy. Methods: 171 baseline specimens from ADJUVANT (n=95, gefitinib arm; n=76, vinorelbine plus cisplatin [VP] arm) underwent targeted sequencing (Geneseeq 422- gene panel). Predictive biomarkers were identified by Cox regression with gene-bytreatment interactions, and a multi-gene composite score was developed to compare the benefits of these treatments. Results: EGFR mutations were confirmed in all cases. TP53, NKX2-1, CDK4, MYC and RB1 were identified as predictive biomarkers. Specifically, gefitinib-favoring biomarkers include TP53 exon4/5 mutations (interaction HR [iHR] 0.33, 95% CI 0.12- 0.93, p=0.035), and copy number gain of NKX2-1 (iHR 0.26, 95% CI 0.098-0.68, p=0.006), CDK4 (iHR 0.14, 95% CI 0.025-0.77, p=0.024) and MYC (iHR 0.10, 95% CI 0.011-0.98, p=0.048). RB1 alterations strongly favored VP (iHR 4.07, 95% CI 1.56- 10.53, p=0.004). The MEDUSA model was developed based on the above, and stratified patients into 3 groups: Strong Gefitinib-favoring (SG, n=60), Moderate Gefitinib-favoring (MG, n=87), and VP-favoring (VP, n=24). Notably, the SG group demonstrated significant OS benefit with adjuvant gefitinib as well: HR of OS was 0.44 (95% CI 0.2-0.98, p=0.04). Conclusions: Incorporating alterations in TP53, NKX2-1, CDK4, MYC and RB1, MEDUSA score could guide personalized adjuvant therapy for resected stage II-IIIA EGFR-mutant NSCLC patients. (Table Presented) .