Analysis of outcomes in resected early-stage non-small cell lung cancer (NSCLC) with rare targetable driver mutations.

8052 Background: As postoperative adjuvant NSCLC treatment has evolved with EGFR and ALK targeted therapies, consideration may be given to treating other NSCLC with targetable mutations. Thus, having baseline outcomes for early-stage NSCLC with these targetable mutations is crucial, given their rarity. This study reports on recurrence-free survival (RFS) and overall survival (OS) in patients with resected NSCLC with treatable rare driver mutations. Methods: A retrospective single centre study identified stage I-III NSCLC patients with rare targetable mutations who underwent curative surgery. Tissue based next-generation sequencing identified mutations in KRAS G12C, EGFR Exon20, ERBB2, ALK, ROS1, BRAFV600E, MET exon14 skipping, RET. Baseline characteristics, adjuvant chemotherapy, mutation subtype, and TP53 co-mutation were correlated with RFS and OS using Cox regression. The KRAS G12C cohort was used as the reference for survival comparisons. Results: Among 201 patients (mean age: 66.4, 63% female) 61% had stage I, 19% stage II, 20% stage III. Predominant histology was adenocarcinoma (95%) and lobectomy (77%) was the most common surgery. Adjuvant chemotherapy was given to 37% (median 4 cycles). Mutations identified included: KRAS G12C (87, 43%), EGFR Exon 20 (27, 13%), ERBB2 (23, 11%), MET (20, 10%), ALK (14, 7%), ROS1 (13, 6%), BRAF (10, 5%) and RET (5, 2%). For all patients, 5-year survival probabilities were 75% stage I, 56% stage II (Hazard ratio HR: 2.17, p=0.038), 55% for stage III (HR: 2.38, p=0.015). Stage was also a significant predictor of RFS: stage II (HR: 1.90, p=0.04), stage III (HR: 2.26, p=0.006) vs stage I. TP53 co-mutation was associated with poorer OS (stage-adjusted HR (aHR): 2.50, p=0.004) and RFS (aHR: 1.70, p=0.037). All patients with rare mutations had shorter RFS compared to those with KRAS G12C mutation (Table). The difference for ERBB2 was significant (aHR 2.26, p=0.014), with only 37% of patients relapse free at 5 years. In contrast, except for ERBB2, other mutations were associated with better OS, with fusion mutations having the greatest difference (aHR 0.24, p=0.021). ERBB2 mutation had the highest cumulative incidence of brain metastasis, 29% at 5 years. Conclusions: Despite consistently poorer RFS, our study shows that, with the exception of ERBB2, OS of all other rare mutations was superior to KRAS G12C mutated NSCLC. TP53 co-mutation was demonstrated to be prognostic of poorer outcome. These dichotomous results may be explained by the use of targeted treatments at relapse, and suggest a potential role of targeted agents in the adjuvant setting. Table: see text