Mapping incidence and survival of non-small-cell lung cancers stratified by mutation type in southern Appalachia: A single center retrospective study.

Abstract

e20568 Background: Non-small cell lung cancer (NSCLC) is often a mutation-driven disease with variability in incidence and survival depending on location as seen in population-based cohort studies. Our study aims to delineate these differences based on molecular testing for these mutations in our study cohort at East Tennessee State University - Quillen College of Medicine, which is based on patients from east Tennessee, western North Carolina, and southwestern Virginia. Methods: We obtained pathology results of 294 patient samples with advanced/metastatic NSCLC from our cancer center/health system database over a 5 year period. PDL-1 was tested in all patients. Fluorescent in-situ hybridization (FISH) testing for common pathogenic driver mutations such as EGFR, ALK, ROS1, RET, MET and PTEN were performed as is standard of care. Molecular sequencing was performed and included over 40 genetic mutations and 2 biomarkers, several of which today are considered targetable for advanced NSCLC, per standard of care. Incidence of different mutations was measured in our patient population. Overall survival was calculated with date of diagnosis available based on pathology report and date of patient expiration obtained from electronic medical records. Data analysis was performed to ascertain which mutations carried an increased risk of death. Results: The TP53 mutation was very common, seen in 85% of squamous cell carcinoma (SCC) histology and 55% of adenocarcinoma. Similarly, incidence of PTEN deletion by FISH was noted to be 52% in SCC and 31% in adenocarcinoma. EGFR mutations were rare, seen in only 3% of tested patients. We had an incidence of KRAS G12C mutations at 11% overall, but it was present in up to 18% of adenocarcinomas. BRAF was seen in 4.8% of all cases, 8% of adenocarcinomas and 2% of SCC. Since there were more than 2 predictors, Cox proportional-hazards regression model was fit. Due to either having all negative values or very few positive values, the variables CCND1, CDKN2B, ERBB2, ERBB4, FGFR, FGFR2, MET, MSI, NRAS, PDGFRA, RET, RS01, SCR were excluded from analysis. A total of 156 samples were used in final analysis. After model building, it was found that BRAF and KRAS G12C mutation were significant predictors of survival by hazard ratios for death. Conclusions: In our unique patient population, there was a greater incidence of deleterious TP53 mutations and PTEN deletions compared to the general North American advanced/metastatic NSCLC population. Although BRAF and KRAS G12C mutations were not at a greater incidence in comparison to the rest of North America, they were notably significant predictors of survival, conferring nearly 3.4 and 2.7 times increased risk of death in our study. [Table: see text]