OBJECTIVES: The study was designed to evaluate co-occurrence patterns of genes known to play a role in the development of non-small cell lung cancer (NSCLC).

METHODS: The Flatiron Clinico-Genomic Database (CGDB) is a linked data source including electronic medical record and genomics data from Foundation Medicine, Inc. Patients with advanced or metastatic NSCLC were eligible for this study if they received anti-cancer therapy within 180 days after diagnosis. The frequencies and patterns of 14 of the more common genes known to have activating alterations in NSCLC (ALK, BRAF, EGFR, ERBB2, FGFR1, FGFR2, HRAS, KRAS, MAP2k1, MET, NF1, NRAS, RET and ROS1) were analyzed using pie chart and Spearman Rank correlation matrix.

RESULTS: A two-layer oncogene pie chart was created based on the percentage of patients who had at least one mutation among all 5807 eligible patents. KRAS (29.7%) and EGFR (17.9%) were the most common, followed by NF1 (6.5%), BRAF( 5.4%), ERBB2 (4.8%), MET (4.3%), ALK (3.8%), FGFR1 (3.2%), ROS1 (1.2%), MAP2K1 (1.0%), NRAS (0.9%), RET (0.8%), FGFR2 (0.6%) and HRAS (0.4%).The co-occurrence pattern between KRAS/EGFR positive and other oncogenes was graphically presented in the second layer of the pie chart. A paired correlation matrix plot showed that no strong positive correlations were observed across above genes, except a medium negative one between KRAS and EGFR (-0.229, p<0.0001).

CONCLUSIONS: There is low co-occurrence of these 14 genes among patients with advanced NSCLC. Graphic presentation of genetic alterations may help to visually communicate tumor biology.