Targetable mutations in gastrointestinal malignancies: A comparison of RAS mutant and RAS wild type tumors.

Abstract

129 Background: KRAS and NRAS (RAS) mutations are considered driver mutations in gastrointestinal malignancies such as colorectal and pancreatic cancer. Our institution obtains broad molecular testing (commercial panels with full exon coverage of at least 300 genes) for all stage IV gastrointestinal malignancies that are reviewed at an internal molecular tumor board (MTB). The MTB subjectively felt there was less benefit from comprehensive molecular testing in RAS mutant tumors and wished to quantify this using standardized analysis according to validated guidelines. Methods: We performed a retrospective cohort study of 209 consecutive genomic sequencing results of advanced gastrointestinal malignancies at our institution dating from March 2016 until December 2019. We compared the number of “targetable” mutations in the RAS mutant and wild type (WT) malignancies, as analyzed by an interpretation service according to the Association of Molecular Pathology (AMP) guidelines. A lower AMP score corresponds to a higher level of evidence as a predictive biomarker. We also compared molecular tumor board specific recommendations for each group (excluding recommendations regarding RAS such as EGFR mAbs). Results: There were 134 RAS mutant and 75 RAS WT cases. Alterations with AMP scores of 1A,1B, and 1C were more commonly seen in the RAS WT population 18/75 and 9/134 respectively (24.0% versus 6.7% p-value=0.0004). 39 of 75 cases in RAS WT and 27/134 in RAS mutant (52.0 versus 20.1%, p-value <0.0001) cohort had at least one alteration that was deemed actionable by our institution’s current MTB criteria. Conclusions: Actionable mutations were significantly less common in the RAS mutant versus RAS WT population, and further studies assessing the value of comprehensive genomic testing in RAS mutant gastrointestinal malignancies may be warranted. [Table: see text]