CMET-15. WHOLE EXOME SEQUENCING OF BRAIN METASTASES FROM COLORECTAL PRIMARY CANCERS REVEALS CLINICALLY ACTIONABLE MUTATIONS Free

Current understanding of the underlying genetic evolution in metastatic colorectal cancer to the central nervous system is lacking. Further, there are no specific treatments for brain metastasis derived from primary gastrointestinal malignancies. Here, we report preliminary data using a next generation sequencing approach to characterize actionable genomic targets in matched colorectal primaries and their associated extra- and intracranial metastases.

A growing cohort of metastatic colorectal cancer is being assembled that consists of a primary tumor, at least a single extracranial metastasis, an intracranial metastasis, and a normal tissue control for each patient. Nucleic acid was extracted for use in whole exome sequencing from twelve samples. Somatic variants in the primary tissue were identified relative to the matched control sample and further compared relative to the metastatic samples. Initial analysis has focused on cancer genes that have been established to have clinical implications to identify variants with potential therapeutic value.

Complex variant signatures were found across the primary colorectal tumors. At least one clinically actionable variant was identified in each case. Alterations in AKT1, POLE, BRAF and GNAS were detected in the brain metastasis samples and not in the extracranial sites. Alterations in GNAS, ARID1A, RET, and FGFR2 occurred at a higher frequency in metastatic samples (extracranial and brain metastases) compared to primary. KRAS and PIK3CA status was concordant across all tumors while PTEN and BRAF status was variable. CONCLUSIONS: Clinically actionable mutations can be found in brain and extracranial metastases that are not detected in their respective clinically sampled colorectal primary tumor. This provides support for the development of combined targeted therapeutic strategies that may be more successful in the metastatic setting. Further investigation is required in larger cohorts to fully characterize the genetic landscape and potential drivers of brain metastasis from colorectal cancer.