Abstract 3414: Determining microsatellite instability (MSI) status of colorectal cancers through next-generation sequencing (NGS)

Introduction: The MSI status of a tumor is often a marker of deoxyribonucleic acid (DNA) mismatch repair deficiency. Recent clinical trials have shown MSI-High (MSI-H) tumors are more likely to respond to checkpoint inhibitor immunotherapy. The United States Food and Drug Administration recently granted approval for a checkpoint inhibitor in metastatic solid tumor that demonstrates high microsatellite instability. Here we show the accuracy of determining the MSI status of colorectal cancers (CRC) using either TruSight™ Tumor 170 (TST170), a NGS-based 170 gene panel for solid tumor profiling, or whole exome sequencing (WES). Further, we report the accuracy of identifying the MSI status through sequencing of tumor-only samples without the subject-matched normal DNA. Experimental Method: The MSI status of 52 CRC tumors (51 formalin-fixed, paraffin embedded and 1 fresh-frozen sample) were evaluated using a microsatellite instability (MSI) assay (Promega), which is a detection method that uses capillary electrophoresis to analyze PCR-products. For comparison to WES, the tumors and subject-matched normal DNA libraries were prepared and enriched using TruSight™ Oncology reagents (Illumina). Targeted NGS libraries were prepared and enriched for the TST170-gene panel using the TST170 library kit (Illumina). NGS libraries were sequenced with the HiSeq™ 2000, HiSeq™ 2500, or NextSeq™ 550 instruments (Illumina). MSI scores of the 52 tumors were calculated through a novel internally developed bioinformatics algorithm. Data Summary: From the 52 subjects assessed through Promega's MSI detection kit, 28 and 24 were found to be MSI-H and microsatellite stable (MSS), respectively. These data were used as the benchmark for the MSI status of these subjects. Both the NGS-based tumor-normal data from WES and TST170 achieved a 100% concordance with PCR-based MSI detection. In addition, when only the sequencing data of the tumor samples were used without the subject-matched normal DNA, the Illumina MSI algorithm could identify the MSI status of the 52 subjects with 98% concordance with Promega's MSI detection kit. Conclusion: Collectively, this data indicates that NGS can be utilized for determining tumor MSI status using WES or the TST170 panel. In addition, a bioinformatics algorithm was developed that successfully categorized tumors according to their MSI status without requiring subject-matched normal DNA sequence data. For Research Use Only. Not for use in diagnostic procedu