Variable DNA mismatch repair-associated gene profiles in colorectal versus uterine cancers.

11610 Background: DNA mismatch repair (MMR) plays an important role in maintaining DNA synthesis fidelity in the genome. Mutation in MMR genes occurs in colorectal and uterine cancers and leads to increased mutation burden that is associated with response to immune checkpoint inhibitors. It is unknown if there is an MMR gene-specific mutation signature in MMR-deficient tumors, or whether mutations in MMR genes drive specific mutation patterns. Methods: The study cohort consisted of 1060 uterine cancer (UtCa), and 797 colorectal cancer (CRC) cases consecutively submitted to Caris Life Science for molecular profiling using multiple technologies, including next generation sequencing (NGS), immunohistochemistry (IHC), and in situ hybridization (ISH). Mutation, IHC-positive, and ISH-positive frequencies were compared using Fisher’s exact test (p-value < 0.05 considered significant). Results: In total, 1,857 tumors were examined. Of the 797 CRC cases, 115 (14.4%) had at least one mutation in MLH1, MSH2, or MSH6. Nineteen (19; 2.3%) of the CRC cases had mutations in multiple MMR related genes. Of the 1060 UtCa cases, 52 (4.9%) had at least one mutation in MLH1, MSH2, or MSH6. Twenty-two (22; 2.1%) of the UtCa cases had mutations in multiple MMR related genes. Colorectal cancers that were MLH1, MSH2, and MSH6 mutated enriched for rare, lineage specific co-mutations, including KRAS A146T (4/32 MLH1-mutated cases; 12.5%). Uterine cancers that were MLH1, MSH2, and MSH6 mutated also enriched for several co-mutations, including ARID1A (8/9 MLH1-mutated cases; 88.9%), a SWI-SNF chromatin remodeling complex family member. Further analyses revealed differences in PD-L1 positivity between MMR mutated CRCs versus UtCa (8/131; 6.1% versus 8/51; 15.7%). Tumor mutational load (defined as the total number of non-synonymous mutations per Mb sequenced) was 35 mutations per Mb in CRC and 51 mutations per Mb in UtCa. Conclusions: There are differences in mutation signatures between uterine and colorectal cancer, and possible additional molecular targets for combination with immune checkpoint therapies. Further analysis of MMR gene-specific differences in molecular profiles is ongoing and will be discussed.