Circulating tumor DNA (ctDNA) is increasingly being investigated as a tool to detect minimal residual disease in resected, stage I-III colorectal cancer. Recent ctDNA studies have indicated that detection of ctDNA following surgery for resectable colorectal cancer confers a significantly higher risk of recurrence than those with negative ctDNA postoperatively. In those with postoperative ctDNA positivity, clearance of minimal residual disease with adjuvant chemotherapy is a positive prognostic indicator. Lastly, ctDNA has demonstrated superior sensitivity to the conventional blood tumor marker carcinoembryonic antigen (CEA) and can offer median lead times of up to 11 months for radiographic detection of recurrence during the surveillance of resected, stage I-III colorectal cancer. In metastatic colorectal cancer (mCRC), there is growing evidence to suggest that plasma ctDNA can be used to monitor tumor response to conventional chemotherapy as well. The present case series demonstrated that plasma ctDNA is a predictor of tumor response to immunotherapy in patients with mCRC that are microsatellite stable or microsatellite instability high. Plasma ctDNA could serve as a dynamic marker of immunotherapy response even in colorectal tumors that were CEA non-producers. Overall, these findings add to ongoing efforts to establish the role of plasma ctDNA in monitoring response to immunotherapy in CRC.

No abstract

Lynch syndrome (LS), caused by heterozygous pathogenic variants affecting one of the mismatch repair (MMR) genes (MSH2, MLH1, MSH6, PMS2), confers moderate to high risks for colorectal, endometrial, and other cancers. We describe a four-generation, 13-branched pedigree in which multiple LS branches carry the MSH2 pathogenic variant c.2006G>T (p.Gly669Val), one branch has this and an additional novel MSH6 variant c.3936_4001+8dup (intronic), and other non-LS branches carry variants within other cancer-relevant genes (NBN, MC1R, PTPRJ). Both MSH2 c.2006G>T and MSH6 c.3936_4001+8dup caused aberrant RNA splicing in carriers, including out-of-frame exon-skipping, providing functional evidence of their pathogenicity. MSH2 and MSH6 are co-located on Chr2p21, but the two variants segregated independently (mapped in trans) within the digenic branch, with carriers of either or both variants. Thus, MSH2 c.2006G>T and MSH6 c.3936_4001+8dup independently confer LS with differing cancer risks among family members in the same branch. Carriers of both variants have near 100% risk of transmitting either one to offspring. Nevertheless, a female carrier of both variants did not transmit either to one son, due to a germline recombination within the intervening region. Genetic diagnosis, risk stratification, and counseling for cancer and inheritance were highly individualized in this family. The finding of multiple cancer-associated variants in this pedigree illustrates a need to consider offering multicancer gene panel testing, as opposed to targeted cascade testing, as additional cancer variants may be uncovered in relatives.

10518 Background: Mismatch repair deficiency (MMRd) occurs in ~15% of colorectal (CRC) and ~30% of endometrial (EC) cancers, of which ~20% are caused by Lynch syndrome (LS). Universal screening of all incident CRC and EC for MMRd uses MLH1 methylation (in tumors with MLH1-loss by immunohistochemistry) to omit common sporadic cases from follow-up germline testing for LS. However, this overlooks unusual cases with constitutional MLH1 methylation (epimutation), a poorly-recognized high-risk mechanism for LS-type cancers that are MLH1 methylated. We aimed to determine the role and prevalence of constitutional MLH1 methylation among CRC and EC cases with tumor MLH1 methylation from clinical and population-based series and if age limits could be used to triage cases warranting additional germline methylation testing. Methods: We screened blood DNA for constitutional MLH1 methylation by pyrosequencing and methylation-specific qPCR in patients with MLH1-loss, MLH1 methylated CRC and EC ascertained from: i) cancer clinics, referred by oncologists and genetic counsellors; ii) Ohio population-based “Columbus” and “OCCPI” cohorts. Bisulfite sequencing confirmed constitutional MLH1 methylation. Results: In clinical cases, constitutional MLH1 methylation was identified in 3/7 EC and 3/8 CRC cases < 60 years. In population-based cohorts, constitutional MLH1 methylation was identified in 4/95 (4.2%) and 4/281 (1.4%) CRC cases of all ages from Columbus and OCCPI, respectively. Highest rates of detection with the majority of cases detected were at age ≤55 years, with 3/4 (75%) in Columbus and 4/17 (23.5%) in OCCPI ≤55 years. For EC, constitutional MLH1 methylation was identified in 0/68 Columbus cases of all ages and 1/24 OCCPI cases < 60 years (age-limited testing). She was one of six (~17%) cases < 50 years in the combined cohorts. EC was the first/dual-first presenting “sentinel” cancer in three female patients with constitutional MLH1 methylation. Of the 14 cases with constitutional MLH1 methylation, nine had hemiallelic methylation (~50% alleles methylated, affecting a single parental allele) and five had mosaicism across various normal tissues. Somatic “second-hits” affecting the unmethylated allele were found in the tumors of mosaic cases, demonstrating causation. Conclusions: A correct diagnosis of constitutional MLH1 methylation at first cancer presentation is important as it will significantly alter clinical management of these high-risk patients. Although rare overall, high rates of detection were found ≤55 years for CRC and ≤50 years for EC cases with a MMRd, MLH1 methylated tumor, entailing minimal extra screening. Testing for constitutional MLH1 methylation is warranted in patients with young-onset CRC or EC, or syn/metachronous cancers at any age, with MMRd and MLH1 methylated tumor(s).

181 Background: KRAS mutations have been widely characterized as markers of poor prognosis in CRC. In stage IV CRC, KRAS mutations are predictive of benefit to anti-EGFR therapy. ctDNA has increasingly been recognized as a prognostic biomarker in CRC as well. We evaluated the association between plasma metabolites and KRAS mutation or ctDNA status in a longitudinal, observational cohort of patients with stage I-IV CRC. Methods: This was a retrospective analysis of prospectively collected blood samples from a single-institute cohort of patients with stage I-IV CRC. All blood samples were collected at pre-chemotherapy baseline. A modified Epi proColon 2.0 CE (Epigenomics AG) assay was used for plasma ctDNA testing on the methylated SEPTIN9 gene (mSEPT9). ctDNA positivity was defined as a mSEPT9 percentage of methylation reference (PMR) value greater than zero. Up to 150 metabolites of central carbon metabolism were analyzed by mass spectrometry and high-performance liquid chromatography. Analytes were compared by relative area under the curve (AUC) and differences evaluated by ANOVA. The mean AUC was used in patients with metabolites measured from > 1 timepoint of collection. Patients were stratified by ctDNA status (positive or negative) and KRAS mutant (MT) or wildtype (WT) status. Results: A total of 32 patients were included with median age 65 years (range 20-90). The majority were female (53%) and had stage IV disease (78%). Of 25 patients with stage IV CRC, 88% had pre-chemotherapy blood samples collected in the first-line setting. Most patients were KRAS MT (44%) compared to KRAS WT (37%) or unknown KRAS status (19%). The most common KRAS MT subtypes were G12D (29%), G12V (29%), G13D (21%), and G12C (14%). The mean overall survival in this cohort was 27.4 months while the mean mSEPT9 PMR value was 2553.6. When stratified by ctDNA status, ctDNA positivity was associated with decreased levels of essential amino acids (lysine, methionine, threonine) and the non-essential amino acid arginine (all p < 0.05). Compared to KRAS WT tumors, KRAS MT tumors were associated with increased levels of proline, phenylalanine, and intermediates of glycolysis (lactate), MTA cycle (SAM, 5-Methioadenosine), and O-GlcNAcylation (GlcNac, all p < 0.05). Conclusions: We are the first to demonstrate the feasibility of associating central carbon metabolites with ctDNA and KRAS mutation status. As ctDNA positivity and KRAS MT status have evolving prognostic potential in CRC, associated metabolic signatures may identify metabolic pathways for novel biomarker development. Our findings also show that KRAS MT CRC appears to be enriched in intermediates of glycolytic, methyl donor, and O-GlcNAcylation pathways.