Abstract 3391: Methylation analysis using TriMeth indicate that asymptomatic colorectal cancers (CRCs) release less circulating tumor DNA (ctDNA) compared to symptomatic cancers

• Published in: Cancer research (Chicago, Ill.) Ill.), 2022-06, Vol.82 (12_Supplement), p.3391-3391
• Main Authors: Jensen, Sarah Østrup, Øgaard, Nadia, Ørntoft, Mai-britt Worm, Rasmussen, Mads Heilskov, Bramsen, Jesper Bertram, Therkildsen, Christina, Andersen, Claus Lindbjerg
• Format: Article
• Language: English
• ISSN: 1538-7445; 1538-7445
• DOI: 10.1158/1538-7445.AM2022-3391

Background: Early detection of cancer is key to identify tumours at a time when patients can be offered curatively intended treatments and outcomes are superior. Previous studies have presented sensitive methods for detection of ctDNA in early-stage cancer, but very few have been validated in asymptomatic individuals e.g. a true screening setting. Furthermore, many methods rely on expensive sequencing-based approaches that may not be cost-effective for population-based screening. Recently, we developed TriMeth, a simple CRC-specific ctDNA detection test based on three DNA methylation markers, which showed a sensitivity of 85% (stage I: 80%, stage II: 85%, stage III: 89%, stage IV: 88%) and a specificity of 99% when evaluated in CRC patients and controls (n=434). Objective: This study aimed to assess the performance of TriMeth cell-free DNA methylation analysis in asymptomatic individuals. Methods and materials: TriMeth was evaluated in 727 average-risk, asymptomatic individuals from the Danish population-based CRC screening program aged 50-74 years. This cohort included 105 CRC patients (stage I: n=53, stage II: n=24, stage III: n=19, stage IV: n=9) and 622 age and gender-matched controls enriched for comorbidities common to the screening population, such as, diabetes (n=47), arthritis (n=9), hypertension (n=279), arteriosclerosis (n=90), inflammatory bowel disease (n=6), adenomas (n=49), and other cancers (n=91). For comparison, TriMeth was also applied to a cohort of 43 symptomatic CRC patients and 42 cancer-free controls. Results: In the asymptomatic cohort, the specificity was consistently high (99%) despite the enrichment for controls with comorbidities in this cohort. The sensitivity was 36%, and was reduced for all disease stages compared to symptomatic cancers, though most markedly reduced for stage I CRC (stage I: 9%, stage II: 67%, stage III: 53%, stage IV: 78%). For a subset of asymptomatic individuals (n=91), an additional aliquot of plasma was analyzed, revealing a very high reproducibility of TriMeth (Spearman corr.coeff. r=0.95, p=7.6e-14). However, merging data from the paired samples did not significantly increase sensitivity. TriMeth showed consistently high accuracy in the symptomatic cohort (sensitivity: 84% and specificity: 100%). Conclusion: TriMeth is highly specific and not affected by comorbidities commonly found in 50-74 years old individuals invited for CRC screening. TriMeth showed high sensitivity for all stages when evaluated