Blood-based detection of early-stage colorectal cancer using multiomics and machine learning.

Abstract

66 Background: Despite population screening efforts, screening rates for colorectal cancer (CRC) remain suboptimal. A non-invasive, blood-based screening test with high sensitivity and specificity in early-stage disease should improve adherence and ultimately reduce mortality; however, tests based only on tumor-derived biomarkers have limited sensitivity. Here we used a multiomic, machine learning platform to discover, refine, and combine tumor- and immune-derived signals to develop a blood test for the detection of early-stage CRC. Methods: Samples from 591 participants enrolled in a prospective study including average-risk screening and case-control cohorts (NCT03688906) were included in this analysis (CRC: n = 43; colonoscopy-confirmed CRC-negative controls: n = 548). Participants with CRC were 60% male with a mean age of 63, and controls were 55% male with a mean age of 60. Stage distribution was 54% early (I/II) and 34% late (III/IV) with 11% unknown. Plasma was analyzed by whole-genome sequencing, bisulfite sequencing, and protein quantification methods. Computational methods were used to assess and infer the performance of individual and combined assays. Results: For colorectal adenocarcinoma, which represents ~95% of all CRCs, our multiomic test achieved a mean sensitivity of 92% in early stage (n = 17) and 84% in late stage (n = 11) at a specificity of 90%. Across all CRC pathological subtypes, our test achieved a mean sensitivity of 80% in early stage (n = 19) and 83% in late stage (n = 12) at a specificity of 90%; the test detected the single squamous cell carcinoma but missed both neuroendocrine tumors. Individual assays achieved a mean sensitivity of 50% in early stage and 66% in late stage at a specificity of 90%. Conclusions: In a prospective cohort, we demonstrated high sensitivity and specificity for early-stage adenocarcinoma by combining tumor- and immune-derived signals from cfDNA, epigenetic, and protein biomarkers. While most CRCs are adenocarcinomas, detection of all pathological subtypes is required to maximize sensitivity in a screening population. Further analysis of molecular and pathological subtypes, as well as the entire ~3000 patient cohort, is underway. Clinical trial information: NCT03688906.

Hydroxymethylation profile of cell-free DNA is a biomarker for early colorectal cancer

“…The study showed that 5hmC profiles of cfDNA are a strong predictor of cancer, including high sensitivity for the detection of early-stage cancer. The performance achieved in this large well-controlled study with internal validation is at least comparable to performance in CRC reported for other analytes 8 , 12 , 32 – 34 and for 5hmC in smaller studies 13 , 14 . Our data suggest that the 5hmC signal derived from epigenetic changes in cfDNA is more sensitive at an early stage than that from other analytes.…”

“…( C ) The 5hmC-based classifier performs comparably to reported classifiers for stage 1 CRC. To gain approximately comparable confidence intervals, 95% binomial confidence intervals were computed for all classifiers using publicly available information 8 , 12 , 33 , 34 , 38 . The CRC classifier from Putcha et al contains both Stage 1 and Stage 2 samples.…”

Hydroxymethylation profile of cell free DNA is a biomarker for early colorectal cancer

“…reported validation (not yet peer reviewed) of the classi er in a small validation set (n = 17) with a mix of stage 1 and 2 samples 33 . We include both the training and validation reports in Fig.…”

Use of Molecular Assays and Circulating Tumor DNA in Early-Stage Colorectal Cancer: A Roundtable Discussion of the Gastrointestinal Cancer Therapy Expert Group

“…The group is optimistic on the future of MRD testing in CRC and anticipates that this type of testing will emerge as a standard component of clinical decision making as data continues to rapidly emerge. ctDNA based blood test (LUNAR-2) currently under investigation for use in detecting CRC in an average risk population ECLIPSE (NCT04136002): CRC screening study initiated in 2019; goal to enroll 10,000 patients ages 45 -84 planning to have a colonoscopy Grail 3,4 Pan-cancer early detection blood test utilizing cell-free (cf) nucleic acids (DNA and RNA) PATHFINDER (NCT04241796): prospective, multicenter study enrolling~6,200 participants; primary goal is to guide appropriate diagnostic workup for different cancer types Freenome 5,6 Multiomics platform utilizing cfDNA, cfRNA, proteomics, epigenetics, as well as computational biology and machine learning methods to generate early tumor and immune cancer signatures PREEMPT CRC (NCT04369053): clinical study enrolling~14,000 average risk participants undergoing colonoscopy; primary goal is to validate blood based test for CRC screening Exact Sciences 7,8,9 ColoGuard Table 5. Use of Molecular Assays in CRC: Summary of GICTEG Practice Points…”

“…Although not intended to be a comprehensive list of all CRC and/or pan-cancer screening tests that are currently under investigation, Table 1 provides an overview of some of the genomic tests and technologies that have been used to screen for CRC, including commercially available and investigational tests [1][2][3][4][5][6][7][8][9]. These noninvasive tests aim to assess for genomic and epigenomic changes in bodily fluids (i.e., "liquid biopsies"), which can detect the presence of cancer.…”