Abstract 1598: Deep targeted sequencing of colorectal cancer cases to study associations of molecular subtypes with clinical, genetic, and lifestyle risk factors
Bibliographic record
Abstract
Abstract Colorectal cancer (CRC) is the 3rd most common cancer in men and women, excluding skin cancer, and the 2nd leading cause of cancer death in the United States. In this biologically heterogeneous disease, a comprehensive molecular characterization is valuable for understanding tumorigenesis and studying associations with clinical, lifestyle, environmental, and germline genetic factors. The Colon Cancer Family Registry (CCFR) and Genetics and Epidemiology of Colorectal Cancer Consortium (GECCO) collected clinical, germline genetic, and epidemiologic data from over 125,000 CRC cases. In a subset of 2,500 CRC cases, we conducted targeted deep sequencing of tumor and matching normal DNA. The custom AmpliSeq sequencing panel includes 205 genes, which were prioritized from literature and analyses of existing CRC tumor whole exome sequencing data. Next-generation sequencing was conducted on Illumina Hi-Seq 2500, and mean sequence coverages of 748x and 287x were attained for tumors and normal DNA samples, respectively. So far, 1,705 cases have been analyzed. Strelka and MuTect were used to identify somatic single nucleotide variants, and VarsScan2, Verdict, and Strelka were used to call indels. Orthogonal technologies were used for validation and improving mutation calling algorithms. Approximately 18% of tumors were hypermutated, among which 70% exhibited microsatellite instability (MSI). Hypermutated tumors with MSI occurred more frequently in proximal colon compared to distal colon and rectum (OR, 10.10; 95% CI, 6.98-14.63; P<0.0001). Hypermutated tumors carried non-silent mutations in DNA mismatch repair genes (MSH2, MSH6, MLH1, MLH3, and PMS2), POLE, and POLD1. In POLE, in microsatellite stable hypermutated tumors, non-silent mutations were frequent in the exonuclease domain. Approximately 96% of non-hypermutated and 99% of hypermutated tumors contained non-silent mutations in genes in the Wnt/beta-catenin, p53, receptor tyrosine kinases/RAS, transforming growth factor-beta, and phosphatidylinositide 3-kinases pathways. Most tumors carried non-silent mutations in genes in more than one of these signaling pathways. The APC gene was the most significantly mutated gene in non-hypermutated tumors followed by TP53, KRAS, and PIK3CA. In hypermutated tumors, RNF43 was the most significantly mutated gene followed by BMPR2, APC, and BRAF. CRC-specific survival was significantly more favorable among individuals with hypermutated tumors, regardless of POLE mutation status (HR=0.36, 95% CI: 0.23-0.57). This survival benefit was attenuated in analyses of overall survival (HR=0.82, 95% CI: 0.65-1.02). This large dataset is being used to study associations with clinical, lifestyle, and environmental factors. This study will provide valuable information to develop better strategies for the prevention, diagnosis, and treatment of CRC. Citation Format: Syed H. Zaidi, Amanda I. Phipps, Tabitha A. Harrison, Robert S. Steinfelder, Quang Trinh, Barbara L. Banbury, Adilya Rafikova, Megan Van Tassel, Emma Reid, Stefanie Brezina, Marios Giannakis, Charles S. Fuchs, Li Hsu, Andrea Gsur, Shuji Ogino, Steven Gallinger, Polly A. Newcomb, Peter T. Campbell, Wei Sun, Thomas J. Hudson, Ulrike Peters. Deep targeted sequencing of colorectal cancer cases to study associations of molecular subtypes with clinical, genetic, and lifestyle risk factors [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 1598.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.002 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.001 |
| Research integrity | 0.001 | 0.000 |
| Insufficient payload (model declined to judge) | 0.003 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".