Abstract
Abstract
BACKGROUND: The genetic events affecting progression from early-stage to advanced/metastatic colorectal cancer remain poorly understood. There is preclinical evidence that germline, rather than tumor somatic, variants in key mechanistic areas (host immunity, paracrine signaling, extracellular matrix, and lymphangiogenesis) can affect the ability of an early lesion to metastasize. This led us to assess these variants in pooled data from 13 epidemiologic studies within the Genetics and Epidemiology of Colorectal Cancer Consortium (GECCO).
METHODS: Genetic variants were compared in early-stage (AJCC stage 1-2; n = 4,978) and advanced (AJCC stage 3-4; n = 3,263) colorectal cancer cases in GECCO. Both common (minor allele frequency [MAF] >5%) and rare (MAF <5%) single nucleotide polymorphisms (SNPs) were evaluated in 88 candidate genes (range: 5kb upstream to 500bp downstream) relevant to the mechanisms of interest. A log-additive genetic model adjusting for age, sex, study, and 3 principal components was used to assess the association for each common variant. To improve power for rare variants association, these were aggregated by gene and gene-level associations were evaluated using the Mixed effects Score Test (MiST). The Bonferroni correction was used for multiple testing, resulting in two-sided p-value significance thresholds of p<3.2 × 10-5 ( = 0.05/1582 SNPs) for common variants and p<5.7 × 10-4 ( = 0.05/88 aggregates) for rare variants.
RESULTS: 1,582 common variants were analyzed after linkage disequilibrium-based SNP pruning, with top 5 results in MMP10 (p = 1.24 × 10-3), EGFR (1.41 × 10-3), MMP7 (1.45 × 10-3), ANGPT2 (2.54 × 10-3), and VEGFR2 (3.16 × 10-3). None reached a priori statistical significance. The most promising aggregated rare variant candidates were CD247 (p = 3.01 × 10-3), HGF (4.44 × 10-3), CD8A (1.57 × 10-2), CXCL2 (1.86 × 10-2), FGFR2 (2.07 × 10-2), and VEGFB (3.52 × 10-2), though again none reached statistical significance.
CONCLUSIONS: Despite the preclinical evidence for these candidate “progression genes,” large-scale genomic analysis did not identify statistically significant germline genetic variation in advanced vs. early-stage colorectal cancer. Follow-up evaluation will include additional survival and environmental data.
Citation Format: Stacey A. Cohen, Conghui Qu, Tabitha A. Harrison, Li Hsu, Polly A. Newcomb, Stephen B. Gruber, Stephane Bezieau, Graham Casey, Andrew T. Chan, Jenny Chang-Claude, Martha L. Slattery, Emily White, Ashley J. Vargas, Rami Nassir, Ulrike Peters, William M. Grady. Assessment of candidate host “progression genes” in the development of advanced colorectal cancer. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 1622.