The rs3803185 variant was not significantly associated with CRC risk in an external cohort (MCC-Spain), but it still showed some borderline significance in the pooled analysis of both cohorts (OR = 1.08, 95% CI = 0.98-1.18, P-value = 0.09, log-additive 0, 1, 2 alleles).
Our results suggest that Trp149Stop is not a predisposition allele in breast, prostate, or colorectal cancer in the Finnish population, and, while the Gly65Val variant may increase familial prostate cancer risk and the Cys148Arg change may affect both breast and prostate cancer risk, the evidence is not strong in these data.
Whereas P131L and W149X did not seem to affect CRC risk, C148R did show, for the first time in CRC, statistically significant differences between cases and controls [odds ratio (OR) = 1.45, 95% confidence interval (95% CI) = 1.13-1.86, P = 0.003], sporadic cases and controls (OR = 1.59, 95% CI = 1.13-2.23, P = 0.007) and familial cases and controls (OR = 1.55, 95% CI = 1.10-2.19, P = 0.01) in agreement with a hypothetical moderate increase of the cancer risk linked to the C148R ARLTS1 variant, both in sporadic and familial CRC cases.
As the first to evaluate the association between Cys148Arg and Trp149Stop and colorectal cancer (CRC) risk, we genotyped 611 cases with CRC (including 77 cases with a first-degree family history) and 539 controls recruited from the German DACHS study.