Inflammation and one of its mediating transcription factors, NF-kappa B signaling (NFκB) have been implicated in prostate cancer (PrCa) carcinogenesis. We sought to define whether germline gene polymorphisms that interact with NFκB are associated with metastatic disease after prostatectomy (RP) or radiation (XRT) for localized disease.
Using a bioinformatics approach interrogating publicly available datasets, we defined a genome-wide functional association network specific to lethal PrCa consisting of 351 genes and 8,154,133 high-confidence functional associations related to the NFκB pathway. The dense module searching (DMS) method was used to analyze 419,461 SNPs from a previously conducted genome wide association study (GWAS) case-only study of 196 lethal PrCa cases compared to 368 indolent controls in the Harvard School of Public Heath (HSPH) Cohorts. Top hits from DMS were then tested in two independent PrCa cohorts: (i) ECOG/DFCI (n = 254 cases, 256 controls) and (ii) Fred Hutchinson Cancer Research Center (FH, n = 570 cases, 103 controls). In all 3 studies, “controls” were men with PrCa who are alive with no evidence of metastasis at least 8-years after RP or XRT and “cases” were men who developed metastatic disease after RP or XRT (FH, HSPH, ECOG) or with de novo presentation (ECOG).
From the DMS, 40 SNPs with a minor allele frequency > 0.1 were associated with lethal PrCa. Of these, rs1910301 in the promoter region of FRAS1 was nominally associated with lethal disease in all 3 studies with similar size effects: the odds ratio (OR) for the A allele was 1.40 (p = 0.02) in HSPH, 1.35 in ECOG/Gelb (p = 0.04), and borderline significant in FH [OR 1.3, p = 0.07]. Fixed effects meta-analysis of all three cohorts found a significant association: OR = 1.38 95% CI: 1.15-1.66; p-value 0.005.
A SNP in the promoter region of FRAS1, which forms a gene unit with FREM2 and together regulate epidermal-basement membrane adhesion and cell migration, is associated with metastatic PrCa. FREM2 is an NFκB regulated gene and mutations in FREM2 and FRAS1 are associated with the Fraser syndrome. Further work is needed to determine the effect of rs1910301 on FRAS1 function and cellular adhesion and the metastatic process.
Clinical trial identification
Legal entity responsible for the study
US Department of Defense, NIH
C.J. Sweeney: Consultant with compensation and research: Janssen (C, R); Astellas (C, R); Sanofi (C, R); Bayer (C, R), Sotio (R), Pfizer (C). All other authors have declared no conflicts of interest.