European Urology

Kiemeney LA, Thorlacius S, Sulem P, et al
Nat Genet 2008;40:1307–12
Expert's summary:
The authors performed an exhaustive study to identify single-nucleotide polymorphisms (SNPs) associated with bladder cancer. They initially genotyped 1803 cases and 34 336 controls for 302 140 SNPs. Ten SNPs were significant at a lower-than-defined threshold (p < 5 × 10⁻⁵) and were then genotyped in a new cohort of 2165 cases and 3800 controls. The strongest associations with bladder cancer were found with a SNP at 8q24.21 (rs9642880; odds ratio [OR]: 1.22; p = 9 × 10⁻¹²) and one on 3q28 (rs710521; OR: 1.19; p = 1.15 × 10⁻⁷). The 8q24 SNP is 30 kb downstream of the c-Myc oncogene and close to a predicted gene (known as BC042052). Approximately 20% of European individuals are homozygous for rs9642880[T], and their risk of developing urinary bladder cancer is 1.49 times that of noncarriers.
Expert's comments:
Genetic epidemiology is undergoing a revolution. Textbooks illustrate inheritance and disease risk using Mendelian principles with uncommon high-risk genes (eg, BRCA1). Inherited mutation of these genes usually results in loss of function and cancer. However, for most cancers, it is common and low-risk gene variants (known as SNPs, which alter rather than ablate function) that determine cancer risk (plus environmental factors) [1]. This hypothesis supports observations that most cancers occur outside cancer families, and an individual's risk of cancer increases modestly if a direct relative is affected (eg, OR of 1.24 for bladder cancer [2]). Kiemeney et al studied a large European population for 302 000 SNPs of the 14.7 million currently known. They did not identify a coding SNP (one that changes a gene's sequence) at their statistical threshold but did find several noncoding SNPs highly associated with the disease. Whilst at first this may appear to be disappointing, many positive findings can be elicited. First, these data suggest that a familial bladder cancer gene does not exist. Second, the modest risk attributed to the SNP suggests that environmental factors are far more important than genetic factors for bladder cancer (changing behaviour could reduce burden). Third, one can identify homozygous individuals at whom to target health promotion. Finally, these findings potentially point to new methods of genetic susceptibility. Studies in breast, colon, and prostate cancer (referenced in Ghoussaini et al [3]) have all identified SNPs within the 8q24 region that predispose to their respective cancers. Why these tumours all share this region, in which there are few genes, is unclear. Could this region mark a distant genetic event or represent part of the machinery of an unknown molecular control mechanism?

Conflicts of interest

The author has nothing to disclose.