Abstract
The recent large genotyping studies have identified a new repertoire of cancer susceptibility genes and loci which are characterized by common risk alleles and low relative risks. Because of these properties, these loci explain a much larger proportion of the etiology of the particular cancers, described by the population attributable fraction (PAF), than of their familial risks (FRRs). For breast cancer, the 9 established loci gave a joint PAF of >60%, but explaining only some 8% of the empirical FRR. For prostate cancer, 6 independent loci at chromosome 8q conferred a joint PAF of 35% but the loci explained no more than 1.9% of the empirical excess familial risks. For colorectal cancer, the contributions of the 2 identified loci to PAF and FRR were somewhat lower. The genome‐wide array platforms have been built for common variants, constraining the results to variants with high PAFs and low FRRs. However, the common variants are likely to tag rarer causative variants with much higher FRRs. The detected loci are noncoding and the underlying genetic mechanisms have not been worked out. The data suggest, in spite of the reservations for combining data on PAFs across populations, that the published first‐generation genome‐wide scans on breast, prostate and colorectal cancers have made successful inroads into genomics of common cancers, yet leaving the mechanisms to be explained. © 2008 Wiley‐Liss, Inc.