News

Prostate cancer is the most common cancer in men, with an average absolute lifetime risk of around 12% across the population. However, there is continuing controversy over whether to implement screening programmes for this disease. Despite numerous studies, it remains unclear whether screening for prostate specific antigen (PSA) does more harm than good (see previous news). Great hope has therefore been placed in finding common genetic variants that might be used to stratify the population into risk categories, so that screening could be targeted at those considered to be at highest risk (see previous news).

 

Four new studies have been published online in the journal Nature Genetics, identifying over a dozen new prostate cancer susceptibility loci. Taken together, these studies involve over 100,000 individuals enrolled as either cases or controls in studies across the globe. Two studies focus on a region of chromosome 8 (8q24.21), which has previously been associated with numerous cancers including prostate, showing that multiple SNPs in this region are independently associated prostate cancer risk [Yeager M et al. (2009) Nat Genet doi:10.1038/ng.444; Al Olama AA et al. (2009) Nat Genet doi:10/1038/ng.452]. Although this region contains no known protein-coding genes, the oncogene MYC is nearby and appears to be the most attractive candidate for the underlying causal association. This gene encodes the transcription factor c-myc, which is linked with the regulation of numerous genes involved in cell proliferation.

 

The other two papers report large genome-wide association studies (GWAS), between them finding or verifying SNPs associated with prostate cancer on chromosomes 2, 3, 4, 8, 11, 19 and 22 [Eeles RA et al. (2009) Nat Genet doi:10.1038/ng.450; Gudmundsson J et al. (2009) Nat Genet doi:10.1038/ng.448]. Although most of the risks associated with each variant are modest, with per allele odds ratios of around 1.1-1.3, there are now over 20 prostate cancer susceptibility variants. When combined, these variants confer a risk of approximately double for the top 10% of the population, and triple for the top 1%, relative to the population average. Moreover, taken together, these loci may explain around 20% of the familial risk of prostate cancer.

 

Comment: Various commentators have suggested that these studies are likely to herald a new era of targeted prostate cancer screening (for example, see The Times). Whilst such claims are likely to be somewhat premature – due the requirement for robust and systematic evidence of the benefits, harms and costs of rolling out a national screening programme – there is no doubt that relative risks of 2-3 are large enough to be potentially clinically useful, as they are comparable with the twofold relative risk conferred by first-degree relatives. Further research is needed to understand whether those at a high genetic risk of developing prostate cancer are also at highest risk of developing the aggressive form of the disease, for which early treatment is undoubtedly beneficial, or the more common benign form, for which treatment is unnecessary and potentially harmful.