Unanswered questions in screening for prostate cancer

Prostate cancer fulfils some of the conditions required of a disease that might be managed by population screening. In a cohort of 50- to 60-year-old men, carrying out a rectal examination and prostate specific antigen (PSA) test will detect clinically suspicious areas within the prostate in approximately 5%, and approximately 10% will have a raised PSA. We are however unsure which of the prostate cancers that are known to be present in approximately 30-40% of men aged over 60 years will be detected. Eventually after such screening, around 4% of men with an otherwise normal prostate will be found to have prostate cancers. The use of rectal examination may increase the number of tumours found, but will reduce compliance. The use of free/total PSA ratios will reduce the number of unnecessary biopsies at the expense of missing some tumours. Of more concern, we remain uncertain how effective aggressive local treatment is in altering the natural history of the disease. The risk of a 50-year-old man with a 25 year life expectancy of having microscopic cancer is 42%, of having clinically evident cancer is 9.5%, and of dying of prostate cancer 2.9%. Only a small proportion of cancers known to be present become clinically evident: more men die with prostate cancer than of it. Screening will identify some men with cancer who will not benefit from treatment. It is unclear whether screening would be followed by a reduction in morbidity and mortality. Recent data suggest a screening effect has been observed in the USA with: an increase in incidence, a decrease in men with distant metastases. The small decrease in mortality recently observed (many times smaller than the increase in incidence) may be confounded by inappropriate 'attribution' of cause of death, the detection of men with better prognosis distant metastatic disease responsive to hormonal ablation and changes in social factors such as diet. Future changes may incorporate molecular markers that might aid identification of men best treated aggressively because of a risk of progression. Tests to identify genetic pre-disposition may also allow targeted screening. New treatments and early chemoprevention or dietary strategies will again shift the ground on which these arguments are being rehearsed. The most urgent evidence required concerns the effectiveness of treatment strategies.