More research suggesting that the widespread use of the prostate specific antigen (PSA) test leads to the overdiagnosis of prostate cancer was published last week (British Journal of Cancer 2006 Jul 11, doi:10.1038/sj.bjc.6603246).
“This study shows that a substantial minority of the observed excess of cases in recent years is likely to be due to PSA testing. A majority of the observed excess in the tested group is estimated to be overdiagnosis,” say the authors from Cambridge University and the Wolfson Institute of Preventive Medicine.
Overdiagnosis was defined by the authors as the detection of prostate cancer through testing that otherwise would not have been diagnosed within the patient’s lifetime. The aim of the study was to estimate the extent to which any overdiagnosis could be put down to the use of the PSA test.
In the study, records of tests at Addenbrooke’s Hospital were linked to prostate cancer registrations, and the differences in prostate cancer registration rates between those receiving and not receiving prediagnosis tests—any testing at least six months before any diagnosis of prostate cancer—were calculated. Probabilities of overdiagnosis were calculated using data on the expected remaining lifetimes for the UK male population.
In the period looked at, 8894 men had at least one test, of whom 23% (2053) were 50-59, 30% (2701) were 60-69, and 38% (3352) were 70-89 years old.
Prostate cancer was detected in only 0.4% (32/8894) men tested. The rate of diagnosis of prostate cancer was 45% higher (rate ratios 1.45, 95% confidence intervals 1.02 to 2.07) in men with a history of prediagnosis testing.
The number of excess cases in the tested men was then calculated. “In this study, a history of testing was associated with a 45% excess risk of prostate cancer diagnosis,” say the authors.
They add, “More than half of the cases with prediagnosis testing are likely to have been overdiagnosed, that is they would not have been detected clinically in their lifetime without the PSA testing. Assuming average lead times of five to 10 years, 40-64% of the PSA-detected cases were estimated to be overdiagnosed.”
Researchers found that the proportion of men over 40 who had the test increased from 1.4% in 1996 to 5.2% in 2002.
The authors cite other research (British Journal of Cancer 2006 Jul 11, doi:10.1038/sj.bjc.6603247) that showed, for East Anglia a 6% excess of prostate cancer registrations in 1991-2000 relative to expectations based on trends before 1991. Around a quarter of these could be attributed to PSA testing.
Dr Muir Gray, director of the National Screening Committee, said, “With population based PSA testing, it is undoubtedly detecting cancers that would not reveal themselves in the lifetime of the individual. That is the reason we have not recommended population screening.”