Prostate cancer screening can save lives but it is too early for a national programme, study findsBMJ 2014; 349 doi: https://doi.org/10.1136/bmj.g5055 (Published 07 August 2014) Cite this as: BMJ 2014;349:g5055
Prostate specific antigen (PSA) screening for prostate cancer can reduce the number of deaths from the disease, a long term study has confirmed.1 But the authors added that “the time for population based screening has not arrived” and that further work was needed to reduce the harms from overdiagnosis and overtreatment of prostate cancer.
The European Randomised Study of Screening for Prostate Cancer began in 1993 and included centres in eight European countries. It randomised 163 388 men to PSA screening every four years (two years in Sweden) or no screening. The latest results detailed the prostate cancer mortality data after 13 years’ follow-up.
The absolute risk reduction of PSA testing in death from prostate cancer after 13 years was 1.28 deaths in every 1000 men randomised—the equivalent of one fewer death from prostate cancer in every 781 men invited for screening. For each of the 27 additional cancers identified through screening, one death was averted.
The study found no difference in all cause mortality between the screening and control groups, but the authors said that no reduction in overall mortality was anticipated because prostate cancer caused only a small proportion of all deaths.
The relative risk reduction from PSA screening remained similar with 11 or 13 years of follow-up, at 22% for 11 years and 21% for 13 years. However, the extra two years’ data meant that the authors could demonstrate a convincing improvement in absolute risk reduction, with a commensurate reduction in the numbers needed to screen and numbers needed to identify.
There are reasons to be cautious about the study. Some heterogeneity existed between study centres; for example, only Swedish and Dutch centres showed a significant reduction in prostate cancer mortality, meaning that the largest component country, Finland, still showed no significant reduction. Also, the length of follow-up from randomisation may not reflect the full benefit of the intervention: the median follow-up after diagnosis was only 6.4 years in the screening group and 4.3 years in the control group, and more than 70% of the study participants were still alive. In addition, the control group showed varying but high levels of contamination, with 23-40% of participants undergoing PSA testing at some time.
Most of these limitations would tend to underestimate the effects of PSA screening on mortality, so it is notable that, despite the study’s success, the authors did not call for the introduction of population level PSA screening. This has been debated for many years, with increasing acceptance that universal screening identifies too many dormant cancers.
In an accompanying editorial Ian Thompson and Catherine Tangen, of the Cancer Therapy and Research Center at the University of Texas in San Antonio, said, “With an enormous reservoir of cancers in ageing men, there is a major risk of detection of many cancers that will never cause symptoms or death.”2 The study showed that 60% of cancers detected by screening were classified as low risk, they added.
Instead of recommending population screening, the study authors pointed to the use of risk stratification tools that could predict a man’s chances of finding a low or high grade tumour on biopsy. They also discussed the use of multiparametric magnetic resonance imaging to differentiate between aggressive and non-significant tumours.
They wrote, “Informed decision making, with well designed decision aids, is necessary for individuals who consider PSA based screening for prostate cancer . . . Despite our findings, further quantification of harms and their reduction [is] still considered a prerequisite for the introduction of population based screening.”
Iain Frame, director of research at the charity Prostate Cancer UK, said that the study highlighted the need for tests that could distinguish between dangerous cancers and those that caused less harm.
He added, “Getting an accurate diagnostic test that can be delivered relatively cheaply and simply could mean that the UK can start thinking about the introduction of a national screening programme, and our research is working towards that. However, in the meantime, men most at risk of prostate cancer—black men, men over 50, and men with a family history of the disease—should speak to their GP about their risk and whether the PSA test is right for them.”
Cite this as: BMJ 2014;349:g5055