The Daily Mail has reported that “routine testing for prostate cancer in middle-aged men does not save enough lives to justify the harm”.
The news is based on a 20-year Swedish study that found screening did not reduce overall death rates or death due to prostate cancer when offered to 50-69 year olds every three years. While the study was well-conducted it was relatively small and its findings may not be applicable to other populations given the size of the study and the screening methods used, which included manual examination of the rectum and the contested method of PSA testing.
These findings support the conclusions of the UK National Screening Committee that there is no evidence that the benefits of prostate screening outweigh the harms. Also, the findings highlight an important point that men should be fully informed about the possible side effects from prostate cancer treatment, particularly as some prostate tumours would not progress to become problematic if they were left untreated.
In the UK, PSA testing is carried out on an individual basis (rather than being offered to the general population). The test is done only after a full assessment by a medical practitioner and discussion about the circumstances supporting its use and possible adverse effects, given a patient’s individual circumstances.
Where did the story come from?
The study was carried out by researchers from the Karolinska Institute and Linköping University Hospital in Sweden, and the Norwegian University of Life Sciences. It was funded by the Research Council of the South-East Region of Sweden, the Swedish Cancer Foundation and the County Council of Östergötland. The study was published in the peer-reviewed British Medical Journal.
Newspapers have covered the story well, although some have not made it clear that UK men are not offered routine prostate cancer screening because research has indicated that the benefits of screening do not outweigh the harms. Headlines suggesting that ‘prostate screening has no benefit’ and that ‘prostate testing is ‘doing more harm than good’’ should be viewed in context as they do not apply to the various situations where UK men are given prostate testing due to their specific medical circumstances.
What kind of research was this?
This was a randomised controlled trial in which healthy Swedish men aged 50- 69 years were randomly selected to either be invited for screening every three years or left unscreened. At the end of the study, the researchers compared these groups for their incidences of prostate cancer, its severity, its treatment and mortality overall, and from the disease within the 20 years since screening commenced. This particular publication reported the mortality outcomes from the study.
What did the research involve?
All men aged between 50-69 years living in Norrköping in Sweden were identified through a population register. There were 9,026 of them. The men were listed by their dates of birth and every sixth one was selected to be screened from 1987 onwards. This led to 1,494 men being invited for screening, with the remaining 7,532 men forming the control group.
The screening programme was advertised through the local media. Men in the screening group were invited to attend a screening appointment where they underwent a digital rectal examination. In the final two screenings (conducted in 1993 and 1996), levels of prostate specific antigen (PSA) were also measured.
Men in the control group were not contacted and offered screening. However, if prostate cancers were detected incidentally or through symptoms, they were treated in the same way that men in the screening group were treated. A needle biopsy was taken for analysis if any suspicious nodules were detected. Men with positive cytology, i.e. evidence of cancer, were then treated according to the standard management protocols for the region.
As the men had been identified from a population register it was possible to trace them all for the duration of the study. This was because all cases of prostate cancer, their date of diagnosis, severity, treatment and date and cause of death were recorded on the region’s prostate cancer registry.
This study reported the mortality outcomes for the two groups, although other outcomes within this population were measured and reported in other publications. A statistical technique called Cox regression was used to analyse whether there was a significant difference between the screened and unscreened groups in terms of overall and prostate cancer-specific mortality 20 years after the study began.
What were the basic results?
The compliance rate within the screening group was about 70-78%. Over the course of the study, 85 out of 1,494 men in the screening group (5.7%) developed prostate cancer compared with 292 out of 7,532 in the control group (3.9%). Only 50% of the tumours in the screened group were detected at a screening examination, the remainder being detected between screens.
Tumours in screened men were more likely to be localised than in unscreened men (57% versus 27%). But although deaths from prostate cancer were numerically lower in the screening group (35% versus 45%) the difference between the groups was not statistically significant (RR 1.16, 95% CI 0.78 to 1.73). When the researchers adjusted for participants’ age at the start of the study, men in the screening group were slightly more likely to die from prostate cancer within the 20 years following study commencement.
In terms of the treatments received, the cancers detected through screening were more likely to be treated using radical prostatectomy (i.e. removal of the entire prostate gland) than cancers in the non-screened men or those detected between screenings for the intervention group.
How did the researchers interpret the results?
The researchers conclude that before they undergo screening, men should be advised that if prostate cancer is diagnosed, it will be treated with an intention to cure their cancer. Treatment might include measures such as performing a radical prostatectomy, a procedure associated with side effects including erectile dysfunction and incontinence.
Importantly, the researchers note that their study was not big enough to draw definitive conclusions on the issue of prostate cancer screening.
This was a well-conducted study, although small for a screening study. It has shown that the particular blanket-screening programme being assessed did not reduce the overall or prostate-cancer specific mortality in a group of men in Sweden.
There are several points to bear in mind when interpreting these findings:
- This is a relatively small study. Population-based screening programmes would include thousands of healthy men, so research assessing the effects of screening in only 1,494 men may not reflect the outcomes seen in larger groups or in national screening programmes.
- In this study, the early screening appointments for the intervention group consisted only of digital rectal examination (i.e. using a finger to feel for abnormalities). Only in the later two screenings were men also offered PSA screening. The rectal examination is not a recommended approach for screening healthy men so the applicability of the findings to the reality of a modern screening programme may be limited. However, the findings do not contradict the approach of the Department of Health in this country, which is to not offer routine prostate cancer screening. The UK National Screening Committee says there is no evidence that the benefits of a PSA-based screening programme will outweigh the harms.
The researchers note that the next goal for prostate screening is to find a way to discriminate between tumours that may never develop into serious disease and those that may progress. If this can be achieved, then only men whose tumours are high risk for progression could be offered radical curative treatments. Men with lower-risk tumours may be offered a choice to forego these treatments and their associated side effects.
Analysis by Bazian
Edited by NHS Website
Links to the headlines
BBC News, 1 April 2011
Daily Mail, 1 April 2011
The Daily Telegraph, 31 March 2011
Links to the science
BMJ 2011; 342:d1539