NPC Archive Item: Prostate cancer screening: no compelling evidence that benefits outweigh harms.

NOTE – This is an archive post from the NPC and has not been updated since first publication. Therefore, some hyperlinks may no longer be working.
MeReC Rapid Review NPC Logo

2nd April 2009

Interim results from a large European randomised controlled trial suggest that, over nine years, 1410 men would need to be screened with PSA testing to prevent one of them dying from prostate cancer. However, interim data from a US study identified no reduction in deaths from prostate cancer from screening compared with usual care over a ten-year period. Neither study has yet provided information on the effect of screening on quality of life and further follow up in both studies is planned.

Action
As advised by the Chief Medical Officer after the publication of this new data, only men who have made an informed choice based on the Prostate Cancer Risk Management Programme materials should receive a prostate specific antigen (PSA) screening test.

NICE advises that men who present with symptoms suggestive of prostate cancer should have a digital rectal examination (DRE) and a PSA test after counselling. The results of these investigations will guide referral and further action.

Further information on the background, diagnosis and treatment of prostate cancer can be found on the prostate cancer section of NPC.

What is the background to these studies?
At present there is no national screening programme for prostate cancer in the UK, as there is no clear evidence to show this would bring more benefit than harm. Prostate cancer is an important health problem, but this criterion is only one of the ten Wilson and Jungner WHO criteria which must be satisfied to justify a national screening programme. These two randomised controlled trials (RCTs), one in the US and one in seven European countries (not including the UK), were carried out to provide information on the effects of screening on mortality  and quality of life, and when complete should  be helpful in re-assessing the benefits of a national screening programme.

What do the studies claim?
In the European study of 162,243 men over nine years, the death rate from prostate cancer was reduced by a fifth in men who were screened (PSA testing once every four years) compared with those who were not (rate ratio [RR] 0.80, 95% confidence interval [CI] 0.65 to 0.98, P=0.04). However, the absolute reduction in death from prostate cancer was 71 per 100,000 men per nine years;. This equates to a “number needed to screen” (analogous to a number needed to treat [NNT] of 1410 (95%CI 1142 to 1721) to prevent one death from prostate cancer over nine years, with an average of 1.7 screening visits per subject in that time.  Prostate cancer was detected in 70% more men in the screening group than in the control group (cumulative incidence 8.2% vs. 4.8%). Of those men diagnosed, an extra 48 would need to be treated to prevent one death from prostate cancer.

The US study randomised 76,693 men to annual PSA testing for six years plus annual DRE for four years, or usual care.

Although more cancers were detected in the screening group, after seven to ten years of follow-up, there was no significant difference between groups in the rate of death from prostate cancer (RR at 7 years 1.13, 95%CI 0.75 to 1.70; RR at 10 years 1.11, 95% CI 0.83 to 1.50). After seven years, the incidence of prostate cancer was 7.4% in the screening group compared with 6.1% in the control group who received usual care (RR 1.22, 95%CI 1.16 to 1.29). Results after 10 years were consistent with these findings.

Neither study reported on quality of life measures, but in the European study the rate of over-diagnosis in the screening group (diagnosis of prostate cancer in men who would not have clinical symptoms in their lifetime) has been estimated to be as high as 50%.

How do these relate to other studies?
A Cochrane review from 2006 identified only two RCTs of screening for prostate cancer, with a total of 55,512 participants. Both trials had methodological weaknesses. Intention-to-treat analysis identified no statistically significant difference in prostate cancer mortality between men randomised for prostate cancer screening and controls (RR 1.01, 95% CI: 0.80 to 1.29). Neither study assessed the effect of prostate cancer screening on quality of life, all-cause mortality or cost effectiveness. The Cochrane reviewers concluded that there was insufficient evidence either to support or refute the routine use of mass, selective or opportunistic screening compared to no screening for reducing prostate cancer mortality. The authors pointed to the results of the two studies reported here to make evidence-based decisions regarding prostate cancer screening.

So what?
The studies clearly show that screening with PSA testing (combined with DRE in the US study) detects many more cancers than would have been done otherwise. However, whether or not this translates into a survival benefit, from prostate cancer or otherwise, remains uncertain. It is important to consider possible benefits arising from early detection and treatment in the context of the potential harms on quality of life arising from over-diagnosis and over-treatment.

Screening of prostate cancer by PSA testing with or without DRE is an imprecise screening method. Screening undoubtedly identifies many more prostate cancers than would otherwise become apparent. The recently published Annual Report from the Chief Medical Officer includes an excellent review on the dilemmas posed by prostate cancer entitled “What to do with the pussycats”.

These trials are not yet complete, and publication might be considered too early for precise estimates of long-term benefits. Up to now, only the European study suggests that there is a significant effect on disease-specific mortality. Taking 95% confidence intervals into account the relative reduction after nine/ten years could be as low as 2% or as high as 35% in the European study and anything between a relative reduction of 17% and a 50% increase in the US study. Unlike the disease-specific survival curve for the European study, in the US study there appears to be no divergence of the two curves in favour of screening after about seven years. Both publications can be criticised for not providing information on quality of life, which is also being studied, with which to balance any effects on survival.

There are several reasons to explain the differences between the results from the two studies. There are differences in methodology (e.g. eligibility criteria, randomisation schemes, screening strategies, intervals and follow-up), rates of PSA screening in the control groups, and there could be differences in the subsequent investigations and treatments received. Many of these limitations are discussed in an accompanying editorial to the two articles.

The UK was not represented in the European study, and it is not known to what extent the results of the study can be extrapolated to the UK, where clinical practices regarding screening, detection and treatment of prostate cancer may differ. However, this study does provide new information to be considered when evaluating the merits of a national screening programme, and the Chief Medical Officer has stated that the Department of Health will formally ask the UK National Screening Committee to review the evidence and make recommendations. The ongoing ProtecT trial, which is evaluating the effectiveness of treatment for clinically localised prostate cancer, may provide further useful UK-specific information to inform a decision regarding the merits of a national screening programme, but is not due to finish until 2013.

In the meantime, NHS Prostate Cancer Risk Management Programme materials are available to support health professionals in counselling asymptomatic men who are considering having a PSA test. It is important that men requesting a PSA test are given information concerning the benefits, limitations and risks associated with having the test. Support materials are available in an information pack, these include a summary sheet, a booklet outlining the evidence, and a leaflet for men

Study details

European study: Schröder FH, et al, for the ERSPC investigators. Screening and prostate-cancer mortality in a randomised European study. N Engl J Med 2009;360:1320–8

Design: RCT in seven European countries (The Netherlands, Belgium, Sweden, Finland, Italy, Spain, Switzerland); blind assessment of causes of deaths.

Patients: 162,243 men aged 55 to 69 years (mean age 61).

Intervention: offered PSA screening (average once every four years). Most centres used a cut-off point of 3.0ng/ml as an indication for biopsy.

Comparison: men not offered screening.

Outcomes: primary outcome was rate of death from prostate cancer; quality of life not reported.

Results: median follow up of nine years; 82% accepted offer of screening (number of men receiving PSA testing in control group not reported); overall 16% of PSA tests were positive; cumulative incidence of prostate cancer 8.2% in the screening group and 4.8% in control group; false positive rate for PSA testing 76% based on biopsy results; positive predicted value for biopsy 24%, absolute reduction in death rate from prostate cancer from screening was 0.71 per 1000 patients; RR 0.80, 95% CI 0.65 to 0.98, P=0.04; number needed to screen 1410, 95%CI 1142 to 1721; estimated that an additional 48 need to be treated to prevent one death from prostate cancer; death rates from other causes not reported.

Sponsorship: supported by the Europe Against Cancer and fifth and sixth framework program for the European Union, by grants from agencies or health authorities in the participating countries, by unconditional grants from Beckman Coulter, and by numerous local grants.

US study: Andriole GL, et al, for the PLCO Project Team. Mortality results from a randomized prostate-cancer screening trial. N Engl J Med 2009;360:1310–9

Design: RCT in 10 US centres; blind assessment of causes of deaths.

Patients: 76,693 men aged 55 to 74 years.

Intervention: offered annual PSA testing for 6 years and DRE for 4 years; a cut-off point of 4.0ng/ml was considered as positive for prostate cancer; those with positive PSA results or suspicious findings on DRE were advised to seek diagnostic evaluation.

Comparison: usual treatment, which could include screening. Note that in the screening group, rates of compliance were 85% for PSA testing and 86% for DRE. Rates of screening in the control group increased from 40% in the first year to 52% in the sixth year for PSA testing and ranged from 41 for DRE.

Outcomes: primary outcome was cause specific mortality; quality of life not reported.

Results: median follow up of 11.5 years; compliance 85% for PSA testing, 86% for DRE; after 7 years, cumulative incidence of prostate cancer 7.4% in the screening group and 6.1% in control group (RR 1.22, 95% CI 1.16 to 1.29) and, after 10 years, 9.0% and 7.8%, respectively (RR 1.17, 95%CI 1.11 to 1.22); after seven years, death attributed to prostate cancer was 0.13% in the screening group and 0.11% in the control group (RR 1.13, 95%CI 0.75 to 1.70) and, after 10 years, they were 2.4% and 2.1%, respectively (RR 1.11, 95%CI 0.83 to 1.50); deaths from all causes, excluding prostate, lung or colorectal cancer, occurred in 6.6% of the screening group and 6.8% of the control group (RR 0.98, 95%CI 0.92 to 1.03) after seven years, and 10.3 and 10.6%, respectively (RR0.97, 95%CI 0.93 to 1.01) after 10 years; among men with prostate cancer at 10 years, more died from other causes in the screening group than in the control group (0.8% vs. 0.6%).

Sponsorship: supported by the US National Cancer Institute.

Feedback
Please comment on this blog in the NPC discussion rooms, or using our feedback form