Wednesday March 18 2009
The PSA test is a blood test
The BBC reports that “routine prostate cancer screening could cut death rates from the disease by 20%”. It said the results from a major study suggest that 2,000 lives a year could be saved in the UK.
The study, which involved over 160,000 men aged 55 to 69 from seven European countries, found that men who were screened every four years with a PSA test were 20% less likely to die from prostate cancer compared to men who received routine care.
Despite these preliminary results, however, the researchers say that it is too soon to advocate a PSA screening programme. They say that with the benefits comes a "high risk" of over diagnosis and over treatment.
They note that to save one life, 1,410 men would need to be screened and 48 treated. Apart from the man whose life has been saved, it’s impossible to say how many of the treated men would benefit.
In the UK men over 45 can ask their GP for a PSA test, but it is not offered as standard. Currently, only about 6% of men ask for the test.
The health minister for England, Ann Keen, said that she will ask the UK National Screening Committee to review the evidence and make recommendations.
Where did the story come from?
The research was carried out by a group called the European Randomised Study of Screening for Prostate Cancer (ERSPC) investigators, which was headed up by Dr Fritz H. Schröder. This was a multicentre European study supported by grants from Europe Against Cancer, the European Union, and other agencies and health authorities in the participating countries. The study was published in the peer-reviewed New England Journal of Medicine.
What kind of scientific study was this?
The aim of this randomised trial was to determine if a 25% reduction in prostate cancer mortality could be achieved by prostate specific antigen (PSA) based screening. PSA is a protein produced by the cells of the prostate gland and is picked up by a blood test. Although PSA is present in small quantities in healthy men, there are often elevated levels in men with enlarged prostate glands due to benign disorders or cancer.
The ERSPC trial began in the early 1990s and carried on until 2006. It involved 182,000 men between 50 and 74 years of age. The trial took place in several european countries, each of which conducted the trials in its own way. In Finland, Sweden and Italy, the researchers used population registries to identify potential trial subjects and randomly allocated them to different groups before asking for their consent. In the Netherlands, Belgium, Switzerland and Spain, the researchers only enrolled the participants after they had provided consent. Portugal was unable to provide the necessary data and withdrew from the study in 2000. France only entered the trial in 2001, and so did not have enough follow up data for inclusion in this report. Belgian data from a pilot study from 1991 to 1994 were also included in the analysis. Most centres began the study after 1994.
After excluding those from some countries who did not give their consent and those outside the “core age” group, the researchers were left with 162,243 men between 55 years and 69 years for randomisation.
Different countries used different protocols for enrolling and screening men. For instance, in Sweden the researchers only enrolled men between 50-54 years of age, while other countries enrolled men up to the age of 74. In Finland men were recruited at 55, 59 63 and 67 years of age and were screened up to 71 years of age.
Most centres used a cut off PSA value of 3.0 nanogrammes (ng) per ml (of blood) to determine whether a man needed to have further investigations, whereas some used 4.0ng per ml, and the Belgian pilot study used a 10ng per ml cut off. Some countries based the decision on the ratio of free (active) PSA to total PSA. Some centres referred men who were above the chosen threshold straight for a biopsy whereas others performed a rectal examination and ultrasound in borderline cases before deciding if a biopsy was justified. Up to 1997, Dutch and Belgian centres performed all three procedures at the same time. The type of biopsy performed and the treatments offered (surgery, radiotherapy or hormone therapy) were determined by local policies. The screening interval varied, from four years in 87% of subjects to two years in Sweden and up to seven years in Belgium.
Data was analysed on an intention to screen basis, meaning that all those offered screening (including those that declined) were included in the screened group for analysis, even if they didn’t actually receive screening.
Causes of death were classified by an independent committee, who were aware of the treatments received. Deaths were categorised according to whether the death was definitely, probably or possibly due to prostate cancer, due to complications the prostate screening intervention (for example due to the biopsy) or due to other causes with or without prostate cancer as a contributory factor. The categories definitely, probably and causes related to screening were grouped together for the analysis.
What were the results of the study?
The researchers say that in the screening group, 82% of men accepted at least one offer of screening. The overall rate of new prostate cancers (cumulative incidence) was 8.2% in the screening group and 4.8% in the control group.
The risk of death from prostate cancer in the screening group was reduced by 20% compared to the control group (RR 0.80, 95% confidence interval, 0.65 to 0.98). This represented a reduction of 0.71 deaths per 1,000 men with screening.
The researchers go on to say that that 1,410 men would need to be offered screening, with 48 additional men being treated to prevent one death from prostate cancer over 10 years.
During the trial 126,462 PSA-based tests were performed (an average of 2.1 per person). 20,437 of these tests were positive (16.2%) and 17,543 biopsies were performed on men testing positive for the blood test (85.8% of men testing PSA-positive). From these 17,543 biopsies, 10,297 prostate cancers were detected and there were 540 deaths from prostate cancer. Based on these figures the researchers report that 13,309 men (75.9%) had a false positive result, meaning that about three quarters of men who had the biopsy for an elevated PSA turned out not to have cancer.
What interpretations did the researchers draw from these results?
The researchers say that “PSA-based screening reduced the rate of death from prostate cancer by 20%, but was associated with a high risk of overdiagnosis”.
What does the NHS Knowledge Service make of this study?
Whether or not men should be routinely screened for prostate cancer is a controversial issue. This is because the balance or risk and harm is delicate for this cancer. In this trial for example, three quarters of men were told they had a raised PSA blood test, went on to have a biopsy to be told that they did not have cancer. An accompanying editorial calls it a “controversy “that refuses to die.” As such, this large study presents interim results that have been eagerly anticipated by the research and clinical community.
Despite the outcome of the research, an apparent 20% reduction in deaths from prostate cancer there are several features to the study, highlighted by both the researchers and the editorial, that suggest it is too soon to advocate a PSA screening programme based on this research:
- Firstly, as the editorial says, within the same journal are published the results of a US trial of PSA screening with a longer follow up time, but with fewer prostate cancer deaths (174 deaths compared to 540 in the European trial). The US study found that screening for prostate did not have a significant effect on the number of deaths from the disease. This may be because the trial was smaller, but could also be due to the high rates of PSA testing in the control group in the US, which could reduce any differences between the screened and usual care groups.
- The collection of trials reported by the ERSPC had different eligibility criteria, randomisation schemes, screening strategies, intervals and follow-up. If a screening programme was to be set up, it would difficult to say from this study alone which protocol should be followed, such as what age should men begin screening, and how frequently should they be screened.
- The researchers in the ERSPC trial do not report how many of the control group were screened as part of usual care. By testing men with a PSA blood test as part of routine care the difference in detection rates between the population screening programme and usual care group may be reduced.
- The researchers do not report how many biopsies were performed in the screened group and how the treatments offered to the screened group compared to the control group, this means it is difficult to estimate the extent of ‘overtreatment’. The authors do say that those diagnosed by biopsy in the screened group did get more aggressive treatment than those diagnosed by biopsy on the control group. The author of the editorial in the NEJM tried to estimate the extent of this and said that 277 men in every 10,000 had radical prostatectomy in the screened group compared to 100 in the usual care group. This is a measure of the ‘overtreatment’ referred to, but not quantified by, the researchers. It is unclear whether the extent of this treatment was appropriate for the stage of cancer detected by screening and the next set of results should help to clarify this. This is an important point for the researchers to clarify as it may be that more aggressive treatment of the screen-detected cancers could have resulted in the improved survival.
The researchers say that although the results of their trial did indicate a reduction in prostate cancer mortality with screening, “the introduction of population-based screening must take into account population coverage, overdiagnosis, overtreatment, quality of life, cost and cost-effectiveness”. They will report on these aspects later.