A Critical Analysis of Two Randomized Trials
In March 2009, the results of two long awaited trials were published in the New England Journal of Medicine.1,2 One said that screening with PSA reduced deaths from prostate cancer by up to 27% and the other claimed it didn’t work.
What are we to believe? Like many things “the devil is in the details” and understanding these details could save your life!
First, the positive trial was a landmark study carried out in seven European countries that studied 162,000 men who were randomized to PSA screening every four years versus no screening.1 With long-term follow up out to fourteen years, there was a 20% decrease in deaths from prostate cancer in the group of men assigned to screening. However, since only 85% of these patients actually underwent screening, if one includes only the men who were actually tested, the decrease in prostate cancer deaths is 27%.
This reduction in death from prostate cancer is similar to the 30% reduction in mortality from breast cancer in women who undergo mammography and the 33% reduction in prostate cancer mortality that occurred in the United States between 1994 and 2003 following the introduction of PSA screening. Thus, the results from the European study support other findings and unequivocally demonstrate that PSA testing can save lives.
The second trial,2 which was carried out in the United States, was half the size of the European trial. It compared screening with PSA every year for six years with no screening thereafter versus no planned screening. It showed no improvement in prostate cancer mortality at 7 years.
In the many sound bites on television and reports in print media proclaiming “no effect”, the words “seven years” were conveniently deleted. This is the major flaw in this study. Death from prostate cancer at seven years is meaningless: 1) screening and aggressive treatment are typically reserved for individuals with at least a 10 year life expectancy; 2) any patient who dies within seven years of diagnosis has advanced non-curable disease at the time of diagnosis and would not benefit from PSA screening; and 3) in the positive European trial mentioned above, there was also no improvement in survival at seven years.
The U.S. trial also failed to achieve some of the important milestones that one would expect from a screening trial if it were successful. For example in trials for breast or colon cancer, screening leads to an increase in the number of cases diagnosed and a decrease in the number who are not curable. In the European prostate cancer trial, with screening there was a 71% increase in the number of cases and a 41% decrease in the number of men with incurable disease. For this reason with longer follow up, the favorable impact on mortality is likely to increase. In contrast, in the U.S. study where there was only a 17% increase in the number of new cases and no decrease in advanced disease. For this reason, with longer follow up of the men in the U.S. trial, unfortunately the results will not change.
What’s the problem with the study carried out in the United States? First, it did not test screening versus no screening – it just compared more screening versus a little less screening. In the screening arm, 85% of patients underwent PSA testing compared to 52% in the controls. That’s right, only a 33% difference!
Furthermore, 44% of the men who entered the trial already had 1 or more prior PSA tests. Consequently these men were not only less likely to have cancer, but also less likely to have life-threatening disease. This explains why there were so few cancer deaths in either arm of the study.
Finally, the American study used an outdated cut point for PSA to trigger a biopsy (greater than 4.0 ng/ml versus 3.0 ng/ml in the European trial) and only 30% of the men who developed a PSA greater than 4.0 ng/ml while in the trial actually underwent a biopsy! If most of the men with elevated PSA levels never underwent a biopsy, how can anyone expect this trial to show that screening saves lives!
Why does this study have so many flaws? If they had set out to design a study to discredit PSA testing it would have been difficult to do a better job. However, giving them the benefit of the doubt, maybe they were just trying to simulate what would happen if every man in the United States had a PSA performed but few followed up with a biopsy or treatment. If that were the question, I think we could have already guessed at the answer without spending $110 million.
Granted, it was more difficult to carry out a trial in the U.S. where PSA testing was already well established compared to Europe where random PSA testing was infrequent. However, why didn’t the investigators from the National Cancer Institute admit the shortcomings of their study rather than promote their flawed findings as the Holy Grail? I don’t know why anyone would do this but you can understand where they are coming from by looking at their many interviews on television and in the print media where they state their objections loud and clear.
They believe that the wide-spread use of PSA testing will encourage many men who may never need treatment to suffer unnecessary side effects. In the European study it is estimated that to prevent one prostate cancer death at ten years, 1,400 men would need to be screened and an additional 48 men would need to be treated. Ten years is the earliest time at which one would expect any benefit and if one looks at a 50 year old man who is going to be alive for another 35 years, those odds would be entirely different. Also, with longer follow up in the European trial these odds will improve.
However, if screening for prostate cancer is ever going to receive popular support, it will be necessary to avoid over-diagnosis in men who are unlikely to have a survival benefit (men who are too old or too ill to live longer than ten years) and to avoid over-treatment in men over age 65 who have low volume disease. Most of all, it is imperative for us as physicians to continue to improve the quality of treatments to reduce their morbidity. If one day it were possible to reduce these side effects to a minimum, the debate would end.
What is the take home message? If you are the kind of person who doesn’t wear a seat belt nor goes regularly to the dentist or your family doctor for a check-up and are not worried about dying from prostate cancer, do not undergo PSA testing. On the other hand if you are a healthy man age 55-69 who does not want to die from prostate cancer, the European trial provides conclusive evidence that PSA testing can save your life.
- European Randomized Study of Screening for Prostate Cancer (ERSPC): Schroder FH, Hugosson J, Roobol MJ, et al. Screening and prostate-cancer mortality in a randomized European study. N Engl J Med 2009;360:1320-8.
- Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial (PLCO): Andriole GL, Grubb, III RL, Buys SS, et al. Mortality results from a randomized prostate-cancer screening trial. N Engl J Med 2009;360:1310-9.