Have the Rules for PSA Testing Changed?

A series of studies were recently published in the New England Journal of Medicine that question the efficacy of widespread PSA testing for the early detection of prostate cancer. One study, The European Randomized Study of Screening for Prostate Cancer (ERSPC Trial), involved 182,000 men in Europe. An American study from the Prostate, Lung Colorectal, and Ovarian (PLCO) Cancer Screening Trial looked at nearly 77,000 men in 10 medical centers.

For the last 20 years, urologists have been practicing in an era that advocates the early detection of prostate cancer. This has led to a decrease in the death rate from prostate cancer. With the decrease in death rates that we have been experiencing, we have to somehow explain what we as physicians are doing that is contributing to this. These studies may suggest that PSA screening may not be responsible for this.

However, several major flaws have been identified with these studies.

1. Most importantly, in the PLCO study, the PSA level threshold for recommending a biopsy was 4 ng/ml. For many years, urologists have been advocating the use of age-adjusted PSA values in order to determine who should undergo a transrectal ultrasound guided biopsy. Using an age-adjusted PSA level results in an increased number of biopsies for patients who are younger; it is this group of patients who have the most to gain by early detection because they have a longer time to survive (with regards to life expectancy). What these studies did is they ignored the group of younger men who had a PSA level below 4 ng/ml, but who may have had prostate cancer that could have been detected early if a biopsy were allowed.

Eliminating this subset of patients from the study group may be skewing the outcome of the study. If a larger number of younger patients were identified and treated for prostate cancer, the data may support benefits of PSA testing.

2. For many years, urologists have been advocating treatment on the basis of survival benefits and disease free status 12-15 years after treatment. By this measure, the two recent studies look at immature data. If the patients were followed for a longer period of time, survival advantages and thus the benefit of PSA testing may become more obvious.

3. The European study has identified a 20% decrease in mortality as a result of widespread PSA testing, a finding they called "marginally statistically significant." Most patients who I have queried since the publication of this study unequivocally wish to proceed with a simple blood test that may result in a 20% decrease in mortality.

4. Lastly, in one of the studies, 52% of the men in the "non-screened" portion had recent PSA testing. Including these patients with normal PSAs in the non-screened portion of the studies substantially reduces the likelihood of prostate cancer related deaths in this part of the study. This potential flaw in the design of the study may be contributing to some of the data analysis.