Prostate cancer: Too much dogma, not enough data
The article on prostate-specific antigen (PSA) testing from Drs. Jones and Klein1 in this issue of the Cleveland Clinic Journal of Medicine illustrates an important phenomenon in our recent approaches to management of prostate cancer: dogma often outweighs real data.
DOGMA 1: PSA ≤ 4 IS NORMAL AND PSA > 4 IS ABNORMAL
As Drs. Jones and Klein emphasize, a single PSA value does not necessarily indicate cancer is present or absent, although we should note that they are speaking predominantly of PSA values lower than 10 μg/L.
In reality, however, a confirmed blood PSA concentration of 100 μg/L is effectively diagnostic of prostate cancer, and I would be quite prepared to treat a patient for prostate cancer in an urgent setting (eg,spinal cord compression from sclerotic bone metastases) based on that confirmed PSA level without a tissue diagnosis. It is important to consider the costs and benefits of treatment and the impact of delay when making decisions of this type. In the setting of imminent spinal cord compression, the results of waiting for a diagnosis by conventional means (ie, by biopsy) are disappointing,2 and delay in care can be an important factor. Thus, we should not ignore the implications of a markedly raised PSA level when the clinical context is appropriate. The conundrum is determining at what cutoff the PSA allows that type of decision to be made without a tissue diagnosis.
DOGMA 2: PROSTATE SCREENING IS BENEFICIAL
An equally vexing issue is community-wide screening for prostate cancer. Screening is the assessment of symptom-free people in the general population for a particular disease, and for it to be successful, it must identify disease early in its course, and early identification of the disease must result in decreased morbidity of treatment or a reduced overall mortality rate.Current dogma is that prostate cancer screening is good for the community at large.
It seems intuitively sensible and logical tha tassessing healthy, symptom-free men for prostate cancer should be a good idea and should lead to earlier diagnosis and an increased chance of cure. The evidence in favor of routine screening includes “first principles,” common sense, the suggestion that death rates from prostate cancer have fallen in various countries since such approaches have been introduced, and the observation of stage migration (with a greater proportion of initial presentations with earlier-stage disease) in association with these screening exercises.
However, level-1 evidence to support this hypothesis is simply nonexistent—there have been no completed, well-designed randomized trials that demonstrate improved survival from the introduction of routine community screening for prostate cancer with digital rectal examination or PSA measurement. To know the true usefulness of community screening for prostate cancer, we must wait until the ongoing European randomized trial of screening is completed.
DOGMA 3: PROSTATE SCREENING IS WORKING
Although the concept of screening for prostate cancer is very appealing, we should not lose sight of the fact that absolute death rates from prostate cancer have fallen remarkably little in the United States since the introduction of our current screening techniques.
The absolute number of deaths from prostate cancer in the United States has hovered in the range of 26,000 to 30,000 per year since the 1980s, when PSA testing became widespread. In 1985, the American Cancer Society estimated that there were 25,500 deaths from prostate cancer3; in 2007, the estimate was 27,050 deaths,4 hardly a quantum leap forward!
In addition, even if one introduces changes in the incidence of prostate cancer and the aging of the community into the argument and thus increases the denominator for calculation of death rates, the diagnosis and treatment of prostate cancer have improved in many other ways besides screening, including better noninvasive imaging and staging techniques, refined methods for pathological classification, advances in surgery and radiotherapy, hormonal adjuvant therapy for locally advanced tumors, improved chemotherapy, and better support technologies. Thus, it is difficult to attribute any perceived major improvement only to screening.
