Prostate-Specific Antigen, Prostate Cancer Screening, and the Pathologist: What Should Be Our Role?

Abstract

The heated national controversy about prostate-specific antigen (PSA)–based prostate cancer screening raises important questions about guidelines and patient-centered care. It also suggests opportunities for pathologists and laboratory medicine. Most readers will be familiar with the following background: Since the introduction of the PSA test in the 1970s, it has become one of the most popular preventive care tests in the country. Yet the evidence for reduced mortality attributable to PSA screening is, at best, weak. Two recently published, large, randomized studies reported conflicting results. The European Randomized Study of Screening for Prostate Cancer found a modest reduction in mortality with screening; the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial did not. In addition to the uncertainty regarding benefits, there has also been increasing appreciation of harms attributable to prostate cancer screening. Biopsies and prostatectomies are not benign procedures. Large numbers of patients are required to undergo these procedures with their associated risks of infection, impotence, and incontinence for every one patient whose life may be extended through early detection of a cancer. Accordingly, guideline development organizations, notably including the American Cancer Society and the American Urologic Association, have issued expert opinions recommending that prostate cancer screening should only be entertained in men who might reasonably stand to benefit and who are in a position to make a fully informed choice that incorporates their personal risk/benefit profile with their personal values. These appear on the surface to be very reasonable recommendations, but, in fact, they reflect an almost impossible aspiration in today’s health care environment. A recent study assessed whether primary care physicians could properly interpret basic statistics (mortality and survival time) of hypothetical cancer screening tests to advise their patients. Most could not. How do the American Cancer Society and the American Urologic Association think that patients will make fully informed choices if even their doctors misunderstand the implications of screening test results? Clinical laboratories, like most entities in US health care, are designed primarily for ‘‘cookbook’’ medicine, not personalized medicine. Laboratories and pathologists are paid based on volume, not outcomes, and certainly not based on time spent educating physicians and patients. Physicians and patients are expected to integrate complex, probabilistic data into sound decisions yet, like all humans, are innately, cognitively ill-equipped to do so. Even if physicians and patients could process such data, highquality effectiveness data based on real-world clinical experience isn’t even available for most laboratory tests. The path forward will depend on leadership from pathology. If pathology does not create solutions, other specialty organizations may step up to fill the gaps, to the detriment of the professional status of pathologists. First, we need comprehensive systems of diagnosticinformation support for doctors and patients. Laboratories and pathologists need to move beyond our historic role of simply providing test results and develop ways to communicate results in clinical and probabilistic context. This might include models of risk and benefit that combine available patient-specific data, such as age and history (and, in the future, genomics), together with the results of clinical effectiveness studies to estimate risk and benefit for individual patients. This, in turn, will require a major commitment to measuring clinical effectiveness. Most tests on our laboratories’ menus were initially popularized based on preliminary, clinical correlations, and relatively few have attained high-level clinical evidence to guide their use. Randomized, controlled studies for all diagnostic tests and scenarios may be an impractical goal, but other methods including registry-based studies and health services research can often shine a great deal of light on quantitative, clinical impact. Building such systems will likely require a number of years. Something we can do more quickly is to discontinue Accepted for publication May 25, 2012. Published as an Early Online Release July 16, 2012. From the Department of Pathology and ARUP Laboratories, University of Utah, Salt Lake City. The author has no relevant financial interest in the products or companies described in this article. doi: 10.5858/arpa.2012-0241-ED Reprints: Brian R. Jackson, MD, MS, Department of Pathology and ARUP Laboratories, University of Utah, Salt Lake City, 500 Chipeta Way, Salt Lake City, UT 84108 (e-mail: brian.jackson@aruplab.com).