Untangling Evidence and Guidelines in Prostate Cancer Screening

Untangling Evidence and Guidelines in Prostate Cancer Screening

Ethan Basch, MD
Phillip Febbo, MD
Andrew Vickers, PhD

Screening of healthy men for prostate cancer with the prostate-specific antigen (PSA) test has been among the most debated areas in cancer prevention research and policy. Clinical practice guideline recommendations from major professional organizations and government panels in the United States have shifted over time and have appeared to disagree with each other, although a nuanced reading of recent recommendations finds more agreement than disagreement. (See Sidebar and Table.)

Assessment of the value of a screening test after it is so widely disseminated in clinical practice has been a particular challenge. Nevertheless, the current evidence can be used to guide clinical decisions that minimize the risk of harms and maximize potential benefits for an individual man.

Evidence of Benefits and Harms

After PSA screening came into practice in the United States in the 1980s, the incidence of prostate cancer diagnosis immediately increased, with mortality rates subsequently declining after a lag of approximately 10 years.1 This observation has been cited by some as evidence of a survival benefit, while others are skeptical of any association.

The recently reported European Randomized Study of Screening for Prostate Cancer (ERSPC) found that PSA screening reduces death from prostate cancer, with a 21% relative risk reduction at 11 years median follow-up (or one prostate cancer death per 1,000 men randomly assigned to screening).2 Although it has been noted that there was no difference in overall mortality, this is a virtually infeasible endpoint for screening trials because the large amount of statistical noise of deaths from other causes means that sample sizes in the millions would be required. The Göteborg screening trial found approximately 300 screenings were needed to prevent one death from prostate cancer at 14 years.3

A statistical modeling study estimated that a program that screened men from age 55 to 69 would require 110 men to be screened and five diagnosed in order to avoid a single prostate cancer death over the course of a man’s lifetime.4 The Prostate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screening Trial was a randomized controlled trial that found no survival benefits after 13 years of follow-up.5 This was not a trial of screening versus no screening, but rather of “systematic” versus “opportunistic” screening, and there were high rates of screening in the control group. As such, the PLCO does not provide an estimate of the effects of choice of screening compared with avoidance of screening.

PSA screening detects many indolent cancers that do not need immediate treatment (i.e., overdiagnosis).6 Despite recent recommendations encouraging surveillance in such cases,7 most of these indolent cancers are treated with surgery or radiation (i.e., overtreatment),8 leading to serious harms including urinary incontinence, erectile dysfunction, pain, infections, hospital readmission, and, rarely, death.9 Moreover, PSA screening generates many false positives because of low specificity, which can lead to unnecessary biopsies that carry a risk of bleeding, infection, pain, urinary symptoms, and hospitalization.10,11 Given these harms, any screening approach must be carefully implemented to limit the negative consequences.

Making Individual Clinical Decisions Based on Evidence

Despite limitations, current evidence can help guide clinical decisions that minimize the risk of harms and maximize potential benefits for an individual man, as well as make use of information from recent clinical practice guidelines.

Older Men: Avoid Screening

Studies consistently demonstrate that routine PSA screening and treatment are not beneficial in older men or those with a limited life expectancy. On a population level, the harms clearly outweigh potential benefits.

In the ERSPC trial, men who started screening at age 70 or older did not benefit;2 in the statistical modeling study, screening men older than age 69 did not lead to additional benefit.4 There is also evidence from randomized trials that the benefits of surgery decrease in men older than 65.12,13 Risk of harms is substantial in this population, and risk of overdiagnosis increases with age.14

Interpreting Differences between PSA Screening Guidelines

In 2012, ASCO issued recommendations against screening in older men but in favor of informed decision making in younger men (see Table). Last month, the American Urological Association (AUA) published a new guideline that substantially changed its previous pro-screening stance, now discouraging screening in all men except those between ages 55 and 69, for whom informed decision making is recommended. The AUA recommendations are remarkably similar to a recently published guidance statement from the American College of Physicians (ACP).

Approximately 1 year ago, the U.S. Preventive Services Task Force (USPSTF) issued a recommendation discouraging routine PSA screening in all men. This controversial recommendation was criticized for grouping younger with older men.24 Notably, the underlying message of this recommendation was softened in a statement by the USPSTF Co-Chair, which suggested that men should undergo informed decision making before considering a PSA test: “…before getting a PSA test, all men deserve to know what the science tells us about PSA screening: there is a very small potential benefit and significant potential harms. We encourage clinicians to consider this evidence and not screen their patients with a PSA test unless the individual being screened understands what is known about PSA screening and makes the personal decision that even a small possibility of benefit outweighs the known risk of harms.”25

The common threads of these recommendations include careful attention to scientific evidence, consideration of harms as well as potential benefits, and appreciation of challenges associated with informed decision making. The ASCO and AUA guidelines particularly address potential differences in the risk–benefit ratio between different age groups, with different perspectives on younger men (a population for which scientific research is lacking, necessitating the use of expert consensus to form recommendations).

The distinction between advising against screening in particular men versus advising for informed decision making to appreciate that harms may outweigh benefits is subtle, and depends to some extent on the judgment of guideline developers about the risk–benefit balance based on available evidence. The equivocal nature of the evidence and uncertain balance of harms and benefits (i.e., clear evidence of harms but less clear magnitude of benefits) is attributable for some of the differences in guideline language.

ASCO recommends against screening in men with a life expectancy of fewer than 10 years,15 and the American Urological Association (AUA),16 and American College of Physicians (ACP)17 recommend against screening in men older than 70 or with a life expectancy of fewer than 10 to 15 years. The U.S. Preventive Services Task Force (USPSTF) recommended against PSA screening regardless of age,9 but has been criticized for this grouping of younger and older men, as well as for methodological limitations.25

Younger Men: Informed Decision Making, PSA Testing

In younger men, the considerable harms associated with PSA screening must be weighed against the modest survival benefits observed in existing studies, with the understanding that trials are not yet fully mature and that, because of methodological problems such as crossover, measured benefits may be underestimated.

Decision aids can assist patients in understanding the tradeoffs, and can foster patient–clinician communication. Decision aids have been shown to improve understanding of risk, increase patient knowledge, and better align patients’ decisions with their values.18-19

ASCO recommends shared decision making in younger men; however, the AUA now recommends this for men ages 55 to 69 but not in younger men unless there is an increased risk of disease related to family history or African American race (Table). The AUA (and ACP) caveat is largely based on the lack of inclusion of men younger than 55 in the major randomized trials, and the relatively low prevalence of prostate cancer in this age group. Evidence is lacking in this population, which should be explained in discussions.

Treatment for Those Who Will Benefit

Many screen-detected cancers are low risk and there is growing consensus that these do not necessitate immediate treatment.8 A recent National Institutes of Health State-of-the-Science Conference Statement recommended offering active surveillance to men with low-risk cancers (e.g., PSA below 10 ng/ml; stage T1c or T2a; Gleason 6).7 By improving selectivity about who is treated, the downstream harms related to screening will be reduced, shifting the harm–benefit ratio.

Studies consistently demonstrate superior treatment outcomes associated with higher case volume of practitioners and centers.20 For those patients who choose treatment after an informed discussion, referral should be offered to a high-volume center.

Emerging Tests

Despite its challenges, PSA remains the best serum-based biomarker in prostate cancer. Variations of PSA such as free-to-total PSA ratio may be used to risk stratify patients. Assessment of urine PCA3 (Progensa PCA3 assay), or assessment of prostate biopsy tissue GSTP1 methylation (ConfirmMDx assay), are marketed to help avoid unnecessary repeat biopsies.

The Oncotype DX prostate biopsy-based genomic test recently became commercially available, and is marketed to identify the presence of an aggressive genotype to aid initial treatment decision making.

Notably, none of these novel tests has been prospectively evaluated to assess effect on clinically meaningful outcomes such as survival or quality of life.

In the United States, there is limited regulatory oversight of molecular diagnostic tests. The FDA regulates “companion” diagnostic assays that are used to predict patient responses to approved products (e.g., KRAS). But for non-companion tests, there is no requirement prior to marketing that benefits be demonstrated in survival or quality of life.

Across cancer types, screening and diagnostic tests continue to appear on the market without demonstration of meaningful benefits. But as we have seen with PSA, the ability to detect the presence of a cancer does not necessarily equate with beneficial outcomes—indolent cancers may be detected and the harms of testing can outweigh the benefits.

To address this regulatory loophole, recommendations were recently issued by the Center for Medical Technology Policy (CMTP) to prospectively evaluate the impact of tests on meaningful outcomes prior to payers reimbursing for such tests.21 Similarly, the Patient-Centered Outcomes Research Institute (PCORI) requires that studies of tests funded by PCORI include assessments of meaningful outcomes.22

Without such demonstrations of value prior to marketing, it is difficult to justify paying for new screening or diagnostic tests and the risk of the “next PSA” is real, which is wide proliferation of a test with unclear clinical benefits that is so widely used that controlled trials are challenging or infeasible.

About the Authors: Dr. Basch is an associate professor of Medicine and Director of the Cancer Outcomes Research Program at the Lineberger Comprehensive Cancer Center at the University of North Carolina School of Medicine, Chapel Hill, NC. Dr. Febbo is an Associate Professor of Urology and Medicine at the University of California, San Francisco, San Francisco, CA. Dr. Vickers is an Attending Research Methodologist at Memorial Sloan-Kettering Cancer Center, New York, NY.

ASCO has developed a patient decision aid for prostate cancer screening. Additional decision aids are also available.