By Annette M. Boyle
SEATTLE, WA – For the VA, the ongoing discussion about use of the controversial prostate-specific antigen (PSA) screening for prostate cancer is far from academic.
Veterans Health Administration physicians diagnose more than 12,000 new cases of prostate cancer each year. Yet, VA’s decision on how to proceed with PSA screening likely will be informed by the spirited debate being waged in medical journals.
With the value of screening in question, the VHA is considering revision of its current recommendations to offer PSA screening to asymptomatic men, according to Michael Kelley, MD, national program director for Oncology, Specialty Care Services, Patient Care Services for the VHA and chief of Hematology-Oncology at the Durham, NC, VAMC.
Some researchers, such as those at Seattle’s Fred Hutchinson Cancer Research Center (FHCRC), suggest that revision is a better approach than abandonment of screening. Two strategies — raising the threshold for referral for biopsy and increasing the interval between PSA screens — achieve similar reductions in prostate cancer deaths as traditional annual screening, while substantially reducing the harms from overdiagnosis, according to a recent report in the Annals of Internal Medicine. 1
Other researchers disagree, however.
In a Point-Counterpoint exchange in Medical Care, Joy Melnikow, MD, MPH, director of the Center for Healthcare Policy and Research, University of California, Davis, and her colleagues point out that clinical trials in the United States and Europe followed 250,000 individuals for 11 to 13 years, with the U.S. cancer screening trial (Prostate, Lung, Colorectal and Ovarian or PLCO) , which they note was highly representative of the population, showing no reduction in mortality as a result of PSA testing. 2
Model vs. Trials
Ruth Etzioni, PhD, of FHCRC and her colleagues responded to the challenge imbedded in the U.S. Preventive Services Task Force’s (USPTF) recommendation last year against PSA-based screening for prostate cancer. At the time, the task force noted that “it is not known whether an alternative approach to screening and management of screen-detected disease could achieve the same or greater benefits while reducing the harms.” 3
The FHCRC researchers set out to see whether a different strategy could produce better results by constructing a statistical model that enabled comparison of 35 screening strategies that differed by ages at which screening started and stopped, the threshold for biopsy referral, and screening intervals. The incidence component of the model linked increase in PSA with disease progression. The mortality component included disease-specific and other-cause survival , with disease-specific survival depending on age, stage, and grade of cancer at diagnosis.
“The model enables us to study the comparative harms and benefits of competing policies based on PSA. It’s unique in that it allows us to understand how the biomarker links to curability. You can’t do a clinical study that compares dozens of strategies with different PSA levels and different rates of change. A mathematical representation in which marker growth is linked to whether cancer is localized in the prostate or distant allows you to move the point of detection earlier, and we know how that corresponds to curability,” Etzioni told U.S. Medicine.
The model-based survival benefit on the results of the European Randomized Study of Screening for Prostate Cancer (ERSPC) which showed a reduction in prostate cancer from 2.86% to 2.15% for annual screening of men aged 50 to 74 years.
“Trials can only test one policy, but the European trial results are supported by population results. If you don’t believe the trial, don’t believe our model,” said Etzioni.