No Definitive Answers from Initial Results of Prostate Screening Trials
Prostate-specific antigen is a substance produced by the prostate. It can be detected in the blood of adult men and may be found in higher levels in men who have prostate cancer and other prostate conditions.
Regular screening for prostate cancer with the prostate-specific antigen (PSA) test did not reduce deaths from the disease after 7 years of follow up, according to the long-awaited results from two large randomized trials, one conducted in the United States and the other in Europe.
After 9 years of follow up, however, the European trial found that PSA screening every few years did produce a small mortality benefit. The 10-year follow-up from the American trial is not yet complete, although investigators said the results to date are consistent with the 7-year follow-up data.
The findings, from the NCI-led Prostate, Lung, Colorectal and Ovarian (PLCO) Screening Trial and the European Randomized Study of Screening for Prostate Cancer (ERSPC) trial, were published online in the New England Journal of Medicine on March 18.
In both studies many more men were diagnosed with prostate cancer in the active screening arms than in the control groups, which PLCO leaders said confirms long-standing concerns that PSA screening often leads to “overdiagnosis” and overtreatment of cancers that would never have been life-threatening.
“There’s no question that there are men who are getting the side effects [of treatment] without getting any of the benefit,” said PLCO co-author Dr. Edward Gelmann of the Herbert Irving Comprehensive Cancer Center in New York, during a press briefing on the results.
The findings, he continued, support recommendations issued last August by the U.S. Preventive Services Task Force against routine PSA screening of men aged 75 and older, who are likely to have a life expectancy of less than 10 years. For younger men, he continued, the findings offer a good foundation for doctors and patients to make individualized decisions based on factors such as PSA results and the risks and benefits of biopsy and treatment.
The prostate component of the PLCO trial randomized nearly 77,000 men to annual screening with a PSA test for 6 years and digital rectal exam (DRE) for 4 years, or to the usual care of their physicians, which included PSA screening in more participants than the study leaders had expected. Mortality due to prostate cancer was uncommon and effectively the same in both groups at 7 and 10 years of follow up.
NIH Challenge Grant: Biomarkers and the PLCO Biorepository
Under an NIH Challenge Grant available as a result of the American Recovery and Reinvestment Act, NCI is soliciting applications for funding to use available biospecimens from the PLCO Screening Trial for prospective studies focused on the discovery and validation of cancer early detection biomarkers (RFA 03-CA-109).
According to Dr. Christine Berg, the PLCO biorepository is ideally suited for use in nested, case-control biomarker discovery and validation studies. “This is an excellent example of NIH using new funds to leverage an unparalleled resource that we’ve already developed,” she said.
Deadline for applications is April 27. For more information, contact Dr. Berg at email@example.com.
That doesn’t mean that with longer follow-up a mortality benefit might not be seen, said Dr. Christine D. Berg, the trial’s lead investigator. Prostate cancers among men in the “usual care” group tended to have higher Gleason scores (between 8 and 10), an important predictor of disease aggressiveness. “So, there may be a survival difference that emerges after more than 10 years, but it would be a small difference,” she said.
At 9 years of follow-up, the ERSPC showed a 20 percent reduction in mortality among men aged 55 to 69 (at enrollment into the study) who underwent PSA screening every 4 years. Much like what was seen in PLCO, the mortality rates between men in the screening and control arms were nearly identical for the first 7 years. It was around that time in ERSPC, however, when the rates began to diverge.
The trials’ discrepant results could be attributed to several key differences in their designs, said PLCO co-author Dr. Gerald Andriole of the Washington University School of Medicine. Not only was ERSPC significantly larger (182,000 participants) and did it use a different screening interval than the PLCO, but most centers involved in the study also had a lower PSA cutoff to refer participants for clinical follow up (3 ng/ml versus 4 ng/ml).
In an accompanying commentary in NEJM, Dr. Michael J. Barry of Harvard Medical School wrote that even if the small mortality benefit seen in the ERSPC is real, it is “important to remember that the key question is not whether PSA screening is effective, it’s whether it does more good than harm.”
Proponents and skeptics of regular PSA screening alike have warned about the overdiagnosis and overtreatment of prostate cancers linked to PSA screening. Estimates of overdiagnosis have ranged from 20 to 80 percent in various studies. A modeling study published March 10 estimated that PSA-related overdiagnosis in the United States could be as high as 42 percent. In the PLCO trial the estimated risk of overdiagnosis at seven years was 22 percent, while in the ERSPC trial it was more than 70 percent, a difference that Dr. Barry said was most likely the result of the smaller difference between screening intensities in the two arms of the PLCO than in ERSPC.
The extent of overdiagnosis, stressed Dr. Andriole, should help to “temper those knee-jerk reactions” to immediately perform biopsies and pursue aggressive treatments.
Looking forward, Dr. Berg said she is hopeful that clinical decision models can be developed based on data from these two trials and other studies, and that new molecular biomarkers will be found that accurately identify aggressive cancers that need to be treated. NCI’s Early Detection Research Network is supporting research on a number of these markers.
Dr. Berg also stressed that the PLCO biorepository houses 2,000 tumor samples as well as blood samples, making it a valuable tool for continued biomarker discovery and validation. (See sidebar.)
Although there has been significant progress to date on the biomarker front, she said, “I think the pace of discovery and the rapidity with which we will get these answers will improve.”