PSA Screening Does Reduce Deaths, New Analysis Shows

Roxanne Nelson, BSN, RN

September 05, 2017

A new analysis of data from two large prostate cancer screening trials shows that testing for prostate-specific antigen (PSA) does reduce deaths from the disease.

The new analysis was published online September 4 in Annals of Internal Medicine,

"Our estimation of the common effect of screening suggests that it can significantly reduce the risk for prostate cancer death," conclude the authors, led by Ruth Etzioni, PhD, of the Fred Fred Hutchinson Cancer Research Center, Seattle, Washington.

These new data "have finally put to rest the question of whether PSA screening reduces prostate cancer mortality," writes Andrew J. Vickers, PhD, from Memorial Sloan Kettering Cancer Center in New York City, in an accompanying editorial.

[New data] have finally put to rest the question of whether PSA screening reduces prostate cancer mortality. Dr Andrew Vickers

"I say this not to close debate but to refocus it," he continues. "Screening with PSA testing does good by saving lives, but it also causes harm in terms of overdiagnosis and overtreatment. Thus, we need to determine how to screen so that the benefits outweigh the harms.

"Unfortunately, the way screening has been implemented in the United States leaves much to be desired," says Dr Vickers. "The controversy about PSA-based screening should no longer be whether it can do good but whether we can change our behavior so that it does more good than harm."

The controversy about PSA-based screening should no longer be whether it can do good but whether we can change our behavior so that it does more good than harm. Dr Andrew Vickers

Conflicting Results

The two large randomized trials, the ERSPC (European Randomized Study of Screening for Prostate Cancer) and the PLCO (Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial), had offered conflicting results as to whether PSA screening reduces prostate cancer mortality, as previously reported by Medscape Medical News.

The ERSPC showed a significant reduction in prostate cancer death, whereas the PLCO showed none.

But the new analysis concludes that both studies do, in fact, show a reduction in the risk for death from prostate cancer among men who were screened.

The new analysis of data used a mathematical model to account for differences in the two studies. Adjusting for implementation compliance and practice settings, it found that both the ERSPC and PLCO provide compatible evidence that screening reduces prostate cancer mortality.

The new analysis showed a 25% to 31% lower risk in the ERSPC and a 27% to 32% lower risk in the PLCO for dying from prostate cancer among men who were screened compared to men who were not screened.

This is a big turnabout. The PLCO was originally reported as not showing any effect on prostate cancer mortality, a finding that was a big factor on the US Preventative Services Task Force's (USPSTF's) decision to recommend against routine PSA screening in 2012.

However, when the results were originally reported in 2009, the differences between the studies and the studies' limitations were cited. Of particular note, the issue of contamination in the PLCO was raised by Philip Kantoff, MD, director of the Lank Center for Genitourinary Oncology at the Dana Farber Cancer Institute, in Boston, Massachusetts, who participated in a perspective roundtable discussion of both articles held by the New England Journal of Medicine.

Dr Kantoff noted that in the PLCO study, half (52%) of the men who were in the nonscreened (control) arm had in fact undergone PSA testing, as compared to 85% in the screened arm. "As a result, it's not surprising to me that there was only a modest increase in the number of cancers [only 20%] that were diagnosed in the screened arm," he said.

Dr Vickers also pointed to the issue of contamination, referring to it as the elephant in the room. The authors of the PLCO trial described it as comparing "opportunistic vs systematic screening" rather than screening vs no screening, he writes.

"This has led many commentators, myself included, to see different study questions as the explanation for the conflicting results," Dr Vickers says. "Comparing the PLCO and ERSPC is like comparing a trial of 1000 mg of aspirin vs placebo with a trial of 1000 mg vs 900 mg of aspirin."

No Evidence of Screening Differences

In the current analysis, Dr Etzioni and colleagues assessed whether the effects of screening on prostate cancer mortality differed between the two studies after accounting for variations in implementation and practice settings. The also evaluated the effects of screening in both trials relative to no screening.

They first conducted a traditional statistical analysis, in which they combined data from both trials. They then compared hazards of prostate cancer death for the intervention arm vs the control arm.

To overcome differences in implementation between the trials, the authors next conducted extended analyses that accounted for variable screening and diagnostic workups for each group. They used mean lead times (MLTs), defined as the average time by which diagnosis is advanced by screening relative to the date of diagnosis without screening, which they estimated empirically and by use of analytic or microsimulation models.

The authors found that the estimated MLTs were similar in the ERSPC and PLCO intervention groups but were longer for patients in the PLCO control group vs control patients in the ERSPC trial.

Their extended analyses showed no evidence that the effects of screening differed between trials (P = .37 to 0.47 [range across MLT estimation approaches]), but there was strong evidence that the benefit of screening increased with MLT (P = .0027 to .0032).

Overall, their analysis showed that screening conferred an estimated 7% to 9% reduction in the risk for prostate cancer death per year of MLT.

Translation to Clinical Practice?

Medscape Medical News approached several experts for their reaction to this latest article and what they saw as the implications of the study for clinical practice.

Ryan P. Terlecki, MD, FACS, director, Men's Health Clinic, and associate professor of urology and Obstetrics and Gynecology at Wake Forest Baptist Health, Winston-Salem, North Carolina, told Medscape Medical News that although the methodology used in this study was interesting, it will not put this issue "to rest."

"I echo many of the sentiments by Dr Vickers, especially that screening needs to be appropriately tailored. PSA is still a valuable instrument, and we shouldn't 'throw the baby out with the bath water,' " he said.

"One could argue that if the PLCO mirrors US practice and found no benefit to screening, and now it has been found that the benefits are not different between ESRPC/PLCO, that perhaps the USPSTF initial recommendation was reasonable," Dr Terlecki said. He noted that he is not making that argument and that it it still is "all in how you look at numbers."

An important question is what percentage of men died from prostate cancer prior to PSA screening relative to after the initiation of such testing. For both groups, it is on the order of 3% of patients, he pointed out. "This is despite the fact that prostate cancer is nearly ubiquitous as men reach the later years of life, based on autopsy studies performed and published years ago in Detroit by a pathologist named Wael Sakr."

A second issue with this study is that the percentage reduction in prostate cancer deaths is based on regression analyses, whereas absolute numbers are needed when considering population-based screening.

"Another area that has yet to be adequately studied is how men are labeled as dying 'from' prostate cancer," Dr Terlecki said. "I suspect that once they've been given the diagnosis, there is a chance to artificially associate the disease with their death once they expire, even if it was unrelated or only loosely related."

Finally, as the study authors acknowledge, they did not account for stage and grade, which are important factors.

He concurred with Dr Vickers that PSA remains a valid instrument in individualized screening, "especially as we get better at managing competing causes of death and patients are living longer.

"As someone who manages a prostate cancer survivorship program, the quest towards quality improvement remains at full throttle," Dr Terlecki added.

Jonathan E. Shoag, MD, from New York-Presbyterian Hospital, New York City, who conducted a reanalysis of the PLCO trial last year and whose findings raised more doubt about the trial and its fundamental conclusions, also commented on the current study.

"Up until recently, there was the widespread belief in the medical community that PSA screening did not save lives," he said. "This idea was based nearly entirely on underestimation of the degree of PSA testing in the control group of PLCO, which was only recently clarified."

Dr Shoag emphasized that this latest study is important, "as it shows that when accounting for control-group contamination and other study differences, both trials actually show a comparable mortality benefit to screening. This should serve as the nail in the coffin to the argument that screening does not prevent prostate cancer mortality."

This should serve as the nail in the coffin to the argument that screening does not prevent prostate cancer mortality. Dr Jonathan Shoag

He added that although the current study focused on mortality, men with metastatic prostate cancer may suffer symptoms for years, so it is important to look beyond just death as an endpoint.

Dr Shoag also agreed with Dr Vickers that there are ways to optimize the benefits of screening. "There are certainly harms, and those must be weighed. However, the question is, by whom?," he said. "While patient preference is paramount, I do think we are obligated to provide clear guidance on the harm/benefit calculus ― of course with the caveat that this assessment is somewhat based on opinion, and is not definitively proven."

Therefore, rather than use the terminology of "shared decision making," Dr Shoag explained that he would "argue that otherwise healthy men should obtain at least a baseline PSA level with subsequent screening tailored to their individualized risk."

Kenny Lin, MD, MPH, a family physician at Georgetown University Medical Center in Washington, DC, does not feel that this article changes much. "The USPSTF's draft recommendation statement already assumed that PSA screening reduces prostate cancer mortality by the degree that was seen in the ERSPC trial, so what this study adds is an explanation of why PLCO was not able to detect a similar benefit," he told Medscape Medical News.

"However, I still think that data from real-life study participants should trump model projections," Dr Lin explained. He pointed to the latest publication from the PLCO trial (Cancer 2017;123:592-9), which showed a total of 255 deaths from prostate cancer in the screening arm and 244 deaths from prostate cancer in the control arm.

Although is was possible that contamination could "obscure a potential mortality benefit, it shouldn't lead to more people dying in the screening group than the control group," he pointed out.

It's an interesting analysis, but I don't think it changes the benefit-harm equation of PSA screening. Dr Kenny Lin

The bottom line is that "it's an interesting analysis, but I don't think it changes the benefit-harm equation of PSA screening," Dr Lin surmised. "Small absolute benefit at best is still outweighed by harms of unnecessary treatment for most men."

The study was funded by the National Cancer Institute. The editorial was funded in part by the Sidney Kimmel Center for Prostate and Urologic Cancers, a Research Excellence grant from the National Cancer Institute, and a National Institutes of Health/National Cancer Institute Cancer Center support grant to Memorial Sloan Kettering Cancer Center. Several of the study authors report relationships with industry, as noted in the article.

Ann Intern Med. Published online September 4, 2017. Abstract, Editorial


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.