Gone Astray: US Prostate Cancer Screening Trial

Deaf Ears About 'Contamination' in 2009?

Nick Mulcahy

January 06, 2012

January 6, 2012 — New longer-term follow-up from the American Prostate, Lung, Colorectal, and Ovarian Cancer Screening (PLCO) Trial continues to indicate that annual prostate-specific antigen (PSA) screening for 6 consecutive years does not provide men with a mortality benefit.

The massive study now has 13-year data on 57% of the men who originally enrolled. The updated data are reported online today in the Journal of the National Cancer Institute.

Maybe there is a benefit.

Despite this conclusion, the study's senior author said in an interview with Medscape Medical News this week that "maybe there is a benefit."

"The lack of a reduction in mortality could be explained, in part, by screening in the control arm," said Philip Prorok, PhD, from the National Institutes of Health in Bethesda, Maryland.

Dr. Prorok reported that the 76,000-men trial "didn't have a pure control group," that many of men in the control group underwent PSA testing, and that some might have gained a mortality benefit from the testing and subsequent biopsy and treatment.

The investigators did not plan it that way.

Dr. Prorok explained that, beginning in 1993, men were randomized to an intervention group (receiving annual PSA tests for 6 years and 4 digital rectal exams) or to a control group (receiving no testing). But then "PSA testing took off" in the United States, he said. By year 6 of the study, 53% of the control group had had at least 1 PSA test. The authors call this "opportunistic testing."

Dr. Prorok said that when the PLCO trial was designed in the early 1990s, PSA testing was rarely used. But then urologists promoted the testing nationally "without any justification of benefit."

PLCO essentially became a different randomized trial, comparing annual PSA testing plus digital rectal exams (intervention group) with usual care and opportunistic testing (control group).

Because of this "contamination" of opportunistic testing in the control group, the trial data are so flawed that the study lacks clinical significance, critics of the study told Medscape Medical News.

[PLCO is] a study that is uninterpretable.

"People are putting a lot of weight on a study that is uninterpretable," said Anthony D'Amico, MD, from Brigham and Women's Hospital in Boston, Massachusetts, referring, in part, to the US Preventative Services Task Force, which recently recommended that healthy men not routinely be offered screening, using the PLCO data as part of their argument.

Dr. D'Amico also pointed out that 15% of the men randomized to screening in PLCO never attended a screening. Thus, the trial was a comparison of 2 groups: one in which 85% of men were screened and one in which 53% of men were screened. In the end, the PLCO study lost its statistical power, he noted. "It's a dataset that you cannot make any conclusions about," Dr. D'Amico summarized.

The problem of opportunistic testing of men in the control group is probably worse than reported, said another expert.

"The contamination rate has probably increased," said Philip Kantoff, MD, from the Dana-Farber Cancer Institute in Boston, referring to the fact that the study's published rate of opportunistic testing was recorded by the investigators only after 6 years.

PSA testing is "enormous" in the United States, Dr. Kantoff observed, and would have afforded the men in the control group repeated opportunities after those first 6 years to learn their PSA level and possibly act on that information to seek biopsy and treatment.

Investigator Dr. Prorok said that the PLCO team will update information on the contamination rate; he was unsure if it had worsened.

He also disagreed with Dr. D'Amico's assessment that the trial is uninterpretable.

"The study provides information about what happens when you add annual screening on top of what a population is already getting in terms of screening," he said, referring the study's eventual and accidental design.

New Results

The PLCO results indicate, among other things, that PSA screening leads to overdiagnosis of prostate cancer, said Dr. Prorok.

At 13 years, 4250 participants have been diagnosed with prostate cancer in the intervention group, compared with 3815 in the control group.

The cumulative incidence rates for prostate cancer in the intervention and control groups were 108.4 and 97.1 per 10 000 person-years, respectively, resulting in a relative increase of 12% in the intervention group (relative risk [RR], 1.12; 95% confidence interval [CI], 1.07 to 1.17), the authors report. "This overdiagnosis from screening is probably an underestimate because there was probably some overdiagnosis in the control group too [because of opportunistic PSA testing]," said Dr. Prorok.

There were also more prostate cancer deaths in the screening group than in the control group (158 vs 145), Dr. Prorok noted. This raises the question of whether or not screening might actually increase the risk for death, he said. However, the difference was slight and could be related to chance, he added.

Specifically, the cumulative mortality rates from prostate cancer in the intervention and control groups were 3.7 and 3.4 deaths per 10,000 person-years, respectively, resulting in a nonstatistically significant difference between the 2 groups (RR, 1.09; 95% CI, 0.87 to 1.36).

Authors Always Pointed Out the Contamination

The PLCO investigators first published their results in 2009, and concluded that "after 7 to 10 years of follow-up, the rate of death from prostate cancer was very low and did not differ significantly between the 2 study groups." (N Engl J Med. 2009; 360:1310-1319).

This conclusion echoes the findings in the new paper by the PLCO team.

However, back in 2009, they also pointed out that many of the men in the control group received prostate cancer screening, which might have obscured the results and hidden any benefit of screening.

"The level of screening in the control group could have been substantial enough to dilute any modest effect of annual screening in the screening group," they wrote.

But this message was largely lost in the media coverage of the study, which generally declared that the PLCO study was a "negative" study, and that its counterpart, the European Randomized Study of Screening for Prostate Cancer (ERSPC), which was published simultaneously, was a "positive" study.

The ESPRC found a significant decrease in the rate of prostate cancer mortality between the screening and control groups (rate ratio, 0.80; 95% CI, 0.65 to 0.98; = .04), or a relative reduction of 20% in the rate of death from prostate cancer.

Importantly, the ESPRC had less contamination because the percentage of men in the control group who got an opportunistic PSA test in the European trial was significantly lower, said Dr. Kantoff. "PSA testing is not as widely available there as it is here," he said.

There is a reduction in mortality from prostate cancer with screening.

"The aggregate information is that there is a reduction in mortality from prostate cancer with screening," said Dr. Kantoff, referring to the 2 major randomized trials of prostate cancer screening.

However, the ESPRC also has multiple flaws, all of the experts interviewed for this article pointed out.

The jury is still out on the efficacy of the whole process.

Dr. Kantoff summarized what he believes is the current state of thinking about prostate cancer screening. "The jury is still out on the efficacy of the whole process," he said. That process includes not only PSA testing but subsequent biopsies, monitoring, and the many decision points that lead to treatment.

J Natl Cancer Inst. Published online January 6, 2012. Abstract


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.