New Study Offers Support for Prostate Testing

For men who are weighing the pros and cons of prostate cancer screening, a new study strengthens the evidence that testing can reduce deaths from this cancer, something two earlier large landmark clinical trials appeared to reach different conclusions about.

The findings do not resolve many of the questions that remain about prostate cancer screening, since routine testing can lead to unneeded and potentially harmful treatments in men who do not need it. But it provides more information that experts can use to assess the benefits and risks of screening.

One of the earlier trials, conducted in Europe, found that screening reduced deaths from prostate cancer by 21 percent, though it led many men down a bumpy road toward harmful or even unnecessary interventions. The other trial, based in the United States, found no difference in death rates between men who were randomly assigned to screening and those who were not. Both reports were published in 2009 in the New England Journal of Medicine.

For the new paper, published Monday in Annals of Internal Medicine, a broad consortium of scientists, including some of the investigators in the original studies, reanalyzed the data from the two trials using three different mathematical models. When they made a rigorous comparison between death rates among men who had actually undergone screening and men who had received no screening, the researchers concluded that screening tests reduced prostate cancer deaths by 25 to 32 percent. The reduction was primarily a result of the earlier detection of cancer, the researchers said.

“By the time the U.S. trial started, a lot of the population was already being screened for prostate cancer” as part of routine care in their doctors’ offices, said Ruth Etzioni of the Fred Hutchinson Cancer Research Center, the senior author of the new paper. About three-quarters of the men in the group that were in the comparison group that was not assigned to screening in the trial were still tested as part of their regular medical care, she said.

Meanwhile, some men assigned to the trials’ screening groups did not actually get screened, further muddying comparisons between the two groups, she said.

“Comparing the groups to one another wasn’t answering the question everyone really wanted answered,” Dr. Etzioni said. “What our analysis amounts to is the comparison that people really wanted: screening versus not screening.” (The Annals provides a summary of information for patientson its website.)

Several investigators from the earlier trials, who made their records available for the new analysis, are also authors on the new paper, and scientists involved in the American trial, which goes by the acronym PLCO, said the analysis was an important contribution.