New data from the first prospective, randomized, large-scale trial of prostate cancer ever reported indicate screening in the general population may significantly decrease prostate cancer deaths.
"PSA testing saves lives," declared Fernand Labrie, MD, PhD, lead investigator of the Quebec Prospective and Randomized Study, and director of research at Laval University, Quebec. The 8-year study, begun in 1988, randomly divided 46,193 men, ranging in age from 45 to 80 years, between screening and no screening. "There was a 69% decrease in deaths from prostate cancer in the group of men who were screened and treated early for prostate cancer at the time of diagnosis," said Labrie.
Labrie delivered the first presentation of the study findings at the plenary session of the annual meeting of the American Society of Clinical Oncology, held in Los Angeles, Calif, in May.
The study was immediately lauded by some physicians as "innovative" and "forward-looking" and "the first of its kind," while other physicians criticized the design of the trial as "flawed" and "unorthodox."
Study Controversial
"It's a very controversial study," said Peter Boyle, MD, director of the division of epidemiology and biostatistics at the European Institute of Oncology in Milan, Italy. "There are those who are going to love it and those who are going to hate it." But he said the Canadian trial showed both "significant foresight and insight" and has to get a "lot of credit for starting 5 years before any other trial was even contemplated, an excellent idea that was ahead of its time." (A large European study of screening for prostate cancer is expected to complete recruitment next year, while a similar US study also is in the recruitment stage.)
"We still need to know more about whether the Canadian study's methodology and patient selection process influenced the results, as this is a highly controversial issue," said Derek Raghavan, MD, PhD, chief of medical oncology at the University of Southern California Norris Comprehensive Cancer Center, Los Angeles. "It would be premature to set national guidelines based on any 1 study, especially as routine implementation of screening asymptomatic patients will be very expensive to the general community," he said. "This is encouraging research that needs to be confirmed and it is important work."
The study population was drawn from the Quebec City electoral rolls and randomly divided into 2 groups. According to Labrie, age and geographical location were taken into account to avoid bias. Men were invited by letter for screening. Those who responded to the invitation received a prostate-specific antigen (PSA) test and a digital rectal examination (DRE) at the first visit.
Between January 1, 1989, and December 31, 1996, the 8 years covered by this analysis of the study, 8137 men were screened at the clinic, while 38,137 men were not screened, according to Labrie. Deaths from prostate cancer were identified in the official death registry of the Minister of Health of the Province of Quebec.
One point of controversy is that, of men who were invited to come in for screening, only 23% responded. The remainder of those who were invited but did not come in for screening were pooled in the analysis with those who were not invited for screening.
Selection Bias?
Boyle commented there was a "selection bias" in the study. He said the 23% acceptance rate for screening raised a number of questions. "The true comparison should be between those who were invited to screening and those who were not invited to screening. This will be the focal point of the debate for the next year about this important study," Boyle said.
Labrie said there was no difference between those who were invited and did not come to screening compared with those who were not invited.
In the Canadian study, there was a PSA cutoff point of 3.0 ng/mL, which is not considered a reliable cutoff in the United States, according to Raghavan.
Labrie said, "We could observe that 3.0 ng/mL PSA should be taken as the upper limit of normal and not 4.0 as routinely done." He also said "transrectal echography of the prostate was performed only if serum PSA or DRE was abnormal, above 3.0 ng/mL (or an increase of more than 20% in 1 year), thus greatly reducing the costs and facilitating the procedure."
"Only 5 deaths from prostate cancer occurred in the screened group of 8137 men, an incidence of 15.0 deaths per 100,000 men-years," said Labrie. "In the group of 38,056 men who were not screened at our clinic and were followed by usual medical practice, 137 deaths occurred, an incidence of 48.7 deaths per 100,000 men-years."
Labrie said the difference between the 2 groups was highly significant (P<.01) and demonstrated a 3.25-fold reduction or a 69% decrease in prostate cancer death.
Of the 5 men who died from prostate cancer in the screened group, 4 received their diagnosis at their first visit, 3 of them with metastatic disease. Labrie noted that among 23,423 follow-up visits, 122 cancers were found to be clinically localized, thus potentially curable. He said "prostate cancer can be diagnosed at a clinically localized stage in close to 100% of cases with screening, thus practically eliminating the diagnosis of metastatic and noncurable disease."
Screening Recommendations Vary
Labrie's recommendation is that PSA measurements alone, annually, should be started at the age of 50 years in the general population and at the age of 40 years in those at high risk, such as having a family history of prostate cancer or in African Americans. Those with an abnormal PSA level, above 3.0 ng/mL, should have further tests, especially DRE, transrectal echography of the prostate, and, when appropriate, biopsy. He said with "screening started at 50 years and appropriate early treatment, death from prostate cancer can become rare."
Boyle said that of 15 medical groups in the United States who have made recommendations about PSA testing, one third have sat on the fence, one third said no, and one third said yes.
Labrie estimated the cost for the diagnosis of 1 case of prostate cancer at the first visit is $US 2665, less than the cost of the diagnosis of 1 case of breast cancer, estimated to be between $15,000 and $30,000.
He concluded that the study demonstrated the "efficacy, reliability, feasibility, and acceptability of diagnosis and early detection of localized prostate cancer in the general population." Since screening and early treatment permit a 69% decrease in deaths from prostate cancer, this would correspond to 27,050 lives saved per year in the United States alone, according to Labrie.
Raghavan also made the point that "one of the hot issues of the 1990s, and probably of the next millennium, is what we can try to do to identify prostate cancer patients at an earlier time." However, he said the Canadian study is "controversial" because of the selection bias in randomization.
"All of us want to diagnose prostate cancer as early as possible, and the problem is that we need to be sure that a screening test such as PSA actually saves lives," he said. "I'm not yet convinced."
—by Pat Phillips, JAMA contributor
(JAMA. 1998;280:8-9)
Comments on this posting?
Click here to post a public comment on the Trash Talk Bulletin Board.
Click here to send a private comment to the Junkman.
Copyright © 1998 Steven J. Milloy. All rights reserved on original material. Material copyrighted by others is used either with permission or under a claim of "fair use." Site developed and hosted by WestLake Solutions, Inc.
Material presented on this home page constitutes opinion of Steven J. Milloy.