Study Devalues a Popular Idea on Evaluating Medical Trials


January 23, 2004

A new report by researchers at the Dana-Farber Cancer Institute in Boston has found little evidence to support a widely held belief about clinical trials of experimental cancer treatments: that patients who enroll in them fare better than those who do not.

"There may be a difference in outcomes, but to our view it hasn't been shown," said Dr. Steven Joffe, a pediatric cancer specialist at Dana-Farber and an author of the paper, which is being published today in the British journal Lancet.

Many cancer specialists have long assumed that being in a clinical trial leads to a better outcome. The studies often take place at leading cancer centers, where care tends to be of the highest quality.

In a clinical trial, patients are usually assigned randomly to receive an experimental therapy or the standard one. Dr. Joffe said he and his colleagues undertook the study after hearing "prominent folks in the field" speaking out at medical meetings "and saying that being in a trial contributes to better outcomes."

The common view, he said, was that "extra care, extra monitoring, extra attention leads to better outcomes," regardless of whether the patients receive the standard therapy or the experimental one. Searching the medical literature, the investigators found two dozen published studies that compared outcomes among cancer patients who enrolled in a clinical trial to those who did not.

The data were not ideal, Dr. Joffe said, because cancer patients who took part in the trials were often hard to compare with those who did not; patients in the trial group might be younger and healthier, for example. But over all there was no evidence that they had better outcomes.

The only exceptions came in a few studies of children and of adults with blood cancers.

"Those are settings where a lot of cancers are more treatment-responsive, so treatment makes more of a difference," Dr. Joffe said.

Also, the pediatric studies tended to be from years when researchers were making enormous strides in treatment and when those patients getting the experimental treatments tended to do much better.

"If you happened to catch a disease at one of those moments," Dr. Joffe explained, "then your study would show a difference."

But for other cancers, like those of breast and lung, the study found no such effects. That may be because when experimental treatments helped in these diseases, their benefits have been modest, he said, adding, "To assume that being in a trial would add a quantum leap is not realistic."

Cancer experts not connected with the Dana-Farber study said they found it convincing.

"I'm in general accepting of the conclusion" that there is no good evidence that being in a trial improves patients' outcomes, said Dr. Harmon Eyre, chief medical officer of the American Cancer Society. "We always need to keep ourselves honest about the communication between doctors and patients. You can't overpromise."

Dr. Archie Bleyer, director of community oncology and a professor of pediatrics at the M. D. Anderson Cancer Center in Houston, said the paper tended to confirm his suspicions about the overall benefits of clinical trials, but he noted that it did not ask whether quality of life might be better for patients in such trials.

"The premise is, with the greater care in a trial, the quality of survival is likely to be better," Dr. Bleyer said.

Rebecca Dresser, who teaches law and medical ethics at Washington University, said a crucial lesson of the findings was the importance of helping patients understand the purpose of clinical trials.

"You want people to understand that the primary aim is to produce knowledge to benefit people in the future," Dr. Dresser said. "The notion that trials are done to benefit individual participants is incorrect."

"I don't want to say I'm antiresearch," she went on. "But we really want people in studies to understand what it is they are doing."

Dr. Joffe said he expected that patients would be willing to enroll in trials even if without the message that they would do better simply by being in one. They will even enroll their children.

"My personal experience is that no one is going to sacrifice their child for the purpose of research, and we don't ask that," he said. "We say, `We think it is a reasonable thing to do for your child's care."