Release of Cancer Research Debated
How quickly cancer patients and their doctors learn about significant new treatment advances depends on a struggle between some of the nation’s leading researchers and the National Cancer Institute.
The battle is over control: Who will decide when a piece of research is conclusive enough to be disclosed, especially when it shows that a new treatment works? Who will decide whether changes in cancer treatment can be announced before scientists publish the results in a medical journal?
The Cancer Institute, a public health agency charged by Congress with fighting the war on cancer, wants to speed things up. Its target: the delay between the end of a study and its publication in the medical journals. Normally, reports sent to journals are given to experts in the same field to review for accuracy--a process that can take many months.
“How many people have to die while a (medical journal) reviewer leisurely looks at a (scientific) paper?” demanded Vincent T. DeVita Jr., former NCI head and now physician-in-chief at Memorial Sloan-Kettering Cancer Center in New York.
Some research physicians, however, are unhappy with the Cancer Institute’s pushing a particular result. Instead, they want to preserve the existing system for presenting data at medical meetings--and then publishing them in scientific journals.
In part, this system is a key to obtaining professional credit and academic advancement. But opponents also worry that the premature release of studies might confuse and mislead doctors and the public about which is the best treatment, especially since about 5% of studies with positive results turn out to be wrong.
Saul A. Rosenberg, chief of oncology at the Stanford University Medical Center, defends what he calls “a traditional and time-tested system of peer review, editorial review and necessary confirmation by other investigators to set the standards of care and to inform physicians. I am not persuaded that we should bypass that.”
But the Cancer Institute has bypassed the old system by mailing study results directly to doctors twice in the last two years.
The first letter, sent to 13,000 physicians in May, 1988, was called a “clinical alert” and described a new advance in breast cancer. It caused an enormous controversy.
In February the New England Journal of Medicine published the results of a large study on the use of drug treatments after surgery for colon cancer that reduced deaths by one-third in patients in a certain stage of the disease. About 22,000 Americans suffer that stage of colon cancer every year.
Preliminary results were known last summer when the Cancer Institute asked the Food and Drug Administration to make a drug used in the study, levamisole, widely available. This discussion led to newspaper stories in June, 1989.
But the institute delayed putting out a letter to about 35,000 doctors until October, when the scientists were ready to publish a preliminary report in a medical journal. To minimize controversy caused by a previous notice to physicians, the Cancer Institute toned down the letter, calling it a “clinical update” instead of a “clinical alert.”
On Jan. 31, cancer researchers and NCI officials gathered to try to reach agreement on whether the federal agency that funds most cancer studies in this country should continue to issue clinical alerts and under what circumstances these alerts should be sent out.
By the end of the all-day and sometimes heated debate, most of the opponents conceded that the alerts could continue if they were used judiciously.
The public probably can expect one or two clinical alerts a year, said Bruce A. Chabner, head of NCI’s division of cancer treatment. “I don’t think we will have something important to say once a month. I wish we did.”
A number of criteria were suggested for handling an alert, including: that it be initiated by the researchers; that before it goes out, the data be reviewed by independent experts; that the treatment being recommended affect a large number of patients; that the NCI try to speed up peer review and publication in a scientific journal.
The experts also thought that before an alert is sent out it should be approved by the National Cancer Advisory Board, an independent body appointed by the President that includes lay members as well as cancer experts.
If that is done, the alert “will hit the front page of The Washington Post before physicians get the information,” said Charles G. Moertel of the Mayo Clinic, who ran the study that led to the colon cancer update. Since the advisory board’s meetings are open to the public, the press can be expected to attend and report on the board’s vote on any clinical alerts.
“That’s life,” said DeVita, who in the past has recommended taking information over the heads of doctors and directly to the public. “We live in that kind of world.”
The clinical alerts were DeVita’s idea and they, like the former director, have been controversial.
The first was issued in May, 1988, when the Cancer Institute sent out a recommendation on a new treatment for breast cancer. The alert described studies by the Eastern Cooperative Oncology Group and the National Surgical Adjuvant Project for Breast and Bowel Cancers, two organizations that test new drug treatments on patients in many medical schools.
Basically, the alert advised all doctors to treat all breast cancer patients who had surgery to remove tumors with follow-up chemotherapy--even if there was no sign that the cancer had spread and was likely to recur. In the past, if there was no sign of cancer cells in the woman’s lymph nodes, no additional treatment was given even though half the women would have a recurrence.
The alert angered physicians, in part because the actual details of the research results were not available. The study on which the alert was mostly based had been sent to the New England Journal of Medicine. Publication was delayed for 14 months, which led DeVita to issue the clinical alert.
Many doctors also were angered because they felt they were being told in a public way how to treat their patients.
Although the Cancer Institute’s Chabner says this was not the intent, many physicians think the alerts establish a standard of care that they must follow when treating their cancer patients. If they don’t prescribe such new treatments, the doctors fear they could be open to a malpractice suit if their patient does not do well.
“There was a lot of hostility with the first clinical alert,” said Charles A. Coltman Jr., chairman of the Southwest Oncology Group and director of oncology at the University of Texas Health Science Center in San Antonio.
In fact, there was a mutiny. Cancer experts in Boston banded together, including those in the Harvard University system, to produce the “Statement of Boston Physicians.” The Boston doctors concluded that “there is no evidence that these (recommended) therapies will be helpful” and suggested that the best thing for patients “is participating in the next generation of trials.”
They essentially refused to comply with what they perceived as an NCI edict. Even now, breast-cancer patients in certain stages of their disease do not routinely receive additional chemotherapy to prevent recurrence.
Stanford’s Rosenberg justifies this position by some doctors, saying, “You can do a lot of harm with chemotherapy. One of the roles (of the doctor) is to say that there is no role for chemotherapy in this type of cancer.”
