Placebos break taboo in cancer drug tests
Study seeks hope for desperately ill
For 10 years, drugs and surgeries kept Betty Rosenbluth alive, but the cancer in her gut persisted and soon the 73-year-old woman could scarcely make her beloved Sunday stroll up the road to Mt. Zion Lutheran Church.
As a last resort, she journeyed from her home in York, Pa., to the Dana-Farber Cancer Institute here to participate in a medical experiment on a cutting-edge cancer drug that has pushed the envelope of research ethics.
As part of the experiment, her doctors explained that she could get an inactive pill called a placebo. They said giving a placebo at random to some of the patients, who all were dying, was the only foolproof method for showing that the drug worked and thereby convincing the federal government to approve its use for many thousands of desperate patients.
The ongoing experiment and a few other clinical trials of cancer drugs are breaking what had been a taboo in cancer medicine: Denying dying patients promising though unproven drugs that may prolong their lives.
Many physicians say the increasingly successful war on cancer depends on these sorts of experiments. But the practice has riled some in the cancer research community, with two leading medical centers refusing to participate and some patient-advocate groups lodging complaints. They say the experiments jeopardize the well-being of patients in the trials for the benefit of future patients. Yet placebo use in cancer trials is expected to increase because the new generation of precision cancer drugs works through subtle biological means that often elude the more crude methods of drug testing typically used with dying patients.
Supporters of the Dana-Farber trial, which is testing a drug called SU11248 for a rare but lethal digestive tract cancer, defend it as ethical, saying patients are fully informed of the risks and that, ultimately, thousands of patients worldwide will benefit.
''I'm trying to put patients' interest up front -- my patients now, my patients later, and patients who won't be my patients but someone else's," said Dr. George D. Demetri, who is running the experiment. ''This is probably the best way for American patients to get this drug."
As it turned out, Rosenbluth got a placebo. Despite the advance warning, it still haunts her.
''If they would have offered me a choice, I would have gone on the drug right away," said Rosenbluth, who was allowed to start taking the real drug in June after her cancer worsened during the three months she took a placebo. It is too soon to tell whether the drug is helping her.
Another patient in the same trial, Sandi Merriman, took a differing view. ''I knew it was about trying to get the drug to market to help other patients," said Merriman, 56, an auto marketing consultant from Minneapolis. ''I knew I was making a sacrifice."
Placebo-controlled experiments are considered the gold standard by federal regulators, who must approve new treatments before they can be widely distributed. Demetri was initially wary of using placebos. But when he brought his early-stage data to the US Food and Drug Administration last summer, he said, regulators questioned the drug's effectiveness against gastrointestinal stromal tumor, or GIST. The problem was that SU11248 does not shrink the tumors of most patients, which the FDA considers strong proof that a cancer drug works, he said. Instead, it seems to halt the progression of disease. But patients' tumors stay put.
The drug may be holding GIST in check in some more subtle biological way. Or these patients could be getting lucky, with something other than the drug helping them. The only way to know reliably whether the drug is the cause is testing with a placebo, so researchers can make detailed biological comparisons between those on the drug and those on the placebo.
Also, without a placebo wing, Dana-Farber researchers said they lacked patients to compare with their SU11248 patients. In many cancer trials, doctors give all patients a drug, then measure how they do against similar patients elsewhere that did not get the drug. But there are few GIST patients around the country that have also failed on Gleevec, as all the SU11248 trial patients have.
''Without the placebo arm, the FDA would question if this drug was doing anything," Demetri said.
FDA spokeswoman Kathleen Quinn said: ''The FDA has developed accelerated approval processes to speed drugs to patients without placebo-based trials. We have not insisted that trials be placebo controlled." She would not comment specifically on the Dana-Farber trial.
Between 5,000 and 10,000 Americans develop GIST annually, and only a third live five years after being diagnosed. Surgery can help some patients, and a drug called Gleevec is the only non-surgical treatment now, shrinking some tumors. But many patients eventually fail on Gleevec.
But S11248 disrupts cancer cells' ability to multiply and cuts off their blood supply, a powerful combination far more targeted than most cancer treatments.
Last month, Demetri unveiled new SU11248 data: In 92 GIST patients who failed on Gleevec, seven had their tumors shrink and 53 had stable disease, meaning their health stabilized, though its unclear how much longer they will live. Sixty-five percent of patients got some benefit from the drug, enough evidence to launch a bigger, final-stage trial with about 400 patients at 60 hospitals worldwide.
But two prestigious medical centers, MD Anderson Cancer Center in Houston and the University of Michigan medical center, refused to join in.
''When patients have an advanced cancer and the cancer is growing, there isn't any way the placebo can be helpful" to the patient, said Dr. Laurence Baker, associate director of Michigan's cancer center. ''To argue that a placebo trial is in society's interests has nothing to do with helping these patients."
Some patient advocates, after initially championing SU11248 trials, have become sharp-elbowed critics of the placebo trial used to test it.
''That drug is the only option they have to live. Under those circumstances, we think a placebo is unnecessary and dangerous and unethical," said Norman Scherzer, executive director of the Life Raft Group, a GIST advocacy organization.
The SU11248 experiment was carefully reviewed by a Harvard ethics panel before it began. One panel member, Dr. Steven Joffe of Dana-Farber, said, ''I can understand the desire of patients with this terrible cancer to have these drugs now, but at the same time if ineffective drugs get out there, it's the patients that pay the price."
As a safeguard, Demetri insisted on a provision that ''unblinds" patients who continue to deteriorate past a certain threshold, revealing if they are taking the drug or a placebo. Those on placebos get the drug, as Rosenbluth eventually did.
''I said I will not run a placebo-controlled trial without a crossover. Period," Demetri said. ''If you sneeze, I'll do a CAT scan. And if you're getting worse, you'll have access to the drug."
Some GIST patient-advocates have argued that
Of that allegation, Pfizer's vice president for cancer drug development, Dr. Bill Slichenmyer, said: ''Absolutely not . . . the expedient thing would be to do no trial at all. Pfizer is not going to make money off this drug." He explained that GIST is a relatively rare cancer representing a small market. The Dana-Farber researchers have no investments in Pfizer, they said.
George Annas, a medical ethicist at Boston University, understood the desire for placebo data. ''Placebo-controlled trials are the gold standard," he said. ''From the scientific point its what you want to do, but you can't always do it.
''You have to give people in that situation something that might help them," he said. ''I don't see any justification for withholding that drug from someone."
Rosenbluth said she still holds her Dana-Farber doctors in high esteem but wishes she had received the drug from the outset. ''I would have had three months under my collar with the drug. Definitely, if I had a choice I'd have taken the pill right away," she said. ''I don't know why they do this with things as serious as cancer."
Raja Mishra can be reached at rmishra@globe.com.
