| NEW YORK
NEW YORK (Reuters Health) - Trying a new, experimental cancer drug may offer a glimpse of hope for very sick patients, but often does more harm than good, a new study shows.
Researchers said cancer doctors regularly resort to drugs still undergoing testing, as long as they have been approved for other diseases or in different combinations or doses.
But because the science is still up in the air, nobody really knows what the consequences of taking such drugs are.
"Many of these drugs end up not being the tremendous improvement that we hoped they would be," said Dr. Otis Brawley, chief medical officer of the American Cancer Society, who was not involved in the new study.
"People need to realize that because the trials have not been completed there is a great deal that is not known about the treatments," he told Reuters Health. "There are people who get these treatments and get hurt."
The new study, published in the Journal of Clinical Oncology, looked at 172 clinical trials published over two years.
Less than a third of the clinical trials showed the experimental drugs improved patient survival, and less than half found the drugs helped other clinical outcomes.
Those numbers could even be too high, researchers say, because negative findings tend not to be published.
On the other hand, two thirds of the trials reported increases in at least one severe or life-threatening side-effect.
Most of the drugs under scrutiny had already been approved in different doses or combinations -- or for different diseases -- and were available on the U.S. market.
It's unclear if doctors were prescribing them for unapproved indications, but it's entirely possible as one-fifth of US drugs are prescribed in such a way -- a legal practice known as off-label prescription.
"We know there is a high demand for new treatments in oncology," said Dr. Jeffrey Peppercorn, a cancer doctor at Duke University Medical Center in Durham, North Carolina, who led the new research.
In an earlier survey, he said, he had found that more than eight in 10 cancer doctors prescribe medications that are being tested.
"This sounds a cautionary note," Peppercorn said about his new findings. "In most cases, we should refrain from using experimental drugs outside of clinical trials."
And it's not only patients taking the experimental drugs that may feel the impact.
Clinical trials done in the US, the researchers found, had trouble recruiting participants and took nearly twice as long to complete when the drug under scrutiny was already on the market.
"The concern is that you have fewer people choosing to enter the trial if the drug is already available," Peppercorn said, cautioning that the trial didn't prove that to be the case.
Trials are designed to compare new drugs with existing ones that have been deemed safe and effective by the US Food and Drug Administration.
Signing a patient up for a trial therefore often means only a 50-50 chance that he or she will get the new drug. So doctors may choose to prescribe it off label instead, and insurance companies often pay for it.
"Most of the doctors who do this are thinking that they are doing something in the best interest of the patients," said Brawley. But in reality, he added, they end up slowing down the development of new drugs.
Dr. Steven Joffe, an ethicist at Dana-Farber Cancer Institute in Boston who has worked with Peppercorn on earlier studies, agreed.
"I perfectly understand why a patient with a poor prognosis might want to have access to these drugs and why doctors would give them the drugs," he told Reuters Health.
"They may think they know that it works," he said. "But the fact is, somebody is running a study to find out if it does. Almost by definition it hasn't been proved safe and effective."
As an example of how doctors and patients can be fooled, Joffe mentioned that scores of women were treated for breast cancer with stem cells in the 1990s, when the treatment was still in clinical trials.
"It took about ten years to do the trial because it was so hard to get women to enroll," said Joffe. And when the results finally came, the treatment turned out to be a dud -- a toxic and expensive one, too.
Joffe said prescribing drugs when they are still in clinical trials might not only slow the development of good medical evidence, but could even halt it.
"We know that many studies fail to ever get completed," he explained. "These data suggest that the use of drugs outside of trial protocols could contribute to the failure of studies ever to get completed."
Sometimes, of course, new drugs do work, and no one argues that doctors shouldn't be allowed to prescribe medications they think will help patients.
But doctors should be very clear about the high risks involved. One way to do that, said Peppercorn, would be to require that cancer doctors get informed consent from patients before they start them on experimental drugs.
In the end, Brawley said, the decision to use such treatment should be left to the patient and his or her doctor.
"There are times when it can be justified," he noted, "but it is done far more often than it should be."
SOURCE: link.reuters.com/vyz32q Journal of Clinical Oncology, online October 25, 2010.