If further study bears this out, more than 20,000 women each year in the United States alone might be spared the side effects of this drug or similar ones without significantly raising the risk their cancer will return. That would be roughly half of all breast cancer patients who get chemo now.
"We want to make sure these data are correct before withholding it (Taxol) from some patients ... the stakes are high," said the lead researcher, Dr. Daniel Hayes of the University of Michigan. "On the other hand, we don't want to keep a therapy that doesn't work."
Interestingly, in the study, Taxol did the most good for women who had overactive HER-2 genes the target of the newer breast cancer drug Herceptin. These women were about 40% less likely to have a recurrence if they received Taxol.
"The days of 'one size fits all' therapy for patients with breast cancer are coming to an end," Dr. Anne Moore of Weill Cornell Medical College wrote in an editorial accompanying the study in Thursday's New England Journal of Medicine.
"Oncologists have a responsibility to their patients to be aware of this report."
The original study involved more than 3,000 women whose cancer had spread to nearby lymph nodes but not widely throughout the body. This is the situation of about one-fourth of the 175,000 women diagnosed with breast cancer in the U.S. each year.
Researchers tested adding paclitaxel, sold as Taxol by New York-based Bristol-Myers Squibb Co. and now also in generic form. They gave it after surgery to remove the cancer and treatment with the chemo drugs Adriamycin and Cytoxan.
Taxol improved survival and became a new standard of care. But the drug frequently causes neurological side effects including numbness and tingling in the hands and feet. In the original study, 18% of women had this problem months and even years after taking Taxol.
Even more worrisome has been the growing evidence that some women do not benefit as much from chemo as others. Hayes and other researchers wondered whether that was true in their Taxol study.
They retrieved frozen tissue samples from 1,500 of the original participants, did genetic tests to better identify their types of cancer, and discovered big differences in who had responded to the drug.
The study was paid for by grants from the federal government and a breast cancer foundation. Several researchers consult for Bristol-Myers Squibb.
"We should have done this a long time ago," but the tools were lacking and researchers now have the advantage of longer follow-up of these women, said another senior author, Donald Berry.
Berry is reanalyzing another earlier Taxol study, and Moore urged other scientists to do the same.
With more evidence, "we can begin to use the biology of the cancer to decide whether the chemotherapy will work" before subjecting women to it, Hayes said.
The typical four-cycle treatment with generic paclitaxel costs $7,000 or more, including infusion fees that doctors charge. Insurance typically pays most of this.
For now, many doctors will be reluctant to skip Taxol or other chemo, said Dr. Julie Gralow, a cancer specialist at the University of Washington School of Medicine. Some may fear lawsuits if the cancer recurs and the chemo wasn't given, she said.
"It's just so much easier to give the chemotherapy and know you've been super-aggressive."
However, Kris Miller, a 54-year-old former nurse from Chelsea, Michigan, said patients should be given the choice. She has had problems since taking Taxol two years ago for a type of breast cancer that the new research suggests would not respond to the drug.
"Most people recover from it, and I guess I'm one of those unfortunate ones that did not," she said of the side effects. "I have severe numbness and tingling, mostly in my feet. It becomes painful by the end of the day. It never goes away."
"I hope they give people that option," to weigh the risks and benefits and possibly skip Taxol, she said. "If I was going through it now, I would like to have that information."