When William Gradishar began his career as a breast cancer specialist 30 years ago, chemotherapy was recommended after surgery for all women with early-stage disease.
"It was reflexive," said Gradishar, an oncologist at Northwestern University.
More judicious use of chemo began about a decade ago, with the advent of genomic tests that can gauge each patient's risk of recurrence based on the pattern of gene activity in tumors.
Now, the power of such tests to individualize treatment is clear. A federally sponsored study that used Oncotype DX found most women with the most common form of early stage breast cancer can safely skip chemo. An estimated 65,000 women a year will be spared the time-consuming, costly, unpleasant — and occasionally debilitating — rigors of the cancer-killing therapy.
Next month, a committee Gradishar chairs will consider the data to update the breast cancer treatment guidelines of the National Comprehensive Cancer Network (NCCN), based in Fort Washington.
"Clearly, this is a landmark study," he said. "I think it will cause some changes in the guideline language."
A number of companies have developed genomic tests to customize diagnosis and treatment of different cancers, but Oncotype DX, made by Genomic Health and introduced in 2004, is the most validated by studies.
"You have to use the test in many populations and then reproduce the results," said Angela M. DeMichele, a University of Pennsylvania breast oncologist and epidemiologist. "That takes a long time. In most other cancers, it's at a point where we're just validating the tests."
Oncotype DX is only for women with early breast cancer that is driven by hormones, has not spread to any lymph nodes, and doesn't overproduce the Her2 protein, a marker of aggressive tumors. Such women usually have surgery — mastectomy or a lumpectomy and radiation — followed by years of a hormone-blocking drug aimed at keeping cancer from returning. But many women are also advised to have chemo to kill undetectable cancer cells that may have escaped the breast tumor.
Based on an analysis of the activity of 21 genes, Oncotype DX gives women a low, intermediate or high score estimating the chance of recurrence. The $4,000 test has been part of the NCCN treatment guidelines since 2008 (other expert groups also recommend it). Medicare and most private insurers cover it.
But the test left many women with a quandary. While those with high scores clearly benefited from chemo and those with low scores clearly did not, there were not enough data to advise those who fell in between.
"It's been really challenging for women in the middle," said Jean Sachs, CEO of Living Beyond Breast Cancer, the Bala Cynwyd-based support and advocacy group.
"If a woman had an intermediate score, there would be a lengthy discussion about the risks and benefits of chemotherapy," said Fox Chase Cancer Center oncologist Angela Jain, who specializes in women's cancers. "Some patients would go forward with it, others would hold back."
Chemotherapy almost always causes transient nausea and hair loss. But some side effects are permanent. In younger women, it can trigger infertility and premature menopause. Neuropathy — numbness in the extremities — is another possible consequence. In rare cases, the therapy leads to leukemia or heart failure.
To clarify the value of chemo, the landmark study, called TailorRX, randomly assigned 6,700 women with intermediate scores to get chemo and hormone-blocking therapy, or just the hormone blocker.
After nine years of follow-up, 94 percent of both groups were alive, and about 84 percent of both groups were cancer-free — clear evidence that adding chemo was overtreatment.
An exception was women under age 50 with intermediate scores on the higher side, who did better with chemo.
"The overall view is that this will eliminate chemotherapy in many cases, but not all," Gradishar said of the data, presented Sunday at a cancer conference and published in the New England Journal of Medicine. "There are nuances."
There are also unanswered questions about how to further reduce overtreatment. A study called RXponder is underway to see whether some women with low Oncotype DX scores can safely skip chemo, even though their cancer has spread to one to three lymph nodes. Genomic tests may also clarify who should take hormone-blocking therapy — with its hot flashes and other side effects — for five years rather than seven or 10 years years, Gradishar said.
Finding answers could help women like Sheryl Greene, 50, a fiber optics saleswoman in Roanoke, Va. She was 44 when she was diagnosed with breast cancer that had spread to a single lymph node. Her Oncotype DX score was intermediate, but on the border of low.
Rather than a decade of anti-hormone therapy, she opted for a total hysterectomy to stop her natural hormone production, although that meant premature menopause. She also decided to have chemo because of a family history of ovarian cancer that was not linked to a known genetic susceptibility.