Genomic Tests Improve Prediction of Breast Cancer Response to Chemotherapy, Hormonal Therap
M. D. Anderson News Release 12/14/06
Researchers at The University of Texas M. D. Anderson Cancer Center have developed two genomic tests to better predict how breast cancer patients will respond to chemotherapy or hormonal therapy.
In presentations Dec. 14-15 at the San Antonio Breast Cancer Symposium, the research team reports on a highly sensitive multiple-gene test designed to predict response to chemotherapy, performed before surgery at M. D. Anderson, and a genomic index that gauges 10-year survival for patients who receive post-operative hormonal therapy with the drug tamoxifen.
"We are moving these tests toward clinical trials, where we can measure improvements in treatment response and track how physicians and patients use this information to make better decisions about treatment," says one of the team leaders, Lajos Pusztai, M.D., Ph.D., associate professor in M. D. Anderson's Department of Breast Medical Oncology.
The team developed a multi-gene predictor for pathological complete response (absence of cancer after chemotherapy), by analyzing 82 women who received preoperative paclitaxel and fluorouracil-doxocrubicin-cyclophsophamide (T/FAC), the standard of care for chemotherapy. Pusztai says the team analyzed 780 different combinations before winnowing the test down to a small set of genes.
The test was then applied to 51 new patients who received identical T/FAC treatment. It accurately predicted 12 of 13 patients who achieved a complete response. Of the 28 patients the test predicted would have residual cancer, 27 had breast cancer after treatment, indicating a high negative predictive value. "That's important, because if you knew in advance that you were unlikely to benefit from this standard chemotherapy, then you might choose a different treatment or an investigational drug," Pusztai says.
The test also predicted that 23 patients would achieve complete response, when only 12 did, mainly because the remainder had minimal residual disease after treatment. The team is studying minimal residual disease further. Pusztai said the positive predictive value is an improvement over current predictions based on pathologic characteristics of the cancer.
The chemotherapy findings were published earlier in the Journal of Clinical Oncology.
About 70% of breast cancers express the estrogen receptor (ER), but tamoxifen and other anti-estrogen therapies only benefit about half of these patients. The challenge is to predict exactly who will be helped and who should seek other treatment.
Researchers focused on the influence of estrogen receptor activity on the expression of downstream genes. "We defined the 200 strongest ER-related genes and developed from those a Sensitivity to Endocrine Therapy (SET) index to actually measure this activity of ER," says W. Fraser Symmans, M.D., associate professor in the Department of Pathology.
The ability of the SET index to predict the benefit of hormonal therapy was then tested in 267 patients who received tamoxifen without chemotherapy for five years after surgery. "There was a strong relationship between their SET index score and the probability of there being no disease 10 years later," Symmans said.
Scoring in the top 35% of the index was significantly associated with excellent distant relapse-free survival at 10 years. Patients who scored in the lower 50% of the index derived little benefit from tamoxifen alone and, in a separate analysis, were shown to more likely benefit from chemotherapy.
An intermediate group, who scored in the 50th-65th percentile, initially benefited from five years of tamoxifen, but their relapse rate began to increase about two years after treatment ended and eventually approached that of the SET low-scorers.
At five years, about 92% of both high and intermediate SET index scorers had relapse-free survival, compared with 70% for the low SET scorers. At 10 years, however, about 88% of high SET scorers had avoided relapse, compared to 65% for both intermediate and low scorers.
"The intermediate group may be very important," Symmans said, "because recent clinical trials have indicated that there is a subset of patients who benefit from continuation of hormonal therapy with newer drugs after their five years of tamoxifen. This intermediate group might be showing us who those patients are."
Symmans and colleagues are validating the SET Index in a second group of patients.
Taken together, the SET Index and the T/FAC test of chemotherapy provide a platform to help triage and guide breast cancer treatment, Symmans and Pusztai said. They plan to add predictors for other available treatments.
Both chemotherapy and tamoxifen response tests are being developed in close collaboration with Nuvera Biosciences, a start-up company launched by M. D. Anderson for commercialization of molecular diagnostic technologies. The University of Texas Board of Regents owns stock in Nuvera. These arrangements are managed in accordance with M. D. Anderson's conflict of interest policies.