CORRECTION: Gene expression profiles may improve cancer prognosis
[Editor's note: The study on which this 2008 story - originally filed as 20080401elin033 -- was based has been retracted. Please see reut.rs/xjBYkv for more details.]
NEW YORK (Reuters Health) - Integrating genomic information with traditional clinical risk factors can refine the prognosis and help optimize treatment strategies for women with early breast cancer, a research team at Duke University reports in the Journal of the American Medical Association this week.
Dr. Anil Potti and colleagues in Durham, North Carolina, took a look back at clinical and genomic data from a National Institutes of Health repository containing tumor data for 964 patients with early stage breast cancer.
Within each risk category (low, intermediate, or high risk), the investigators identified prognostically significant clusters, "representing clinically important" genomic subtypes of breast cancer.
Specifically, in the low-risk category, a worse prognosis was associated with a gene expression profile that included activation of wound healing, invasiveness, chromosomal instability and deregulation of a key cancer-associated pathway.
Median relapse-free survival time in patients with this gene signature was 16 months less than among those with tumors exhibiting the inverse pattern, they found.
The prognostic clusters also have unique sensitivity patterns to commonly used chemotherapy drugs, the investigators report. .
In an editorial published with the study, Dr. Chiang-Ching Huang and Dr. Markus Bredel, from the Feinberg School of Medicine in Chicago, note that genes that have mechanistic implications for breast cancer "represent potential targets for specific molecular therapy."
This strategy is "an advance in the changing landscape of oncology toward individualized patient management," they write.
