Predicting Breast Cancer Risk In African American Women Improved

Featured Article
Main Category: Breast Cancer
Also Included In: Women's Health / Gynecology;  Clinical Trials / Drug Trials
Article Date: 28 Nov 2007 - 7:00 PDT

email to a friend   printer friendly   view / write opinions   rate article

Researchers in the US have developed a better way to predict breast cancer risk in African American women. Until now scientists were using a model based on data from white women, but this new model was developed using data from studies on African American women.

The study was carried out by Dr Mitchell Gail, from the National Cancer Institute in Bethesda, Maryland, and colleagues, and is published in the 27th November online issue of the Journal of the National Cancer Institute.

Until now, researchers have used the Breast Cancer Risk Assessment Tool or Gail model, that was developed by Mitchell Gail himself, to estimate breast cancer risk and thereby select women for breast cancer prevention trials. But this model was developed using data on 240,000 white women and it has never been clear how well it estimates risk for other racial groups, including African American women.

Gail said that he and others had been concerned for some time about the assumptions they were making for women of African American and other racial and ethnic origin. He said it "turns out that we have been underestimating the risk for African American women".

To build a new model, Gail and colleagues used data on African American women with and without breast cancer, which they got from a study called the Women's Contraceptive and Reproductive Experiences (CARE) study, and data from the Surveillance, Epidemiology, and End Results (SEER) program. Altogether, the data collected came from 3,200 black women, of which more than half had breast cancer.

They tested the new model on data from two trials: the Women's Health Initiative (WHI), and the Study of Tamoxifen and Raloxifene (STAR).

The new model, now known as the CARE model, was able accurately to predict the overall number of cancers detected in African American women participating in the WHI trial, and also in subgroups.

Overall, the researchers found that the old Gail model underestimated the risk in at least 90 percent of the scenarios tested.

The researchers also showed that the CARE model was a more reliable predictor than the Gail model for African American women aged 45 and older.

The CARE model estimated that 30 per cent of African American women would have a 5 year breast cancer risk of 1.66 per cent. This is double the proportion predicted by the Gail model, which put the percentage at 14.5 per cent. Using the CARE model therefore would make twice as many African American women eligible for breast cancer prevention trials as the old model.

The researchers cautioned that the CARE model is not recommended for women with a history of breast cancer, and it could underestimate the risk in certain groups, for instance women who have a BRCA gene variant.

However, in spite of these drawbacks, the CARE model seems to offer more reliable (and larger) estimates of the risk of invasive breast cancer among African American women when compared with the currently used Gail model.

The researchers said that while more studies should be done to replicate these results, in the meantime they would recommend that the CARE model is used in deciding eligibility, and helping African American women decide whether to undergo trials for breast cancer prevention.

More than 19,000 cases of breast cancer are diagnosed in African American women every year, and nearly 6,000 die from it.

"Projecting Individualized Absolute Invasive Breast Cancer Risk in African American Women."
Mitchell H. Gail , Joseph P. Costantino , David Pee , Melissa Bondy , Lisa Newman , Mano Selvan , Garnet L. Anderson , Kathleen E. Malone , Polly A. Marchbanks , Worta McCaskill-Stevens , Sandra A. Norman , Michael S. Simon , Robert Spirtas , Giske Ursin , and Leslie Bernstein.
Journal of the National Cancer Institute Advance Access published on November 27, 2007
DOI 10.1093/jnci/djm223.

Click here for Abstract.

Written by: Catharine Paddock
Copyright: Medical News Today
Not to be reproduced without permission of Medical News Today