Greater risks. African Americans have poorer prognoses for certain cancers.

Simon Jarratt/Corbis

Cancer Does Discriminate

For years, researchers have known that race is a factor in cancer survival. Black women are less likely than whites to get breast cancer, for example, but much more likely to die from it (Science, 2 February 2007). How much of this difference stems from unequal access to health care, such as regular screening and aggressive treatment, and how much is rooted in biology? A new study that looks back at dozens of cancer clinical trials concludes that for some cancers, such as lung and leukemia, race makes no difference, while for others, such as breast, prostate, and ovarian, it does.

Oncologist Kathy Albain of Loyola University Chicago in Illinois felt that one way to get a big-picture view of the link between race and cancer survival was by combing through a vast network of clinical trials. So she turned to the Southwest Oncology Group (SWOG), one of several cooperative groups in the United States that oversees large, multicenter trials for a variety of cancers in which all the participants get the same level of care, and many receive the same treatment. Albain and her colleagues focused on 35 large SWOG trials that ran at some point between 1974 and 2001, and that included nearly 20,000 adults. The trials covered eight cancers: breast, lung, colon, ovarian, prostate, multiple myeloma, lymphoma, and leukemia. Albain's team examined how African-American participants fared, adjusting for potential confounding factors such as weight and socioeconomic status (estimated from ZIP codes).

For most of the cancers, race made no difference in survival. But for three--breast, prostate, and ovarian--it mattered. In early-stage premenopausal breast cancer, for example, 10-year survival rates were 68% for African Americans versus 77% for all other patients (mostly Caucasian) in the trials. For advanced ovarian cancer, median survival was 1.3 years for African Americans and 2.3 years for the rest of the participants. And for advanced prostate cancer, it was 2.2 years versus 2.7 years, the group reports online today in the Journal of the National Cancer Institute.

On one hand, the results are reassuring because for many common cancers, race made no difference to the outcome, says Albain. She believes the answer to why differences exist for three cancers lies in an interaction among tumor biology, hormonal responses in the patient, and genes that could affect how drugs, such as certain chemotherapies, are metabolized.

One of the questions that's dogged the study of racial disparities in cancer has been how much is due to access to care, notes Timothy Rebbeck, a cancer epidemiologist at the University of Pennsylvania. But here, "disparities persist even in the presence of a standard treatment." Furthermore, the three, sex-specific cancers have been fingered in other racial-disparity studies, he says.

One concern, says Peter B. Bach, a physician and epidemiologist at Memorial Sloan-Kettering Cancer Center in New York City, is whether deaths in the trials were invariably due to cancer--as opposed to some unrelated disease to which African Americans are more susceptible, such as diabetes. But Albain notes that if this were the case, her group would have seen disparities across all cancers--not just the few they detected. She's now scrutinizing gene expression patterns and other features of banked tumors for more clues.