Listen to the audio recording of this call (MP3 Format)

This teleconference took place on December 14, 2006, and was a roundtable discussion about important new developments in breast cancer epidemiology, including a look at the "Report to the Nation" on U.S. cancer data (http://www.cancer.org/docroot/STT/stt_0.asp), the issues surrounding breast cancer statistics, and racial/ethnic disparities in breast cancer research. The call lasted one hour.

Call Notes
Participants: Dick Clapp, Devra Davis, Guy Dauncey, Pat Donnelly, Nancy Evans, Bill Graham, Janet Gray, Karen Jacobs, Judy Lane, Michael Lerner, Marcia Marks, Leo Petrilli, Jeanne Rizzo, Deborah Shields, Susan Shinagawa, Julia Varshavsky, Lisa Wanzor

Nancy Evans, Health Science Writer & Consultant, Breast Cancer Fund
Nancy Evans introduced the piece that she wrote with Brynn Taylor at Breast Cancer Fund, commenting that getting the science right is critical, yet it is proving difficult to analyze breast cancer data. We may need to talk about breast cancer in the larger context of cancer because we have no national registry and the SEER data has many flaws. How can we best talk about breast cancer – can we really say the risk was 1 in 20 in 1940? If so, whom do we cite?

Devra Davis, Ph.D., MPH, Executive Director, Center for Environmental Oncology, University of Pittsburgh Medical Center, Cancer Pavilion
Devra Davis commented that although breast cancer statistics are flawed, we should not ignore them. Once we characterize the limitations of the data, we can learn something. It is important to pay attention to the statistics because most cases of breast cancer occur in women with little known risk factors.

When looking at the data, it is important to look at the age adjusted rates because median ages change, and age adjustment varies depending on the average age of the population during a certain time. If we do not look at the age adjusted rates and instead group all ages together, it appears that breast cancer rates have not changed much overtime. It is also important to point out that numbers increased dramatically from 1975 to 1990 because mammography became widely available during that time.

SEER registries are taken at different time periods in different states, and therefore cannot be grouped together. Also, SEER under-represents minorities. In age adjusted data though, older age groups show a slight decline in breast cancer rates, which some speculate may be due to a decline in hormone therapy use.

Janet Gray, Ph.D., Professor, Department of Psychology, Participant, Program in Science, Technology & Society and Neuroscience & Behavior, Vassar College
Janet Gray discussed the history of breast cancer registries in the United States: the U.S. did not keep any cancer data at all until the turn of the 20th century, when ten states began to include cancer as a single disease in their mortality data.

In 1913, the precursor to the American Cancer Society (ACS), the American Society for Cancer Control, began collecting more reliable cancer mortality data, taking things like population size and age adjustment into account. Even then, there were many social and economic issues that affected the way cancer was reported.

It was not until 1935 that the first statewide registry in Connecticut was created. The second was New York, which did not even include the five boroughs of New York City – a huge part of the state - until the 1970s. These are two examples of the problems we face in the data. It has only been in the last five to ten years that we have looked at a national registry approach.

Susan Shinagawa, Past Chair, Intercultural Cancer Council, and Co-founder/Co-chair, Asian & Pacific Islander National Cancer Survivors Network
Susan Shinagawa discussed the reasons that cancer data is flawed when it comes to minority populations, the main reason being how the National Cancer Institute (NCI) collects and reports cancer statistics based on aggregate populations. Even in 2006, there are many reports coming out of NCI that only look at black, white, and other, which falsely groups minorities together. Even when looking at Asian Pacific Islander (API) populations, they do not divide that group up properly.

API populations should be divided into Asians and native Hawaiian and other Pacific Islanders. Both of these groups have over 35 populations with different languages and cultures that should be taken into account. Of the API population that NCI looks at, 93% are Asians, and 70% of them are first generation immigrants that come from countries with the lowest breast cancer rates in the world. Although after coming to the U.S. their cancer rates increase dramatically, the NCI data does not take this into account. Grouping all API together hides the data. For example, breast cancer rates for native Hawaiians are much higher than for white people, but this fact gets lost in the data when everything is grouped together.

Susan Shinagawa also commented that another big problem is that NCI does not take into account poverty as part of their equation.

Discussion / Q&A

•    We discussed lobular breast cancer and that alcohol consumption and Hormone Replacement Therapy (HRT) may influence risk, which may be possible explanations for high breast cancer rates in Marin County, CA.

•    We discussed the difficulty of analyzing DCIS data, because it has been inconsistently recorded. After a detailed analysis of the registry data, researchers found that DCIS incidence is in fact increasing in postmenopausal (especially ages 50-59) women (the report is on Devra Davis’s website, http://www.environmentaloncology.org). The most recent data, however, shows a slight decline.

•    We discussed potential research and communication solutions – clear messaging (perhaps an educational primer), focusing on the broader context of cancer rather than breast cancer, talking about why we cannot be more precise, supporting a national cancer registry, implementing more community-based participatory research programs, planning an international 2008 conference, pushing for a paradigm shift, etc.

•    Dick Clapp added that the North American Association of Central Cancer Registries (NAACCR, http://www.naaccr.org) is attempting to standardize what everybody reports, and to include southern and northeast states. CDC funding is responsible for this organization, so making sure the CDC state cancer registry funding continues is key.