Sister study hopes to answer breast cancer questions. (NIEHS News).

Working out of their office in Research Triangle Park, North Carolina, a pair of NIEHS researchers are laying the groundwork for what may prove to be a landmark study of the possible interplay between genetics and the environment in the development of breast cancer. Principal investigator Dale Sandler, acting chief of the institute's Epidemiology Branch, and colleague Clarice Weinberg, chief of the Biostatistics Branch, plan to recruit 50,000 female volunteers aged 35-70 whose sisters have been diagnosed with breast cancer and follow their health over a period of time. Eventually the researchers will evaluate the 1,500 or so of the women that they estimate will develop breast cancer during the initial five years of follow-up, analyzing environmental, genetic, and health data captured at the outset of the study.

Sandler and Weinberg are absorbed in the extensive work necessary to develop a study of this size, scope, and importance. Besides developing an overall plan for the study, the researchers are meeting with focus groups, doing feasibility studies, developing applications and questionnaires, and marshaling the considerable technical, financial, and professional resources necessary to undertake the study, which will run at least 10 years.

"Despite great interest in trying to discover environmental causes of breast cancer, we don't have a lot of evidence," Sandler says. "We don't have a lot of dues; there are no obvious places to start." Some evidence suggests that pesticides and solvents may cause breast cancer. Electromagnetic fields have been the focus of other studies, though Sandler cautions that evidence from those studies is inconsistent. Phthalates, compounds used as plasticizers and fixatives and found in diverse products including cosmetics and lotions, may be a culprit as well. A recent study by the Centers for Disease Control and Prevention that measured phthalate concentrations in urine samples found evidence of widespread exposure. The highest levels of exposure to monobutyl phthalate, a reproductive toxicant in animals, were found in women of reproductive age. Phthalates are classified among chemicals thought to disrupt reproductive hormones. From that information, the scientists suspect a breast cancer link, too.

Sandler says retrospective and case-control studies such as the Long Island Breast Cancer Study Project and the Carolina Breast Cancer Study have the advantage of including large numbers of breast cancer cases, but because they are retrospective, they may not accurately characterize levels of exposure prior to the onset of breast cancer. Instead, she and Weinberg believe a prospective study will be better able to address factors that cannot easily be measured retrospectively or that may change once a woman develops breast cancer.

Besides collecting biologic and environmental samples such as blood, urine, tap water, and household dust from subjects at the outset of the study, researchers will also use questionnaires to gather a multitude of data about health history, past environmental exposures, and lifestyle. Sandler believes analyses based on samples collected well before the onset of disease will yield more accurate data for at least two reasons. "One," she explains, "women's exposure patterns could change once they are sick, and if you are measuring something that reflects recent exposures, like diet in the last twenty-four hours, what you get may not represent usual exposure levels. Two, the illness itself or its treatment may lead to changes in the levels.... If you are measuring something intrinsic [for example, hormones or immune markers], you would not know if the levels were high because the women were sick or if high levels in the past caused the women to become sick."

To execute such an ambitious study, the research team will need large numbers of participants. Though Sandler and Weinberg expect recruiting the large cohort necessary to generate a statistically viable group to be a challenge, they plan to surmount this hurdle by going after precisely the participants with the most motivation to enroll and stay enrolled--women who have a sister with breast cancer.

"First-degree relatives, especially sisters, have about two times the risk of developing breast cancer as the average woman," Sandler says. Sisters of breast cancer patients are also likely to be within the same age range (and therefore at an age when they are at risk for developing breast cancer) and likely to have been exposed to the same environmental risk factors during early childhood. And of course they share many of the same genes. Sandler says sisters are also likely to share genetic polymorphisms that might affect the way a person's system handles carcinogens or repairs DNA.

The study should also fuel rewarding professional collaborations within the NIEHS, the NIH, and eventually with extramural colleagues too. "We're already building partners [such as the National Cancer Institute] into the study's design," Sandler asserts.

In focus groups of women of diverse socioeconomic, racial, and lifestyle backgrounds, Sandler and Weinberg have discovered that the women have one thing in common: as sisters of breast cancer patients, they share a common concern over the disease. That, says Sandler, makes them more likely to want to participate in the study and more willing to comply with the demands of the study, which will lead to a better response rate and higher retention rate over time.

The researchers are building a Web site for the project, developing a marketing plan to attract study participants (including advertisements in women's magazines and cancer publications), finalizing questionnaires, and securing preliminary scientific reviews to validate the integrity of the study's design. "We also have to work things out like standardized blood sampling," Weinberg adds.

"We're hoping by the first of next year to phase in the study," Sandler says. "We'll go into three or four test cities to try the marketing plan and see if it gets women to respond. Then we'll work on getting the bugs out and assess what we need to do to get an ethnically diverse group of women," she adds. In this first phase, 2000 women will be enrolled in the study. Sandler and Weinberg expect it to take three years to enroll all 50,000 participants.

Women enrolled in the study will provide periodic updates during the study and will be given follow-up questionnaires to account, for example, for diet changes or menopause. "By starting with a cohort of women and following them over time, we'll be able to look at many conditions of concern to women in addition to breast cancer, like osteoporosis and thyroid and autoimmune disorders," Sandler says. "This [database of study information] will be a rich resource for lots of nested studies," she adds, "and for addressing new questions that come up over time."

Over time, the study cohort's value to the researchers will increase. "If we do the study for five years, we'll have enough [breast cancer] cases to look at the interaction between many of the exposures of interest and the genes of interest," Sandler explains. "Going ten years for rare exposures or rare genes ... will give us even more statistical power."