7
Recommendations for
Future Research
Although much has been learned about breast cancer and its relation to environmental exposures, much remains unclear. As the preceding chapters have illustrated, this reflects a mixture of circumstances. First, the scientific community is faced with conflicting and inconclusive results from past studies of some risk factors. Second, growing knowledge of the complex biology of breast cancer suggests a need to reframe hypotheses by focusing more on exposures in early life, examining associations with tumors of specific types, and considering mechanistically driven gene–environment interactions. Third, for a wide array of exposures, data are simply inadequate because exposure assessment methodologies have not been developed, informative studies may be nearly impossible to conduct in humans, and/or the existing tools and resources to conduct relevant research in animals or in vitro systems are limited.
With the complexity of breast cancer as a disease and of the combinations of biological and environmental factors that are potential contributors to it, the committee is persuaded that no one perspective will be sufficient to guide the future research that is needed to reduce the toll of this disease. Bringing together the perspectives of many disciplines into a transdisciplinary approach will be needed to generate innovative and cost-effective approaches to framing research questions, designing and conducting studies, developing new tools for data collection and analysis, and translating the results of research on risk factors into interventions that can reduce the risk of breast cancer.
Drawing on the insights developed in the previous chapters, the committee presents in this final chapter recommendations for research that
range from further examination of elements of the biology of breast development and carcinogenesis to tests of potential interventions to reduce risk. Important components of the work recommended here provide support for the research necessary to develop better tools for assessing the carcinogenicity of chemicals and pharmaceuticals as well as tools needed to strengthen epidemiologic research. The importance of a life course perspective runs throughout these recommendations.
Many of these recommendations are directed to both researchers and research funders. Researchers will have to conduct the work described here, but they will need the resources that come from a variety of sources. The National Institutes of Health and other federal agencies are major funders of research on breast cancer or they have unique authority or responsibility in certain areas. But the nation’s portfolio of research on breast cancer is also shaped in important ways by funders and other organizations in the private sector, such as Susan G. Komen for the Cure, that have the flexibility to pursue research topics and approaches that federal agencies may not. The committee urges effective and innovative collaborations to answer the many unresolved questions about the causes of breast cancer.
APPLYING A LIFE COURSE PERSPECTIVE TO RESEARCH ON BREAST CANCER
Progress has been made in understanding the biology of breast development, molecular mechanisms of carcinogenesis, the influence of the tissue microenvironment on breast cancer development, and some aspects of risk and prevention. But gaps remain in understanding of the etiology of breast cancer and the extent of environmental influences on breast cancer development.
Most epidemiologic studies have been obliged to focus on events in the few years or perhaps one to two decades before a breast cancer diagnosis. As described in Chapter 5, however, growing evidence suggests that events associated with breast carcinogenesis may occur much earlier—in young adulthood, puberty, childhood, and in utero. The effect of radiation, for instance, is greater when exposure occurs around the time of puberty or earlier. Although information about some early life events, such as age when first giving birth or age at menarche, can be reliably retrieved, few studies have collected information on nonreproductive environmental exposures that may influence the occurrence of clinically detectable breast cancer many decades later.
To address gaps in knowledge about the origins of breast cancer, the committee determined that research should increasingly focus on the influence of environmental factors during potential windows of susceptibility over the life course. It is possible that some exposures later in life, after
childbearing is complete, have little effect on breast cancer risk whereas similar exposures, if incurred early in life, before completion of breast development, may increase risk for breast cancer. On the other hand, exposures later in life may increase the growth of cancerous cells that have lain dormant for years and that would, without the exposure, have continued to be dormant. Thus the committee recommends that future research address the timing of exposures in relation to a woman’s life course and explore vulnerable windows for specific exposures of concern.
Recommendation 1: Breast cancer researchers and research funders should pursue integrated and transdisciplinary studies that provide evidence on etiologic factors and the determinants of breast cancer across the life course, with the goal of developing innovative prevention strategies that can be applied at various times in life.
• Such studies should seek to integrate animal models that capture the whole life course and human epidemiologic cohort studies that follow individuals over long periods of time and allow for investigation of so-called “windows of susceptibility” wherein breast tissue may be especially sensitive to environmental influences (e.g., prenatal, childhood, and adolescent, and childbearing periods). Long-term follow-up of cohorts is critical because new, unexpected evidence frequently arises with extended follow-up.
• Topics warranting attention include (but are not limited to) the biology of breast development; the mechanisms of carcinogenesis early in life, including the role of the tissue microenvironment in tumor suppression and development, and differences that may be related to tumor type; differences in risk by tumor type; the potential contribution of timing of exposure to variation in risk; and analytical tools for investigating the potential for interactions among exposures and the impact of mixtures of environmental agents on biologic processes.
Other work to aid investigation of environmental influences on breast cancer risk includes
• identifying cellular, biochemical, or molecular biomarkers of early events leading to breast cancer and validating their predictive value for future risk for breast cancer;
• determining whether intermediate endpoints, such as indicators of breast development or peak height growth velocity, are valid and predictive biomarkers of risk for breast cancer so that research can
effectively identify predictors of change in risk earlier in life or with shorter study periods;
• investigating the role that environmental factors may have in the origins of breast cancers of different types (e.g., estrogen or progestin receptor positive [ER+, PR+] or receptor negative [ER–, PR–]; HER2/neu positive or negative; or triple negative, meaning being negative for all three types of receptors) to better understand the potential contribution of these factors to disparities in the incidence of types of breast cancers among racial and ethnic groups;
• exploring the value of linking information across cohort studies focused on different stages of life as a way to overcome the challenges of mounting single long-term follow-up studies; and
• ensuring that cohorts established primarily to study genetic determinants of cancer and other diseases improve their capacity to capture information about environmental exposures over the life course.
Rationale: From its examination of evidence on a selection of environmental factors, the committee sees particular benefit in further research to clarify the mechanisms underlying breast cancer.
Recommendation 2: Breast cancer researchers and research funders should pursue research to increase knowledge of mechanisms of action of environmental factors for which there is provocative, but as yet inconclusive, mechanistic, animal, life course, or human health evidence of a possible association with breast cancer risk.
High-priority topics include the following:
• Shift work: There is growing evidence that shift work resulting in the disruption of circadian rhythm is probably associated with increased risk for breast cancer. Currently, there are no known effective interventions other than avoidance of shift work, which will not be an option for many workers. The biological mechanisms and the potential contribution of light exposure during normal sleep periods are poorly understood. More needs to be learned about the biological processes and pathways through which shift work and circadian rhythm disruption, or other factors arising from shift work, relate to breast cancer. This includes investigation of hormonal effects of circadian disruption, the role of “clock genes” and signaling pathways in breast tissue development, how
disruption of those signaling pathways may contribute to initiation or progression of breast tumors, developing more detailed and standardized approaches to exposure assessment for use in epidemiologic research, and developing and testing the effectiveness of interventions that could mitigate the carcinogenic effects that may be associated with shift work.
• Endocrine activity: Exposure to chemicals with estrogenic or other properties relevant to sex steroid activity, such as bisphenol A (BPA), polybrominated diphenyl ethers (PBDEs), zearalenone, and certain dioxins and dioxin-like compounds, may influence breast cancer risk, especially if those exposures occur at certain life stages or in combination with exposure to other similar chemicals, certain dietary components, or other factors. Although the evidence on the association between breast cancer risk and individual chemicals in this category is not conclusive, current mechanistic hypotheses warrant further research to examine their activity, to investigate additive or greater potency across multiple chemicals, to explore the effects of timing of exposure, and to evaluate interactions with diet, body mass index, and other factors that may influence the relationship of these types of compounds to breast cancer risk.
• Genotoxicity: Animal studies have demonstrated that some mutagenic chemicals are capable of inducing malignant mammary tumors, and numerous animal models of breast carcinogenesis routinely use the potent mutagens 7,12-dimethylbenz[a]anthracene (DMBA) and N-methyl-N-nitrosourea (MNU) as reproducible initiators of those tumors. But these studies have shown that the effect is highly sensitive to the timing of the exposures and can be influenced by other factors. More research is needed to understand the degree to which mutagenic chemicals, such as polycyclic aromatic hydrocarbons (PAHs), benzene, and ethylene oxide, acting alone or in combination with other exposures at specific life stages, may contribute to breast cancer risk at current levels of exposure.
• Epigenetic activity: Recent studies have demonstrated that some chemicals, including BPA, while not genotoxic per se, can have important influences on gene expression that may be relevant to breast cancer risk. Relatively little is known about the importance for breast cancer risk of such epigenetic modifications by environmental chemicals. More fundamental research on the role of epigenetic modifications in breast cancer risk is needed.
• Gene–environment interactions: Although few such interactions have been identified, to some extent this may reflect the small number of discrete exposures for which relevant genes are currently identifiable. Limited evidence indicates, for example, that
genes governing acetylation efficiency may describe a susceptible subset of the population for which exposure to tobacco smoke has substantial influence on breast cancer risk. Likewise, isozymes of different enzymes involved in alcohol metabolism may affect breast cancer risks, particularly among those with high alcohol intake.
Studies of Occupational Cohorts and Other Highly Exposed Populations
Rationale: Many known human carcinogens were first identified as a result of studies carried out in occupational settings where workers were subject to chemical and physical exposures that were typically higher than those experienced by the general population. When many of the early occupational studies were carried out, relatively few women were in the workforce. Changes in the typical workplace and the presence of more women in the workforce, both in the United States and internationally, make it appropriate to revisit occupational studies as a possible means to identify some exposures that increase risk for breast cancer. These studies should account for not only comparisons of breast cancer incidence associated with various work assignments or job titles, but also the distribution of known breast cancer risk factors among workers to ensure that the analyses of exposure-related risk are not confounded by differences among types of workers in the prevalence of these other known risk factors.
Outside the workplace, other events such as industrial accidents or contamination episodes can lead to high exposures for specific population groups. Sometimes these events provide opportunities to investigate the impact of specific timing of exposures, as in the case of the survivors of the atomic bombs in Hiroshima and Nagasaki, or the population living in the vicinity of the industrial accident in Seveso, Italy, and exposed to high levels of dioxin. High-dose or long-term medical exposures have also lent themselves to study through the assembly of cohorts from records of patients treated for specific diseases or conditions.
Recommendation 3: Breast cancer researchers and research funders should pursue studies of populations with higher exposures, such as occupational cohorts, persons with event-related high exposures, or patient groups given high-dose or long-term medical treatments. These studies should include collection of information on the prevalence of known breast cancer risk factors among the study population. Support for these studies should include resources for the development of improved exposure assessment methods to quantify chemical and other
environmental exposures potentially associated with the development of breast cancer.
New Exposure Assessment Tools
Rationale: A life course perspective on breast cancer suggests that critical periods of vulnerability may exist during in utero development, in childhood, adolescence, and early adulthood, and at older ages. Exposure assessment becomes particularly challenging if the interval between critical exposure events and the point at which breast cancer can be diagnosed extends over decades.
If evidence of exposure is retained in either environmental media or the human body, measurements made long after exposure may provide an adequate basis for estimating an earlier exposure. To be able to do so requires sufficient knowledge of the patterns of persistence of chemical compounds and their metabolites, the determinants of variability in retention, and the variation in exposure levels over time. If evidence of exposure is not retained, one-time measurements are unlikely to be an adequate basis for assessing true exposure unless it is known that an individual’s exposure is consistent over long periods.
To effectively study exposures over long time periods, research protocols may need to obtain measurements of exposure at multiple time points. However, because repeated measurements can be prohibitively burdensome, it may be necessary to develop alternative strategies that rely on external indicators of exposure. For instance, if, hypothetically, 50 percent of the body burden of a chemical exposure is from consumption of liquids from plastic bottles, then questionnaires about such behavior patterns may be a more reliable basis for assessing exposure than measurements of urinary metabolites. If, additionally, persons who consume fluids from plastic bottles do so consistently over years or decades, then this approach may be reasonable for establishing past exposures as well.
Recommendation 4: Breast cancer and exposure assessment researchers and research funders should pursue research to improve methodologies for measuring, across the life course, personal exposure to and biologically effective doses of environmental factors that may alter risk for or susceptibility to breast cancer.
Such research should encompass
• improving measurements in the environment and assessing variation over time and space;
• determining routes of exposures and how they vary over time and over the life course;
• evaluating how products are used and the extent to which actual usage deviates from label instructions (e.g., home pesticide applications) as a critical component of exposure assessment, and focusing on the impact on personal exposures;
• incorporating use of advanced environmental dispersion modeling techniques with accurate emissions and air monitoring data to characterize specific population exposures;
• measuring compounds and their metabolites in biospecimens, including specimens obtained by noninvasive means;
• understanding pharmacodynamics and pharmacokinetics and how they vary by lifestage, body weight, nutrition, comorbidity, or other factors;
• developing other biomarkers of exposure through early biologic effects (DNA adducts, methylation, tissue changes, gene expression, etc.);
• using existing and yet-to-be-established human exposure biomonitoring programs (e.g., breast milk repositories) by geographic areas; and
• validating exposure questionnaires through various strategies.
RESEARCH TO ADVANCE PREVENTIVE ACTIONS
Minimizing Exposure to Ionizing Radiation
Rationale: As discussed in Chapters 3 and 6 and Appendix F, some of the strongest evidence reviewed by the committee indicated a strong causal association between breast cancer and ionizing radiation. However, population exposures to ionizing radiation in medical imaging are increasing. Chapter 6 sets forth a series of steps that can be taken by various groups and in various settings to reduce exposures to ionizing radiation and therefore reduce risks for breast and other cancers. However, many unknowns remain about the best ways to achieve these reductions. This work might include investigation of the feasibility of developing cost-effective forms of imaging that do not rely on ionizing radiation. Further research is warranted to clarify the extent of population risks, unnecessary uses of medical radiography, and the best means to maximize its benefits and minimize its harms.
Recommendation 5: The National Institutes of Health, the Food and Drug Administration, and the Agency for Healthcare Research and Quality should support comparative effectiveness research to assess
the relative benefits and harms of imaging procedures and diagnostic/follow-up algorithms in common practice. This research effort should also assess the most effective ways to fill knowledge gaps among patients, health care providers, hospitals and medical practices, industry, and regulatory authorities regarding practices to minimize exposure to ionizing radiation incurred through medical diagnostic procedures.
Developing and Validating Preventive Measures
Rationale: Some breast cancer risk factors appear to be modifiable, but it is important to determine what modifications of these environmental exposures can be most effective in reducing risk and when during the life course these changes need to occur. For example, overweight and obesity are recognized as increasing risk for postmenopausal breast cancer, but the contribution of weight loss to reducing risk is much less clear.
Recommendation 6: Breast cancer researchers and research funders should pursue prevention research in humans and animal models to develop strategies to alter modifiable risk factors, and to test the effectiveness of these strategies in reducing breast cancer risk, including timing considerations and population subgroups likely to benefit most.
Particular aspects of prevention that require attention include
• when weight loss is most likely to be beneficial in reducing risk for postmenopausal breast cancer;
• effective strategies for achieving and maintaining weight loss in different risk groups;
• effective and sustainable methods to prevent obesity;
• the feasibility of interventions in early life and development that may influence breast cancer risk in adult life such as preventing childhood obesity, increasing physical activity, and minimizing exposures to potentially harmful environmental carcinogens;
• approaches to prevention that respond to the differing breast cancer experience of various racial and ethnic groups; and
• dissemination and adoption of effective prevention strategies.
Chemoprevention—Research on Medications to Reduce Breast Cancer Risk
Rationale: Breast cancer is likely to remain a major source of morbidity for many decades to come. However, if early life events are critical in breast cancer carcinogenesis, then most women may have already had some
critical exposures by mid-life, when the incidence of breast cancer increases. Avoiding other exposures later in life, such as hormone therapy, may delay or even prevent breast cancer in some women, but it may be that further reductions in risk later in life are most efficiently achieved through pharmaceutical interventions.
Research has demonstrated that drugs that alter responses to estrogen (e.g., tamoxifen, raloxifene) or production of estrogen (e.g., aromatase inhibitors) can substantially reduce risk of ER+ breast cancer (Cummings et al., 2009; Nelson et al., 2009; Goss et al., 2011). The Food and Drug Administration (FDA) has approved use of tamoxifen and raloxifene for this purpose by women who are considered at increased risk of breast cancer and are not at increased risk for cerebrovascular disease. Other medications, such as bisphosphonates and metformin, are under study to assess their potential role in reducing the risk of either ER+ or ER– breast cancer (Cuzick et al., 2011). But relatively few eligible women have chosen to use tamoxifen and raloxifene, at least in part because they are associated with increased risk for serious adverse health effects, including endometrial cancer and stroke (Fisher et al., 2005; Vogel et al., 2010).
The desirability of drugs that can reduce breast cancer risk must be balanced against any potential dangers associated with the use of those drugs. These dangers are of particular concern for the large numbers of women who would not have developed breast cancer even without medication, as well as for the smaller numbers of women who develop breast cancer despite using them.
Additional research into medications that can reduce risk for breast cancer with minimal added risk of other serious adverse health effects should be fostered and accelerated. Studies should include sufficient follow-up, both during the study when the medications are being used and after what is anticipated to be the typical period of use, to provide an adequate basis for determining the benefits and risks that may be associated with the medication. Furthermore, because the approved drugs only reduce the risk of ER+ breast cancer, research is critically needed to find effective ways to reduce the risk of other forms of breast cancer, including triple negative breast cancer and other hard-to-treat forms of breast cancer that may have a disproportionate impact at younger ages or among African American, Asian, or Hispanic women.
Recommendation 7: Breast cancer researchers and research funders should pursue continued research into new breast cancer chemoprevention agents that have minimal risk for other adverse health effects. This work should include efforts to identify chemopreventive approaches for hormone receptor negative breast cancer.
Adequately sized primary prevention studies will be needed to allow for estimation of both benefits and risks. Research plans should also include long-term follow-up to identify any changes in risk patterns for types of breast cancer or other effects that only become evident beyond the time frame of the initial study and analyses.
TESTING TO IDENTIFY POTENTIAL BREAST CARCINOGENS
In Vivo Testing for Carcinogenicity
Rationale: Testing in animals is currently an established component of the evaluation of the carcinogenicity of chemicals in industry and commerce, but it is unclear which whole-animal test protocols are best suited for screening for possible human breast carcinogens. Human sensitivity to breast cancer has been demonstrated for exposures in utero (e.g., diethylstilbestrol [DES]), before and during puberty (e.g., radiation), and postmenopausally (e.g., combination hormone therapy). Studies in animals have also demonstrated that some exposures early in life that are not themselves carcinogenic may alter susceptibility to carcinogens encountered later in life.
But these age windows are typically not included in standard cancer bioassays such as those used in conjunction with the registration of pesticides and pharmaceuticals. The standard protocols commonly begin exposures when animals are 7 to 8 weeks of age. Thus they miss the rapid mammary ductal growth and branching during pubertal development, a period of heightened sensitivity in the rat to adverse effects from chemical exposures. These protocols also miss gestational exposures and terminate the experiments at 2 years, which omits the older age period, a time of increasing incidence of breast cancer in humans.
Interpretation of rodent bioassays for mammary carcinogenicity is complicated by certain characteristics of the animals typically used for these studies. The mouse strains appear generally insensitive to hormonally induced mammary tumors. Conversely, a commonly used rat strain is overly sensitive to the occurrence of constant estrus and early reproductive senescence. Constant estrus and early reproductive senescence can tend to increase the incidence of mammary tumors, but this phenomenon may not be relevant for humans. Thus results of bioassays of hormonally active agents are confounded when mammary tumors are increased concomitantly with constant estrus in the treated rats. With the insensitivity of mice, negative results from tests in mice are not necessarily a reliable indicator of lack of mammary carcinogenicity.
To increase the ability to detect statistically significant increases in cancer rates in the limited number of animals that can be used in toxicity and carcinogenicity testing, chemicals are typically administered at dose
equivalents that are far higher than the exposures humans would normally have. Pharmacokinetic and metabolic differences between high- and low-dose chemical exposures complicate the prediction of risks at lower doses that would be more comparable to human experience.
Finally, standard bioassay protocols for regulatory testing generally test individual chemicals. However, humans are generally exposed throughout life to a myriad of hormonally active and genotoxic chemicals. Some experimental protocols used in cancer research employ mixed exposures (e.g., in utero exposure to one agent and subsequent high-dose exposure to a genotoxic chemical during a period of rapid ductal growth). Other tests look for abnormal development of the mammary gland following in utero or early in life exposure, to identify early predisposing events. In reports from some research studies, it is difficult to assess the level of attention devoted to important design issues such as randomization, blinded assessment of endpoints, and standardization of endpoints.
Recommendation 8: The research and testing communities should pursue a concerted and collaborative effort across a range of relevant disciplines to determine optimal whole-animal bioassay protocols for detection and evaluation of chemicals that potentially increase the risk of human breast cancer.
The development of these protocols should address several issues, including the following:
• potential differences in sensitivity to carcinogenic effects and during different life stages;
• the appropriateness and limitations of the rodent strains and species used for testing, and potential alternatives;
• the frequency, magnitude, and route of dosing, and the possible need for alternative protocols that provide improved relevance for predicting human risk;
• the utility of genetically engineered mouse models, which show promise for studying breast tumor formation and progression and the effectiveness of treatments; and
• standard practices for conducting and reporting results of animal studies.
This work will probably also require targeted mechanistic and pharmacokinetic studies to assess appropriate dosing levels in test protocols to better address human exposure circumstances, including the influence of life stage, genetic variability, and multiple chemical exposures.
New Approaches to Toxicity Testing
Rationale: Most of the thousands of chemicals used in industry and commerce have not been tested for their potential to contribute to breast and other cancers. Screening all chemicals with the standardized approaches used for pharmaceuticals and pesticides is impracticable because of the time and resources (including large numbers of test animals) that would be required (NRC, 2006, 2007). Furthermore, the tests are done chemical-by-chemical, which does not address the potential consequences of exposures to mixtures of chemicals or interactions with other ongoing exposures (e.g., dietary components). The high doses used in testing also introduce uncertainty and limitations for predicting risks at lower doses that are relevant to human exposures.
Under the broad umbrella of the Tox21 (EPA, 2011) and National Toxicology Program initiatives, new toxicity testing approaches are being developed to more rapidly and accurately screen and identify the toxicity of chemicals encountered in human environmental, occupational, and product exposures. This effort relies on the elucidation of key toxicity pathways involved in human disease, and on the development of sensitive, rapid testing approaches to determine a chemical’s potential to perturb such pathways and at what concentrations. A variety of tests are being developed and considered: high-throughput in vitro screens that use cell components and engineered cells; toxicogenomic responses following cellular, tissue, and organism exposures; novel animal systems (e.g., the roundworm, Caenorhabditis elegans); and limited, targeted testing in laboratory animals to anchor test results and understand mechanisms, new chemistries, and pharmacokinetics (Dix et al., 2007; NRC, 2007).
The new approach also calls for the use of pharmacokinetic evaluations, human biomonitoring data, and epidemiologic results to establish the predictive ability of the tests. Pharmacokinetics will be an important consideration in understanding test results, in studying uptake and distribution to target cells, and in examining the biochemical transformations that make the chemical biologically active or inactive. This aspect of the effort is currently a significant challenge in the development of high-throughput and other in vitro tests.
Because breast cancer is a major contributor to morbidity among women, these tests should address pathways that underlie the basic mechanisms of breast cancer—mutagenesis, estrogen receptor signaling, epigenetic programming, growth promotion via mitogenic cell signaling, and modulation of immune functioning—with particular attention to cell types and environments relevant to breast cancer. They should also take into account alterations at the whole-organ level, and they should be relevant to typical human exposures, which often occur at low doses and as mixtures.
Recommendation 9:
a. The research and testing communities should ensure that new testing approaches developed to serve as alternatives to long-term rodent carcinogenicity studies include components that are relevant for breast cancer.
To be relevant for breast cancer, it will be necessary to be able to assess changes in susceptibility through the life course and mechanisms characteristic of hormonally active agents. The test development should also include exploring the predictive value of in vitro and in vivo experimental testing for site-specific cancer risks for humans.
b. A research initiative should assess the persistence and consequences for mammary carcinogenicity of abnormal mammary development and related intermediate outcomes observed in some toxicological testing.
As useful predictors of increased mammary cancer risk become available, intermediate outcomes may aid in identifying chemicals that may pose increased risk of human breast cancer when exposures occur early in life.
c. Research should be conducted to improve understanding of the potential cumulative effects of multiple, small environmental exposures on risk for breast cancer and the interaction of these exposures with other factors that influence risk for breast cancer.
Improved understanding of both mixed and serial low-dose exposures is critical for the interpretation of in vivo results and is of heightened importance for understanding the results of the emerging in vitro tests. Relevant exposures may come from sources that include food, pharmaceuticals, and the general environment. It is also critical for the understanding of epidemiologic and in vivo and in vitro experimental research results on the health effects of chemical mixtures that are characteristic of human environmental exposures.
Identifying Breast Cancer Risks Associated with Hormonally Active Pharmaceutical Products
The committee sees a need to ensure that mechanisms for detection and assessment of breast cancer risks associated with use of drugs regulated by FDA are adequate. It also recognizes that enhanced methods to detect breast cancer risks represent only one specific dimension of a more general interest in strengthening FDA’s ability to ensure the safety and timely avail-
ability of prescription and over-the-counter drugs (IOM, 2007a,b) and in strengthening the science to support FDA’s regulatory work (e.g., IOM, 2011).
Menopausal hormone therapy was originally developed to control menopausal symptoms. Some health professionals advocated long-term and substantially expanded use in anticipation that it would reduce age-related health problems, including cardiovascular disease and memory disorders, even before clear evidence was in hand. Although these products are effective in reducing menopausal symptoms and osteoporotic fractures while women are taking them, evidence from the Women’s Health Initiative examining multiple health outcomes in a randomized trial design showed that use of a combination of estrogen and progestin in postmenopausal hormone preparations increases risk of breast cancer and stroke and does not provide overall benefits for cardiovascular risk or memory disorders (Writing Group for the Women’s Health Initiative Investigators, 2002). This experience is an illustration of the dangers of exposing millions of healthy women to pharmacological doses of exogenous hormones without sufficient evidence of net benefit. Decades of study have also confirmed a small excess risk of breast cancer among current users of oral contraceptives (Collaborative Group on Hormonal Factors in Breast Cancer, 1996; Marchbanks et al., 2002; Strom et al., 2004; IARC, 2011). Although the increased risk of breast cancer that is associated with use of combination hormone therapies, including oral contraceptives, declines after treatment stops, women should be aware of the full range of potential harms as well as the benefits when they decide whether to use any form of hormone therapy, including those touted as safe because they are “bioidentical” or “natural.”
New Approaches to Testing Hormonally Active Candidate Pharmaceuticals
Rationale: Given the evidence for hormonal influences on the development of breast cancer, the committee is concerned that testing required to gain marketing approval for various hormonally active pharmaceuticals that are already on the market or that are being developed does not adequately address the potential impact on the risk for breast cancer. For example, the 2-year rodent carcinogenicity studies done for Prempro, the combined estrogen–progestin product used in the Women’s Health Initiative, showed a reduction in mammary tumors in rats (Ayerst Laboratories, 2003), and premarketing human safety and efficacy studies are generally too small and too brief to detect an effect on the incidence of breast cancer. Given that some hormonal products have been found to increase the risk of breast cancer, it is important that new postmenopausal hormone preparations, including those advertised as bioidentical or natural hormones, have
an adequate evidence base to support any claims that they do not cause breast cancer. It is also important to have an adequate understanding of the implications for breast cancer risk of the hormone composition and dosing schedules of new oral contraceptives (e.g., a preparation that causes a woman to have only four menstrual periods per year).
Identifying hormonally active substances is complex, in that various models are used to measure hormonal activity and the activity levels detected for a substance may differ depending on the model and dose used. It is important to assess the effectiveness of current testing protocols for hormonally active products in providing indicators of the potential for increased risk of breast cancer, and to develop and validate new testing practices where needed.
Recommendation 10: The pharmaceutical industry and other sponsors of research on new hormonally active pharmaceutical products should support the development and validation of better preclinical screening tests that can be used before such products are brought to market to help evaluate their potential for increasing the risk of breast cancer.
A suite of in vitro and in vivo tests will likely be needed to address the different mechanisms of action that may be relevant over the life course (in utero, early infancy, pre- and postpuberty, pregnancy, and pre- and postmenopause). If such tests can be developed and validated, FDA should require submission of the results as part of the process for approving the introduction of new hormonal preparations for prescription or over-the-counter use. These tests may also prove useful in testing environmental chemicals.
Postmarketing Studies of Hormonally Active Products
Rationale: With the demonstration that use of certain hormonally active prescription drugs is associated with an increased risk of breast cancer and other adverse health effects, it is important to investigate whether use of other hormonally active drugs is also associated with increased risk. The Food and Drug Administration Amendments Act of 2007 gave FDA the authority to require postmarketing studies or clinical trials for approved drugs when adverse event reporting would not be sufficient to assess a known or suspected serious risk (FDA, 2011). Because adverse event reporting systems are generally better suited to the detection of adverse events that occur soon after use of a drug than to events such as breast cancer that take years to develop, formally conducted studies appear necessary to assess the potential breast cancer risk.
Recommendation 11: FDA should use its authority under the Food and Drug Adminisration Amendments Act of 2007 to engage the pharmaceutical industry and scientific community in postmarketing studies or clinical trials for hormonally active prescription drugs for which the potential impact on breast cancer risk has not been well characterized.
Study oversight should be designed to mitigate against bias and conflict of interest of study sponsors. Special attention should be accorded to those products that represent a substantial change in pharmacologic composition or dosage schedule from products currently on the market. The studies should be adequately powered to quantitatively explore the possible contribution of the products to breast cancer risk, as well as other risks that have been associated with these classes of drugs (e.g., cardiovascular effects).
UNDERSTANDING BREAST CANCER RISKS
Researchers, health care providers, and the public all have an incomplete picture of the components of breast cancer risk. Further work is needed to clarify the contribution of recognized risk factors to differences and changes in the incidence of breast cancer and to determine the most effective ways to convey information about breast cancer risk.
Risk Modeling
Rationale: Public health messages about ways to reduce risk should rest on strong science on the attribution of risks to various causal factors. Systematic modeling approaches are needed to refine the estimates of the proportion of breast cancer in the United States and other countries that can be attributed to known factors, especially modifiable factors. Substantial proportions of the increase in breast cancer incidence rates in the United States over the past century, and of the differences in rates of breast cancer between less developed countries and more affluent countries, are probably due in large part to differences over time and between countries in the prevalence of established breast cancer risk factors (e.g., age at menarche, age at first birth, parity, use of menopausal hormone therapy, physical activity, weight and weight change). Few reliable estimates of these temporal and international differences in risk factor prevalence exist.
Developing data on changes in the prevalence of known risk factors, along with changes in breast cancer incidence, should permit statistical modeling of the size of these proportions associated with individual risk factors and combinations of these risk factors. This information will also help in determining the magnitude of risk associated with other unidentified
factors, which may include other environmental exposures. Of particular interest are the modifiable risk factors.
Risk modeling on both the individual and population levels will benefit greatly from improved understanding of the etiology of breast cancer. As the science improves, risk models can also help guide future research investments and policy decisions for population-level interventions. A collaborative approach, such as that used by the Cancer Intervention and Surveillance Modeling Network (CISNET) consortium, may be a cost-effective way to pursue some of this work.
Recommendation 12: Breast cancer researchers and research funders should support efforts to (1) develop statistical methodology for the estimation of risk of breast cancer for given sets of risk factors and that takes the life course perspective into account, (2) determine the proportion of the total temporal and geographic differences in breast cancer rates that can be plausibly attributed to established risk factors, and (3) develop modeling tools that allow for calculation of breast cancer risk, in both absolute and relative terms, with the goal of assessing potential risk reduction strategies, at both personal and public health levels.
Communicating About Breast Cancer Risks
Rationale: Accurate and effective communication of breast cancer risks is important for individuals, the public at large, and policy makers and public health officials. Individuals need to be able to make informed choices regarding risk factors, prevention opportunities, and health care appropriate to their risk circumstances. Research indicates that women may have a poor understanding of their risk of breast cancer, with both over- and underestimates of risk observed (Lipkus et al., 2001; Apicella et al., 2009; Waters et al., 2011). A systematic review under the auspices of the Cochrane Collaborative found that both health care providers and consumers understood risks of health outcomes better when those risks were presented as frequencies rather than as probabilities (Akl et al., 2011). Both thought the risks were lower when presented as absolute risk reduction than as relative risk reduction, and both were more persuaded by relative than absolute risks in terms of potential behavioral change. To allow a fair comparison of risks and benefits, supplementing presentation of relative measures with absolute ones is useful because other disease endpoints may be more or less common than breast cancer.
From a public health policy and practice perspective, it is important to determine where risks lie and the potential for benefit and risk at a population level. Uncertainty is inherent in risk prediction, and it can be difficult or impossible to establish that an exposure is not associated with cancer
risk. However, moderate or large risks can be ruled out with reasonable confidence when studies with robust and appropriate research designs and analyses have been conducted in populations with relevant exposures. Meaningful differences in risk need to be effectively communicated to the public, health care providers, and policy makers so that limited funds can be invested in the most promising research and intervention strategies.
Recommendation 13: Breast cancer researchers and research funders should pursue research to identify the most effective ways of communicating accurate breast cancer risk information and statistics to the general public, health care professionals, and policy makers.
Because people differ in their health literacy, their numeracy (ability to understand numerical information), and in their preferred modes of learning, multiple communication strategies, modes, and messaging tactics will be needed to reach diverse communities and stakeholders. Among the topics that should receive attention in this research are
• perception and comprehension of different ways to present messages (numbers, graphs, text), modalities of communication (audio, video, print, face-to-face, and multiple modalities, etc.), as well as the content of the messages themselves;
• ways that personal experiences (e.g., family history) affect the ability to absorb messages;
• determination of the similarities and differences in how individuals from diverse racial, ethnic, educational, and occupational groups understand and respond to breast cancer risk information that is presented various ways;
• comprehension of terms such as relative risks, absolute risks, and hazards;
• ways to improve translation of research results into messages that can effectively convey the implications of the results for women in different risk categories, women from diverse racial and ethnic groups, health care providers, and public health decision makers; and
• ways to convey information about chemicals for which there is suggestive evidence of risk from experimental studies.
Breast cancer is a leading cause of cancer morbidity among women in the United States and many other countries. Major advances have been made in understanding its biology and diversity, but more needs to be
learned about the causes of breast cancer and how to prevent it. Familiar advice about healthful lifestyles appears relevant, but it remains difficult to discern what contribution a diverse array of other environmental factors may be making. Important targets for research are the biologic significance of life stages at which environmental risk factors are encountered, what steps may counter their effects, when preventive actions can be most effective, and whether opportunities for prevention can be found for the variety of forms of breast cancer.
Akl, E. A., A. D. Oxman, J. Herrin, G. E. Vist, I. Terrenato, F. Sperati, C. Costiniuk, D. Blank, et al. 2011. Using alternative statistical formats for presenting risks and risk reductions. Cochrane Database Syst Rev (3):CD006776.
Apicella, C., S. J. Peacock, L. Andrews, K. Tucker, M. B. Daly, and J. L. Hopper. 2009. Measuring and identifying predictors of women’s perceptions of three types of breast cancer risk: Population risk, absolute risk and comparative risk. Br J Cancer 100(4):583–589.
Ayerst Laboratories. 2003. Prempro™. [product labeling] http://www.accessdata.fda.gov/drugsatfda_docs/label/2003/20527s30bl.pdf (accessed October 18, 2011).
Collaborative Group on Hormonal Factors in Breast Cancer. 1996. Breast cancer and hormonal contraceptives: Collaborative reanalysis of individual data on 53,297 women with breast cancer and 100,239 women without breast cancer from 54 epidemiological studies. Lancet 347(9017):1713–1727.
Cummings, S. R., J. A. Tice, S. Bauer, W. S. Browner, J. Cuzick, E. Ziv, V. Vogel, J. Shepherd, et al. 2009. Prevention of breast cancer in postmenopausal women: Approaches to estimating and reducing risk. J Natl Cancer Inst 101(6):384–398.
Cuzick, J., A. DeCensi, B. Arun, P. H. Brown, M. Castiglione, B. Dunn, J. F. Forbes, A. Glaus, et al. 2011. Preventive therapy for breast cancer: A consensus statement. Lancet Oncol 12(5):496–503.
Dix, D. J., K. A. Houck, M. T. Martin, A. M. Richard, R. W. Setzer, and R. J. Kavlock. 2007. The ToxCast program for prioritizing toxicity testing of environmental chemicals. Toxicol Sci 95(1):5–12.
EPA (Environmental Protection Agency). 2011. Tox21. http://www.epa.gov/ncct/Tox21/ (accessed July 29, 2011).
FDA (Food and Drug Administration). 2011. Guidance for industry. Postmarketing studies and clinical trials—implementation of Section 505(o)(3) of the Federal Food, Drug, and Cosmetic Act. http://www.fda.gov/downloads/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/UCM172001.pdf (accessed April 26, 2011).
Fisher, B., J. P. Costantino, D. L. Wickerham, R. S. Cecchini, W. M. Cronin, A. Robidoux, T. B. Bevers, M. T. Kavanah, et al. 2005. Tamoxifen for the prevention of breast cancer: Current status of the National Surgical Adjuvant Breast and Bowel Project P-1 study. J Natl Cancer Inst 97(22):1652–1662.
Goss, P. E., J. N. Ingle, J. E. Ales-Martinez, A. M. Cheung, R. T. Chlebowski, J. Wactawski-Wende, A. McTiernan, J. Robbins, et al. 2011. Exemestane for breast-cancer prevention in postmenopausal women. N Engl J Med 364(25):2381–2391.
IARC (International Agency for Research on Cancer). 2011. IARC monographs on the evaluation of carcinogenic risks to humans. Vol. 100, Part A: Pharmaceuticals. Lyon, France: IARC.
IOM (Institute of Medicine). 2007a. Challenges for the FDA: The future of drug safety, workshop summary. Washington, DC: The National Academies Press.
IOM. 2007b. The future of drug safety: Promoting and protecting the health of the public. Washington, DC: The National Academies Press.
IOM. 2011. Building a national framework for the establishment of regulatory science for drug development: Workshop summary. Washington, DC: The National Academies Press.
Lipkus, I. M., W. M. Klein, and B. K. Rimer. 2001. Communicating breast cancer risks to women using different formats. Cancer Epidemiol Biomarkers Prev 10(8):895–898.
Marchbanks, P. A., J. A. McDonald, H. G. Wilson, S. G. Folger, M. G. Mandel, J. R. Daling, L. Bernstein, K. E. Malone, et al. 2002. Oral contraceptives and the risk of breast cancer. N Engl J Med 346(26):2025–2032.
Nelson, H. D., R. Fu, J. C. Griffin, P. Nygren, M. E. Smith, and L. Humphrey. 2009. Systematic review: Comparative effectiveness of medications to reduce risk for primary breast cancer. Ann Intern Med 151(10):703–715, W-226–W-735.
NRC (National Research Council). 2006. Toxicity testing for assessment of environmental agents: Interim report. Washington, DC: The National Academies Press.
NRC. 2007. Toxicity testing in the 21st century: A vision and a strategy. Washington, DC: The National Academies Press.
Strom, B. L., J. A. Berlin, A. L. Weber, S. A. Norman, L. Bernstein, R. T. Burkman, J. R. Daling, D. Deapen, et al. 2004. Absence of an effect of injectable and implantable progestin-only contraceptives on subsequent risk of breast cancer. Contraception 69(5):353–360.
Vogel, V. G., J. P. Costantino, D. L. Wickerham, W. M. Cronin, R. S. Cecchini, J. N. Atkins, T. B. Bevers, L. Fehrenbacher, et al. 2010. Update of the National Surgical Adjuvant Breast and Bowel Project Study of Tamoxifen and Raloxifene (STAR) P-2 Trial: Preventing breast cancer. Cancer Prev Res (Phila) 3(6):696–706.
Waters, E. A., W. M. Klein, R. P. Moser, M. Yu, W. R. Waldron, T. S. McNeel, and A. N. Freedman. 2011. Correlates of unrealistic risk beliefs in a nationally representative sample. J Behav Med 34(3):225–235.
Writing Group for the Women’s Health Initiative Investigators. 2002. Risks and benefits of estrogen plus progestin in healthy postmenopausal women: Principal results from the Women’s Health Initiative randomized controlled trial. JAMA 288(3):321–333.
