The USPSTF Draft Recommendations on Medication Use to Prevent Breast Cancer: Comments by Women’s Health Experts

National Center for Health Research, February 11, 2019

Thank you for the opportunity to express our views on the draft recommendations for the use of tamoxifen, raloxifene, and aromatase inhibitors to prevent breast cancer in women at increased risk.

We are writing on behalf of five nonprofit health organizations that have examined issues pertaining to primary prevention of breast cancer for decades:

  • The National Center for Health Research is a nonprofit think tank that conducts, analyzes, and scrutinizes research, policies, and programs on a range of issues related to health and safety. We do not accept funding from companies that make products that are the subject of our work.
  • The Jacobs Institute of Women’s Health’s mission is to identify and study aspects of healthcare and public health, including legal and policy issues, that affect women’s health at different life stages and to foster awareness of and facilitate dialogue around issues that affect women’s health.
  • The National Women’s Health Network (NWHN) is a non-profit advocacy organization that works to improve the health of all women by supporting informed consumer decision-making.
  • Our Bodies Ourselves is a nonprofit educational and advocacy organization that has advanced the health and human rights for girls and women for almost 50 years.
  • PharmedOut is a Georgetown University Medical Center project that exposes inappropriate pharmaceutical marketing tactics and advances evidence-based prescribing practices.

We have several very strong concerns about the draft proposal and strongly urge USPSTF to revise the recommended guidelines for the following reasons:

  1. The procedures and analytic framework used in this evidence review are insufficient to provide useful recommendations for patients or their physicians. In fact, they focus too heavily on efficacy of drugs to prevent breast cancer with too little attention to balancing risks and benefits for healthy women who might consider these prevention strategies.  The Key Questions and Contextual Questions established for the review do not adequately address the issues we raise below.  The USPSTF questions should be revised to better assess cancer risk, potential benefit, and potential harm, as well as to better address the importance of patients participating in decision-making, the role of education and counseling, and patient-centered care.  A more rigorous and comprehensive assessment of existing data and remaining gaps in knowledge would lead to more appropriate recommendations for both providers and patients.
  2. The recommendations are based on the impact of hormonal drugs on relative rather than absolute risk of breast cancer as well as adverse events.  According to calculations by the National Center for Health Research, the lifetime risk of breast cancer attributed to tamoxifen would be reduced from 12% to 8% if tamoxifen is taken over 5 years.[1] [2] At the same time, however, tamoxifen would more than double the risk of endometrial cancer, which NCI describes as 2.9% for the average woman[3],  and therefore that would increase to 6.5%2.  The lifetime risk of thromboembolism for women is 20% and with a relative risk of almost 2[4], tamoxifen would increase that risk to 39%2.  Similarly, raloxifene reduces the lifetime risk of breast cancer from 12% to 5%1 2 but the 1.5 relative risks described by the USPSTF increase the risk of thromboembolism for patients taking raloxifene from 20% to approximately 31%4.  Aromatase inhibitors (AI) also lower the absolute risk of breast cancer from 12% to 5%1 2, while the lifetime risk of thromboembolism for AI patients increases from 20% to 25%4.  The risk of fractures increases with AI and decreases with raloxifene, but those measures are primarily based on diagnostic tests such as x-rays and bone mineral density, rather than pain, other quality of life issues, or disability. 2 [5]  Overall, these statistics indicate that the increases in absolute risk for several serious outcomes are considerably higher than the decrease in absolute risk of breast cancer.
  3. The recommendations have switched from shared decision-making to “offer to prescribe.” In 2013, USPSTF recommended that physicians engage in informative conversations with their high-risk patients to determine together whether the benefits of hormonal therapy outweigh the risks.[6]  In contrast, the proposed new recommendation is that physicians “offer” these hormonal treatments to high-risk patients, wording far closer to recommending a treatment than to discussing the risks and benefits in a shared decision-making conversation.2
  4. The definition of women at “high risk of breast cancer” is questionable.  Based on the NCI breast cancer risk calculator, a woman who has a ‘predicted’ risk of breast cancer higher than 1.7% over 5 years is considered high risk.[7] However, that includes almost all white women over 65 who have at least one additional risk factor or Hispanic women over 65 with just 2 risk factors such as menses starting before age 12 or having a first child born after age 30.7  In 2013, the USPSTF used a standard of 3% risk of breast cancer over the next 5 years, rather than 1.7%.6  In the proposed new recommendations, high risk is defined as at least 1.7% or 3%.2  Given the modest benefits of these hormonal treatments, we strongly urge that high-risk be defined as 3% or higher.
  5. Adverse effects of treatments on quality of life are not addressed or evaluated.  For example, women on tamoxifen have increased rates of hot flashes, arthralgia, vaginal dryness, and sexual dysfunction.[8]

In addition to the above issues, we urge the USPSTF to consider nonpharmacologic factors that can decrease the risk of breast cancer.  Major research studies conducted on tens of thousands of women indicate that habits such as maintaining a healthy weight, regular exercise, and a diet low in red meat and alcohol and high in fresh fruits and vegetables are associated with a lower risk of breast cancer.[9]

Overall, there are problems with the questions asked as well as major flaws in the proposed recommendations because they are inconsistent with the current evidence review.  Most importantly, there is no clear evidence that the modest benefits of these drugs or individual women outweigh the risks for women with a 1.7% or higher risk of breast cancer over the next five years, since the women are also at risk of thromboembolism, endometrial cancer, or bone fractures.  The 2013 recommendations defined high risk of breast cancer as a 3% risk in the following five years, which is a more appropriate standard for considering drugs with such well-established serious adverse events.  Only the women at very high risk of breast cancer and low risk of endometrial cancer and vascular disease should consider these drugs.  That requires a meaningful conversation with a knowledgeable physician, not an “offer” of a prescription.

We strongly urge USPSTF to change their recommendations by revising the framework for analysis and key questions and conducting a new evidence review as well as correcting the flaws identified above.  We anticipate that this would lead to recommendation that physicians should engage in shared decision-making rather than “offering” chemoprophylaxis. Both benefits and risks of these medications should be presented in absolute terms.  These drugs should be considered only for the highest-risk patients (at least 3% risk over five years).  And adverse effects of these drugs, including quality of life, should be adequately described as part of the informed consent process.

For questions or more information, please contact Dr. Varuna Srinivasan at

[1]. NCHR calculated the absolute risk based on the statistics provided by the National Cancer Institute. (2012). Breast Cancer Risk in American Women.

[2]. NCHR calculated the absolute risk based on the statistics provided by the United States Preventative Services Task Force Draft Recommendation Statement. (2019). Breast Cancer: Medication Use to Reduce Risk.

[3]. NCHR calculated the absolute risk based on the statistics provided by the National Cancer Institute. (2013). Uterine Cancer – Cancer Stat Facts.

[4]. NCHR calculated the absolute risk based on the statistics provided Bell EJ, Lutsey PL, et al. (2015). Lifetime Risk of Venous Thromboembolism in Two Cohort Studies. American Journal of Medicine.

[5]. NCHR calculated the absolute risk based on the statistics provided by International Osteoporosis Foundation. (2017). Facts and Statistics.

[6]. USPSTF. Breast Cancer: Medications for Risk Reduction. (2013).

[7]. National Cancer Institute. The Breast Cancer Risk Assessment tool.

[8]. Day, R. (2001). Quality of Life and Tamoxifen in a Breast Cancer Prevention Trial. Annals of the New York Academy of Sciences.

[9]. Thomson, CA et al. (2014). Nutrition and Physical Activity Cancer Prevention Guidelines, Cancer Risk, and Mortality in the Women’s Health Initiative. Cancer Prevention Research.