Psychotropic Medication Use and Postmenopausal Breast Cancer Risk

An estimated 268,600 new female breast cancer cases are expected in 2019 (1). Psychotropic medications have been associated with modest increases in breast cancer risk in some (2–4), but not other (5–7), epidemiologic investigations. Elevated prolactin levels are a common adverse effect of psychotropic drugs, especially with typical antipsychotics (first-generation drugs). In contrast, atypical antipsychotics (second-generation drugs), except for risperidone, cause smaller prolactin elevations. Higher circulating prolactin levels are associated with higher breast cancer risk, especially for hormone receptor–positive and postmenopausal disease (8). Given prior inconsistent results, we prospectively evaluated associations between psychotropic medication use and postmenopausal breast cancer risk within the Women's Health Initiative (WHI) cohort, a large prospective population-based cohort with high quality data on medication use and adjudicated breast cancer outcomes.

WHI enrolled postmenopausal women ages 50–79 years into observational study (OS) or clinical trial (CT) components from 1993 to 1997 at 40 clinical centers nationwide (N = 161,808). Participants provided written informed consent at enrollment, and institutional review board approval was obtained at each clinical center. For this analysis we excluded participants with a personal breast cancer history (N = 5,397) or <1 day follow-up time (N = 674), giving a final analytic cohort of 155,737 postmenopausal women.

Participants brought all current prescription and nonprescription medications and supplements to their baseline visit. A research nurse recorded each medication name and dosage. We classified reported antipsychotic medications as “typical” (fluphenazine, chlorpromazine, haloperidol, thiothixene, flupenthixol, and molindone) or “atypical” (risperidone, clozapine, olanzapine, and aripiprazole) based on their structures and mechanisms of action using UpToDate. Participants were categorized as using any psychotropic medications (no and yes) and separately by use of typical (no and yes) or atypical (no and yes) antipsychotics or lithium (no and yes); users of typical antipsychotics who did not also use atypical antipsychotics were classified as “no” for atypical use, and vice versa. Breast cancer cases were centrally adjudicated using medical records.

We compared baseline descriptive statistics between users and nonusers of psychotropic medications. Hazard ratios (HR) and 95% confidence intervals (CI) examining associations of psychotropic medications with breast cancer were estimated using Cox proportional hazards regression models. Follow-up time began at enrollment; participants were censored at either breast cancer diagnosis, death, loss to follow-up, or March 31, 2018, whichever came first. We decided a priori to adjust for age and WHI study arm (OS vs. CT) and arm of hormone therapy clinical trial given known differences in breast cancer risk across these groups; these adjusted HRs changed <2% when additional variables were included (i.e., characteristics summarized in Table 1). Thus, our final model adjusted for age and the WHI study participation variables.

We performed sensitivity analyses among the subgroup of women with regular mammograms during the first 10 years of follow-up, as determined by study protocol for CT participants or self-report of ≥6 mammograms during the 10-year period for OS participants (N = 133,754). We repeated analyses restricting to estrogen receptor–positive (ER⁺) breast cancer s. We also incorporated medication use at the year 3 follow-up visit, and repeated analyses as described above, starting follow-up time at year 3, and also estimating HRs for the consistency of psychotropic medication use between baseline and year 3 (never used, initiated use, stopped use, and consistent use).

Prevalence of psychotropic medication use was low (n = 642; 0.4%), with most users taking either a typical antipsychotic (n = 272; 42.4%) or lithium (n = 326; 50.8%; Table 1). During an average 14.8 (SD 6.5) years of follow-up, 10,097 invasive and 2,285 in situ breast cancers were diagnosed (Table 2). The average age at breast cancer diagnosis was 72.0 years (range 50–99). No association between any psychotropic medication use and invasive breast cancer was observed (HR 0.82; 95% CI, 0.57–1.18); likewise, there was no association of typical or atypical antipsychotics or lithium with invasive breast cancer risk. Psychotropic medication use was positively associated with increased in situ breast cancer risk (HR 1.66; 95% CI, 0.98–2.81), which likely was driven by typical antipsychotic use (HR 2.05; 95% CI, 0.97–4.30); results were similar when restricted to participants with regular mammograms (HR 1.87; 95% CI, 0.84–4.16). Results were similar when restricting to ER⁺ cancers and when modeling psychotropic medication use at year 3. No associations were observed between consistency of psychotropic medication use at baseline and year 3 and invasive or in situ breast cancer risk (data not shown).

Our results do not support an association between psychotropic medication use and subsequent invasive breast cancer risk. We did observe a suggestive two-fold increase in in situ breast cancer risk associated with typical antipsychotic use, which persisted among the subgroup of women with regular mammograms. The consistency of these results suggests that screening differences between users and nonusers may not fully account for the elevated risk. However, our findings were limited by a small number of psychotropic medication users, including only seven typical antipsychotic users later diagnosed with in situ breast cancer, and thus should be interpreted cautiously. We are unaware of a potential biologic mechanism that would result in psychotropic medications increasing only in situ breast cancer risk. Additional limitations include the potential for underreporting of psychotropic medications if women selectively chose not to bring such medications to their clinical visit, as well as the inability to distinguish between diagnostic and screening mammograms for OS participants. Prior studies have either included only invasive cases (2) or have not stratified analyses by invasiveness (3–7), thus additional evaluations, perhaps with pooled data across multiple studies, are needed. Overall, our findings contribute to knowledge of the safety profile of psychotropic medications and may be useful to clinicians and patients considering use of these medications.

A.H. Shadyab is a consultant for Rancho Biosciences LLC. No potential conflicts of interest were disclosed by the other authors.

Conception and design: K.W. Reeves

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): R.B. Wallace

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): A. George, S.R. Sturgeon, A.H. Shadyab, K.W. Reeves

Writing, review, and/or revision of the manuscript: A. George, S.R. Sturgeon, S.E. Hankinson, A.H. Shadyab, R.B. Wallace, K.W. Reeves

The WHI program is funded by the National Heart, Lung, and Blood Institute, NIH, and U.S. Department of Health and Human Services through contracts HHSN268201600018C, HHSN268201600001C, HHSN268201600002C, HHSN268201600003C, and HHSN268201600004C. The authors would like to thank the following: Program Office: (National Heart, Lung, and Blood Institute, Bethesda, MD) Jacques Rossouw, Shari Ludlam, Joan McGowan, Leslie Ford, and Nancy Geller; Clinical Coordinating Center: (Fred Hutchinson Cancer Research Center, Seattle, WA) Garnet Anderson, Ross Prentice, Andrea LaCroix, and Charles Kooperberg; Investigators and Academic Centers: (Brigham and Women's Hospital, Harvard Medical School, Boston, MA) JoAnn E. Manson; (MedStar Health Research Institute/Howard University, Washington, DC) Barbara V. Howard; (Stanford Prevention Research Center, Stanford, CA) Marcia L. Stefanick; (The Ohio State University, Columbus, OH) Rebecca Jackson; (University of Arizona, Tucson/Phoenix, AZ) Cynthia A. Thomson; (University at Buffalo, Buffalo, NY) Jean Wactawski-Wende; (University of Florida, Gainesville/Jacksonville, FL) Marian Limacher; (University of Iowa, Iowa City/Davenport, IA) Jennifer Robinson; (University of Pittsburgh, Pittsburgh, PA) Lewis Kuller; (Wake Forest University School of Medicine, Winston-Salem, NC) Sally Shumaker; (University of Nevada, Reno, NV) Robert Brunner; (University of Minnesota, Minneapolis, MN) Karen L. Margolis.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.