Skip to main content
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2016 Jun 20.
Published in final edited form as: Cancer Causes Control. 2015 Oct 30;27(1):125–136. doi: 10.1007/s10552-015-0689-y

Breast cancer recurrence in relation to antidepressant use

Jessica Chubak 1,2, Erin JA Bowles 1, Onchee Yu 1, Diana SM Buist 1,2,3, Monica Fujii 1, Denise M Boudreau 1,4
PMCID: PMC4913547  NIHMSID: NIHMS791528  PMID: 26518198

Abstract

Purpose

Women with breast cancer frequently use antidepressants; however, questions about the effect of these medications on breast cancer recurrence remain.

Methods

We identified 4216 women ≥18 years with an incident stage I or II breast cancer diagnosed between 1990–2008 in a mixed model healthcare delivery system linked to a cancer registry. Recurrences were ascertained from chart review. Medication exposures were extracted from electronic pharmacy records. We used multivariable Cox proportional hazards models to estimate hazard ratios (HR) and 95% confidence intervals (CI) to assess the association between antidepressant use and breast cancer recurrence and mortality. We also conducted analyses restricted to tamoxifen users.

Results

Antidepressants overall, tricyclic antidepressants, and selective serotonin reuptake inhibitors were not associated with risk of breast cancer recurrence or mortality. Women taking paroxetine only (adjusted HR: 1.66; 95% CI: 1.02, 2.71) and trazadone only (adjusted HR: 1.76; 95% CI: 1.06, 2.92), but not fluoxetine only (adjusted HR: 0.92; 95% CI: 0.55, 1.53), had higher recurrence risks than antidepressant non-users. There was some suggestion of an increased recurrence risk with concurrent paroxetine and tamoxifen use compared to users of tamoxifen only (adjusted HR: 1.49; 95% CI: 0.79, 2.83).

Conclusions

In general, antidepressants did not appear increase risk of breast cancer recurrence; though there were some suggested increases in risk that warrant further investigation in other datasets. Our results combined systematically and quantitatively with results from other studies may be useful for patients and providers making decisions about antidepressant use after breast cancer diagnosis.

Keywords: breast carcinoma, breast carcinoma recurrence, breast carcinoma mortality, antidepressants

INTRODUCTION

Antidepressant use is common among women with breast cancer in the United States, with upwards of one quarter to one third of women using these medications after their cancer diagnosis [13]. Thus, understanding the safety of antidepressants is important for the health of the growing number of breast cancer survivors.

Tamoxifen is an important treatment for breast cancer hormone receptor-positive breast cancer [4]. Inhibitors of the enzyme CYP2D6, such as the selective serotonin reuptake inhibitors (SSRIs) paroxetine and fluoxetine, may impair conversion of tamoxifen into its active form, endoxifen [58]. The clinical importance of their potential inhibitory effects is unknown and questionable, especially since studies suggest that even genetic variants of CYP2D6 encoding reduced enzymatic activity do not result in worse outcomes [913]; however, there is much debate on the effects of these medications in women with breast cancer [1419] and the Food and Drug Administration (FDA) has issued precautions regarding concomitant use of tamoxifen and paroxetine [20,21].

Epidemiologic research on antidepressant use and breast cancer outcomes varies considerably in exposures and outcomes studied [1,2231]. Studies of antidepressants that weakly inhibit CYP2D6 (such as citalopram) have generally not been associated with breast cancer recurrence risk [22,23,31]. However, there has been some suggestion of increased risk of breast cancer recurrence [22] and breast cancer mortality [27] among patients who use tamoxifen with the strong CYP2D6 inhibitor paroxetine in some but not all studies [25]. Studies of CYP2D6 inhibitors in general (including medications other than antidepressants) [26], antidepressants with moderate/strong CYP2D6 affinity [30], and SSRIs in as a class [1,28,31] have generally not suggested an association with breast cancer outcomes among tamoxifen users. Of note, studies lacked statistical precision and little information was available on the commonly used antidepressant fluoxetine, which also inhibits CYP2D6. We therefore assessed the association between different classes of antidepressants and individual medications and the risk of breast cancer recurrence.

METHODS

Study overview

This cohort study, Commonly Used Medications and Breast Cancer Outcomes (COMBO), is described in detail elsewhere [32,33]. Briefly, we conducted this study within the western Washington region of Group Health, a mixed model health plan in Washington state and northern Idaho. Study participants had to reside in one of the 13 western Washington counties covered by the western Washington Surveillance, Epidemiology, and End Results program (SEER) registry. We used cancer registry files linked to Group Health enrollment files to identify women aged 18 years and older who were diagnosed with a first primary stage I or II invasive breast cancer between 1990 and 2008, inclusive. Participants had to be enrolled in Group Health’s Integrated Group Practice model for the year before and after their incident breast cancer diagnosis (unless they died during that year). Medical records of potentially eligible participants (N=4426) were reviewed. We excluded participants with no medical record (N=72), bilateral disease (N=6), recurrent or second primary breast cancers that were incorrectly identified as incident first breast cancers (N=79), and no definitive surgery (N=44). We required women be alive and recurrence-free for 120 days after surgery and therefore excluded people who died (N=5) or had metastases (N=4) within 120 days of surgery. The final cohort consisted of 4216 women. Five-year outcomes from a subset of this cohort (N=1306) were included in an earlier report on antidepressants and breast cancer outcomes [1]. The Group Health human subjects review committee approved all study procedures.

The SEER registry provided data on tumor characteristics and certain patient characteristics. When these were not available in SEER, we abstracted them from medical records as part of a detailed chart abstraction [31,32]. Charts were abstracted from one year before diagnosis through patient death, disenrollment from Group Health, or the date of chart abstraction. Data elements included treatment of the incident breast cancer, breast cancer recurrences and second primaries [34]. Data on comorbidity diagnoses (including mental health diagnoses [Supplemental Table 1] and the Charlson comorbidity score [35]), health history and medication use, height, weight, and date and cause of death, came primarily from administrative data sources and the electronic medical record at Group Health. A list of data elements and their sources was previously published by Boudreau and colleagues [32].

Exposure

Group Health’s pharmacy database provided information on prescription fills at Group Health-owned pharmacies and pharmacy claims throughout the entire study period [36,37]. Each record contained information on the drug dispensed, including national drug code (NDC), quantity, strength, days’ supply, date dispensed, and prescriber. Drug exposures of interest included use of the following after breast cancer diagnosis: 1) any antidepressant, 2) any SSRI, 3) any tricyclic antidepressant (TCA), and 4) any miscellaneous antidepressant. We further classified SSRI use into strong CYP2D6 inhibitors (paroxetine, fluoxetine, and buproprion) and weak/moderate CYP2D6 inhibitors (sertraline, citalopram, fluvoxamine, and escitalopram) [2,30].

For each drug and drug class, we defined episodes of use. We used the days’ supply field to compute the intended duration of each prescription [1,32]. Assuming 80% adherence, we calculated prescriptions’ run-out date by multiplying days’ supply by 1.25 [38]. If a new prescription for the drug or class of interest was dispensed <60 days after the run-out date of the preceding prescription, it was included as part of the same episode of use. Episodes started on the dispensing day of the first prescription and ended on the run-out day of the last prescription [32]. Once someone became a user after breast cancer diagnosis, she remained a user, regardless of when she discontinued use. The rationale for this was that any use – not necessarily current use – could be related to recurrence risk. Concurrent use between antidepressants and endocrine therapy (tamoxifen or aromatase inhibitors) was defined as an overlap of ≥60 days of use [28].

Information on endocrine therapy (tamoxifen and aromatase inhibitors) was obtained from prescription fills. Like all other exposure variables, concurrent antidepressant use was time-varying and unidirectional. Once someone became a concurrent user, she remained a concurrent user, even if she discontinued use of either the antidepressant or tamoxifen. A non-concurrent user could become a concurrent user if the 60 days of overlap was achieved (Supplemental Figure 1).

Outcomes

Recurrence was defined as a ductal carcinoma in situ or invasive cancer of the ipsilateral breast or in any regional or distant sites [3234]. A cancer was classified as a recurrence only if it occurred more than 120 days after definitive surgery [34]. Date and cause of death were obtained from Washington State death files. Second cancers in the contralateral breast were obtained from the SEER program registry and chart abstraction. They were considered censoring events in the analysis of recurrence.

Statistical analysis

Analysis of breast cancer recurrence

We compared users and non-users of any antidepressant after diagnosis with respect to patient characteristics and features of their cancer using the Chi-square test for categorical variables and the t-test for continuous variables. We used multivariable Cox proportional hazards models to estimate hazard ratios (HR) and their 95% confidence intervals (CI) to assess the association between antidepressant use and breast cancer recurrence risk while accounting for competing risks [39]. Time from diagnosis was the analytic time scale [40] and women entered the analysis when they became at risk for recurrence, which was defined as 120 days after definitive surgery. Women were followed until the earliest of recurrence, second primary breast cancer diagnosis, death, disenrollment from Group Health, or end of study (chart abstraction date). Women were censored when they developed a second primary breast cancer and thus, any recurrence occurring after the second primary was not included as a recurrence outcome. Covariates in all models were selected a priori and included age (18–49, 50–59, 60–69, 70–79, ≥80 years) and year of diagnosis (1990–1994, 1995–1999, 2000–2004, 2005–2008), stage (I, IIA, IIB) [41], estrogen receptor (ER) and progesterone receptor (PR) status (ER+/PR+, ER+/PR−, ER−/PR+, ER−/PR−), primary treatment (mastectomy, breast conserving surgery [BCS] with radiation, BCS without radiation), any chemotherapy treatment, any endocrine therapy, body mass index (BMI) (<18.5, 18.5–24.9, 25.0–29.9, 30.0–34.9, ≥35 kg/m2) in the year before diagnosis, smoking in the year before diagnosis (current, past, never/unknown), menopausal status in the year before diagnosis (peri- or pre-menopausal, post-menopausal), and Charlson comorbidity score (0, 1, ≥2). All variables were included as categorical variables. Endocrine treatment and Charlson comorbidity score were included as time-varying covariates. Persons with unknown ER/PR status and BMI were excluded from the main recurrence and mortality models. We conducted an exploratory analysis with depression as a time-varying covariate.

In analyses of drug class and recurrence risk, we used two approaches to control for confounding by other antidepressants. We conducted one analysis in which we restricted (in a time-varying manner) to non-users of any antidepressant and users of only the antidepressant class of interest. In this analysis, exposure was time-varying in that women could switch from being a non-user to a user of one class, but were censored when they used an antidepressant from another class. In the tables, these are referred to as “restriction” analyses. Separate models were conducted for each antidepressant of interest and women were censored when they used another antidepressant. In the second approach, we included each class of antidepressants as a separate exposure variable in the model so that they were mutually adjusted for one another. In the tables, these are referred to as “adjustment” analyses.

We conducted an analysis limited to tamoxifen users (N=1902) to address the question of whether antidepressants modify tamoxifen’s effectiveness. For the analysis of concurrent tamoxifen use, women entered the analysis once they became at risk of recurrence and had used 60 days of tamoxifen (Supplemental Figure 1). Women with aromatase inhibitor use before tamoxifen use were excluded. Women could be non-users of antidepressants (i.e., tamoxifen only users), non-concurrent users of tamoxifen and antidepressants, or concurrent users of tamoxifen and antidepressants. Concurrent use of antidepressant and tamoxifen was time-varying and occurred once a woman had 60 days of overlap of tamoxifen and antidepressant use. We conducted separate models for concurrent use of tamoxifen and any antidepressant as well as with each antidepressant drug and drug class. Due to small sample size, we adjusted only for diagnosis age, diagnosis year, and stage at diagnosis in these models. Women were censored at the earliest of the start of an aromatase inhibitor, second primary breast cancer, death, disenrollment, and end of study follow-up.

Analysis of breast cancer mortality

In a secondary analysis, we examined the risk of breast cancer-specific mortality in relation to antidepressant use, using multivariable Cox models with the same exposures of interest as in the main analysis. Once a person started using an antidepressant, she was classified as a user; however, new use after a recurrence diagnosis or second primary was not included, as recurrences and second primaries were likely to be a strong time-varying confounders [42]. We followed women until the earliest of death from another cause, Group Health disenrollment, or end of study follow-up.

Proportional hazards assumptions

We evaluated proportional hazards assumptions by testing the interaction between the medication classes of interest and the logarithm of follow-up time. Separate tests were run for breast cancer recurrence and mortality outcomes. The assumptions held for all exposure-outcome pairs except for miscellaneous antidepressants and breast cancer mortality and any antidepressant and breast cancer mortality. To further assess the non-proportional hazards for, we divided the follow-up time into two periods each containing half the breast cancer deaths (≤4.24 years since diagnosis and >4.24 years). Separate HRs for exposures were estimated for each time period by including an interaction term between time period and the exposure in the multivariate Cox proportional hazards model. For both any antidepressant and miscellaneous antidepressants models, the HRs in the two time periods were not statistically significant and the confidence intervals overlapped. Therefore, we report the overall results for both miscellaneous antidepressants and any antidepressant.

RESULTS

Among the 4216 women who met study eligibility criteria, more than half (N=2302, 54.6%) filled at least one prescription for an antidepressant sometime after their incident breast cancer diagnosis (but before recurrence or second primary). Antidepressant users were more often white, more likely to smoke at baseline, have a mental health diagnosis and filled prescriptions for antidepressants in the year before diagnosis, have lower education levels, more comorbidities, and higher BMI than participants who did not use antidepressants after breast cancer diagnosis (Table 1). Users and non-users were generally similar with respect to risk factors for recurrence, such as stage and other features of initial cancer. However, compared to non-users, antidepressant users were slightly more likely to receive mastectomy over breast conserving surgery with radiation and to receive chemotherapy.

Table 1.

Characteristics of users and non-users of any antidepressants after a diagnosis of breast cancer and before recurrence (Group Health, 1990–2008), N=4,216

All Any antidepressant use after diagnosis
No Yes
(N=4216)
n (%)a
(N=1914)
n (%)a
(N=2302)
n (%)a
Age at diagnosis (years)
 Mean (SD) 62.2 (13.3) 62.5 (13.2) 62.6 (13.4)
 18–39 139 (3.3) 65 (3.4) 74 (3.2)
 40–49 646 (15.3) 285 (14.9) 361 (15.7)
 50–59 995 (23.6) 438 (22.9) 557 (24.2)
 60–69 1018 (24.1) 491 (25.7) 527 (22.9)
 70–79 940 (22.3) 436 (22.8) 504 (21.9)
 ≥80 478 (11.3) 199 (10.4) 279 (12.1)
Year of diagnosis
 1990–1994 950 (22.5) 442 (23.1) 508 (22.1)
 1995–1999 1191 (28.2) 520 (27.2) 671 (29.1)
 2000–2004 1201 (28.5) 513 (26.8) 688 (29.9)
 2005–2008 874 (20.7) 439 (22.9) 435 (18.9)
Follow-up time
 Mean (SD) 7.1 (4.3) 6.6 (4.3) 7.4 (4.3)
 Median 6.3 5.7 6.7
Race
 White 3719 (88.5) 1640 (86.1) 2079 (90.5)
 African American 136 (3.2) 73 (3.8) 63 (2.7)
 American Indian/Alaska Native 113 (2.7) 44 (2.3) 69 (2.7)
 Asian/Pacific Islander 233 (5.5) 148 (7.8) 85 (3.7)
 Other 1 (0) 0 (0) 1 (0)
 Unknown 14 9 5
Ethnicity
 Not Hispanic 3976 (94.6) 1798 (94.2) 2178 (94.8)
 Hispanic 229 (5.4) 110 (5.8) 119 (5.2)
 Unknown 11 6 5
Menopausal status at diagnosis
 Peri- or Premenopausal 1145 (27.2) 522 (27.3) 623 (27.1)
 Postmenopausal 3071 (72.8) 1392 (72.7) 1679 (72.9)
Education
 High school or less 418 (23.4) 188 (22.1) 230 (24.5)
 Some college 634 (35.4) 281 (33.0) 353 (37.6)
 College or post graduate 737 (41.2) 382 (44.9) 355 (37.8)
 Unknown 2427 1063 1364
Cancer characteristics
AJCC stage at diagnosis
 I 2648 (62.8) 1175 (61.4) 1473 (64.0)
 IIA 1078 (25.6) 518 (27.1) 560 (24.3)
 IIB 490 (11.6) 221 (11.5) 269 (11.7)
Lymph node status
 Unknown 451 195 256
 Negative 2847 (75.6) 1304 (75.9) 1543 (75.4)
 Positive (1–3 nodes) 680 (18.1) 309 (18) 371 (18.1)
 Positive (≥4 nodes) 234 (6.2) 104 (6.1) 130 (6.4)
 Positive (unknown number of nodes) 4 (0.1) 2 (0.1) 2 (0.1)
Tumor size
 <2.0 cm 3110 (73.8) 1392 (72.8) 1718 (74.7)
 ≥2.0 cm 1104 (26.2) 521 (27.2) 583 (25.3)
 Unknown 2 1 1
Tumor histology
 Ductal 3315 (78.6) 1506 (78.7) 1809 (78.6)
 Lobular 336 (8) 147 (7.7) 189 (8.2)
 Mixed/other 565 (13.4) 261 (13.7) 304 (13.3)
Tumor grade
 Well differentiated 1041 (26.9) 471 (26.7) 570 (27.1)
 Moderately differentiated 1600 (41.3) 714 (40.4) 886 (42.1)
 Poorly/undifferentiated 1233 (31.9) 582 (33.0) 651 (30.9)
 Not determined or stated 342 147 195
ER/PR status
 ER & PR unknown 217 93 124
 ER−/PR− 667 (16.7) 313 (17.2) 354 (16.3)
 ER+/PR− 383 (9.6) 177 (9.7) 206 (9.5)
 ER−/PR+ 61 (1.5) 24 (1.3) 37 (1.7)
 ER+/PR+ 2888 (72.2) 1307 (71.8) 1581 (72.6)
Her-2 test (diagnosis in 1998+ only)
 Test done 2074 (79.7) 950 (81.0) 1124 (78.7)
Her-2 test result among women with test done
 Positive/Borderline 353 (17.0) 152 (16.0) 201 (17.9)
 Negative 1714 (82.6) 795 (83.7) 919 (81.8)
 No result 7 (0.3) 3 (0.3) 4 (0.4)
Cancer treatment
Primary therapy
 Mastectomy with or without radiation 1521 (36.1) 635 (33.2) 886 (38.5)
 Breast conserving surgery with radiation 2172 (51.5) 1039 (54.3) 1133 (49.2)
 Breast conserving surgery without radiation 523 (12.4) 240 (12.5) 283 (12.3)
Adjuvant treatment (not mutually exclusive)
 Chemotherapy 1376 (32.6) 586 (30.6) 790 (34.3)
 Tamoxifen 2057 (48.8) 884 (46.2) 1173 (51.0)
 Aromatase Inhibitor 849 (20.1) 353 (18.4) 496 (21.6)
Health history in year prior to diagnosis
Charlson Index
 0 3229 (76.6) 1541 (80.5) 1688 (73.3)
 1 704 (16.7) 263 (13.7) 441 (19.2)
 ≥2 283 (6.7) 110 (5.7) 173 (7.5)
Body mass index (kg/m2)
 <18.5 69 (1.6) 29 (1.5) 40 (1.7)
 18.5 to <25 kg/m2 1453 (34.6) 698 (36.6) 755 (33.0)
 25 to <30 kg/m2 1362 (32.5) 636 (33.4) 726 (31.7)
 30 to <35 kg/m2 766 (18.3) 330 (17.3) 436 (19.0)
 ≥35 kg/m2 546 (13.0) 212 (11.1) 334 (14.6)
 Unknown 20 9 11
Smoking status
 Current 253 (6.0) 83 (4.3) 170 (7.4)
 Past 352 (8.3) 169 (8.8) 183 (7.9)
 Never/Unknown 3611 (85.6) 1662 (86.8) 1949 (84.7)
Diagnosis of depression 402 (9.5) 39 (2.0) 363 (15.8)
Diagnosis of anxiety 174 (4.1) 30 (1.6) 144 (6.3)
Mental health diagnosis 795 (21.7) 166 (10.1) 629 (31.1)
Antidepressant use 935 (22.2) 53 (2.8) 882 (38.3)
 SSRIs 452 (10.7) 17 (0.9) 435 (18.9)
 TCAs 422 (10.0) 28 (1.5) 394 (17.1)
 Miscellaneous 280 (6.6) 12 (0.6) 268 (11.6)

AJCC = American Joint Committee on Cancer; ER = estrogen receptor; PR = progesterone receptor; SD = standard deviation; SSRI = selective serotonin reuptake inhibitor; TCA = tricyclic antidepressant

a

Column percents may not add to 100% due to rounding

The median duration of use after breast cancer diagnosis for all antidepressants was 23 months (Table 2) and median time to first use was 7.4 months (not shown). SSRIs were the most commonly used class of antidepressants. Of all individual antidepressants, fluoxetine and trazodone were the two most frequently used agents. Overall, use of antidepressants after incident breast cancer diagnosis was not associated with risk of recurrence (adjusted HR: 1.16, 95% CI: 0.94, 1.44) compared to non-users (Table 3). HRs for both SSRIs and TCAs were near 1.0 and were not statistically significant. Miscellaneous antidepressants were associated with a non-significant increased risk, possibly driven by the most commonly used of these medications, trazodone (adjusted HR: 1.76; 95% CI: 1.06, 2.92). We observed an increased recurrence risk among women taking paroxetine only (adjusted HR: 1.66; 95% CI: 1.02, 2.71), but not fluoxetine only (adjusted HR: 0.92; 95% CI: 0.55, 1.53). Risk estimates were elevated for citalopram users, but were based on only three events in the user group.

Table 2.

Antidepressant prescriptions filled after breast cancer diagnosis and before recurrence in 4216 women diagnosed with early stage breast cancer (Group Health, 1990–2008)

Among users

Number of users (%)a Median duration of use after diagnosis (months)
Any antidepressant 2302 (54.6) 23.1
Selective serotonin reuptake inhibitors 1498 (35.5) 18.7
 Strong CYP2D6 inhibitors 1245 (29.5) 13.8
  Paroxetine 551 (13.1) 8.5
  Fluoxetine 904 (21.4) 11.4
 Weak/Moderate CYP2D6 inhibitors 579 (13.7) 12.0
  Sertraline 301 (7.1) 10.0
  Citalopram 337 (8.0) 9.3
  Escitalopram 17 (0.4) 2.3
  Fluvoxamine 3 (0.1) 35.0
Tricyclic antidepressants 1128 (26.8) 6.7
  Amitriptyline 408 (9.7) 5.9
  Nortriptyline 537 (12.7) 3.5
  Desipramine 72 (1.7) 5.4
  Doxepin 206 (4.9) 3.8
  Imipramine 135 (3.2) 6.4
  Protriptyline 1 (0.02) 1.3
Miscellaneous antidepressants 1099 (26.1) 6.8
  Buproprion 261 (6.2) 6.5
  Trazodone 785 (18.6) 4.4
  Duloxetine 21 (0.5) 4.2
  Mirtazapine 96 (2.3) 6.2
  Nefazodone 11 (0.3) 6.3
  Tranylcypromine 2 (0.1) 23.7
  Venlafaxine 127 (3.0) 11.7
a

Categories of antidepressants are not mutually exclusive

Table 3.

Risk of breast cancer recurrence in relation to the use of antidepressants after a diagnosis of early stage breast cancer (Group Health, 1990–2008), N=4,216

Person-years Number of recurrences Recurrence rate per 1,000 person-years (95% CI) Unadjusted hazard ratio (95% CI) Adjusted a hazard ratio (95% CI)
Overall 28063 415 14.8 (13.4, 16.3)
No antidepressant use 15662 228 14.6 (12.8, 16.6) Reference Reference
Any antidepressant use 12401 187 15.1 (13.1, 17.4) 1.13 (0.93, 1.38) 1.16 (0.94, 1.44)
SSRI use
 Only – restriction 2917 45 15.4 (11.5, 20.7) 1.05 (0.76, 1.45) 1.11 (0.79, 1.55)
 Any – adjustmentb 7315 110 15.0 (12.5, 18.1) 1.11 (0.89, 1.38) 1.14 (0.88, 1.48)
 Paroxetine only 790 18 22.8 (14.4, 36.2) 1.55 (0.96, 2.50) 1.66 (1.02, 2.71)
 Fluoxetine only 1235 18 14.6 (9.2, 23.1) 0.96 (0.60, 1.56) 0.92 (0.55, 1.53)
 Sertraline only 255 3 11.8 (3.8, 36.5) 0.80 (0.26, 2.51) 0.88 (0.28, 2.76)
 Citalopram only 183 3 16.4 (5.3, 51.0) 1.09 (0.35, 3.40) 1.60 (0.52, 4.92)
TCA use
 Only – restriction 3096 45 14.5 (10.9, 19.5) 1.04 (0.76, 1.44) 0.96 (0.67, 1.38)
 Any – adjustmentb 6527 93 14.2 (11.6, 17.5) 1.09 (0.87, 1.38) 1.04 (0.80, 1.35)
 Amitriptyline only 1057 14 13.2 (7.8, 22.4) 0.94 (0.55, 1.61) 0.89 (0.50, 1.60)
 Nortriptyline only 834 13 15.6 (9.0, 26.8) 1.11 (0.63, 1.94) 0.99 (0.53, 1.85)
Miscellaneous antidepressant use
 Only – restriction 1279 23 18.0 (12.0, 27.1) 1.27 (0.83, 1.95) 1.41 (0.89, 2.23)
 Any – adjustmentb 5125 75 14.6 (11.7, 18.4) 1.12 (0.87, 1.44) 1.12 (0.83, 1.51)
 Bupropion only 180 0 0
 Trazodone only 899 19 21.1 (13.5, 33.1) 1.55 (0.97, 2.48) 1.76 (1.06, 2.92)

SSRI = selective serotonin reuptake inhibitor; TCA = tricyclic antidepressant; CI = confidence interval

a

All estimates are from separate models accounting for competing risks. Adjusted for age at diagnosis (18–49, 50–59, 60–69, 70–79, 80+ years), year of diagnosis (1990–1994, 1995–1999, 2000–2004, 2005–2008), American Joint Commission on Cancer stage (I, IIA, IIB), estrogen receptor (ER) and progesterone receptor (PR) status (ER+/PR+, ER+/PR−, ER−/PR+, ER−/PR−), primary treatment (mastectomy, breast conserving surgery with radiation, breast conserving surgery without radiation), chemotherapy treatment (yes, no), endocrine therapy (yes, no), body mass index (<18.5, 18.5–24.9, 25.0–29.9, 30.0–34.9, 35+ kg/m2), smoking at diagnosis (current, past, never/unknown), menopausal status at diagnosis (peri- or pre-menopausal, post-menopausal), and Charlson score (0, 1, 2+). Total number of participants in adjusted model was 3979 due to exclusion of women with unknown ER/PR status or BMI.

b

Unexposed groups for these analyses, i.e., non-users of specific antidepressants, are not shown in this table.

In the analysis restricted to tamoxifen users, concurrent use of most antidepressants was not associated with recurrence risk (Table 4). There was some suggestion of a decreased risk associated with concurrent weak CYP2D6-inhibitor use (adjusted HR: 0.51; 95% CI: 0.16, 1.62) and increased risk with concurrent paroxetine use (adjusted HR: 1.49; 95% CI: 0.79, 2.83). However, these estimates were based on only 6 and 16 events in the concurrent user groups, respectively. There were even fewer events among non-concurrent antidepressant users. For fluoxetine, recurrence risk was elevated with non-concurrent (adjusted HR: 2.21, CI: 1.04, 4.66) tamoxifen use, but not concurrent use.

Table 4.

Risk of breast cancer recurrence in relation to antidepressant use after a diagnosis of hormone receptor positive breast cancer among 1902 tamoxifen users (Group Health, 1990–2008)a

Person-years Number of recurrences Recurrence rate per 1,000 person-years (95% CI) Adjusted hazard ratio (95% CI)c
Any antidepressant
 Non-user (tamoxifen only) 7406 125 16.9 (14.1, 20.1) Reference
 Non-concurrent w/tamoxifen 799 12 15.0 (7.8, 26.2) 1.48 (0.75, 2.91)
 Concurrent w/tamoxifenb 3464 54 15.6 (11.7, 20.3) 0.90 (0.64, 1.37)

SSRIs
 Non-user (tamoxifen only) 9240 150 16.2 (13.7, 19.1) Reference
 Non-concurrent w/tamoxifen 651 12 18.4 (9.5, 32.2) 1.82 (0.92, 3.61)
 Concurrent w/tamoxifenb 1778 29 16.3 (10.9, 23.4) 1.10 (0.68, 1.79)
SSRI, Strong CYP2D6 inhibitord
 Non-user 9654 155 16.1 (13.6, 18.8) Reference
 Non-concurrent w/tamoxifen 585 11 18.8 (9.4, 33.7) 1.84 (0.91, 3.73)
 Concurrent w/tamoxifenb 1430 25 17.5 (11.3, 25.8) 1.15 (0.69–1.92)
SSRI, Weak/Moderate CYP2D6 inhibitorf
 Non-user (tamoxifen only) 10881 184 16.9 (14.6, 19.5) Reference
 Non-concurrent w/tamoxifen 241 1 4.2 (0.1, 23.1) 0.43 (0.06, 3.15)
 Concurrent w/tamoxifenb 547 6 11.0 (4.0, 23.9) 0.51 (0.16, 1.62)
Paroxetine
 Non-user (tamoxifen only) 10659 172 16.1 (13.8, 18.7) Reference
 Non-concurrent w/tamoxifen 340 3 8.8 (1.8, 25.8) 0.95 (0.30, 3.05)
 Concurrent w/tamoxifenb 670 16 23.9 (13.6, 38.8) 1.49 (0.79, 2.83)
Fluoxetine
 Non-user (tamoxifen only) 10367 170 16.4 (14.0, 19.1) Reference
 Non-concurrent w/tamoxifen 459 11 24.0 (12.0–42.9) 2.21 (1.04, 4.66)
 Concurrent w/tamoxifenb 843 10 11.9 (5.7, 21.8) 0.73 (0.34, 1.59)

TCAs
 Non-user (tamoxifen only) 9451 153 16.2 (13.7, 19.0) Reference
 Non-concurrent w/tamoxifen 546 11 20.2 (10.1, 36.1) 1.26 (0.58, 2.74)
 Concurrent w/tamoxifenb 1672 27 16.2 (10.6, 23.5) 0.79 (0.48, 1.31)

Miscellaneous antidepressant
 Non-user (tamoxifen only) 10229 170 16.6 (14.2, 19.3) Reference
 Non-concurrent w/tamoxifen 412 10 24.3 (11.6, 44.7) 1.99 (0.90, 4.40)
 Concurrent w/tamoxifenb 1028 11 10.7 (5.3, 19.1) 0.74 (0.36, 1.52)

SSRI = selective serotonin reuptake inhibitor; TCA = tricyclic antidepressant; CI = confidence interval; w/ = with

a

Restricted to 1902 women using tamoxifen for ≥60 days and no prior aromatase inhibitor use

b

Concurrent use ≥60 days

c

All estimates are from separate models accounting for competing risks. Censoring events were start of aromatase inhibitors, second primary breast cancer, death, Group Health disenrollment, and end of study follow-up. Adjusted for age at diagnosis (18–49, 50–59, 60–69, 70–79, 80+ years), year of diagnosis (1990–1994, 1995–1999, 2000–2004, 2005–2008), and American Joint Commission on Cancer stage

d

Paroxetine and fluoxetine

f

Sertraline, citalopram, fluvoxamine, and escitalopram

Antidepressant use was not associated with breast cancer mortality (adjusted HR: 95% CI: 1.09; 95% CI: 0.83, 1.43) (Table 5). Confidence intervals for classes of antidepressants and individual drugs were wide. Post-hoc adjustment for depression did not change results meaningfully and are therefore not presented.

Table 5.

Risk of breast-cancer specific death in relation to the use of antidepressants after a diagnosis and before recurrence (Group Health 1990–2008), N=4,216

Person-years Number of deaths due to breast cancer Breast cancer mortality rate per 1,000 person-years (95% CI) Unadjusted hazard ratio (95% CI) Adjusteda hazard ratio (95% CI)
Overall 29900 264 8.8 (7.8, 10.0)
No use 16724 145 8.7 (7.4, 10.2) Reference Reference
Any antidepressant use 13176 119 9.0 (7.5, 10.8) 1.03 (0.81, 1.32) 1.09 (0.83, 1.43)
SSRI use
 Only – restriction 3024 18 6.0 (3.7, 9.4) 0.85 (0.52, 1.40) 0.89 (0.52, 1.50)
 Any – adjustment 7714 70 9.1 (7.2, 11.5) 1.03 (0.78, 1.36) 1.14 (0.83, 1.56)
 Paroxetine only 834 4 4.8 (1.8, 12.8) 0.76 (0.28, 2.08) 0.99 (0.35, 2.83)
 Fluoxetine only 1268 7 5.5 (2.6, 11.6) 0.82 (0.38, 1.77) 0.71 (0.30, 1.67)
 Sertraline only 258 3 11.7 (3.8, 36.1) 1.98 (0.63, 6.26) 2.42 (0.83, 7.02)
 Citalopram only 189 2 10.6 (2.6, 42.3) 1.69 (0.42, 6.89) 2.52 (0.57, 11.20)
TCA use
 Only – restriction 3269 20 6.1 (3.9, 9.5) 0.92 (0.57, 1.49) 0.82 (0.48, 1.42)
 Any - adjustment 7035 65 9.2 (7.2, 11.8) 1.06 (0.80, 1.41) 1.04 (0.77, 1.43)
 Amitriptyline only 1101 5 4.5 (1.9, 10.9) 0.75 (0.30, 1.83) 0.71 (0.28, 1.81)
 Nortriptyline only 889 6 6.7 (3.0, 15.0) 1.05 (0.46, 2.40) 0.81 (0.31, 2.08)
Miscellaneous antidepressant use
 Only – restriction 1312 10 7.6 (4.1, 14.2) 1.22 (0.64, 2.34) 1.55 (0.79, 3.04)
 Any – adjustment 5407 43 8.0 (5.9, 10.7) 0.88 (0.63, 1.22) 0.92 (0.61, 1.38)
 Bupropion only 180 1 5.6 (0.8, 39.5) 0.94 (0.13, 6.71) 1.70 (0.22, 13.10)
 Trazodone only 929 6 6.5 (2.9, 14.4) 1.07 (0.47, 2.43) 1.25 (0.51, 3.00)

SSRI = selective serotonin reuptake inhibitor; TCA = tricyclic antidepressant; CI = confidence interval

a

All estimates are from separate models. Adjusted for age at diagnosis (18–49, 50–59, 60–69, 70–79, 80+ years), year of diagnosis (1990–1994, 1995–1999, 2000–2004, 2005–2008), American Joint Commission on Cancer stage (I, IIA, IIB), estrogen receptor (ER) and progesterone receptor (PR) status (ER+/PR+, ER+/PR−, ER−/PR+, ER−/PR−), primary treatment (mastectomy, breast conserving surgery with radiation, breast conserving surgery without radiation), chemotherapy treatment (yes, no), endocrine therapy (yes, no), body mass index (<18.5, 18.5–24.9, 25.0–29.9, 30.0–34.9, 35+ kg/m2), smoking at diagnosis (current, past, never/unknown), menopausal status at diagnosis (peri- or pre-menopausal, post-menopausal), and Charlson score (0, 1, 2+). Total number of participants in adjusted model was 3924 due to exclusion of women with unknown ER/PR status or BMI.

b

Unexposed groups for these analyses, i.e., non-users of specific antidepressants, are not shown in this table.

DISCUSSION

Our study provides some reassurance that many commonly used antidepressant medications are not associated with an increased risk of breast cancer recurrence. Results from this large, population-based retrospective cohort study are generally consistent with other studies that have generally not observed an increased risk of breast cancer recurrence with antidepressant use [6].

We hypothesized that paroxetine and fluoxetine would increase recurrence risk among tamoxifen users because these antidepressants inhibit the conversion of tamoxifen to its active form [57]. There was a suggestion of increased risk with concurrent paroxetine use among tamoxifen users. This finding is consistent with point estimates in one study (odds ratio for recurrence: 2.4, 95% CI: 0.6–9.5) [22], but not another (HR for disease free survival: 0.84; 95% CI: 0.34–2.05) [25]. Of note, all three studies (including ours) had wide confidence intervals.

Among tamoxifen users, we also observed that fluoxetine was associated with an increased risk of recurrence when taken separately from tamoxifen (i.e., non-concurrently) but not at the same time (i.e., concurrently). This is surprising and inconsistent with the biological hypothesis, which would suggest that fluoxetine, a strong CYP2D6 inhibitor, would affect breast cancer recurrence risk by reducing tamoxifen’s effectiveness if the two medications were used concurrently. In general, our findings for non-concurrent antidepressant use among tamoxifen users should be cautiously interpreted due to small sample size.

We saw an increased risk of recurrence with trazodone use. Trazodone may be a weak/moderate CYP2D6 inhibitor [30,43], but to our knowledge, other studies have not looked specifically at the risk of recurrence in relation to trazodone use. Given trazadone is commonly used for insomnia [44], it is unclear whether our findings are due to chance or confounding by indication. The suggested increased risk of recurrence with citalopram should also be interpreted cautiously as it was based on only three events and is not consistent with findings from larger studies [22,23,31]. Similarly, the small number of deaths due to breast cancer in each drug exposure category limits our ability to draw conclusions.

Our study had relatively long follow-up and high-quality longitudinal, prospectively collected data on exposures, outcomes, and confounders with gold-standard chart-review for many of these variables. Variables were not subject to recall bias. A limitation was that we had very small number of events when examining individual drugs, particularly for analyses of non-concurrent use among patients taking tamoxifen. This has been a common issue in other studies of specific antidepressant use and breast cancer recurrence [22,23,25,26,31] and motivates the need for pooled meta-analyses on specific drugs. We anticipated this problem, but believed that these analyses were still important to conduct. Women will benefit from having information about the risk profiles of individual antidepressants to make decisions about which may be safer to use after breast cancer diagnosis. Last, COMBO uses data from a single health plan and includes an educated, primarily white population, and individuals with access to both medical care and prescription drug benefits. This may limit generalizability to some populations if the biological effect of antidepressants differs by patient characteristics.

Having access to the most important predictors of cancer recurrence (i.e., disease characteristics) helped reduce confounding, but residual confounding may have been present. For example, there may be lifestyle risk factors for recurrence that are also associated with antidepressant use. Confounding by indication is also a possibility in this study; even though all medications studied are used for depression, some have other indications (e.g., anxiety, pain, sleep). We did not have information on the reasons for antidepressant use and therefore could not adjust for it. It is also possible that antidepressants were used for symptoms of an undiagnosed recurrence. Under such circumstances, they could appear to increase the risk of recurrence due to reverse causality (i.e., protopathic bias). To protect against confounding by indication in the breast cancer mortality analysis (i.e., a woman started using antidepressants because she had a recurrence), we counted only exposure initiated before recurrence [42].

Conclusions

Our results provide important information on antidepressant use after breast cancer diagnosis including information on individual drugs, which – when combined systematically and quantitatively with results from other studies – may be useful for patients and providers making decisions about whether to use these medications.

Supplementary Material

Suppl Table

Acknowledgments

The authors would like to thank Dr. Heidi Wirtz for helpful discussions and input on preliminary analyses.

FUNDING

This study was funded by the National Cancer Institute at the National Institutes of Health, Award Number 1R01CA1205621 to Denise M. Boudreau. Part of the data collection was supported by a grant from the National Cancer Institute: U01CA63731, PI: Buist. The collection of cancer incidence data used in this study was supported by the Cancer Surveillance System of the Fred Hutchinson Cancer Research Center, which is funded by Contract No. N01-CN-67009 and N01-PC-35142 from the Surveillance, Epidemiology and End Results (SEER) Program of the National Cancer Institute with additional support from the Fred Hutchinson Cancer Research Center and the State of Washington. Funders did not participate in the design, data collection, manuscript drafting, or decision to submit the manuscript for publication.

Footnotes

CONFLICTS OF INTERST

Ms. Yu has received funding from unrelated research grants awarded to Group Health Research Institute from Amgen and Bayer. Dr. Boudreau has received funding from unrelated research grants awarded to Group Health Research Institute from Amgen, Sanofi Aventis, and Johnson & Johnson.

ETHICAL APPROVAL

For this type of study formal consent is not required.

References

  • 1.Chubak J, Buist DS, Boudreau DM, Rossing MA, Lumley T, Weiss NS. Breast cancer recurrence risk in relation to antidepressant use after diagnosis. Breast Cancer Res Treat. 2008;112:123–32. doi: 10.1007/s10549-007-9828-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Dusetzina SB, Alexander GC, Freedman RA, Huskamp HA, Keating NL. Trends in co-prescribing of antidepressants and tamoxifen among women with breast cancer, 2004–2010. Breast Cancer Res Treat. 2012 doi: 10.1007/s10549-012-2330-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Fisch MJ, Zhao F, Manola J, Miller AH, Pirl WF, Wagner LI. Patterns and predictors of antidepressant use in ambulatory cancer patients with common solid tumors. Psycho-Oncology. 2015;24:523–32. doi: 10.1002/pon.3606. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Burstein HJ, Temin S, Anderson H, et al. Adjuvant endocrine therapy for women with hormone receptor-positive breast cancer: american society of clinical oncology clinical practice guideline focused update. J Clin Oncol. 2014;32:2255–69. doi: 10.1200/JCO.2013.54.2258. doi: 10.1200/JCO.2013.54.2258 JCO.2013.54.2258 [pii] [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Stearns V, Johnson MD, Rae JM, Morocho A, Novielli A, Bhargava P, Hayes DF, Desta Z, Flockhart DA. Active tamoxifen metabolite plasma concentrations after coadministration of tamoxifen and the selective serotonin reuptake inhibitor paroxetine. J Natl Cancer Inst. 2003;95:1758–64. doi: 10.1093/jnci/djg108. [DOI] [PubMed] [Google Scholar]
  • 6.Cronin-Fenton DP, Lash TL. Clinical epidemiology and pharmacology of CYP2D6 inhibition related to breast cancer outcomes. Expert Rev Clin Pharmacol. 2011;4:363–77. doi: 10.1586/ecp.11.18. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Jin Y, Desta Z, Stearns V, et al. CYP2D6 genotype, antidepressant use, and tamoxifen metabolism during adjuvant breast cancer treatment. J Natl Cancer Inst. 2005;97:30–9. doi: 10.1093/jnci/dji005. [DOI] [PubMed] [Google Scholar]
  • 8.Desmarais JE, Looper KJ. Interactions between tamoxifen and antidepressants via cytochrome P450 2D6. J Clin Psychiatry. 2009;70:1688–97. doi: 10.4088/JCP.08r04856blu. [DOI] [PubMed] [Google Scholar]
  • 9.Rae JM, Drury S, Hayes DF, et al. CYP2D6 and UGT2B7 genotype and risk of recurrence in tamoxifen-treated breast cancer patients. J Nat Cancer Inst. 2012;104:452–60. doi: 10.1093/jnci/djs126. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Regan MM, Leyland-Jones B, Bouzyk M, et al. CYP2D6 genotype and tamoxifen response in postmenopausal women with endocrine-responsive breast cancer: The Breast International Group 1–98 Trial. J Nat Cancer Inst. 2012;104:441–51. doi: 10.1093/jnci/djs125. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Lum DW, Perel P, Hingorani AD, Holmes MV. CYP2D6 genotype and tamoxifen response for breast cancer: a systematic review and meta-analysis. PLoS One. 2013;8:e76648. doi: 10.1371/journal.pone.0076648. doi: 10.1371/journal.pone.0076648 PONE-D-13-30278 [pii] [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Province MA, Goetz MP, Brauch H, et al. CYP2D6 genotype and adjuvant tamoxifen: meta-analysis of heterogeneous study populations. Clin Pharmacol Ther [Articles] 2014;95:216–27. doi: 10.1038/clpt.2013.186. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Mwinyi J, Vokinger K, Jetter A, Breitenstein U, Hiller C, Kullak-Ublick GA, Trojan A. Impact of variable CYP genotypes on breast cancer relapse in patients undergoing adjuvant tamoxifen therapy. Cancer Chemother Pharmacol. 2014;73:1181–8. doi: 10.1007/s00280-014-2453-5. [DOI] [PubMed] [Google Scholar]
  • 14.Brauch H, Schroth W, Goetz MP, Mürdter TE, Winter S, Ingle JN, Schwab M, Eichelbaum M. Tamoxifen use in postmenopausal breast cancer: CYP2D6 matters. J Clin Oncol. 2013;31:176–80. doi: 10.1200/jco.2012.44.6625. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Kelly CM, Pritchard KI. CYP2D6 genotype as a marker for benefit of adjuvant tamoxifen in postmenopausal women: lessons learned. J Nat Cancer Inst. 2012;104:427–8. doi: 10.1093/jnci/djs139. [DOI] [PubMed] [Google Scholar]
  • 16.Kelly CM, Pritchard KI. Response. J Nat Cancer Inst. 2012;104:1268. doi: 10.1093/jnci/djs303. [DOI] [PubMed] [Google Scholar]
  • 17.Berry D. CYP2D6 genotyping and the use of tamoxifen in breast cancer. J Nat Cancer Inst. 2013;105:1267–9. doi: 10.1093/jnci/djt221. [DOI] [PubMed] [Google Scholar]
  • 18.Pharoah PD, Abraham J, Caldas C. Re: CYP2D6 genotype and tamoxifen response in postmenopausal women with endocrine-responsive breast cancer: The Breast International Group 1–98 Trial and Re: CYP2D6 and UGT2B7 genotype and risk of recurrence in tamoxifen-treated breast cancer patients. J Natl Cancer Inst. 2012 doi: 10.1093/jnci/djs312. doi: djs312 [pii] 10.1093/jnci/djs312. [DOI] [PubMed] [Google Scholar]
  • 19.Berry DA. CYP2D6 genotype and adjuvant tamoxifen. Clin Pharmacol Ther [Point/Counterpoint] 2014;96:138–40. doi: 10.1038/clpt.2014.96. [DOI] [PubMed] [Google Scholar]
  • 20.U.S. Food and Drug Administration. Safety: Paxil CR (paroxetine hydrochloride) Controlled-Release Tablets October 2008. Silver Spring, MD: U.S. Food and Drug Administration; 2008. [updated September 24, 2013; cited 2014 December 28]; Available from: http://www.fda.gov/Safety/MedWatch/SafetyInformation/Safety-RelatedDrugLabelingChanges/ucm122147.htm. [Google Scholar]
  • 21.U.S. Food and Drug Administration. Safety: Paxil (paroxetine HCl) tablets, oral solution and Paxil CR tablets. Silver Spring, MD: U.S. Food and Drug Administration; 2011. [updated August 11, 2011; cited 2014 December 28]; Available from: http://www.fda.gov/Safety/MedWatch/SafetyInformation/ucm233693.htm. [Google Scholar]
  • 22.Ahern TP, Pedersen L, Cronin-Fenton DP, Sørensen HT, Lash TL. No increase in breast cancer recurrence with concurrent use of tamoxifen and some CYP2D6-inhibiting medications. Cancer Epidemiol Biomarkers Prev. 2009;18:2562–4. doi: 10.1158/1055-9965.epi-09-0516. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Lash TL, Pedersen L, Cronin-Fenton D, et al. Tamoxifen’s protection against breast cancer recurrence is not reduced by concurrent use of the SSRI citalopram. Br J Cancer. 2008;99:616–21. doi: 10.1038/sj.bjc.6604533. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Lash TL, Rosenberg CL. Evidence and practice regarding the role for CYP2D6 inhibition in decisions about tamoxifen therapy. J Clin Oncol. 2010;28:1273–5. doi: 10.1200/JCO.2009.26.7906. doi: JCO.2009.26.7906 [pii] 10.1200/JCO.2009.26.7906. [DOI] [PubMed] [Google Scholar]
  • 25.Siegelmann-Danieli N, Kurnik D, Lomnicky Y, Vesterman-Landes J, Katzir I, Bialik M, Loebstein R. Potent CYP2D6 Inhibiting drugs do not increase relapse rate in early breast cancer patients treated with adjuvant tamoxifen. Breast Cancer Res Treat. 2011;125:505–10. doi: 10.1007/s10549-010-1008-7. [DOI] [PubMed] [Google Scholar]
  • 26.Lehmann D, Nelsen J, Ramanath V, Newman N, Duggan D, Smith A. Lack of attenuation in the antitumor effect of tamoxifen by chronic CYP isoform inhibition. J Clin Pharmacol. 2004;44:861–5. doi: 10.1177/0091270004266618. [DOI] [PubMed] [Google Scholar]
  • 27.Kelly CM, Juurlink DN, Gomes T, Duong-Hua M, Pritchard KI, Austin PC, Paszat LF. Selective serotonin reuptake inhibitors and breast cancer mortality in women receiving tamoxifen: a population based cohort study. BMJ. 2010;340:c693. doi: 10.1136/bmj.c693. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Dezentje VO, van Blijderveen NJC, Gelderblom H, Putter H, van Herk-Sukel MPP, Casparie MK, Egberts ACG, Nortier JWR, Guchelaar H-J. Effect of concomitant CYP2D6 inhibitor use and tamoxifen adherence on breast cancer recurrence in early-stage breast cancer. J Clin Oncol. 2010;28:2423–9. doi: 10.1200/jco.2009.25.0894. [DOI] [PubMed] [Google Scholar]
  • 29.Wernli KJ, Hampton JM, Trentham-Dietz A, Newcomb PA. Use of antidepressants and NSAIDs in relation to mortality in long-term breast cancer survivors. Pharmacoepidemiol Drug Saf. 2011;20:131–7. doi: 10.1002/pds.2064. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Azoulay L, Dell’Aniello S, Huiart L, du Fort G, Suissa S. Concurrent use of tamoxifen with CYP2D6 inhibitors and the risk of breast cancer recurrence. Breast Cancer Res Treat. 2011;126:695–703. doi: 10.1007/s10549-010-1162-y. [DOI] [PubMed] [Google Scholar]
  • 31.Lash TL, Cronin-Fenton D, Ahern TP, et al. Breast cancer recurrence risk related to concurrent use of SSRI antidepressants and tamoxifen. Acta Oncol. 2010 doi: 10.3109/02841860903575273. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Boudreau DM, Yu O, Chubak J, Wirtz HS, Bowles EJ, Fujii M, Buist DS. Comparative safety of cardiovascular medication use and breast cancer outcomes among women with early stage breast cancer. Breast Cancer Res Treat. 2014;144:405–16. doi: 10.1007/s10549-014-2870-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Wirtz HS, Buist DS, Gralow JR, et al. Frequent antibiotic use and second breast cancer events. Cancer Epidemiol Biomarkers Prev. 2013;22:1588–99. doi: 10.1158/1055-9965.EPI-13-0454. doi: 10.1158/1055-9965.EPI-13-0454 1055-9965.EPI-13-0454 [pii] [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Geiger AM, Thwin SS, Lash TL, et al. Recurrences and second primary breast cancers in older women with initial early-stage disease. Cancer. 2007;109:966–74. doi: 10.1002/cncr.22472. doi: [DOI] [PubMed] [Google Scholar]
  • 35.Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol. 1992;45:613–9. doi: 10.1016/0895-4356(92)90133-8. doi: http://dx.doi.org/10.1016/0895-4356(92)90133-8. [DOI] [PubMed] [Google Scholar]
  • 36.Boudreau DM, Doescher MP, Saver BG, Jackson JE, Fishman PA. Reliability of Group Health Cooperative automated pharmacy data by drug benefit status. Pharmacoepidemiol Drug Saf. 2005;14:877–84. doi: 10.1002/pds.1119. [DOI] [PubMed] [Google Scholar]
  • 37.Saunders KW, Davis RL, Stergachis A. Group Health Cooperative. In: Strom BL, editor. Pharmacoepidemiology. 4th. Chichester: John Wiley & Sons Ltd; 2005. pp. 223–39. [Google Scholar]
  • 38.Smith NL, Heckbert SR, Lemaitre RN, Reiner AP, Lumley T, Weiss NS, Larson EB, Rosendaal FR, Psaty BM. Esterified estrogens and conjugated equine estrogens and the risk of venous thrombosis. JAMA. 2004;292:1581–7. doi: 10.1001/jama.292.13.1581. [DOI] [PubMed] [Google Scholar]
  • 39.Varadhan R, Weiss CO, Segal JB, Wu AW, Scharfstein D, Boyd C. Evaluating health outcomes in the presence of competing risks: a review of statistical methods and clinical applications. Med Care. 2010;48(6 Supplement):S96–S105. doi: 10.1097/MLR.0b013e3181d99107. doi: [DOI] [PubMed] [Google Scholar]
  • 40.Chubak J, Yu O, Buist DS, Wirtz HS, Boudreau DM. Time Scale in Follow-up Studies: Considering Disease Prognosis. Epidemiology. 2013;24:628–9. doi: 10.1097/EDE.0b013e3182961708. doi: 10.1097/EDE.0b013e3182961708 [doi] 00001648-201307000-00030 [pii] [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Greene FL, Page DL, Fleming ID, Fritz AG, Balch C, Haller DG, Morrow M, editors. AJCC cancer staging manual. 6th. New York: Springer; 2002. [Google Scholar]
  • 42.Chubak J, Boudreau DM, Wirtz HS, McKnight B, Weiss NS. Threats to validity of nonrandomized studies of postdiagnosis exposures on cancer recurrence and survival. J Natl Cancer Inst. 2013;105:1456–62. doi: 10.1093/jnci/djt211. doi: djt211 [pii] 10.1093/jnci/djt211. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Wu AHB, Lorizio W, Tchu S, et al. Estimation of tamoxifen metabolite concentrations in the blood of breast cancer patients through CYP2D6 genotype activity score. Breast Cancer Res Treat. 2012;133:677–83. doi: 10.1007/s10549-012-1963-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Bertisch SM, Herzig SJ, Winkelman JW, Buettner C. National use of prescription medications for insomnia: NHANES 1999–2010. Sleep. 2014;37:343–9. doi: 10.5665/sleep.3410. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Suppl Table

RESOURCES