Abstract
Background: Concerns have been raised about the risk of dementia associated with antiestrogen adjuvant therapy in breast cancer, but study results have been inconsistent. We examined whether tamoxifen or other endocrine therapy was associated with dementia risk in a large population of patients with breast cancer.
Methods: We used Danish nationwide medical registries to identify patients with breast cancer diagnosed between 1990 and 2004, use of endocrine therapy, and subsequent diagnoses of dementia. We used Cox regression to estimate the risk of dementia among patients who received five years of tamoxifen or other endocrine therapies.
Results: The study included 16,419 patients with breast cancer. In this cohort, 37% were unexposed to endocrine therapy, 9% had five years of tamoxifen therapy, and 54% had other endocrine regimens, some of them containing tamoxifen for less than five years with subsequent aromatase inhibitor therapy. Tamoxifen therapy was associated with a near-null risk of dementia [HR, 1.4; 95% confidence interval (CI), 1.0–1.9], and a null association was observed after death was taken into account as a competing risk (sub-HR, 1.0; 95% CI, 0.76–1.4).
Conclusions: No clinically relevant association between use of tamoxifen or other endocrine therapy and risk of dementia was observed.
Impact: Our result contradicts earlier research findings suggesting tamoxifen and other endocrine therapies increase the risk of dementia in breast cancer patients. Cancer Epidemiol Biomarkers Prev; 22(5); 993–6. ©2013 AACR.
Introduction
Adjuvant tamoxifen and aromatase inhibitors reduce the risk of recurrence in patients with breast cancer with estrogen receptor (ER)–positive tumors. Concerns have been raised about the potential for tamoxifen to impair cognitive function and increase dementia risk independently or in combination with chemotherapy, but studies have been conflicting (1, 2). We examined whether tamoxifen or other endocrine therapy was associated with risk of dementia in a large population of patients with breast cancer.
Materials and Methods
As described previously (3), we used the Danish Breast Cancer Cooperative Group (DBCG) Registry to identify all women diagnosed with stage I or II, ER-positive breast tumors between 1990 and 2004 (4). On the basis of DBCG data, we identified women prescribed with tamoxifen treatment for 5 years, other endocrine treatments including tamoxifen followed by aromatase inhibitor for a total of 5 years or more, or aromatase inhibitor for 5 years, and women not receiving endocrine therapy. We used the civil registration number, a unique identifier assigned to all Danish inhabitants, to link cohort members to the Danish National Patient Registry, The Danish Cancer Registry, and the Danish Psychiatric Central Research Register, to collect all diagnoses of dementia, osteoporosis, endometriosis, endometrial cancer, and brain cancer, as well as Charlson Comorbidity Index (CCI) scores (5–7).
We defined dementia as an incident diagnosis of Alzheimer's disease, Parkinsonism, vascular dementia, frontotemporal dementia, atypical Parkinsonism, or mild cognitive impairment using the first recorded diagnosis—the diagnostic criteria for dementia published by the Danish Dementia Research Centre. Follow-up began 1 year after breast cancer diagnosis to allow for an induction period and continued until breast cancer recurrence, diagnosis of brain cancer, death, emigration, or December 31, 2009.
We calculated crude dementia rates. Cox proportional hazard regression models were constructed to calculate HRs of the association between endocrine therapy, or endocrine therapy combined with chemotherapy, and dementia and took into account death as a competing risk. The study was approved by the Danish Data Protection Agency (2011-41-6705).
Results
The study included 16,419 women, after excluding those with preexisting dementia (n = 659) or brain cancer (n = 4). As shown in Table 1, 37% were unexposed, 9.0% were tamoxifen-exposed only, and 54% were exposed to other endocrine therapy. Only 2.1% of unexposed patients received chemotherapy compared with 41% of patients exposed to tamoxifen and 30% of patients exposed to other endocrine therapy. Among unexposed patients, 87% had no comorbidity compared with 86% of patients exposed to tamoxifen or other endocrine therapy–exposed patients. Crude rate ratio estimates were homogeneous within strata of CCI score and osteoporosis diagnoses before breast cancer. There were too few observations to stratify on history of endometriosis or endometrial cancer. Overall, there were 501 cases of dementia diagnosed during follow-up (Table 2). Among unexposed patients, 4.0% developed dementia, 3.5% among tamoxifen only exposed patients, and 2.3% among patients exposed to other endocrine therapy.
Descriptive characteristics of the patients with breast cancer diagnosed with UICC stage I or II, ER-positive breast cancer between 1900 and 2004
| . | No endocrine therapyaN (%) . | Tamoxifena, 5 y N (%) . | Other endocrine therapyaN (%) . |
|---|---|---|---|
| Number of patients | 6,141 (37) | 1,479 (9.0) | 8,799 (54) |
| Age group/years | |||
| 45–50 | 1,190 (19) | 294 (20) | 2,016 (23) |
| 51–55 | 1,070 (17) | 306 (21) | 1,947 (22) |
| 56–60 | 1,431 (23) | 304 (21) | 1,881 (21) |
| 61–65 | 1,385 (23) | 340 (23) | 1,739 (20) |
| 66–70 | 1,065 (23) | 235 (16) | 1,216 (14) |
| Menopausal status | |||
| Premenopausal | 2,260 (37) | 600 (41) | 3,963 (45) |
| Postmenopausal | 3,881 (63) | 879 (59) | 4,836 (55) |
| CCI score | |||
| 0 | 5,354 (87) | 1,269 (86) | 7,603 (86) |
| 1 | 503 (8.2) | 140 (9.5) | 743 (8.4) |
| 2 | 197 (3.2) | 45 (3.0) | 309 (3.5) |
| ≥3 | 87 (1.4) | 25 (1.7) | 144 (1.6) |
| Chemotherapy | |||
| No | 6,012 (98) | 878 (59) | 4,140 (70) |
| Yes | 129 (2.1) | 601 (41) | 2,659 (30) |
| . | No endocrine therapyaN (%) . | Tamoxifena, 5 y N (%) . | Other endocrine therapyaN (%) . |
|---|---|---|---|
| Number of patients | 6,141 (37) | 1,479 (9.0) | 8,799 (54) |
| Age group/years | |||
| 45–50 | 1,190 (19) | 294 (20) | 2,016 (23) |
| 51–55 | 1,070 (17) | 306 (21) | 1,947 (22) |
| 56–60 | 1,431 (23) | 304 (21) | 1,881 (21) |
| 61–65 | 1,385 (23) | 340 (23) | 1,739 (20) |
| 66–70 | 1,065 (23) | 235 (16) | 1,216 (14) |
| Menopausal status | |||
| Premenopausal | 2,260 (37) | 600 (41) | 3,963 (45) |
| Postmenopausal | 3,881 (63) | 879 (59) | 4,836 (55) |
| CCI score | |||
| 0 | 5,354 (87) | 1,269 (86) | 7,603 (86) |
| 1 | 503 (8.2) | 140 (9.5) | 743 (8.4) |
| 2 | 197 (3.2) | 45 (3.0) | 309 (3.5) |
| ≥3 | 87 (1.4) | 25 (1.7) | 144 (1.6) |
| Chemotherapy | |||
| No | 6,012 (98) | 878 (59) | 4,140 (70) |
| Yes | 129 (2.1) | 601 (41) | 2,659 (30) |
aBecause of rounding, percentages may not add to 100%.
Number of dementia cases and associated crude and age-adjusted HRs associating type of endocrine therapy with dementia overall and with type of dementia
| . | No dementiaaN (%) . | DementiaaN (%) . | Age-adjusted HRb (95% CI) . | Age-adjusted HRc (95% CI) . |
|---|---|---|---|---|
| Endocrine therapy | ||||
| Nonendocrine therapy | 5,895 (96) | 246 (4.0) | 1.00 | 1.00 |
| Tamoxifen, 5 y | 1,428 (97) | 51 (3.5) | 1.4 (1.0–1.9) | 1.0 (0.76–1.4) |
| Other endocrine therapy | 8,595 (98) | 204 (2.3) | 1.1 (0.88–1.3) | 0.83 (0.69–0.99) |
| Tamoxifen and chemotherapyd | ||||
| None | 5,768 (96) | 244 (4.1) | 1.00 | 1.00 |
| Tamoxifen+, chemotherapy− | 830 (95) | 48 (5.5) | 1.5 (1.1–2.1) | 0.92 (0.74–1.1) |
| Tamoxifen−, chemotherapy+ | 127 (98) | 2 (1.6) | 0.93 (0.22–3.9) | 0.93 (0.23–3.8) |
| Tamoxifen+, chemotherapy+ | 598 (100) | 3 (0.5) | 0.48 (0.15–1.6) | 0.55 (0.27–1.1) |
| Alzheimer's diseasee | ||||
| Nonendocrine therapy | 6,082 (99) | 59 (1.0) | 1.00 | 1.00 |
| Tamoxifen, 5 y | 1,473 (100) | 6 (0.4) | 0.70 (0.30–1.6) | 0.51 (0.22–1.2) |
| Other endocrine therapy | 8,751 (99) | 48 (0.6) | 1.2 (0.80–1.7) | 0.88 (0.60–1.3) |
| Parkinsonisme | ||||
| Nonendocrine therapy | 6,121 (100) | 20 (0.3) | 1.00 | 1.00 |
| Tamoxifen, 5 y | 1,473 (100) | 6 (0.4) | 1.8 (0.71–4.6) | 1.4 (0.58–3.5) |
| Other endocrine therapy | 8,778 (100) | 21 (0.2) | 1.3 (0.69–2.4) | 1.0 (0.55–1.9) |
| Atypical Parkinsonisme | ||||
| Nonendocrine therapy | 6,136 (100) | 5 (0.1) | 1.00 | 1.00 |
| Tamoxifen, 5 y | 1,479 (100) | 0 (0) | — | — |
| Other endocrine therapy | 8,798 (100) | 1 (0) | — | — |
| Vascular dementiae | ||||
| Nonendocrine therapy | 5,983 (97) | 158 (2.6) | 1.00 | 1.00 |
| Tamoxifen, 5 years | 1.442 (98) | 37 (2.5) | 1.6 (1.1–2.3) | 1.2 (0.82–1.7) |
| Other endocrine therapy | 8,667 (99) | 132 (1.5) | 1.1 (0.83–1.3) | 0.82 (0.65–1.0) |
| Frontotemporal dementiae | ||||
| Nonendocrine therapy | 6,136 (100) | 5 (0.1) | 1.00 | 1.00 |
| Tamoxifen, 5 y | 1,479 (100) | 0 (0) | — | — |
| Other endocrine therapy | 8,795 (100) | 4 (0.1) | — | — |
| Mild cognitive impairmente | ||||
| Nonendocrine therapy | 6,126 (100) | 15 (0.2) | 1.00 | 1.00 |
| Tamoxifen, 5 y | 1,476 (100) | 3 (0.2) | 1.1 (0.31–4.0) | 0.90 (0.26–3.1) |
| Other endocrine therapy | 8,787 (100) | 12 (0.1) | 0.97 (0.45–2.1) | 0.77 (0.36–1.6) |
| . | No dementiaaN (%) . | DementiaaN (%) . | Age-adjusted HRb (95% CI) . | Age-adjusted HRc (95% CI) . |
|---|---|---|---|---|
| Endocrine therapy | ||||
| Nonendocrine therapy | 5,895 (96) | 246 (4.0) | 1.00 | 1.00 |
| Tamoxifen, 5 y | 1,428 (97) | 51 (3.5) | 1.4 (1.0–1.9) | 1.0 (0.76–1.4) |
| Other endocrine therapy | 8,595 (98) | 204 (2.3) | 1.1 (0.88–1.3) | 0.83 (0.69–0.99) |
| Tamoxifen and chemotherapyd | ||||
| None | 5,768 (96) | 244 (4.1) | 1.00 | 1.00 |
| Tamoxifen+, chemotherapy− | 830 (95) | 48 (5.5) | 1.5 (1.1–2.1) | 0.92 (0.74–1.1) |
| Tamoxifen−, chemotherapy+ | 127 (98) | 2 (1.6) | 0.93 (0.22–3.9) | 0.93 (0.23–3.8) |
| Tamoxifen+, chemotherapy+ | 598 (100) | 3 (0.5) | 0.48 (0.15–1.6) | 0.55 (0.27–1.1) |
| Alzheimer's diseasee | ||||
| Nonendocrine therapy | 6,082 (99) | 59 (1.0) | 1.00 | 1.00 |
| Tamoxifen, 5 y | 1,473 (100) | 6 (0.4) | 0.70 (0.30–1.6) | 0.51 (0.22–1.2) |
| Other endocrine therapy | 8,751 (99) | 48 (0.6) | 1.2 (0.80–1.7) | 0.88 (0.60–1.3) |
| Parkinsonisme | ||||
| Nonendocrine therapy | 6,121 (100) | 20 (0.3) | 1.00 | 1.00 |
| Tamoxifen, 5 y | 1,473 (100) | 6 (0.4) | 1.8 (0.71–4.6) | 1.4 (0.58–3.5) |
| Other endocrine therapy | 8,778 (100) | 21 (0.2) | 1.3 (0.69–2.4) | 1.0 (0.55–1.9) |
| Atypical Parkinsonisme | ||||
| Nonendocrine therapy | 6,136 (100) | 5 (0.1) | 1.00 | 1.00 |
| Tamoxifen, 5 y | 1,479 (100) | 0 (0) | — | — |
| Other endocrine therapy | 8,798 (100) | 1 (0) | — | — |
| Vascular dementiae | ||||
| Nonendocrine therapy | 5,983 (97) | 158 (2.6) | 1.00 | 1.00 |
| Tamoxifen, 5 years | 1.442 (98) | 37 (2.5) | 1.6 (1.1–2.3) | 1.2 (0.82–1.7) |
| Other endocrine therapy | 8,667 (99) | 132 (1.5) | 1.1 (0.83–1.3) | 0.82 (0.65–1.0) |
| Frontotemporal dementiae | ||||
| Nonendocrine therapy | 6,136 (100) | 5 (0.1) | 1.00 | 1.00 |
| Tamoxifen, 5 y | 1,479 (100) | 0 (0) | — | — |
| Other endocrine therapy | 8,795 (100) | 4 (0.1) | — | — |
| Mild cognitive impairmente | ||||
| Nonendocrine therapy | 6,126 (100) | 15 (0.2) | 1.00 | 1.00 |
| Tamoxifen, 5 y | 1,476 (100) | 3 (0.2) | 1.1 (0.31–4.0) | 0.90 (0.26–3.1) |
| Other endocrine therapy | 8,787 (100) | 12 (0.1) | 0.97 (0.45–2.1) | 0.77 (0.36–1.6) |
aBecause of rounding, percentages may not add to 100%.
bAdjusting for menopausal status, calendar period of breast cancer diagnosis, CCI score, osteoporosis, endometriosis, or endometrial cancer did not change the estimates.
cSub-HR incorporating death as a competing risk.
dExcluding patients treated with other endocrine therapy.
eDementia types are not mutually exclusive.
Tamoxifen therapy was weakly associated with dementia risk [adjusted HR, 1.4; 95% confidence interval (CI), 1.0–1.9], but not after accounting for death as a competing risk (sub-HR, 1.0; 95% CI, 0.76–1.4). Tamoxifen use was not associated with a substantially increased risk of individual dementia diagnoses, except for a weak positive association with vascular dementia (HR, 1.6; 95% CI, 1.1–2.3), which diminished after accounting for death as a competing risk (sub-HR, 1.2; 95% CI, 0.82–1.7). Changes in the length of the induction period did not alter our findings. Chemotherapy did not increase the risk of dementia among tamoxifen-exposed patients.
Discussion
In this large population-based cohort study, we found no association between exposure to endocrine therapy and risk of dementia. Study strengths included an unselected breast cancer cohort. On the basis of DBCG treatment guidelines, all patients with breast cancer were assigned treatment and followed up with regular clinical examinations in a setting similar to a randomized controlled trial, which limited confounding from noncompliance and other factors (4). Implementation of new national treatment guidelines for breast cancer may take longer for some hospital departments than others; consequently, some patients might have been misclassified on the category of endocrine therapy. In addition, validity and completeness varies for dementia diagnoses in the registries and could bias results (8). We lacked information on potential confounders; however, none of these variables are likely associated with receipt of endocrine therapy, which is almost entirely determined by ER status of the tumor in patients with breast cancer.
Our study does not support previous research suggesting adverse effects of endocrine therapy alone or combined with chemotherapy on dementia risk, but rather is consistent with research rejecting these hypotheses (1, 2). Our near-null results are likely driven by better survival and longer follow-up time for patients treated with tamoxifen or other endocrine therapy compared with patients not treated with endocrine treatment.
In conclusion, this study does not support adverse effects of tamoxifen or other endocrine therapies on risk of dementia in Danish patients with breast cancer.
Disclosure of Potential Conflicts of Interest
No potential conflicts of interest were disclosed.
Authors' Contributions
Conception and design: A.G. Ording, A.B. Jensen, L. Pedersen, H.T. Sørensen, T.L. Lash
Development of methodology: A.G. Ording, L. Pedersen, H.T. Sørensen, T.L. Lash
Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): A.G. Ording, D. Cronin-Fenton, L. Pedersen, H.T. Sørensen
Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): A.G. Ording, A.B. Jensen, D. Cronin-Fenton, L. Pedersen, T.L. Lash
Writing, review, and/or revision of the manuscript: A.G. Ording, A.B. Jensen, D. Cronin-Fenton, H.T. Sørensen, T.L. Lash
Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): A.G. Ording
Study supervision: H.T. Sørensen
Acknowledgments
The authors thank the board of the DBCG for its permission to conduct the study and the DBCG secretariat for assistance with updating the dataset.
Grant Support
This study was supported by the Clinical Epidemiology Research Foundation, Aarhus University, Arkitekt Holger Hjortenberg og hustru Dagmar Hjortenbergs Foundation, Inge og Jørgen Larsens Mindelegat Foundation, and Aalborg Sygehus, Region North, Denmark.
RSS Feeds