Cognitive Dysfunction in Postmenopausal Breast Cancer Patients on Aromatase Inhibitors

Michael Batalo; Govardhanan Nagaiah; Jame Abraham


Expert Rev Anticancer Ther. 2011;11(8):1277-1282. 

In This Article

AIs & Cognition

Studying the possible association between AI and cognitive dysfunction has accelerated over the past 7 years. A pilot study using neuropsychological testing from the Arimidex, Tamoxifen, Alone or in Combination (ATAC) trial showed that patients receiving hormone therapy performed lower on processing speed and verbal memory tasks than a control group.[13] Since this 2004 study, aromatase inhibitor use and any effect on cognition remain under consideration. A 2009 ASCO Breast Cancer Symposium review of five studies evaluated the majority of this research and reported that associations are seen with endocrine treatment and cognition.[14] Just as these five studies show conflicting results for hormone therapies including tamoxifen and aromatase inhibitors, they all vary in study design and analysis. However, significant data exist in these studies that aromatase inhibitors may not negatively affect cognitive function as purportedly as much seen by the ATAC trial pilot study. Two prospective studies report such findings. Although still having a negative effect on cognition, data from the Breast International Group (BIG) 1–98 trial suggest that letrozole use, as measured through size effect, has a small-to-moderate difference in having a less negative association with poorer cognition when compared with tamoxifen use. Perhaps the most relevant study, the International Breast Intervention Study (IBIS)-II trial reports that anastrozole use has no statistically significant associated negative effect on cognition over placebo in postmenopausal women who were known to be breast cancer-free.

The IBIS-II study is ideal to control for two relevant confounding factors such as cancer and any effect that prior chemotherapy may have on cognition.[15,16] IBIS-II comprised a randomized, double-blind study of anastrozole compared with placebo in postmenopausal patients without known breast cancer or prior chemotherapy use. Such a study involving patients without known breast cancer or chemotherapy provides the needed control for these factors in evaluating the role of hormonal therapy in breast cancer patients. The study construct involved a set of standardized cognitive tasks. Overall, there were no significant differences between anastrozole and placebo groups when analysis was restricted to the baseline and 6-month data. In addition, the objective cognitive tasks did not seem to be influenced by time. After 2 years, 151 women were available for full-duration data analysis. A significant weak point of this study includes that prior hormone-replacement therapy use was more common in study participants who had received placebo instead of anastrozole. Even though the average age (57 years) of this group was below 65 years old and greater that the WHIMS observed, this group showed more cognitive decline at baseline than participants who would receive anastrozole.[4] Not controlling for this factor could explain the study's finding of no difference between placebo and anastrozole by potentially matching any decrease on cognition that anastrozole may potentially cause. Since the standard duration of aromatase inhibitor therapy is currently at least 5 years, and possibly longer after future studies, the short time course of this study limits how clinically applicable these findings currently remain. In addition, missing values at baseline required their imputation by using linear regression analysis of the measure of intelligence and age in order to calculate mean change from baseline.

While the IBIS-II study controlled for breast cancer and prior chemotherapy as confounders, the following studies all include breast cancer patients with only a few controlling for prior chemotherapy treatment.

A more recent analysis of a prospective trial involved the aromatase inhibitor letrozole under a 5-year study period and analyzed data from 120 breast cancer patients of the adjuvant endocrine treatment BIG 1–98 trial that included four treatment arms of letrozole alone, tamoxifen alone, letrozole for 2 years followed by 3 years of tamoxifen, or tamoxifen for 2 years followed by 3 years of letrozole.[17] Some patients in this study had received prior chemotherapy. After controlling for language, no statistically significant difference existed for any individual task between tamoxifen and letrozole users but the letrozole group had statistically significant higher composite scores than the tamoxifen group. Although not shown in the article, control measures were taken to address a significant confounding set of factors that could explain the difference in composite scores: tamoxifen users were more likely to take antidepressants and had recently consumed alcohol prior to the study. Some of the same authors in a subsequent study report improved cognitive function after treatment cessation in postmenopausal patients who received either adjuvant letrozole or tamoxifen alone or in sequence.[18]

Another prospective study analyzed aromatase inhibitor use in breast cancer patients and cognitive function more generally by observing for associations of cognitive dysfunction amid induced menopause and antiestrogen therapy (AIs and tamoxifen).[19] While no differences in terms of cognitive functioning with antiestrogen treatment were observed, this study did not control for prior chemotherapy. In addition, menopause status was subjectively determined by study participants instead of laboratory testing, which remains significant in how group differences were established. Interestingly, patients who underwent induced menopause had a statistically significant favorable influence for specific word-fluency testing.[19]

One repeatedly cited cross-sectional pilot study of 31 postmenopausal women with early-stage breast cancer reported statistically significant differences in cognitive performance between patients who received anastrozole and tamoxifen.[20] In particular, it was observed that anastrozole users had poorer visual and verbal learning and memory performance than women receiving tamoxifen. Although controlling for breast cancer staging remains a strength of this analysis, women in this study still received chemotherapy, which could confound evaluating for any associations between antiestrogen therapy and cognitive performance. While this study controlled for marital status, number of children, number of years of education and breast cancer staging, it did not initially control for age, which is significant as aromatase inhibitor users were older than women receiving tamoxifen by an average of approximately 9 years. Reportedly, the authors secondarily controlled for age and still report a difference in cognitive performance between the two groups. While these limitations, including small study size, are acknowledged by the authors, the call for more controlled prospective studies with baseline evaluations in the study population remains relevant.

While no statistically significant difference between tamoxifen and exemestane users was observed in a cross-sectional study of 178 postmenopausal breast cancer patients who underwent hormonal therapy after doxorubicin/cyclophosphamide (AC) chemotherapy, it was determined that study participants who received AC chemotherapy and either of the antihormonal agents were observed to have a statistically significant worse cognitive function than healthy controls.[21] However, the authors note that this cross-sectional study did not allow for control of factors that occurred before time of treatment. Second, the authors also acknowledge that the sequential treatment limits our understanding if the chemotherapy or antihormonals caused the reported objective cognitive dysfunction. In addition, they suggest that even a more novel explanation could be considered that two separate sequential hits against the brain occurred, affecting cognition on a clinically measurable level.

In an extended study from their original evaluation of chemotherapy on cognition, Collins et al. evaluated the cognitive effects of hormonal therapy in breast cancer patients.[22] They purport that anastrozole causes cognitive decline (processing speed and verbal memory) in females taking this medication. However, the authors acknowledge that the majority of the scores in the hormonal treatment groups were still within normal range after treatment. While the authors understood that this study would have a greater ability to evaluate the cognitive effects of hormonal therapy with a planned design involving carefully chosen controls and pretreatment assessment, they report that subtle differences between hormonal treatment groups and healthy controls should not be assumed to be insignificant. Another concerning aspect of this study is the small population of anastrozole users tested in only 14 participants.

Moving towards controlling prior chemotherapy's possible effect on cognitive functioning, Schilder et al. separately examined the effects of tamoxifen and exemestane in postmenopausal breast cancer patients that did not undergo chemotherapy.[7] This study demonstrates the significant finding that no cognitive dysfunction was observed in postmenopausal hormone receptor-positive breast cancer patients on the aromatase inhibitor exemestane. Although this study aims to control for the cognitive dysfunction confounder of the presence of breast cancer through the inclusion of baseline cognitive testing, patients were only assessed with 1 year of exemestane use, which again is 4 years less than the fully recommended treatment duration. Also reported in this study is that tamoxifen use was associated with cognitive dysfunction in an age-dependent manner, with tamoxifen users 65 years or younger performing worse than healthy controls on executive functioning. Tamoxifen users older than 65 years had worse performance in verbal memory and information processing speed when compared with healthy controls. Although baseline or pretreatment testing was performed to attempt to control for the cognitive dysfunction confounder of having breast cancer, one cannot fully eliminate having breast cancer as a source of cognitive dysfunction in these patients. The authors note that their use of statistical adjustment cannot fully correct this difference.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.