An updated version of this post can be found HERE.

Two thirds of all breast cancer tumors are known to have receptors that respond to hormones; that is they can grow in response to estrogen. Tamoxifen is a medication which selectively blocks estrogen receptors on breast cancer tumors and, in this fashion, reduces the risk of recurrence by half in women with estrogen receptor-positive breast cancer (1).  

Tamoxifen, however, is a “prodrug”.  That means that after it is taken, it needs to be “activated” or converted into its active form, endoxifen (most prevalent active form), by a liver enzyme called CYP2D6.

Treatment of Depression in Women with Breast Cancer

Roughly half of women diagnosed with breast cancer report symptoms of depression, anxiety or both during the first year of receiving the diagnosis (1).  It is estimated that one in four women taking tamoxifen are also taking an antidepressant, most commonly serotonin reuptake inhibitors (SRIs).  Given that so many women with breast cancer take antidepressants, studying the interaction between medications used to treat either of these disorders is of importance, as they are often prescribed together.

The major concern is that some antidepressants inhibit the function of the CYP2D6 enzyme (described above) which is required for the conversion of tamoxifen to its active form.  Examples of strong CYP2D6 inhibitors are bupropion (Wellbutrin), fluoxetine (Prozac), and paroxetine (Paxil).  Other SSRI medications, such as citalopram (Celexa), escitalopram (Lexapro), venlafaxine (Effexor) and sertraline (Zoloft) are considered to be less potent inhibitors of CYP2D6.  If taken with tamoxifen, these CYP2D6 inhibitors could possibly decrease the available amount of the active metabolite (endoxifen).

Naturally, the following question arises: does taking CYP2D6-inhibiting antidepressants along with tamoxifen decrease the effectiveness of tamoxifen?

Measuring the Interaction Between Tamoxifen and Antidepressants

There have been many studies on interaction between antidepressants which have yielded conflicting results.  In an article published in the September 2016 issue of BMJ, Donneyong et al. set out to compare differences in mortality (risk of death) in women who were treated with the combination of tamoxifen and strong CYP2D6 inhibitors (fluoxetine or paroxetine) vs. women who were treated with tamoxifen and weaker CYP2D6 inhibitors (citalopram, escitalopram, fluvoxamine, sertraline).    

This study used data collected from health insurance programs from 1995 to 2013. The women were divided into two cohorts: Cohort 1 had 6,067 women who were taking an SSRI and then started treatment with tamoxifen.  Cohort 2 included 8,465 patients who were taking tamoxifen and then started taking an SSRI.

The use of tamoxifen along with potent CYP2D6-inhibiting SSRIs versus other SSRIs was not associated with an increased risk of death.  After about two years of follow-up, pooled hazard ratios for death were similar (58.6/1000 for inhibitors vs. 57.9/1000 for other SSRIs). They also took into account which medication was started first (SSRI or tamoxifen) with no change in results reported.  The authors extrapolate that CYP2D6-inhibiting SSRIs do not block the conversion of tamoxifen completely and might still leave enough active metabolite to adequately block estrogen receptors.

It is important to note, however, that the above study is not conclusive in answering whether tamoxifen is just as effective if taken with strong CYP2D6 inhibitors versus other SSRI medications.  The authors list the following limitations in analyzing their data.  Since this is very large population based data they were not able to identify the cause of death for all patients.  That is, was the cause of death cancer related or not? They also noted that, as in other large observational studies, certain variables such as socioeconomic status, smoking, and obesity can have confounding effects that are difficult to fully account for during data analysis.  Additional limitations include the fact that the researchers assumed that prescribers did not preferentially select SSRI medications based on the prognosis of breast cancer.  They also did not have data on blood levels of medication to indicate whether or not the patients actually took the prescribed medications.

Perhaps the most important limitation of this study is that the mean follow-up time  was 2.4 years.  As the authors indicate, this limits the conclusiveness of their analysis as it may not pertain to the long-term safety profile of concomitant use of tamoxifen and the named SSRIs.  This point should be further considered when taking into account that the authors were limited in their ability to distinguish stages of breast cancer due to using claims data.  

In an editorial in the same issue of BMJ, Juurlink outlines the importance of this issue and the controversial nature of such studies.  Such studies are important, he indicates, as tamoxifen has been monumental in reducing the recurrence and mortality of breast cancer.  The SSRI medications are often co-prescribed with this medication for prolonged periods for treatment of depression and/or anxiety, as well as vasomotor symptoms associated with tamoxifen treatment.  He points out that the possible consequences of said interaction are not acute, and are delayed by years. Tamoxifen pharmacokinetics, he notes, involves processes other than CYP2D6 which adds to the complexity of such studies.  

Juurlink noted that  the findings of this study were “unsurprising, if for no other reason than follow-up was too brief for any differential survival to show up.” Furthermore, “studying total mortality rather than cancer specific outcomes further diminishes the investigators’ ability to discern signal from noise.  Consequently this study does little to disprove a meaningful interaction between tamoxifen and CYP2D6 inhibitors.”   The editorial points out that other effects not studied include: non-adherence, therapeutic switching, dose-response effects, CPY2D6 polymorphisms, mechanisms and degrees of inhibition, and endoxifen threshold needed for successful treatment.  This study does illustrate, however, just how challenging it can be to conduct such studies.

Guidelines for Treating Depression in This Population

In short, drawing any conclusions related to interaction between tamoxifen and antidepressants would be premature at this stage.  This is especially true considering the life saving effects of tamoxifen and the fact that we can likely sidestep any possible harm due to possible medication interactions.  The editorial points out the following three logical steps to guide co-prescription of these medications (2):

  • First, antidepressants should only be prescribed to those who have likely benefits for these medications.  
  • Second, prescribers should avoid co-prescribing antidepressants which are strong CYP2D6 inhibitors (i.e. fluoxetine, bupropion, and especially paroxetine) to the extent that it is possible and to instead prescribe reasonable alternatives such as sertraline, citalopram, escitalopram, and venlafaxine which have lesser degrees of inhibition.  
  • Third, for patients who are currently prescribed tamoxifen and a known CYP2D6 inhibitor the cross-taper of antidepressants should be gradual in order to reduce the risk of withdrawal or adverse effects during the transition to a new medication (2).

Of note, while the above recommendations are quite reasonable, there are many situations in which a women has responded to a potent CYP2D6 medication well and alternatives have not proven effective.  A discussion of risks, benefits and alternatives in such situations is essential so that patients may navigate the course of treatment in the most informed fashion.

Edwin Raffi, MD, MPH

  1. Donneyong, M.M., Bykov, K., Bosco-Levy, P., Dong, Y.H., Levin, R. and Gagne, J.J., 2016. Risk of mortality with concomitant use of tamoxifen and selective serotonin reuptake inhibitors: multi-database cohort study. BMJ354, p.i5014.
  2. Juurlink, D., 2016. Revisiting the drug interaction between tamoxifen and SSRI antidepressants. BMJ 2016; 354: i5309.