Tamoxifen undergoes oxidative biotransformation to its metabolites primarily by the CYP enzymes CYP3A4 and CYP2D6. In particular, endoxifen formation, the main active metabolite thought to be responsible for tamoxifen’s therapeutic effect, is catalyzed by the CYP2D6 enzyme [4, 10]. It was therefore traditionally hypothesized that the concurrent use of potent CYP3A4 inducers, such as rifampin, may increase the amount of endoxifen formed, and in fact have a beneficial effect on the clinical outcomes of tamoxifen users, particularly among CYP2D6 poor metabolizers (PM). The effect of potent inducers such as rifampin can be multifactorial. We have shown that nuclear receptors such as Pregnane X Receptor (PXR) and Hepatocyte Nuclear Receptor 4α (HNF4α) regulate CYP3A4 expression [11]. Therefore, in principle, activation of PXR would be predicted to increase the level of endoxifen. However, we now know that endoxifen clearance may be further enhanced through the action of phase II enzymes as well as drug transporters. Indeed, our group has shown that endoxifen is an excellent substrate for P-glycoprotein and that central nervous system entry of endoxifen was nearly 20-fold greater in P-glycoprotein deficient mice, due to its absence at the blood brain barrier [12]. P-glycoprotein is also highly expressed in the apical domain of enterocytes thus reducing substrate drug absorption, while its expression on the canalicular domain of hepatocytes facilitates biliary excretion [13]. In addition, induction of glucuronidation is also likely involved in the enhanced clearance of tamoxifen and its metabolites [14]. Rifampin has been shown to induce several conjugating enzymes, including the uridine 5’-diphospho-glucuronosyltransferase (UGT) enzyme that catalyzes glucuronidation [15]. An interaction between rifampin and tamoxifen was noted in the late 1990s when a study showed a marked reduction in tamoxifen levels [16]. However, in terms of clinical impact, only recently have reports demonstrated drugs such as phenytoin and rifampin may have a deleterious effect on endoxifen formation [7, 8]. Based on the previous hypothesis that concurrent use of rifampin and tamoxifen would enhance endoxifen levels, the clinical trial published by Binkhorst et al was designed to show a beneficial effect among breast cancer patients on tamoxifen therapy [8]. However, the investigators terminated the study prematurely due to interim analysis data that showed profound reduction in endoxifen levels.
Here, we had the opportunity to investigate a potential DDI in a patient requiring both tamoxifen and rifampin. Through measurement of tamoxifen and metabolite levels prior to and following rifampin administration we were able to clearly document the reduction in metabolite formation. As the bioactivation of tamoxifen is complex, involving multiple drug metabolizing enzymes and transporters for metabolite clearance, we suggest that rifampin induction may play a role in multiple, non-mutually exclusive pathways resulting in lower systemic exposure of endoxifen. Induction of phase II enzymes and P-glycoprotein may result in increased clearance of endoxifen (Fig. 2). Additionally, the large reduction in the ratio of either primary metabolite to tamoxifen suggests that tamoxifen may in fact be metabolized to an alternate metabolite, reducing the formation of NDM-tam and 4-OH-tam, which are necessary for endoxifen formation (Fig. 2). It is important to note that the absorption and bioavailability of many drugs may be affected in the setting of inflammatory bowel disease. However, here we were able to measure the ratio of endoxifen to tamoxifen at several time points during the UC flare, including just prior to rifampin initiation (considered as baseline) and two weeks post rifampin discontinuation. The profound reduction in the endoxifen/tamoxifen ratio following rifampin treatment suggests that endoxifen is being cleared more rapidly than at baseline. Therefore, although we can’t rule out altered gut absorption, we predict this effect is primarily due to the drug interaction and not the disease setting.
Rifampin is used in the treatment of a number of infections, including those caused by tuberculous or non-tuberculous mycobacterium, and methicillin-susceptible or methicillin-resistant staphylococci. This broad spectrum of coverage likely means that its use remains consistent, particularly in the setting of a large, tertiary hospital site. Additionally, the use of tamoxifen is widespread in those who have received a diagnosis of breast cancer, as well as for specific populations of women at high-risk for cancer. Therefore, it is likely that a significant number of patients may be receiving concurrent therapy with both agents. We are struck by the magnitude of the inductive effect of rifampin on tamoxifen metabolism. A near 3-fold reduction in endoxifen level after only 10 days of rifampin therapy would suggest most patients on such combinations would be predicted to lose therapeutic benefit from tamoxifen placing them at a higher risk of recurrence.