Tamoxifen (TAM) is an effective selective estrogen receptor modulator (SERM) widely used for the treatment and chemoprevention of estrogen receptor (ER) positive tumors. Notwithstanding the advent of aromatase inhibitors (AIs) and newer SERMs, TAM remains an important member of the armamentarium against breast cancer. Despite its advantages and a strong track record of efficacious use, TAM is associated with significant clinical problems. Its use is associated with a large inter-subject variability in its pharmacokinetics (PK) and resulting therapeutic activity/side effects. It may affect the clearance of co-administered drugs by inducing and⁄or inhibiting CYP3A. CYP3A plays a central role in TAM metabolism to its major metabolite, N-desmethyl tamoxifen (N-DMT), and minor metabolites including α-hydroxy tamoxifen (α-OHT), which is implicated in endometrial toxicity. The role of CYP3A5 and the impact of its polymorphic expression on the formation of α-OHT and N-DMT are not known. Furthermore, previous work in our laboratory indicated that TAM induces CYP3A in human hepatocytes by activating pregnane X receptor (PXR). Therefore, we investigated the genetic influence on TAM metabolism in vitro and CYP3A induction in breast cancer patients undergoing TAM therapy. Using cDNA expressed CYP3A4, CYP3A5, CYP3A5∗8, CYP3A5∗9 microsomes and a panel of human liver microsomes genotyped for CYP3A4⁄5 variants, enzyme kinetics for the formation of α-OHT and N-DMT were determined. Our in vitro findings suggest that the formation of α-OHT is primarily mediated by CYP3A4, and is not affected by CYP3A5 genotype. This is further supported by the lack of association between CYP3A5 genotypes and plasma concentrations of α-OHT in our clinical study. On the other hand, both CYP3A4 and CYP3A5 contribute to the formation of N-DMT. The intrinsic clearance for N-DMT formation by cDNA expressed CYP3A4 and CYP3A5 were 0.57 and 0.35 ml⁄min⁄pmol P450, respectively. Furthermore, there was a significant difference in the Vmax of N-DMT formation between CYP3A5∗1⁄ ∗1 and CYP3A5∗3⁄∗3 microsomes (267.63 Vs 101.03 pmol⁄min⁄mg).
Employing midazolam (MDZ) as a CYP3A probe, we assessed the induction of CYP3A by TAM in breast cancer patients. Apparent oral clearance of midazolam was determined at the baseline (prior to TAM administration), Day 1 and Day 42 after administration of a TAM (20 mg⁄day). In 6 out of 13 patients, we observed an increase in the clearance of MDZ (mean: 70%, range: 26 to 161%) indicative of CYP3A induction. Furthermore, there was a marked inter-subject variability in the extent of CYP3A induction. Blood samples were genotyped for polymorphisms in CYP3A (CYP3A4∗1B and CYP3A5∗3) and PXR (−25385C>T, −24381 A>C and 63396 C>T) to investigate their association with CYP3A induction. We observed a trend towards higher induction in subjects with CYP3A5∗1⁄∗3 than CYP3A5∗3⁄∗3 genotypes. One patient with maximum induction (2.5 fold) was homozygous to TT and CC variants of PXR (−25385C>T and −24381 A>C). No association was found with CYP3A4∗1B, PXR 63396 C>T and CYP3A induction. Overall, our studies provide novel insights into genetic contributions to the formation of CYP3A derived metabolites such as N-DMT and α-OHT, and to the inter-subject variability in CYP3A inductive effect of TAM.