3.6 Effect of ningetinib on pharmacokinetics of D6-M1 in mice.
To investigate whether ningetinib could affect the pharmacokinetics of M1 by inhibiting efflux transporters, D6-M1 was intravenously injected to ICR mice to prevent the effect of M1 generated by ningetinib metabolism. The results are shown in Fig. 7 and Table 4. Compared with the control group, the blood exposure of D6-M1 in the ningetinib-combination group was increased by approximately 75%. In addition, the blood concentration of D6-M1 was not affected by ningetinib within 2 h after administration. From 4 h to 24 h, the blood exposure of D6-M1 in the ningetinib group increased nearly by 200%, which further demonstrated that ningetinib could inhibit the elimination of D6-M1.
Discussion
In pharmacokinetic studies, M1 was identified as the primary circulating metabolite of ningetinib in humans. However, in clinical trials on multi-target-based combination medication for NSCLC therapy, both the plasma exposure and Cmax of M1 dropped dramatically and were reduced by more than 80% when ningetinib was co-administered with gefitinib, whereas the pharmacokinetic parameters (AUC, Cmax and Tmax) of ningetinib were not affected. The elimination behaviours of both ningetinib and M1 were not obviously changed. These data suggested a DDI between ningetinib and gefitinib, mostly likely through metabolic mechanism.
The mechanism of M1 formation was initially investigated. The enzymatic kinetic data showed that M1 formation was mainly mediated by CYP1A1 and to a lesser extent by CYP1B1, CYP2C9 and CYP3A4. When examining the potency of gefitinib as an inhibitor of CYP450s in vitro , our results were somewhat different from those reported in the literature (Filppula, Neuvonen, & Backman, 2014; Rahman et al., 2014). The present study revealed that gefitinib was a strong inhibitor of CYP1A1, moderate inhibitor of CYP1B1 and weak inhibitor of CYP2C9 but had no influence on CYP3A4. Gefitinib significantly inhibited M1 production in HLMs, but the concentration of the parent drug was not influenced by gefitinib, since the amount of M1 produced in HLMs and in patients was less than 5%, andN -desmethylation was not a major metabolic pathway of ningetinib in humans.
The high plasma exposure of M1 could be related to tissue distribution and/or clearance. Unlike in plasma, the concentrations of M1 in tissues were less than 30% of the parent drug, suggesting that the metabolite tended to be retained in the plasma and its penetration into tissues was more difficult than that of ningetinib. One possible explanation for the findings was that M1 exhibited much higher protein binding rates both in mouse and human plasma than the parent drug (99.9% versus 90%).
In vitro incubations showed that the liver was the primary site of M1 formation. Thus, the hepatic clearance pathway of M1 was evaluated subsequently. The transport studies demonstrated that M1 was a substrate of P-gp, BCRP and MRP2. The parent drug ningetinib happened to be a potent inhibitor of P-gp and moderate inhibitor of BCRP and MRP2. Therefore, it was speculated that ningetinib obstructed the biliary excretion pathway of M1 and resulted in its high plasma exposure. The D6-M1 pharmacokinetic experiment further proved this conclusion. In addition, the literature (Galetti et al., 2015b; Kitazaki et al., 2005) and our results confirmed that gefitinib was also an inhibitor of P-gp and BCRP. It might inhibit the efflux of M1 mediated by P-gp and BCRP. However, the IC50 value of gefitinib on P-gp was significantly higher than that of ningetinib (5.40 μM versus 0.413 μM) and gefitinib had no inhibitory effect on MRP2. Though the IC50 value of gefitinib on BCRP was lower than that of ningetinib (9.09 μM versus 18.7 μM), both of them were much higher than the peak plasma concentrations of ningetinib (1.37 μg·mL-1) and gefitinib (272 ng·mL-1) when therapeutic doses of these drugs were administered to NSCLC patients. The plasma Cmax of gefitinib was less than one third of ningetinib. In this situation, it was speculated that the efflux pathway of M1 was inhibited by ningetinib, whereas the inhibition of efflux by gefitinib was negligible due to its higher IC50 value and much lower in vivo concentrations. When co-administrated with gefitinib, the pharmacokinetic alteration of M1 showed the inhibitory effect of gefitinib on M1 formation, not on its efflux transports. The interaction of ningetinib and gefitinib mediated by P-gp and BCRP was also studied in the present work. The ER values of ningetinib were decreased by not more than 50% under the effect of 50 μM of gefitinib (Table S3). Therefore, transporter-mediated effect of gefitinib on ningetinib could be ignored.
To our knowledge, it was the first time to report that gefitinib was a strong inhibitor of CYP1A1. CYP1A1 was the most actively studied human pulmonary CYP enzyme involved in polycyclic aromatic hydrocarbons metabolism, and might play a role in the development of lung cancer (Oyama et al., 2008). It was also reported that CYP1A1 was highly expressed in human brain, colon and bladder tumour tissues compared to the normal tissues, as determined by mRNA level expression (Androutsopoulos et al., 2013; Wahid, Mahjabeen, Baig, & Kayani, 2013). In addition, CYP1A1 polymorphisms (rs4646903 and rs1048943) were positively correlated with prostate cancer, breast cancer, lung cancer, et,al (Lu et al., 2020). Therefore, gefitinib’s strong inhibitory effect on CYP1A1 may provide possibilities for its application in tumour prevention and reversal of terminal tumour resistance. In vitroexperiments in our laboratory indicated that the amount of M1 produced was positively correlated with the protein expression of CYP1A1 in HLMs and HLUMs from smokers of varying degrees. The high selectivity of ningetinib for CYP1A1 may provide its potential as an indicator of CYP1A1.
In conclusion, this study illustrated the DDI mechanism of ningetinib and gefitinib in patients with NSCLC. Ningetinib was metabolized by CYP1A1 to produce a small amount of N -demethylated metabolite M1. The low tissue affinity of M1 and the inhibitory effect of ningetinib on M1 canalicular efflux resulted in its high plasma exposure and long elimination half-life. When co-administrated, as a strong inhibitor of CYP1A1, gefitinib inhibited the formation of M1 and reduced its plasma exposure by more than 80%. However, the pharmacokinetics of the parent drug ningetinib was not influenced because of the low metabolic yield of M1. Due to the relatively high expression of CYP1A1 in pulmonary tumour tissues, the concentration of the parent drug ningetinib in the target tissue may be increased by gefitinib. As a consequence, the long-term safety profile and efficacy of ningetinib combined with gefitinib should be concerned in NSCLC patients. The findings also suggested that in-depth analysis of the pharmacokinetics of metabolites in DDI studies can help enrich the DDI mechanism and discovery of novel inhibitors (or inducers) of metabolic enzymes or transporters.