DDIs with AUC Changes ≥2-Fold.

There were approximately 60 moderate-to-strong drug interaction studies involving 15 NMEs as substrates, with more than half (N = 8) being oncology drugs. Inhibition and induction of CYP3A explained most of these interactions (about 90%). All drug interactions with AUC change ≥5-fold are presented in Table 1, with the maximum AUCR observed listed.

Based on the results of mechanistic studies with clinical index inhibitors, three drugs were identified as sensitive substrates, namely avapritinib and lonafarnib for CYP3A and relugolix for P-glycoprotein (P-gp). When concurrently administered with the strong CYP3A inhibitor itraconazole (200 mg once daily for 14 days) in healthy subjects, avapritinib, a kinase inhibitor indicated for the treatment of gastrointestinal stromal tumor (GIST), exhibited a 4.32-fold increase in AUC (after 200-mg single dose). A higher change of 7.00-fold in healthy subjects and 7.90-fold in patients with GIST in avapritinib steady-state AUC (after the clinical dose of 300 mg once daily) was predicted using PBPK modeling and simulations. The effect of moderate and weak CYP3A inhibitors was also predicted using the same approach. Erythromycin (500 mg three times daily), fluconazole (200 mg once daily), and verapamil (80 mg three times daily), all moderate CYP3A inhibitors, were predicted to increase avapritinib steady-state AUC 2- to 3-fold in patients with GIST, whereas the weak CYP3A inhibitor cimetidine (400 mg three times daily) was not predicted to affect avapritinib exposure (FDA, 2020b). Because higher plasma concentrations of avapritinib may increase the incidence and severity of adverse reactions, concomitant administration with strong and moderate CYP3A inhibitors should be avoided. If coadministration with a moderate CYP3A inhibitor cannot be avoided, the dose of avapritinib should be reduced (FDA, 2020b). Lonafarnib is an orphan drug used to reduce the risk of death due to Hutchinson-Gilford progeria syndrome and for the treatment of certain processing-deficient progeroid laminopathies in patients 12 months of age and older. Coadministration with ketoconazole (200 mg once daily for 5 days), a strong CYP3A inhibitor, resulted in a 5.07-fold increase in lonafarnib AUC in healthy subjects (FDA, 2020p). CYP3A inhibitors with lower potency were not evaluated but were expected to increase lonafarnib exposure to a clinically meaningful extent. Therefore, because of safety concerns, the use of lonafarnib with strong or moderate CYP3A inhibitors is contraindicated. Concomitant use of lonafarnib with weak CYP3A inhibitors should be avoided as well; however, if coadministration is unavoidable, the dose of lonafarnib should be reduced. Additionally, patients should be monitored for adverse reactions, such as arrhythmias and other cardiovascular events like syncope and heart palpitations, as the effect of lonafarnib on QT interval is unknown (FDA, 2020p). Finally, exposure to the oncology drug relugolix was found to increase when coadministered with erythromycin (6.25-fold increase in AUC and 6.18-fold increase in C_max), primarily due to inhibition of intestinal P-gp. In vitro, relugolix was a P-gp substrate, with an efflux ratio of 16.4 in Caco-2 cells. Although contribution of CYP3A cannot be fully ruled out because relugolix was metabolized by CYP3A in vitro and erythromycin is also a moderate CYP3A inhibitor, CYP3A is expected to play a minimal role given that a much smaller increase (1.21- to 1.51-fold) in the exposure of relugolix was observed in the presence of moderate and strong CYP3A inhibitors (FDA, 2020i).

Compared with the inhibition results, more drugs (N = 6) were sensitive to induction. In addition to avapritinib and lonafarnib, fostemsavir (prodrug, active moiety temsavir), rimegepant, pemigatinib, and selpercatinib were all sensitive to CYP3A induction, with AUC decreases of 81%–87% after concomitant administration of multiple doses of rifampin, an index inducer of CYP3A. Notably, induction of P-gp may also be involved in these interactions, as all were P-gp substrates in vitro except avapritinib (FDA, 2020a,b,g,j,k,p). To avoid reduced efficacy associated with the significant decrease in drug exposure, concomitant use with strong CYP3A inducers is either contraindicated or should be avoided for these drugs (FDA, 2020a,b,g,j,k,p).

Regarding moderate inhibition, 10 drugs were found to be moderate sensitive substrates (AUCRs 2–5) based on inhibition or PGx results: fostemsavir (CYP3A), oliceridine (CYP2D6), ozanitib [breast cancer resistance protein (BCRP)], pemigatinib (CYP3A), pralsetinib (CYP3A), rimegepant (CYP3A), selpercatinib (CYP3A), tazemetostat (CYP3A), tucatinib (CYP2C8), and vibegron (P-gp). Oliceridine, an opioid agonist indicated for the management of severe acute pain, was primarily metabolized by CYP2D6 and CYP3A4 in vitro. The moderate sensitivity to CYP2D6 inhibition was identified through PGx studies, wherein CYP2D6 poor metabolizers showed approximately 2-fold higher exposure to oliceridine compared with normal metabolizers. The effect of concomitant administration of a CYP2D6 inhibitor on oliceridine pharmacokinetics has not been evaluated but is expected to be similar to what was observed in CYP2D6 poor metabolizers based on population pharmacokinetic analyses. An approximate 4-fold increase in oliceridine AUC and 70% reduction in clearance were expected in CYP2D6 normal metabolizers in the worst-case scenario of concomitant inhibition of both CYP3A4 and CYP2D6 enzymes, whereas coadministration with itraconazole, a strong CYP3A inhibitor, was estimated to reduce oliceridine clearance by 45% (FDA, 2020h). On the other hand, based on induction studies, three kinase inhibitors (capmatinib, pralsetinib, and selumetinib) were found to be moderate sensitive substrates of CYP3A, with coadministration of multiple doses of rifampin significantly reducing their exposure by 51%–68% (FDA, 2020c,f,l). Most of DDIs mentioned above led to specific label recommendations when concomitantly administered with known inhibitors or inducers. For instance, it is recommended to avoid concomitant use of selpercatinib with strong CYP3A inhibitors; if coadministration is unavoidable, the dose of selpercatinib should be reduced by half. In addition, because selpercatinib can cause QT prolongation, it is suggested to monitor the QT interval with ECGs more frequently (FDA, 2020a,o). For ozanimod, a drug indicated for the treatment of multiple sclerosis, coadministration with inhibitors of BCRP is not recommended (FDA, 2020o).

DDIs with AUC Changes <2-Fold but with Clinical Implications.

There were about 50 studies with AUCR 1.25–2 or 0.5–0.8 involving NMEs as substrates, but less than half led to clinical recommendations. Inhibition or induction of CYP3A explained more than 60% of these results, and the majority of studies were evaluations of sensitive substrates with less potent inhibitors or inducers. Additionally, the following seven drugs were found to be weak substrates based on studies with strong inhibitors (itraconazole for CYP3A, gemfibrozil for CYP2C8, and cyclosporine for P-gp and BCRP): berotralstat (P-gp, BCRP), capmatinib (CYP3A), oliceridine (CYP3A), ozanimod (CYP2C8), relugolix (CYP3A), rimegepant (CYP2C9; PGx study), and ripretinib (CYP3A). Given the smaller change in the drugs’ exposure, label recommendations are mostly to monitor for adverse events and reduce dose or dosing frequency as needed. However, a few drugs have stricter recommendations because of a narrower therapeutic range. For instance, it is recommended to avoid coadministration of strong or moderate CYP3A4 inhibitors or fluconazole with selumetinib, a kinase inhibitor to treat neurofibromatosis type 1. If coadministration cannot be avoided, the dose of selumetinib should be reduced (FDA, 2020f). Similarly, for ozanimod, “Coadministration of ZEPOSIA with strong CYP2C8 inhibitors (e.g., gemfibrozil) is not recommended” (FDA, 2020o).

DDIs with AUC Changes ≥2-Fold.

There were only 10 moderate-to-strong drug interactions involving NMEs as precipitants, and all were related to inhibition, with no strong or moderate inducer of enzymes or transporters identified. Inhibition of CYP3A explained about half of the interactions. A total of seven drugs were involved, with more than half (N = 4) indicated for cancer treatment. All interaction results led to label recommendations to mitigate the risk of DDI in clinical settings.

Three drugs were considered strong inhibitors of enzymes (victim drug AUCR ≥5): cedazuridine for cytidine deaminase, lonafarnib for CYP3A, and tucatinib for CYP3A (Table 1). No drug exhibited strong inhibition of transporters. Notably, cedazuridine, indicated in combination with decitabine for the treatment of myelodysplastic syndromes, is used to prevent the rapid metabolism of decitabine in the gastrointestinal tract to allow for oral administration and therefore increase systemic exposure of decitabine. Following different dosing regimen, cedazuridine was found to increase the AUC of decitabine up to 12-fold (FDA, 2020d). Because cedazuridine is a strong inhibitor of cytidine deaminase, coadministration of the combination drug with drugs metabolized by cytidine deaminase may result in increased exposure with potential for increased toxicity of these drugs. Therefore, coadministration of cedazuridine and decitabine with drugs that are metabolized by cytidine deaminase should be avoided (FDA, 2020d). Regarding lonafarnib and tucatinib, both caused a significant increase in the AUC of midazolam, an index substrate of CYP3A (7.39- and 5.74-fold, respectively). The clinical trials were performed because in vitro results suggested that both drugs have the potential to inhibit CYP3A at clinically relevant concentrations. In vitro, both were mechanism-based inhibitors of CYP3A (K_I = 0.54 µM and k_inact = 0.011/min for tucatinib; values not provided for lonafarnib in the NDA review), and tucatinib was found to also reversibly inhibit CYP3A with a K_i value of 0.805 µM (FDA, 2020n,p).

Additionally, the following four drugs were found to be moderate inhibitors: berotralstat (CYP2D6 and CYP3A), capmatinib (CYP1A2 and BCRP), osilodrostat (CYP1A2 and CYP2C19), and selpercatinib (CYP2C8), with increases in the index substrates of 2- to 3-fold. Label recommendations are provided regarding concomitant medications that are substrates of these cytochrome P450 isoforms or BCRP transporter with a narrow therapeutic index (NTI). For example, both osilodrostat (indicated for the treatment of Cushing’s disease) and capmatinib (a kinase inhibitor to treat metastatic non–small cell lung cancer) increased the AUC of the CYP1A2 index substrate caffeine to a similar extent (AUCRs of 2.50 and 2.43, respectively), with no change in its C_max. Both labels give recommendations regarding concomitant administration of CYP1A2 substrates with an NTI. For osilodrostat, it is recommended to use with caution or decrease dose of the substrate drug when coadministered with CYP1A2 substrates with an NTI (FDA, 2020e). If concomitant use is unavoidable between capmatinib and CYP1A2 substrates, in which minimal concentration changes may lead to serious adverse reactions, dose of the CYP1A2 substrates should be decreased in accordance with its approved prescription information (FDA, 2020l).

DDIs with AUC Changes <2-fold but with Clinical Implications.

There were 56 studies showing weak inhibition or induction. Similarly to the substrate studies, only 40% of these interactions were considered clinically relevant. Eight drugs weakly inhibited one or two specific cytochrome P450 enzymes, leading to label recommendations for only two of them, namely lonafarnib (CYP2C19) and selpercatinib (CYP3A). Regarding induction studies, three drugs exhibited weak induction, but only one drug, tazemetostat, had label recommendations based on the results. Indeed, tazemetostat, an orphan drug for epithelioid sarcoma, was found to decrease the AUC of midazolam by 37% when coadministered in patients with cancer at the dose of 800 mg twice daily, due to induction of CYP3A. It is noted in the label that “coadministration of CYP3A substrates, including hormonal contraceptives, with tazemetostat can result in decreased concentrations and reduced efficacy of CYP3A substrates” (FDA, 2020m).

About a third of these weak interactions were mediated by drug transporters, involving P-gp, BCRP, organic anion–transporting polypeptide (OATP) 1B1/1B3, organic cation transporter (OCT) 2, and multiantimicrobial extrusion protein (MATE) 1/2-K. Regarding inhibition of renal transporters, in vitro capmatinib inhibited MATE1 (K_i = 0.28 µM) and MATE2-K (K_i = 0.29 µM), pemigatinib inhibited MATE1 (IC₅₀ = 0.075 µM) and OCT2 (IC₅₀ = 1.1 µM), selpercatinib inhibited MATE1 (IC₅₀ = 0.666 µM), and tucatinib inhibited MATE1 (IC₅₀ = 0.0855 µM) and MATE2-K (IC₅₀ = 0.135 µM) as well as OCT2 (IC₅₀ = 0.107 µM). Because of the potential inhibition of MATEs and OCT2 in vivo by these four oncology drugs, endogenous creatinine was measured as the marker of renal function as well as activity of MATEs and OCT2. For capmatinib, all subjects showed transient grade 1 serum creatinine increase postdose during a 72-hour sampling period. According to the NDA review, these results indirectly suggest that reversible inhibition of renal transporters (i.e., MATE1 and MATE2-K) may explain the increase in serum creatinine. Based on this observation, the label states that “coadministration of capmatinib may increase the exposure of MATE1 and MATE2-K substrates, which may increase the adverse reactions of these substrates. If coadministration is unavoidable between capmatinib and MATE1 or MATE2-K substrates where minimal concentration changes may lead to serious adverse reactions, dose of MATE1 or MATE2K substrate should be decreased in accordance with the approved prescribing information” (FDA, 2020l). For pemigatinib, endogenous serum creatinine was found to increase an average of 0.2 mg/dl within the first 21-day cycle of treatment, whereas for selpercatinib and tucatinib, a mean increase of 1.18- to 1.32-fold in endogenous serum creatinine was observed in clinical studies, and their labels suggest that “alternative markers of renal function should be considered if persistent elevations in serum creatinine are observed” (FDA, 2020a,j,n).