Drug–drug interactions in targeted cancer therapies: a focus on tyrosine kinase inhibitors Authors: Jack C Stromatt, Anika T Chowdhury, Kevin M Huang, Shuiying Hu, Alex Sparreboom, Sharyn D Baker, Eric D Eisenmann Published: 2025 ABSTRACT AbstractIntroduction:Tyrosine kinase inhibitors (TKIs) have significantly improved outcomes for cancer patients. However, TKIs have a narrow therapeutic window, impact a high-risk population, and are frequently combined with additional medications, each of which increases the risk of drug–drug interactions (DDIs). As polypharmacy becomes more common, especially in older and higher-risk groups that are prescribed TKIs, there is an increased risk for harmful DDIs.Areas covered:TKI-related DDIs can lead to toxicity or reduced efficacy, which worsens patient outcomes and quality of life. In this review, we utilized the Drug Interaction Database (DIDB), FDA labels, and PubMed TKI DDI/ADME queries to outline the mechanisms underlying both pharmacokinetic (PK) and pharmacodynamic (PD) DDIs with TKIs. Pharmacokinetic DDIs involve changes in the absorption, distribution, metabolism, and excretion (ADME) of a molecule (victim) caused by another drug (perpetrator). These DDIs may involve direct inhibition, post-translational modification, or transcriptional regulation. Pharmacodynamic DDIs involve overlapping mechanisms resulting in additive, synergistic, or antagonistic effects when two medications are combined.Expert opinion:Understanding the mechanisms underlying DDIs facilitates a better grasp of TKI clinical pharmacology and will help in the development of strategies to optimize drug dosing and avoid DDIs.Keywords: ADME, DDIs, TKIs, pharmacokinetics, pharmacodynamics Introduction Cancer is a collection of heterogeneous diseases characterized by unrestrained proliferation and resistance to cell death [1]. Historical efforts to treat cancer focused on the development of broad cytotoxic chemotherapy, which kills or impairs the growth of rapidly dividing cells. These drugs improve the treatment of certain cancers but are not universally effective. Moreover, cytotoxic chemotherapy is frequently intolerable, in part because it hurts both cancerous and normal cells [2]. To address these challenges, extensive research has focused on identifying therapeutic targets that selectively inhibit cancer cell proliferation. These efforts have established tyrosine kinases (TKs) as a promising class of molecules that can serve as key druggable targets [3,4]. TKs, including at least 58 receptor TKs and 32 non-receptor TKs, transfer a phosphate group from adenosine triphosphate (ATP) to the hydroxyl of tyrosine residues in substrate proteins, which results in signal transduction [3,5]. TKs are involved in signaling cascades that mediate physiological cellular functions, including proliferation and survival. Aberrant TK activity has been identified as a key driver of oncogenesis in numerous cancers, leading to the development of TK inhibitors (TKIs) as targeted therapeutic agents. Since the accelerated approval of imatinib, a TKI targeting the fusion protein BCR-ABL in chronic myeloid leukemia (CML), TKIs have dominated the landscape of anticancer drug development, with over 80 small-molecule TKIs approved by the FDA since 2001 [3]. While TKIs are designed to target unique kinases, most TKIs act similarly by competing with ATP at the binding pocket of the catalytic domain of the kinase, which prevents phosphorylation of downstream targets [6]. Despite intended selectivity, TKIs invariably inhibit kinases or other proteins beyond their intended targets, including drug-metabolizing enzymes and transporters (DMETs) [7,8]. Although TKIs have revolutionized the treatment of certain cancers, they share some of the same challenges as traditional cytotoxic chemotherapy, including the emergence of resistance and toxicity, in part due to a lack of selectivity [9–11]. Moreover, small-molecule TKIs present unique concerns because they have been developed as oral formulations. While oral delivery is preferred by patients and facilitates treatment adherence, the chronic use of orally administered TKIs results in a substantial risk of drug–drug interactions (DDIs) with concurrently administered medications [12–14]. What is a DDI? A DDI occurs when one drug (perpetrator) impacts another drug (victim). These interactions can be classified as pharmacokinetic, when a perpetrator affects the systemic concentrations of the victim drug, or pharmacodynamic, when a perpetrator influences the intended, or unintended, effect of the victim drug, or a combination of both [15]. Pharmacodynamic DDIs can be additive, synergistic, or antagonistic. Pharmacokinetic DDIs involve altered absorption, distribution, metabolism, or excretion (ADME) of the victim drug, often resulting from physicochemical changes, like adjusted stomach pH or gastric emptying times, and/or DMET induction or inhibition by the perpetrating drug [13,16–19]. Pharmacokinetic DDIs may lead to decreased victim drug concentrations, which could reduce efficacy, or increased victim drug concentrations, which could lead to toxicity [20]. In the context of TKIs, DDIs are particularly concerning due to substantial interpatient variability and a relatively narrow therapeutic index [21,22]. TKI DDI clinical implications Cancer patients are at a particularly high risk for DDIs due to polypharmacy, as they often require multiple medications for treatment, symptom management, and comorbid conditions. Patients taking TKIs often take more than four concomitant medications, with this number unsurprisingly serving as an independent predictor for potential DDI risk [23,24]. In particular, patients prescribed TKIs are at a high risk of DDIs. Retrospective studies estimate that more than 60% of patients receiving TKIs are exposed to at least one potentially significant DDI, with one report indicating that a DDI-related clinical effect was observed in up to 20% of patients taking TKIs [23,25,26]. Beyond this report, the clinical significance of DDIs specific to TKIs, or in general, remains largely unreported. However, recent studies that are not specific to TKIs report that DDIs are associated with an increased number of adverse drug events (ADEs) and a reduced health-related quality of life [27,28]. Similarly, the economic cost of DDIs is steep, as it is estimated that pharmacist-directed efforts to reduce DDIs can result in savings that exceed 30 times their cost [29]. While data specific to TKIs is limited, the clinical implications of their DDIs are at least as significant as those of other drugs, considering their narrow therapeutic index and the life-threatening conditions they treat. The ongoing development of new TKIs coupled with inconsistent, and at times contradictory, information regarding their potential for DDIs continues to complicate both our understanding and the systematic study of these interactions. In this review, we provide an overview of the mechanisms underlying both pharmacokinetic and pharmacodynamic DDIs with TKIs as perpetrators or victims, illustrated by examples from the FDA labels of TKIs approved since 2023 (Table 1), and build upon previous comprehensive reviews [14,30–46]. We have searched the Drug Interaction Database (DIDB), FDA drug labels, and PubMed to identify studies and data related to TKI DDIs, clinical impact, ADME, and PK/PD. However, an exhaustive review cataloging every reported TKI-related DDI is outside the scope of this review. Table 1.DDIs associated with TKIs approved between 2023 and 2025 as reported in FDA labeling.TKIReferencesTargetIndicationApproval dateIn vitro victim interactionsClinical victim InteractionsIn vitro perpetrator interactionsClinical perpetrator interactions Fruquintinib[33]VEGFRAdult metastatic colorectal cancer (mCRC) patients who have been previously treated with fluoropyrimidine-, oxaliplatin, and irinotecan-based chemotherapy2023No significant interactionCYP3A4# and ABCB1 induction: 65% AUC decrease with rifampinNo significant interactionNo significant interactionMomelotinib[34]JAK1, JAK2, JAK2 (V617F)Intermediate or high-risk myelofibrosis (MF), including primary MF or secondary MF in adults with anemia2023ABCB1, ABCG2, OATP1B1 and OATP1B3 substrateOATP1B1/1B3$ inhibition: 1.57-fold AUC increase with rifampinCYP2B6, UGT1A1, UGT1A9, and ABCG2 inhibitorABCG2 inhibition: 3.2-fold AUC increase of rosuvastatinPirtobrutinib[35]BTKAdult mantle cell lymphoma (MCL) after at least two lines of systemic therapy. Adult chronic lymphocytic leukemia or small lymphocytic lymphoma (CLL/SLL) who have received at least two prior lines of therapy2023ABCB1 and ABCG2 substrateCYP3A4$ and ABCB1 inhibition: 1.49-fold AUC increase with itraconazole.CYP3A4$ and ABCB1 induction: 71% AUC decrease with rifampinCYP2C8, CYP2C9, CYP3A, CYP1A2, CYP2B6, CYP2C19, CYP2D6, ABCB1 and ABCG2 inhibition%.CYP3A4, CYP3A5, CYP2B6 and CYP2C19 inductionCYP3A4 inhibition: 1.7-fold AUC increase of midazolam. CYP2C8 inhibition: 2.3-fold AUC increase of repaglinide. CYP2C19 inhibition: 1.56-fold AUC increase of omeprazole. ABCB1 inhibition: 1.35-fold AUC increase of digoxin. ABCG2 inhibition: 2.4-fold AUC increase of rosuvastatinQuizartinib[36]FLT3Adult newly diagnosed acute myeloid leukemia (AML) that is FLT3 internal tandem duplication (ITD)-positive2023ABCB1 substrateCYP3A4$ inhibition: 1.94-fold AUC increase with ketoconazole. CYP3A4 # induction: 90% AUC decrease with efavirenzNo significant interactionNo significant interactionRitlectinib[37]JAK3Severe alopecia areata in adults and adolescents 12 years and older2023No significant interactionCYP3A4# and ABCB1 induction: 56% AUC decrease with rifampinNo significant interactionCYP3A4$ inhibition: 2.69-fold AUC increase of midazolam. CYP1A2$ inhibition: 2.65-fold AUC increase of caffeine. OCT1 inhibition: 1.3 – 1.5-fold AUC increase of sumatriptanDeuruxolitinib[38]JAKAdult severe alopecia areata2024ABCB1 and ABCG2 substrateCYP3A4#and ABCB1# induction: 1.78-fold AUC decrease with rifampin. CYP2C9/CYP3A4# inhibition: 3-fold AUC increase with fluconazoleABCG2, ABCB11, OAT3, and MATE2-K inhibitorNo significant interactionEnsartinib[39]ALKAdult anaplastic lymphoma kinase (ALK)-positive locally advanced or metastatic non-small cell lung cancer (NSCLC) who have not previously received an ALK-inhibitor2024ABCB1 and CYP3A4 substrateNo significant interactionNo significant interactionNo significant interactionLazertinib[40]EGFRAdult locally advanced or metastatic non-small cell lung cancer (NSCLC) withh epidermal growth factor receptor (EGFR) exon 19 deletions or exon 21 L858R substitution mutations2024No significant interactionCYP3A4 and ABCB1 inhibition: 1.5-fold AUC increase with itraconazole. CYP3A4 and ABCB1 induction: 83% AUC decrease with rifampinCYP3A4, UGT12A1, ABCG2 and OCT1 inhibitorCYP3A4 inhibition: 1.5-fold AUC increase of midazolam. ABCG2 inhibition: 2-fold AUC increase of rosuvastatinTovorafenib[41]RAFRelapsed or refractory pediatric low-grade glioma (LGG) harboring a BRAF fusion or rearrangement, or BRAF V600 mutation2024CYP2C8# substrateNo significant interactionCYP2C8, CYP2C9, CYP2C19, CYP3A, and ABCG2 inhibitorCYP3A4 inhibition: 20% decreased AUC of midazolamTaletractinib[42]ROS1Adult locally advanced or metastatic ROS1-positive non-small cell lung cancer (NSCLC)2025No significant interactionCYP3A4# and ABCB1 inhibition: 3.9-fold AUC increase with itraconazole.CYP3A# and ABCB1 induction: 42% AUC decrease with rifampin. pH-dependent solubility#, 40% decreased AUC with omeprazoleNo significant interactionNo significant interactionAvutometinib[43]MEK1Adult KRAS-mutated recurrent low-grade serous ovarian cancer2025ABCB1 and ABCG2 substrateCYP3A4# induction: 34% AUC decrease with phenytoinNo significant interactionNo significant interactionDefactinib[44]FAK and PYK2Adult KRAS-mutated recurrent low-grade serous ovarian cancer2025ABCB1 and ABCG2 substrateCYP3A4#and ABCB1 inhibition: 3.9-fold AUC increase with itraconazole. CYP3A4# induction: 1.87-fold AUC decrease with phenytoin. pH-dependent solubility #, 79% AUC decrease with omeprazoleCYP3A4, CYP2C9, and UGT1A1 inhibitor. CYP2B6 and CYP1A2 inducer. ABCB1, ABCG2, OATP1B1, OATP1B3, and MATE2-K inhibitorNo significant interactionMirdametinib[45]MEK1/2Adult and pediatric neurofibromatosis type 1 patients with symptomatic plexiform neurofibromas not amenable to complete resection2025ABCB1 and ABCG2 substrateNo significant interactionNo significant interactionNo significant interactionVimseltinib[46]CSF1RAdult patients with symptomatic tenosynovial giant cell tumor (TGCT) for which surgical resection will potentially cause worsening functional limitation or severe morbidity2025ABCB1 substrateNo significant interactionABCB1#, ABCG2#, ABCB11, OATP1B1, OATP1B3, OCT2#, MATE1, and MATE2-K inhibitorABCB1# inhibition: Predicted 2- to 3-fold AUC increase with vimseltinibNo significant interaction: No significant interaction reported, or no data referenced#DDI informed label warnings: prohibit concomitant use$Recommend dose reduction or drug monitoring%refer to victim drug label. DDIs associated with TKIs approved between 2023 and 2025 as reported in FDA labeling. No significant interaction: No significant interaction reported, or no data referenced DDI informed label warnings: prohibit concomitant use Recommend dose reduction or drug monitoring refer to victim drug label. Pharmacokinetics, DDIs, and ADME Systemic exposure and distribution to the site of action are correlated with both treatment outcomes and the development of adverse effects with both TKIs and other drugs. Pharmacokinetics refers to how the body processes a drug and regulates drug exposure, which refers to the systemic or tissue concentration achieved over time. Pharmacokinetics is governed by the four key ADME processes [47]. TKIs both cause and are susceptible to DDIs mediated by changes in each of these ADME processes (Figure 1). Here, we describe each of the major ADME processes with illustrative examples of TKI DDIs as victim (i.e., substrate) or perpetrator (i.e., inhibitor, inducer), with an emphasis on outlining information about recently approved TKIs. Typical ADME pathways of an orally administered compound potentially modified by DDIs. Pink arrow: Potential physicochemical DDI. Purple arrow: Potential physicochemical and enterocyte transporter DDI. Blue arrow: Potential hepatic transporter DDI. Yellow starburst: Potential drug metabolizing enzyme DDI. Green arrow: Potential renal transporter DDI. Yellow arrow: Potential target distribution DDI. Created in BioRender. Eisenmann, E. (2025). https://BioRender.com/na2h8hn. Absorption Small-molecule TKIs are administered orally, which likely contributes to variability in TKI exposure [48]. Absorption is the process underlying entry into systemic circulation after oral administration and a significant determinant of oral bioavailability. To be absorbed, a solid dosage form must be dissolved and solubilized before being transported or diffusing across the gut lumen. While disintegration and solubilization mainly occur in the mouth and stomach, the transporter-mediated absorption of most drugs is believed to primarily occur in the small intestine due to its exceptionally high surface area. However, the stomach has been validated to contribute to the transporter-mediated absorption of dasatinib, emphasizing the notion that drug absorption is complex [49]. pH-dependent solubility, gastric emptying time, inhibition of uptake and efflux transporters, and inhibition or induction of enterocyte cytochrome p450 (CYP) enzyme CYP3A4 can also contribute to clinically relevant absorption DDIs [13,16]. Acid-reducing agents (ARA), used by 20–55% of cancer patients, may cause absorption-mediated DDIs with TKIs through changes in pH [13,50,51]. Changes in pH will impact the solubility of orally administered TKIs, especially when the pH is near the pKa of the TKI. Changes in pH affect the equilibrium between readily dissolved ionized forms and non-soluble unionized forms. The majority of TKIs are weakly basic, with increased ionization and solubility at lower pH levels. Co-administered ARAs, such as proton-pump inhibitors, H2-receptor antagonists, and antacids, elevate the pH in the stomach, from 1 to as high as 4, which may decrease the bioavailability of TKIs through decreasing solubility. For example, dasatinib, gefitinib, erlotinib, and nilotinib each have decreased exposure with increased gastric pH [50,52]. More recently, two TKIs approved in 2025, defactinib and taletractinib, have FDA labels warning of pH-dependent solubility (Table 1). GET is a physicochemical determinant of absorption. GET typically impacts biopharmaceutics classification system (BCS) class 1 drugs, drugs with high solubility and high permeability, by affecting the maximum plasma concentration (Cmax) and the time to maximum concentration (Tmax) with little impact on area under the concentration–time curve (AUC), a measure of drug exposure over time [53,54]. Most TKIs are BCS class 2 or 4 drugs, with low solubility and high permeability or low solubility and low permeability, suggesting they are less likely to be impacted by GET changes [53,55]. However, changes in GET induced by physical modifications (e.g., bariatric surgery) can impact TKIs, as seen with decreased absorption and efficacy of several TKIs, including nilotinib, dasatinib, bosutinib, ponatinib, and imatinib, in the treatment of CML or cabozantinib for the treatment of clear renal cell carcinoma [18,19]. While several drugs, including GLP-1 agonists, an increasingly important class of medications used for diabetes and weight loss, can modulate GET, a clear DDI study between GET effectors and TKIs has yet to be established [56]. Given the validated impact of physical changes on TKI disposition, we anticipate that drugs that interact with GET will also impact TKI disposition, and that our understanding of GET on TKI disposition will continue to expand with the increasing use of drugs modifying GET. Uptake and efflux transporters are major determinants of the absorption of TKIs. One key efflux transporter involved in the absorption of TKIs is ABCB1, which is also known as P-glycoprotein (P-gp). With ABCB1 inhibitors, like verapamil, the absorption of TKIs that are ABCB1 substrates will be increased, possibly resulting in toxicity. Many TKIs are ABCB1 substrates, including several from 2023 to 2025 (Table 1) and have potential ABCB1-DDI liabilities. For instance, co-administration of afatinib, an ABCB1 substrate, with oral verapamil increased afatinib systemic exposure, likely due to increased oral absorption and decreased biliary excretion [13,57]. This hypothesis was tested using intestinal perfusion studies, Caco-2 uptake, and MDR1-MDCKII permeability studies, which collectively suggested that verapamil increases afatinib accumulation by decreasing efflux back into the intestinal lumen [57]. Conversely, several TKIs inhibit ABCB1, acting as the perpetrator of DDIs and impacting the absorption of other co-medications. A key example of this phenomenon is lapatinib increasing the exposure of digoxin, a narrow therapeutic index drug, by 80% [13]. Beyond ABCB1, other drug uptake or efflux transporters can also alter TKI absorption. Organic anion transporting polypeptides (OATP) are solute carrier (SLC) transporters that frequently act as influx transporters. OATP2B1 (gene name, SLCO2B1) is highly expressed in enterocytes and has been shown to transport several TKIs, including erlotinib, and hydroxymethylglutaryl-coenzyme A-CoA (HMG-CoA) reductase inhibitors (statins), such as fluvastatin [58–60]. Conversely, TKIs may inhibit OATP2B1 and alter the pharmacokinetics of substrates, such as fluvastatin. Erlotinib has been shown to act as an OATP2B1 substrate and inhibitor, highlighting complexities associated with absorption [61]. OATP1A2 (SLCO1A2) is an additional member of the OATP family expressed in enterocytes, although expression is lower than that of OATP2B1 [62]. Substrates of OATP1A2 include select TKIs, such as imatinib, and other drugs, such as pravastatin and fexofenadine [63–65]. Before entering the hepatic vein, oral TKIs must survive metabolism by intestinal enzymes, especially CYP3A4. Given that the majority of TKIs are CYP3A4 substrates, intestinal CYP3A4 likely contributes significantly to determining bioavailability. However, clearly differentiating the activities of liver and intestinal CYP3A4 is difficult. One relatively selective inhibitor of intestinal CYP3A4 is grapefruit juice [66]. When administered with grapefruit juice, the AUC of sunitinib and nilotinib increased by 11% and 29%, respectively [67,68]. Furthermore, rifampin, dexamethasone, and phenobarbital induced jejunum CYP3A4 expression 5.2-, 3.5-, and 4.1-fold [69]. Both inhibition and induction of intestinal CYP3A4 activity likely regulate apparent bioavailability. The recent development of relevant preclinical models, such as conditional knockout models, will help more elegantly differentiate between intestinal and hepatic CYP3A4 activity [70–72]. Distribution Distribution is the process by which drugs move from systemic circulation into peripheral sites, including the intended target (e.g., tumor), sites of elimination (e.g., liver and kidney), and additional organs (e.g., potential sites mediating toxicity). Alterations in distribution will impact efficacy when a drug relies on adequate distribution to poorly penetrated organs. The brain is a key example where DDIs can profoundly impact the distribution of TKIs. Many TKIs, including lapatinib and imatinib, are restricted from penetrating the blood–brain barrier (BBB) in part due to efflux transporters, including ABCB1 and ABCG2, which is also known as breast cancer resistance protein (BCRP) [73,74]. Further highlighting ABCB1 and ABCG2’s relationship to BBB penetration, osimertinib, a third-generation EGFR TKI used in non-small cell lung cancer (NSCLC) with lower affinity for ABCB1 and ABCG2, has higher clinical activity in NSCLC with brain metastases compared to earlier generation EGFR TKIs (erlotinib, gefitinib, and afatinib) [75–77]. In preclinical models, the inhibition of these transporters by perpetrators, such as elacridar, has increased brain concentrations up to 40-fold [73]. This DDI could intentionally boost imatinib to increase brain penetration to improve treatment of malignant gliomas, or could impact potential toxicity within the central nervous system [78–82]. As perpetrators, TKIs may inhibit ABCB1 and increase brain penetration of ABCB1 substrates. Several TKIs, such as erlotinib, nilotinib, dabrafenib, axitinib, crizotinib, dasatinib, imatinib, lapatinib, and vemurafenib, inhibit ABCB1 in vitro, which could increase BBB penetration of substrates [83]. SLC transporters, including OATPs, organic anion transporters (OATs), and organic cation transporters (OCTs), facilitate the distribution of endogenous and xenobiotic compounds throughout the body; differences in the abundance and expression of SLCs in different tissues govern this distribution and whether therapeutics can reach their site of action [84]. Statins rely on transport by OATP1B1 (SLCO1B1) and OATP1B3 (SLCO1B3) to enter the liver to inhibit HMG-CoA reductase. Decreases in OATP1B function by polymorphisms or inhibition (e.g., by TKIs) have been shown to prevent liver accumulation and decrease statin efficacy [54,85–87]. Distribution into the liver also plays an important role in metabolism [88]. Most TKIs are hepatically metabolized, thus OATPs, among other hepatic transporters, often play a rate-limiting role in drug metabolism [31,89]. Inhibition of liver transport will indirectly decrease metabolism by preventing exposure to metabolic enzymes in the liver. Nucleoside transporters, including equilibrative nucleoside transporter 1 (ENT1; SLC29A1) and 2 (ENT2; SLC29A2), are relevant to the transport of nucleoside analogs, which are frequently combined with TKIs. Accumulating evidence suggests that these transporters are involved in DDI, including a report that 12 of 24 studied TKIs are moderate to potent inhibitors of ENT1, with lorlatinib being the most potent inhibitor [90]. Collectively, these results suggest that TKIs could impact the distribution and activity of nucleoside analogs, despite the notion that these transporters are not currently recommended for investigation by regulatory agencies. Metabolism Metabolism involves the transformation of a drug by phase 1 or phase 2 enzymes mediated by the liver or other organs. Phase 1 enzymes mediate hydroxylation and facilitate oxidation, reduction, and hydrolysis reactions. Phase 2 enzymes mediate conjugation reactions. Conjugations include glucuronic acid conjugation, sulfate conjugation, amino acid conjugation, glutathione conjugation, acetylation, and methylation. TKIs interact with both phase 1 and 2 enzymes as substrates and perpetrators. Increases in TKI concentrations from inhibition of metabolism may result in increased toxicities [91]. Induction of metabolism could reduce TKI exposure and reduce treatment efficacy. Drug metabolizing enzymes (DMEs), such as CYP3A4, 3A5, 2C8, 2C9, 1A2, 2D6, and UDP-glucuronosyltransferase family 1 member A1 (UGT1A1), contribute to the metabolism of TKIs. Among them, CYP3A4 is the most relevant contributor to pharmacokinetic DDIs and altered metabolism [31,92–96]. Among DMET interactions, CYP3A-mediated interactions are most commonly responsible for TKI DDI prescribing warnings (Table 1). Nearly all TKIs, and nearly 50% of all approved drugs, are metabolized by CYP3A4 to some extent, which has informed FDA and EMA guidelines requiring the characterization of CYP3A4 interactions [13,31,97–100]. Among kinase inhibitors approved through 2018, 41 of 43 were metabolized, to some extent, by CYP3A4. More recently, of the 42 small-molecule kinase inhibitors approved from 2018 to 2023, 35 were TKIs, 22 of which were moderately to severely impacted by strong CYP3A4 inhibition or induction [32]. From 2023 to 2025, 14 TKIs were approved, 9 of which are substrates for CYP3A4 (Table 1). Notably, ensartinib is primarily metabolized by CYP3A4 and is speculated to be impacted by CYP3A4 inhibition, but clinical studies are not currently referenced in the FDA label [39]. Induction of metabolism also poses distinct risks. Treatment efficacy could be decreased with lower plasma concentrations. For example, plasma concentrations of imatinib are predictive of treatment efficacy in CML and gastrointestinal stromal tumors, with non-responders having significantly lower plasma levels [101,102]. This implies that metabolic inducers will reduce treatment efficacy by reducing plasma concentrations. Less intuitively, induction may reduce exposure to an active metabolite. For example, clopidogrel is an antiplatelet medication that relies on the CYP2C19 enzyme to produce its active metabolite and be effective [103]. Induction of metabolism could hypothetically impact TKIs with a substantial active metabolite (e.g., acalabrutinib) [104]. Interestingly, CYP3A4 and CYP2C8 induction have also been linked to hepatotoxicity [47]. TKI-induced hepatotoxicity is likely caused, in part, by the bioactivation of small-molecule kinase inhibitors. For example, dexamethasone-induced CYP3A4 induction increases lapatinib-induced hepatotoxicity in breast cancer patients [105]. In addition to being victims, TKIs perpetrate DDIs mediated by metabolism. Among kinase inhibitors approved up to 2018, of the 35 TKIs, 8 inhibited and 4 induced CYP3A4 activity. Dasatinib, gilterintib, lapatinib, pazopanib, and larotrectinib showed weak inhibition, while crizotinib, imatinib, and nilotinib exhibited strong inhibition. Vemurafenib, erlotinib, dabrafenib, and lorlatinib were weak to moderate inducers [92]. For TKIs approved between 2018 and 2023, nine were identified as time-dependent CYP3A4 inhibitors, including encorafenib, erdafitinib, avapritinib, pralsetinib, tucatinib, pemigatinib, infigratinib, pacritinib, which also inhibits CYP1A2, and lorlatinib, which may inhibit CYP2C9 and CYP1A2 [32,106]. Of the 14 TKIs approved from 2023 to 2025, four include CYP3A4 perpetrator data. Lazertinib, tovorafenib, pirtobrutinib, and ritlectinib reference varying degrees of CYP3A4 inhibition [35,37,40,41]. In addition to CYP3A4, pirtobrutinib inhibits CYP2C8 and CYP2C19, and ritlectinib inhibits CYP1A2 (Table 1). Notably, time-dependent shifts in metabolism pathways can alter DDI liabilities. For instance, imatinib single-dose administration with the inducers rifampin or St. John’s wort reduced AUC by 74% and 30% [107,108], while co-administration with ketoconazole increased AUC by 40% [109]. However, a separate study demonstrated that at steady state, imatinib is insensitive to CYP3A4 inhibition by ritonavir. Subsequent investigation revealed that CYP3A4 inhibition with human liver microsomes had little effect, suggesting that other enzymes may be involved in imatinib metabolism. The authors also showed that imatinib itself is a potent inhibitor of CYP3A4 in vitro [110]. Follow-up studies expanded on this idea and demonstrated that, after initial treatment with imatinib, CYP3A4 and CYP2C8 mediate 60% and 40% of hepatic clearance, respectively [111]. Notably, imatinib is a mechanism-based irreversible CYP3A4 inhibitor. After multiple doses at clinically relevant concentrations, imatinib has been predicted to inhibit up to 90% of hepatic CYP3A4, which shunts metabolism from CYP3A4 to CYP2C8, and hepatic clearance shifts to 25% and 65–75% by CYP3A4 and CYP2C8 [111,112]. In physiologically based pharmacokinetic (PBPK) simulations, one 400 mg dose of imatinib with a CYP3A4 (itraconazole) or CYP2C8 (gemfibrozil) inhibitor resulted in a predicted 40% and 80% AUC increase; however, after multiple daily 400 mg doses of imatinib, AUC increases were predicted to be 20% and 130%, respectively. The combination of itraconazole and gemfibrozil was predicted to cause a 200% AUC increase [111]. Time-dependent shifts in key metabolism pathways introduce additional complexities when predicting metabolism DDIs [110]. Auto-inhibition or complex elimination may warrant non-conventional multi-dose DDI study designs with inhibitor combinations. Notably, observed pharmacokinetic changes induced by combination therapy may differ from predictions, including those specified in the drug labeling. A notable example is osimertinib, which is now approved for first-line treatment in EGFR-mutant non-small cell lung cancer (NSCLC) when combined with platinum-based chemotherapy. Despite a label warning against coadministration with CYP3A4 modulators, osimertinib pharmacokinetics are not substantially impacted by coadministration with cisplatin, which has been suggested to induce CYP3A4 [113–116]. Population pharmacokinetic analyses from osimertinib trials demonstrated that cisplatin does not significantly alter osimertinib exposure, emphasizing the importance of validating predicted DDIs in the context of combination therapies. Nonetheless, despite the notion that preclinical or label-based DDI concerns may not always translate into clinically meaningful interactions, understanding the factors that influence drug disposition remains important for guiding combination therapy decisions and preventing potential DDIs. Excretion Excretion involves the removal of the parent drug and its metabolites from the body, primarily via hepatic and renal elimination [117]. Drug transport plays a significant role in the distribution and elimination of endogenous compounds and therapeutics, including TKIs [88,118]. TKIs interact with drug transporters as both victims and perpetrators. As of 2024, the FDA recommends investigating a drug’s substrate specificity or inhibition of ABCB1, ABCG2, OATP1B1, OATP1B3, OAT1 (SLC22A6), OAT3 (SLC22A8), OCT2 (SLC22A2), MATE1 (SLC47A1), and MATE2-K (SLC47A2), while the EMA recommends the same list with the addition of OCT1 (SLC22A1) and ABCB11 [99,100]. We will discuss these interactions by grouping transporters with similar expression localization, substrate specificity, and DDI liability. TKIs interact with efflux transporters, including ABCB1, ABCG2, and ABCC4, which is also known as multidrug resistance-associated protein 4 (MRP4). Inhibition of efflux transport will increase exposure to TKIs (e.g., imatinib and dasatinib) that are substrates for efflux transporters [119]. TKIs may also inhibit efflux transporters. A study from 2016 determined that 16 TKIs inhibited ABCC4, and 15 of 16 TKIs tested inhibited ABCG2 [120]. The authors argue that ABCC4 inhibition, confounded with ABCB11, will increase patients’ risk for cholestatic liver injury. Of TKIs approved between 2023 and 2025, three are in vitro ABCB1 substrates, six are in vitro ABCB1 and ABCG2 substrates, and three FDA labels lack in vitro data but are impacted by potential ABCB1 inhibition and induction (Table 1). TKIs may interact with hepatic transporters such as OATP-family transporters. In 2024, a systematic evaluation identified 13 of 62 FDA-approved TKIs as substrates of OATP1B1 [121]. For TKIs approved between 2023 and 2025, only momelotinib was identified as an OATP1B1 and OATP1B3 substrate (Table 1). Conversely, several TKIs inhibit OATP1B1 and OATP1B3 [122]. Inhibition of these transporters could reduce statin clearance and increase the risk of statin-associated myopathy and rhabdomyolysis [85–87]. OCT1 exhibits broad substrate specificity but plays a limited role in TKI DDIs. Few TKIs are known OCT1 substrates or inhibitors, suggesting minimal DDI potential via OCT1 interactions. OCT1 inhibitors such as quinidine, ranitidine, and midazolam slightly increase imatinib exposure [119,123]. Although OCT1 may contribute to imatinib exposure, CYP3A4 and ABCB1 have a more clinically relevant impact [119]. Of TKIs approved from 2023 to 2025, lazertinib and ritlectinib are identified OCT1 substrates [37,40]. Ritlectinib also increased the exposure of sumatriptan, another OCT1 substrate; however, this interaction was not considered clinically significant [124]. Transporters primarily localized to the kidney include OAT1, OAT3, OCT2, MATE1, and MATE2-K. Renal excretion of the TKIs is uncommon, as is transport of the TKIs by these primarily renal transporters. As perpetrators, in vitro the TKIs imatinib, dasatinib, gefitinib, erlotinib, and sunitinib inhibit OCT2, MATE1, and MATE2-K while nilotinib inhibits MATE1 and MATE2-K but showed low activity against OCT2 [125]. Despite this, there is limited clinical evidence of OCT2, MATE1, and MATE2-K kidney-related TKI–perpetrator interactions. The most severe example being tucatinib, a breast cancer HER2 inhibitor and potential OCT2, MATE1, and MATE2-K inhibitor [126]. Co-administration of tucatinib with metformin, an OCT2, MATE1, and MATE2-K substrate, induced a 40% AUC increase [127]. This potential DDI did not warrant any related FDA DDI label warnings [126]. Notably, simulations with nilotinib and the OAT3 substrate rivaroxaban predicted AUC increases of 52%. The AUC increase induced by nilotinib was predicted to significantly increase rivaroxaban’s dose-related bleeding risk and justified dose reduction recommendations [128,129]. While current clinical evidence is limited, TKIs may have a limited potential to induce clinically relevant renal DDIs. Mechanisms of ADME modification DMETs are modulated by mechanisms including direct binding, post-translational modifications, and transcriptional regulation. Direct binding to a DMET site of action results in competitive inhibition through steric hindrance. Several TKIs bind directly to CYP3A4, including ceritinib, imatinib, lapatinib, nilotinib, and osimertinib [92]. Post-translational modifications (PTMs) regulate protein structure, function, and availability to interact with the cellular environment. The most common PTMs include phosphorylation, glycosylation, and ubiquitination [130]. Recent work has illuminated the potential for TKIs to modify PTMs and impact transporter function; notable examples include the modification of OATP function through interaction with LYN kinase and OCT1 and OCT2 function through inhibition of Yes1 [131–135]. Ongoing research seeks to characterize the role of additional PTMs in transporter-mediated disposition and excretion [136]. PTMs also regulate DMEs, including CYP3A4, which TKIs could possibly impact [137,138]. Transcriptional regulation (TR) is a complex biological process that dictates gene expression. Transcription factors are a group of diverse proteins responsible for regulating transcription by binding to enhancers and recruiting necessary cofactors. Transcription factors, including constitutive androstane receptor (CAR), pregnane X receptor (PXR), PPARα, LXR, FXR, GR, HNF-1α, HNF-4α, and NF-κβ have been established to impact the expression of DMETs, with CAR and PXR the most relevant for DDIs [139–141]. PXR and CAR are well established to activate CYP3A4, CYP2B6, CYP2C8, CYP2C9, and CYP2C19, along with GST, UGT, and SULT [140–147]. PXR and CAR share significant drug transporter targets, such as OATP1B2, the murine ortholog of OATP1B1 and OATP1B3, ABCB11, NTCP (SLC10A1), ABCB4, ABCC3, and ABCC2 in mouse and rat, and ABCC1 and ABCC2 in humans [148–155]. Xenobiotics, including dexamethasone, mifepristone, and rifampin, may activate PXR and induce CYP3A4 and ABCB1 [141,156]. Induction of ABCB1 and CYP3A4 by rifampin decreases exposure to the substrates glecaprevir and pibrentasvir [157]. The TKIs erlotinib, gefitinib, nilotinib, sorafenib, and vandetanib induce ABCB1 expression through PXR induction [156]. Tovorafenib, which was approved in 2024, induces CYP3A4 expression, likely through PXR and/or CAR induction [41]. Pharmacodynamic DDIs Pharmacodynamic DDIs occur when the physiological effects of one drug are affected by another [158]. Pharmacodynamic DDIs are classified as additive, synergistic, and antagonistic interactions. Despite their clinical significance, pharmacodynamic DDIs are investigated less frequently than pharmacokinetic DDIs, likely because they lack a formal evaluation framework [159]. Most pharmacodynamic DDI research is limited to in vitro screenings, with fewer in vivo or clinical investigations, resulting in a lack of robust models and methodologies to evaluate these interactions in clinically relevant settings. This gap highlights the need for more comprehensive approaches, such as mechanism-based and quantitative systems pharmacology (QSP) modeling, to better predict, interpret, and guide pharmacodynamic DDI studies for research, regulatory, and clinical purposes [160]. In oncology, pharmacodynamic interaction studies have been crucial for designing both sequential and concurrent regimens effectively. For example, sequential regimens involving cell cycle dependent chemotherapies with taxane pre-treatment (e.g., paclitaxel or docetaxel) can reduce platinum agent (e.g., cisplatin) accumulation and platinum-DNA adduct formation in leukocytes, potentially lowering both efficacy and hematopoietic toxicity [161,162]. While paclitaxel/carboplatin combinations reduce toxicity, tumor responses may decline when docetaxel precedes carboplatin, as taxane-induced M-phase arrest interferes with platinum-induced S-phase cytotoxicity [163]. Conversely, concurrent administration was found to enhance carboplatin-DNA adduct formation in urothelial carcinoma cells [164]. Similar complexities are observed when TKIs are incorporated into chemotherapy regimens. TKIs can induce cell cycle arrest at specific phases (often G0/G1), modulate apoptotic signaling pathways, and alter transporter expression or activity, all of which influence chemotherapy sensitivity. For instance, EGFR and VEGFR TKIs have been shown to reduce the efficacy of antimitotic agents when administered prior to chemotherapy, likely due to reduced proliferation rates and diminished entry into vulnerable cell cycle phases. A meta-analysis indicated that the sequence of EGFR TKI and chemotherapy impacts outcomes. The Tarceva OR CHemotherapy (TORCH) study demonstrated that initiating treatment with EGFR TKIs (e.g., erlotinib or gefitinib) followed by chemotherapy upon progression is linked to inferior overall survival compared to the reverse sequence, particularly in unselected populations [165,166]. These data underscore the complex interplay of pharmacodynamic effects, such as cell cycle modulation and downstream signaling inhibition by TKIs that affect chemotherapy sensitivity. These findings highlight the importance of mechanism-driven pharmacodynamic drug–drug interaction studies and integrated PK/PD and PBPK modeling in clarifying seemingly conflicting outcomes across various treatment schedules and biological contexts. Additive interactions Additivity is defined as a pharmacodynamic interaction in which the combined effect of two or more drugs is equal to the sum of the individual effects produced by each agent when administered separately [160]. TKIs are frequently combined with other chemotherapeutics, including other TKIs [167]. These combination regimens often lead to additive interactions with boosted efficacy or toxicity. For example, combination regimens of erlotinib, a TKI that inhibits EGFR, and bevacizumab showed boosted antitumor efficacy in advanced non-small-cell lung cancer (NSCLC) with EGFR+ mutations [168]. Similarly, combining a type I TKI, PLX9486, which targets the active conformation of the kinase, with a type II TKI, sunitinib, which binds the inactive conformation, resulted in enhanced efficacy [169]. A similar phenomenon is observed in the combination of crenolanib and sorafenib, which achieves synergy by simultaneously and durably inhibiting downstream FLT3 signaling pathways, specifically p-STAT5 and p-ERK, for up to 8 hours [170]. This dual blockade not only enhances antileukemic activity but also reduces the emergence of resistance-driving FLT3-KD mutations compared with either agent alone. These approaches leverage the complementary mechanisms of action of the two drugs, targeting different conformational states of the KIT protein, and produce greater tumor inhibition than either agent alone. Additive interactions can occur through restoration of drug efficacy in resistant models. For example, lapatinib, a dual tyrosine kinase inhibitor targeting HER1 (EGFR) and HER2, interacts with ATP-binding cassette (ABC) transporters. In vitro studies demonstrated that lapatinib significantly enhanced the sensitivity of ABCB1- and ABCG2-overexpressing cells to conventional chemotherapeutic agents, including doxorubicin and mitoxantrone, through increasing intracellular accumulation. Consistent with these results, lapatinib potentiated the antitumor activity of paclitaxel in ABCB1-overexpressing xenograft models in vivo [171]. Importantly, efflux transporters impact both the intracellular accumulation and downstream activity of xenobiotics (PD effect) and the distribution into different organs (PK effect). Additive toxicity can also occur when combining TKIs with additional agents [172]. For example, combining multiple TKIs targeting EGFR can increase the risk of skin rashes, as observed in clinical trials combining dasatinib and erlotinib [173,174]. Synergistic interactions Synergy refers to a pharmacological interaction in which the combined effect of two drugs exceeds the sum of their individual effects, indicating an enhanced therapeutic outcome beyond simple additivity [160]. An example of this is the combination of the antimetabolite cladribine with TKIs such as gefitinib or dasatinib, which produces a synergistic inhibitory effect on breast cancer (MCF-7) cells by enhancing G2/M cell cycle arrest, reactive oxygen species production, mitochondrial membrane potential loss, and apoptosis through mitochondrial pathways [175]. Additionally, a sequential regimen with erlotinib (EGFR TKI) followed by doxorubicin sensitizes a subset of triple-negative breast cancer cells to genotoxic drugs by rewiring oncogenic signaling pathways, yielding synergistic inhibition of cell growth [176]. Synergism has also been observed with BTK inhibitors combined with the BCL-2 inhibitor venetoclax. Venetoclax enhances efficacy in preclinical models by targeting complementary apoptotic pathways [177]. However, synergistic interactions can also result in increased toxicity, especially when drugs with overlapping or cumulative toxicological profiles are combined. One example is the combination of TKIs with immune checkpoint inhibitors (ICIs). ICIs are associated with a markedly increased risk of interstitial lung disease [166]. ILD is a serious and potentially fatal pulmonary toxicity. For instance, combination treatment with osimertinib and durvalumab resulted in high rates of grade 3–4 ILD, prompting early termination of clinical trials due to safety concerns. The incidence of ILD was significantly higher in patients receiving combination therapy (25.7%) compared to those on TKI monotherapy (4.59%) in the context of EGFR-mutant NSCLC. Furthermore, sequential administration of ICIs followed closely by TKIs – such as nivolumab followed by osimertinib within 1 month – has also been shown to significantly increase the risk of ILD [178]. Synergistic hepatotoxicity has similarly been reported in trials combining TKIs with ICIs or with other TKIs, often resulting in elevated liver enzymes and liver injury severe enough to warrant treatment discontinuation. For example, the combination of durvalumab and gefitinib led to grade 3–4 hepatic toxicity requiring cessation of therapy in several patients. In the CheckMate 370 trial, five out of thirteen patients receiving the combination of nivolumab and crizotinib developed grade 3–4 hepatic toxicity, including two fatal events. Hepatic toxicity was also found to be dose-dependent in studies evaluating ceritinib in combination with nivolumab, necessitating dose reductions to manage adverse effects [178]. Antagonistic interactions Antagonism refers to a pharmacological interaction in which the combined effect of two drugs is less than the sum of their individual effects, indicating a reduction in therapeutic efficacy when used together [160]. Antagonistic pharmacodynamic interactions are generally undesirable, particularly in the context of cancer therapy, as they can diminish the overall efficacy of treatment regimens. Such interactions not only compromise therapeutic outcomes but also introduce additional risks and complexity in an already vulnerable patient population facing life-threatening disease. An example of an antagonistic interaction is observed with concurrent administration of EGFR-targeting TKIs and cell cycle phase-dependent chemotherapeutic agents due to TKI-induced G1-phase cell cycle arrest [179]. This arrest renders chemotherapeutic agents that act on later phases of the cell cycle, such as the S- or M-phase, less effective [180,181]. Conclusion TKIs are subject to high interpatient variability, narrow therapeutic windows, and inconsistent outcomes. As DDIs contribute to the variability of oral TKIs and small-molecule therapeutics, there is a continued need to understand PK and PD DDI liabilities. DDI mechanisms are often complex, in part due to their occurrence at various ADME stages or through unique PD mechanisms. Our understanding of DDIs, in part gleaned by the development of more elegant methods and tools, will accelerate drug development while saving time and money associated with clinical failures. We are optimistic that continued collaborative efforts to unravel complex DDIs associated with TKIs promise a safer future for patients treated with TKIs. Expert opinion The TKI portfolio has been rapidly developed since imatinib was introduced in 2001. As their clinical applications and combination regimens continue to expand, it is essential to understand potential DDIs and their impact on efficacy and toxicity. Oral TKIs often exhibit unpredictable pharmacokinetics and variable responses. A better understanding of DDIs will help avoid deleterious DDIs and increase effectiveness. Evolving strategies to characterize DDIs include preclinical models, PBPK modeling, and monitoring strategies, including therapeutic drug monitoring (TDM) and DMET probes. Preclinical models have provided valuable insight into the mechanisms underlying the disposition and interaction liability of TKIs. Preclinical models offer several advantages when compared to clinical trials, such as reduced cost, time, and risk to patients. Robust and comprehensive preclinical analysis of DDI liability is critical to the characterization of new molecular entities, to both inform clinical trial decisions and the development of a new drug application [182,183]. These analyses involve both in vitro analyses (e.g., microsomal assays, transporter inhibition assays, and substrate assays) and in vivo analyses (e.g., toxicity and pharmacokinetics) to inform DDI victim and perpetrator studies and determine starting doses [184–187]. While this landscape is quickly shifting in favor of new alternative methods (NaMs), including elegant computational models and human organoid systems, traditional preclinical models remain important for assessing the safety and DDI liability of small-molecule drugs, including TKIs [188,189]. Mechanistic static models have demonstrated success in predicting the magnitude of DDIs with DMETs. These models utilize in vitro hepatocytes and human liver microsomes to perform in vitro-in vivo correlations (IVIVC) [190,191]. Further efforts are being made to optimize IVIVC methods and quantify the predicted AUC ratio with and without inhibitors (AUCR) [16,192]. Additionally, given the promiscuity of ‘selective’ prototypical substrates and inhibitors, genetic knockout models have been invaluable to selectively determining the impact of specific DMETs in the disposition of TKIs and characterizing the DMETs involved in specific DDIs [193]. For example, our group has used mice with a deficiency of CYP3A to demonstrate that cobicistat impacts ibrutinib pharmacokinetics almost exclusively through this pathway [194]. These mechanistic insights helped to inform computational modeling to determine a human starting dose of ibrutinib that, when combined with cobicistat, would achieve similar exposure to clinically approved doses. In contrast to these results, we separately used these mice to determine that itraconazole is only partially dependent on CYP3A to increase gilteritinib exposure [195]. These results implicated OATP1B-type transporters in the disposition of gilteritinib, which would have been difficult without preclinical models. Collectively, preclinical models remain helpful in characterizing the DDI liability of TKIs, including through informing the development of computational models. Computational modeling is valuable to predict the DDI liability of TKIs. PBPK modeling utilizes known or determined physiological parameters and chemical features to simulate ADME processes. PBPK modeling has been invaluable for characterizing the pharmacokinetics of TKIs and for informing first-in-human doses [196]. This approach allows researchers to conduct experiments with higher throughput and/or investigate scenarios that would not be clinically feasible otherwise. PBPK modeling has been used to predict appropriate doses in underrepresented populations, such as pediatric patients [197]. More recently, this approach has been used to accurately predict the tissue distribution of osimertinib [198,199]. It is anticipated that the role of computational modeling in predicting TKI DDI liability will continue to expand, especially with the continued development of artificial intelligence (AI) and machine learning (ML). These methods will complement the data obtained from carefully monitoring TKI exposure or DMET activity. To better understand the disposition of TKIs and their DDI liability, we advocate for the analysis of TKI pharmacokinetics and the continued development of DMET endogenous or exogenous probes. Analyzing the pharmacokinetics and DDI liability of TKIs is an integral component of the drug approval process [200]. We advocate for further analysis of TKI pharmacokinetics beyond the approval requirements. For example, pharmacokinetics should be analyzed in the context of new combination regimens or in optimizing TKI dosing [201,202]. Therapeutic drug monitoring (TDM) is a potential tool to inform dose modulation strategies and achieve more personalized medicine with TKIs, as we have advocated previously [203–205]. However, the adoption of TDM is currently limited by the lack of a well-defined therapeutic window for most TKIs and inadequate infrastructure available to monitor drug exposure [206,207]. The continued integration of pharmacokinetics in ongoing clinical trials studying TKIs, especially in combination regimens, will further our understanding of TKI DDIs and pharmacokinetics, including the therapeutic window. For example, we have previously utilized sorafenib pharmacokinetic analyses and computational modeling to determine that higher exposure is associated with greater toxicity and that sorafenib is sensitive to OATP1B1 inhibition [208]. These data imply that caution should be exercised when combining sorafenib with drugs that inhibit OATP1B1. Despite the value offered by these analyses, we appreciate that this approach requires elegant LC-MS/MS analyses and complicated computational analyses. To more practically inform dosing adjustments, we advocate for the continued development of endogenous and exogenous probes to characterize the DDI liability of drugs during development. For example, our group has characterized chenodeoxycholate-24-glucuronide (CDCA-24 G) as a selective endogenous probe to determine whether drugs inhibit OATP1B-type transport [209]. Recently, a 2024 article summarized efforts to establish various endogenous biomarkers for OATP1B1, OATP1B3, OCT1, OAT1, OAT3, OCT2, MATE1, and MATE-2K and established the tier 1 and tier 2 biomarker designations [210]. Tier 1 options are considered fully validated biomarkers, while tier 2 options are not entirely validated but could be useful to support DDI risk assessment. Coproporphyrin I (CPI) was identified as an effective probe of OATP1B-type transport [211]. CPI was considered fully validated and the only tier 1 biomarker [210]. As a notable addition, recent analysis of genomic and metabolomic databases identified three potential ABCG2 probes, 4-hydroxychlorothalonil (4HC), 3-bromo-5-chloro-2,6-dihydroxybenzoic acid (BCDBA), and 3,5-dichloro-2,6-dihydroxybenzoic acid (DCDBA) [212]. These advancements allow for endogenous DDI biomarkers to be monitored in first-in-human studies to help de-risk or inform whether DDI liabilities exist and warrant a dedicated DDI study [210]. In addition to endogenous probes, exogenous probes are being developed for clinical use. Pitavastatin is a promising exogenous OATP1B probe that offers another safe option for DDI studies assessing OATP1B function [213]. Novel approaches, such as metabolomics-based biomarker discovery approach, provide promising avenues for further probe discovery, as seen with the identification of novel OAT1 and OAT3 biomarkers [214]. Establishing probes for all relevant DMETs will further our understanding of DDI liability, informing dosing adjustments that could reduce variability and improve patient outcomes. Article highlights Polypharmacy is growing in prominence and complexity, increasing the risk of DDIs that impact clinical outcomes. Polypharmacy is growing in prominence and complexity, increasing the risk of DDIs that impact clinical outcomes. Orally administered TKIs offer significant survival benefits but are often accompanied by substantial DDI risk. Orally administered TKIs offer significant survival benefits but are often accompanied by substantial DDI risk. TKI-related pharmacokinetic interactions, both as victim and perpetrator, impact various stages of typical ADME processes that may significantly affect exposure. TKI-related pharmacokinetic interactions, both as victim and perpetrator, impact various stages of typical ADME processes that may significantly affect exposure. TKI-related additive, synergistic, and antagonistic pharmacodynamic DDIs further complicate TKI DDI liabilities. TKI-related additive, synergistic, and antagonistic pharmacodynamic DDIs further complicate TKI DDI liabilities. Understanding the factors underlying the disposition of TKIs will facilitate informed dose modifications and treatment strategies to avoid unfavorable DDIs. Understanding the factors underlying the disposition of TKIs will facilitate informed dose modifications and treatment strategies to avoid unfavorable DDIs. Funding The manuscript was funded by the National Institutes of Health Grants [R37CA299879] (to E. Eisenmann), [R01HL168045] (to A. Sparreboom and S. Baker), and [R01CA272254] (to S. Hu and S. Baker); startup funding from the OSU College of Pharmacy (to E. Eisenmann), a Research and Innovation Faculty Startup Award from the OSU Enterprise for Research, Innovation and Knowledge, [Research and Innovation Faculty Startup Award] (to E. Eisenmann), and the OSU Comprehensive Cancer Center Pelotonia foundation (to S. Baker, E. Eisenmann, S. Hu, and A. Sparreboom). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. The content is solely the responsibility of the authors and does not necessarily represent the official views of the funding agencies, including the National Institutes of Health. Footnotes Declaration of interests The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties. Reviewer disclosures A peer reviewer has been on the Advisory Board for EMD Serono, BMS, Merck, Seattle Genetics, Astellas, Janssen, Bicycle Therapeutics, Pfizer, Gilead, Scholar Rock, Eli Lilly, Loxo Oncology, Vial, Aktis, Daicchi-Sankyo and GSK; a consultant/Scientific Advisory Board (SAB)/trial steering committee for Syapse, Merck, Servier, Ellipses and Biosite Research; received research support from EMD Serono, Jazz Therapeutics, Bayer, Sumitomo Pharma, Blue Earth Diagnostics and Exelixis; a speaker for Seagen, Gilead, Exelixis, Janssen, Astellas, Bayer, Aveo, Pfizer, Merck and Astrazeneca; and received travel fees from BMS and Astellas. This reviewer’s spouse is also employed by Myriad, Exact Sciences. Peer reviewers on this manuscript have no other relevant financial relationships or otherwise to disclose. References Papers of special note have been highlighted as either of interest (•) or of considerable interest (••) to readers.