Impact of Physiologic Conditions on CYP3A4-Activated/Deactivated Anticancer Drugs

With respect to nongenetic factors, a series of studies have reported a sex difference in CYP3A4 activity. It has been confirmed that CYP3A4 activity is higher in women than in men by analyzing plasma 4β-hydroxycholesterol (an endogenous CYP3A marker) in subjects from different populations (Diczfalusy et al., 2011). The protein and mRNA levels as well as the activity of CYP3A4 in HLMs and hepatocytes were higher in females than in males (Liu et al., 2021). Conversely, studies in HLM from diabetes and nonalcoholic fatty liver disease donors suggested an insignificant sex difference in CYP3A4 activity and protein expression (Jamwal et al., 2018). As a result, the effects of sex on CYP3A4 metabolic activity under various conditions need to be further elucidated.

There are significant age differences in the activity and expression of P450s due to age and age-related factors (e.g., height, weight, body surface area, body fat, serum albumin, and total body water), which potentially affect intrinsic drug metabolism and disposition (Maagdenberg et al., 2018; Liu et al., 2021). In children aged from 1 to 18 years, the activity of CYP3A4 per gram of liver decreased slightly with age, as measured by a physiologic population pharmacokinetic modeling method, indicating that the activity of CYP3A4 is affected by age (Brussee et al., 2018). However, no age-related differences in CYP3A4 activity were observed when comparing samples in HLMs and hepatocytes from the 20- to 60-year-old age group with the over 60-year-old age group (Parkinson et al., 2004; Jamwal et al., 2018).

Although information on the physiologic conditions for CYP3A4-mediated metabolism appears to be easily obtained for guiding clinicians in selecting appropriate dose adjustments, in the majority of studies, the effect of sex and age on CYP3A4 metabolic activity remains to be verified in a subpopulation of specific cancers, which may have implications for cancer therapy. In most cases, the fact that cancer patients usually have other diseases and confounding factors makes it difficult to obtain specific data on the CYP3A4 profile of these patients.

Impact of Tumor-Associated Inflammation on CYP3A4 and Drug Responses

For decades, inflammation has been considered a pivotal factor for interindividual variability of CYP3A4 metabolism, which should be taken into account in drug development and clinical practice to avoid undesired drug responses and toxicities (Lenoir et al., 2021). Within the tumor microenvironment, elevated plasma concentrations of the acute-phase reactant C-reactive protein (CRP) and proinflammatory cytokines, particularly interleukin 6 (IL-6), were proven to play important roles in the occurrence and development of a variety of malignancies (e.g., tumor cell proliferation, invasion, progression, and metastasis) (Heikkilä et al., 2008; Allin and Nordestgaard, 2011; Taniguchi and Karin, 2014). Convincing evidence has revealed that the effects of tumor-derived CRP and IL-6 are closely associated with reduced mRNA and protein synthesis of hepatic CYP3A4 and attenuated enzymatic activity in tumors derived from different patients, directly leading to decelerated CYP3A4-mediated metabolism and enhanced toxicity to docetaxel and vinorelbine (Rivory et al., 2002; Charles et al., 2006; Kacevska et al., 2008). Consistent with clinical observations, the repressed transcriptional level of the CYP3A4 gene was greatly associated with increased production of these inflammatory markers in human primary hepatocytes and different animal models of extrahepatic cancer xenografts, and the concentration-inhibition relationship between IL-6 and CYP3A4 appears to be the strongest among the proinflammatory cytokines and P450 isozymes (Charles et al., 2006; Aitken and Morgan, 2007). These observations indicate that the inhibitory effect of proinflammatory cytokines on CYP3A4 expression and activity is recognized as a common feature of tumor-derived inflammatory responses, which thus impairs the biotransformation of a wide range of (pro) drugs that are metabolized through the CYP3A4 enzyme (Kacevska et al., 2008). A physiologically based pharmacokinetic model was recently developed to quantify the effect of inflammatory responses on CYP3A4 activity to predict the pharmacokinetic behavior of its substrates based on systemic CRP (Simon et al., 2021). Mechanistically, previous work revealed that a network signal pathway of transcription factors, including the nuclear receptor-mediated regulatory pathways, nuclear factor-kappaB (NF-κB) signaling pathways, and several other liver-enriched transcription factors, was responsible for tumor-derived proinflammatory cytokines and was likely to be involved in inflammation-altered CYP3A4 repression (Goodwin et al., 1999; Goodwin et al., 2002; Charles et al., 2006; Zhou et al., 2006; Martínez-Jiménez et al., 2007). Recent studies suggest that the pentose phosphate pathway contributes to IL-6-mediated CYP3A decrease and that miR-155 and other microRNAs cause coordinated CYP3A4 downregulation in the context of inflammation (Kugler et al., 2020; Liu et al., 2020).

Furthermore, obesity is a low-grade chronic inflammatory condition that has been linked to tumor-promoting effects (Olefsky and Glass, 2010; Iyengar et al., 2016). It has been reported that CYP3A4 activity and protein expression are suppressed in vitro and in obese humans (Kotlyar and Carson, 1999; van Rongen et al., 2018). Taken together, these studies imply that tumor-mediated inflammation has the potential to affect CYP3A4-mediated drug responses and toxicities. Thus, it is biologically and therapeutically important to pay close attention to the critical relationship between tumor-related inflammation and CYP3A4 activity and expression, as well as to conduct intervention studies to improve the pharmacokinetics and efficacy of these chemotherapeutic drugs.

Impact of Hepatic or Renal Impairment on CYP3A4 and Drug Responses

Hepatic and renal function are considered to be major determinants of drug exposure (Krens et al., 2019). Unfortunately, renal impairment and hepatic impairment in cancer patients are common comorbidities, either because of the disease itself or as a result of previous toxic anticancer regimens.

The liver is known to be the primary organ involved in the metabolism and disposal of the vast majority of drugs. Because the activity of P450s in the liver has a significant impact on the pharmacokinetics and therapeutic effects of the majority of anticancer drugs, when the main organs are damaged, homeostasis is disrupted and the expression and activity of hepatic P450s may change. In most cases, the release of circulating proinflammatory cytokines caused by infection or inflammation, for example, can result in altered intrahepatic and extrahepatic P450 expression and decreased P450 activities. Accordingly, liver dysfunction is associated with decreased CYP3A4 activity (Kolwankar et al., 2007). The catalytic activity and protein expression of CYP3A4 were significantly lower than those in healthy individuals, quantified by using proteomic methods and HLM, which may alter drug clearance, thereby influencing the efficacy and safety of CYP3A4-metabolized drug treatments (Jamwal et al., 2018). Moreover, CYP3A4 activity in the liver decreased with increased fat content, and protein content decreased with increased severity of steatosis in humans (Kolwankar et al., 2007). The etiology (noncholestatic vs. cholestatic) and severity of liver disease are closely related to decreased CYP3A4 activity, and the decreased CYP3A4 activity contributes to an impaired ability of the liver to eliminate therapeutic drugs metabolized by CYP3A4 among patients with hepatic cirrhosis, which potentially affects subsequent therapy response (Furlan et al., 1999; Chalasani et al., 2001; Yang et al., 2003). Prior pharmacokinetics studies have demonstrated that the metabolic clearance of CYP3A4 substrates (lignocaine, quinidine, midazolam, etc.) in patients with liver cirrhosis is significantly reduced (Stanek et al., 1997; Orlando et al., 2003; Vuppalanchi et al., 2013). In HCC patients with a certain degree of liver damage, the altered pharmacokinetics of sorafenib and bosutinib resulted in an unpredictable inhibition–plasma concentration relationship between CYP3A4 and drug exposure in vivo (Fucile et al., 2015; Abbas and Hsyu, 2016).

In addition to hepatic impairment, care should also be given to the activity and expression of CYP3A4 in renal impairment. It is estimated that approximately 55% of cancer patients have a glomerular filtration rate below 90 ml/min, and approximately 15% have a glomerular filtration rate below 60 ml/min (Launay-Vacher et al., 2007). In patients with chronic kidney disease (CKD), the activity and/or expression of drug-metabolizing enzymes (DMEs) may be downregulated or directly inhibited by accumulated toxins and drug metabolic clearance may be altered, which may lead to increased drug exposure at normally administered doses. Renal impairment reduces not only renal clearance but also hepatic clearance of CYP3A4 substrates. CYP3A4 activity is significantly attenuated in patients with severe CKD, which could have a substantial impact on drug therapy. Therefore, for patients with moderate to severe renal impairment, it is necessary to adjust the dose through therapeutic drug monitoring (TDM) to ensure that patients with CKD receive the maximum benefit from drug treatments while minimizing potential toxicity (Yamamoto et al., 2018). For patients with mild renal or hepatic impairment, asciminib and dabrafenib are generally well tolerated, and no dose adjustment or special monitoring is required (Puszkiel et al., 2019b; Hoch et al., 2021). Additionally, the effects of uremic toxins on CYP3A4 activity were diminished in end-stage renal disease patients on maintenance hemodialysis, whereas there was no discernible effect on the pharmacokinetics of CYP3A4-metabolized drugs. However, caution is recommended when administering and performing appropriate TDM (Pai et al., 2019).

Concomitant Drugs Involving Inhibition or Induction of CYP3A4-Activated/Deactivated Anticancer Drugs

In most cases, cancer patients receive combination chemotherapy along with other drugs (such as antiemetics) as needed, which might induce or inhibit CYP3A4 activity, thereby altering the metabolic balance of activation and inactivation of some anticancer drugs. The risk of clinical safety issues is high under these conditions. Therefore, care should be taken to avoid potential DDIs of anticancer drugs, which may modulate the activity of this metabolic pathway in the liver and extrahepatic tissues. Comedication with CYP3A4 inducers might enhance the activity of this enzyme and promote the activation/inactivation of CYP3A4-metabolized anticancer drugs, possibly resulting in enhanced/reduced efficacy and toxicity. Conversely, inhibitors and substrates of CYP3A4 may interfere with the activation/inactivation of these drugs (Kivistö et al., 1995). For example, the CYP3A4 inducer St. John's wort (SJW) could increase docetaxel metabolism, and the effect of SJW on the pharmacokinetics of irinotecan was stronger than that of docetaxel (Komoroski et al., 2005). The coadministered SJW in patients treated with irinotecan reduced the concentration of active metabolites in plasma and affected the efficacy (Mathijssen et al., 2002). Therefore, it is theoretically possible that patients with CYP3A4-metabolized drug treatments would experience increased drug toxicities when combined with CYP3A4 inhibitors. However, a retrospective study did not indicate an increase in irinotecan-induced toxicity when coadministered with clarithromycin (a CYP3A4 inhibitor) in patients with colorectal cancer (Makihara et al., 2017), possibly due to uridine diphosphate glucuronosyltransferase isoform 1A1, organic anion transporter 1B1, and genetic variation factors such as irinotecan-induced wide interindividual variability in drug response and toxicity (Riera et al., 2018). More detailed studies on individualized dose adjustment for cancer patients in clinical practice may be accomplished by developing a combined population pharmacokinetic model of irinotecan and its metabolites to prevent adverse effects caused by excessive drug/metabolite accumulation (Oyaga-Iriarte et al., 2019).

The pharmacokinetic consequences of TKIs are frequently affected by comedication with known CYP3A4 inducers or inhibitors in clinical practice (Teo et al., 2015). Coadministration of erlotinib with a CYP3A4 inhibitor (ketoconazole) significantly increased erlotinib exposure, but pre- or contemporaneous treatment with a powerful CYP3A4 inducer (rifampicin) decreased erlotinib exposure (Rakhit et al., 2008; Hamilton et al., 2014). Hence, dosage adjustment of erlotinib may be necessary when CYP3A4 inducers and inhibitors are concomitantly used. It has been reported that the anti–acute myeloid leukemia mediated by different TKIs (sorafenib, quizartinib, and gilteritinib) in bone marrow mesenchymal stem cells is affected by CYP3A4 activity. Inhibition of CYP3A4 activity by clarithromycin reduced the resistance of bone marrow mesenchymal stem cells to TKIs (Chang et al., 2019). Possible interactions between imatinib and CYP3A4 inhibitors or inducers may result in alterations in the plasma concentration of imatinib and coadministered drugs. Two open-label, fixed-sequence studies showed that coadministration of imatinib with SJW in healthy adult volunteers significantly affected the pharmacokinetics of imatinib (reduced imatinib exposure by 30%∼40% and increased oral clearance by 44%), which may compromise imatinib’s clinical efficacy (Frye et al., 2004; Smith et al., 2004). Aprepitant (a moderate CYP3A4 inhibitor), a drug that prevents chemotherapy-induced nausea and vomiting, largely increased the AUC and Cmax (99% and 53%, respectively) after a single dose of bosutinib in healthy volunteers (Hsyu et al., 2017). Another study showed that sorafenib increased plasma concentration in HCC patients by inhibiting CYP3A4 in combination with felodipine (Gomo et al., 2011), and combined with prednisolone it reduced the plasma concentration by inducing CYP3A4 (Noda et al., 2013). In addition, enhanced exposure to sunitinib or pazopanib was detected in patients coadministered with CYP3A4 inhibitors (amlodipine, amiodarone, and diltiazem), hence increasing the risk of adverse events (Azam et al., 2020).

Increased attention should be given to CYP3A4-metabolized anticancer drugs with narrow therapeutic indices that are also CYP3A4 inhibitors and/or inducers in the clinic (see Table 2). For instance, tamoxifen and its active metabolites desmethyltamoxifen and 4-hydroxytamoxifen inhibit CYP3A4 enzyme-mediated metabolic activity; therefore, tamoxifen and its metabolites may cause DDIs by inhibiting the metabolic activity of CYP3A4, particularly when used in combination with other CYP3A4-metabolized substrates (Bekaii-Saab et al., 2004). In addition, some TKIs such as imatinib, dasatinib, and gefitinib have different degrees of inhibitory effects on CYP3A4 activity (Teo et al., 2015). These TKIs may result in pathway-dependent inhibition of paclitaxel hydroxylation by reducing the ratio of paclitaxel metabolites, and such differential metabolism has been linked to paclitaxel-induced neurotoxicity in cancer patients (Wang et al., 2014). Imatinib has a moderate inhibitory effect on CYP3A4 activity and can competitively inhibit the metabolism of CYP3A4 substrates (Filppula et al., 2012). Dasatinib is a weak time-dependent inhibitor of CYP3A4, which is also subject to DDIs with CYP3A4 inhibitors or inducers. In clinical DDI studies, when coadministered with a potent CYP3A4 inhibitor (ketoconazole) in patients with advanced solid tumors, its exposure increased nearly 5-fold, and the terminal half-life was reached within 3.3∼8.7 hours (Johnson et al., 2010). Concomitant apatinib administration resulted in significant increases in systemic exposure to nifedipine (CYP3A4 substrate) (Zhu et al., 2020), and it greatly hindered the metabolism of gefitinib in vitro and in vivo (Wang et al., 2021).

TABLE 2

Representative CYP3A4-metabolized drugs are also CYP3A4 inhibitors and/or inducers in the clinic

Information from the DrugBank database (https://go.drugbank.com/drugs) and ChemIDplus (https://chem.nlm.nih.gov/chemidplus/). These inducers and inhibitors may interact with any CYP3A4 substrate and lead to an increased risk of adverse events.

These available pharmacokinetic-pharmacodynamic DDI data suggest that coadministered inducers or inhibitors of CYP3A4 may have a considerable impact on drug exposure and overall clinical outcomes. Especially for anticancer medicines with low therapeutic indices and steep dose-response curves, even small alterations in the pharmacokinetic profile might significantly affect the clinical effectiveness of drugs. Due to the increased risk of toxicity, any metabolic interactions resulting from concurrent therapy with CYP3A4 inhibitors or inducers should be recognized and avoided as soon as possible. If necessary, the daily dose must be lowered (Teo et al., 2012; Noda et al., 2013; Azam et al., 2020).

Impact of Genetic Polymorphisms on CYP3A4 and Drug Responses in Cancer Therapy

The presence of single nucleotide polymorphisms (SNPs) is another predominant factor in CYP3A4 expression and activity, causing sharp variations in drug susceptibility and leading to differences in CYP3A4-dependent drug response, pharmacological activity, toxicity, and clearance among individuals and ethnicities (Ozdemir et al., 2000; Rahmioglu et al., 2011). According to an investigation of nucleotide diversity and spectrum, CYP3A4 is predominantly selectively distributed among Africans, Caucasians, and Chinese (Chen et al., 2009; Hu et al., 2017). The current Human Cytochrome P450 (CYP) Allele Nomenclature Database shows that 15 distinct CYP3A4 allelic variations can modify its function (https://www.pharmvar.org/gene/CYP3A4). Variations in CYP3A4 expression and activity can be elicited by some CYP3A4 SNPs, hence affecting the metabolism of CYP3A4 substrate drugs (summarized in Table 3) and the clinical and pharmacokinetic consequences of CYP3A4 gene variants. The predominant CYP3A4 genotype is the wild type (CYP3A4*1/*1). Multiple SNP mutations are associated with alterations in the expression and activity of the CYP3A4 gene, resulting in substantial racial disparities. Among these alleles, the mutation frequency of CYP3A4*1G (g.20230G>A, rs2242480) in intron 10 of the noncoding region of CYP3A4 is high in Chinese, Asian Caucasian and African individuals (Du et al., 2007). This allele acts as an enhancer and promoter to increase the activity of CYP3A4, enhancing the metabolic capacity (Du et al., 2007). Another gene, CYP3A4*1B (rs2740574A>G), has considerable regional differences in distribution, and the allele frequency in the gene’s promoter region is relatively prominent and less distributed in Asians, whereas it is more distributed in Africans. To date, the reported results of pharmacokinetics and drug exposure-safety association studies of CYP3A4*1B gene variants have been inconsistent, and it remains controversial whether CYP3A4*1B exhibits altered enzymatic transcription rates and activity (Spurdle et al., 2002; Atasilp et al., 2020; Torres Espíndola et al., 2020). In non-African populations, CYP3A4*1B mutation moderately increased CYP3A4 expression and activity (Schirmer et al., 2006), whereas this genetic variant had no profound effect on the pharmacokinetics and drug toxicity of erlotinib and its major metabolite OSI-420 in young children with brain tumors (Reddick et al., 2019). Similarly, the CYP3A4*1B allele was marginally correlated with high-grade toxicity in erlotinib-treated patients. Individuals with higher CYP3A4 expression (G/G and A/G) were less likely to develop rash (Rudin et al., 2008).

TABLE 3

Clinical and pharmacokinetic consequences of CYP3A4 gene variants

Some frequent allelic variations (e.g., CYP3A4*18 and *16 alleles) can alter catalytic activities depending on different substrate characteristics in vitro and in vivo, leading to interindividual differences in the pharmacokinetics and pharmacodynamics of these drugs. CYP3A4*18 (rs169068 T>C) results in a shift from leucine to proline at codon 293 due to a variant allele in exon 10 (Leu293Pro). Three genotypes exist: wild type (T/T), heterozygous mutant (T/C), and homozygous mutant (C/C) (Hu et al., 2005). CYP3A4*18 is primarily distributed among East Asians, such as Chinese, Japanese, Koreans, and Malaysians, but not in Caucasians. The existence of CYP3A4*18 SNP loci was associated with an increase in the level of CYP3A4 activity (Fukushima-Uesaka et al., 2004), which could lead to a significant reduction in systemic plasma exposure of CYP3A4 substrates (Zeng et al., 2009). However, no correlation was observed between the increased toxicity of docetaxel and CYP3A4*18 in patients. Similarly, a recent study suggested that the CYP3A4*18 allele is not significantly associated with irinotecan-induced severe neutropenia (Atasilp et al., 2020). Likewise, another study found no significant connection between CYP3A4*18 variation and imatinib mesylate response in Malaysian patients with chronic myeloid leukemia (Maddin et al., 2016). CYP3A4*16 has been commonly observed in Japanese, Korean, and Mexican populations (Lamba et al., 2002; Ruzilawati et al., 2007; Chen et al., 2011; Hu et al., 2017). For paclitaxel and irinotecan, CYP3A4*16 showed a considerable reduction in enzymatic activity (more than 60%) toward paclitaxel and irinotecan compared with the wild type, whereas CYP3A4*18 exhibited a moderate reduction in its catalytic activity (by 34%∼52%) (Maekawa et al., 2010). In a clinical study of 235 Japanese cancer patients treated with paclitaxel, a significant reduction (20%) in paclitaxel metabolite 3-p-hydroxypaclitaxel/paclitaxel exposure was observed in heterozygous CYP3A4*16 carriers compared with wild-type CYP3A4*1/*1 carriers, suggesting a lower metabolic activity of CYP3A4*16 (Nakajima et al., 2006). In contrast, the catalytic activity of CYP3A4*16 toward docetaxel was retained, indicating that this allele has no substantial impact on docetaxel metabolism in vivo (Maekawa et al., 2010).

Several variant alleles are associated with reduced hepatic CYP3A4 gene transcription, and attenuation of enzymatic activity therefore reduces drug clearance and increases the risk of paclitaxel-induced grade 3 neurotoxicity (de Graan et al., 2013; Hannachi et al., 2020). For example, CYP3A4*20 allele (rs67666821) is a rare variant in which enzyme activity is absent. It is found in 1.2% of the Spanish population but harbors a low frequency in most Asian, European, and African populations (0.22%, 0.06%, and 0.26%, respectively) (Apellániz-Ruiz et al., 2015a). This allelic variant is characterized by the insertion of an adenine residue (c.1461_1462insA) that carries a premature stop codon (p. P488Tfs*494), thus synthesizing a truncated and inactive protein (Westlind-Johnsson et al., 2006). Similar to CYP3A4*20, the CYP3A4*22 variant allele (rs35599367 C>T) is a frequently investigated allele variant of CYP3A4 that is mainly due to a downregulation of CYP3A4 mRNA expression and activity both in vivo and in vitro, with an incidence of 0.083 in Caucasians and 0.043 in Asian and African populations (Elens et al., 2011). A population-pharmacokinetic model was recently developed to investigate pazopanib systemic exposure; its clearance rate in CYP3A4*22 mutant patients was significantly lower (35%) than that in wild-type patients, and the incidence of severe toxicity in patients was increased, requiring dose adjustment according to CYP3A4*22 status (Bins et al., 2019). A multivariate analysis indicated that the clearance of another TKI, sunitinib, was reduced by 22.5% in 114 patients with metastatic renal cell carcinoma or gastrointestinal stromal tumors who had the CYP3A4*22 genotype compared with the clearance in CYP3A4*22/*22 participants (Diekstra et al., 2014). Similarly, erlotinib clearance was reduced in people carrying the CYP3A4*22 allele, and CYP3A4*22 heterozygotes were associated with severe early onset myelosuppression in patients with sunitinib-induced renal cell carcinoma (Patel et al., 2018). A pharmacokinetic-pharmacogenetic evaluation showed that the CYP3A4*22 genotype was closely related to the metabolism and efficacy of tamoxifen in 730 breast cancer patients who received tamoxifen adjuvant therapy, resulting in increased plasma concentrations of endoxifen (Puszkiel et al., 2019a).

In addition, some low-frequency/rare alleles (e.g., CYP3A4*2, CYP3A4*4, *5, *6, *8, *11, *12, *13, *17, and *26 alleles) have been reported to be linked to reduced enzymatic activity (Hsieh et al., 2001; García-Martín et al., 2002; Lamba et al., 2002; Chen et al., 2011; Werk and Cascorbi, 2014). Few studies have investigated the impact of such genetic variants on the pharmacokinetic consequences of anticancer drugs specifically metabolized by CYP3A4. There are no changes between the wild-type gene and other gene variants in the levels of expression and enzyme activity, such as CYP3A4*3, *7, *9, and *10 (Dai et al., 2001; Eiselt et al., 2001).

On the whole, the high frequency of these genetic variants may alter enzyme activity in the context of genotype-phenotype association. The genotype or actual phenotype of different populations and substrate drug-dependent characteristics should be elucidated before treatment to predict drug response and toxic effects. Prior pharmacogenetic testing may assist individualized clinical treatment and improve the benefit/risk ratio of pharmaceuticals.