Physiologically Based Pharmacokinetic Modeling To Predict Drug-Biologic Interactions with Cytokine Modulators: Are These Relevant and Is Interleukin-6 Enough? [Commentary (Invited)]

In Vivo P450 Suppression by IL-6

The correlations between raised IL-6 concentrations and P450 activity have been reviewed previously (Morgan et al., 2008; Harvey and Morgan, 2014; Coutant and Hall, 2018). A few examples are detailed here. Raised IL-6 concentrations have been significantly correlated with decreased CYP1A2, CYP2C9, CYP2C19, and CYP3A4 activity in congestive heart failure, cancer, bone marrow transplant, coronavirus disease 2019 (COVID-19) infection, and surgery patients (Chen et al., 1994; Frye et al., 2002; Sato et al., 2016; Trousil et al., 2019; Lenoir et al., 2021a,b). In contrast, raised IL-6 concentrations were associated with increased CYP2E1 activity in patients with ovarian cancer and CYP2C9 activity in hip surgery patients (Trousil et al., 2019; Lenoir et al., 2021b). However, CYP2D6 activity did not change in hip surgery patients within 3 days of surgery or in patients infected with COVID-19 (Lenoir et al., 2021a,b).

Several therapeutic protein-drug interactions relating to IL-6 have also been reported. Administration of sarilumab or tocilizumab to patients with RA led to a decrease in simvastatin, midazolam, omeprazole, and S-warfarin area under the curve (AUC), a slight increase in CYP1A2 AUC, and no effect on CYP2D6 activity (Zhang et al., 2009; Schmitt et al., 2011; Zhuang et al., 2015; Lee et al., 2017). Similarly, patients infected with COVID-19 who received tocilizumab >12 hours prior to lopinavir administration had significantly lower lopinavir (CYP3A4 substrate) exposure (Marzolini et al., 2020). Sarilumab and tocilizumab are monoclonal antibodies (mAbs) that block IL-6 from binding to its receptor and hence remove the suppressive effect of raised IL-6 on P450s in RA and COVID-19 patients, leading to increased P450 activity and coadministered drug clearance.

In Vivo P450 Suppression by Other Cytokines

CYP1A2, CYP2C, CYP2E1, and CYP3A protein expression was reduced to 37%, 45%, 60%, and 39% of control values in hepatic microsomes isolated from surgical samples from hepatectomy patients receiving high doses of IL-2 (9 or 12 × 106 U/m2) prior to surgery (Elkahwaji et al., 1999). Similarly, methoxyresorufin and erythromycin activity was significantly reduced after IL-2 administration (Elkahwaji et al., 1999). Indinavir trough concentrations and AUC significantly increased with a corresponding significant decrease in oral clearance (CLpo) after administration of IL-2 (aldesleukin) in patients with human immunodeficiency virus (HIV) (Piscitelli et al., 1998). However, significant increases in IL-6 concentration were also observed over the 5-day IL-2 infusion, with mean IL-6 concentrations of ∼80 pg/ml by day 5 (Piscitelli et al., 1998). These levels of IL-6 are similar to those observed in patients with RA (Machavaram et al., 2013) where administration of IL-6R antagonists tocilizumab and sirukumab led to a 2.4- and 1.5-fold decrease in AUC of CYP3A4 substrates simvastatin and midazolam (Schmitt et al., 2011; Zhuang et al., 2015). Thus, the similar level of interaction (AUC ratio of 1.9-fold) between aldesleukin and indinavir (CYP3A4 substrate) suggests that IL-2 does not cause an increased suppression of CYP3A4 compared with that caused by IL-6 alone.

Basiliximab and daclizumab are monoclonal antibodies (mAbs) that act as IL-2 receptor α (IL-2Rα) antagonists. After administration of IL-2Rα antagonists (mainly basiliximab) to renal transplant patients, tacrolimus trough concentrations significantly increased (Sifontis et al., 2002; Lin et al., 2015). Similarly, significantly increased cyclosporine trough concentrations, early cyclosporine toxicity, and a lower dose requirement were found in pediatric renal transplant patients after dosing with basiliximab when compared with controls (Strehlau et al., 2000). The reduction in tacrolimus and cyclosporine clearance is thought to be due to basiliximab blocking the binding of circulating IL-2 to IL-2Rα on T cells, and instead IL-2 binds to the IL-2R on hepatic and intestinal cells leading to suppression of CYP3A4 (Sifontis et al., 2002). In contrast, administration of daclizumab to patients with multiple sclerosis had no effect on the exposure of midazolam, S-warfarin, omeprazole, caffeine, or dextromethorphan (Tran et al., 2016). IL-6 concentrations were not monitored in these studies.

There are few reports linking IL-8 to suppression of P450 metabolism in vivo and no data after direct dosing of IL-8 or IL-8 antagonists. In patients with psoriasis, IL-8 concentrations were significantly increased (∼10 pg/ml) when compared with healthy subjects; however, no correlation was found between raised IL-8 levels and venlafaxine [CYP2D6 and P-glycoprotein (P-gp) substrate] metabolic ratios (Godoy et al., 2016). In contrast, raised IL-8 levels in patients with ovarian cancer (71.6 pg/ml) were significantly associated with increased CYP2E1 activity (3-fold) and reduced CYP3A4 activity (42%) when compared with healthy volunteers (Trousil et al., 2019). However, the changes in enzyme activity were also significantly associated with IL-6 and TNF-α levels. In fact, the IL-6 concentration (37.3 pg/ml) and the extent of reduction in CYP3A4 activity in patients with ovarian cancer were comparable to those in patients with RA before sirukumab treatment (Zhuang et al., 2015). Since IL-8, TNF-α, and IL-6 could all contribute to effects in CYP2E1 and CYP3A4 activity, a direct role of IL-8 in a CYP2E1 or CYP3A4-mediated DDIs is inconclusive.

A double-blind crossover study where 8 μg/kg of IL-10 and placebo were administered to healthy volunteers once daily for 6 days has been published (Gorski et al., 2000). On days 4 and 5, tolbutamide (CYP2C9), caffeine (CYP1A2), dextromethorphan (CYP2D6), and midazolam (CYP3A4) were coadministered. The study showed that administration of IL-10 did not alter CYP1A2, CYP2C9, and CYP2D6 activities and that the CYP3A activity was reduced by only 12% ± 17%. The IL-10 concentrations after dosing of 8 μg/kg are likely to be much higher than those observed in patients with immune disorders (89.5 pg/ml in psoriasis, 58.7 pg/ml in RA, and 12.6 pg/ml in systemic lupus erythematosus) (Lacki et al., 1995; Godsell et al., 2016; Sobhan et al., 2016).

There are few reports linking IFN-γ to suppression of P450 metabolism in vivo, and there are no data after direct dosing of IFN-γ or IFN-γ antagonists. In healthy subjects suffering with an acute viral respiratory infection, IFN-α and IFN-γ concentrations were significantly increased (2.7- and 11.3-fold, respectively) and antipyrine clearance was significantly decreased (1.3-fold) compared with controls (Brockmeyer et al., 1992). IFN-α and IFN-γ concentrations are also markedly higher in HIV patients with severe disease. When these patients were treated with zidovudine, the IFN-α and IFN-γ concentrations significantly decreased (60% and 59%, respectively) and antipyrine clearance significantly increased (1.2-fold) (Brockmeyer et al., 1992, 1998). Decreases in theophylline, antipyrine, caffeine, mephenytoin, debrisoquine, chlorzoxazone, and erythromycin metabolism have been reported after direct administration of INF-α to hepatitis and melanoma patients, as reviewed by Lee et al. (2010). Therefore, the P450 suppression observed in respiratory infection and HIV patients may be due to increased IFN-α rather than IFN-γ. Other cytokines such as IL-6 may also be increased in these diseases. Changes in antipyrine clearance in subjects with respiratory infection or in HIV patients treated with zidovudine are generally more limited than the changes in P450 substrate clearance reported upon administration of IL-6R antagonists to patients with RA (Schmitt et al., 2011; Zhuang et al., 2015).

There have been several reports of vaccine-drug interactions, which have been attributed to increases in IFN-γ concentrations after vaccination (Pellegrino et al., 2015). The data for warfarin (CYP2C9) are conflicting between studies, which may reflect the limited effect of IFN-γ on CYP2C9 in vitro (Aitken and Morgan, 2007). Reports for theophylline (CYP1A2) are also conflicting (Jonkman and Upton, 1984; Pellegrino et al., 2015), potentially due to inappropriate timing of some studies, whereby the maximum P450 suppression and effect on theophylline PK were missed due to the sparse sampling used. There are limited reports of vaccine interactions with anticonvulsants (e.g., carbamazepine, phenytoin, and phenobarbital) showing an increase in anticonvulsant exposure after vaccination, although the extent and duration of the drug-vaccine interaction differs widely between reports (Pellegrino et al., 2015). Vaccination also causes a significant transient increase in IL-6 and other cytokine concentrations (Tsai et al., 2005; Wright et al., 2005; Brydon et al., 2008; Harrison et al., 2009; Herrin et al., 2014; Sharpley et al., 2016; Kuhlman et al., 2018). Therefore, the role of IFN-γ in drug-vaccine interactions is not clear.

Frye et al. (2002) reported that in congestive heart failure patients given a metabolic probe cocktail consisting of caffeine (CYP1A2), mephenytoin (CYP2C19), dextromethorphan (CYP2D6), and chlorzoxazone (CYP2E1), a significant inverse relationship was found between both TNF-α and IL-6 plasma concentrations and the activity of CYP2C19. Since both TNF-α and IL-6 could contribute to suppression of CYP2C19 activity, the role of TNF-α in a CYP2C19-mediated DDI is inconclusive. In another study, patients with HIV had lower CYP3A4 and CYP2D6 activity when compared with age- and sex-matched healthy volunteers (18% and 90%, respectively) but no significant difference for CYP1A2 (Jones et al., 2010). Higher TNF-α concentrations in patients with HIV were significantly correlated with the reduced CYP3A4 activity but not CYP2D6 activity (Jones et al., 2010). Raised TNF-α levels in patients with ovarian cancer (45.4 pg/ml) were significantly associated with increased CYP2E1 activity (3-fold) and reduced CYP3A4 activity (42%) when compared with healthy volunteers (Trousil et al., 2019). However, the changes in enzyme activity were also significantly associated with IL-6 and IL-8 levels. Since IL-8, TNF-α, and IL-6 could all contribute to effects on CYP2E1 and CYP3A4 activity, the role of TNF-α in a CYP2E1- or CYP3A4-mediated DDI is inconclusive. In patients with psoriasis, TNF-α concentrations were also significantly increased (∼12 pg/ml) when compared with healthy subjects; however, no correlation was found between raised TNF-α levels and venlafaxine (CYP2D6 and P-gp substrate) metabolic ratios (Godoy et al., 2016). Similarly, TNF-α levels in patients infected with COVID-19 or after hip surgery did not correlate with the decreased CYP1A2, CYP2C19, or CYP3A4 activity observed in these patients, which is likely due to the increased IL-6 levels (Lenoir et al., 2021a,b).

A few therapeutic protein-drug interactions relating to TNF-α have also been reported. Etanercept is a fusion protein that blocks TNF-α from binding to its receptor and hence would remove any suppressive effect of raised TNF-α on P450s in patients. Administration of etanercept to healthy volunteers had no effect on digoxin (P-gp substrate) or warfarin (CYP2C9 substrate) exposure (Zhou et al., 2004a,b); however, healthy volunteers would be expected to have low circulating TNF-α levels, and hence any potential P450 suppression prior to etanercept administration would be minimal. Wen et al. (2020) reported a case study of a patient with ankylosing spondylitis, hypertension, diabetes mellitus, and IgA nephropathy who was receiving etanercept and cyclosporine. Use of etanercept was correlated with increased cyclosporine (CYP3A4 substrate) clearance; however, the authors suggest that this was due to the large decrease in circulating IL-2 concentrations after administration of etanercept rather than a direct effect of TNF-α on CYP3A4 (Wen et al., 2020).

In vivo data regarding P450 suppression by VEGF and IL-17 are extremely limited. In one phase I/II clinical trial, bevacizumab (anti-VEGF mAb) was administered to non–small-cell lung cancer patients with erlotinib (CYP3A4 substrate) and exposure of both drugs was compared with that in patients receiving each drug alone (Herbst et al., 2005). No differences in erlotinib PK were found upon coadministration of bevacizumab, suggesting that VEGF does not have a suppressive effect on CYP3A4. In patients with psoriasis, IL-17 concentrations were significantly increased (∼4 pg/ml) when compared with healthy subjects; however, no correlation was found between raised IL-17 levels and venlafaxine (CYP2D6 substrate and P-gp) metabolic ratios (Godoy et al., 2016).

Finally, it is worth noting the cytokine modulatory effects of some small molecule drugs such as toll-like receptor (TLR)-7 agonists. Jones et al. (2012) hypothesized that the time-dependent PK observed for PF-04878691 was a result of P450 suppression caused by TLR-7 agonism causing elevation of cytokine levels. This example illustrates that cytokine-mediated DDIs are not limited to biologics only, as small molecules that alter cytokine levels can also lead to modulation of P450 expression.

留言 (0)

沒有登入
gif