In vivo assessment of the drug interaction between sorafenib and paracetamol in rats

Sorafenib (CAS number 284461-73-0), sorafenib N-oxide, paracetamol glucuronide and paracetamol sulphate were purchased from LGC Standards (Łomianki, Poland). Lapatinib (CAS number 231277-92-9), paracetamol (CAS number 103-90-2), methanol, acetonitrile, sodium sulphate, ethyl acetate, glacial acetic acid, formic acid, perchloric acid, theophylline, ammonium acetate, sodium hydroxide and dimethyl sulfoxide (DMSO) were purchased from Sigma-Aldrich (Poznań, Poland). Water used in the mobile phase was deionised, distilled and filtered through a Millipore system (Direct Q3, Millipore, USA) before use. Sorafenib tosylate (Nexavar^®, batch number BXHT61) was purchased from Bayer Polska Sp. z o.o. (Warsaw, Poland). Paracetamol (Pedicetamol, batch number K003) was purchased from Sequoia sp. z o.o. (Warsaw, Poland).

The experimental protocol assuming involvement of animals in this study was reviewed and approved by the Local Ethics Committee, Poznań, Poland (no 61/2017). All procedures were performed in accordance with the European Union regulations concerning the handling and use of laboratory animals. Adult male Wistar rats (weight 480–530 g) were used in the study. The animals were maintained under standard breeding conditions with a 12/12 h light–dark cycle (lights on at 06.00, lights off at 18.00) at constant room temperature (23 ± 2 °C), relative humidity of 55 ± 10% and given ad libitum access to food and water. The animals were allowed to acclimatise for a week before the beginning of the experiments. The rats were divided into three groups. One group received sorafenib and paracetamol (I_S+PA), another group received sorafenib (II_S), whereas the last group received paracetamol (III_PA). Sorafenib (100 mg/kg b.w [27]) was dissolved in 1 mL 10% DMSO and administered directly into the animals’ stomachs using a gastric probe. To make sure that the animals received the entire dose of the drug, 1 mL of 10% DMSO was then administered to rinse the probe. 100 µL of blood was collected from each rat by cutting off a piece of his tail. The blood samples were collected into heparinised test tubes at the following time points: 0, 0.5, 1, 2, 3, 4, 5, 6, 7, 8, 10, 12, 24, 30, 48, 72, 96 h. Paracetamol was administrated as an oral solution (Pedicetamol^®, 100 mg/ml) at a dose of 100 mg/kg b.w [28]. to the I_S+PA and III_PA groups. Blood samples for paracetamol analysis were collected before (0′) and 5, 15, 30, 60, 90, 120, 240, 360 and 480 min after the drug administration. All the blood samples were centrifuged at 2.880 g for 10 min at 4 °C.

The concentrations of paracetamol, paracetamol glucuronide and paracetamol sulphate were assayed using the high-performance liquid chromatography (HPLC) method with ultraviolet (UV) detection [29]. Separation was achieved by isocratic elution of the mobile phase, comprising sodium sulphate 0.05 M pH 2.2 (adjusted with 85% orthophosphoric acid) and acetonitrile (93:7, v/v), at a flow rate of 1.0 mL/min through an BDS Hypersil^® C18 column (150 mm × 4.6 mm, 5.0 μm particle size) (Thermo Electron Corporation^®, Waltham, MA, USA). The total time of analysis for each run was 10 min. The column temperature was maintained at 25 °C, the UV detection wavelength was set at 254 nm and the injection volume was 50 μL. Theophylline was used as the internal standard (IS).

The concentrations of sorafenib, sorafenib N-oxide were assayed using the modified HPLC method with UV detection [30]. Separation was achieved by gradient elution of the mobile phase, comprising ammonium acetate 0.1 M pH 3.4 (adjusted with glacial acetic acid)—eluent A and acetonitrile—eluent B, at a flow rate of 1.0 mL/min through an reversed phase C8 column (Symmetry^® C8, 250 mm × 4.6 mm, 5.0 μm particle size) (Waters Corporation^®, Milford, MA, USA). The total time of analysis for each run was 22 min. Linear gradient started at 60% eluent A and 40% eluent B to 29% eluent A and 71% eluent B. The column temperature was maintained at 25 °C, the UV detection wavelength was set at 265 nm and the injection volume was 20 μL. Lapatinib was used as IS.

The Phoenix^® WinNonlin version 8.1 software (Certara L.P.) was used for the calculation of the pharmacokinetic parameters based on the plasma concentrations paracetamol, paracetamol glucuronide, paracetamol sulphate, of sorafenib, sorafenib N-oxide. The maximum plasma concentration (C_max) and the time to reach the C_max (t_max) were derived directly from the observed plasma concentrations. The total area under the concentration–time curve (AUC) was estimated by the trapezoidal rule with extrapolation from time zero to infinity (AUC_0-∞). The elimination half-life (t_0.5) was estimated from ln2/k_el. The apparent plasma drug clearance (Cl/F) was calculated by dividing the dose (D) by the AUC_0-∞. The apparent volume of distribution (V_d/F) was estimated from D/(k_el × AUC_0-∞).

The traits were tested for departure from normality using the Shapiro–Wilk test. The traits which did not show significant deviation from normality were subject to the heterogeneity of variance test, followed by pooled (heterogeneity of variance test p-value > 0.05) or Satterthwaite (heterogeneity of the variance test p-value < 0.05) t-tests to verify the significance of differences between the I_S+PA and II_S or I_S+PA and III_PA. Differences between the I_S+PA and II_S or I_S+PA and III_PA in the characteristics which showed significant deviation from normality were tested with the Kruskal–Wallis test. The analysis was performed using capability, t-test and npar1way procedures of SAS (SAS Institute Inc. 2002–2012. The SAS System for Windows version 9.4. Cary, NC, USA).

The arithmetic means of plasma concentrations for paracetamol and its metabolites: glucuronide and sulphate after oral administration to the groups are shown in Fig. 1. The main pharmacokinetic parameters from non-compartmental methods are summarized in Table 1.

Sorafenib significantly increased paracetamol C_max by 33% (p = 0.0372). When paracetamol and sorafenib were coadministered, the AUC_0-∞ of paracetamol grew from 88.62 to 151.80 μg × h/mL (p = 0.0018). In the group of rats receiving the both drugs, paracetamol t_max was longer when compared to the group receiving paracetamol alone, but there was no statistical significance (p = 0.0639). Statistically significant differences were revealed for Cl/F (p =  < 0.0001), V_d/F (p = 0.0005), t_0,5 (p = 0.0204).

The C_max of paracetamol glucuronide tended to be higher in the paracetamol group (38.38 vs. 19.83 μg/mL, p =  < 0.0001). The exposure to paracetamol glucuronide was significantly lower in the presence of sorafenib, what was reflected by decreased values of AUC_0-t (p = 0.0012), AUC_0-∞ (p = 0.0012).

Sorafenib elevated paracetamol sulphate C_max by 152% (p = 0.0012). Statistically significant differences were revealed for AUC_0-t (p = 0.0012), AUC_0-∞ (p =  < 0.0001).

The arithmetic means of plasma concentrations for sorafenib and sorafenib N-oxide after oral administration to the groups are shown in Fig. 2. The main pharmacokinetic parameters from non-compartmental methods are summarized in Table 2.

Paracetamol significantly increased sorafenib C_max by 60% (p = 0.0068). When paracetamol and sorafenib were coadministered, the AUC_0-∞ of sorafenib increased from 67.02 to 95.93 μg × h/mL, but the elevation was not statistically significant (p = 0.0929). There were no significant differences between groups with respect to the following parameters: AUC_0-t (p = 0.0742), t_0.5 (p = 0.5935), t_max (p = 0.6330), Cl/F (p = 0.0986), and V_d/F (p = 0.0783).

The C_max of sorafenib N-oxide was increased by 83% (p = 0.0023) in the I_S+PA group. Statistically significant differences were revealed for AUC_0-t (p = 0.0002) and AUC_0-∞ (p = 0.0065). The mean t_max of sorafenib N-oxide was similar in the both groups (16.3 vs. 15.5 h, p = 0.9121).

As it is necessary to use painkillers in oncological therapy, the likelihood of simultaneous use of sorafenib and paracetamol increases. In view of the fact that studies have proved strong inhibition of some UGTs by sorafenib (UGT1A9 and UGT1A1) [21] and the fact that glucuronidation is one of the metabolic pathways of paracetamol, there is a risk of interaction between these two drugs.

Conjugation with glucuronic acid is the most common mechanism of metabolism of xenobiotics and endogenous compounds (e.g. bilirubin and steroid hormones). This process is catalysed by UDP-glucuronyl transferase enzymes, mostly the UGT1A, UGT2A and UGT2B subfamilies [19]. UGT1A6, UGT1A9, and UGT2B15 participate in the metabolism of paracetamol [32]. Liu et al. [26] investigated the influence of various kinase inhibitors on the paracetamol glucuronidation process in vitro, using recombinant UDP-glucuronyl transferases and liver microsomes. The researchers observed that sorafenib, dasatinib and imatinib inhibited UGT1A9 and UGT2B15—the isoenzymes responsible for paracetamol glucuronidation. Additionally, the FDA and EMA recommend a detailed description of the inhibition of the UGT enzyme by TKIs [21]. The inhibition of drug metabolism by the UGT enzyme causes a wide range of clinically relevant DDIs. Regorafenib and sorafenib are the strongest human inhibitors of UGT enzymes that have been identified so far [21, 33]. Miners et al. [21] conclude that the extrapolation of in vitro-in vivo studies indicates that the inhibition of UGT1A1 significantly contributes to hyperbilirubinemia observed in patients treated with sorafenib.

When paracetamol and sorafenib were co-administered, the C_max and AUC_0-∞ of paracetamol increased in I_S+PA group by 33% and 71%, respectively (Table 1). After a single dose of 100 mg/kg of paracetamol administered orally Pingili et al. [34]. observed a lower C_max (5.04 µg/ml), longer t_max (1.16 h) and longer t_0.5 (4.43 h). When Mekjaruskul et al. [35] administered an analogous dose orally, they noted a higher C_max (19.10 µg/ml), longer t_max (1.0 h) and shorter t_0.5 (21.29 min).

Moreover, the study has revealed the significantly lower exposure to paracetamol glucuronide (decreased C_max and AUC_0-∞ by 48% and 42%, respectively) in the presence of sorafenib (Table 1). However, the ratios of paracetamol glucuronide/paracetamol did not reach statistical significance, suggesting the lack of influence of sorafenib on the glucuronidation of paracetamol. Nonetheless, it cannot be ruled out that the lack of significant differences in the paracetamol glucuronide/paracetamol ratios may result from the compensation of the glucuronidation pathway by the UGT1A6 form of the isoenzyme or insufficient power of the experiment. In the I_S+PA group we also observed increased C_max and AUC_0-∞ of paracetamol sulphate by 2.1- and 2.7-fold, respectively with a simultaneous increase in paracetamol sulphate/paracetamol ratio for C_max and AUC_0-∞ (p = 0.0003 and p = 0.0003, respectively). Karbownik et al. [36‐38] conducted an in vivo study and proved that other TKIs, i.e. erlotinib, lapatinib and sunitinib significantly affected the PK of paracetamol. After the administration of erlotinib, lapatinib and sunitinib the C_max of paracetamol dropped by: 18.9%, 55.7% and 63.95% and the AUC_0-∞ decreased by 35.5%, 48.8% and 68.19%, respectively. That research also proved that all these TKIs inhibited the paracetamol glucuronidation process.

The current analysis of the PK of paracetamol and its metabolites was significantly limited by the lack of NAPQI measurements.

The concomitant application of sorafenib and paracetamol contributed to 1.6-fold increase of sorafenib C_max. The comparison of the PK data of sorafenib in our project with the data published by Wang [27] and Wang [39] showed that in our study there were higher C_max values: 1.562 µg/mL, 230.86 µg/L, 338 ng/mL, respectively. The changes in the concentration may have been caused by the drug carrier but they also confirmed the high PK variability. The T_max in our study was also shorter than in the studies by Wang and Wang (5.125, 8.14, 8.0 h, respectively). We did not compare the AUC data because the blood samples were collected at different times. In our study it was up to 96 h, whereas it was up to 48 h in the study by Wang and up to 36 h in the study by Wang. In our project the t_0.5 was also longer than in the studies by Wang and Wang. Moreover, coadministration of paracetamol increased C_max and AUC_0-∞ of sorafenib active metabolite—N-oxide by 55 and 79%, respectively. However, the values of N-oxide sorafenib/sorafenib ratios for C_max and AUC_0-∞ in the II_S and I_S+PA groups (p = 0.5632 and p = 0.5952, respectively) did not confirm significant influence of paracetamol on sorafenib oxidation. The increased exposure of sorafenib and its active metabolite is most likely to have been caused by the inhibition of P-gp transport of this drug by paracetamol. Nevertheless, higher concentrations of sorafenib and sorafenib N-oxide may improve the response to therapy, but on the other hand potentiate adverse drug reactions. Both of these possibilities should be taken into account during therapy. These findings deserve further study and confirmation in humans.

As the resulting direction and extent of this interaction may be individual and influenced by dosing schedule, the randomized clinical trials after single and multiple doses of paracetamol are required.

The lack of differences in the values of the k_a constant parameters of paracetamol and sorafenib (2.35 ± 0.99 vs. 2.09 ± 1.11 h⁻¹ (p = 0.6089); 0.81 ± 0.32 vs. 0.74 ± 0.23 h⁻¹ (p = 0.6110), respectively) shows that the drugs did not influence the absorption rate of each other.

Karbownik et al. examined the influence of paracetamol on the pharmacokinetics of lapatinib and erlotinib [36, 37]. The co-administration of lapatinib and paracetamol increased the AUC_0-t and C_max of lapatinib by 240% (p = 0.0030) and by 184% (p = 0.0011), respectively. When erlotinib was applied concomitantly with paracetamol, the AUC_0-t and C_max also increased (by 31%, p = 0.0329 and by 88%, p = 0.0004, respectively). The research results indicated that the drugs used in oncological therapy may increase the intensity of adverse reactions, including those reducing the quality of life, e.g. diarrhoea.

The selection of analgesic drug is an important aspect during the sorafenib therapy of HCC patients. Li et al. [40] proved that the combination therapy with acetylsalicylic acid (ASA) and sorafenib gave a synergistic anticancer effect against liver tumours both in vitro and in vivo. The combination of these drugs induced apoptosis in the tumours without causing a weight loss, hepatotoxicity or inflammation. The authors have suggested that ASA overcomes the resistance to sorafenib and the combination of the drugs may be an effective approach to improve HCC treatment. Fujioka et al. [41] observed that the administration of a strong opioid (morphine) for anesthesia may promote the progression of lung cancer in the EGFR (epidermal growth factor receptor) phosphorylation pathway. This phenomenon decreases the efficacy of drugs, e.g. erlotinib. Therefore, the choice of the analgesic drugs in cancer patients should be based not only on its potency to reduce pain, but also on the possible influence on the activity and efficacy of anticancer drug.

The study was significantly limited by the lack of assay of sorafenib glucuronide and the free fraction of sorafenib. The free fraction of the drug was not measured because sorafenib binds to human plasma proteins in 99.8% in an in vitro study [42].