*3.5. E*ff*ect of Metformin on Verapamil Metabolism in Rat Hepatic and Intestinal Microsomes*

The Vmax, K<sup>m</sup> and CLint for the metabolism of verapamil with and without metformin in rat hepatic and intestinal microsomes are shown in Figure 5 and Table 4, respectively. There was no change of Vmax, K<sup>m</sup> or CLint for verapamil metabolism in the presence of metformin, indicating that metformin did not affect verapamil metabolism in rat hepatic and intestinal microsomes under these conditions.

○ without (●) metformin **Figure 5.** Nonlinear regression for mean values (± S.D.) of verapamil metabolism activity with (#) or without (•) metformin in rat hepatic (**A**, *n* = 5 for each) or intestinal (**B**, *n* = 4 for each) microsomes.



*K*m, the concentration at which the rate is one-half of the *V*max; *V*max, maximum velocity; CLint , intrinsic clearance.

#### *3.6. Rat Plasma Protein Binding of Verapamil and Metformin Using Equilibrium Dialysis*

The concentration of 5 g/mL of each drug was chosen based on previous reports [17,42]. Protein binding values of metformin with and without verapamil were 13.5% ± 5.39% and 11.6% ± 6.10%, respectively. The corresponding values for verapamil with and without metformin were 93.0% ± 38.9% and 85.1% ± 26.3%, respectively. Metformin and verapamil did not affect the rat plasma protein binding to each other.

#### **4. Discussion**

The ratio of AUC of a drug with an inhibitor (AUC<sup>i</sup> )/ AUC of a drug without an inhibitor (AUC0) of over 1.25 is classified as a relevant drug interaction by an inhibitor in U.S. FDA criteria [43]. In our study, the ratios of AUC<sup>i</sup> /AUC of metformin with and without verapamil were 1.67 and 1.74 (i.e., relative bioavailability) after the intravenous and oral administration of both drugs (Table 1), indicating that verapamil might act as an inhibitor to cause a pharmacokinetic interaction with metformin.

In the intravenous study, the contribution of CL<sup>R</sup> to CL, 66.5%, was a large portion of the metformin elimination pathway (Table 1), indicating that renal excretion is the main route of metformin elimination. The estimated CLRs of metformin considering the free fractions of metformin in the plasma (CLR,fus) were 4.08 and 14.2 mL/min/kg with and without verapamil, respectively. The CLR,fu of metformin without verapamil was faster than the reported glomerular filtration rate (GRF, represented by creatinine clearance), 5.24 mL/min/kg, in rats [44], indicating that active secretion of metformin as its renal excretion mechanism is changed by verapamil to glomerular filtration. Verapamil slowed the CLR,fu of metformin, 4.08 mL/min/kg, to the creatinine clearance level in rats, indicating that verapamil might cause metformin reabsorption in the renal tubules. The inhibited renal excretion pathway caused a dramatic increase in the systemic exposure (e.g., AUC) of metformin in this study similar as other references [18,45].

To investigate the inhibitory mechanism of verapamil on renal excretion of metformin, the IC<sup>50</sup> of verapamil against metformin uptake in HEK-293 cells overexpressing OCT2 was conducted based on the known facts that metformin is an OCT2 substrate and verapamil inhibits OCT2 [15,16]. The inhibitory effect of verapamil on OCT2-mediated metformin uptake in HEK-293 cells overexpressing OCT2 (Figure 2) supported the reduced CL<sup>R</sup> of metformin when co-administered with verapamil (Table 1). Verapamil significantly reduced the metformin concentration in the kidneys after intravenous and oral administration of metformin with verapamil (Table 2 and Figure 3), probably due to the inhibition of OCT2-mediated metformin uptake into the proximal renal tubules, as shown in the IC<sup>50</sup> of verapamil in HEK-293 cells over-expressing OCT2 (Figure 2).

Considering that the metformin concentration in the liver is important to preserve the glucose-lowering effect of metformin [11,46,47], the metformin concentration in the liver was also measured after intravenous and oral administration of both drugs (Table 2 and Figure 3). However, the metformin concentration in the liver was not changed by verapamil, which might be due to verapamil not sufficiently inhibiting OCT1-mediated metformin uptake in the sinusoidal membrane of hepatocytes. Since OCT1 in the basolateral membrane uptakes metformin from the sinusoidal blood into hepatocytes, comparable metformin concentrations in the liver with and without verapamil could be supported by the relatively high IC<sup>50</sup> of verapamil for inhibiting metformin uptake by HEK-293 cells over-expressing OCT1 (Figure 2). In other words, verapamil might have a stronger potential to inhibit OCT2 activity than OCT1 activity. Although verapamil inhibited OCT1 and OCT2-mediated metformin uptake in vitro (Figure 2), the inhibitory effect of verapamil on OCT1-mediated metformin uptake in hepatocytes might be almost negligible in in vivo studies (Table 2 and Figure 3). Therefore, verapamil is an OCT2 inhibitor in the renal proximal tubules, resulting in reduced renal excretion and increased systemic exposure of metformin.

On the other hand, the contribution of gastrointestinal (including biliary) excretion of unchanged metformin to its CLNR was almost negligible; the GI24 h was less than 0.471% of the intravenous dose (Table 1). Similarly, it has been reported that metformin is mainly eliminated via renal excretion, but the biliary excretion of metformin as a parent form into feces was negligible [11,17]. In the aspect of metformin and verapamil interactions, the unchanged CLNR and GI24 h of metformin by verapamil indicted that verapamil did not influence the non-renal elimination pathway (e.g., biliary excretion and metabolism) of metformin in rats.

After oral administration of metformin with and without verapamil, oral absorption of metformin was rapid and extensive regardless of co-administration of verapamil. For comparison, we estimated the mean 'true' unabsorbed fractions ('*F*unabs') after oral metformin administration to rats with and without verapamil based on the following reported equation [48]:

$$\begin{aligned} \text{(0.0357 = 'F\_{\text{unabs}} ' + (0.00471 \times 0.297) &\text{ without verapamil} \\ \text{(0.0153 = 'F\_{\text{unabs}} ' + (0.00831 \times 0.309) &\text{with verapamil} \end{aligned} \tag{1}$$

where 0.0357 (0.0153), 0.00471 (0.00831) and 0.297 (0.309) are the oral GI24 h, intravenous GI24 h and *F*, respectively, of metformin without verapamil (with verapamil). The '*F*unabs' values thus estimated were 1.27% and 3.43% with and without verapamil, respectively, indicating that verapamil probably does not affect metformin absorption in the intestine (Table 1). Thus, the reduced AUC of metformin

with verapamil could be due to the inhibition of renal excretion of metformin by verapamil, for the same reason as in the intravenous study.

In contrast, metformin did not change any pharmacokinetic profile of verapamil after intravenous and oral administration of metformin and verapamil together compared to verapamil alone. Metformin also did not affect the formation of norverapamil as an active metabolite of verapamil (Table 3). As hepatic metabolism via CYP3A is the main route of elimination of verapamil [28,49], and a suppressive effect of metformin on PXR-regulating CYP3A4 has been reported [21], the change of verapamil metabolism by metformin was evaluated with greater focus in this study. In parallel to the unchanged CLNR of verapamil by metformin in the intravenous co-administration of metformin and verapamil (Table 3), metformin showed a negligible interaction with verapamil metabolism in in vitro hepatic and microsomal studies (Table 4). Considering that verapamil is a drug with a high (or intermediate) hepatic extraction ratio in rats [44], its hepatic clearance (metabolism) depends on the hepatic CLint and free fraction (unbound to plasma proteins) of verapamil, and hepatic blood flow rate [50]. In our study, the unchanged CLNRs of verapamil with and without metformin (Table 3) were supported by the comparable hepatic CLints and free unfound fractions of verapamil, and constant hepatic blood flow rate with and without metformin in rats. Metformin might not affect the hepatic blood flow rate based on studies in humans [51]. In other words, these findings indicated that metformin might inhibit the hepatic metabolism of verapamil including the formation of norverapamil (Table 3). Additionally, the unchanged *K*ms and *V*maxs of verapamil with metformin compared to those without metformin in in vitro hepatic microsomal studies indicated that metformin did not affect the affinity between metabolic enzyme and verapamil and maximum rate of metabolism of verapamil.

Although the contribution of renal excretion of verapamil is minor in regards to its elimination, the renal excretion of verapamil with and without metformin was estimated as follows: the CLR,fus of verapamil with and without metformin adjusted by the free fraction of verapamil in the plasma were 0.00756 and 0.219 mL/min/kg, respectively. Both CLR,fus of verapamil were significantly slower than the reported GFR, indicating that glomerular filtration was a renal excretion mechanism of verapamil regardless of the presence of metformin and metformin did not inhibit its renal excretion in rats (Table 3).

After oral administration of both drugs, no effect of metformin on the pharmacokinetic profiles of verapamil and norverapamil was observed. The comparable AUCs of verapamil with and without metformin (Table 3) were likely due to the unchanged absorption of verapamil; the estimated '*F*unabs' values of verapamil were 0.00179 and 0.00256% for with and without metformin, respectively, from the equations [48]:

$$\begin{aligned} 0.00291 &= 'F\_{\text{unabs}} '+ (0.00148 \times 0.234) & \text{without metallicity} \\ 0.00200 &= 'F\_{\text{unabs}} '+ (0.00117 \times 0.182) & \text{with metallicity} \end{aligned} \tag{2}$$

where 0.00291 (0.00200), 0.00148 (0.00117) and 0.234 (0.182) are the oral GI24 h, intravenous GI24 h and *F* of verapamil, respectively, without metformin (with metformin). Thus, metformin might not affect verapamil absorption in the intestine (Table 3). Orally administered metformin might also not inhibit verapamil metabolism in the liver and intestine, as supported by the unchanged verapamil metabolism by metformin in the in vitro hepatic and intestinal microsome studies (Table 4).

Although Cho et al. [32] reported the inhibitory effect of verapamil on the glucose tolerance activity of metformin without any change of metformin's AUC in healthy adults, verapamil has recently emerged as a new indication for the treatment of hypertension in DM patients [22,23]. Extrapolating the rat dose to human equivalent dose based on the equation by FDA [52], there is a slight difference: the estimated human equivalent doses of metformin and verapamil in our study are 340 mg/70 kg and 227 mg/70 kg, respectively, and the corresponding doses in Cho et al. [32] were 1000 mg and 180 mg per patient (average body weight of patients was 70.5 kg). This inconsistency of doses can be one reason for the result in the different PK interaction pattern in metformin and verapamil combination. In addition, the relatively lower doses were used to emphasize the potential for the occurrence of metformin and

verapamil interaction, which can provide a clue to cause OCT mediated drug interaction as underlying mechanism for the further investigations using various dosage regimens in preclinical as well as clinical studies.
