In vitro evaluation of the impact of Covid-19 therapeutic agents on the hydrolysis of the antiviral prodrug remdesivir

Abstract

Remdesivir (RDV, Veklury®) is an FDA-approved prodrug for the treatment of hospitalized patients with COVID-19. Recent in vitro studies have indicated that human carboxylesterase 1 (CES1) is the major metabolic enzyme catalyzing RDV activation. COVID-19 treatment for hospitalized patients typically also involves a number of antibiotics and anti-inflammatory drugs. Further, individuals who are carriers of a CES1 variant (polymorphism in exon 4 codon 143 [G143E]) may experience impairment in their ability to metabolize therapeutic agents which are CES1 substrates. The present study assessed the potential influence of nine therapeutic agents (hydroxychloroquine, ivermectin, erythromycin, clarithromycin, roxithromycin, trimethoprim, ciprofloxacin, vancomycin, and dexamethasone) commonly used in treating COVID-19 and 5 known CES1 inhibitors on the metabolism of RDV. Additionally, we further analyzed the mechanism of inhibition of cannabidiol (CBD), as well as the impact of the G143E polymorphism on RDV metabolism. An in vitro S9 fraction incubation method and in vitro to in vivo pharmacokinetic scaling were utilized. None of the nine therapeutic agents evaluated produced significant inhibition of RDV hydrolysis; CBD was found to inhibit RDV hydrolysis by a mixed type of competitive and noncompetitive partial inhibition mechanism. In vitro to in vivo modeling suggested a possible reduction of RDV clearance and increase of AUC when coadministration with CBD. The same scaling method also suggested a potentially lower clearance and higher AUC in the presence of the G143E variant. In conclusion, a potential CES1-mediated DDI between RDV and the nine assessed medications appears unlikely. However, a potential CES1-mediated DDI between RDV and CBD may be possible with sufficient exposure to the cannabinoid. Patients carrying the CES1 G143E variant may exhibit a slower biotransformation and clearance of RDV. Further clinical studies would be required to evaluate and characterize the clinical significance of a CBD-RDV interaction.

Keywords: CES1; COVID-19; Cannabidiol; G143E; Remdesivir.

Fig. 1

Metabolic pathway of RDV.

Fig. 2

Chromatograph of GS-704277, the IS phenacetin, and RDV. The standard solution was made with 6.25 μM GS704277 and 625 μM RDV in water mixed with 3:1 (v/v) 100 μM phenacetin in acetonitrile as the internal standard.

Fig. 3

S9 fraction incubation of RDV. Model: Eq. (2). r² = 0.95. Each data point was average of the duplication. Values are shown with best fit value ± SE.

Fig. 4

In vitro coincubation of the COVID-19 therapeutic agents. The coadministration drug's final concentration is 10 μg/mL. Among all compounds assessed, CBD produced the strongest inhibition of RDV hydrolysis at 10 μg/mL.

Fig. 5

Clarithromycin weakly inhibits RDV hydrolysis. Eq. (2) was used for regression. r² = 0.94.

Fig. 6

IC50 of CBD-associated RDV hydrolysis and determination of the inhibition type. Panel A: IC50 of CBD vs RDV. Eq. (2) was used for regression. r² = 0.94. Panel B: determination of the inhibition type. A paired t-test was utilized to compare the preincubation group and control group. Control group r² = 0.94; Pre-incubation group r² = 0.95. Panel C: paired t-test of control group vs preincubation group, P = 0.21 (nonsignificant). Each data point was the average of the duplicate determinations. Values are shown with best fit value ± SE.

Fig. 7

Mechanism of CBD inhibiting RDV hydrolysis. Model: Eq. (4). r² = 0.85. Each data point represented the average of duplication. Values are shown with best fit value ± SE.

Fig. 8

S9 fraction incubation for RDV hydrolysis with CES1 G143E. r² for the wild-type condition is > 0.99, r² for G143E is 0.97. Each data point was the mean of the duplicate time points. Values are shown with best fit value ± SE.