Interactions of the protease inhibitor, ritonavir, with common anesthesia drugs

Abstract

The protease inhibitor, ritonavir, is a strong inhibitor of CYP 3A. The drug is used for management of the human immunovirus and is currently part of an oral antiviral drug combination (nirmatrelvir-ritonavir) for the early treatment of SARS-2 COVID-19-positive patients aged 12 years and over who have recognized comorbidities. The CYP 3A enzyme system is responsible for clearance of numerous drugs used in anesthesia (e.g., alfentanil, fentanyl, methadone, rocuronium, bupivacaine, midazolam, ketamine). Ritonavir will have an impact on drug clearances that are dependent on ritonavir concentration, anesthesia drug intrinsic hepatic clearance, metabolic pathways, concentration-response relationship, and route of administration. Drugs with a steep concentration-response relationship (ketamine, midazolam, rocuronium) are mostly affected because small changes in concentration have major changes in effect response. An increase in midazolam concentration is observed after oral administration because CYP 3A in the gastrointestinal wall is inhibited, causing a large increase in relative bioavailability. Fentanyl infusion may be associated with a modest increase in plasma concentration and effect, but the large between subject variability of pharmacokinetic and pharmacodynamic concentration changes suggests it will have little impact on an individual patient, especially when used with adverse effect monitoring. It has been proposed that drugs that have no or only a small metabolic pathway involving the CYP 3A enzyme be used during anesthesia, for example, propofol, atracurium, remifentanil, and the volatile agents. That anesthesia approach denies children of drugs with considerable value. It is better that the inhibitory changes in clearance of these drugs are understood so that rational drug choices can be made to tailor drug use to the individual patient. Altered drug dose, anticipation of duration of effect, timing of administration, use of reversal agents and perioperative monitoring would better behoove children undergoing anesthesia.

Keywords: COVID-19; anesthesia; antiviral; children; drug interactions; pharmacodynamics.

FIGURE 1

The sigmoid Emax curve showing response curves. Both drugs have the same maximum response (E _max), but the slope of Drug B is steeper and described by a Hill coefficient (N) of 10. Small changes in concentration between the C₂₀ and C₈₀ have pronounced effect. Small changes in concentration when the concentration is above the C₉₅ or if the slope is gentle (Drug A) have far less effect.

FIGURE 2

Simulation of plasma concentration (C _p) and sedation score in a 9‐year‐old, 30‐kg child given intravenous ketamine 1.5 mg kg⁻¹ when clearance is doubled. The arousal score is graded from 0 to 5 where a score of 2 indicates the child arouses slowly to consciousness, with sustained painful stimulus and a score of 3 indicates the child arouses with moderate tactile or loud verbal stimulus. Pharmacokinetic and pharmacodynamic parameter estimates were from Herd et al.

FIGURE 3

Simulated plasma concentration (C _p) and effect profiles of rocuronium 1.2 mg kg⁻¹ (4 × ED₉₅) with and without ritonavir. The first twitch of the train‐of‐four response has been used to characterize the time‐effect profile. Recovery is slower when rocuronium is given in the presence of ritonavir and the duration of dense blockade is also longer. Concentrations achieved neuromuscular blockade on the lower flattened upper part of the sigmoid response curve, representing maximum effect (inverted Emax response). PKPD parameter estimates from Vega et al.

FIGURE 4

Fentanyl plasma concentration (C _p) is related to pupillary constriction with and without co‐administration of retinovar. Intravenous fentanyl 1 μg kg⁻¹ was given as a loading dose. Maintenance was 1 μg kg⁻¹ h⁻¹ for 2 h. Simulation revealed that small changes in concentration after infusion resulted in small changes in pupillary effect. PK parameter estimates from Shafer et al. PD parameter estimates from Asbury et al.

FIGURE 5

A child was given midazolam 0.5 mg kg⁻¹ orally. The effect compartment concentration (C _e) is linked to plasma concentration by a rate constant (keo). Simulation shows the impact of increased bioavailability (twofold) and slower clearance (30% reduction). Pharmacodynamics were described using electroencephalographic amplitudes in the 11.5–30 Hz (beta) frequency band; these were used as a proxy for sedation. Not only is electroencephalographic effect prolonged due to slower clearance, but a ceiling effect is also achieved because of higher concentrations. Pharmacodynamic parameter estimates from Mandema et al.