Quantifying the sensitivity and specificity of CYP450 probe drugs in predicting in vivo enzyme activity under (patho)physiological conditions
A. M. (Marinda) van de Kreeke (1)*, Laura M. de Jong (1)*, Mariam Ahmadi (1), Jesse J. Swen (2), Catherijne A.J. Knibbe (1,3), J.G. Coen van Hasselt (1), Martijn L. Manson (1), Elke H.J. Krekels (1,4)
*both authors contributed equally, (1) Division of Systems Pharmacology and Pharmacy, Leiden Academic Centre for Drug Research, Leiden, Netherlands, (2) Department of Clinical Pharmacy and Toxicology, Leiden University Medical Center, Leiden, Netherlands, (3) Department of Clinical Pharmacy, St. Antonius Hospital, Nieuwegein, The Netherlands, (4) Certara Inc, Princeton, NJ, USA
Introduction: CYP1A2, CYP2B6, CYP2C9, CYP2D6, and CYP3A4 are involved with more than 85% of drugs that are metabolized. Disease-related change in the activity of these enzymes could therefore be of relevance for dosing regimen of many drugs. CYP450 (CYP) phenotyping, which is the quantification of an individual’s plasma clearance of CYP-specific probe drugs as a proxy for in vivo CYP enzyme activity, is increasingly applied to study alterations in CYP enzyme activity under various (patho)physiological conditions, like inflammation, obesity, or pregnancy [1-4]. This approach assumes that changes in plasma clearance of probe drugs are solely driven by changes in CYP enzyme activity (CLint). However, drug plasma clearance may also be influenced by the unbound drug fraction (fu), blood-to-plasma ratio (B/P), and hepatic blood flow (Qh), all of which may change under (patho)physiological conditions [5]. Using a physiologically-based pharmacokinetic (PBPK) workflow, we quantify how plasma clearance of CYP probes changes as a result of alterations in CYP enzyme activity (sensitivity) and of alterations in protein binding, blood-to-plasma ratio, and hepatic blood flow (specificity). Additionally, we illustrate how the results can be used to assess suitability of probe drugs for the assessment of in vivo enzyme activity under (patho)physiological conditions using three inflammatory conditions: chronic inflammation (rheumatoid arthritis), acute inflammation (surgery), and acute infection (COVID-19).
Methods: Thirteen commonly used probe drugs for the six most clinically relevant CYP enzymes were selected. This resulted in the following enzyme-probe combinations for the analysis: CYP3A4: midazolam and quinine, CYP2D6: dextromethorphan and metoprolol, CYP2C19: omeprazole, CYP2C9: diclofenac, flurbiprofen, losartan, s-warfarin, and tolbutamide, CYP2B6: bupropion and efavirenz, and CYP1A2: caffeine. Their plasma clearance was calculated with the dispersion model, for which system-specific and drug-specific parameter values were extracted from published sources. The impact of clinically relevant univariate changes of -90% to +150% for CLint and fu and -50% - +50% for B/P and Qh on plasma clearance was quantified. The sensitivity of plasma clearance of the probe drugs to changes in CLint was assessed by evaluating the proportionality of the change in plasma clearance to the change in CLint. Selectivity of plasma clearance to changes in CLint, was assessed by quantifying how much the change in plasma clearance deviated from 0 upon change in the remaining three parameters. We exemplify how to assess the suitability of probe drugs for three inflammatory conditions, by calculating the change in plasma clearance upon reported changes in fu, B/P, and Qh.
Results: Plasma clearance of investigated probe drugs proved sensitive to alterations in CLint, except for midazolam in case CYP3A4 is >50% induced. For example, an induction of 150% results in an increase of plasma clearance of only 41%. Importantly, plasma clearance was equally sensitive to variations in fu, diminishing the specificity in quantifying in vivo CYP enzyme activity. This is particularly relevant for high protein-bound probe drugs, as their relative change in fraction unbound upon changes in the abundance of drug binding plasma proteins tends to be larger. Alterations in B/P and Qh limitedly impact plasma clearance of the studied probe drugs. Assuming that fu is only dependent on human serum albumin concentration, the three inflammatory conditions cause increases in fu for all probe drugs, leading to possible non-specificity in quantifying the in vivo CYP enzyme activity based on change in plasma clearance of the probe drugs. However, whereas dextromethorphan has high increases in fu (e.g., 46% during COVID-19), metoprolol has smaller alterations in fu (e.g., 4% during COVID-19), leading to more specificity and suitability as CYP2D6 probe.
Conclusions: CYP probe drugs are sensitive to alterations in both CYP enzyme activity and protein binding. Because drug-binding plasma proteins concentrations may change under (patho)physiological conditions, changes in plasma clearance can therefore not be assumed to be directly correlated with changes in in vivo CYP enzyme activity. Consequently, alterations in protein binding should be measured and/or corrected for when using probe drug plasma clearance as a proxy for in vivo CYP enzyme activity in patient populations.
References:
[1] Breimer DD, Schellens JH. A “cocktail” strategy to assess in vivo oxidative drug metabolism in humans. Trends Pharmacol Sci. 1990;11(6):223-225. doi:10.1016/0165-6147(90)90245-4
[2] Lenoir C, Daali Y, Rollason V, et al. Impact of Acute Inflammation on Cytochromes P450 Activity Assessed by the Geneva Cocktail. Clin Pharmacol Ther. 2021;109(6):1668-1676. doi:10.1002/cpt.2146
[3] Ghasim H, Rouini M, Safari S, et al. Impact of Obesity and Bariatric Surgery on Metabolic Enzymes and P-Glycoprotein Activity Using the Geneva Cocktail Approach. J Pers Med. 2023;13(7). doi:10.3390/jpm13071042
[4] Tracy TS, Venkataramanan R, Glover DD, Caritis SN. Temporal changes in drug metabolism (CYP1A2, CYP2D6 and CYP3A Activity) during pregnancy. Am J Obstet Gynecol. 2005;192(2):633-639. doi:10.1016/j.ajog.2004.08.030
[5] Espié P, Tytgat D, Sargentini-Maier ML, Poggesi I, Watelet JB. Physiologically based pharmacokinetics (PBPK). Drug Metab Rev. 2009;41(3):391-407. doi:10.1080/10837450902891360