Muhammad Waqar Ashraf (1), Satu Poikola (2,3), Mikko Neuvonen (4,5), Johanna I. Kiiski (4,5), Janne T. Backman (4,5,6), Klaus T. Olkkola (3), Mikko Niemi (4,5,6), Vesa Kontinen (2,3), Teijo I. Saari (1,7)
1. Department of Anaesthesiology and Intensive Care, University of Turku, Turku, Finland; 2. Division of Anaesthesiology, Department of Anaesthesiology, Intensive Care and Pain Medicine, Jorvi Hospital, University of Helsinki and HUS Helsinki University Hospital, Helsinki, Finland; 3. Department of Anaesthesiology, Intensive Care and Pain Medicine, University of Helsinki, HUS Helsinki University Hospital, Helsinki, Finland; 4. Department of Clinical Pharmacology, Faculty of Medicine, University of Helsinki, Helsinki, Finland; 5. Individualized Drug Therapy Research Program, Faculty of Medicine, University of Helsinki, Helsinki, Finland; 6. Department of Clinical Pharmacology, HUS Diagnostic Center, Helsinki University Hospital, Helsinki, Finland; 7. Division of Perioperative Services, Intensive Care and Pain Medicine, Turku University Hospital, Turku, Finland.
Objectives: Codeine is an analgesic prodrug commonly used in multimodal analgesia (1). CYP2D6 phenotype leads to marked differences in O-demethylation of codeine to morphine in vivo (2). Individuals lacking CYP2D6 activity acquire insufficient analgesia from codeine, whereas ultra-fast metabolisers encounter potentially life-threatening adverse effects (3).
CYP2D6 activity score can be used as a patient covariate in population models (4). Our primary aim was to quantify the effect of CYP2D6 genotype on the biotransformation of codeine in ambulatory care patients. We have developed a population pharmacokinetic model using plasma concentration-time profiles of codeine, morphine, codeine-6-glucuronide (C6G), and morphine-3-glucuronide (M3G).
Methods: Data were obtained from a prospective study in 1000 ambulatory surgery patients receiving 60 mg of oral codeine preoperatively. Two blood samples were drawn at 20-60 and 180-360 minutes after dosing, and plasma was analysed with LC-MS. Genotyping was carried out, and copy number variation was quantified using TaqManR genotyping assays and the QuantStudio™ 12K Flex Real-Time PCR System. Revised CPIC/DPWG guidelines were used to decipher activity score (AS) and phenotype (poor (PM), intermediate (IM), normal (NM), or ultra-rapid (UM)) of the patients for CYP2D6 activity. Non-linear mixed-effects modelling was conducted with NONMEM® (version 7.4.3).
Pharmacokinetic models were developed and validated in a stepwise manner for codeine, morphine, C6G and M3G using MU-referencing, ADVAN13 subroutine and first-order conditional estimation with interaction (FOCE-i) method. Then, CYP2D6 phenotype or AS was used to implement genetic covariate models, while final model performance was tested using standard diagnostic procedures. The influence of relevant covariates on model performance was tested using stepwise covariate modelling (SCM). Model simulations were conducted using a typical patient with AS variation between 0 – 4, or a patient collective using covariate set of the study population at standard dosing regimens.
Results: The pharmacokinetics of codeine, morphine, C6G and M3G were explained with one-compartmental models respectively. Codeine was observed to be rapidly (Ka = 8.74/h) and nearly completely (F = 0.84) absorbed from the gut. Weight scaling was observed to significantly improve the model fit (∆OFV = -55). In a parent-metabolite model, the addition of codeine to morphine metabolic ratio parameter (RM) in conjunction with first-order codeine clearance significantly improved model performance (∆OFV = -113, AIC = 733), while the ratio parameter was estimated at a biologically plausible value (RM = 8%).
One compartmental model was fitted to secondary metabolites (C6G/M3G) and these produced a good model fit (∆OFV = -1809, AIC = -1064) with plausible parameter estimates. The fraction of C6G from codeine elimination was numerically assigned to be 1 – RM, while the metabolic fraction of M3G from morphine elimination was fixed at 60% (5). Out of tested covariate models for CYP2D6 genetic effect, the exponential model provided the best fit for study data with numerical stability (∆OFV = -102, AIC = -1869). Implementation of the covariate model reduced BSV in RM from 52% to 33%, while the median (interquartile range, %) of apparent CYP2D6 activity (RM) was 0.55 (0.34 – 0.75) for CYP2D6 PMs, 6.82 (5.39 – 8.67) for IMs, 13.8 (10.9 – 16.7) for NMs and 19.9 (16.8 – 23.1) for UMs, of the total codeine clearance respectively.
No further improvements in model performance were observed in the stepwise covariate modelling (SCM) protocol. Simulations from the final model demonstrate model performance in predicting the effect of AS on increasing morphine exposure. A theoretical therapeutic window of 9 – 14 ng/mL was used in the assessment of exposure (6), and model performance showed that the AS of 3 progressively approaches morphine EC50 (9.1 ng/mL) for respiratory depression as steady-state concentration develops, while the AS of 4 results in morphine exposure that is much higher than EC50 at standard dose regimens.
Conclusions: Our final model accounts for the pharmacokinetics of codeine, morphine, C6G and M3G. It uses CYP2D6 genetic information (activity scores) for the individualization of codeine to morphine metabolism and demonstrates the high utility of using genetic makeup as a determinant of exposure in a genetically heterogeneous population when a therapeutic failure or contrarily, a toxic exposure is possible.
References:
[1] Schug S, Chandrasena C. Postoperative pain management following ambulatory anaesthesia: challenges and solutions. Ambulatory Anesthesia. 2015;2:11-20, https://doi.org/10.2147/AA.S54869.
[2] Kirchheiner J, Schmidt H, Tzvetkov M, Keulen JT, Lötsch J, Roots I, Brockmöller J. Pharmacokinetics of codeine and its metabolite morphine in ultra-rapid metabolizers due to CYP2D6 duplication. Pharmacogenomics J. 2007 Aug;7(4):257-65. doi: 10.1038/sj.tpj.6500406. Epub 2006 Jul 4. PMID: 16819548.
[3] Gasche Y, Daali Y, Fathi M, Chiappe A, Cottini S, Dayer P, Desmeules J. Codeine intoxication associated with ultrarapid CYP2D6 metabolism. N Engl J Med. 2004 Dec 30;351(27):2827-31. doi: 10.1056/NEJMoa041888. Erratum in: N Engl J Med. 2005 Feb 10;352(6):638. PMID: 15625333.
[4] Crews KR, Monte AA, Huddart R, Caudle KE, Kharasch ED, Gaedigk A, Dunnenberger HM, Leeder JS, Callaghan JT, Samer CF, Klein TE, Haidar CE, Van Driest SL, Ruano G, Sangkuhl K, Cavallari LH, Müller DJ, Prows CA, Nagy M, Somogyi AA, Skaar TC. Clinical Pharmacogenetics Implementation Consortium Guideline for CYP2D6, OPRM1, and COMT Genotypes and Select Opioid Therapy. Clin Pharmacol Ther. 2021 Jan 2. doi: 10.1002/cpt.2149. Epub ahead of print. PMID: 33387367.
[5] Gammal RS, Crews KR, Haidar CE, Hoffman JM, Baker DK, Barker PJ, Estepp JH, Pei D, Broeckel U, Wang W, Weiss MJ, Relling MV, Hankins J. Pharmacogenetics for Safe Codeine Use in Sickle Cell Disease. Pediatrics. 2016 Jul;138(1):e20153479. doi: 10.1542/peds.2015-3479. PMID: 27335380; PMCID: PMC4925073.
[6] Coetzee, J. (2010). Safety of pain control with morphine: New (and old) aspects of morphine pharmacokinetics and pharmacodynamics. Southern African Journal of Anaesthesia and Analgesia, 16(2), 7-15. doi:10.1080/22201173.2010.10872660
Reference: PAGE 29 (2021) Abstr 9735 [www.page-meeting.org/?abstract=9735]
Poster: Drug/Disease Modelling - Other Topics