Population pharmacokinetic analysis of nonlinear behavior of adefovir
Qian Dong (1), Chunli Chen (1), Max Taubert (1), Muhammad Bilal (1), Martina Kinzig (2), Fritz Sörgel (2), Oliver Scherf-Clavel (3), Uwe Fuhr (1), Charalambos Dokos (1)
(1) Faculty of Medicine and University Hospital Cologne, Center for Pharmacology, Department I of Pharmacology, University of Cologne, Cologne, Germany. (2) Institute for Biomedical and Pharmaceutical Research, Nürnberg-Heroldsberg, Germany. (3) Institute for Pharmacy and Food Chemistry, University of Würzburg, Würzburg, Germany.
Objectives:
Adefovir dipivoxil, an anti-hepatitis B virus drug[1], is recommended as a phenotyping agent to estimate the activity of the renal organic anion transporter 1 (OAT1) in humans[2]. Along with metformin, sitagliptin, pitavastatin, and digoxin, we selected adefovir as the OAT1 probe drug in a transporter cocktail phenotyping study[3], also as it is highly selective for OAT1 in vitro[4]. Renal adefovir clearance (CLR) was used as the primary parameter to estimate OAT1 activity because it only depends on renal excretion driven by plasma concentrations and thus can be assessed reliably. While there were no major transporter-mediated drug-drug interactions (TDDIs) between the cocktail components, a slightly higher systemic exposure of adefovir when given with other cocktail components was observed[3]. The non-compartmental analysis (NCA) applied in this trial could not explain the changes of underlying pharmacokinetic (PK) processes in more detail.
The objectives of the present study was to describe the PK processes contributing to the renal elimination of adefovir in detail and to identify any effect of the other cocktail components thereon using a population PK (popPK) modeling approach.
Methods:
Plasma and urine data of adefovir from 24 healthy volunteers in the previous clinical trial were reanalyzed[3]. In this trial, a single 10 mg dose of adefovir dipivoxil was given alone and in combination with the aforementioned probe drugs in the reference and test periods, respectively. A popPK model of adefovir was developed using NONMEM 7.4.2. After the identification of a reasonable base model, the effects of potential covariates including the coadministration of the other drugs were incorporated into the final model using standard forward inclusion and backward elimination procedures.
Results:
A total of 1101 adefovir plasma and urine concentrations were available for the popPK analysis. Less than 1% of post-dose samples with concentrations lower than the lower limit of quantification were removed from the analysis.
A one-compartment model with two parallel first-order absorption processes (absorption rate constants (lag times) were 4.5 h-1 (0.15 h) and 0.76 h-1 (1.6 h), respectively) and combined nonrenal and renal elimination best-described plasma and urine data. A significantly higher bioavailability was identified in the test period (point estimate 74%) compared to the published value of 59%[5] which was used for the reference period (drop in objective function value [OFV] by 437). Implementation of saturable renal elimination decreased the OFV by 8.2 compared to linear renal elimination. Inclusion of the absolute estimated glomerular filtration rate (GFR) as a component of renal adefovir clearance further improved OFV by 15, where GFR was estimated via the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) 2012 equation[6] and was transformed to absolute GFR by taking individual body surface area[7] into account. Nonlinear renal clearance contributed to about 51% of median empirical Bayes estimates of CLR obtained from the final model (13 L/h). The estimate of the Michaelis-Menten constant (Km) of the saturable renal elimination was 142 nmol/L, which is beyond the observed range of maximum plasma concentrations (5.6 – 91 nmol/L).
The model could not be improved further by incorporating an effect of the concomitant administration with the other cocktail components on further PK parameters of adefovir, while there was evidence that such effects may exist. Thus, we were not able to identify which PK process would be altered by co-administration, while an apparent increase in bioavailability as observed during co-administration with the cocktail indicates that mainly absorption of adefovir dipivoxil or formation of adefovir from the prodrug might be altered.
Conclusions:
In conclusion, beyond elimination by glomerular filtration, a saturable renal elimination pathway of adefovir was identified using a popPK modeling approach, which probably reflects active renal secretion of the drug. Parameters describing this process may be superior to CLR to assess OAT1 activity, which can reliably be estimated only if a reliable assessment of GFR is available. The high Km value suggests that the application of this pathway to assess OAT1 activity is not compromised by saturability, while using a lower adefovir dose may provide an additional safety margin.
References:
[1] Dando, T.M. and G.L. Plosker, Adefovir dipivoxil. Drugs, 2003. 63(20): p. 2215-2234.
[2] U.S. Food and Drug Administration. Clinical drug interaction studies-study design, data analysis, and clinical implications guidance for industry. Final Guidance as of January 2020.
[3] Trueck, C., et al., A clinical drug‐drug interaction study assessing a novel drug transporter phenotyping cocktail with adefovir, sitagliptin, metformin, pitavastatin, and digoxin. Clinical pharmacology & therapeutics, 2019. 106(6): p. 1398-1407.
[4] Maeda, K., et al., Inhibitory effects of p-aminohippurate and probenecid on the renal clearance of adefovir and benzylpenicillin as probe drugs for organic anion transporter (OAT) 1 and OAT3 in humans. European Journal of Pharmaceutical Sciences, 2014. 59: p. 94-103.
[5] Gilead Sciences International Limited Fachinformation (summary of product characteristics, SPC) Hepsera 10mg Tabletten as of April 2021.
[6] Inker, L.A., et al., Estimating glomerular filtration rate from serum creatinine and cystatin C. New England Journal of Medicine, 2012. 367(1): p. 20-29.
[7] Mosteller, R., Simplified calculation of body surface area. New Engl J Med, 1987. 317: p. 1098.