From physiology to disease: A quantitative framework for system-disease-drug interaction in cortisol replacement therapy
Davide Bindellini (1, 2), Robin Michelet (1), Linda B.S. Aulin (1), Johanna Melin (1, 2), Wilhelm Huisinga (3), Uta Neumann (4), Oliver Blankenstein (4, 5), Martin J. Whitaker (6), Richard Ross (6), Charlotte Kloft (1)
(1) Dept. of Clinical Pharmacy and Biochemistry, Institute of Pharmacy, Freie Universitaet Berlin, Germany, (2) and Graduate Research Training program PharMetrX, Germany, (3) Institute of Mathematics, University of Potsdam, Germany (4) Charité-Universitätsmedizin Berlin, Berlin, Germany (5) Labor Berlin, Charité Vivantes GmbH, Berlin, Germany (6) University of Sheffield, United Kingdom
Objectives: Congenital adrenal hyperplasia (CAH) is characterised by a decreased or absent cortisol production, mainly caused by reduced 21-hydroxylase enzyme activity. Cortisol production is stimulated by adrenocorticotropic hormone (ACTH) following a circadian rhythm, with cortisol suppressing ACTH secretion through feedback inhibition. In CAH, the lack of cortisol results in ACTH and subsequently androgens overproduction [1]. The severity of CAH depends on the percentage of remaining 21-hydroxylase activity and is typically classified as: Salt wasting (SW, 0%), simple virilising (SV, 1%-2%) and nonclassic (NC, 20%-50%). Typically, SW and SV patients show symptoms of cortisol deficiency and signs of androgens excess early in life, while NC patients are often asymptomatic at first, showing signs of androgens excess later in life [2].
Hydrocortisone (synthetic cortisol, HC) is the drug-of-choice for cortisol replacement therapy in CAH patients, aiming to mimic the physiological cortisol circadian rhythm while ideally avoiding ACTH and androgens overproduction. Since remaining endogenous cortisol production differs largely among SW, SV and NC patients, to optimise cortisol replacement therapy it is essential to understand the ACTH-cortisol system physiology, the impact of CAH and how this is modulated by HC therapy. Both modified-release (MR) [3] and immediate-release (IR) HC formulations are available, yet which one is most beneficial for a patient type and how to optimally dose them is understudied.
By developing a quantitative framework based on healthy adults data this work aims to:
1) Quantify the ACTH-cortisol system regulation in the healthy state
2) Characterise IR and MR HC PK
3) Evaluate the disease impact of different severity on the ACTH-cortisol system in virtual CAH patients
4) Evaluate the interaction of the diseased system with IR and MR HC therapy in virtual CAH patients
Ultimately, this work aims to move the field towards cortisol replacement therapy individualisation.
Methods: Clinical trial data from healthy adult males was leveraged to develop the framework: Endogenous ACTH and cortisol concentrations (N=13), as well as cortisol concentrations following HC administration as IR oral granules (N=29), MR oral granules (N=22) and intravenous bolus infusion (N=13) were available. To allow for a characterisation of HC PK, dexamethasone (DEX) was administered to suppress endogenous ACTH and cortisol production.
A previously developed IR HC PK model was used as starting point for the modelling activities [4]. The endogenous ACTH and cortisol, IR and MR HC submodels were developed sequentially and integrated into a nonlinear mixed-effects modelling framework. Cortisol/HC disposition kinetics were assumed to be identical.
1) To characterise ACTH pulsatile secretion, different numbers of surge functions [5] were evaluated in combination with baseline secretion. To describe ACTH-dependent cortisol production, linear, loglinear, and (sigmoidal) Emax models were evaluated, while (sigmoidal) Imax models were evaluated for cortisol-driven ACTH suppression.
2) Linear, saturable, dual absorption and transit compartment absorption models (TCAM) [6] were evaluated for IR HC. To model MR HC absorption, the IR absorption model was extended with drug release and dissolution processes.
3) To generate virtual CAH patients with different disease severity, the estimated maximum cortisol production in the healthy state was reduced to 0% (SW), 2% (SV) and 20% (NC), mimicking reduced enzymatic activity: ACTH and cortisol profiles were simulated (n=1000) and interpreted in relation to typical clinical phenotypes of SW, SV and NC patients.
4) To evaluate the impact of HC therapy, ACTH and cortisol concentrations were simulated (n=1000) in three different HC single dosing scenarios per patient type: SW and SV patients given 20 mg MR HC at 23:00, 10 mg IR HC given at 05:00 or 07:00, and NC patients given 5 mg MR HC at 23:00, 5 mg IR HC given at 05:00 or 07:00. To evaluate treatment benefit, reaching cortisol concentrations similar to healthy individuals and reducing ACTH overproduction compared to untreated virtual CAH patients, were used as metrics.
Results: Cortisol/HC disposition was best described by a two-compartment model with linear elimination and theory-based allometric scaling.
1) Endogenous ACTH secretion was best characterised using a baseline secretion and two surge functions for pulsatile secretion (estimated morning peak time=06:18), while the ACTH-dependent cortisol production was best described using a sigmoidal Emax model (Emax=5400 nmol/h, EC50=6.63 pmol/L, γE=2.94). Cortisol-driven suppression of ACTH pulsatile secretion was best characterised using a sigmoidal Imax model (Imax=100%, IC50=160 nmol/L, γI=5.33).
2) In the HC PK submodels, DEX was assumed to fully suppress ACTH pulsatile secretion (IDEX=100%). The absorption of HC after release from the IR formulation was best characterised by a TCAM. The drug release and dissolution processes of the MR formulation were described using two surge functions (two distinct absorption phases) to mimic pH-dependent drug release followed by linear dissolution.
3) The estimated Emax value was scaled to 0%, 2% and 20% to generate virtual SW, SV and NC patients, respectively. In SW and SV patients, cortisol concentrations were negligible while ACTH morning peak concentrations were ~100-fold higher than in healthy individuals. By contrast, NC patients showed cortisol profiles similar to healthy individuals and ACTH morning peak concentrations were ~10-fold higher than in healthy individuals. The results were relatable to typical clinical phenotypes of SW, SV and NC patients.
4) For SW and SV patients, 20 mg MR HC at 23:00 was the most beneficial as it achieved night and morning ACTH and cortisol concentrations similar to healthy individuals. NC patients benefited similarly from 5 mg MR HC at 23:00 and 5 mg IR at 05:00: Both reduced ACTH overproduction compared to untreated NC patients. In all patients, IR doses given at 07:00 did not reduce ACTH overproduction showing the importance of dosing HC prior to ACTH secretion peak time.
Conclusions: We developed a quantitative framework characterising the healthy ACTH-cortisol system, the impact of CAH and the interaction with HC therapy. As shown, the framework can be used to simulate ACTH and cortisol concentrations in healthy individuals and virtual CAH patients with different disease severity, and to evaluate HC dosing regimens in different types of CAH patients with the aim of avoiding ACTH overproduction besides mimicking cortisol circadian profiles. In the future, CAH patient simulations will be validated using real-world data. The inclusion of multiple dosing simulations will be essential to evaluate HC dosing regimens over multiple days. The framework has the potential to be used to individualise therapy in a clinical setting based on the remaining enzymatic activity of a CAH patient.
References:
[1] D.P. Merke, S.R. Bornstein. Congenital adrenal hyperplasia. The Lancet 365: 2125–2136 (2005).
[2] M.I. New, M. Abraham, B. Gonzalez, M. Dumic, M. Razzaghy-Azar, D. Chitayat, L. Sun, M. Zaidi, R.C. Wilson, T. Yuen. Genotype-phenotype correlation in 1,507 families with congenital adrenal hyperplasia owing to 21-hydroxylase deficiency. Proc Natl Acad Sci U S A 110: 2611–2616 (2013).
[3] M.J. Whitaker, M. Debono, H. Huatan, D.P. Merke, W. Arlt, R.J. Ross. An oral multi-particulate, modified release, hydrocortisone replacement therapy that provides physiological cortisol exposure. Clin Endocrinol (Oxf) 80: 554 (2014).
[4] J. Melin, Z.P. Parra-Guillen, N. Hartung, W. Huisinga, R.J. Ross, M.J. Whitaker, C. Kloft. Predicting Cortisol Exposure from Paediatric Hydrocortisone Formulation Using a Semi-Mechanistic Pharmacokinetic Model Established in Healthy Adults. Clin Pharmacokinet 57: 515–527 (2018).
[5] B. Charles, Y. Touitou, B. Selmaoui. A population Pharmacokinetic Turnover and Surge-Function Model for Describing Melatonin Biological Rhythm in Healthy Male Subjects. J Pharm Sci 98: 782–790 (2009).
[6] R.M. Savic, D.M. Jonker, T. Kerbusch, M.O. Karlsson. Implementation of a transit compartment model for describing drug absorption in pharmacokinetic studies. J Pharmacokinet Pharmacodyn 34: 711–726 (2007).