Christian Hollensen1
1Novo Nordisk
Objective: The rapid initiation of phase 2 studies is increasingly vital in modern drug development. Pharmacokinetic/pharmacodynamic (PKPD) modelling provides early-stage prediction tools to aid internal decision-makers in assessing risk, potential trial outcomes, and facilitating interactions with regulatory authorities. This case study details the development of a PKPD model using data from a single-dose phase 1 trial with subcutaneous (s.c.) administration, informing a 12-week phase 2b trial with oral administration. This approach harnesses PKPD modelling as a pivotal tool for informed decision-making and dosage regimen optimization. Methods: PK, total PCSK9, and LDL-C data from a first human dose study involving s.c. administration of NNC0385-0434 (PCSK9i: PCSK9 inhibitor peptide) were used to develop a PKPD model. The trial included three different dose arms ranging from 10 to 250 mg in healthy subjects, and one arm with 200 mg in patients with hypercholesterolemia. Data from all 28 individuals treated with PCSK9I were included in the model. The PK data of PCSK9I was described using a 2-compartment model with fast and slow first-order absorption from the subcutaneous tissue and first-order clearance. The total PCSK9 data was described using the quasi-steady state approximation to the target-mediated drug disposition (TMDD) model[1]. The LDL-C data was described using an indirect effect model, with free PCSK9 affecting the clearance of LDL-C. All model parameters were simultaneously estimated. The model’s validity for trial simulation was assessed using standard goodness-of-fit evaluations and visual predictive checks. Oral absorption of PCSK9i was assumed to be similar to that of another orally administered peptide[2] for simulating the phase 2b trial with daily oral dosing of 15, 40, and 100 mg of PCSK9i. The assumption was supported with non-clinical data from a comparability trial with dogs. The model predictions for both PK and LDL-C were used both in internal decisions of optimal trial design as well as in the argumentation leading to regulatory acceptance of the clinical development approach. Results: The data from the s.c. trial was well described with the PKPD model, and the evaluation showed its suitability for trial simulations. The phase 2b trial was simulated with baseline covariates reflecting a patient population with hypercholesterolemia, between-subject variation in PKPD parameters, and within-individual variation in bioavailability. The simulations results were shared with internal stakeholders to iteratively optimise the trial design and the simulation of the final trial design was shared with regulatory authorities prior to initiation of the phase 2b trial. After 1000 simulations of the trial design, they were predicted to yield a mean LDL-C efficacy of -28% [95% prediction interval (PI): -32;-23%], -46% [95% PI: -51;-42%], and -60% [95% PI -65;-56%] for 15, 40, and 100 mg, respectively. For the actual trial outcome, the observed mean responses were -26 [standard error (SE): 4%], -39% [SE: 4%], and 56% [SE: 4%] for 15, 40, and 100 mg, respectively[3]. Conclusion: This study shows that model-informed drug development using PKPD modelling can support faster transition from a single-dose phase 1 trial with s.c. administration to a 12-week phase 2b trial with daily oral dosing. The simulations demonstrated the possibility of extrapolation from early phases to later stage trials with confidence.
[1]: Gibiansky, L., Gibiansky, E., Kakkar, T. et al. Approximations of the target-mediated drug disposition model and identifiability of model parameters. J Pharmacokinet Pharmacodyn 35, 573–591 (2008). [2]: Overgaard, R.V., Navarria, A., Ingwersen, S.H. et al. Clinical Pharmacokinetics of Oral Semaglutide: Analyses of Data from Clinical Pharmacology Trials. Clin Pharmacokinet 60, 1335–1348 (2021) [3]: PCSK9 inhibition with orally administered NNC0385-0434 in hypercholesterolaemia: a randomised, double-blind, placebo-controlled and active-controlled phase 2 trial. The Lancet Diabetes & Endocrinology, ISSN: 2213-8587, Vol: 12, Issue: 3, Page: 174-183 (2024)
Reference: PAGE 33 (2025) Abstr 11471 [www.page-meeting.org/?abstract=11471]
Poster: Drug/Disease Modelling - Other Topics