Population dose-Hgb modelling of Jesduvroq (daprodustat) across five phase 3 studies in chronic kidney disease patients with anemia
Martijn van Noort (1), Paul van den Berg (1), Teun M. Post (1), Misba Beerahee (2), Kelly M. Mahar (3), Shuying Yang (2)
(1) Leiden Experts on Advanced Pharmacokinetics and Pharmacodynamics (LAP&P), The Netherlands, (2) Clinical Pharmacology Modelling & Simulation, GSK, London, UK, (3) Clinical Pharmacology Modeling & Simulation, GSK, Collegeville, PA, USA
Objectives: Jesduvroq (daprodustat) is a hypoxia-inducible factor prolyl hydroxylase inhibitor (HIF-PHI). It is approved for the treatment of anemia in adult patients with chronic kidney disease (CKD) in Japan. Furthermore, approval was obtained in the United States for the treatment of anemia in adult patients who have been receiving dialysis for at least four months. A dose-hemoglobin (Hgb) model was established and updated continuously based on data obtained from the development program. In the current analysis, the existing daprodustat population dose-Hgb model was updated with the aim to support its dosing algorithm in CKD patients with anemia.
Methods:
Previously, a dose-Hgb model based on global phase 2 data supported the starting doses and titration algorithms in phase 3 studies through simulation. Subsequently, this model was updated in 2019 with the inclusion of three phase 3 studies in Japanese patients[1] and was the starting point for this analysis. The model consisted of a precursor cell compartment and 12 transit compartments to describe the red blood cell lifespan[1,2]. Treatment was modelled as the stimulation of the precursor cell production rate by a power of allometrically scaled dose[1,3]. In the current analysis, the model was updated using five global phase 3 studies with once daily (QD) and three times weekly (TIW) titrated dosing. It consisted of four major steps:
- Prediction of 2/3 of the current data set by the 2019 model,
- Restructuring/optimization of this model,
- Covariate analysis, including backward deletion of pre-existing covariates, and
- Prediction of the remaining 1/3 of the data.
Titration-based visual predictive checks (VPCs)[4] were used to assess model adequacy in addition to the usual goodness-of-fit diagnostics.
Results:
Model development used 53,535 Hgb values from 2770 patients. The final dose-Hgb model adequately described this data and predicted the remaining 1/3 of the data (26,317 values, 1384 patients).
The model had the following features:
- Separate Hgb baselines were estimated for different patient groups (depending on dialysis status and erythropoiesis-stimulating agent [ESA] use).
- Prior ESA dose was a covariate on true Hgb baseline (ie, baseline after washout of prior ESA treatment; −14% at 95th percentile of the ESA dose distribution).
- Disease progression was included as exponential decline of the Hgb production rate over time:
- o Faster progression in patients starting dialysis (+96.8%) vs nondialysis (ND) patients
- o Slower progression in ESA users (−31.8%) vs ESA nonusers, and chronic hemodialysis (HD)/peritoneal dialysis (PD) patients (−57.3%) vs ND patients
- The daprodustat treatment effect in ESA hyporesponders did not differ from other patients, in line with statistical analyses.
- Red blood cell life span was longer in HD patients (+13.8%) and patients just starting dialysis (+30.9%) as compared to ND or PD patients.
Other covariates (effects <13%) were CKD stage, hepcidin, body weight, ESA hyporesponder, race on (true) Hgb baseline, clopidogrel on dose, body weight and prior ESA use on red blood cell life span, and race on drug-effect slope.
Conclusions: The dose-Hgb model captured the daprodustat effect on Hgb in ND, HD, PD, ESA nonuser, ESA user, and ESA hyporesponder patients, supporting the daprodustat dosing algorithm. Disease progression varied with dialysis status, reflecting the variable decline in Hgb over the disease course. The analysis showed ESA hyporesponders and normoresponders benefited equally well from daprodustat treatment. TIW dosing response was comparable to equivalent QD dosing.
References:
[1] van den Berg P et al. Clin Pharm Therap 2021;109:S5–88. Abstract PI-033.
[2] Lledó-García M et al. J Pharmacokinet Pharmacodyn 2012;39:453–62.
[3] Companion PAGE 2023 Presentation 1, Goulooze B et al. Previously presented at ACoP 2022. https://www.go-acop.org/default.asp?abstract=434.
[4] Companion PAGE 2023 Presentation 2, Goulooze B et al. Previouslay presented at ACoP 2022. https://www.go-acop.org/default.asp?abstract=440.