Population Pharmacokinetics and Extrapolation of NP-011 from Animal to Humans: A Novel Therapeutic Candidate Protein for Metabolic Dysfunction-Associated Steatohepatitis

Abstract

Background/Objective: Metabolic dysfunction-associated steatohepatitis (MASH), an advanced stage of metabolic dysfunction-associated steatotic liver disease (MASLD), is characterized by progressive liver fibrosis and represents an important therapeutic target. NP-011 is a truncated recombinant human milk fat globule EGF factor 8 protein developed for the treatment of MASH-associated liver fibrosis. This study extrapolated the pharmacokinetics of NP-011 from animals to humans and retrospectively evaluated the predictions using Phase I clinical data. Methods: Population pharmacokinetic models were developed using plasma concentration–time data following the intravenous administration of NP-011 in mice, rats, and monkeys. Model performance was evaluated using goodness-of-fit diagnostics, visual predictive checks, and nonparametric bootstrap analyses. The human pharmacokinetic parameters were predicted using simple allometric scaling. The maximum lifespan potential-corrected scaling was evaluated as a supplementary sensitivity analysis. Predicted human concentration–time profiles and exposure parameters were retrospectively compared with observed data from the single ascending dose component of the Phase I data. The Phase I study was registered at ClinicalTrials.gov (NCT05387499). Results: A two-compartment model with first-order elimination described NP-011 pharmacokinetics in all three species. For a 70-kg adult, simple allometry predicted clearance, central volume of distribution, intercompartmental clearance, and peripheral volume of distribution values of 3.80 L/h, 2.88 L, 0.25 L/h and 24.61 L, respectively. Across the 0.25–4 mg dose range, the predicted median AUCinf values were 1.62- to 1.96-fold higher than the corresponding observed medians, whereas the predicted median Cmax values were 1.33- to 1.48-fold higher. Overall 17.1% of the observed concentrations were contained within the simulated 5th–95th percentile prediction intervals. The maximum lifespan potential correction resulted in greater overprediction of AUCinf, with predicted-to-observed ratios of 2.38–2.89, whereas Cmax predictions were unchanged. Conclusions: Simple allometry provided closer agreement with the observed Phase I exposure than maximum lifespan potential-corrected scaling. However, the systematic overprediction of exposure and limited coverage of the simulated prediction intervals indicate that these predictions should be interpreted cautiously. These findings support simple allometry as a practical approach for extrapolating the pharmacokinetics of NP-011 to healthy adults in early clinical development, although broader generalization requires further evaluation.