First-in-Human Dose Predictions


This case study features Applied BioMath's collaboration with Alexion Pharmaceuticals. 

Crigler-Najjar syndrome type 1 (CN1)  is an an autosomal recessive disease caused by a marked decrease in uridine-diphosphateglucuronosyltransferase (UGT1A1) enzyme activity.  Delivery of a modified messenger RNA encoding for UGT1A1 (hUGT1A1-modRNA) as a lipid nanoparticle (LNP) is anticipated to restore hepatic expression of UGT1A1, allowing normal glucuronidation and clearance of bilirubin in CN1 patients.  In vivo mechanism of action and clinical pharmacology is unknown.

The goal of this collaboration was to develop a QSP model to support translation from preclinical to clinical studies, and first-in-human (FIH) studies and to use this QSP model to provide a deeper understanding of the mechanisms of hUGT1A1-modRNA and to guide design of the first-in-human clinical studies.  


  1. Preclinical: This model helped guide the selection of preclinical candidates, accelerate the termination of a program based on predicted low therapeutic window, reduced the number of in vivo studies, and provided insight into the mechanism of action of mRNA LNP.

  2. Clinical: This model helped inform recommended safe starting dose, impacted clinical strategy in term of patient populations for Ph1, and helped project consequences of nonlinear parameters that do not follow typical allometric scaling rules to inform first-in-human dose.

The Model

The QSP model was based on first principles as a system of elementary mass-action, mechanistic PKPD, ordinary differential equations.

The model reactions describe elimination of LNP from the plasma due to liposomal instability, uptake of LNP by endocytosis into liver hepatocytes, release of mRNA from the endosome into the cytoplasm, transcription of the mRNA to produce UGT1A1 protein, and glucuronidation of bilirubin by UGT1A1 leading to the increased clearance of bilirubin.

lipid nanoparticle model diagram

First in human predictions



  1. The QSP Model provided first-in-human predictions of bilirubin lowering for a dose range of a single dose, or every 1, 2, 3, or 4 weekly doses of ALXN1540 (a, b, c, d, e, respectively).

  2. Model predicted human dose response curve for steady-state, trough bilirubin levels for each dosing schedule. Dose times are marked with a black triangle. Time point for dose response value is denoted by the dashed lines in subplots a–e.

  3. QSP Model provided mechanistic explanation for various phases of PD response.

Download the Case Study