(T-101) Comparison of Predictive Performance of Complex IS/ID and Simple KPD Models on durability projections of Neutralizing Antibodies of LNP-mRNA-Based Vaccines

Miao Zhang – Northeastern University; Mansoor Amiji – Northeastern University; Linh Van – Moderna TX, Inc.

Objectives: The IS/ID is an extension of PKPD model to describe the immunostimulatory (IS) on the immunodynamics (ID) by incorporating adaptive immune responses of the antibody dynamics. This model has successfully guided dose selection for the COVID-19 vaccine in pediatric populations1. However, this complex IS/ID model requires the estimation of a significant number of parameters to model the immune response mechanism, which can lead to model instability in the absence of supporting experimental data. Alternatively, a simplified KPD model capturing the antibody kinetics was compared to the highly mechanistic model on the predictive performance of LNP-mRNA-based vaccines.

Methods: The complex IS/ID model incorporates the adaptive immune response mediated by B cell populations that is induced by the LNP-mRNA-based vaccine encoded antigen. The KPD model focuses solely on antigen-induced antibody kinetics. Both models were implemented using NONMEM software with FOCEI and SAEM estimation algorithms to assess their predictive performance, design dosing regimens, and compare antibody durability responses.

Results: Diagnostics of both models, including GOF, VPC, and model validation, showed consistent predictive performance and parameters accuracy. For example, the ratios of predicted AUC and Cmax compared to observed data fell within a 2-fold range. However, the IS/ID model had limitations in the precision of the estimates due to number of parameters involved compared to KPD, yet it was inferior in capturing the Cmax upon subsequent dose administration.

Conclusions: Both the IS/ID and KPD models exhibit comparable predictive performance, suggesting that a simple model may be feasible for describing antibody durability.

Citations: 1 Ivaturi V, Attarwala H, Deng W, Ding B, Schnyder Ghamloush S, Girard B, Iqbal J, Minnikanti S, Zhou H, Miller J, Das R. Immunostimulatory/Immunodynamic model of mRNA-1273 to guide pediatric vaccine dose selection. CPT Pharmacometrics Syst Pharmacol. 2025 Jan;14(1):42-51. doi: 10.1002/psp4.13237.

Keywords: LNP-mRNA-based vaccine; IS/ID and KPD models; Model structure comparison