Biomarkers of mRNA vaccine efficacy derived from mechanistic modeling of tumor-immune interactions

Abstract

The success of mRNA vaccines against infectious diseases such as COVID-19 has opened new avenues for their application in oncology. In cancer immunotherapy, mRNA vaccines-typically encapsulated in lipid nanoparticles (LNPs) 100-200 nm in size-enable delivery of tumor-specific antigens to activate immune responses. Here, we investigated the efficacy of mRNA vaccines in cancer by modeling tumor-immune interactions and tumor microenvironment (TME) dynamics to identify predictive biomarkers. Using a mechanistic mathematical model, we simulated tumor growth, immune cell dynamics, and vaccine pharmacokinetics in virtual cohorts of 1,635 patients generated via Latin hypercube sampling. Our simulations demonstrated a 45% average tumor size reduction and a 60% increase in CD8 + T cell infiltration in responsive tumors. Multiple regression analyses validated the predictive power of both pre- and on-treatment biomarkers. Key predictors of vaccine efficacy included antigen-presenting cell (APC) density and cytotoxic T cell fraction. Specifically, an APC density above 500 cells/mm³ in lymph nodes correlated with a 55% increase in vaccine response rates, while a cytotoxic T cell fraction above 20% in tumors was associated with a 60% reduction in tumor volume. A reduced M2/M1 macrophage ratio further improved treatment outcomes by 50%, highlighting the role of reprograming immunosuppressive macrophages. TME characteristics significantly influenced vaccine efficacy. Low extracellular matrix (ECM) density-modeled as a 5-10 × increase in hydraulic conductivity-combined with medium cytokine levels (IL-2 and TNF-α at 10-50 pg/ml), created optimal conditions for immune activation. Under these conditions, vaccine uptake improved by 35% and cytotoxic T cell infiltration increased by 65%, resulting in up to a 50% improvement in therapeutic outcomes. Model predictions aligned with pre-clinical data from melanoma and breast cancer models. These findings provide a framework for optimizing mRNA vaccine strategies and advancing personalized cancer immunotherapy.