Un método de cribado sin marcadores y multiparamétrico para predecir la estabilidad de las formulaciones de ARNm-LNP

With the rapid advancement of messenger RNA (mRNA) technologies, lipid nanoparticles (LNPs) have been widely adopted for vaccine and gene therapy applications. The stability of mRNA-LNP formulations is a critical determinant of their therapeutic efficacy; however, conventional stability prediction methods are often time-consuming and operationally complex. In this study, a label-free and real-time multi-parameter thermal profiling method was established to provide a rapid assessment of colloidal stability, simultaneously monitoring turbidity, cumulant radius (r_h), and scattering. Six formulations sharing the same mRNA cargo and lipid composition but differing in buffer, preservative, and pH were evaluated to examine the discriminatory power of this assay. A redefined aggregation-onset temperature (T_agg) was introduced as the primary thermal stability readout and combined with Δr_h to capture early structural perturbations under thermal stress, thereby enabling efficient parallel screening during early formulation development. Compared with conventional accelerated stability studies, this method markedly shortened the evaluation period while still capturing thermally induced colloidal-stability changes. Using T_agg and Δr_h in a logistic regression model, thermal-profiling-based "Qualified/Not Qualified" calls demonstrated good concordance with 40 °C accelerated stability outcomes, supporting the use of this strategy as an early-stage screening tool within this mRNA-LNP platform. Nevertheless, its broader applicability across different cargos and lipid chemistries will require further validation.