A translational mathematical model linking systemic biomarkers to disease recurrence in diabetic macular edema: a proof-of-concept analysis

This study aimed to develop a translational mathematical model that links systemic biomarkers to diabetic macular edema (DME) dynamics and to provide a proof‑of‑concept assessment of its plausibility for informing anti‑vascular endothelial growth factor (anti‑VEGF) treatment strategies. A hybrid modeling approach was employed, combining mechanistic reasoning with a retrospective analysis of DRCR.net Protocol H data. A mechanistic model formalized the biological interaction between systemic drivers -glycated hemoglobin (HbA1c), systolic blood pressure (BP), and glycemic variability- and retinal VEGF dynamics, while an empirical penalty model translated these drivers into a framework for estimating personalized injection intervals (I_model). Using data from 97 eyes treated with bevacizumab, we assessed the association between Imodel and a theoretical time to disease recurrence (T_rec), defined as a >10% increase in central subfield thickness or a >5‑letter loss in visual acuity. The model’s hierarchical structure, which prioritized HbA1c, produced a shorter or equivalent interval compared to T_rec in 63.6% of the overall cohort, and in a targeted subgroup where the model’s logic was fully applicable, the concordance rate was 77.6%. In Group E, simulating a post‑loading phase with two injections plus laser, the mean deviation between I_model and T_rec was minimal (-0.1 weeks) with reduced variability (SD=5.6 weeks). Sensitivity analysis confirmed the designed hierarchy, showing HbA1c exerted a 36‑fold greater influence on the model’s output than BP. These findings present a novel mathematical framework linking systemic metabolic and hemodynamic control to DME progression. The proof‑of‑concept analysis supports the biological plausibility of using HbA1c and BP to stratify recurrence risk and provides a transparent, interpretable foundation for future research. This model bridges systemic and ocular care and proposes a structured hypothesis for personalized treatment scheduling that warrants prospective validation in appropriately designed clinical cohorts.