Development of a Risk Stratification Nomogram for Predicting Generalization in Adult Ocular Myasthenia Gravis: A Multicenter Study

Purpose: The aim of this study is to investigate factors influencing the progression of ocular myasthenia gravis (OMG) to generalized myasthenia gravis (GMG) in adults and to construct a multidimensional risk stratification nomogram to identify individuals at high risk.Methods: A retrospective analysis was conducted on clinical data from 322 adults diagnosed with OMG. Feature selection utilized the Boruta algorithm and Lasso regression. Independent predictors of generalization were identified through Cox regression, and a nomogram was developed. Model performance was internally validated using the Bootstrap method. Population stratification was performed with the X-tile program, based on scores derived from the nomogram.Results: Multivariate Cox regression analysis identified sex, age at onset, body mass index (BMI), antibody type, Repetitive Nerve Stimulation (RNS), thymic status, thymectomy, concomitant autoimmune diseases, the Charlson Comorbidity Index (CCI), and treatment status as significant independent predictors of generalization in adult OMG. The nomogram developed from these ten factors demonstrated lower Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC) values compared to any individual factor. The bootstrap-corrected concordance index (C-index) for the nomogram was 0.802 (95% CI 0.771, 0.833). The calibration curve showed a strong correlation between predicted and observed values, supporting the model's accuracy. Decision curve analysis confirmed its potential clinical utility. Kaplan-Meier curve analysis demonstrated that the nomogram effectively distinguishes among low-, medium-, and high-risk groups for generalization in adult OMG (P<0.0001), validating its efficacy in predicting clinical outcomes.