The Value of Combined Detection of Systemic Inflammation Response Index and Prognostic Nutritional Index in Predicting Short-Term Prognosis of Myasthenia Gravis.

Purpose: Although systemic inflammation response index (SIRI) and prognostic nutritional index (PNI) are associated with prognosis in various diseases, their role in myasthenia gravis (MG) remains unclear. This study aims to evaluate the predictive value of SIRI combined with PNI for MG prognosis. Methods: 260 MG patients were enrolled in this retrospective study and were categorized into clinical improvement and non-improvement groups based on changes in MG-ADL and QMG scores after 6 months' treatment. Lymphocyte-to-monocyte ratio (LMR), platelet-to-lymphocyte ratio (PLR), neutrophil-to-lymphocyte ratio (NLR), SIRI and PNI were calculated from admission blood indices. Clinical differences between groups were compared. Logistic regression was used to identify independent predictors of clinical non-improvement. The ROC curve was utilized to assess the prognostic predictive value of SIRI, PNI, and their combination. Interaction effects and stratified analyses were used to explore the relationship between SIRI, PNI and MG prognosis across distinct subgroups. Results: Patients without clinical improvement exhibited significantly elevated SIRI, NLR, and PLR, whereas LMR and PNI were reduced (p < 0.001). Multivariate logistic regression demonstrated that both SIRI and PNI significantly predicted clinical non-improvement (OR = 9.108, 95% CI: 3.412-24.317, p < 0.001; OR = 0.695, 95% CI: 0.601-0.804, p < 0.001). The area under the curve (AUC) of SIRI combined with PNI for predicting clinical non-improvement in MG was 0.928 (95% CI:0.896-0.961, sensitivity: 0.873, specificity: 0.851), which is higher than SIRI (AUC: 0.841, 95% CI: 0.783-0.899, sensitivity: 0.772, specificity: 0.845) and PNI (AUC: 0.822, 95% CI: 0.770-0.875, sensitivity: 0.759, specificity: 0.740) alone. A statistically significant interaction was identified between SIRI and thymoma (p = 0.009). Conclusion: SIRI and PNI are independently associated with MG prognosis, particularly in thymoma cases, where SIRI shows a stronger correlation. Furthermore, the combination of SIRI and PNI can serve as a valuable predictor of clinical non-improvement in MG.