Interpretable Machine Learning Models for Predicting Short-term Prognosis in AChR-Ab+ Generalized Myasthenia Gravis Using Clinical Features and Systemic Inflammation Index

Overview

Journal Front Neurol

Specialty Neurology

Date 2024 Oct 24

PMID 39445190

Authors

Yanan Xu

Qi Li

Meng Pan

Xiao Jia

Wenbin Wang

Qiqi Guo

Liqin Luan

Affiliations

Soon will be listed here.

Abstract

Background: Myasthenia Gravis (MG) is an autoimmune disease that causes muscle weakness in 80% of patients, most of whom test positive for anti-acetylcholine receptor (AChR) antibodies (AChR-Abs). Predicting and improving treatment outcomes are necessary due to varying responses, ranging from complete relief to minimal improvement.

Objective: Our study aims to develop and validate an interpretable machine learning (ML) model that integrates systemic inflammation indices with traditional clinical indicators. The goal is to predict the short-term prognosis (after 6 months of treatment) of AChR-Ab+ generalized myasthenia gravis (GMG) patients to guide personalized treatment strategies.

Methods: We performed a retrospective analysis on 202 AChR-Ab+ GMG patients, dividing them into training and external validation cohorts. The primary outcome of this study was the Myasthenia Gravis Foundation of America (MGFA) post-intervention status assessed after 6 months of treatment initiation. Prognoses were classified as "unchanged or worse" for a poor outcome and "improved or better" for a good outcome. Accordingly, patients were categorized into "good outcome" or "poor outcome" groups. In the training cohort, we developed and internally validated various ML models using systemic inflammation indices, clinical indicators, or a combination of both. We then carried out external validation with the designated cohort. Additionally, we assessed the feature importance of our most effective model using the Shapley Additive Explanations (SHAP) method.

Results: In our study of 202 patients, 28.7% (58 individuals) experienced poor outcomes after 6 months of standard therapy. We identified 11 significant predictors, encompassing both systemic inflammation indexes and clinical metrics. The extreme gradient boosting (XGBoost) model demonstrated the best performance, achieving an area under the receiver operating characteristic (ROC) curve (AUC) of 0.944. This was higher than that achieved by logistic regression (Logit) (AUC: 0.882), random forest (RF) (AUC: 0.917), support vector machines (SVM) (AUC: 0.872). Further refinement through SHAP analysis highlighted five critical determinants-two clinical indicators and three inflammation indexes-as crucial for assessing short-term prognosis in AChR-Ab+ GMG patients.

Conclusion: Our analysis confirms that the XGBoost model, integrating clinical indicators with systemic inflammation indexes, effectively predicts short-term prognosis in AChR-Ab+ GMG patients. This approach enhances clinical decision-making and improves patient outcomes.

References

Jiang T, Gradus J, Rosellini A . Supervised Machine Learning: A Brief Primer. Behav Ther. 2020; 51(5):675-687. PMC: 7431677. DOI: 10.1016/j.beth.2020.05.002. View

Zhong H, Ruan Z, Yan C, Lv Z, Zheng X, Goh L . Short-term outcome prediction for myasthenia gravis: an explainable machine learning model. Ther Adv Neurol Disord. 2023; 16:17562864231154976. PMC: 9969443. DOI: 10.1177/17562864231154976. View

Wu X, Wang H, Xie G, Lin S, Ji C . Increased systemic immune-inflammation index can predict respiratory failure in patients with Guillain-Barré syndrome. Neurol Sci. 2021; 43(2):1223-1231. DOI: 10.1007/s10072-021-05420-x. View

Jannusch K, Dietzel F, Bruckmann N, Morawitz J, Boschheidgen M, Minko P . Prediction of therapy response of breast cancer patients with machine learning based on clinical data and imaging data derived from breast [F]FDG-PET/MRI. Eur J Nucl Med Mol Imaging. 2023; 51(5):1451-1461. PMC: 10957677. DOI: 10.1007/s00259-023-06513-9. View

Nost T, Alcala K, Urbarova I, Smith Byrne K, Guida F, Sandanger T . Systemic inflammation markers and cancer incidence in the UK Biobank. Eur J Epidemiol. 2021; 36(8):841-848. PMC: 8416852. DOI: 10.1007/s10654-021-00752-6. View

Hu Y, Wang J, Rao J, Xu X, Cheng Y, Yan L . Comparison of peripheral blood B cell subset ratios and B cell-related cytokine levels between ocular and generalized myasthenia gravis. Int Immunopharmacol. 2020; 80:106130. DOI: 10.1016/j.intimp.2019.106130. View

Bi Y, Xiang D, Ge Z, Li F, Jia C, Song J . An Interpretable Prediction Model for Identifying N-Methylguanosine Sites Based on XGBoost and SHAP. Mol Ther Nucleic Acids. 2020; 22:362-372. PMC: 7533297. DOI: 10.1016/j.omtn.2020.08.022. View

Nohara Y, Matsumoto K, Soejima H, Nakashima N . Explanation of machine learning models using shapley additive explanation and application for real data in hospital. Comput Methods Programs Biomed. 2021; 214:106584. DOI: 10.1016/j.cmpb.2021.106584. View

Silvestri N, Wolfe G . Treatment-refractory myasthenia gravis. J Clin Neuromuscul Dis. 2014; 15(4):167-78. DOI: 10.1097/CND.0000000000000034. View

10.

Omar E, Mat H, Abd Karim A, Sanaudi R, Ibrahim F, Omar M . Comparative Analysis of Logistic Regression, Gradient Boosted Trees, SVM, and Random Forest Algorithms for Prediction of Acute Kidney Injury Requiring Dialysis After Cardiac Surgery. Int J Nephrol Renovasc Dis. 2024; 17:197-204. PMC: 11283789. DOI: 10.2147/IJNRD.S461028. View

11.

Wang Z, Yan Y . Immunopathogenesis in Myasthenia Gravis and Neuromyelitis Optica. Front Immunol. 2018; 8:1785. PMC: 5732908. DOI: 10.3389/fimmu.2017.01785. View

12.

Yoshimoto Y, Ishida S, Hosokawa T, Arawaka S . Assessment of clinical factors affecting outcome of myasthenia gravis. Muscle Nerve. 2021; 64(1):90-94. DOI: 10.1002/mus.27247. View

13.

Schiefer S, Wirsik N, Kalkum E, Seide S, Nienhuser H, Muller B . Systematic Review of Prognostic Role of Blood Cell Ratios in Patients with Gastric Cancer Undergoing Surgery. Diagnostics (Basel). 2022; 12(3). PMC: 8947199. DOI: 10.3390/diagnostics12030593. View

14.

Lefeuvre C, Payet C, Fayet O, Maillard S, Truffault F, Bondet V . Risk factors associated with myasthenia gravis in thymoma patients: The potential role of thymic germinal centers. J Autoimmun. 2019; 106:102337. DOI: 10.1016/j.jaut.2019.102337. View

15.

Qin B, Ma N, Tang Q, Wei T, Yang M, Fu H . Neutrophil to lymphocyte ratio (NLR) and platelet to lymphocyte ratio (PLR) were useful markers in assessment of inflammatory response and disease activity in SLE patients. Mod Rheumatol. 2015; 26(3):372-6. DOI: 10.3109/14397595.2015.1091136. View

16.

Chen L, Jiang J, Dou B, Feng H, Liu J, Zhu Y . Machine learning study of the extended drug-target interaction network informed by pain related voltage-gated sodium channels. Pain. 2023; 165(4):908-921. PMC: 11021136. DOI: 10.1097/j.pain.0000000000003089. View

17.

Liu L, Liu W, Jia Z, Li Y, Wu H, Qu S . Application of machine learning algorithms to predict lymph node metastasis in gastric neuroendocrine neoplasms. Heliyon. 2023; 9(10):e20928. PMC: 10622622. DOI: 10.1016/j.heliyon.2023.e20928. View

18.

Almasan O, Leucuta D, Hedesiu M . Blood Cell Count Inflammatory Markers as Prognostic Indicators of Periodontitis: A Systematic Review and Meta-Analysis. J Pers Med. 2022; 12(6). PMC: 9225277. DOI: 10.3390/jpm12060992. View

19.

Chen J, Yang L, Han J, Wang L, Wu T, Zhao D . Interpretable Machine Learning Models Using Peripheral Immune Cells to Predict 90-Day Readmission or Mortality in Acute Heart Failure Patients. Clin Appl Thromb Hemost. 2024; 30:10760296241259784. PMC: 11146004. DOI: 10.1177/10760296241259784. View

20.

Ding R, Deng M, Wei H, Zhang Y, Wei L, Jiang G . Machine learning-based prediction of clinical outcomes after traumatic brain injury: Hidden information of early physiological time series. CNS Neurosci Ther. 2024; 30(7):e14848. PMC: 11228354. DOI: 10.1111/cns.14848. View