Establishment and verification of a nomogram and a preliminary study on predicting the clinical response of conventional synthetic disease-modifying antirheumatic drugs (csDMARDs) in rheumatoid arthritis patients

Machine learning-based remission prediction in rheumatoid arthritis patients treated with biologic disease-modifying anti-rheumatic drugs: findings from the Kuwait rheumatic disease registry

Background

Rheumatoid arthritis (RA) is a common condition treated with biological disease-modifying anti-rheumatic medicines (bDMARDs). However, many patients exhibit resistance, necessitating the use of machine learning models to predict remissions in patients treated with bDMARDs, thereby reducing healthcare costs and minimizing negative effects.

Objective

The study aims to develop machine learning models using data from the Kuwait Registry for Rheumatic Diseases (KRRD) to identify clinical characteristics predictive of remission in RA patients treated with biologics.

Methods

The study collected follow-up data from 1,968 patients treated with bDMARDs from four public hospitals in Kuwait from 2013 to 2022. Machine learning techniques like lasso, ridge, support vector machine, random forest, XGBoost, and Shapley additive explanation were used to predict remission at a 1-year follow-up.

Results

The study used the Shapley plot in explainable Artificial Intelligence (XAI) to analyze the effects of predictors on remission prognosis across different types of bDMARDs. Top clinical features were identified for patients treated with bDMARDs, each associated with specific mean SHAP values. The findings highlight the importance of clinical assessments and specific treatments in shaping treatment outcomes.

Conclusion

The proposed machine learning model system effectively identifies clinical features predicting remission in bDMARDs, potentially improving treatment efficacy in rheumatoid arthritis patients.

Galectin-9 as an indicator of functional limitations and radiographic joint damage in patients with rheumatoid arthritis

Background

Previous studies have revealed that Galectin-9 (Gal-9) acts as an apoptosis modulator in autoimmunity and rheumatic inflammation. In the present study, we investigated the potential role of Gal-9 as a biomarker in patients with rheumatoid arthritis (RA), especially as an indicator of functional limitations and radiographic joint damage.

Methods

A total of 146 patients with RA and 52 age- and sex-matched healthy controls were included in this study. Clinical data including disease activity, physical function, and radiographic joint damage were assessed. Functional limitation was defined as the Stanford Health Assessment Questionnaire (HAQ) disability index >1. Subjects with joint erosion >0 or joint space narrowing >0 were considered to have radiographic joint damage. Serum Gal-9 levels were detected by an enzyme-linked immunosorbent assay. Univariate and multivariate logistic regression analysis were used to evaluate the association between Gal-9 and high disease activity and functional limitations, and a prediction model was established to construct predictive nomograms.

Results

Serum levels of Gal-9 were significantly increased in patients with RA compared to those in healthy controls (median 13.1 ng/mL vs. 7.6 ng/mL). Patients with RA who were older (>65 years), had a longer disease duration (>5 years), longer morning stiffness (>60mins), elevated serum erythrocyte sedimentation rate and C-reactive protein, and difficult-to-treat RA had significantly higher Gal-9 levels than those in the corresponding control subgroups (all p <0.05). Patients with RA were divided into two subgroups according to the cut-off value of Gal-9 of 11.6 ng/mL. Patients with RA with Gal-9 >11.6 ng/mL had a significantly higher core clinical disease activity index, HAQ scores, Sharp/van der Heijde modified Sharp scores, as well as a higher percentage of advanced joint damage (all p<0.05) than patients with Gal-9 ≤11.6 ng/mL. Accordingly, patients with RA presenting either functional limitations or radiographic joint damage had significantly higher serum Gal-9 levels than those without (both p <0.05). Furthermore, multivariate logistic regression analysis showed that a serum level of Gal-9 >11.6 ng/mL was an independent risk factor for high disease activity (OR=3.138, 95% CI 1.150-8.567, p=0.026) and presence of functional limitations (OR=2.455, 95% CI 1.017-5.926, p=0.046), respectively.

Conclusion

Gal-9 could be considered as a potential indicator in patients with RA, especially with respect to functional limitations and joint damage.

Knockdown of Galectin-9 alleviates rheumatoid arthritis through suppressing TNF-α-induced activation of fibroblast-like synoviocytes

The role of Galectin-9 (Gal-9) in the pathogenesis of rheumatoid arthritis (RA) remains unclear. This study aimed to investigate the mechanism of action and therapeutic potential of Gal-9 in RA. We detected Gal-9 expression in clinical samples, explored the mechanism of function of Gal-9 by knockdown and overexpression in fibroblast-like synoviocytes (FLSs), and further verified it in collagen-induced arthritis (CIA) model. We found that the levels of Gal-9 were considerably elevated in RA synovium than in osteoarthritis (OA) patients. A substantial decrease of Gal-9 was demonstrated after tumor necrosis factor (TNF-α) inhibitor treatment in the plasma of patients with RA. Additionally, transcriptome sequencing revealed that Gal-9 was involved in the regulation of the TNF-α pathway. Gal-9 was considerably upregulated after TNF-α stimulation in FLSs, and knockdown of Gal-9 substantially inhibited TNF-α activated proliferation, migration and inflammatory response. According to cell transcriptome sequencing results, we further confirmed that Gal-9 could achieve these effects by interacting with MAFB and affecting PI3K/AKT/mTOR pathway. Finally, we knocked down Gal-9 on the CIA model and found that it could alleviate the progression of arthritis. In conclusion, our study revealed that the knockdown of Gal-9 could inhibited TNF-α induced activation in RA through MAFB, PI3K/AKT/mTOR.