A NEW scoping review suggests that machine learning models may help predict treatment response to biologic and targeted synthetic disease-modifying antirheumatic drugs in rheumatoid arthritis, potentially supporting more personalised treatment strategies in clinical practice.
Biologic and targeted synthetic DMARDs have transformed outcomes for many patients with rheumatoid arthritis, but response remains highly variable. Identifying which patients are most likely to benefit from a given therapy continues to be a major challenge. Machine learning approaches offer a potential solution by integrating complex clinical, laboratory, and patient-reported data to generate predictive models, yet evidence in this area has remained fragmented.
Mapping the Evidence Base
To address this gap, researchers conducted a scoping review following PRISMA-ScR guidelines, analysing studies that applied machine learning methods to predict treatment response to biologic or targeted synthetic DMARDs in rheumatoid arthritis. Searches of PubMed, MEDLINE, and Embase identified 24 eligible studies published up to March 2024.
Most studies relied on real-world data sources, particularly national or regional disease registries, while others used electronic health records. Sample sizes varied widely, ranging from fewer than 50 patients to more than 7,000, reflecting substantial heterogeneity in study design and data availability.
Model Performance and Key Predictors
A range of machine learning techniques were employed, with boosted decision trees, random forests, support vector machines, and regularised regression models used most frequently. Outcomes commonly assessed included remission, low disease activity, and treatment non-response.
Across studies, model performance was variable, with reported area under the curve values ranging from 0.54 to 0.92, and a mean AUC of 0.71. Boosted trees and neural networks tended to demonstrate the strongest predictive performance. Common predictors included baseline disease activity, inflammatory biomarkers, functional status, and patient-reported outcomes.
However, external validation was uncommon, reported in fewer than one in five studies, limiting confidence in generalisability. While most studies were assessed as having low-to-moderate risk of bias, reporting quality and methodological consistency varied considerably.
Barriers to Clinical Translation
The authors conclude that machine learning holds promise for predicting treatment response in rheumatoid arthritis, but significant barriers remain before these tools can be implemented in routine care. Greater standardisation of outcome definitions, improved transparency in reporting, and robust external validation are needed to ensure reliability and clinical utility.
As interest in precision medicine continues to grow, future work integrating machine learning into prospective studies and clinical trials may help translate these predictive models into practical decision-support tools for rheumatologists.
Reference
Eriakha EB et al. Machine learning for predicting treatment response to biologic and targeted synthetic disease-modifying antirheumatic drugs in rheumatoid arthritis: a scoping review. BMC Rheumatol. 2025;DOI: 10.1186/s41927-025-00584-x.
