Prediction of ineffectiveness of biological drugs using machine learning and explainable AI methods: data from the Austrian Biological Registry BioReg.

OBJECTIVES: Machine learning models can support an individualized approach in the choice of bDMARDs. We developed prediction models for 5 different bDMARDs using machine learning methods based on patient data derived from the Austrian Biologics Registry (BioReg). METHODS: Data from 1397 patients and 19 variables with at least 100 treat-to-target (t2t) courses per drug were derived from the BioReg biologics registry. Different machine learning algorithms were trained to predict the risk of ineffectiveness for each bDMARD within the first 26 weeks. Cross-validation and hyperparameter optimization were applied to generate the best models. Model quality was assessed by area under the receiver operating characteristic (AUROC). Using explainable AI (XAI), risk-reducing and risk-increasing factors were extracted. RESULTS: The best models per drug achieved an AUROC score of the following: abatacept, 0.66 (95% CI, 0.54-0.78); adalimumab, 0.70 (95% CI, 0.68-0.74); certolizumab, 0.84 (95% CI, 0.79-0.89); etanercept, 0.68 (95% CI, 0.55-0.87); tocilizumab, 0.72 (95% CI, 0.69-0.77). The most risk-increasing variables were visual analytic scores (VAS) for abatacept and etanercept and co-therapy with glucocorticoids for adalimumab. Dosage was the most important variable for certolizumab and associated with a lower risk of non-response. Some variables, such as gender and rheumatoid factor (RF), showed opposite impacts depending on the bDMARD. CONCLUSION: Ineffectiveness of biological drugs could be predicted with promising accuracy. Interestingly, individual parameters were found to be associated with drug responses in different directions, indicating highly complex interactions. Machine learning can be of help in the decision-process by disentangling these relations.

Validity of data collected in BIOREG, the Austrian register for biological treatment in rheumatology: current practice of bDMARD therapy in rheumatoid arthritis in Austria.

BACKGROUND: The purpose of the present study was to check the validity of data collected in BIOREG, the Austrian register for biological treatment in rheumatology, and to elucidate eventual differences with respect to disease activity (DA) in patients with rheumatoid arthritis (RA) on established biological DMARDs (bDMARDs) before inclusion into the register (EST) and beginners at the time point of inclusion (NEW) after 1 year of treatment. METHODS: RA patients with a complete follow-up of 1 year in BIOREG were divided into EST and NEW and compared with respect to DA, remission rates, concomitant synthetic DMARDs (csDMARDs) and glucocorticoid therapy (GC) at baseline and after 1-year follow-up. Safety concerns are listed. Descriptive statistics are applied. RESULTS: For 346 RA patients (284 EST, 62 NEW) out of 970 RA patients included into BIOREG, a full data set for a 1-year follow-up was available. No differences in DA were observed after 1 year as expressed by DAS28 or RADAI-5, and small differences as expressed by remission rates according to DAS28, RADAI-5 or Boolean criteria (namely approximately 1/2, 1/3 to 1/4 and 1/4 to 1/5 of the patients respectively). Sixty-four adverse events (AEs) were noted in 56 (20 %) of EST and 20 in 19 (31 %) of NEW patients. Malignancy occurred in four patients. After 1 year, 48 % of EST patients but only 16 % of NEW patients were on bDMARD monotherapy. CONCLUSION: Regarding DA, the date collected in BIOREG appeared to be valid. After 1 year of bDMARD therapy, all patients, whether EST or NEW, achieved a similar level of DA. AEs occurred more frequently during the early phase of bDMARD treatment. Austrian rheumatologists initiate bDMARD therapy in patients with lower disease levels than in other European countries, leading to high remission rates.