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.

2022 update of the Austrian Society of Rheumatology and Rehabilitation nutrition and lifestyle recommendations for patients with gout and hyperuricemia.

BACKGROUND: Gout is the most frequent inflammatory joint disease in the western world and has a proven genetic background. Additionally, lifestyle factors like increasing life span and wealth, sufficient to excess nutritional status and a growing prevalence of obesity in the population, as well as e.g. alcohol consumption contribute to the rising incidence of hyperuricemia and gout. Apart from an adequate medication, medical advice on nutrition and lifestyle is an essential part of the management of gout patients, being at high risk of internal comorbidities. OBJECTIVE: In 2015, the ÖGR (Österreichische Gesellschaft für Rheumatologie und Rehabilitation) working group for osteoarthritis and crystal arthropathies already published nutrition and lifestyle recommendations for patients with gout and hyperuricemia. Since then, a multitude of literature has been published addressing this topic, what required an update. METHODS: First, the authors performed a hierarchical literature search to screen the meanwhile published literature. Also considering references of the first publication, the relevant literature was selected, and the 2015 recommendations were either kept as published, reformulated or newly produced. Finally, the evidence level and the level of agreement with each recommendation were added. RESULTS: Following this process, ten recommendations were generated instead of the initial nine. Like in the original publication, a colored icon presentation was provided to complement the written text. CONCLUSION: The Austrian nutrition and lifestyle recommendations for patients with gout and hyperuricemia were updated incorporating the most recent relevant literature, serving as education material for patients and updated information for physicians.

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.

Erythropoietin requirement in patients with type 2 diabetes mellitus on maintenance hemodialysis therapy.

Diabetes is known to be a risk factor for the severity of anemia in non-dialyzed patients with renal failure. The aim of this study was to evaluate differences in hemoglobin (Hb) response to erythropoietin (EPO) in diabetic and nondiabetic patients on chronic hemodialysis (CHD). Sixty-four patients on CHD were included in the study: 24 type 2 diabetics (mean age, 59+/-11 years; 10 men, 14 women) and 40 nondiabetics (age, 53+/-14 years; 21 men, 19 women). All patients received a fixed dose of 50 mg ferric saccharate and EPO per week, dosed individually to achieve a target Hb level of 12 g/dl. Hb levels, ferritin, transferrin saturation (TSAT), EPO requirement (IU/kg/week), folic acid, vitamin B12 and C-reactive protein (CRP) were measured every two months. Additionally, the incidence of infectious diseases during the observation period of six months was evaluated, and a univariate correlation analysis of CRP and EPO requirements was performed in both groups. Patients with and without diabetes were divided into two groups each: those with normal CRP and those with elevated CRP. The EPO requirements of these groups were compared. Under identical iron substitution the mean Hb level increased more, but not significantly, in non-diabetic patients than in diabetic patients. After 6 months the mean Hb levels were 12.1+/-1.2 versus 11.5+/-1.2 g/dl (NS), although the actual EPO requirement was higher in diabetic than in non-diabetic subjects (244+/-122 versus 183+/-118 IU/kg/week; p<0.05). CRP after 6 months was significantly higher in diabetic than in non-diabetic patients (2.6+/-2.2 versus 1.5+/-1.3 mg/dl; p<0.05), as was the incidence of infectious disease (n/patient/month) (0.24 versus 0.08; p<0.05). The correlation coefficient between CRP and EPO requirements was statistically significant in both diabetic (r=0.547 p<0.01) and non-diabetic subjects (r=0.577; p<0.001). All other laboratory indices were similar in both groups. In the diabetic patients with normal CRP (n=6) the Hb levels achieved after six months were similar to those of non-diabetic patients (n=10) with normal CRP (11.9+/-1.1 versus 12.1+/-1.2%), and the required EPO was comparable. We conclude that the Hb response to EPO is reduced in diabetic patients on CHD. This elevated EPO requirement may be explained by a greater prevalence of infectious diseases, characterized by a significantly higher CRP level, in these patients. Other causes for the elevated EPO requirement could be excluded.

Predialysis management and predictors for early mortality in uremic patients who die within one year after initiation of dialysis therapy.

Despite improvements in dialysis therapy, the mortality rate of patients with end stage renal disease (ESRD) has remained high. A relatively high proportion of uremic patients dies within one year after the initiation of dialysis treatment. The aim of this study was to evaluate predictors for this early mortality in patients with ESRD. A total of 66 uremic patients were included in the study. Patients were divided in those who survived < 1 year (n = 17) and those who survived > or = 1 year (n = 49). We compared the prevalence of diabetes and hypertension and of vascular diseases as well as the prevalence of heart insufficiency (EF < 30%) and left ventricular hypertrophy (LVH). Additionally, we estimated the laboratory parameters serum creatinine, creatinine clearance, BUN, cholesterol, triglycerides, fibrinogen, serum protein, serum albumin and hemoglobin, and evaluated the indications for the initiation of dialysis therapy in both patient groups. The patients with survival < 1 year were significantly older (64+/-12 vs. 54+/-14 years, p<0.01) and showed a lower BMI (22+/-3 vs. 25+/-3, p<0.01) than those who survived > 1 year. The prevalence of diabetes (70% vs. 31%, p<0.05), cardiac insufficiency (70% vs. 16%, p<0.025), cardiovascular disease (65% vs. 28%, p<0.05) and peripheral vascular diseases (70% vs. 28%, p<0.05) was significantly higher in the patients with early mortality. The prevalence of hypertension was similar in both groups, however, the prevalence of LVH was significantly higher in the patients who survived < 1 year (88% vs. 37%, p<0.05). Laboratory parameters were not significantly different in the two groups of patients, with the exception of serum albumin, which was significantly lower in the patients with early mortality (3.5+/-0.6 vs. 3.9+/-0.4 g/l, p<0.02). Hyperhydration was the most common indication for the start of dialysis in patients with early mortality (59% vs. 13%, p<0.025). Cardiac insufficiency was the most common cause of death in these subjects (n = 10, 59%). Six individuals (12%) died within four weeks after initiating dialysis therapy. Thus, there are several predictors for early mortality in end-stage renal disease patients, including high age, low BMI, the presence of diabetes, coronary heart disease, heart insufficiency and LVH, as well as low serum albumin levels. A relatively high percentage of patients die shortly after the start of dialysis therapy. Heart insufficiency is the most common cause of early death in these patients.