Unsupervised cluster analysis of clinical and ultrasound features reveals unique gout subtypes: Results from the Egyptian College of Rheumatology (ECR).

BACKGROUND: Gout comprises a heterogeneous group of disorders; however, comorbidities have been the focus of most efforts to classify disease subgroups. OBJECTIVES: We applied cluster analysis using musculoskeletal ultrasound (MSUS) combined with clinical and laboratory findings in patients with gout to identify disease phenotypes, and differences across clusters were investigated. PATIENTS AND METHODS: Patients with gout who complied with the ACR/EULAR classification criteria were enrolled in the Egyptian College of Rheumatology (ECR)-MSUS Study Group, a multicenter study. Selected variables included demographic, clinical, and laboratory findings. MSUS scans assessed the bilateral knee and first metatarsophalangeal joints. We performed a K-mean cluster analysis and compared the features of each cluster. RESULTS: 425 patients, 267 (62.8 %) males, mean age 54.2 ± 10.3 years were included. Three distinct clusters were identified. Cluster 1 (n = 138, 32.5 %) has the lowest burden of the disease and a lower frequency of MSUS characteristics than the other clusters. Cluster 2 (n = 140, 32.9 %) was mostly women, with a low rate of urate-lowering treatment (ULT). Cluster 3 (n = 147, 34.6 %) has the highest disease burden and the greatest proportion of comorbidities. Significant MSUS variations were found between clusters 2 and 3: joint effusion (p < 0.0001; highest: cluster 3), power Doppler signal (p < 0.0001; highest: clusters 2), and aggregates of crystal deposition (p < 0.0001; highest: cluster 3). CONCLUSION: Cluster analysis using MSUS findings identified three gout subgroups. People with more MSUS features were more likely to receive ULT. Treatment should be tailored according to the cluster and MSUS features.

Development of machine learning models for detection of vision threatening Behçet's disease (BD) using Egyptian College of Rheumatology (ECR)-BD cohort.

BACKGROUND: Eye lesions, occur in nearly half of patients with Behçet's Disease (BD), can lead to irreversible damage and vision loss; however, limited studies are available on identifying risk factors for the development of vision-threatening BD (VTBD). Using an Egyptian college of rheumatology (ECR)-BD, a national cohort of BD patients, we examined the performance of machine-learning (ML) models in predicting VTBD compared to logistic regression (LR) analysis. We identified the risk factors for the development of VTBD. METHODS: Patients with complete ocular data were included. VTBD was determined by the presence of any retinal disease, optic nerve involvement, or occurrence of blindness. Various ML-models were developed and examined for VTBD prediction. The Shapley additive explanation value was used for the interpretability of the predictors. RESULTS: A total of 1094 BD patients [71.5% were men, mean ± SD age 36.1 ± 10 years] were included. 549 (50.2%) individuals had VTBD. Extreme Gradient Boosting was the best-performing ML model (AUROC 0.85, 95% CI 0.81, 0.90) compared with logistic regression (AUROC 0.64, 95%CI 0.58, 0.71). Higher disease activity, thrombocytosis, ever smoking, and daily steroid dose were the top factors associated with VTBD. CONCLUSIONS: Using information obtained in the clinical settings, the Extreme Gradient Boosting identified patients at higher risk of VTBD better than the conventional statistical method. Further longitudinal studies to evaluate the clinical utility of the proposed prediction model are needed.