AB0228 COMPREHENSIVE RHEUMATOID HAND ASSESSMENT THROUGH PATTERN OF DEFORMITIES USING CLUSTER ANALYSIS

Background: The treatment of rheumatoid hand, which is characterized by thumb deformity, finger deformities, and ulnar drift (UD), is challenging. Its pathophysiology is complex, and a comprehensive understanding of the optimal intervention for this condition requires high technical skill and extensive clinical experience. Moreover, the natural course of rheumatoid hand itself remains unclear. Objectives: This study was performed to comprehensively evaluate rheumatoid hand through the specific parameters of each deformity. Methods: A rheumatoid hand cohort was established in 2004. In total, 134 hands of 67 patients were registered and underwent clinical evaluations. All hands surgically treated during follow-up were excluded from the study, but the contralateral hands were assessed. Evaluations were repeated in 2009 (100 hands of 52 patients) and in 2015 (63 hands of 37 patients) among all available patients. Therefore, among the data obtained from the 3 study endpoints, 297 hands were available for the cross-sectional analysis and 43 hands were available for the longitudinal analysis. Thumb deformities and finger deformities (swan-neck and boutonniere) were semi-quantitated by the Nalebuff classification score, and UD was quantified using a metacarpophalangeal joint condition scoring method1). A two-step cluster analysis was performed with entered parameters, and the distribution of each parameter was considered to clarify the characteristics of each cluster. The hands with different clusters at each endpoint were recruited for the following longitudinal analysis. The natural course of rheumatoid hand was considered based on the cluster change. Results: Seven clusters were used in this study to emphasize the impact of thumb deformity on function. The characteristics of each cluster were as follows. Cluster 1: mild finger deformities and various severities of UD; Cluster 2: type 1 thumb deformity and various severities of UD; Cluster 3: type 2 thumb deformity and severe UD; Cluster 4: type 3 or 4 thumb deformity, low or moderate level of swan-neck deformity, and various severities of UD; Cluster 5: various types of thumb deformity, severe boutonniere deformity, and various severities of UD; Cluster 6: type 1 thumb deformity, severe swan-neck deformity, and various severities of UD; and Cluster 7: type 6 thumb deformity. The longitudinal analysis showed that Cluster 1 mainly changed to Cluster 2 or 4, indicating progression of thumb deformity. Cluster 2 changed to Cluster 3, indicating that thumb type 1 progressed to type 2 (Figure 1). When the affected period was shorter than 10 years, the incidence of severe hand deformity (including two or more affected joint areas and low hand function) was 30% (Figure 2). Conclusion: This study suggests the presence of seven patterns of deformity enabling a comprehensive understanding of rheumatoid hand. Furthermore, the results of the longitudinal analysis suggest a natural course of rheumatoid hand progression. Therefore, from the distribution of parameters of each deformity and its severity, rheumatologists can easily classify rheumatoid hand and determine its pathophysiology to choose the most effective intervention. References: [1]Toyama S, Oda R, Tokunaga D et al. A new assessment tool for ulnar drift in patients with rheumatoid arthritis using pathophysiological parameters of the metacarpophalangeal joint. Modern rheumatology 2019, 29: 113-8. Acknowledgments: This work was supported by JSPS KAKENHI Grant Numbers JP19K19914. Disclosure of Interests: None declared