E. Prokhorova, E. Zhilyaev, G. Zhilyaev
2021 Annals of the Rheumatic Diseases  
Background:Joint bone scintigraphy is a promising method for diagnosing of inflammatory joint diseases. However, its scintigraphic semiotics remains to be developed.Objectives:To develop an algorithm for the differential diagnosis of the most common arthropathies based on quantitative osteoscintigraphy.Methods:This single-center study included inpatients with clinical diagnoses of rheumatoid arthritis (RA), axial spondyloarthritis (including ankylosing spondylitis - aSPA), peripheral
more » ... hritis (including reactive arthritis - pSpA), psoriatic arthritis (PsA), and osteoarthritis (ОА) established by a rheumatologist and meeting the relevant criteria, Three hours after the injection of pyrophosphate, labeled with Tc-99m, scintigraphy of the skeletal bones was carried out according to the "whole body" program. The joint / bone accumulation ratio (AR) was calculated as the ratios of counts in relevant areas. All AR were recalculated into T-score for each joint based on data from control group. The CHAID algorithm for classification trees constructing was used. The significance of the division in the nodes was estimated with Bonferroni adjustment.Results:266 patients were included in the study aged 46.6±14.3 years, men - 134 (50.4%). aSPA was diagnosed in 40 patients, pSpA in 87, RA in 45, ОА in 68, PsA in 26. 2279 joints were analyzed. A classification tree for differential diagnosis of arthropathies has been built (Pict.). Key indicators for identifying subgroups in the algorithm: AR in the wrist, knee and hip joints.At the first step of the classification tree, the sum of the T-scores of the wrist joints is determined. If this amount exceeds 7.76 (node 2), the most likely diagnosis is RA (58.5% of patients in the subgroup of this node). Patients with the sum of the T-scores in the wrist joints less than 7.76 (node 1), in turn, were divided depending on the sum of the T-scores of hip joints. Those with this value less than 3.25 (node 3), mainly suffered from OA (48.7%), less often in this subgroup were diagnosed pSpA (16.7%) and PsA (14.1%). Patients in whom the sum of the T-score in the hips was more than 3.245 (node 4) were further subdivided depending on the sum of the T-scores in the shoulder joints. If it was more than 4.21 (node 8), then pSpA was most likely (58.7%). If the sum of the T-scores of the shoulder joints was in the range from 2.05 to 4.21 (node 7), then these are mainly patients with spondyloarthritides (56.2% with aSPA and 31.2% with pSpA) and a small number of patients (10%) with PsA. In the subgroup with the sum of T-scores of the shoulder joints <2.05 (node 6), patients with PSA (40%) prevailed, with a significant proportion of persons with OA (27.5%). Thus, out of the final branches formed by the tree (node 2, 3, nodes 6-8), four correspond to certain diseases (node 2 - PA, node 3 - OA, node 7 - aSPA, node 8 - pSpA). Node 6 includes patients with various diseases with a predominance of PsA. The level of significance of differences between the formed groups at all branch points of the tree is p = 0.001 or less (Bonferroni adjusted). In the training sample, 51.5% of observations are correctly classified. According to cross-validation data, the expected rate of correct classifications in real application of the algorithm is 38.0%.Conclusion:An algorithm for the differential diagnosis of the most common inflammatory diseases of the joints has been developed, which makes it possible to use the data of quantitative osteoscintigraphy in the process of diagnosing arthritis.Acknowledgements:MD. PhD Constantin V. Kushnir, Main hospital of police.Disclosure of Interests:None declared
doi:10.1136/annrheumdis-2021-eular.2931 fatcat:pzgoinh3qraxrfzwup5llzrg24