Article Text
Abstract
Background Axial Spondyloarthritis (axSpA) is characterised by progressive loss of spinal mobility, due to inflammation and structural damage. Conventional mobility tests have been used alongside radiographic structural damage and disease activity scores in order to explore the overall relationship between spinal mobility, structural damage and disease activity. This has rarely been done at the segmental level e.g. at the level of the lumbar spine only. The ViMove inertial motion sensor-based system allows segmental spinal mobility to be directly measured, namely at the lumbar spine level.
Objectives To analyse the relationship between mobility, structural damage and disease activity in the lumbar spine of patients with axSpA.
Methods Lumbar spinal mobility was measured using the the ViMove system with a pair of sensors located at L1 and the sacrum. Lumbar ROM in the three planes (anterior flexion - LFF, extension - LFE, lateral flexion – LL, and rotation - LR) was measured. Radiographs were obtained to calculate the lumbar part of the mSASSS (LmSASSS). Other outcome measures were: BASMI, Schober and lateral flexion (measured with a tape), BASDAI and ASDAS-CRP. Subgroups of patients with high and low levels of lumbar structural damage were defined, by dichotomising the population based on the median LmSASSS (5.5). Pearson correlations between measures, Student T tests for significant differences between groups and linear regression analysis, to analyse relationship between structural damage and mobility, were obtained.
Results 44 axSpA patients were recruited from the COSPAR cohort of the Reina Sofia University Hospital (39% females, age 43±12 years, disease duration 17±14 years, BMI 26±3). All mobility measurements (except for LFE) in the severe damage group were significantly reduced compared to the group with less damage (p<0.001). BASDAI and ASDAS were not significantly different between these groups. LmSASSS correlated (p<0.001) with all mobility measures obtained by IMU system and conventional metrology, especially with LFF, but not with activity indexes. Only a weak relation between BASDAI and LFE was found (p<0.05). A linear regression for analysing relationship between LmSASSS and LFF including as variables: age, sex, evolution time, BMI, PCR, ASDAS, LFF, LFE, LL, LR and LmSASSS was done. LmSASSS linear regression using only LFF obtained an adjusted R2 of 0.72 (Figure-A). Introducing the rest of variables, the most significant for LmSASSS was LFF (p<0.001) and the other mobility measures (LL, LR with p<0.05) and age, PCR and ASDAS (p<0.01) with an adjusted R2 of 0.81(Figure-B). For LFF, LmSASSS was the most significant (p<0.001), with Sex (p<0.01) and PCR (p<0.05) included, obtaining an adjusted R2 of 0.76 (Figure-C). Differences between models and fitted values (RMSE) were 3.8 units for LmSASSS and 7.7° for LFF.
Conclusion Other studies have shown a relationship between mobility, structural damage and disease activity using a more global assessment. Our study focuses on the analysis of structural damage and mobility in the lumbar vertebrae, and our results suggest that structural damage is the dominant factor determining spinal mobility in this region (especially lumbar frontal flexion). Further studies are needed combining these results with MRI.
Acknowledgement This study was funded by FOREUM (www.foreum.org), Junta de Andalucia (PIN-0079-2016) and XXI Research Promotion Program of the University of Córdoba (Spain).
Disclosure of Interests Juan L. Garrido-Castro: None declared, Inmaculada Concepcion Aranda-Valera: None declared, Cristina Gonzalez-Navas: None declared, Pedro Machado Consultant for: Abbvie, BMS, Celgene, Janssen, MSD, Novartis, Pfizer, Roche and UCB, Speakers bureau: Abbvie, BMS, Celgene, Janssen, MSD, Novartis, Pfizer, Roche and UCB, Philip Gardiner: None declared, Joan Condell: None declared, Eduardo Collantes Estevez: None declared