Marie Fechtenbaum1, Anna Molto1, Christian Roux2, Philippe Goupille3, Sylvie Chevret4, Karine Briot5. 1. Epidemiology and Biostatistics Unit, Sorbonne Paris Cité Research Center Inserm U1153, Paris, France; Department of Rheumatology, Cochin Hospital, Paris, France. 2. Epidemiology and Biostatistics Unit, Sorbonne Paris Cité Research Center Inserm U1153, Paris, France; Department of Rheumatology, Cochin Hospital, Paris, France; Paris Descartes University, Paris, France. 3. François-Rabelais University, CNRS 7292, Tours, France. 4. Biostatistics and Clinical Epidemiology research, Inserm U1153, Paris Diderot University, Paris, France. 5. Epidemiology and Biostatistics Unit, Sorbonne Paris Cité Research Center Inserm U1153, Paris, France; Department of Rheumatology, Cochin Hospital, Paris, France. Electronic address: karine.briot@aphp.fr.
Abstract
OBJECTIVE: The aim of this study was to assess the effect of baseline inflammation on Magnetic Resonance Imaging (MRI) on the change in Bone Mineral Density (BMD) over 5 years in patients with early spondyloarthritis (SpA). METHODS: From the patients of the DESIR cohort (an early axial SpA cohort), patients with BMD data at both baseline and 5 years, and baseline spine and sacroiliac joints MRI were included. Inflammation was assessed with the SpondyloArthritis Research Consortium of Canada (SPARCC) spine score. Significant BMD loss was defined by a change of >0.03g/cm2. No patients had received TNF blockers before inclusion in the cohort. Univariate and multivariable prognostic analyses were performed. An inverse propensity score weighting method was used to handle confounders. RESULTS: One hundred and eighty-three patients were included (mean age 33.9±8.7 years, 58.5% men). A significant bone loss was reported in 51% (n=92) of patients at either lumbar spine or hip. Fourteen (7%) patients had low BMD (Z-score <-2) at the end of the follow-up vs. 28 (15%) at baseline. In multivariable analysis, age was a protective factor of 5 year-BMD loss at any site (OR=0.96, 95% CI [0.93-0.99]). Baseline MRI inflammation has no significant effect on BMD change at any site (OR=0.84, 95% CI [0.46-1.53]). CONCLUSION: Half of patients with early SpA have a significant bone loss at either lumbar spine or hip over 5 years. Baseline MRI inflammation is not a determinant of this bone loss.
OBJECTIVE: The aim of this study was to assess the effect of baseline inflammation on Magnetic Resonance Imaging (MRI) on the change in Bone Mineral Density (BMD) over 5 years in patients with early spondyloarthritis (SpA). METHODS: From the patients of the DESIR cohort (an early axial SpA cohort), patients with BMD data at both baseline and 5 years, and baseline spine and sacroiliac joints MRI were included. Inflammation was assessed with the SpondyloArthritis Research Consortium of Canada (SPARCC) spine score. Significant BMD loss was defined by a change of >0.03g/cm2. No patients had received TNF blockers before inclusion in the cohort. Univariate and multivariable prognostic analyses were performed. An inverse propensity score weighting method was used to handle confounders. RESULTS: One hundred and eighty-three patients were included (mean age 33.9±8.7 years, 58.5% men). A significant bone loss was reported in 51% (n=92) of patients at either lumbar spine or hip. Fourteen (7%) patients had low BMD (Z-score <-2) at the end of the follow-up vs. 28 (15%) at baseline. In multivariable analysis, age was a protective factor of 5 year-BMD loss at any site (OR=0.96, 95% CI [0.93-0.99]). Baseline MRI inflammation has no significant effect on BMD change at any site (OR=0.84, 95% CI [0.46-1.53]). CONCLUSION: Half of patients with early SpA have a significant bone loss at either lumbar spine or hip over 5 years. Baseline MRI inflammation is not a determinant of this bone loss.