Receptor activator of nuclear factor kappa-Β ligand (RANKL) serum levels are associated with progression to seropositive/negative rheumatoid arthritis

Burska, A.N., El-Jawhari, J.J. ORCID: 0000-0002-0580-4492, Wu, J., Wakefield, R.J., Marzo-Ortega, H., Conaghan, P.G., Emery, P., Ponchel, F. and Freeston, J.E., 2021. Receptor activator of nuclear factor kappa-Β ligand (RANKL) serum levels are associated with progression to seropositive/negative rheumatoid arthritis. Clinical and Experimental Rheumatology, 39 (3), pp. 456-462. ISSN 0392-856X

[img]
Preview
Text
1339506_a901_El_Jawhari.pdf - Post-print

Download (373kB) | Preview

Abstract

Objective: The aim of this study was to establish whether serum RANKL levels in early infammatory arthritis (IA) were associated with rheumatoid arthritis (RA) diagnosis at follow-up, and to evaluate the added value of RANKL for RA diagnosis.

Methods: Serum from 298 patients was collected. Demographic and clinical (swollen/tender joint counts, CRP, DAS28-CRP, RF, ACPA and shared-epitope data were recorded. Baseline ultrasound of 26 joints was performed, including total power Doppler (PD). An ELISA was used to measure RANKL. Predictors of progression were identifed using multivariable logistic regression analysis. Area under the receiver operating characteristics (AUROC) was used to assess the performance of the prediction models and quantify the added value of RANKL in RA diagnosis.

Results: 151 patients developed RA and 147 were non-RA (undifferentiated IA, other infammatory diagnoses or non-persistent infammation). RANKL levels were signifcantly higher in RA (median [IQR]: 474.1 [270.8–1430.6]) than in non-RA (median [IQR]: 301.0 [174.1–477.5]. Three clinical factors (age, SJC and PD) were identifed by multivariable logistic regression with model performance AUROC of 77.9% (95% CI 72.1–83.8%). Adding RANKL resulted in a relative increase of 6.5% in the model classifcation performance of an AUROC of 83.0% (95% CI 77.9–88.1%). In ACPA-negative patients, the model performance increased from 77.6% (95% CI 69.5–85.7%) with clinical data only to 81.9% (95% CI 73.7–89.8%) with added value of RANKL and imaging.

Conclusion: RANKL levels can predict RA diagnosis over clinical biomarkers alone, both seropositive and particularly in seronegative IA patients.

Item Type: Journal article
Publication Title: Clinical and Experimental Rheumatology
Creators: Burska, A.N., El-Jawhari, J.J., Wu, J., Wakefield, R.J., Marzo-Ortega, H., Conaghan, P.G., Emery, P., Ponchel, F. and Freeston, J.E.
Publisher: Pacini Editore SpA
Date: 21 May 2021
Volume: 39
Number: 3
ISSN: 0392-856X
Identifiers:
NumberType
1339506Other
Divisions: Schools > School of Science and Technology
Record created by: Linda Sullivan
Date Added: 30 Jun 2020 08:12
Last Modified: 02 Aug 2021 14:28
URI: http://irep.ntu.ac.uk/id/eprint/40134

Actions (login required)

Edit View Edit View

Views

Views per month over past year

Downloads

Downloads per month over past year