Article Text

THU0084 Can Autoantibody Level Changes Predict Arthritis in Arthralgia Patients?
  1. M.V. Beers1,
  2. M. Stuiver2,
  3. M.D. Koning3,
  4. L.V.D. Stadt4,
  5. D.V. Schaardenburg5
  1. 1Amsterdam Rheumatology and immunology Center, Reade
  2. 2Dept Medical Statistics, Academic Medical Centre
  3. 3Laboratory, Onze Lieve Vrouwe Gasthuis
  4. 4Amsterdam Rheumatology and immunology Center, VU University Medical Center
  5. 5Amsterdam Rheumatology and immunology Center, Academic Medical Centre, Amsterdam, Netherlands


Background A preclinical phase of rheumatoid arthritis (RA) has been recognized in which abnormalities in autoantibodies exist, frequently accompanied by arthralgia. IgM rheumatoid factor (IgM-RF) and anti-citrullinated protein antibodies (ACPA) can be present many years before the diagnosis of RA, and blood donor cohorts show a rise in these markers 1-3 years preceding diagnosis.

Objectives The present study was undertaken to develop individual prediction models using the autoantibody levels over time.

Methods Data came from 263 patients in an ongoing prospective cohort of arthralgia patients with ACPA and/or IgM-RF positivity and with repeated measurements of ACPA and IgM-RF, who were followed for arthritis development. These patients were used to fit a joint model of IgM-RF/ACPA levels over time combined with time-to-event data for arthritis.

Results Ninety-six patients (26%) developed arthritis after a median of 22 months (IQR 15-37). A total of 986 antibody measurements were included with a median number of 4 (IQR 3-5). The course of antibody levels over time was highly variable. On a group level, there was no statistically significant change of RF levels over time between the groups of patients who developed arthritis compared to those who did not. ACPA levels over time on the group level showed a small but statistically significant difference between the groups, with a rise of 10log transformed ACPA of 0.004 per month in the arthritis group and 0 per month in the patients who did not develop arthritis. The combination of longitudinally measured ACPA with time-to-event data for arthritis in the joint models showed a corrected hazard ratio of 2.6 (95%CI: 1.9-3.6, p<0.001) for the development of arthritis, with an Area Under the Curve (AUC) of 0.769. However, when using only baseline levels of ACPA the AUC was 0.748. RF levels over time did not show any significant associations between the groups and in the joint models.

Conclusions ACPA levels over time had limited additional predictive value over baseline measurement of ACPA levels only, while longitudinally measured RF had no predictive value for development of arthritis in seropositive arthralgia patients. The use of repeated antibody measurements for the individual prediction of arthritis development in the symptomatic at risk phase does not seem useful.

Disclosure of Interest None declared

Statistics from

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.