Article Text

THU0170 Predicted Versus Observed Radiographic Progression in a RA Randomized Trial
  1. A. Levitsky1,
  2. K. Forslind2,
  3. R. F. van Vollenhoven1,
  4. the SWEFOT trial group
  1. 1Unit for Clinical Therapy Research, Inflammatory Diseases (ClinTRID), Karolinska Institutet, Stockholm
  2. 2Rheumatology, Helsingborgs Lasarett and Lund University, Helsingborg and Lund, Sweden


Background Radiological progression is a key outcome in early RA. We previously developed a radiological progression prediction method based on the patient’s report of symptom duration and the radiological damage at first examination1. We previously presented data on predicted radiological progression for a subset of patients in the SWEFOT trial. Here, we applied the analysis of predicting radiological progression (prediction of progression in early RA, POPERA) to the full data set from this randomized trial.

Objectives To analyze predicted versus observed radiological progression in early RA patients treated with MTX or combination therapies.

Methods In the SWEFOT trial2, 487 patients with early RA were given MTX, and non-responders after 3-4 months (DAS28>3.2) were randomized to MTX+SSZ+HCQ (“Triple therapy”) vs. MTX+infliximab (“anti-TNF”). The others continued on MTX (“MTX-responders”). Hand and foot X-rays (baseline, 1, and 2 years) were analyzed with the Sharp-van der Heijde scoring (SHS) method. Predicted progression at 1 and 2 years was calculated as the baseline SHS divided by symptom duration in months multiplied by 12 or 24 respectively. The analyses were based on intention-to-treat (ITT) without imputations of scores. In order to allow inclusion of patients with SHS 0 at baseline in the mathematical model all SHS values in the entire data set were increased by 1. Additional analyses focused on randomized patients who discontinued the trial (i.e. treatment “failure”) and on patients who maintained treatment for 2 years (“completers”). Comparisons between observed progressions were done by non-parametric, Mann-Whitney U testing.

Results The prediction model was applied to 362 patients out of 405 who were eligible. Reasons for exclusion were primarily missing baseline data. In patients who had failed both MTX monotherapy and subsequent triple therapy, the radiological reduction was numerically the least at 12 months (40.7%). In all three groups of patients, observed radiological progression was reduced from predicted by 50-94% (Table I). After 12 months, there were no significant between-group differences. After 24 months, progression was reduced more in the anti-TNF arm than in both the MTX and triple therapy arms.

Conclusions The progression observed in patients who failed their first 2 antirheumatic therapies was 59.3% of predicted, indirectly supporting the validity of this method. The overall results demonstrated significant reductions from predicted radiological progression in patient groups who were successfully treated with MTX and in those who were included in either group of the randomization. Over 24 months, anti-TNF therapy was superior to triple therapy.


  1. Wick et al. Ann Rheum Dis 64:134-137, 2005

  2. Van Vollenhoven et al. Lancet 374:459-466, 2009.

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.