Predicted vs. observed radiographic progression in early rheumatoid arthritis (POPeRA): results from a randomized trial

Objectives: The aim of this study was to apply a previously published method for evaluating radiographic progression, namely, predicted vs. observed radiographic progression in early rheumatoid arthritis (POPeRA), to the Swedish pharmacotherapy (SWEFOT) trial. Method: In SWEFOT, 487 patients with eRA were given methotrexate (MTX), and non-responders were randomized to group A [triple therapy: MTX + sulfasalazine (SSZ) + hydroxychloroquine (HCQ)] and group B [anti-tumour necrosis factor (anti-TNF) therapy: MTX + infliximab]. Responders continued on MTX. Predicted progression for 343 eligible patients was calculated based on the baseline total Sharp/van der Heijde score (SHS) divided by symptom duration, compared to observed progression at 12 and 24 months. Results: Observed radiographic progression was reduced from predicted by a mean of 50.1% (A), 72.3% (B), and 73.9% (MTX) at 12 months and by 87.2, 89.8, and 87.8% at 24 months, respectively. Among completers, reductions of 56.7% (A) and 76.5% (B) at 12 months and of 91.0% and 96.0% at 24 months, respectively, were observed. At 12 months, there were no significant between-group differences. At 24 months, progression was reduced more in group B than in group A (first quartile difference 8.5% favouring group B) and in MTX [n = 316, 89.8% (sd +/- 32.0) vs. 87.2% (+/- 32.2), p = 0.021; vs. 87.8% (+/- 27.8), p = 0.013, respectively]. Conclusions: The POPeRA method confirms the original SWEFOT finding in that anti-TNF therapy was statistically marginally superior (2.6%) to triple therapy in preventing radiographic progression at 24 months among initial MTX nonresponders. The simulation provided through POPeRA may facilitate comparisons of the relative efficacy of various treatments in preventing radiographic progression.