Objectives: To compare the 48-week drug survival, efficacy and toxicity of monotherapy with a fully human anti-tumour necrosis factor-alpha (TNF-alpha) monoclonal antibody (moAb) and methotrexate (MTX) in patients with active long-standing rheumatoid arthritis (RA). Secondary aims were to identify potential predictors for clinical response.

Methods: Patients with RA, enrolled in phase I trials with a human anti-TNF-alpha moAb and followed for at least 48 weeks at our centre, were compared with patients receiving MTX monotherapy without folate supplementation. The first 6 weeks of anti-TNF therapy were placebo-controlled and followed by an open-label study. Patients treated with MTX participated in a 48-week, double-blind, phase III study of MTX alone vs MTX with folate supplementation, which was co-ordinated by our department. The studies with anti-TNF-alpha and MTX were performed in the same period and had very similar inclusion, exclusion, response and stop criteria.

Results: Sixty-one patients treated with anti-TNF-alpha moAb were compared with 137 receiving MTX monotherapy. At baseline, patients in the anti-TNF-alpha group had a longer disease duration (median 108 vs 50 months, P=0.0001) and a more protracted history of second-line anti-rheumatic drugs than those treated with MTX (median 4 vs 1, P=0.0001). The 48-week dropout rate was lower among patients treated with anti-TNF (23 vs 45% in the MTX group, P<0.005). Proportional hazard analysis showed a significantly lower dropout risk among anti-TNF-treated patients [relative risk (95% confidence interval): 0.28 (0.12-0.6) uncorrected and 0.17 (0.06-0.45) corrected for confounders). The 48-week area under the curve for the disease activity score (DAS) was smaller in the anti-TNF-alpha group than in the MTX group (P=0.005). The percentage of responders was higher in the anti-TNF-alpha group over the whole study period. The median percentage of visits in which a patient fulfilled the European League Against Rheumatism (EULAR) response criteria was 83% in the anti-TNF-alpha group vs 40% in the MTX group (P=0.0001). Clinical and demographic characteristics were, in general, poor predictors for response to therapy at week 48. The clinical response after the first anti-TNF-alpha dose tended to increase the chance of prolonged efficacy of this approach [relative risk (95% confidence interval): 2 (0.75-6.0)]. The previous number of second-line drugs was the only predictive variable for response to MTX to which it was inversely related [relative risk (95% confidence interval): -0.71 (-0.57 to -0.88)].

Conclusions: In patients with active, long-standing RA, blocking TNF-alpha is more effective and better tolerated than MTX monotherapy. An early response increases the chance of a sustained effect of anti-TNF-alpha. In contrast to MTX, the response to anti-TNF-alpha is not affected by previous disease-modifying anti-rheumatic drug history.