Background/Purpose:

Step-down therapeutic strategies are often considered in established rheumatoid arthritis (RA) once remission is achieved but are associated with a potential risk of relapse or structural damage progression.

A 18-month equivalence randomized controlled PROBE (prospective open blinded endpoint) trial was conducted in established RA patients treated with etanercept (ETA) or adalimumab (ADA) in stable DAS28 remission, comparing the impact of a DAS28-driven step-down strategy based on progressive injection spacing (S arm) to a strategy maintaining the therapy at full regimen (M arm).

Methods:

Inclusion criteria were: RA according to the ACR1987 criteria, treatment with ETA or ADA for ³ 1 year either as monotherapy or in combination, prednisone at a daily dose £ 5 mg/d, DAS28 remission (DAS28 £ 2.6) for ³ 6 months, no structural damage progression on X-rays since last X-ray assessment. Patients were randomized (blocks of variable sizes, after stratification on centre and TNF-blocker) in the 2 arms M or S and followed every 3 months for 18 months. In the S arm, the interval between 2 subcutaneous injections was increased by 50% every 3 months up to a complete stop in 4^th step. If DAS28 remission was not maintained, dose tapering was suspended or reversed to the previous interval based on DAS28 level. The primary endpoint was disease activity, based on repeated DAS28 measures (every 3 months over 18 months) under the hypothesis of non-inferiority of the S arm, based on a mixed linear model. Secondary endpoints were DAS28 and HAQ evolution over 18 months and relapse (DAS28 increase > 0.6 and DAS28 > 2.6).

Results :

137 patients were included, 64 and 73 in the S and M arm respectively. Their main baseline characteristics were the following (mean ±sd or %): age 55 ±11 years, female 78%, RA duration 9.5 ±8.0 years, RF+ 68%, ACPA+ 78%, erosive disease 88%, DAS28 1.8 ±0.6, DAS44 1.0 ±0.5, HAQ 0.4 ±0.5, number of previous DMARDs 2.7 ±1.7, ETA 54 %, ADA 46 %. At 18 months, 15 % of the S-arm patients had completely stopped and 67% tapered TNF-blockers, other treatments remaining stable. The remaining 18% patients were unable to taper their treatment and remain at the initial injection interval.

Disease activity on DAS28 and functional status (HAQ) were not significantly different between the 2 arms (figure). However, we failed to demonstrate the equivalence between the 2 strategies (p = 0.6) and due to the step-down strategy, a relapse occurred more frequently in the S than in the M arm (81% vs. 56%, p=0.0009).

Figure

Conclusion :

Spacing of TNF-blockers was feasible in 82% of patients. Although we failed to demonstrate the equivalence between the 2 strategies, the spacing strategy did not result in a significant increase in disease activity or functional impairment. The impact of the spacing strategy on X-ray structural damage is currently being analysed.

ClinicalTrials.gov n°: NCT00780793