Background Erosive OA involves interphalangeal joints (IPJs), resulting in a high disease burden, for which treatment options are limited. Although it is characterized by joint inflammation, earlier treatment with TNF blockers was equivocal.
Objectives To investigate the 1-year efficacy of ETN in erosive OA.
Methods In a European multicentre study (NTR 1192) patients were equally randomized to subcutaneous ETN (24 weeks 50 mg weekly, thereafter 25 mg weekly) or placebo. Patients with erosive (≥1 IPJ with radiographic pre(erosive) anatomical phase (“J”/“E”) according to Verbruggen-Veys system) inflammatory (≥1 IPJ with soft swelling/erythema and with positive power Doppler at US) symptomatic (VAS pain >30/100 on NSAID use, flare after NSAID washout) OA were included. VAS pain, hand function (FIHOA), quality of life (SF-36), no. of tender joints and grip strength were assessed after 4, 8, 12, 24, 36 weeks and 1 year. Radiographic progression of IPJs was scored blindly in paired order (baseline, 24 weeks and 1 year) following the quantitative Ghent University Scoring System (GUSS, 0–300 per IPJ).
With linear mixed models VAS pain was compared between treatment groups at 24 weeks (primary outcome), and 1 year in intention-to-treat (ITT) analyses. With general estimated equations secondary outcomes were analysed. Adjustments were made for centre, baseline values and patient effects were appropriate. Completers fulfilling the extensive inclusion criteria were included in per-protocol (PP) analyses.
Results Of 284 screened patients, 90 (mean age 60 years, 81% women, 96% fulfilled ACR hand OA criteria) were randomized; 22 discontinued the study prematurely. At baseline patient characteristics did not differ between the groups. VAS pain in all patients decreased -24.8 mm (95%CI -29.2;-20.5 (P<0.001)) at 24 weeks. In ITT differences in pain between the groups were in favour of ETN, but did not reach statistical significance. However, in PP the difference reached statistical significance (table). No differences were seen on secondary clinical outcomes. The PP analysis of GUSS showed a mean difference in favour of ETN (table), indicating more remodelling in the ETN group. Additional analyses showed an interaction between soft swelling/erythema and ETN treatment on GUSS, resulting in a statistical significant (P<0.05) mean difference between ETN and placebo groups. More patients dropped out on placebo than on ETN (6 vs. 3) due to inefficacy, whereas more on ETN than on placebo (6 vs. 1) due to adverse effects.
Conclusions In erosive OA ETN was not superior over placebo on VAS pain at 24 weeks. However in the symptomatic and inflammatory patients completing the study ETN was superior over placebo both on pain and structural damage assessed by GUSS; ETN was especially effective in joints with signs of inflammation.
Acknowledgement Pfizer for supply of study medication and research grant.
Disclosure of Interest None declared