Background Chemically crosslinked hylan G-F 20 (hylan) is perceived to have longer efficacy than natural non-chemically modified sodium hyaluronate (SH) in the treatment of osteoarthritis (OA).
Objectives To demonstrate that efficacy of one intra-articular (i.a.) injection of 2% SH is non-inferior to hylan in patients with symptomatic tibiofemoral OA.
Methods This was a randomised, double-blind, controlled, parallel-group, multicentre, non-inferiority trial evaluating patients with tibiofemoral OA (KL grade Ib to III) and a WOMAC A pain ≥40 mm. After an NSAID washout period, eligible patients were randomised to receive one i.a. injection of either 40 mg/2.0 ml SH (Ostenil Plus) or 48 mg/6.0 ml hylan (Synvisc-One). Efficacy parameters were collected by a blinded assessor on Days 0 (D0), 30, 90 and 180. The primary endpoint was the change from D0 in WOMAC A at Day 180. The lower margin of non-inferiority was pre-specified at -8 mm.
Results On the 292 randomised patients (SH =144, hylan =148), 142 were injected with SH, 146 with hylan, and 4 did not receive the injection due to consent withdrawal (2 patients) or presence of effusion (2 patients). A total of 266 patients (91.1%) completed the study. Both groups were homogeneous at D0 for demographic and disease characteristics (SH: 143 patients, age =67.1±9.7; hylan: 147 patients, age =66.6±10.7), except for a lower male ratio in the SH group (SH: 27.3%, hylan: 39.5%; p=0.0279) which had no significant effect on the results. The efficacy analysis performed on the 280 patients of the full analysis set (SH =139, hylan =141) showed that all parameters decreased from D0 in both groups. The mean WOMAC A change from D0 at Day 180 was -33.2±20.1 mm and -36.1±21.5 mm for SH and hylan, respectively (p=0.2613). The mean observed difference between groups was -2.9±20.8 mm with a 95% confidence interval (CI) of [-7.9; 2.2] mm. Therefore, the lower margin of the 95% CI of the difference in mean WOMAC A was higher than the pre-specified bound for non-inferiority. At Day 180, improvement from D0 of the Lequesne score was 4.2±3.7 and 4.7±4.0 in the SH and hylan groups, respectively. The OMERACT-OARSI responder rate at Day 180 was 81.1% in the SH group and 86.3% in the hylan group (p=0.2543). Results were found to be similar in the per protocol set (225 patients, SH =113, hylan =112) with a mean observed difference between groups of -1.9±20.5 mm and a 95% CI of [-7.3; 3.5] mm.
From a safety point of view, 24.0% of the patients reported at least one adverse event. In the hylan group, 4 patients (2.7%) received a corticosteroid injection, 1 had knee prosthesis after about 6 months and 1 had a meniscectomy. In the SH group, 2 patients (1.4%) received a corticosteroid injection. Local reactions to the injections occurred in 8.4% of patients in the SH group versus 13.0% in the hylan group. No serious reaction related to the injection was reported.
Conclusions This clinical trial demonstrated the non-inferiority of one i.a. injection of 2 ml SH compared with 6 ml hylan for the primary efficacy outcome. Both preparations reduced pain and improved functional activity in patients with painful tibiofemoral OA over 6 months.
Disclosure of Interest None declared