Patient disposition and demographics
A total of 351 patients were screened from 14 sites. Of these, 84 failed screening, and finally 267 patients were randomized, including 133 in the NRD101 group (test group) and 134 in the Artz group (control group). Among all randomized patients, 259 patients (97%) completed the study and 8 patients (3%) discontinued the study (Fig. 1). The reasons for early discontinuation were as follows: five patients withdrew consent, including two from the NRD101 group and three from the Artz group; two patients discontinued due to the occurrence of an AE, all in the Artz group; one patient in the NRD101 group was withdrawn due to being mistakenly injected with Artz (this discontinuation was at the suggestion of the site’s ethics committee). In whole, the persistence rate of subjects in this study remained high, and there was no clear difference in early discontinuation between the two groups.
For analysis set classification, 264 subjects were included in the safety analysis set, including 131 in the NRD101 group and 133 in the Artz group; 250 subjects were included in the full analysis set, including 127 in the NRD101 group and 123 in the Artz group; 230 subjects were included in the per protocol set, including 118 in the NRD101 group and 112 in the Artz group. Three subjects were excluded from the safety analysis set due to a violation of the informed consent process (an unblinded investigator signed the informed consent form as an impartial witness of these illiterate subjects).
Of the 250 patients in the full analysis set, 82.0% were female, the mean age was 60.9 ± 8.19 years old, and the mean body mass index was 25.3 ± 3.39 kg/m2. The mean baseline of WOMAC NRS pain was 6.60 ± 1.135, and the median duration of chronic pain of study knee was 1.88 years. For K-L grade, 48.8% of subjects were judged as Grade 2. There was no difference in baseline characteristics between the two groups (Table 1).
Efficacy outcomes
In the full analysis set, the change in WOMAC NRS pain from baseline on a Last Observation Carried Forward basis at Week 5 was −2.98 ± 0.193 points in the NRD101 group and −2.66 ± 0.194 points in the Artz group; the difference between the two groups was 0.33 ± 0.258 points with a 95% CI of −0.18 to 0.83 points. The lower limit of 95% CI was −0.18 points, which was not below the non-inferiority margin of −0.925 points. Therefore, it is considered that the NRD101 group was not inferior to the Artz group, and the trial attained the primary efficacy end point. The improvement in WOMAC pain in the NRD101 group tended to be greater than in the Artz group from Week 3 to the end of the study (Fig. 2). The change in WOMAC NRS stiffness from baseline at Week 5 was −2.63 ± 2.393 points in the NRD101 group and −2.21 ± 2.422 points in the Artz group, and the change in WOMAC NRS physical function was −2.48 ± 2.155 in the NRD101 group and −2.31 ± 1.986 in the Artz group. For most items of the WOMAC, the NRD101 group consistently tended to show greater improvement from baseline at Week 5 than did the Artz group (Table 2). On the other hand, rates of improvement of local pain were similar between the NRD101 and Artz groups (Table 3). The range of motion in the joint was significantly improved compared with baseline in both treatment groups although no difference between the groups was observed. The mean change in maximum extension angle (degree) from baseline at Week 5 was −1.13 (95% CI −1.70, −0.55) in the NRD101 group and −1.22 (95% CI −1.76, −0.68) in the Artz group; the mean change in maximum flexion angle (degree) from baseline at Week 5 was 7.48 (95% CI 5.66, 9.30) in the NRD101 group and 5.93 (95% CI 4.13, 7.72) in the Artz group. The number of patients showing an overall improvement ([number showing ≤2-point change from baseline]/[total number of patients]) at Week 5 was higher in the NRD101 group (60.0%) than in the Artz group (53.4%), but no statistically significant difference was observed between the two groups (odds ratio: 1.31, [95% CI 0.79, 2.18]).
The results of the pre-specified subgroup analyses of primary outcome (WOMAC NRS pain change at Week 5) were essentially consistent with the results in the overall population, with NRD101 tending to be more effective than Artz in most subgroups (Table 4). Among them, NRD101 showed the tendency of greater pain reduction than Artz in the subgroup of patients whose duration of chronic pain was >1 year and ≤5 years (difference: −1.01 [95% CI −1.83, −0.19]) and patients whose body weight was >50 kg and ≤60 kg (difference: −1.22 [95% CI −2.23, −0.21]).
Overall, there was no difference in the use of acetaminophen as a rescue therapy between the two groups.
Safety
Among all subjects in the safety analysis set, the percentage of subjects with AEs was 32.2% (85/264): 26.0% (34/131) in the NRD101 group and 38.3% (51/133) in the Artz group (Table 5). The most common AEs were viral upper respiratory tract infection 5.3% (14/264), upper respiratory tract infection 3.8% (10/264), and arthralgia 3.0% (8/264). There was no difference between the two groups in the incidence of arthralgia (3.1% [4/131] for NRD101 and 3.0% [4/133] for Artz). As for injection site–related events, there was only one subject with injection site erythema and one subject with injection site pain in the NRD101 group, and one subject with injection site rash in the Artz group. The proportion of subjects who experienced severe AEs was 0.8% in the NRD101 group and 2.3% in the Artz group. Two out of 264 subjects (0.8%) developed serious AEs: sleep apnea syndrome (one event in one subject in the NRD101 group) and hypertension worsened (one event in one subject in the Artz group). Neither serious AE was related to the study drug, and the dose of study drug remained unchanged. Among all subjects in the safety analysis set, 14 (5.3%) experienced 20 adverse reactions: 4.6% (6/131) in the NRD101 group and 6.0% (8/133) in the Artz group. AEs leading to discontinuation occurred only in the Artz group (one subject due to arthralgia and plica syndrome, one subject due to rash). No subject died during the study. Thus, no new safety concerns were identified throughout the study.