A total of 308 patients were screened, of which 118 were excluded, and 190 patients fulfilled the study eligibility criteria. Ninety-five patients were randomly assigned to each study group, receiving biosimilar or reference denosumab (Figure 1). The baseline characteristics and demographics of the patients are presented in Table 1.
BMD changes were evaluated in 164 patients (84 patients in the Arylia group and 80 in the Prolia® group) at the lumbar spine (L1–L4) and in 165 patients (84 patients in the Arylia group and 81 in the Prolia® group) at the total hip and femoral neck. At month 18, there were 24 patients with missing BMD at the total hip and femoral neck and 25 patients with missing BMD at the lumbar spine. The corresponding reason for missing BMD values is provided in Figure 1. One patient in the Arylia group was excluded from the PP population because of protocol deviation, and the BMD was not missing at month 18. The mean (SD) baseline BMD at the lumbar spine, total hip, and femoral neck were 0.71 (0.06), 0.79 (0.10), and 0.64 (0.09), respectively, in the Arylia group and 0.71 (0.06), 0.78 (0.11), and 0.64 (0.08), respectively, in the Prolia® group. The mean (SD) final (month 18) BMD at the lumbar spine, total hip, and femoral neck were 0.75 (0.07), 0.80 (0.09), and 0.66 (0.08), respectively, in the Arylia group and 0.75 (0.07), 0.81 (0.10), and 0.65 (0.08), respectively, in the Prolia® group. No significant difference was noticed in baseline and final BMD values at either site between the two study groups (p = 0.85, 0.90, and 0.75 at the baseline visit and p = 0.92, 0.83, and 0.64 at the final visit at the lumbar spine, total hip, and femoral neck, respectively).
The mean (SD) percent changes in BMD at the lumbar spine, total hip, and femoral neck were 5.91 (5.58), 2.32 (5.24), and 1.91 (6.32), respectively, in the Arylia group and 5.52 (5.59), 2.28 (5.52), and 1.50 (6.62), respectively, in the Prolia® group in the PP set. The differences between the treatment groups were not statistically significant at either site (p = 0.66, 0.96, and 0.68, respectively; Figure 2). Detailed comparisons of the mean percent changes in BMD at all measured sites in the PP and ITT analysis sets are provided in Additional file 1.
Mean percentage change in BMD of the lumbar spine (L1–L4) (a), total hip (b), and femoral neck (c). Error bars show standard error. The plot is based on the PP set
In the ANCOVA, the least-squared mean (SE) percent changes in BMD at the lumbar spine, total hip, and femoral neck were 5.89 (0.60), 2.28 (0.54), and 1.94 (0.67), respectively, in the Arylia group and 5.53 (0.62), 2.31 (0.55), and 1.46 (0.68), respectively, in the Prolia® group in the PP set. The differences between the treatment groups were not statistically significant at either site (p = 0.67, 0.97, and 0.62, respectively). Detailed comparisons of least-squared mean percent changes in BMD at all measured sites in PP and ITT analysis sets are provided in Additional file 2.
The primary endpoint of the study was met, as the lower limits of 95% two-sided confidence intervals of differences in the mean percent changes in BMD at the lumbar spine, total hip, and femoral neck were all greater than the predefined margin of − 1.78 in the PP analysis set (Figure 3). The results were similar when we adjusted the data for baseline BMD in the ANCOVA (Figure 3). Similarly, the lower limits of 95% CIs of differences in the mean percent changes in BMD at the lumbar spine and femoral neck were greater than the − 1.78 margin in the ITT analysis set. The results for the ITT analysis set are provided in Additional file 3.
Forest plot for comparing Arylia versus Prolia® in terms of mean percent changes in BMD of the lumbar spine (L1–L4), total hip, and femoral neck at 18 months of the study. Forest plot demonstrating both t test analysis and ANCOVA model for PP set
New Vertebral Fractures
Regarding the occurrence of new vertebral fractures, no new vertebral fractures occurred among the 190 patients who participated in the study.
Changes in BTMs were not affected by the treatment group and followed a similar trend in both treatment groups. Bone formation markers, including BSAP, OC, and P1NP, and bone resorption markers, including CTX and NTX, all decreased over 18 months of the study (Figure 4).
In this study, a total of 135 AEs were reported. The most common AEs in both arms by system organ class (SOC) were metabolism and nutrition disorders and musculoskeletal and connective tissue disorders. Hypocalcemia and hypertension had the highest incidence among all AEs, and there were no significant differences between the Prolia® and Arylia arms.
The severity of AEs was evaluated using CTCAE v5.0. According to this classification, 119 AEs were categorized as grade 1 or 2, and 16 were categorized as grade 3. No grade 4 AEs were reported in this study. Among all AESIs, nine patients reported AEs in infections and infestations SOC, and three reported AEs in neoplasms benign, malignant, and unspecified (including cysts and polyps) SOC. Further details about these AEs are mentioned in Additional file 4. No report was found regarding eczema, ONJ, AFF, bone fracture, cardiovascular disorder, and pancreatitis acute.
Throughout this study, 13 serious adverse events (SAEs) were recorded in 12 patients, and all events resulted in patient hospitalization. All SAEs were considered unrelated to treatment. Additional file 4 demonstrates the summary of key safety results of the study.
Among all the samples, two samples from months 0 and 6 were positive for one patient, and the samples from months 12 and 18 of this patient were negative. Other patient samples were negative for anti-denosumab antibodies at all time points.
Arthritis Res Ther. 2022;24(161) © 2022 BioMed Central, Ltd.
Copyright to this article is held by the author(s), licensee BioMed Central Ltd. This is an Open Access article: verbatim copying and redistribution of this article are permitted in all media for any purpose, provided this notice is preserved along with the article's original citation.