Background/Purpose: In April 2019, FDA approved the first treatment, intravenous (IV) belimumab (BEL), specifically for children 5 to 17 years of age with active, seropositive SLE receiving standard care (SOC). The approval was supported by a single post-marketing required randomized, controlled trial (NCT01649765) that evaluated the efficacy, safety and pharmacokinetics (PK) of 10 mg/kg IV BEL vs placebo in 93 pediatric patients. Due to the rarity of childhood onset SLE limiting enrollment, by design, the study was not adequately powered to make formal statistical inferences on its own. Determination of efficacy was therefore based on PK and efficacy results from the study, as well as extrapolation of the established efficacy of IV BEL from the two phase 3 adult studies. To provide more reliable efficacy estimates in this context, FDA also employed a novel post-hoc Bayesian analysis which borrowed information from the phase 3 adult IV studies, under a reasonable assumption about the similarity between the adults and pediatric subjects.

Methods: A Bayesian logistic regression model was used to analyze the treatment effect in SLE Responder Index (SRI) response in pediatric patients, which adjusted covariates for treatment group, baseline SELENA SLEDAI score (< 13 vs. ≥13), and age group (5-11 vs. 12-17 years of age). A prior for the treatment effect in the pediatric population was constructed using a weighted combination of a skeptical prior and the treatment effect estimate distribution in adults, where the weight represented the degree of belief in the similarity of the pediatric and treatment effects estimated using two phase 3 adult studies.

Results: The key baseline disease characteristics and PK were similar between the pediatric and adult IV study populations. A plot of the point estimates and 95% credible intervals for varying weights is shown in Figure 1. A prior weight of 55% or larger provided 95% credible intervals that excluded an odds ratio of one, corresponding to a 97.5% posterior probability of efficacy (positive estimate of treatment effect) which is analogous to a rejection of the null hypothesis with a one-sided type I error of 0.025. These results were further supported by the secondary efficacy endpoints and safety, described in the FDA-approved belimumab labeling.

Conclusion: We applied a Bayesian approach to support approval of IV BEL, the first treatment for children with SLE in the United States. Given the similarity between the adult and pediatric populations and studies, a weight of at least 55% weight on the relevance of the adult information to the pediatric population is reasonable, ensuring that there is at least 97.5% posterior probability that belimumab 10 mg/kg has a positive treatment effect in pediatric subjects. The results of this unique analysis supported a conclusion that the treatment effect of IV BEL in the pediatric population favored BEL 10 mg/kg as compared to placebo under reasonable assumption. This innovative approach, especially if pre-specified, may help expedite clinical development in pediatric rheumatic diseases, and address some of the challenges with conducting trials in the setting of these rare conditions.

« Back to 2019 ACR/ARP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/application-of-bayesian-statistics-to-support-approval-of-intravenous-belimumab-in-children-with-systemic-lupus-erythematosus-in-the-united-states/