Systematic review and meta-analysis to estimate the antibacterial treatment effect of nitrofurantoin for a non-inferiority trial in uncomplicated urinary tract infection. (September 2020)

• Record Type: Journal Article
• Authors: Mitrani-Gold, Fanny S.; Raychaudhuri, Aparna; Rao, Sapna
• Is Part Of: Journal of global antimicrobial resistance. Volume 22(2020)
• Page Start: 68
• Page End: 77
• Publication Date: 2020-09
• Subjects: Non-inferiority trial -- Uncomplicated urinary tract infection -- uUTI -- Acute cystitis -- Nitrofurantoin -- Antibacterial treatment effect; Drug resistance -- Periodicals; Drug resistance -- Periodicals; Drug resistance; Periodicals; 616.9041
• DOI: 10.1016/j.jgar.2020.01.027
• Languages: English
• ISSNs: 2213-7165

Abstract

Highlights: Historical treatment effects of placebo and an active comparator are needed for non-inferiority (NI) trials. Evident gaps exist in estimating therapeutic response from historical nitrofurantoin trials. Systematic literature review/meta-analysis was used to estimate the treatment effect of nitrofurantoin and placebo. A NI margin of 12.5% was supported based on the microbiological response endpoint. Heterogeneity in uncomplicated UTI trials is evident and should be addressed in future trials. Abstract: Objectives: Active-comparator, non-inferiority study designs are used in uncomplicated urinary tract infection (uUTI) to establish the efficacy of a new antibacterial, given the availability of effective antibiotics. Here we estimated the treatment effect of a planned antimicrobial comparator (nitrofurantoin) from historical trial data to properly design an upcoming non-inferiority trial in uUTI. Methods: A systematic literature review and meta-analysis was conducted in compliance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines, which incorporate recommendations for standardised data quality assessment, reporting of results, risk of bias assessment and sensitivity analyses. To account for interstudy variability, a weighted, non-iterative, random-effects model was fit using R software to obtain estimates of the microbiological response rate and corresponding 95% confidence interval (CI) for nitrofurantoin and placebo treatment. Interstudy heterogeneity was assessed with Cochran's χ 2 test for interstudy heterogeneity; I 2 statistic and P -values were computed and included in the forest plot of the meta-analysis. Results: Twelve unique studies met the final eligibility criteria for meta-analysis inclusion; three trials assessed placebo efficacy, eight trials assessed nitrofurantoin efficacy, and one study assessed both nitrofurantoin and placebo efficacy in uUTI. The overall microbiological response (95% CI) was 0.766 (0.665–0.867) for nitrofurantoin and 0.342 (0.288–0.397) for placebo. Conclusion: The corresponding treatment effect estimate for nitrofurantoin was 26.8%, which supports a conservative non-inferiority margin of 12.5% and is consistent with the recently published draft FDA guidance. The findings from this systematic review and meta-analysis may inform future antibacterial trials by providing non-inferiority margin justification.