A competitive BB-ELISA for anti-toxin B neutralising antibodies in assessing risk of recurrent Clostridioides difficile infection: a retrospective cohort study from the MODIFY I and II trials

Background Recurrent Clostridioides difficile infection (rCDI) is a major clinical challenge. Host neutralising antibodies (NAbs) against toxin B (TcdB) are key correlates of protection, but the absence of a simple, standardised assay to measure this functional immunity prevents risk-adapted patient management. We aimed to develop and clinically validate a rapid, bispecific antibody-based competitive ELISA for measuring anti-TcdB NAbs.Methods In this retrospective cohort study of the MODIFY I and II phase 3, randomised, placebo-controlled trials of bezlotoxumab (conducted at 322 sites in 30 countries between Oct 10, 2011, and May 22, 2015), we assessed post-treatment serum samples from placebo-arm participants with CDI after initial clinical cure. We engineered a bispecific antibody (BB-IgG) targeting two conserved neutralising TcdB epitopes and developed a rapid competitive ELISA (BB-ELISA). The assay was standardised into a commercial-grade kit and validated through correlation with a cell-based neutralisation assay. We measured patients' anti-TcdB NAb levels after antibiotic therapy by using BB-ELISA and evaluated their association with the primary outcome (rCDI) within 12 weeks. The prognostic cutoff was determined using classification and regression tree analysis.Findings The BB-ELISA results correlated strongly with cell-based neutralising titres (R2=0·7245, pInterpretation The BB-ELISA represents a simple, standardised, and scalable tool for quantifying functional anti-TcdB immunity at the point of cure, filling a long-standing diagnostic gap. This assay identifies patients with CDI who are at high risk of recurrence, thus supporting rapid risk stratification. The assay could help guide personalised management decisions and future targeted therapeutic strategies for rCDI.Funding National Institutes of Health and Merck.