<i>Post Hoc</i> Analysis of a Randomized Controlled Trial
OBJECTIVES: Cost-effective guidance of therapeutic strategy in Crohn's disease patients with secondary infliximab (IFX) treatment failure may be achieved by serum IFX and anti-IFX antibody (Ab) measurements by radioimmunoassay (RIA). This study investigated implications of using other techniques for this purpose. METHODS: This is a post hoc analysis of randomized clinical trial including 66 Crohn's disease patients with IFX failure in whom IFX and anti-IFX Ab measurements by RIA had been used for therapeutic guidance. Samples were additionally assessed by enzyme-linked immunosorbent assay (ELISA), homogeneous mobility shift assay (HMSA), and functional cell-based reporter gene assay (RGA). RESULTS: IFX detection was comparable between assays (82% RIA, 76% ELISA, 88% HMSA, and 74% RGA), and it correlated significantly (Pearson's r=0.91-0.97, P<0.0001). However, IFX concentrations varied systematically between all pair of assays except RIA-RGA. Anti-IFX Ab detection was variable (27% RIA, 9% ELISA, 33% HMSA, and 11% RGA), but correlated significantly (Pearson's r=0.77-0.96; P<0.0001). Anti-IFX Abs detected by RIA and HMSA were often from sera without drug-neutralizing activity (RGA). Assays agreed on classification of underlying mechanism for treatment failure in most cases (79-94%). The majority (74-88%) failed IFX owing to pharmacodynamic problems, or had noninflammatory pathophysiology for symptoms resembling relapse. Applied threshold for therapeutic vs. subtherapeutic IFX level influenced classifications. The four different assays did not differ in terms of the ability to predict response to interventions defined by the algorithm. CONCLUSIONS: Despite variable analytical properties, common assays result in similar classifications and interventions in patients with IFX treatment failure, and with comparable clinical outcomes. Implications are, however, profound for the minority classified differently.
American Journal of Gastroenterology, 2014, Vol 109, Issue 7, p. 1055-64