An evaluation of the FDA Responder Endpoint for IBS-C clinical trials: analysis of data from linaclotide Phase 3 clinical trials

Authors


Address for Correspondence

Anthony J. Lembo, 330 Brookline Ave, Boston, MA 02215, USA.

Tel: +617-667-2138; fax: +617-667-1171;

e-mail: alembo@bidmc.harvard.edu

Abstract

Background

Our objective was to evaluate the performance of the Food and Drug Administration (FDA) Responder Endpoint for clinical trials in IBS-C, using data from two large Phase 3 clinical trials of linaclotide. The FDA interim endpoint requires that, for 50% of trial weeks, patients report ≥30% decrease in Abdominal Pain at its worst and (in the same week) an increase in Complete Spontaneous Bowel Movements (CSBMs) of ≥1 from baseline.

Methods

Anchor-based methodology was used to estimate thresholds of clinically meaningful change using symptom-specific patient rating of change questions (PRCQs) and symptom severity questions. The diagnostic accuracy of the FDA Responder Endpoint was assessed using sensitivity/specificity-based methods.

Key Results

Using anchor-based methods, the estimates of the clinically meaningful improvement thresholds for Abdominal Pain ranged from 25.9% to 32.4% and thresholds for increase in weekly CSBM rate ranged from 1.4 to 1.6 CSBMs per week. Compared with the symptom-specific PRCQs for patient rating of relief, the FDA Responder Endpoint has a sensitivity of 60.7%, a specificity of 93.5%, and an accuracy of 82.0%. Changing the number of weeks required to be a responder or the percentage improvement in the Abdominal Pain criteria did not result in notable improvement in the accuracy of the FDA Responder Endpoint.

Conclusions & Inferences

The FDA Responder Endpoint for IBS-C clinical trials represents clinically meaningful improvements in IBS-C symptoms for patients with excellent specificity and reasonable sensitivity.

Ancillary