Dissertation - Doctoral Facilitating Clinical Trials of Parenteral Lipid Strategies for the Prevention of Intestinal Failure Associated Liver Disease (IFALD) in Infants

Author

Ivan Russell Diamond

Abstract

Objective: The objective of this thesis was to facilitate clinical trials of the optimal lipid based approach (e.g.: omega-3 containing lipid emulsions or minimization of conventional lipid) for the prevention of Intestinal Failure Associated Liver Disease (IFALD). This was achieved through 3 related projects.

Project 1: The first project examined the risk of advanced IFALD associated with exposure to conventional intravenous lipid in a logistic regression model. The study demonstrated that each day of conventional lipid (> 2.5 g/kg/day) was associated with a significant increase in the risk of advanced IFALD [Odds Ratio: 1.04 95% CI: 1.003 – 1.06].

Project 2: The second project surveyed experts in Intestinal Failure regarding their beliefs of the efficacy of lipid minimization and lipid emulsions containing omega-3 fatty acids relative to conventional emulsions. The goal of the project was to develop prior distributions of the treatment response for these therapies that can be used in Bayesian analyses of clinical trials. Our results demonstrated consistent expert opinion that the novel lipid based approaches are superior to conventional therapy. Estimates of the treatment effect were similar for the two approaches (median elicited treatment response, relative to conventional lipid, was a relative risk of 0.53 for omega-3 lipid and 0.45 for lipid minimization).

Project 3: The final project was a pilot randomized controlled trial of an omega-3 emulsion. The study demonstrated that the randomized design is a feasible strategy for evaluating lipid based approaches for the prevention of IFALD. A Bayesian preliminary assessment of the results of the trial, suggests a high likelihood that the trial will demonstrate a difference between the conventional and omega-3 emulsion evaluated in the trial. However, since the analysis was blinded, the direction of the difference is not known.

Conclusion: This thesis will contribute to the design and analysis of high quality and feasible randomized trials that will allow investigators to address the optimal lipid based approach to the management of IFALD.


Supervisor

Brian Feldman


Co-Supervisor

Paul Wales


Committee

Aideen Moore