Impact of Different Visual Field Testing Paradigms on Sample Size Requirements for Glaucoma Clinical Trials

Zhichao Wu,Felipe A Medeiros

doi:10.1038/s41598-018-23220-w

Abstract

Visual field testing is an important endpoint in glaucoma clinical trials, and the testing paradigm used can have a significant impact on the sample size requirements. To investigate this, this study included 353 eyes of 247 glaucoma patients seen over a 3-year period to extract real-world visual field rates of change and variability estimates to provide sample size estimates from computer simulations. The clinical trial scenario assumed that a new treatment was added to one of two groups that were both under routine clinical care, with various treatment effects examined. Three different visual field testing paradigms were evaluated: a) evenly spaced testing, b) United Kingdom Glaucoma Treatment Study (UKGTS) follow-up scheme, which adds clustered tests at the beginning and end of follow-up in addition to evenly spaced testing, and c) clustered testing paradigm, with clusters of tests at the beginning and end of the trial period and two intermediary visits. The sample size requirements were reduced by 17–19% and 39–40% using the UKGTS and clustered testing paradigms, respectively, when compared to the evenly spaced approach. These findings highlight how the clustered testing paradigm can substantially reduce sample size requirements and improve the feasibility of future glaucoma clinical trials.

Highlights

The helpful notion of clustering visual field tests at the beginning and end of the trial period was incorporated into the design of the United Kingdom Glaucoma Treatment Study (UKGTS)[7]
As the UKGTS still included burdensome intermediary visits occurring every couple of months, it remains to be determined if using a testing paradigm that includes fewer of those intermediary visits, but which would employ clustering of tests in each one could be beneficial in terms of further reducing sample size requirements and decreasing the number of visits
This study demonstrated that the sample size requirements for a short-term clinical trial were reduced using the UKGTS testing paradigm when compared to a conventional evenly spaced paradigm, but even further reduced by the clustered testing paradigm described in this study when between-group trend-based analysis was used as the outcome measure

Summary

Introduction

The helpful notion of clustering visual field tests at the beginning and end of the trial period was incorporated into the design of the United Kingdom Glaucoma Treatment Study (UKGTS)[7]. Paradigm randomized clinical trial designed to evaluate the beneficial effect of latanoprost versus placebo in delaying visual field progression. The study included one regular visual field test at each follow-up visit occurring at approximately 2–3 month intervals[7]. Whilst this paradigm is likely to be an improvement from an evenly spaced approach, the extent to which it would reduce sample size requirements is unknown. The purpose of this study was to evaluate sample size requirements for clinical trials using visual field endpoints using different testing paradigms. We demonstrate that a hybrid paradigm modeled after the “wait-and-see” approach, but modified to include few intermediary visits, is a powerful design for detecting endpoint differences when investigating the effect of new therapies compared to existing standard of care

Objectives

Methods

Results

Conclusion