Session MC.4

Gated Intravascular Ultrasound Reduces the Population Size in Clinical Trials: A Computer Simulation Study

SA De Winter*, R Hamers, JRTC Roelandt, PWJC Serruys, N Bruining

Erasmus University
Rotterdam, Netherlands

During imaging of coronary vessels with intravascular ultrasound (IVUS), cardiac cycle-dependent motion of the IVUS catheter introduces artifacts which are prominently visible in reconstructed longitudinal views (L-views). These artifacts could result in possible measurement inaccuracies. Either ECG- or image-based gating can reduce this, but the effects on quantitative IVUS (QCU) endpoints in clinical trials is difficult to assess due to the lack of a golden standard.
To investigate the impact of these motion induced artifacts and the differences between gated and non-gated (QCU), we developed a computer simulation model. This model includes the motion dynamics of the IVUS catheter (longitudinal shift) and the dimensional changes of the coronary vessel during the cardiac cycle. We simulated a clinical atherosclerosis progression-regression trial (n=400) in which 200 patients received a drug and 200 a placebo. The treatment effect of the drug was estimated at a 3% vessel volume (e.g. plaque) decrease of the coronary segment against 0% in the placebo group. Segment lengths were 38±12 mm and vessel diameters ranged from 3.5 mm to 5.7 mm. All variations were parameterized using real-life data from a previous clinical IVUS study (i.e. the Perspective study, evaluating the effects of an ACE-Inhibitor on progression/regression of atherosclerosis). Estimated longitudinal catheter shift was 2.6±1.3 mm, heart rate 72±10 bpm and vessel volume was determined at baseline and follow up with 3 methods: 1) calculation of the exact volume (simulated golden standard); 2) conventional non-gated QCU and 3) gated QCU.
To rule out possible random effects of one single study, we repeated the simulation 150 times which thus equals 150 studies of 400 patients. The treatment effect of 3% decrease was detected in all patients with method 1 (p<0.05). Gated QCU detected a statistical significant difference when n=26 and for non-gated QCU when n=254.
An estimated treatment effect of 3% plaque reduction is already detected in 26 patients if IVUS is gated in contrast to non-gated QCU in which at least 254 patients must be included. This computer simulation study shows that by applying gating in IVUS driven progression/regression trials the population size could be significantly reduced.

(Abstract Control Number: 214)