期刊
MULTIPLE SCLEROSIS
卷 14, 期 6, 页码 764-769出版社
SAGE PUBLICATIONS LTD
DOI: 10.1177/1352458507087842
关键词
MRI; persistent black hole; sample size; negative binomial distribution
Background MRI is often used as primary outcome measure in phase II clinical trials in multiple sclerosis (MS). Since persistent T1 hypointense lesions are a surrogate parameter for axonal damage and demyelination, they may serve as a marker for monitoring the efficacy of neuroprotective drugs. At present, a power analysis using black hole (BH) evolution as primary outcome measure has not been performed. Objective To assess the feasibility of using BH evolution on serial brain MR images as primary outcome measure in proof of concept studies in MS. Methods MRI-data obtained from 169 active RRMS patients were analysed for BH evolution by determining the cumulative number of contrast enhancing lesions (CEL) evolving into a persistent black hole (PBH) after 3 months. With a parametric simulation procedure, based on a statistical distribution fitting the data, sample sizes were calculated. Results 21.2% of the total number of CELs observed during the study period evolved into a PBH. Ring enhancing lesions evolved most frequently into a PBH (59.4%), followed by lesions larger than 10 mm (57.4%) and periventricular CELs (30.6%). The simulation procedure, based on the statistical negative binomial (NB) model resulted in a sample sizes between 200 subjects and 30 subjects per arm, for treatment effects ranging from 50% to 90% reduction of the number of CELs evolving into a PBH, respectively. Conclusion To perform a MRI monitored phase II clinical trial with a feasible sample size, using the evolution of CELs into PBHs as primary outcome parameter, a potent drug is required to obtain sufficient power.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据