Bidirectional Contrast agent leakage correction of dynamic susceptibility contrast (DSC)-MRI improves cerebral blood volume estimation and survival prediction in recurrent glioblastoma treated with bevacizumab
Bidirectional Contrast agent leakage correction of dynamic susceptibility contrast (DSC)-MRI improves cerebral blood volume estimation and survival prediction in recurrent glioblastoma treated with bevacizumab
复制标题
DOI：
10.1002/jmri.25227
复制
发表时间：
2016-11-01
影响因子：
4.4
通讯作者：
Ellingson, Benjamin M.
中科院分区：
文献类型：
--
作者：
Leu, Kevin;Boxerman, Jerrold L.;Ellingson, Benjamin M.
目的评价动态磁敏感对比磁共振成像（DSC-MRI）中由造影剂外渗引起的T-1和T2* 伪影的泄漏校正算法，该算法考虑了双向造影剂流量，并比较该模型与单向和未校正模型在复发性胶质母细胞瘤患者中的相对脑血容量（CBV）估计值和总生存期（OS）分层。材料和方法我们测定了患者在贝伐单抗治疗前后对比增强肿瘤内的中值rCBV（在1.5T上扫描75次，在3.0T上进行19次扫描），复发性GBM无泄漏校正，并应用单向和双向泄漏校正算法来确定rCBV是否分层OS。使用双向渗漏校正算法计算的贝伐珠单抗相对于基线的rCBV与较长OS显著相关（考克斯，P=0.01），而使用单向模型（P=0.43）或未校正的rCBV值（P=0.28）的rCBV变化没有。两种渗漏校正算法计算的rCBV估计值平均相差14.9%.ConclusionAccounting T1和T2* 渗漏污染DSC-MRI使用二室，双向而不是单向交换模型可能会改善贝伐单抗治疗后复发GBM患者的生存分层。J.磁共振Imaging 2016;44：1229-1237.
PurposeTo evaluate a leakage correction algorithm for T-1 and T2* artifacts arising from contrast agent extravasation in dynamic susceptibility contrast magnetic resonance imaging (DSC-MRI) that accounts for bidirectional contrast agent flux and compare relative cerebral blood volume (CBV) estimates and overall survival (OS) stratification from this model to those made with the unidirectional and uncorrected models in patients with recurrent glioblastoma (GBM).Materials and MethodsWe determined median rCBV within contrast-enhancing tumor before and after bevacizumab treatment in patients (75 scans on 1.5T, 19 scans on 3.0T) with recurrent GBM without leakage correction and with application of the unidirectional and bidirectional leakage correction algorithms to determine whether rCBV stratifies OS.ResultsDecreased post-bevacizumab rCBV from baseline using the bidirectional leakage correction algorithm significantly correlated with longer OS (Cox, P=0.01), whereas rCBV change using the unidirectional model (P=0.43) or the uncorrected rCBV values (P=0.28) did not. Estimates of rCBV computed with the two leakage correction algorithms differed on average by 14.9%.ConclusionAccounting for T-1 and T2* leakage contamination in DSC-MRI using a two-compartment, bidirectional rather than unidirectional exchange model might improve post-bevacizumab survival stratification in patients with recurrent GBM. J. Magn. Reson. Imaging 2016;44:1229-1237.