Computed tomography perfusion evaluation after extracranial-intracranial bypass surgery

Vos, Pieter C., Riordan, Alan J., Smit, Ewoud J., de Jong, Hugo W. A. M., van der Zwan, Albert, Velthuis, BK, Viergever, Max A., Dankbaar, Jan Willem


Clinical Neurology and Neurosurgery 136 p. 139-146


Objective: Perfusion imaging is increasingly used for postoperative evaluation of extracranial to intracranial (EC-IC) bypass surgery. Altered hemodynamics and delayed arrival of the contrast agent in the area fed by the bypass can influence perfusion measurement. We compared perfusion asymmetry obtained with different algorithms in EC-IC bypass surgery patients.

Methods: We retrospectively identified all patients evaluated with computed tomography perfusion (CTP) between May 2007 and May 2011 after EC-IC bypass surgery at our institution. CTP images were analyzed with three perfusion algorithms that differ among their ability to anticipate for delayed arrival time of contrast material: the delay-sensitive first-moment mean transit time (fMTT), the semi-delay-sensitive standard singular value decomposition (sSVD) and the delay-insensitive block-circulant SVD (bSVD). The interhemispheric difference in bolus arrival time (Delta BAT) was determined to confirm altered hemodynamics. Interhemispheric asymmetry in perfusion values (mean transit time (MU) difference, cerebral blood flow (CBF) ratio and cerebral blood volume (CBV) ratio) was compared between the three algorithms. Presence of a new infarct in the treated hemisphere was evaluated on follow-up imaging and perfusion asymmetry was compared between patients with and without infarction.

Results: Twenty-two patients were included. The median interhemispheric difference in Delta BAT was 0.98 s. The median MTT difference was significantly smaller when calculated with the delay-insensitive algorithm than with the other algorithms (0.44 s versus 0.90 s and 0.93 s, p

Conclusion: In patients with EC-IC bypass surgery, delay-sensitive algorithms showed larger MTT asymmetry than delay-insensitive algorithms. Furthermore, only the delay-insensitive method seems to differentiate between patients with and without infarction on follow-up. (C) 2015 Elsevier B.V. All rights reserved.