Clinical utility of intraoperative Arterial Spin Labeling for resection control in brain tumor surgery at 3 T
Marta Calvo-Imirizaldu 1 , Verónica Aramendía-Vidaurreta 1 2 , Carmen Sánchez-Albardíaz 1 , Marta Vidorreta 3 , Reyes García de Eulate 1 , Pablo D Domínguez Echávarri 1 2 , Josef Pfeuffer 4 , Bartolomé Bejarano Herruzo 5 , Lain H Gonzalez-Quarante 5 , Antonio Martinez-Simon 6 , María A Fernández-Seara 1 2
Resection control in brain tumor surgery can be achieved in real time with intraoperative MRI (iMRI). Arterial spin labeling (ASL), a technique that measures cerebral blood flow (CBF) noninvasively without the use of intravenous contrast agents, can be performed intraoperatively providing morpho-physiological information.
This study aimed to evaluate the feasibility, image quality and potential to depict residual tumor of a pseudo-continuous-ASL (PCASL) sequence at 3T. 17 patients with brain tumors, primary (16) or metastatic (1), undergoing resection surgery with iMRI-monitoring were prospectively recruited (9 men, age:56±16.6 years).
A PCASL sequence with long labeling duration (3000ms) and postlabeling delay (2000ms) was added to the conventional protocol that consisted of pre- and post-contrast 3D-T1-weighted (T1w) images, optional 3D-FLAIR, and diffusion. Three observers independently assessed image quality (4-point scale) of PCASL-derived CBF maps. In those with diagnostic quality (scores 2-4) they evaluated the presence of residual tumor using the conventional sequences first, and the CBF maps afterwards (3-point scale). Inter-observer agreement for image quality and the presence of residual tumor was assessed using Fleiss-kappa statistics. Intraoperative CBF-ratio of the surgical margins (i.e., perilesional CBF values normalized to contralateral gray matter CBF) was compared to preoperative CBF-ratio within the tumor (Wilcoxon's test).
Diagnostic ASL image quality was observed in 94.1% of patients (inter-observer Fleiss-κ=0.76). PCASL showed additional foci suspicious of high-grade residual component in three patients, and a hyperperfused area extending outside the enhancing component in one patient. Interobserver agreement was almost perfect in the evaluation of residual tumor with the conventional sequences (Fleiss-κ=0.92) and substantial for PCASL (Fleiss-κ=0.80). No significant differences were found between pre and intraoperative CBF-ratios (p=0.578) in patients with residual tumor (n=7). iMRI-PCASL perfusion is feasible at 3T and is useful for the intraoperative assessment of residual tumor, providing in some cases additional information to the conventional sequences.