Intraoperative computed tomography, navigated ultrasound, 5-amino-levulinic acid fluorescence and neuromonitoring in brain tumor surgery: overtreatment or useful tool combination?

Abstract

Brain tumor surgery is routinely supported by several intraoperative techniques, such as fluorescence, brain mapping and neuronavigation, which are often used independently. Efficacy of navigation is limited by the brain-shift phenomenon, particularly in cases of large or deep-sited lesions. Intraoperative imaging was introduced also to update neuronavigation data, to try and solve the brain-shift phenomenon-related pitfalls and increase overall safety. Nevertheless, each intraoperative imaging modality has some intrinsic limitations and technical shortcomings, making its clinical use challenging. We used a multimodal intraoperative imaging protocol to update neuronavigation, based on the combination of intraoperative Ultrasound (i-US) and intraoperative Computed Tomography (i-CT) integrated with 5-ALA fluorescence and neuromonitoring-guided resection. This is a pilot study on 52 patients (29 men), including four children, with a mean age of 57.67 years, suffering from brain low- (N.=10) or high-grade (N.=34) glioma or metastasis (N.=8), prospectively and consecutively enrolled. They underwent 5-ALA fluorescence-guided microsurgical tumor resection and neuromonitoring was used in cases of lesions located in eloquent areas, according to preoperative clinical and neuroradiological features. Navigated B-mode ultrasound acquisition was carried out after dural opening to identify the lesion. After tumor resection, i-US was used to identify residual tumor. Following further tumor resection or in cases of unclear US images, post-contrast i-CT was performed to detect and localize small tumor remnants and to allow further correction for brain shift. A final i-US check was performed to verify the completeness of resection. Clinical evaluation was based on comparison of pre- and postoperative Karnofsky Performance Score (KPS) and assessment of overall survival (OS) and progression-free survival (PFS). Extent of tumor resection (EOTR) was evaluated by volumetric postoperative Magnetic Resonance performed within 48 h after surgery. Forty-one of the 52 (78.8%) patients were alive and still under follow-up in December 2017. 5-ALA was strongly or vaguely positive in 45 cases (86.5%). Seven lesions (four low-grade glioma, one high-grade glioma, and two metastases) were not fluorescent. i-US visualized residual tumor after resection of all fluorescent or pathological tissue in 22 cases (42.3%). After i-US guided resection, i-CT documented the presence of further residual tumor in 11 cases (21.1%). Mean EOTR was 98.79% in the low-grade gliomas group, 99.84% in the high-grade gliomas group and 100% in the metastases group. KPS changed from 77.88, preoperatively, to 72.5, postoperatively. At the last follow-up, mean KPS was 84.23. The combination of different intraoperative imaging modalities may increase brain tumor safety and extent of resection. In particular, i-US seems to be highly sensitive to detect residual tumors, but it may generate false positives due to artifacts. Conversely, i-CT is more specific to localize remnants, allowing a more reliable updating of navigation data.