Overview
Brain tumor surgery in so-called "eloquent" brain areas aims to remove as much tumor as possible while preserving neurological functions. Standard surgical planning typically focuses on discrete, anatomically defined cortical regions. However, modern neuroscience demonstrates that most brain functions arise from distributed networks of interconnected areas rather than isolated spots - a concept that standard navigation tools do not fully capture.
The CORTEX study evaluates a surgical workflow - termed "connectome-guided network-based navigation" - in which advanced diffusion MRI processing is used to reconstruct patient-specific maps of white matter pathways and large-scale brain networks. These maps are imported into a clinical neuronavigation system to guide preoperative planning and intraoperative decision-making for patients with gliomas or brain metastases in eloquent regions.
The primary aims are to determine how often network-based information leads to meaningful changes in surgical strategy compared with conventional anatomy-based planning, and to assess early neurological outcomes. Secondary objectives include characterizing the extent of tumor removal, the proximity of the resection to critical white matter tracts, and the feasibility of implementing this pipeline in a high-volume clinical setting.
Description
The CORTEX study is a prospective, single-centre cohort study conducted at the Unit of Neurosurgery of A.R.N.A.S. Civico Di Cristina Benfratelli, Palermo, Italy. Consecutive eligible patients are enrolled from January 2022 onwards.
Background and Rationale:
Contemporary understanding of brain organization emphasizes the distributed, network-based nature of neurological and cognitive functions. Surgical planning centered exclusively on anatomical landmarks - a "localist" approach - may fail to account for the role of long-range white matter pathways, association fasciculi, and large-scale cortico-subcortical networks in sustaining higher-order functions. The concept of "extended eloquence" extends surgical risk stratification beyond classical primary cortices to include associative and integrative networks, whose disruption may produce clinically relevant higher-order deficits even in the absence of damage to traditional eloquent areas.
Diffusion MRI Processing Pipeline:
Preoperative high-direction diffusion MRI is processed using an open-source pipeline integrating MRtrix3 (denoising, Gibbs correction, multi-tissue constrained spherical deconvolution, anatomically constrained tractography with the iFOD2 algorithm, SIFT2 tractogram filtering), FSL (eddy current and motion correction, susceptibility distortion correction), and FreeSurfer (cortical and subcortical segmentation, atlas-based parcellation). In a subset of patients with optimal data quality, an HCP-style surface-based analysis is performed using the Ciftify framework. Workflow automation is achieved through custom Bash and Python scripts, reducing operator-dependent variability.
Neuronavigation Integration:
Tractograms and volumetric overlays of clinically relevant white matter tracts - including the corticospinal tract, arcuate and superior longitudinal fasciculi, inferior fronto-occipital fasciculus, optic radiations, and frontal aslant tract - are co-registered to anatomical space and imported into neuronavigation platform. These overlays are used during preoperative planning to define craniotomy location, surgical corridor, and intended extent of resection relative to critical network architecture.
Intraoperative Integration:
Where applicable, connectome-guided navigation is integrated with intraoperative neurophysiological monitoring and, in selected cases, awake craniotomy with direct electrical stimulation. Concordance between tractographic predictions and intraoperative stimulation findings is recorded prospectively.
Outcome Assessment:
Postoperative MRI is obtained within 48-72 hours of surgery. Neurological assessment is performed at discharge and at 3-month follow-up by the treating neurosurgical team.
Eligibility
Inclusion Criteria:
- Indication for supratentorial brain tumor surgery (glioma or metastasis)
- Lesion located within or adjacent to eloquent cortical or subcortical regions (motor, language, visual, or higher-order associative networks, as defined by clinical and neuroimaging criteria)
- Availability of standardized preoperative and early postoperative brain MRI (including high-direction diffusion tensor imaging, ≥64 directions)
- Consistent postoperative clinical follow-up planned at the treating centre
- Provision of informed consent for use of anonymized clinical and imaging data
Exclusion Criteria:
- Inability to undergo pre- or postoperative MRI
- Significant comorbidities precluding surgery
- Incomplete imaging or clinical data
- Purely infratentorial lesions or non-tumoral pathologies
