Functional Magnetic Resonance Imaging and Diffusion Tensor Imaging-Tractography in Resective Brain Surgery: Lesion Coverage Strategies and Patient Outcomes

Abstract

Diffusion tensor imaging (DTI)-tractography and functional magnetic resonance imaging (fMRI) have dynamically entered the presurgical evaluation context of brain surgery during the past decades, providing novel perspectives in surgical planning and lesion access approaches. However, their application in the presurgical setting requires significant time and effort and increased costs, thereby raising questions regarding efficiency and best use. In this work, we set out to evaluate DTI-tractography and combined fMRI/DTI-tractography during intra-operative neuronavigation in resective brain surgery using lesion-related preoperative neurological deficit (PND) outcomes as metrics. We retrospectively reviewed medical records of 252 consecutive patients admitted for brain surgery. Standard anatomical neuroimaging protocols were performed in 127 patients, 69 patients had additional DTI-tractography, and 56 had combined DTI-tractography/fMRI. fMRI procedures involved language, motor, somatic sensory, sensorimotor and visual mapping. DTI-tractography involved fiber tracking of the motor, sensory, language and visual pathways. At 1 month postoperatively, DTI-tractography patients were more likely to present either improvement or preservation of PNDs (p = 0.004 and p = 0.007, respectively). At 6 months, combined DTI-tractography/fMRI patients were more likely to experience complete PND resolution (p < 0.001). Low-grade lesion patients (N = 102) with combined DTI-tractography/fMRI were more likely to experience complete resolution of PNDs at 1 and 6 months (p = 0.001 and p < 0.001, respectively). High-grade lesion patients (N = 140) with combined DTI-tractography/fMRI were more likely to have PNDs resolved at 6 months (p = 0.005). Patients with motor symptoms (N = 80) were more likely to experience complete remission of PNDs at 6 months with DTI-tractography or combined DTI-tractography/fMRI (p = 0.008 and p = 0.004, respectively), without significant difference between the two imaging protocols (p = 1). Patients with sensory symptoms (N = 44) were more likely to experience complete PND remission at 6 months with combined DTI-tractography/fMRI (p = 0.004). The intraoperative neuroimaging modality did not have a significant effect in patients with preoperative seizures (N = 47). Lack of PND worsening was observed at 6 month follow-up in patients with combined DTI-tractography/fMRI. Our results strongly support the combined use of DTI-tractography and fMRI in patients undergoing resective brain surgery for improving their postoperative clinical profile.

Keywords: DTI-tractography; brain tumor surgery; fMRI; neuronavigation; presurgical evaluation.

Figure 1

Flowcharts of the fMRI tasks. (A) Verbal fluency task: patients were presented with a single letter and were instructed to silently generate words starting with that letter. (B) Semantic description task: patients were presented with a single word of objects and tools of household and urban life and were instructed to silently generate a complete sentence with a definition for this object or a description of its common use. (C) Reading comprehension task: patients were presented with sentences of brief factual statements and were instructed to read them silently as soon as they appeared on screen. (D) Listening comprehension task: patients were asked to listen carefully to a pre-recorded narrative and be prepared to answer questions on the content. (E) Motor, sensory and sensorimotor tasks: the patients were presented with the simple instructions “Go” to begin the activity and “Stop” to stop the activity. During the sensory task, “Go” and “Stop” were instructions for the functional neuroimager who was performing somatosensory stimulation. (F) Visual task: the patient was instructed to visually explore the complex colorful images. Durations for each item appear in green on the left lower corner of each item; this was not part of the presentation to the patients.

Figure 2

Examples highlighting the strategy and rationale of functional and structural peri-lesional coverage. Anatomical localization and morphological depiction (A), along with fMRI and DTI-tractography mapping (B), on post-G Gadolinium T1-weighted images for a precentral tumor. Anatomical localization and morphological depiction (C), along with fMRI and DTI-tractography mapping (D), on post-Gadolinium T1-weighted images for a central tumor.

Figure 3

Examples highlighting the strategy and rationale of functional and structural peri-lesional coverage. Anatomical localization and morphological depiction (A), along with fMRI and DTI-tractography mapping (B), on post-G Gadolinium T1-weighted images for an inferior frontal tumor. Anatomical localization and morphological depiction (C), along with fMRI and DTI-tractography mapping (D), on post-G Gadolinium T1-weighted images for an inferior parietal tumor.

Figure 4

Examples highlighting the strategy and rationale of functional and structural peri-lesional coverage. Anatomical localization and morphological depiction (A), along with fMRI and DTI-tractography mapping (B), on post-G Gadolinium T1-weighted images for a posterior quadrant tumor. Anatomical localization and morphological depiction (C), along with fMRI and DTI-tractography mapping (D), on post-G Gadolinium T1-weighted images for a multi-lobar (frontal/insular/temporal) tumor.