PURPOSE:Interstitial photodynamic therapy (iPDT) is an emerging loco-regional treatment for glioblastoma based on intracerebral illumination delivered through stereotactically implanted optical fibers. Despite encouraging oncological results, treatment-related morbidity remains significant, and the contribution of unintended light-induced thermal effects has never been experimentally investigated. This study aimed to preclinically characterize temperature changes during illumination using MRI-based thermometry. METHODS:Fifty-seven minutes of illumination were performed in ex vivo calf brain model at room temperature baseline. The temperature was monitored using the proton resonance frequency shift (PRFS) method on a 1.5 T MRI system. Thermometric measurements were first validated against fiber-optic temperature probes during single diffusing fiber illumination. Spatio-temporal thermal distributions were then evaluated for three diffusing fiber configurations with inter-fiber spacings of 2.0, 1.0, and 0.6 cm. Macroscopic assessment of thermal tissue damage was performed after illumination. RESULTS:MRI thermometry showed excellent agreement with fiber optic temperature probes. Cumulative heating strongly depended on inter-fiber spacing. At the barycenter of the diffusing part of the three fibers, the mean ± SD final temperature estimated was 43.7 ± 0.7 °C at 2.0 cm spacing, 46.6 ± 5.9 °C at 1.0 cm spacing, and 53.5 ± 10.2 °C at 0.6 cm spacing. No macroscopic thermal tissue damage was observed. CONCLUSION:MRI thermometry provides a robust and translatable tool to quantify thermal effects during iPDT. Diffusing fiber spacing critically influences temperature distribution, supporting the use of thermometric monitoring to enhance procedural safety. Future clinical thermometric data could be implemented into treatments planning systems to refine fiber placement and enable patient-specific strategies aimed at maximizing therapeutic efficacy while minimizing the risk of thermal injury.

BACKGROUND:Awake surgery and intraoperative magnetic resonance imaging optimize the extent of resection while preserving neurological function. Diffusion tensor imaging (DTI) enables reconstruction of subcortical white matter bundles, allowing assessment of compression, partial invasion, or destruction. PRIMARY ENDPOINT:evaluate the feasibility of intraoperative fiber tractography (iFT); secondary endpoints: identify factors influencing tractography quality and assess its clinical value. METHODS:A two-phase study was conducted. A retrospective phase: all patients undergoing intraoperative magnetic resonance imaging-guided surgery at the Neurosurgery Department of Lille University Hospital between 2014 and 2021 were screened. Inclusion criteria included age >18 years, intraparenchymal pathology requiring resection, and availability of both preoperative and intraoperative DTI. In this proof-of-concept cohort, resections were systematically performed during awake surgery allowing clinical correlation. Prospective validation phase (November 2021-September 2022): resections were guided using the established intraoperative tractography protocol. Subcortical bundles of interest were identified, and qualitative and quantitative factors affecting tract reconstruction were analyzed. RESULTS:Fifty-nine DTI sequences from 24 patients (mean age 41 years; range 22-66) were analyzed, including drug-resistant epilepsy, low- and high-grade gliomas, and metastases. Evaluated bundles were: arcuate fasciculus, inferior fronto-occipital fasciculus, corticospinal tract, and optic radiations. Awake surgery was performed in 75% of cases. Intraoperative DTI was interpretable in 83%, and iFT was reconstructed in 79%. Air presence was the main limiting factor, correlating with blood artifacts, ischemic changes, and saline filling of the resection cavity. Intraoperative FT correlated with color map visualization and preoperative tract reconstruction. CONCLUSION:iFT is feasible and clinically valuable, although its accuracy is significantly affected by air-related artifacts.

INTRODUCTION:The combined use of intraoperative MRI and awake surgery is a tailored microsurgical resection to respect functional neural networks (mainly the language and motor ones). Intraoperative MRI has been classically considered to increase the extent of resection for gliomas, thereby reducing neurological deficits. Herein, we evaluated the combined technique of awake microsurgical resection and intraoperative MRI for primary brain tumours (gliomas, metastasis) and epilepsy (cortical dysplasia, non-lesional, cavernomas).PATIENTS AND METHODS:Eighteen patients were treated with the commonly used "asleep awake asleep" (AAA) approach at Lille University Hospital, France, from November 2016 until May 2020. The exact anatomical location was insular with various extensions, frontal, temporal or fronto-temporal in 8 (44.4%), parietal in 3 (16.7%), fronto-opercular in 4 (22.2%), Rolandic in two (11.1%), and the supplementary motor area (SMA) in one (5.6%).RESULTS:The patients had a mean age of 38.4 years (median 37.1, range 20.8-66.9). The mean surgical duration was 4.1 hours (median 4.2, range 2.6-6.4) with a mean duration of intraoperative MRI of 28.8 minutes (median 25, range 13-55). Overall, 61% (11/18) of patients underwent further resection, while 39% had no additional resection after intraoperative MRI. The mean preoperative and postoperative tumour volumes of the primary brain tumours were 34.7 cc (median 10.7, range 0.534-130.25) and 3.5 cc (median 0.5, range 0-17.4), respectively. Moreover, the proportion of the initially resected tumour volume at the time of intraoperative MRI (expressed as 100% from preoperative volume) and the final resected tumour volume were statistically significant (p= 0.01, Mann-Whitney test). The tumour remnants were commonly found posterior (5/9) or anterior (2/9) insular and in proximity with the motor strip (1/9) or language areas (e.g. Broca, 1/9). Further resection was not required in seven patients because there were no remnants (3/7), cortical stimulation approaching eloquent areas (3/7) and non-lesional epilepsy (1/7). The mean overall follow-up period was 15.8 months (median 12, range 3-36).CONCLUSION:The intraoperative MRI and awake microsurgical resection approach is feasible with extensive planning and multidisciplinary collaboration, as these methods are complementary and synergic rather than competitive to improve patient oncological outcomes and quality of life.

Cavernous malformations are clusters of abnormal and hyalinized capillaries without interfering brain tissue. Here, we present a cavernous malformation operated under awake conditions, due to location, in an eloquent area and using intraoperative magnetic resonance imaging due to patient’s movement upon the awake phase. We present the pre-, per-, and postoperative course of an inferior parietal cavernous malformation, located in eloquent area, in a 27-year-old right-handed Caucasian male, presenting with intralesional hemorrhage and epilepsy. Preoperative diffusion tensor imaging has shown the cavernous malformation at the interface between the arcuate fasciculus and the inferior fronto-occipital fasciculus. We describe the microsurgical approach, combining preoperative diffusion tensor imaging, neuronavigation, awake microsurgical resection, and intraoperative magnetic resonance imaging. Complete microsurgical en bloc resection has been performed and is feasible even in eloquent locations. Intraoperative magnetic resonance imaging was considered an important adjunct, particularly used in this case as the patient moved during the “awake" phase of the surgery and thus neuronavigation was not accurate anymore. Postoperative course was marked by a unique, generalized seizure without any adverse event. Immediate and 3 months postoperative magnetic resonance imaging confirmed the absence of any residue. Pre- and postoperative neuropsychological exams were unremarkable.

Abstract IntroductionEpilepsy is a major feature of multiple types of lesional epilepsy (LE) and non-lesional (NLE). However, there is an important heterogeneity in the use of such nomenclature. A favorable outcome is correlated to an identifiable cause and the extent of resection. Here, we evaluate the use of intraoperative MRI (iMRI) combined with neuronavigation, in improving the complete and/or adequate resection rates in patients with intractable epilepsy of various etiologies. We further discuss whether this results in better clinical outcomes. Material and methodsA medical librarian performed a comprehensive review of literature, searching the Medline®, Embase®, Index Medicus® and Cochrane® databases. Two reviewers independently applied following inclusion criteria: reports of at least two cases of any age undergoing microsurgical resection for epilepsy surgery, the use of iMRI as mandatory for intraoperative assessment, evaluation of LE versus NLE, describing if second resection after iMRI was necessary, detailing the extent of resection, qualitative reports of seizure outcome (Engel class). ResultsWe report 15 studies, encompassing 867 patients, with various types of LE and NLE. Most common surgical indications for the use of iMRI in LE were dysembrioplastic neuroepithelial tumor (DNET) and gangliogliomas, cavernomas, hypothalamic hamartomas, or primary brain tumors. For NLE these were sclerosis, dysplasia, ischemia and gliosis, which are further separately detailed in a supplementary material. With the use of iMRI, the rate of complete or adequate resection significantly increased from 74.8% (95% confidence interval (CI) 67.3-82.3%) to 97.6% (95% confidence interval (CI) 95.8-99.3%). Hundred and twenty-six cases out of 820 underwent additional resection after iMRI, for an overall rate of 24.4% (range 16.9-31.8). This was translated in an overall clinical benefit with 77.1% (95% CI 71-83.2%) Engel class I at last follow-up (p<0.001).DiscussionIntraoperative MRI might improve the extent of resection in LE and NLE. This might translate into a favorable clinical outcome (Engel class I as high as 77%). This is fairly comparable to what one expects in temporal lobe epilepsy surgery. We suggest that iMRI is beneficial when proposed in selected cases of intractable epilepsy. However, the reader should take into account that such data are not provided by randomized controlled trials and that such conclusions should be carefully interpreted.

Microsurgical resection of primary brain tumors located within or near eloquent areas is challenging. Primary aim is to preserve neurological function, while maximizing the extent of resection (EOR), to optimize long-term neurooncological outcomes and quality of life. Here, we review the combined integration of awake craniotomy and intraoperative MRI (IoMRI) for primary brain tumors, due to their multiple challenges. A systematic review of the literature was performed, in accordance with the Prisma guidelines. Were included 13 series and a total number of 527 patients, who underwent 541 surgeries. We paid particular attention to operative time, rate of intraoperative seizures, rate of initial complete resection at the time of first IoMRI, the final complete gross total resection (GTR, complete radiological resection rates), and the immediate and definitive postoperative neurological complications. The mean duration of surgery was 6.3 h (median 7.05, range 3.8-7.9). The intraoperative seizure rate was 3.7% (range 1.4-6; I<^>2 = 0%, P heterogeneity = 0.569, standard error = 0.012, p = 0.002). The intraoperative complete resection rate at the time of first IoMRI was 35.2% (range 25.7-44.7; I<^>2 = 66.73%, P heterogeneity = 0.004, standard error = 0.048, p < 0.001). The rate of patients who underwent supplementary resection after one or several IoMRI was 46% (range 39.8-52.2; I<^>2 = 8.49%, P heterogeneity = 0.364, standard error = 0.032, p < 0.001). The GTR rate at discharge was 56.3% (range 47.5-65.1; I<^>2 = 60.19%, P heterogeneity = 0.01, standard error = 0.045, p < 0.001). The rate of immediate postoperative complications was 27.4% (range 15.2-39.6; I<^>2 = 92.62%, P heterogeneity < 0.001, standard error = 0.062, p < 0.001). The rate of permanent postoperative complications was 4.1% (range 1.3-6.9; I<^>2 = 38.52%, P heterogeneity = 0.123, standard error = 0.014, p = 0.004). Combined use of awake craniotomy and IoMRI can help in maximizing brain tumor resection in selected patients. The technical obstacles to doing so are not severe and can be managed by experienced neurosurgery and anesthesiology teams. The benefits of bringing these technologies to bear on patients with brain tumors in or near language areas are obvious. The lack of equipoise on this topic by experienced practitioners will make it difficult to do a prospective, randomized, clinical trial. In the opinion of the authors, such a trial would be unnecessary and would deprive some patients of the benefits of the best available methods for their tumor resections.

Fronto-temporo-insular (FTI) gliomas continue to represent a surgical challenge despite numerous technical advances. Some authors advocate for surgery in awake condition even for non-dominant hemisphere FTI, due to risk of sociocognitive impairment. Here, we report outcomes in a series of patients operated using intraoperative magnetic resonance imaging (IoMRI) guided surgery under general anesthesia, using no cortical or subcortical mapping. We evaluated the extent of resection, functional and neuropsychological outcomes after IoMRI guided surgery under general anesthesia of FTI gliomas located in the non-dominant hemisphere. Twenty patients underwent FTI glioma resection using IoMRI in asleep condition. Seventeen tumors were de novo, three were recurrences. Tumor WHO grades were II:12, III:4, IV:4. Patients were evaluated before and after microsurgical resection, clinically, neuropsychologically (i.e., social cognition) and by volumetric MR measures (T1G+ for enhancing tumors, FLAIR for non-enhancing). Fourteen (70%) patients benefited from a second IoMRI. The median age was 33.5 years (range 24–56). Seizure was the inaugural symptom in 71% of patients. The median preoperative volume was 64.5 cm 3 (min 9.9, max 211). Fourteen (70%) patients underwent two IoMRI. The final median EOR was 92% (range 69–100). The median postoperative residual tumor volume (RTV) was 4.3 cm 3 (range 0–38.2). A vast majority of residual tumors were located in the posterior part of the insula. Early postoperative clinical events (during hospital stay) were three transient left hemiparesis (which lasted less than 48 h) and one prolonged left brachio-facial hemiparesis. Sixty percent of patients were free of any symptom at discharge. The median Karnofsky Performance Score was of 90 both at discharge and at 3 months. No significant neuropsychological impairment was reported, excepting empathy distinction in less than 40% of patients. After surgery, 45% of patients could go back to work. In our experience and using IoMRI as an adjunct, microsurgical resection of non-dominant FTI gliomas under general anesthesia is safe. Final median EOR was 92%, with a vast majority of residual tumors located in the posterior insular part. Patients experienced minor neurological and neuropsychological morbidity. Moreover, neuropsychological evaluation reported a high preservation of sociocognitive abilities. Solely empathy seemed to be impaired in some patients.

Le défi de la prise en charge neurochirurgicale des tumeurs gliales consiste à réaliser une ablation tumorale aussi complète que possible sans déficit fonctionnel postopératoire. La technique de tenseur de diffusion (DTI) permet l’identification non-invasive des fibres de substance blanche et leurs rapports avec la tumeur. Cependant, il reste soumis à de nombreuses limitations techniques et sa fiabilité en tant qu’outil de planification pré- et peropératoire est remise en question. La plupart des études de validation ont été réalisées en comparaison avec la stimulation corticale peropératoire, mais aucune n’a été réalisée sur l’analyse anatomopathologique de la tumeur pour évaluer l’intégrité des fibres de la substance blanche dans l’environnement tumoral. L’objectif de l’étude est donc d’évaluer la faisabilité d’une corrélation anatomo-radiologique entre le DTI et l’analyse histologique des tumeurs gliales. 12 patients atteints de tumeurs gliales de grade II à IV ont bénéficié d’une IRM avec séquence de DTI préopératoire. Les fibres de substance blanche détruites et infiltrées ont été identifiées à l’aide de la carte d’anisotropie fractionnelle à codage colorimétrique directionnel (carte RVB) par évaluation visuelle. Les corrélations DTI et histologiques ont été réalisées sur 31 régions, classées selon le degré d’infiltration tumorale, la destruction des fibres de myéline et des neurofilaments, la présence de fibres de substance blanche organisée et leur orientation dans l’espace. Les corrélations observées entre le DTI et les analyses histologiques concernant l’orientation et la destruction des fibres sont élevées. Selon le niveau de myéline intacte évalué à l’aide de l’histologie et des fibres détectées sur les images DTI, une concordance est observée pour 23 à 26 des 31 zones étudiées. La concordance directionnelle des faisceaux est également observée chez 9 patients sur les 10 présentant des fibres organisées en DTI. Cette étude a démontré qu’il est possible d’obtenir une corrélation entre les images DTI et l’étude histologique des fibres de myéline dans les tumeurs gliales et a confirmé qu’une analyse visuelle de la carte RVB peut être une méthode utile pour évaluer l’infiltration de substance blanche dans les tumeurs gliales. Cependant, d’autres études sont nécessaires pour évaluer la corrélation avec les données cliniques afin de déterminer le seuil de destruction auquel le faisceau perd sa fonction (Fig. 1).