Hugues Duffau, Peggy Gatignol, Emmanuel Mandonnet, Laurent Capelle and Luc Taillandier
Despite better knowledge of cortical language organization, its subcortical anatomofunctional connectivity remains poorly understood. The authors used intraoperative subcortical stimulation in awake patients undergoing operation for a glioma in the left dominant hemisphere to map the language pathways and to determine the contribution of such a method to surgical results.
One hundred fifteen patients harboring a World Health Organization Grade II glioma within language areas underwent operation after induction of local anesthesia, using direct electrical stimulation to perform online cortical and subcortical language mapping throughout the resection.
After detection of cortical language sites, the authors identified 1 or several of the following subcortical language pathways in all patients: 1) arcuate fasciculus, eliciting phonemic paraphasia when stimulated; 2) inferior frontooccipital fasciculus, generating semantic paraphasia when stimulated; 3) subcallosal fasciculus, inducing transcortical motor aphasia during stimulation; 4) frontoparietal phonological loop, eliciting speech apraxia during stimulation; and 5) fibers coming from the ventral premotor cortex, inducing anarthria when stimulated. These structures were preserved, representing the limits of the resection. Despite a transient immediate postoperative worsening, all but 2 patients (98%) returned to baseline or better. On control MR imaging, 83% of resections were total or subtotal.
These results represent the largest experience with human subcortical language mapping ever reported. The use of intraoperative cortical and subcortical stimulation gives a unique opportunity to perform an accurate and reliable real-time anatomofunctional study of language connectivity. Such knowledge of the individual organization of language networks enables practitioners to optimize the benefit-to-risk ratio of surgery for Grade II glioma within the left dominant hemisphere.
Laurent Capelle, Denys Fontaine, Emmanuel Mandonnet, Luc Taillandier, Jean Louis Golmard, Luc Bauchet, Johan Pallud, Philippe Peruzzi, Marie Hélène Baron, Michèle Kujas, Jacques Guyotat, Remi Guillevin, Marc Frenay, Sophie Taillibert, Philippe Colin, Valérie Rigau, Fanny Vandenbos, Catherine Pinelli, Hugues Duffau and for the French Réseau d'Étude des Gliomes (REG)
The spontaneous prognostic factors and optimal therapeutic strategy for WHO Grade II gliomas (GIIGs) have yet to be unanimously defined. Specifically, the role of resection is still debated, most notably because the actual amount of resection has seldom been assessed.
Cases of GIIGs treated before December 2007 were extracted from a multicenter database retrospectively collected since January 1985 and prospectively collected since 1996. Inclusion criteria were a patient age ≥ 18 years at diagnosis, histological diagnosis of WHO GIIG, and MRI evaluation of tumor volume at diagnosis and after initial surgery. One thousand ninety-seven lesions were included in the analysis. The mean follow-up was 7.4 years since radiological diagnosis. Factors significant in a univariate analysis (with a p value ≤ 0.1) were included in the multivariate Cox proportional hazard regression model analysis.
At the time of radiological diagnosis, independent spontaneous factors of a poor prognosis were an age ≥ 55 years, an impaired functional status, a tumor location in a nonfrontal area, and, most of all, a larger tumor size. When the study starting point was set at the time of first treatment, independent favorable prognostic factors were limited to a smaller tumor size, an epileptic symptomatology, and a greater extent of resection.
This large series with its volumetric assessment refines the prognostic value of previously stressed clinical and radiological parameters and highlights the importance of tumor size and location. The results support additional arguments in favor of the predominant role of resection, in accordance with recently reported experiences.