Patients with functionally eloquent brain lesions are at risk of functional decline in the course of resection. Given tumor-related plastic reshaping and reallocation of function, individual data are needed for patient counseling and risk assessment prior to surgery. This study evaluates the utility of mapping by navigated transcranial magnetic stimulation (nTMS) and nTMS-based diffusion tensor imaging fiber tracking (DTI FT) for individual risk evaluation of surgery-related decline of motor or language function in the clinical setting. In total, 250 preoperative nTMS mappings (100 language and 150 motor mappings) derived from 216 patients (mean age: 57.0 ± 15.5 years, 58.8% males; glioma World Health Organization (WHO) grade I & II: 4.2%, glioma WHO grade III & IV: 83.4%, arteriovenous malformations: 1.9%, cavernoma: 2.3%, metastasis: 8.2%) were included. Deterministic tractography based on nTMS motor or language maps as seed regions was performed with 25%, 50%, and 75% of the individual fractional anisotropy threshold (FAT). Lesion-to-tract distances (LTDs) were measured between the tumor mass and the corticospinal tract (CST), arcuate fascicle (AF), or other closest language-related tracts. LTDs were compared between patients and correlated to the functional status (no/transient/permanent surgery-related paresis or aphasia). Significant differences were found between patients with no or transient surgery-related deficits and patients with permanent surgery-related deficits regarding LTDs in relation to the CST (p < 0.0001), AF (p ≤ 0.0491), or other closest language-related tracts (p ≤ 0.0435). The cut-off values for surgery-related paresis or aphasia were ≤ 12 mm (LTD—CST) and ≤ 16 mm (LTD—AF) or ≤25 mm (LTD—other closest language-related tract), respectively. Moreover, there were significant associations between the status of surgery-related deficits and the LTD when considering the CST (range r: −0.3994 to −0.3910, p < 0.0001) or AF (range r: −0.2918 to −0.2592, p = 0.0135 and p = 0.0473 for 25% and 50% FAT). In conclusion, this is the largest study evaluating the application of both preoperative functional mapping and function-based tractography for motor and language function for risk stratification in patients with functionally eloquent tumors. The LTD may qualify as a viable marker that can be seamlessly assessed in the clinical neurooncological setup.

Navigated repetitive transcranial magnetic stimulation (nrTMS) is effective therapy for stroke patients. Neurorehabilitation could be supported by low-frequency stimulation of the non-damaged hemisphere to reduce transcallosal inhibition. The present study examines the effect of postoperative nrTMS therapy of the unaffected hemisphere in glioma patients suffering from acute surgery-related paresis of the upper extremity (UE) due to subcortical ischemia. We performed a randomized, sham-controlled, double-blinded trial on patients suffering from acute surgery-related paresis of the UE after glioma resection. Patients were randomly assigned to receive either low frequency nrTMS (1 Hz, 15 min) or sham stimulation directly before physical therapy for 7 consecutive days. We performed primary and secondary outcome measures on day 1, on day 7, and at a 3-month follow-up (FU). The primary endpoint was the change in Fugl-Meyer Assessment (FMA) at FU compared to day 1 after surgery. Compared to the sham stimulation, nrTMS significantly improved outcomes between day 1 and FU based on the FMA (mean [95% CI] +31.9 [22.6, 41.3] vs. +4.2 [-4.1, 12.5]; P = .001) and the National Institutes of Health Stroke Scale (NIHSS) (-5.6 [-7.5, -3.6] vs. -2.4 [-3.6, -1.2]; P = .02). To achieve a minimal clinically important difference of 10 points on the FMA scale, the number needed to treat is 2.19. The present results show that patients suffering from acute surgery-related paresis of the UE due to subcortical ischemia after glioma resection significantly benefit from low-frequency nrTMS stimulation therapy of the unaffected hemisphere. Local institutional registration: 12/15; ClinicalTrials.gov number: NCT03982329.

When using preoperative tractography intraoperatively, inaccuracies due to brain shift might occur. Intraoperative tractography is rarely performed. Elastic fusion (EF) is a tool developed to compensate for brain shift, gravity, and tissue resection based on intraoperative images. Our hypothesis was that preoperative tractography is accurate and adjustments of tractography by intraoperative magnetic resonance imaging (ioMRI)-based EF (IBEF) compensate for brain shift. Between February 2018 and June 2019, 78 patients underwent eloquent (46 motor, 32 language) glioma resection in our department using intraoperative MRI. Mean distances between the resection cavity and tractography were analyzed and correlated with clinical outcomes. The mean ± standard deviation (range) distance after the application of IBEF was 5.0 ± 2.9 mm (0–10 mm) in patients without surgery-related motor deficits compared with 1.1 ± 1.6 mm (0–5 mm) in patients who showed new permanent surgery-related motor deficits postoperatively (p < 0.001). For language, the distance was 0.7 ± 1.2 mm (0–2 mm) in patients with new permanent deficits compared with 3.1 ± 4.5 mm (0–14 mm) in patients without new permanent surgery-related language deficits (p = 0.541). Preoperative tractography corrected by IBEF for subcortical resection of gliomas is highly accurate. However, at least for such subcortical anatomy, the severity of brain shift was considerably overestimated in the past.

Objective: A considerable number of gliomas require resection via direct electrical stimulation (DES) during awake craniotomy. Likewise, the feasibility of resecting language-eloquent gliomas purely based on navigated repetitive transcranial magnetic stimulation (nrTMS) has been shown. This study analyzes the outcomes after preoperative nrTMS-based and intraoperative DES-based glioma resection in a large cohort. Due to the necessity of making location comparable, a classification for language eloquence for gliomas is introduced. Methods: Between March 2015 and May 2019, we prospectively enrolled 100 consecutive cases that were resected based on preoperative nrTMS language mapping (nrTMS group), and 47 cases via intraoperative DES mapping during awake craniotomy (awake group) following a standardized clinical workflow. Outcome measures were determined preoperatively, 5 days after surgery, and 3 months after surgery. To make functional eloquence comparable, we developed a classification based on prior publications and clinical experience. Groups and classification scores were correlated with clinical outcomes. Results: The functional outcome did not differ between groups. Gross total resection was achieved in more cases in the nrTMS group (87%, vs. 72% in the awake group, p = 0.04). Nonetheless, the awake group showed significantly higher scores for eloquence than the nrTMS group (median 7 points; interquartile range 6–8 vs. 5 points; 3–6.75; p < 0.0001). Conclusion: Resecting language-eloquent gliomas purely based on nrTMS data is feasible in a high percentage of cases if the described clinical workflow is followed. Moreover, the proposed classification for language eloquence makes language-eloquent tumors comparable, as shown by its correlation with functional and radiological outcomes.

Preoperative language mapping with navigated transcranial magnetic stimulation (nTMS) is currently based on the disruption of performance during object naming. The resulting cortical language maps, however, lack accuracy when compared to intraoperative mapping. The question arises whether nTMS results can be improved, when another language task is considered, involving verb retrieval in sentence context. Twenty healthy German speakers were tested with object naming and a novel action naming task during nTMS language mapping. Error rates and categories in both hemispheres were compared. Action naming showed a significantly higher error rate than object naming in both hemispheres. Error category comparison revealed that this discrepancy stems from more lexico-semantic errors during action naming, indicating lexico-semantic retrieval of the verb being more affected than noun retrieval. In an area-wise comparison, higher error rates surfaced in multiple right-hemisphere areas, but only trends in the left ventral postcentral gyrus and middle superior temporal gyrus. Hesitation errors contributed significantly to the error count, but did not dull the mapping results. Inclusion of action naming coupled with a detailed error analysis may be favorable for nTMS mapping and ultimately improve accuracy in preoperative planning. Moreover, the results stress the recruitment of both left- and right-hemispheric areas during naming.