Noriko Tamura, Motohiro Hayashi, Mikhail Chernov, Manabu Tamura, Ayako Horiba, Yoshiyuki Konishi, Yoshihiro Muragaki, Hiroshi Iseki and Yoshikazu Okada
The focus of the present study was the evaluation of outcomes after unstaged and staged-volume Gamma Knife surgery (GKS) in children harboring intracranial arteriovenous malformations (AVMs).
Twenty-two children (median age 9.5 years) underwent GKS for AVMs and were followed up for at least 2 years thereafter. The disease manifested with intracranial hemorrhage in 77% of cases. In 68% of patients the lesion affected eloquent brain structures. The volume of the nidus ranged from 0.1 to 6.7 cm3. Gamma Knife surgery was guided mainly by data from dynamic contrast-enhanced CT scans, with preferential targeting of the junction between the nidus and draining vein. The total prescribed isodose volume was kept below 4.0 cm3, and the median margin dose was 22 Gy (range 20–25 Gy). If the volume of the nidus was larger than 4.0 cm3, a second radiosurgical session was planned for 3–4 years after the first one. Nine patients in the present series underwent unstaged radiosurgery, whereas staged-volume treatment was scheduled in 13 patients.
Complete obliteration of the AVM was noted in 17 (77%) of 22 patients within a median period of 47 months after the last radiosurgical session. Complete obliteration of the lesion occurred in 89% of patients after unstaged treatment and in 62.5% after staged GKS. Four (67%) of 6 high-grade AVMs were completely obliterated. Complications included 3 bleeding episodes, the appearance of a region of hyperintensity on T2-weighted MR images in 2 patients who had no symptoms, and reappearance of the nidus in the vicinity of the completely obliterated AVM in 1 patient.
Radiosurgery is a highly effective management option for intracranial AVMs in children. For larger lesions, staged GKS may be applied successfully. Initial targeting of the nidus adjacent to the draining vein and application of a sufficient radiation dose to a relatively small volume (≤ 4 cm3) provides a good balance between a high probability of obliteration and a low risk of treatment-related complications.
Masayuki Nitta, Yoshihiro Muragaki, Takashi Maruyama, Soko Ikuta, Takashi Komori, Katsuya Maebayashi, Hiroshi Iseki, Manabu Tamura, Taiichi Saito, Saori Okamoto, Mikhail Chernov, Motohiro Hayashi and Yoshikazu Okada
There is no standard therapeutic strategy for low-grade glioma (LGG). The authors hypothesized that adjuvant therapy might not be necessary for LGG cases in which total radiological resection was achieved. Accordingly, they established a treatment strategy based on the extent of resection (EOR) and the MIB-1 index: patients with a high EOR and low MIB-1 index were observed without postoperative treatment, whereas those with a low EOR and/or high MIB-1 index received radiotherapy (RT) and/or chemotherapy. In the present retrospective study, the authors reviewed clinical data on patients with primarily diagnosed LGGs who had been treated according to the above-mentioned strategy, and they validated the treatment policy. Given their results, they will establish a new treatment strategy for LGGs stratified by EOR, histological subtype, and molecular status.
One hundred fifty-three patients with diagnosed LGG who had undergone resection or biopsy at Tokyo Women's Medical University between January 2000 and August 2010 were analyzed. The patients consisted of 84 men and 69 women, all with ages ≥ 15 years. A total of 146 patients underwent surgical removal of the tumor, and 7 patients underwent biopsy.
Postoperative RT and nitrosourea-based chemotherapy were administered in 48 and 35 patients, respectively. Extent of resection was significantly associated with both overall survival (OS; p = 0.0096) and progression-free survival (PFS; p = 0.0007) in patients with diffuse astrocytoma but not in those with oligodendroglial subtypes. Chemotherapy significantly prolonged PFS, especially in patients with oligodendroglial subtypes (p = 0.0009). Patients with a mutant IDH1 gene had significantly longer OS (p = 0.034). Multivariate analysis did not identify MIB-1 index or RT as prognostic factors, but it did identify chemotherapy as a prognostic factor for PFS and EOR as a prognostic factor for OS and PFS.
The findings demonstrated that EOR was significantly correlated with patient survival; thus, one should aim for maximum tumor resection. In addition, patients with a higher EOR can be safely observed without adjuvant therapy. For patients with partial resection, postoperative chemotherapy should be administered for those with oligodendroglial subtypes, and repeat resection should be considered for those with astrocytic tumors. More aggressive treatment with RT and chemotherapy may be required for patients with a poor prognosis, such as those with diffuse astrocytoma, 1p/19q nondeleted tumors, or IDH1 wild-type oligodendroglial tumors with partial resection.
Taiichi Saito, Yoshihiro Muragaki, Takashi Maruyama, Manabu Tamura, Masayuki Nitta, Shunsuke Tsuzuki, Yoshiyuki Konishi, Kotoe Kamata, Ryuta Kinno, Kuniyoshi L. Sakai, Hiroshi Iseki and Takakazu Kawamata
Identification of language areas using functional brain mapping is sometimes impossible using current methods but essential to preserve language function in patients with gliomas located within or near the frontal language area (FLA). However, the factors that influence the failure to detect language areas have not been elucidated. The present study evaluated the difficulty in identifying the FLA in dominant-side frontal gliomas that involve the pars triangularis (PT) to determine the factors that influenced failed positive language mapping.
Awake craniotomy was performed on 301 patients from April 2000 to October 2013 at Tokyo Women's Medical University. Recurrent cases were excluded, and patients were also excluded if motor mapping indicated their glioma was in or around the motor area on the dominant or nondominant side. Eighty-two consecutive cases of primary frontal glioma on the dominant side were analyzed for the present study. MRI was used for all patients to evaluate whether tumors involved the PT and to perform language functional mapping with a bipolar electrical stimulator. Eighteen of 82 patients (mean age 39 ± 13 years) had tumors that showed involvement of the PT, and the detailed characteristics of these 18 patients were examined.
The FLA could not be identified with intraoperative brain mapping in 14 (17%) of 82 patients; 11 (79%) of these 14 patients had a tumor involving the PT. The negative response rate in language mapping was only 5% in patients without involvement of the PT, whereas this rate was 61% in patients with involvement of the PT. Univariate analyses showed no significant correlation between identification of the FLA and sex, age, histology, or WHO grade. However, failure to identify the FLA was significantly correlated with involvement of the PT (p < 0.0001). Similarly, multivariate analyses with the logistic regression model showed that only involvement of the PT was significantly correlated with failure to identify the FLA (p < 0.0001). In 18 patients whose tumors involved the PT, only 1 patient had mild preoperative dysphasia. One week after surgery, language function worsened in 4 (22%) of 18 patients. Six months after surgery, 1 (5.6%) of 18 patients had a persistent mild speech deficit. The mean extent of resection was 90% ± 7.1%.
Identification of the FLA can be difficult in patients with frontal gliomas on the dominant side that involve the PT, but the positive mapping rate of the FLA was 95% in patients without involvement of the PT. These findings are useful for establishing a positive mapping strategy for patients undergoing awake craniotomy for the treatment of frontal gliomas on the dominant side. Thoroughly positive language mapping with subcortical electrical stimulation should be performed in patients without involvement of the PT. More careful continuous neurological monitoring combined with subcortical electrical stimulation is needed when removing dominant-side frontal gliomas that involve the PT.
Taiichi Saito, Manabu Tamura, Yoshihiro Muragaki, Takashi Maruyama, Yuichi Kubota, Satoko Fukuchi, Masayuki Nitta, Mikhail Chernov, Saori Okamoto, Kazuhiko Sugiyama, Kaoru Kurisu, Kuniyoshi L. Sakai, Yoshikazu Okada and Hiroshi Iseki
The objective in the present study was to evaluate the usefulness of cortico-cortical evoked potentials (CCEP) monitoring for the intraoperative assessment of speech function during resection of brain tumors.
Intraoperative monitoring of CCEP was applied in 13 patients (mean age 34 ± 14 years) during the removal of neoplasms located within or close to language-related structures in the dominant cerebral hemisphere. For this purpose strip electrodes were positioned above the frontal language area (FLA) and temporal language area (TLA), which were identified with direct cortical stimulation and/or preliminary mapping with the use of implanted chronic subdural grid electrodes. The CCEP response was defined as the highest observed negative peak in either direction of stimulation. In 12 cases the tumor was resected during awake craniotomy.
An intraoperative CCEP response was not obtained in one case because of technical problems. In the other patients it was identified from the FLA during stimulation of the TLA (7 cases) and from the TLA during stimulation of the FLA (5 cases), with a mean peak latency of 83 ± 15 msec. During tumor resection the CCEP response was unchanged in 5 cases, decreased in 4, and disappeared in 3. Postoperatively, all 7 patients with a decreased or absent CCEP response after lesion removal experienced deterioration in speech function. In contrast, in 5 cases with an unchanged intraoperative CCEP response, speaking abilities after surgery were preserved at the preoperative level, except in one patient who experienced not dysphasia, but dysarthria due to pyramidal tract injury. This difference was statistically significant (p < 0.01). The time required to recover speech function was also significantly associated with the type of intraoperative change in CCEP recordings (p < 0.01) and was, on average, 1.8 ± 1.0, 5.5 ± 1.0, and 11.0 ± 3.6 months, respectively, if the response was unchanged, was decreased, or had disappeared.
Monitoring CCEP is feasible during the resection of brain tumors affecting language-related cerebral structures. In the intraoperative evaluation of speech function, it can be a helpful adjunct or can be used in its direct assessment with cortical and subcortical mapping during awake craniotomy. It can also be used to predict the prognosis of language disorders after surgery and decide on the optimal resection of a neoplasm.