Mitsutoshi Nakada, Syuichi Akaike and Kazuya Futami
Yutaka Hayashi, Masashi Kinoshita, Mitsutoshi Nakada and Jun-ichiro Hamada
Disturbance of the arcuate fasciculus in the dominant hemisphere is thought to be associated with language-processing disorders, including conduction aphasia. Although the arcuate fasciculus can be visualized in vivo with diffusion tensor imaging (DTI) tractography, its involvement in functional processes associated with language has not been shown dynamically using DTI tractography. In the present study, to clarify the participation of the arcuate fasciculus in language functions, postoperative changes in the arcuate fasciculus detected by DTI tractography were evaluated chronologically in relation to postoperative changes in language function after brain tumor surgery.
Preoperative and postoperative arcuate fasciculus area and language function were examined in 7 right-handed patients with a brain tumor in the left hemisphere located in proximity to part of the arcuate fasciculus. The arcuate fasciculus was depicted, and its area was calculated using DTI tractography. Language functions were measured using the Western Aphasia Battery (WAB).
After tumor resection, visualization of the arcuate fasciculus was increased in 5 of the 7 patients, and the total WAB score improved in 6 of the 7 patients. The relative ratio of postoperative visualized area of the arcuate fasciculus to preoperative visualized area of the arcuate fasciculus was increased in association with an improvement in postoperative language function (p = 0.0039).
The role of the left arcuate fasciculus in language functions can be evaluated chronologically in vivo by DTI tractography after brain tumor surgery. Because increased postoperative visualization of the fasciculus was significantly associated with postoperative improvement in language functions, the arcuate fasciculus may play an important role in language function, as previously thought. In addition, postoperative changes in the arcuate fasciculus detected by DTI tractography could represent a predicting factor for postoperative language-dependent functional outcomes in patients with brain tumor.
Riho Nakajima, Masashi Kinoshita, Tetsutaro Yahata and Mitsutoshi Nakada
Supplementary motor area (SMA) syndrome is defined as temporary paralysis after the resection of brain tumor localized in the SMA. Although in most cases paralysis induced by SMA resection resolves within a short period, the time until complete recovery varies and has not been precisely analyzed to date. In this study, the authors investigated factors for predicting the time required for recovery from paralysis after SMA resection.
Data from 20 cases were analyzed. All 20 patients (mean age 54.9 ± 12.6 years) had undergone resection of frontal lobe glioma involving the SMA. The severity of postoperative paralysis was recorded until complete recovery using the Brunnstrom recovery stage index. To investigate factors associated with recovery time, the authors performed multivariate analysis with the following potentially explanatory variables: age, severity of paralysis after the surgery, resected volume of the SMA, and probability of disconnection of fibers running through or near the SMA. Moreover, voxel-based lesion symptom analysis was performed to clarify the resected regions related to prolonged recovery.
In most cases of severe to moderate paralysis, there was substantial improvement within the 1st postoperative week, but 2–9 weeks were required for complete recovery. Significantly delayed recovery from paralysis was associated with resection of the cingulate cortex and its deep regions. The factors found to influence recovery time from paralysis were stage of paralysis at postoperative day 7 and disconnection probability of the cingulum (adjusted R2 = 0.63, p < 0.0001).
Recovery time from paralysis due to SMA syndrome can be predicted by the severity of paralysis at postoperative day 7 and degree of damage to the cingulum.
Yutaka Hayashi, Mitsutoshi Nakada, Shingo Tanaka, Naoyuki Uchiyama, Yasuhiko Hayashi, Daisuke Kita and Jun-ichiro Hamada
Fluorescence-guided surgery using 5-aminolevulinic acid (5-ALA) has been broadly recognized as a beneficial tool for the resection of glioblastoma multiforme (GBM). Fluorescence in the ventricular walls, which were apparently free of macroscopic tumor or MR imaging enhancement indicative of CSF dissemination, was detected during surgery for GBM. To evaluate the clinical significance of the 5-ALA fluorescence, the authors resected ventricle wall tissue together with the adjacent tumors for pathological examination and then followed up the clinical courses of the patients.
Seven consecutive GBMs located near the lateral ventricle were surgically treated using a fluorescence-guided technique with 5-ALA at the authors' hospital since acquiring instrumentation for the detection of 5-ALA fluorescence in 2007. All of the procedures were performed using a ventricular entry, and 5-ALA fluorescence of the ventricular wall was detected despite the absence of macroscopic tumor invasion of the wall.
A pathological examination of the resected ventricular wall tissues revealed tumor cells in 6 of the 7 cases and disruption of the ependymal cell layer in all 7 cases. Delayed communicating hydrocephalus followed surgery in all 7 patients, and ventricular wall enhancements on MR imaging were demonstrated after hydrocephalus in 2 of the patients.
Data in this study suggest that 5-ALA fluorescence of the ventricular wall may be predictive of postoperative hydrocephalus associated with CSF dissemination even in cases without evidence of CSF dissemination on MR imaging studies before surgery. The authors also speculate that postoperative radiotherapy covering the whole ventricular system may be a better therapeutic option for these patients.
Masashi Kinoshita, Harumichi Shinohara, Osamu Hori, Noriyuki Ozaki, Fumiaki Ueda, Mitsutoshi Nakada, Jun-ichiro Hamada and Yutaka Hayashi
Recently, intraoperative mapping has disclosed that, in addition to the classic language centers (that is, the Broca and Wernicke centers), other cortical regions may also play an important role in language organization. In the prefrontal cortex, although the lateral superior frontal gyrus (LSFG) could have language-related functions, there are no detailed reports that demonstrate the anatomical connection between the LSFG and other well-known language cortices, such as the Broca center. To show the existence of the structural connection, white matter association fibers between the inferior frontal gyrus (IFG) and the LSFG were examined using fiber dissection (FD) and diffusion tensor (DT) imaging–based tractography.
Eight cadaveric cerebral hemispheres were dissected to reveal the association fibers between the IFG and LSFG. The DT imaging–based tractography studies targeting the prefrontal cortex were obtained in 53 right-handed patients who had no organic cerebral lesions.
The association fiber tract between Brodmann area 44/45 (the Broca center in the dominant hemisphere) and LSFG were detected in all specimens by FD. In the DT imaging–based tractography studies, the tract was identified in all patients bilaterally, except for the 4 in whom the tract was detected only in the left hemisphere. This tract was spread significantly wider in the left than in the right hemisphere, and left lateralization was evident in male patients.
Based on its character, this tract was named the Broca-LSFG pathway. These findings suggest a close relationship between this pathway and language organization. The structural anatomy of the Broca-LSFG pathway may explain speech disturbances induced by LSFG stimulation that are sometimes observed during intraoperative language mapping.
Masashi Kinoshita, Riho Nakajima, Harumichi Shinohara, Katsuyoshi Miyashita, Shingo Tanaka, Hirokazu Okita, Mitsutoshi Nakada and Yutaka Hayashi
Although the right prefrontal region is regarded as a silent area, chronic deficits of the executive function, including working memory (WM), could occur after resection of a right prefrontal glioma. This may be overlooked by postoperative standard examinations, and the disabilities could affect the patient's professional life. The right prefrontal region is a part of the frontoparietal network and is subserved by the superior longitudinal fasciculus (SLF); however, the role of the SLF in spatial WM is unclear. This study investigated a persistent spatial WM deficit in patients who underwent right prefrontal glioma resection, and evaluated the relationship between the spatial WM deficit and the SLF.
Spatial WM was examined in 24 patients who underwent prefrontal glioma resection (right, n = 14; left, n = 10) and in 14 healthy volunteers using a spatial 2-back task during the long-term postoperative period. The neural correlates of spatial WM were evaluated using lesion mapping and voxel-based lesion-symptom mapping. In addition, the spatial 2-back task was performed during surgery under direct subcortical electrical stimulation in 2 patients with right prefrontal gliomas.
Patients with a right prefrontal lesion had a significant chronic spatial WM deficit. Voxel-based lesion-symptom mapping analysis revealed a significant correlation between spatial WM deficit and the region that overlapped the first and second segments of the SLF (SLF I and SLF II). Two patients underwent awake surgery and had difficulties providing the correct responses in the spatial 2-back task with direct subcortical electrical stimulation on the SLF I, which was preserved and confirmed by postoperative diffusion tensor imaging tractography. These patients exhibited no spatial WM deficits during the postoperative immediate and long-term periods.
Spatial WM deficits may persist in patients who undergo resection of the tumor located in the right prefrontal brain parenchyma. Injury to the dorsal frontoparietal subcortical white matter pathway, i.e., the SLF I or SLF I and II, could play a causal role in this chronic deficit. A persistent spatial WM deficit, without motor and language deficits, could affect the professional life of the patient. In such cases, awake surgery would be useful to detect the spatial WM network with appropriate task during tumor exploration.
Salvatore Zavarella, Mitsutoshi Nakada, Shawn Belverud, Salvatore J. Coniglio, Amanda Chan, Mark A. Mittler, Steven J. Schneider and Marc Symons
Medulloblastomas are the most common malignant brain tumors in children. These tumors are highly invasive, and patients harboring these lesions are frequently diagnosed with distant spread. In this study, the authors investigated the role of Rac1, a member of the Rho family of small guanosine triphosphatases, in medulloblastoma invasion.
Three established medulloblastoma cell lines were used: DAOY, UW-228, and ONS-76. Specific depletion of Rac1 protein was accomplished by transient transfection of small interfering RNA. Cell invasion through extracellular matrix (Matrigel) was quantified using a transwell migration assay. Mitogen activated protein kinase activation was determined using phospho-MAP kinase–specific antibodies, and inhibition of MAP kinase pathways was achieved by specific small molecule inhibitors. Localization of Rac1 and its expression levels were determined by immunohistochemical analysis using a Rac1-specific antibody, and Rac1 activation was qualitatively assessed by Rac1 plasma membrane association.
Small interfering RNA–mediated depletion of Rac1 strongly inhibited medulloblastoma cell invasion. Although depletion of Rac1 inhibited the proliferation of UW-228 cells, and of ONS-76 cells to a lesser extent, it stimulated the proliferation of DAOY cells. Depletion of Rac1 also inhibited the activation of the ERK and JNK MAP kinase pathways, and inhibition of either pathway diminished invasion and proliferation. Immunohistochemical analysis demonstrated that the Rac1 protein was overexpressed in all medulloblastoma tumors examined, and indicated that Rac1 was hyperactive in 6 of 25 tumors.
The authors' data show that Rac1 is necessary for the invasive behavior of medulloblastoma cells in vitro, and plays a variable role in medulloblastoma cell proliferation. In addition, these results indicate that Rac1 stimulates medulloblastoma invasion by activating the ERK and JNK pathways. The authors suggest that Rac1 and signaling elements controlled by this guanosine triphosphatase may serve as novel targets for therapeutic intervention in malignant medulloblastomas.
Mitsutoshi Nakada, Daisuke Kita, Kazuya Futami, Junkoh Yamashita, Noboru Fujimoto, Hiroshi Sato and Yasunori Okada
Object. Acquisition of invasive and metastatic potentials through proteinase expression is an essential event in tumor progression. Among proteinases, matrix metalloproteinases (MMPs) are thought to play a key role in tumor progression through the degradation of the extracellular matrix. In the present study, the authors examined the role of MMP-2 (gelatinase A) and membrane type 1 MMP (MT1-MMP), an activator of the zymogen of MMP-2, proMMP-2, together with tissue inhibitors of metalloproteinases (TIMP-1 and TIMP-2) in the invasion of astrocytic tumors in humans.
Methods. Analyses performed using sandwich enzyme immunoassays demonstrated that the production levels of proMMP-2 and TIMP-1, but not TIMP-2, are significantly higher in glioblastomas multiforme than in other grades of astrocytic tumors. Quantitative reverse transcription-polymerase chain reaction indicated that MT1-MMP is expressed predominantly in glioblastoma tissues, and its expression levels are significantly enhanced as tumor grade increases. In addition, the expression levels and proMMP-2 activation ratio were remarkably higher in glioblastomas associated with cerebrospinal fluid (CSF) dissemination than in those not associated with CSF dissemination. In contrast, an examination of TIMP-2 levels showed a reverse correlation. Like MT1-MMP, TIMP-1 and TIMP-2 were immunolocalized to neoplastic cells in glioblastoma samples. To study the roles of these molecules in the invasion of astrocytic tumors more fully, stable transfectants expressing the MT1-MMP gene were developed in a U251 human glioblastoma cell line. The MT1-MMP transfectants displayed prominent activation of proMMP-2 and invasive growth in three-dimensional collagen gel; however, mock transfectants and parental cells displayed noninvasive growth without the activation. The invasion and gelatinolytic activity of the transfectants were completely inhibited by addition of recombinant TIMP-2, but not recombinant TIMP-1.
Conclusions. These results indicate that MT1-MMP may contribute to tumor invasion and CSF dissemination of glioblastoma cells on the basis of an imbalance of TIMP-2.