SURGICAL THERAPIES

Neuro-Oncology, Nov 2011

INTRODUCTION: The role of aggressive surgical resection for glioblastoma multiforme (GBM) is controversial, with conflicting evidence published on the effect of extent of resection on outcome. We analyzed our patient population to determine if there was a correlation between greater extent of resection and overall survival or progression-free survival (PFS) in our neurosurgical oncology practice. METHODS: We conducted a retrospective cohort study between 2003 and 2010 of 109 patients treated with surgical resection for GBM at a single institution. Extent of resection was defined as gross-total resection ([GTR], no residual enhancement), near-total resection ([NTR], thin rim of enhancement in resection cavity only) or sub-total resection ([STR], residual nodular enhancement) based on immediate postoperative MRI findings. Variables associated with survival in a univariate analysis (proportional hazards regression analysis) were included in the multivariate Cox model if P < 0.10. Variables with probability values > 0.05 were then removed from the multivariate model in a stepwise fashion. RESULTS: Mean age and KPS score at time of diagnosis was 59.7 ± 15.9 years and 63.8 ± 13.8, respectively. GTR, NTR, and STR was achieved in 26 (24%), 37 (34%), and 46 (42%) cases, respectively. In univariate analysis, GTR was associated with a survival benefit vs STR (P = 0.02) but not NTR (P = 0.19). Median survival for GTR, NTR, and STR was 16.7, 15.8, and 10.5 months, respectively. Adjusting for factors associated with survival (age, radiation therapy, and temozolomide), GTR vs STR showed a trend toward improved survival (P = 0.07). In univariate analysis, GTR was also associated with prolonged PFS vs STR (P = 0.04) but not NTR (P = 0.20). CONCLUSIONS: In our population of 109 patients with GBM, GTR was associated with a survival benefit. When technically feasible, we believe that GTR should be attempted for the optimal treatment of this disease.

A PDF file should load here. If you do not see its contents the file may be temporarily unavailable at the journal website or you do not have a PDF plug-in installed and enabled in your browser.

Alternatively, you can download the file locally and open with any standalone PDF reader:

https://neuro-oncology.oxfordjournals.org/content/13/suppl_3/iii154.full.pdf

SURGICAL THERAPIES

0 Stanford University , Palo Alto, CA 1 ST-04. IMPACT OF INTRAOPERATIVE STIMULATION BRAIN MAPPING ON GLIOMA SURGERY OUTCOME: A META-ANALYSIS Philip C. De Witt Hamer 2 Vanderbilt University , Nashville, TN 3 ST-03. ENDOSCOPIC TRANSPHENOIDAL PITUITARY ADENOMA RESECTION GUIDED BY INTRAOPERATIVE HIGH-FIELD MAGNETIC RESONANCE IMAGING Tao Zhou , Xianghui Meng, Bainan Xu, Shaobo Wei, and Xiaolei Chen; PLA General Hospital , Beijing, China 4 Johns Hopkins University , Baltimore, MD 5 ST-01. GROSS-TOTAL RESECTION IS ASSOCIATED WITH A SURVIVAL BENEFIT IN GLIOBLASTOMA MULTIFORME Lola B. Chambless , Scott L. Parker, Laila Hassam-Malani, Matthew J. McGirt, and Reid C. Thompson; Vanderbilt University , Nashville, TN 6 Brain Tumor Research Center, Department of Neurological Surgery, University of California San Francisco , San Francisco, CA 7 Hopital Guy de Chauliac, CHU Montpellier, Department of Neurosurgery , Montpellier, France 8 ST-06. DEVELOPMENT AND VALIDATION OF A PRE-OPERATIVE PROGNOSTIC SCORING SYSTEM FOR PATIENTS WITH GLIOBLASTOMA Kaisorn Chaichana 9 ST-05. NEUROSURGICAL TREATMENT AS FIRST CHOICE FOR EXTRA-AXIAL EWING'S SARCOMA WITHOUT METASTASES: A CASE REPORT AND LITERATURE REVIEW Jose Daniel S. Ruiz Gonzalez , Olaya Vazquez Alberto, and Herrera Mora Patricia; Hospital Medica Sur , Mexico 10 Yotsuya Medical Cube , Tokyo, Japan 11 Nippon Medical School, Chibahokuso Hospital , Chiba, Japan 12 Nippon Medical School , Tokyo, Japan 13 ST-13. PROTECTION OF MOTOR PATHWAY ADJACENT TO INTRAAXIAL TUMORS BY THE GUIDANCE OF INTRAOPERATIVELY PLACED MARKERS Fumio Yamaguchi 14 Kasugai Rehabilitation Hospital , Yamanashi, Japan 15 Seoul National University Hospital , Seoul, Republic of Korea 16 Seoul National University, Bundang Hospital , Seongnam-si, Republic of Korea 17 ST-19. VISUAL OUTCOME AFTER SURGICAL TREATMENT FOR CRANIOPHARYNGIOMAS Young-Hoon Kim 18 Royal College of Surgeons in Ireland , Dublin, Ireland 19 ST-25. ASSESSING PERFORMANCE IN BRAIN TUMOR DEBULKING USING A NOVEL VIRTUAL REALITY SIMULATOR Nicholas Gelinas-Phaneuf 20 National Research Council of Canada, Industrial Materials Institute , Boucherville, QC, Canada 21 Brain Tumour Research Center, Montreal Neurological Institute and Hospital, McGill University , Montreal, QC, Canada 22 ST-22. COMBINED FLUORESCENCE AND REFLECTANCE SPECTROSCOPY FOR IN VIVO QUANTIFICATION OF CANCER BIOMARKERS IN LOW- AND HIGH-GRADE GLIOMA SURGERY Pablo A. Valdes 23 Dartmouth Hitchcock Medical Center , Lebanon, NH 24 Georgetown University , Washington, DC 25 Centre Hospitalier Universitaire Vaudois , Lausanne, Switzerland 26 Radiation Therapy Oncology Group , Philadelphia, PA 27 The University of Texas MD Anderson Cancer Center , Houston, TX 28 ST-34. EXTENT OF RESECTION IN GLIOBLASTOMA - VARIATION BETWEEN MOLECULAR RISK GROUPS IN RTOG-0525 Daniel Cahill 29 Cleveland Clinic , Cleveland, OH 30 Northwestern University , Chicago, IL 31 Emory University School of Medicine , Atlanta, GA 32 ST-41. THE RELATIONSHIP BETWEEN EXTENT OF RESECTION AND OUTCOME IN NEWLY DIAGNOSED GLIOBLASTOMA: RESULTS FROM RTOG 0525 Michael A. Vogelbaum 33 Cleveland Clinic Foundation , Cleveland, OH 34 ST-40. CONFOCAL MICROSCOPIC FLUORESCENCE IMAGING OF GLIOMAS Sameer A. Sheth 35 Northwestern University School of Medicine , Chicago, IL BACKGROUND: Brain surgery of infiltrative gliomas aims to balance tumor removal with preservation of functional integrity. The usefulness of intraoperative stimulation mapping (ISM) has not been addressed in randomized trials. This study addresses the impact of ISM on glioma surgery outcome based on a meta-analysis of observational studies. METHODS: A systematic search retrieved 90 publications published between 1990 and 2010 with 8,091 adult patients who had resective surgery for supratentorial infiltrative glioma, either with or without ISM. Quality criteria consisted of details of postoperative neurological examination and follow-up timing. New postoperative neurological deficits were categorized based on timing and severity. The summary event rates of neurological deficits were determined by meta-analysis with a Bayesian random effects model, as well as percentages of gross total resections and eloquent locations. Meta-regression analysis explored sources of heterogeneity between studies. RESULTS: Late severe neurological deficits were observed in 3.4% (2.3-4.8%) of 3,230 patients after resections with ISM, and in 8.2% (5.7-11.4%) of 1,731 patients after resections without ISM. The percentage of radiologically confirmed gross total resections was 75% (66-82%) with ISM and was 58% (48-69%) without ISM. Eloquent locations were involved in all resections with ISM and in only a subset of resections without ISM. Relevant sources of heterogeneity between studies were ISM, North American or European continent, and academic study setting. CONCLUSION: Glioma resections utilizing ISM are associated with less late severe neurological deficits, without compromising the extent of resection and while involving eloquent locations more frequently. - INTRODUCTION: The role of aggressive surgical resection for glioblastoma multiforme (GBM) is controversial, with conflicting evidence published on the effect of extent of resection on outcome. We analyzed our patient population to determine if there was a correlation between greater extent of resection and overall survival or progression-free survival (PFS) in our neurosurgical oncology practice. METHODS: We conducted a retrospective cohort study between 2003 and 2010 of 109 patients treated with surgical resection for GBM at a single institution. Extent of resection was defined as gross-total resection ([GTR], no residual enhancement), near-total resection ([NTR], thin rim of enhancement in resection cavity only) or sub-total resection ([STR], residual nodular enhancement) based on immediate postoperative MRI findings. Variables associated with survival in a univariate analysis (proportional hazards regression analysis) were included in the multivariate Cox model if P , 0.10. Variables with probability values . 0.05 were then removed from the multivariate model in a stepwise fashion. RESULTS: Mean age and KPS score at time of diagnosis was 59.7 + 15.9 years and 63.8 + 13.8, respectively. GTR, NTR, and STR was achieved in 26 (24%), 37 (34%), and 46 (42%) cases, respectively. In univariate analysis, GTR was associated with a survival benefit vs STR (P 0.02) but not NTR (P 0.19). Median survival for GTR, NTR, and STR was 16.7, 15.8, and 10.5 months, respectively. Adjusting for factors associated with survival (age, radiation therapy, and temozolomide), GTR vs STR showed a trend toward improved survival (P 0.07). In univariate analysis, GTR was also associated with prolonged PFS vs STR (P 0.04) but not NTR (P 0.20). CONCLUSIONS: In our population of 109 patients with GBM, GTR was associated with a survival benefit. When technically feasible, we believe that GTR should be attempted for the optimal treatment of this disease. To evaluate the effect of intraoperative high-field (1.5 Tesla) magnetic resonance imaging (MRI on the results of endoscopic transsphenoidal surgery of pituitary adenomas, 51 patients (tumor size, 32.1 + 8.5 mm; min-max, 20-52.7 mm) with pituitary adenoma were investigated by intraoperative high-field MRI during endoscopic transsphenoidal surgery. If intraoperative imaging depicted an accessible tumor remnant, resection was continued. A navigation system based on intraoperative magnetic resonance (MR) was used in 14 cases. An intraoperative MR scan was performed in each operation from 1 time to 5 times (range, 2.2 + 0.65). A neuro-navigation system was used in 14 operations and the data renewed in 8 cases by the information from intraoperative MR. In 19 cases of 51 patients, intraoperative MRI revealed residual lesions and resulted in 17 cases of further treatment, and eventually 11 tumors were totally removed and 6 tumors were further removed. There were no intraoperative MR-related safety issues or accidents recorded in this study. High-field-strength MRI provided highquality intraoperative MR images of tumor resection that allowed intraoperative modification of the surgical strategy. Combined with the navigation system based on it, intraoperative MR is helpful in maximizing the resection of the pituitary adenoma and improving the safety of endoscopic operations. We present the case of a 1-year-old male patient who developed increased volume at the left occipital region accompanied by pain. He was evaluated by means of cranial magnetic resonance simple and a contrast skull CT scan, cranial ultrasound, whole body scintigraphy, and whole body positron emission tomography CT. It was concluded that he had an extracranial single lesion compatible with juvenile hemangioma versus rhabdomyosarcoma. Radical excision, including free edges of the lesion, was performed. The transoperative study report showed a tumor of small round cells, basophilic, compatible with rhabdomyosarcoma. The total lesion, with a measurement of 6x5x4 cm, was sent to definitive study. During the immediate postoperative period, he had good outcome and was discharged at 3 days. The pathology report showed Ewings sarcoma-TNE primitive. Immunohistochemistry markers were positive for CD99 very intense and diffusely positive MYO D1, actin, desmine, and CD45 negative. He subsequently received 7 cycles of chemotherapy keeping IK 100. We propose radical surgery as initial treatment in selected patients with adequate KPS indexes, accessible lesions in non-eloquent areas, and absence of extracranial lesions in other body parts. This conditional on local disease control as well as preserve the highest quality of life. Glioblastoma multiforme (GBM) is the most common and aggressive type of primary brain tumor in adults, with an average survival of 12 months. Individual survival is heterogeneous, and the ability to predict short- and long-term survivors is limited. Therefore, the aims of this study were to ascertain pre-operative risk factors associated with survival, develop a preoperative prognostic grading system, and evaluate the utility of this grading system to predict survival for patients with GBM. Adult patient who underwent surgery for an intracranial primary (de novo) GBM at Johns Hopkins Hospital (JHH) between 1997 and 2007 were retrospectively reviewed. Multivariate proportional hazards regression analysis was used to identify pre-operative factors associated with survival, after controlling for extent of resection and adjuvant therapies. The identified associations with survival were then used to develop a grading system based on pre-operative variables. The ability of this scoring system to predict survival was applied to patients from JHH (operated on between 2008 and 2009), Stanford, and Vanderbilt Universities. Survival as a function of time was plotted using the Kaplan-Meier method, and survival rates were compared using log-rank analysis. The pre-operative factors, independent of extent of resection and adjuvant therapies, found to be negatively associated with survival were: age . 60 years (P , 0.0001), KPS 80 (P , 0.0001), motor deficit (P 0.02), language deficit (P 0.001), and periventricular tumor location (P 0.04). Patients possessing 0-1, 2, 3, or 4-5 of these variables were assigned a pre-operative grade of 1, 2, 3, or 4, respectively. Patients with a pre-operative grade of 1, 2, 3, or 4 had a median survival of 17.9, 12.3, 10, or 7.5 months, respectively (P , 0.05). This grading system, based only on pre-operative variables, may provide patients and physicians with prognostic information that may guide medical and surgical therapy before any intervention is pursued. ST-07. AN EXTENT OF RESECTION THRESHOLD FOR RECURRENT GLIOBLASTOMAS Mark E. Oppenlander, Andrew Wolf, Randall Porter, Peter Nakaji, Kris A. Smith, Robert F. Spetzler, and Nader Sanai; Barrow Neurological Institute, Phoenix, AZ For newly-diagnosed glioblastoma patients, mounting evidence suggests that greater extent of resection (EOR) corresponds to better overall survival. In combination with radiation and chemotherapy, initial resection beyond a 78% EOR is associated with improved overall survival. For recurrent glioblastoma patients, however, the value of a second resection at the time of recurrence remains uncertain. Specifically, what proportion of contrast-enhancing recurrent glioblastoma tissue must be removed to improve overall survival and what is the survival benefit beyond this threshold? We identified 100 consecutive, recurrent supratentorial glioblastoma patients treated at the Barrow Neurological Institute from 2005-2010. All patients initially presented with a de novo glioblastoma and, following their initial resection, received standard temozolomide and fractionated radiotherapy. Clinical and radiographic data were collected retrospectively, including volumetric tumor analysis. Patients had a median age of 55.4 years and presented with a median KPS of 80. Mean clinical follow-up was 25.6 months and no patient was unaccounted. At the time of recurrence, the median preoperative tumor volume was 24.8 cm3. Following re-resection, median postoperative tumor volume was 2.5 cm3, equating to an 89% EOR. The median overall survival was 20.7 months, with a median progression-free survival following re-resection of 5.5 months. Using Cox proportional hazards analysis, age, KPS, and EOR were predictive of survival following repeat resection (P , 0.001). A significant survival advantage was seen with as little as a 79% EOR. Recursive partitioning analysis validated these findings and provided additional risk stratification parameters related to age, extent of resection, and tumor burden. For recurrent glioblastomas, an improvement in overall survival can be attained beyond a 79% EOR. Interestingly, this lower limit closely approximates that reported for newly diagnosed glioblastomas, suggesting that, for a subset of patients, the value of microsurgical resection does not diminish despite biological progression. ST-08. SUPRATENTORIAL PRIMITIVE NEUROECTODERMAL TUMORS (stPNETs) IN ADULTS Jeong Hoon Kim; Asan Medical Center, Seoul, Republic of Korea Supratentorial primitive neuroectodermal tumors (stPNETs) are rare, especially in adults, but have a grim prognosis, frequently taking an aggressive course with local relapse and metastatic spread. Furthermore, there is no standard treatment protocol. We report the experiences of a singleinstitution therapeutic trial. From 1996 to 2010, 9 consecutive patients were enrolled. In early periods, adjuvant gamma knife radiosurgery was carried out for recurrent or residual tumor. Since 2007, adjuvant chemotherapy had been applied on recurrent or metastatic cases, and extended to all patients from 2009. Median age was 39 years (range, 19-62); 6 were men, 3 were women. Sites of disease were frontal in 4, parietal in 3, fronto-parietal in 1, and pineal in 1, and the mean diameter of mass of these tumors was 5.3 cm. The characteristic magnetic resonance image finding was a large, well-demarcated lobulating mass with intratumoral cyst, necrosis, and/or hemorrhage. Calcification was seen in 2 of 7 patients who underwent a computed tomography scan. All patients underwent total tumor resection except one, who underwent endoscopic biopsy because of pineal location. Two patients underwent reoperations because of local recurrence. Adjuvant whole brain radiotherapy was used in 2 patients, and craniospinal irradiation plus local boost was used in 5 patients. The other 2 patients underwent local boost. Five patients received additional chemotherapy. Mean survival after diagnosis was 43 months (range, 6-175 months). The patients having intratumoral calcifications are all alive, and 3 of 4 who showed a Ki-67 labeling index greater than 30% died. In clinical features and prognosis, adult stPNETs seemed similar to those of children. Extensive surgical resection following adjuvant radiotherapy and chemotherapy can be an essential treatment option. Intratumoral calcifications and the Ki-67 labeling index might be prognostic factors, however, it should be considered that the sample size is too small and not all patients were evaluated. ST-10. ENDOSCOPIC SURGERY FOR TUBERCULUM SELLAE MENINGIOMAS: A SYSTEMIC REVIEW AND META-ANALYSIS Aaron J. Clark, Arman Jahangiri, Michael E. Sughrue, Michael W. McDermott, and Manish K. Aghi; University of California, San Francisco, San Francisco, CA Recent reports of surgical resection of tuberculum sellae meningiomas through an extended endoscopic endonasal (EEE) approach have provided an alternative to transcranial approaches in selected cases. However, these published reports have been limited by small sample sizes from single institutions. We performed a systemic review of the literature and analyzed pooled data for descriptive statistics on short-term morbidity and outcomes to gain insights into the potential limitations and benefits of the EEE approach for tuberculum sellae meningiomas. Five studies (60 patients) met our inclusion criteria. The pooled rate of gross total resection (Simpson grade 1-2) was 25 of 31 documented cases (81%). There were 27 perioperative complications (45%), including 20 cerebrospinal fluid (CSF) leaks (33%), although the range of CSF leak rates was wide (8-40%); 3 patients were diagnosed with postoperative diabetes insipidus (5%), and there was 1 mortality (2%). Age and tumor size did not predict extent of resection or morbidity (P 0.6-1.0). A pooled analysis of transcranial results reported during a similar period yielded 10 studies (390 patients). There were no differences in the rate of gross total resection or perioperative complications between the 2 groups. The EEE group was associated with a significantly higher rate of visual improvement postoperatively (0.95; 95% confidence interval [CI], 0.91-1.0) compared to the transcranial group (0.60; 95% CI, 0.55-0.65; P , 0.05). A systemic review of the small series of EEE approaches for tuberculum sellae meningiomas published to date revealed similar extent of resection and morbidity but increased postoperative visual improvement compared to transcranial approaches during a similar period. Long-term follow-up will be needed to define recurrence rates of the EEE approach as compared to transcranial approaches. The cautious use of EEE approaches for the removal of smaller tuberculum sellae meningiomas after formal endoscopic training may be warranted. ST-11. PALLIATIVE ENDOSCOPIC THIRD VENTRICULOSTOMY VERSUS VENTRICULOPERITONEAL SHUNTING IN THE TREATMENT OF OBSTRUCTIVE HYDROCEPHALUS FROM CEREBRAL METASTASIS Clark Chen1, Ekkehard Kasper2, and Peter Warnke3; 1Dana-Farber Cancer Institute, Boston, MA; 2Beth Israel Deaconess Medical Center, Boston, MA; 3Division of Neurosurgery, University of Chicago, Chicago, IL BACKGROUND: The feasibility of endoscopic third ventriculostomy (ETV) in the treatment of obstructive hydrocephalus related to cerebral metastasis was previously demonstrated. However, the relative indication of ETV vs ventriculoperitoneal shunt (VPS) placement in this setting remains somewhat arbitrary in neurosurgical practice. METHODS: We retrospectively reviewed the efficacy of these modalities in our single-institution experience. In our institution, ETV was only performed in 1) patients who are not candidates for surgical resection, 2) patients with favorable ventricular anatomy (with third ventricle dilated beyond 3 mm and without obstructive ventricular mass), and 3) patients without i) cerebrospinal fluid (CSF) protein level twice normal threshold (when CSF available), ii) a history of ventricular hemorrhage, iii) dilated subarachnoid space, and iv) leptomeningeal disease. Patients failing these criteria underwent either resection or VPS placement. RESULTS: Between 2005 and 2010, 52 of 1266 patients (4%) with brain metastasis suffered obstructive hydrocephalus not amenable to surgical resection. Sixteen patients (31%) underwent ETV and 36 (69%) underwent VPS placement. All patients underwent either radiosurgery or whole brain radiation therapy after ETV or VPS. For both ETV and VPS patients, approximately 70% experienced improvement in the presenting symptoms. Thirteen percent (n 2) of the ETV patients required a second surgery for either wound revision or ventricular drain placement for persistence of symptoms, whereas 19% (n 7) of the VPS patients required revision (P 0.94). CONCLUSION: In this retrospective series, the efficacy and complication rates of ETV and VPS are comparable when strict criteria were applied for ETV patient selection. ST-12. A NEW SCALE FOR THE DECISION OF SURGICAL MANAGEMENT IN RECURRENT GLIOBLASTOMAS Chul-Kee Park, Se-Hoon Lee, Sang Woo Song, Jin Wook Kim, and Tae Min Kim; Seoul National University Hospital, Seoul, Republic of Korea OBJECTIVE: To formulate the benefit of surgical management in recurrent glioblastoma, we analyzed a series of recurrent glioblastoma patients who had undergone surgery and devised a scale to predict their survival. MATERIALS AND METHODS: Clinical and radiographical data of 60 consecutive and histologically confirmed recurrent glioblastoma or gliosarcoma patients after surgical management were evaluated. Kaplan-Meier survival analysis and Cox proportional hazards regression modeling were used for the prognostic variables and their effect on overall survival (OS). RESULTS: Among the tested prognostic variables, performance status (P 0.004) and ependymal involvement (P 0.062) were selected after multivariate analysis for the ingredients of the new prognostic scale. A 3-tier scale (range, 0 to 2 points), composed of additive scores of performance status (0 for KPS 70 and 1 for KPS , 70) and ependymal involvement (0 for no enhancement and 1 for enhancement of ependymal lining in MRI) significantly distinguished prognostic groups with good (0 points; OS, 15.0 months), intermediate (1 point; OS, 10.0 months), and poor (2 points; OS, 4.0 months). While there were no significant differences in survival in the good or poor prognostic groups whether adjuvant treatment was added after surgery, the intermediate prognostic group showed a significant survival gain when adjuvant treatment was added after surgery, rather than just observation (OS, 18.0 vs 8.0 months; P 0.0388). CONCLUSIONS: We developed a new, simple, practical scale for the decision of surgical management in recurrent glioblastoma patients, and this scale provided useful diagnosis of prognostic groups and guidelines of management options. INTRODUCTION: Intraoperative recognition of motor pathways is quite difficult without any landmark. It is ideal to know the 3-dimentional location of motor pathways during surgery in advance of tumor resection. A novel method has been investigated and evaluated to determine whether this technique is precise and reliable. METHODS: Subcortical mappings were done in 25 glioma patients using a bipolar electrode registered as a neuronavigator probe. A twenty-five mm- plastic tube was installed onto each needle electrode that has 30 mm length. The electrode, along with plastic tube, was inserted into the cerebral white matter with the reference of tractography information in a neuro-navigator. The insertion with continuous electrical stimulation was stopped when muscle-motor-evoked potentials were detected. Then, the electrode was removed, leaving 2 plastic tubes in place. Further resection was done until the tube tips were exposed. Motor functions and KPSs were evaluated as well as presence of residual tumor by MRI. RESULTS: This procedure leads successful detections of the pyramidal tract in all cases and could notify the safe resectable margin that is 5 mm apart from motor tract. There were discrepancies between tractography data and actual neurophysiological localization of pyramidal tracts because of brain shift. Postoperative MRIs revealed gross total resection close to pyramidal tracts. None of the patients presented postoperative neurological deterioration. KPSs were maintained or increased than the preoperative state. CONCLUSIONS: Electrophysiologically placed tube markers enabled the recognition of 3-dimensional pathways of motor tracts even under intraoperative brain shift. This technique is easy and reliable in glioma surgery to preserve motor function. ST-14. CEREBRAL LYMPHOMA TO BIOPSY OR NOT TO BIOPSY? Erlick A. Pereira, James Livermore, Olaf Ansorge, and Stana Bojanic; University of Oxford, Oxford, United Kingdom INTRODUCTION: Cerebral lymphoma can be diagnosed by vitreous or cerebrospinal fluid sampling, but brain tissue histopathology is desirable to guide medical treatment. The literature suggests high hemorrhage rates with cerebral lymphoma biopsy (14%). METHODS: Retrospective review of prospectively collected data from a single-center, case series of 47 consecutive adults diagnosed with histologically confirmed cerebral lymphoma over 4 years, from December 2006 to January 2011, compared to a 96-patient cohort receiving cerebral biopsies for other pathologies over 1 year, from August 2009 to July 2010. RESULTS: 37 of 47 lymphoma patients (79%) received biopsy (29 by framed stereotaxy and 8 frameless). The male:female ratio in this group was 19:18 and the mean age at surgery was 60 years (range, 29-80). The remaining 10 patients received craniotomy and surgical resection of lymphoma. There was no preference for laterality of tumor, but 27% of tumors were thalamic and 8% were brainstem. Five symptomatic hemorrhages happened in the lymphoma cohort, 4 of 13 in those receiving thalamic or brainstem biopsies (11%) and 1 (3%) in a patient receiving craniotomy for right frontal lymphoma resection. There was 1 death from thalamic hemorrhage. There were no significant differences in bleed rates between lymphoma biopsies overall (4/37) and other biopsies (5/91). Brainstem and thalamic lymphoma biopsies were higher risk (chi- square test, P , 0.05; odds ratio, 4.4; 95% confidence interval, 1.3-26) than other biopsies overall. CONCLUSIONS: Approximately 12 cerebral lymphomas present per year to our center, which serves 3 million people annually. Histopathology demonstrates vascular endothelial compromise intuitively increasing hemorrhage risk. Thalamic and brainstem lymphoma biopsies are high risk with up to 30% hemorrhage and 7% mortality in this series, which is comparable to the limited published literature. A diagnosis enabling medical treatment should be made from vitreous biopsy and cerebrospinal fluid sampling where possible, and perhaps even on clinical context and imaging alone where only palliative therapy is being considered. ST-15. HIGH-FIELD INTRAOPERATIVE MRI WITH A MOVABLE MAGNET GUIDED TRANSSPHENOIDAL SURGERY FOR PITUITARY ADENOMA Xianghui Meng, Bainan Xu, Xiaolei Chen, Shaobo Wei, Tao Zhou, Huaiyu Tong, Xinguang Yu, Dingbiao Zhou, Yuanzheng Hou, Zhe Zhou, and Jiashu Zhang; Chinese PLA General Hospital, Beijing, China Our objective was to review the preliminary clinical experience with a dual-room intraoperative high-field-strength magnetic resonance (MR) imaging suite with a movable magnet in the pituitary adenoma operation with the transsphenoidal approach. From February 2009 to April 2010, 62 patients (age range, 17-69 years) with pituitary adenoma were operated on with the transsphenoidal approach and examined intraoperatively with a movable 1.5-Tesla MR magnet. There were 57 cases of hormonally inactive pituitary macroadenomas, 3 cases of prolactin-secreting pituitary macroadenomas, and 2 cases of growth-hormone-secreting pituitary macroadenomas. Median tumor size was 1.6-7.27 cm (1-2 cm, 2 cases; 2-4 cm, 47 cases; 4 cm, 13 cases). In some cases, a microscope and/or endoscope, in combination with a ceiling-mounted navigation system, was also used to assist in performance of the operation. Sixty-one resections with the transnasal transsphenoidal approach were performed, and the transoral transsphenoidal approach was performed for 1 patient. In 28 cases of 62 patients, intraoperative MR imaging had revealed tumor residuals. In 4 invasive cavernous sinus cases, no further resection of the tumor was performed because of internal carotid artery encasement; the remaining 24 cases underwent further resection, with total removal in 20. This increased the rate of complete tumor removal from 55% (34/62) to 87% (54/62). There were also no intraoperative MR-related safety issues or accidents recorded in this study. The dualroom intraoperative high-field-strength MR imaging suite with a movable magnet provided high-quality intraoperative MR images and valuable information about tumor resection that allowed intraoperative modification of the surgical strategy. It could be very helpful in maximizing resection of the pituitary adenoma and minimizing injury to neurological function. ST-16. ESTHESIONEUROBLASTOMA RESECTION VIA THE EXPANDED ENDONASAL ENDOSCOPIC APPROACH Andrew J. Fabiano, Nestor Rigual, Stephan Munich, and Robert A. Fenstermaker; Roswell Park Cancer Institute, Buffalo, NY INTRODUCTION: Esthesioneuroblastoma (ENB) is a rare, malignant tumor involving the nasal cavity and anterior skull base. Treatment varies and may include surgical resection, radiotherapy, and chemotherapy. The expanded endonasal endoscopic approach provides a wide exposure to the anterior skull base. This series examines 3 consecutive cases in which ENB resection was attempted via the expanded endonasal endoscopic approach. METHODS: Case series. RESULTS: Three consecutive patients were included; the average age was 54 years and 2 patients were female. In 2 cases, the tumor was completely resected using the expanded endonasal endoscopic approach exclusively. In both of these cases, numerous marginal biopsies were taken that were all negative. In the remaining case, marginal dural biopsies were positive, prompting conversion to a bifrontal craniotomy. Direct inspection revealed a residual tumor nodule attached to the falx cerebri adjacent to the crista galli. CONCLUSIONS: The expanded endonasal endoscopic approach can be used to treat selected cases of ENBs; however, positive marginal biopsies should alert the surgeon to the possible need to convert from the endonasal route to a synchronous craniotomy. Kadish C tumors in particular may be difficult to resect completely by way of the expanded endonasal approach. ST-17. LOW-GRADE INSULAR GLIOMA RESECTION WITH 1.5T INTRA-OPERATIVE MRI: PRELIMINARY RESULTS OF A PROSPECTIVE RANDOMIZED TRIAL Xiaolei Chen, Xianghui Meng, Jiashu Zhang, Fei Wang, Yan Zhao, and Bai-nan Xu; Chinese PLA General Hospital, Beijing, China For low-grade gliomas (LGGs), more aggressive resection and higher extent of resection (EOR) may result in longer survival of the patient. However, aggressive resection of insular LGGs remains challenging due to the tumors relationship with complex surrounding functional structures. To provide highclass evidence regarding the use of 1.5-T intra-operative MRI (iMRI) for the resection of insular LGGs, we conducted a prospective, randomized, controlled trial. Adult patients with non-contrast-enhancing insular lesions suspicious of LGGs who were scheduled to undergo radiologically complete tumor resection are eligible to enter this trial. With their informed consent, patients are randomized to undergo either iMRI-guided or conventional micro-neurosurgical tumor resection. For the iMRI group, pre- and intra-operative (before wound closure) 1.5-T MRI images are obtained to perform volumetric analyses to assess the EOR. For the control group, both pre- and post-operative (within 48 hours after surgery) 1.5-T MR images are acquired to assess the EOR. After the inclusion of 81 patients (iMRI group, 40 cases; control group, 41 cases), we performed a preliminary analysis. There were no significant differences in patient age (P 0.32) or preoperative tumor sizes (P 0.47) between the 2 treatment groups. We observed a significantly higher rate of complete tumor resections in the iMRI group compared to the control group (P , 0.05). EOR was significantly higher, and post-operative tumor volumes were significantly lower in the iMRI group than in the control group (P , 0.05). On the other hand, post-operative long-term (. 3 months) morbidity is significantly lower in the iMRI group than in the control group (P , 0.05).The use of iMRI appears to be associated with a higher rate of gross total as well as sub-total tumor resections compared to conventional microneurosurgery. Furthermore, aggressive resection of insular LGGs with iMRI did not increase the long-term morbidity. ST-18. PREDICTING LONG TERM REMISSION BY MEASURING IMMEDIATE POSTOPERATIVE GROWTH HORMONE LEVELS AND ORAL GLUCOSE TOLERANCE TEST IN ACROMEGALY Eui Hyun Kim, Min Chul Oh, Eun Jig Lee, and Sun Ho Kim; Yonsei University College of Medicine, Seoul, Republic of Korea INTRODUCTION: The suppression of the growth hormone (GH) in an oral glucose tolerance test (OGTT) has been accepted as the most reliable parameter for determining remission of acromegaly. We evaluated the role of immediate postoperative GH level and 1-week postoperative OGTT as early predictive tools of long-term surgical remission. METHODS: One hundred ninety-four acromegalic patients who received a transsphenoidal tumor resection and had at least 3 postoperative OGTTs were followed up for longer than 1.5 years (3.80 + .17 years). GH was measured 2, 6, 12, 18, 24, 48, and 72 hours postoperatively and an OGTT was performed 1 week after surgery, every 6 months for the first 3 years, and annually thereafter. RESULTS: One hundred seventy-seven patients underwent gross total resection, and long-term remission was achieved in 153. The GH level at 24 hours after surgery showed the highest predictive power for long-term remission. Long-term remission was maintained in 125 of 127 patients (98.4%) who had nadir GH levels , 1.0 mg/liter in an early postoperative OGTT. However, when nadir GH levels were higher than 1.0 mg/liter in an early postoperative OGTT, long-term remission was observed in 28 patients (28 of 67, 41.8%) in a delayed fashion. One-week postoperative OGTT had a sensitivity of 81.7% and a specificity of 95.1% for predicting remission. CONCLUSION: Immediate postoperative GH level is a very good predictor of long-term outcome in acromegaly. One-week postoperative OGTT is also a good predictor with high specificity. These findings may provide critical information for the determination of adjuvant treatment after surgery. INTRODUCTION: Craniopharyngioma, a histologically benign intracranial tumor, often presents a clinically dismal course due to its high recurrence rate and close vicinity to critical neurovascular structures. The authors evaluated the longitudinal visual outcome and its prognostic factors of surgically treated craniopharyngiomas. METHODS: From 2004 to 2010, 86 consecutive patients who underwent surgical treatments for newly diagnosed craniopharyngiomas were enrolled at 3 institutions. There were 40 pediatric and 46 adult patients, and the median age was 29 years (range, 1-67 years). Forty-one patients (48%) showed preoperative visual disturbances. The unilateral pterional approach was used in 43 cases (50%), the interhemispheric approach in 29 (34%), and the transsphenoidal approach in 12 (14%). The median follow-up duration was 33 months (range, 6-88 months). RESULTS: Gross or near-total resections were performed in 54 cases (63%) and subtotal resections were performed in 32 cases (37%). At the last follow-up, 26 patients (31%) experienced improvements in visual acuity or field; 19 patients (22%) experienced aggravations. Several prognostic factors of visual outcome, including sex, age, location, subtype, involvement in third ventricle, hydrocephalus, surgical approach, pathology, extent of resection, and adjuvant interventions, were evaluated. Sub-total resection was the only significant risk factor for visual deterioration (P 0.022), and the retrochiasmatic type and adjuvant treatment were associative factors (P 0.080 and P 0.058, respectively). CONCLUSION: Seventy-eight percent of the patients with surgically treated craniopharyngiomas experienced a favorable visual outcome. Grossly total removal and no adjuvant intervention were the important factors in visual improvements. ST-20. DUAL-ROOM INTRA-OPERATIVE MAGNETIC RESONANCE IMAGING SUITE WITH A MOVABLE MAGNET: CLINICAL EXPERIENCE OF 405 CASES Xiaolei Chen, Xianghui Meng, Fei Wang, Yan Zhao, and Bai-nan Xu; Chinese PLA General Hospital, Beijing, China Intraoperative magnetic resonance imaging (iMRI) may become a useful tool for maximizing the extent of resection in glioma surgery. The authors describe a novel dual-room high-field iMRI suite with a movable magnet and analyze its efficacy based on the clinical experience of 405 cases. The iMRI suite consists of an operating room and an adjacent diagnostic room, which has a movable 1.5-T magnet. From February 2009 to June 2010, 405 patients (age range, 6-81 years; mean age, 42.0 years) were operated on in the suites operating room; during the same period, 1,500 diagnostic scans were performed in the diagnostic room. The imaging data from all the operation cases were collected, and the impact of iMRI on surgery was analyzed. Three hundred twenty-seven craniotomies, 65 transsphenoidal surgeries, 10 frameless biopsies, and 3 frameless ablations were performed. Among the 392 cases of lesion resection, in 120 cases (30.6%), iMRI revealed residual lesions and resulted in a change of the surgical strategy (eg, further resection of the lesion). Eventually, 360 lesions (88.9%) were totally removed. Postoperative follow-up periods range from 3-12 months (mean, 7.8 months). Long-term morbidity is 4.9% (20 cases). With the MR suite, high-quality images, together with functional data, could be obtained intraoperatively. The intraoperative 1.5-T MRI and functional neuro-navigation can be successfully integrated into the standard neurosurgical workflow. Intraoperative MRI can provide high-quality images and valuable information for intraoperative modification of the surgical strategy, while the dual-room setting can maximize the efficiency of the system. ST-21. THE IMPACT OF RE-OPERATION FOR RECURRENT GLIOBLASTOMA Dietmar Krex, Claudia Lindner, Tareq Juratli, Claudia Raue, and Gabriele Schackert; University Hospital Carl Gustav Carus Dresden, Dresden, Germany The first-line treatment of glioblastoma has been standardized in recent years. However, the treatment of recurrent glioblastoma is less well defined. It is a matter of debate if, when, how extensive, and in what way a recurrent tumor should be neurosurgically treated. Therefore, we retrospectively analyzed a well-characterized glioblastoma population, including one-third of patients who had repeated surgery for recurrent tumors. The entire population comprised 381 patients who had been treated for primary or recurrent glioblastoma between 2004 and 2010 in our department. One hundred nineteen patients had a second operation, 35 a third, and 4 a fourth. Two hundred forty-nine patients had no second operation, while 13 patients, who had no evidence of tumor after recurrent surgery according to histopathological diagnosis, were disregarded. The median age was 64.5, 60, and 68 years for the entire population, those who had a second operation, and those who did not, respectively. KPS score at admission was . 70 in all cases. Radiotherapy was given to 90%, 95%, and 87% and any kind of chemotherapy to 64%, 70%, and 59% of the entire population, those who had a second operation, and those who did not, respectively. Median overall survival was 13.5 months for the entire population, 18.5 months for those who had a second operation, and 9.7 months for those who did not (P , 0.001). Adjuvant chemotherapy was an independent co-factor for overall survival after multivariate analysis. MGMT methylation status and IDH mutation status were also determined. To conclude, in our series, patients who were operated on for recurrent glioblastoma had a median overall survival that was almost twice as long as those who were not. Our data confirm that close follow-up and aggressive treatment, including repeated surgery, is effective in prolonging overall survival for glioblastoma patients. Gliomas represent a heterogeneous group of brain tumors, with extent of resection shown as a significant factor influencing post-surgical recurrence and prognosis. Recently, use of 5-aminolevulinic acid (ALA)-induced protoporphyrin IX (PpIX) fluorescence-guided resection has shown promise for improving the degree of resection in high-grade gliomas. We have shown that absolute quantification of PpIX biomarker in tissue significantly improves tumor detection across a range of tumor histologies. Nevertheless, despite improved detection for tumor tissue with this biomarker, and as a result of the inter- and intra-tumor heterogeneity of gliomas and the multifaceted nature of neoplastic processes, individual biomarkers are not sufficient for achieving optimal tumor tissue diagnosis. Here, we hypothesized that use of multiple factors (ie, biomarkers) is necessary to maximize the diagnostic and predictive power of optical technologies for tumor tissue delineation. We used fluorescence and reflectance spectral signatures for in vivo quantification of multiple biomarkers during glioma surgery, with fluorescence contrast provided by exogenously-induced PpIX following administration of ALA. We performed light-transport modeling to quantify multiple biomarkers indicative of tumor biological processes intraoperatively, including the local concentration of PpIX and associated photoproducts, total hemoglobin concentration, oxygen saturation, and optical scattering parameters. We developed a diagnostic algorithm for intraoperative tissue delineation that accounts for the combined tumor-specific predictive capabilities of these quantitative biomarkers. Tumor tissue delineation achieved accuracies of up to 94% (specificity 94%, sensitivity 94%) across a range of glioma histologies beyond current state-of-the-art optical approaches, including state-of-the-art fluorescence image guidance. These results suggest the ability of a quantitative multiple biomarker approach to detect both tumor bulk and infiltrating glioma tissues, especially in lowgrade gliomas where the impact on patient prognosis and survival could be substantial. This multiple biomarker strategy opens the door to optical methods for surgical guidance that use in vivo quantification of wellestablished neoplastic processes. ST-23. LOCAL THERAPY IN RECURRENT GLIOBLASTOMA Dietmar Krex, Tareq Juratli, Claudia Lindner, Claudia Raue, and Gabriele Schackert; University Hospital Carl Gustav Carus Dresden, Dresden, Germany Carmustine wafers are the only available local therapy for glioma treatment and are frequently used for recurrent glioblastoma (GBM). The therapeutic effect has been shown for primary and recurrent GBM in phase II and III studies, respectively; however, those series have been performed before the current standard treatment regimen for GBM, including radiotherapy and concomitant and adjuvant chemotherapy. We retrospectively analyzed our patients who had standard treatment for the primary lesion and local therapy using Carmustine-wafer implants for recurrent tumors. Sixty patients with primary GBM who had been treated for recurrent GBM at our department between 2005 and 2010 by repeated surgery and Carmustine-wafer implantation were analyzed. There were 39 men (median age, 58.4; range, 39.8-75.8) and 21 women (median age, 57.7; range, 34.7-73.5). All patients had intended gross total resection of the primary tumor followed by radiotherapy up to 60 Gy and concomitant and adjuvant temozolomide-based chemotherapy. Median overall survival was 18.6 months (range, 7.9-57.9 months). Thirty-two percent of patients had an overall survival . 2 years. Median time-to-progression from first operation to implantation of Gliadel wafers was 7.9 months (range, 3.8-31.7 months). Median survival time from Carmustine-wafer implantation until death or censuring was 8.9 months (range, 1.8-48.4 months). The MGMT promoter was methylated in 35% of all patients. Cerebrospinal fluid leakage and wound infection occurred in 5.7% and 13.2% of cases, respectively. In conclusion, the use of Carmustine-wafer implants after resection of recurrent GBM constitutes an additional available component in an aggressive and multimodal therapy regimen, which, in our series, was associated with an increased number of patients with a 2-year survival. An elevated risk of wound infection should be kept in mind. ST-24. SPINAL CERVICAL INTRAMEDULLARY CAPILLARY HEMANGIOBLASTOMA: AN USUAL TUMOR IN AN UNUSUAL SITE Giuseppe Occhiogrosso, Pasquale Cascardi, Maria Blagia, and Antonio De Tommasi; Neurosurgery, Bari, Italy Capillary hemangioblastoma is a slowly growing, highly vascular tumor with uncertain histogenesis. It may arise either as sporadic lesions (68%) or in the setting of von Hippel-Lindau disease (32%). According to MR findings, this tumor can be divided into 3 types: a) syringeal (the tumor varies in size and is associated with syringobulbia and syringomyelia); b) cystic (the tumor shows a cyst with a small mural nodule); and c) solid (the tumor reveals a giant solid mass). Only 5% are found as the only location at the C1-C2 level. The authors report a rare case of a 32-year-old woman who underwent extensive removal of an intramedullary hemangioblastoma at the C1-C2 level. The relevant literature is also discussed. The patient came to our attention after the onset of chronic headache and numbness of both upper extremities. Cervical MRI showed the presence of an oval mass in the intradural space at the C1-C2 level, with intense enhancement after contrast. The lesion was associated with hydrosyringomyelia extending from the obex up to C5. The patient underwent a suboccipital craniectomy, C1 laminectomy, and total removal of the intradural intramedullary lesion. The intraconal lesion appeared reddish with tense, elastic consistency, richly vascularized, invading the underlying spinal cord reaching the central canal. The patient was discharged neurologically intact 7 days after surgery. Histopathological study confirmed a capillary hemangioblastoma of the cervical spine. Immunohistochemical profiling revealed the presence of stromal cells positive for S100, neuron-specific enolase, vimentin, and aquaporin. Vascular lining cells were negative for CD31 and CD24. The low incidence of intramedullary capillary hemangioblastomas poses problems in the identification of histological variability, diagnostic imaging, clinical manifestations, and natural history of the disease. Problems of differential diagnosis arise mainly with metastatic renal cell carcinoma, resolved by positivity for epithelial membrane antigen, CD10, and inhibin (the latter being positive in hemangioblastoma). NeuroTouch Cranio is a virtual reality (VR) simulator with haptic feedback developed to train and assess technical skills in neurosurgical oncology. The simulator was used in a competitive setting to investigate the performance of medical professionals with varying baseline characteristics. Participants were enrolled at the sixth Top Gun skills competition held during the 2011 American Association of Neurological Surgeons annual meeting. The simulated task performed with NeuroTouch Cranio was the internal debulking of a convexity meningioma. Participants were given 150 seconds to complete the task. The instruments available were an ultrasonic aspirator (right hand) to remove tumor and suction (left hand) to clear blood from the surgical field. Automated performance metrics included the total volume of tumor and normal brain removed, path lengths of tool tips, forces applied on tissue, and efficient use of the ultrasonic aspirator. Results were analyzed according to the participants training level, gender, handedness, number of meningiomas removed, and hours per week playing a musical instrument or video games. Univariate analyses of variance (ANOVAs) were conducted on each factor. ANOVAs were computed to adjust for different covariates. Seventy-two participants were enrolled (10 medical students, 18 junior residents [PGY 1-3)], and 44 senior residents [PGY 4 + ]). Training level was found to have a significant effect on the total score (P 0.028), the percentage of tumor removed (P 0.002), and the efficiency ratio of the ultrasonic aspirator (P 0.003). Post-hoc analyses showed significant differences between the medical students and the residents but not between the junior and senior residents. After covariate adjustment for the number of meningioma cases, only the percentage of tumor removed and the efficiency ratio remained as significant factors. This study is the first to demonstrate that the users characteristics affect their performance on a VR neurosurgical oncology task. ST-26. INTRACAVITARY GELFOAM PLACEMENT DURING INTRAOPERATIVE MRI AIDS IN EVALUATION OF RESIDUAL BRAIN TUMOR AND EXTENT OF RESECTION Nandita Guha-Thakurta and Sujit S. Prabhu; The University of Texas MD Anderson Cancer Center, Houston, TX INTRODUCTION: High-field intraoperative magnetic resonance imaging (iMRI) is being used to maximize resection of brain tumors. The purpose of this study is to determine if intracavitary gelfoam placement improves the evaluation of brain tumor resection in near-real time. METHODS: A retrospective analysis was performed of 10 brain tumor patients in whom gelfoam was placed in the resection cavity during iMRI. The extent of resection (EOR), 3-dimensional morphology of the resection cavity, relationship to the adjacent brain parenchyma, and imaging characteristics of the gelfoam were evaluated. Comparison of the same parameters during iMRI in brain tumor patients without intracavitary gelfoam placement served as controls. RESULTS: In comparison to controls, patients with intracavitary gelfoam demonstrated no significant collapse of the resection cavity and improved delineation of the resection margins. This enabled better evaluation of the EOR. Furthermore, the gelfoam demonstrated homogeneous intermediate signal intensity on T1-weighted images and low signal intensity on T2-weighted images. CONCLUSION: This study supports the usefulness of gelfoam placement in the resection cavity during iMRI by improving visualization of residual tumor and optimizing EOR in brain tumor patients. ST-27. COMBINED BRAIN MAPPING AND INTRAOPERATIVE MRI FOR BRAIN TUMOR RESECTION Michael Schulder, Salvatore Zavarella, Dominic Nardi, and Sarah Schaffer; Hofstra North Shore LIJ School of Medicine, Manhasset, NY INTRODUCTION: Surgery on patients with tumors in the dominant hemisphere for language is best done with awake language mapping, to allow for maximally safe lesion resection. The ideal technique combines awake mapping with intraoperative MRI (iMRI). PATIENTS AND METHODS: Twelve patients with tumors in the dominant hemisphere underwent awake resection with iMRI guidance. Imaging was done using a compact, 0.15-Tesla iMRI unit that is designed to sit under the head of the operating table. Diagnoses included low-grade glioma in 7 patients and high-grade glioma in 5. Tumor location was in the frontal lobe in 7 patients, the temporal lobe in 4, and the parietal lobe in 1. Patients were positioned so they could see the examining neuropsychologist. Pin fixation was used to avoid unnecessary movements during surgery. Local anesthesia was used to anesthetize the pin sites and the incision line, and to block major scalp nerves below the planned surgical site. Intravenous sedation was used without intubation. Patients were awakened in time for mapping after dural opening. Language, memory, and cognitive testing were done, as appropriate for the particular surgery, by a neuropsychologist who had tested the patient before surgery. Imaging was done before surgery and when tumor resection was thought to be complete. RESULTS: The number of images acquired in the operating room ranged from 1 to 5 (mean, 2.7); iMRI added an hour on average to surgery. In 7 of 12 patients, intraoperative imaging led to further resection. A gross total resection was achieved in 8 patients, including 4 patients as a result of iMRI. Language mapping led to a cessation of surgery in 4 of 12 patients. CONCLUSIONS: Combined awake language mapping and iMRI guidance is feasible for resection of dominant hemisphere tumors. There is no need for clinicians and patients to choose between these modalities. ST-28. STEREOTACTIC BRACHYTHERAPY WITH 125IODINE SEEDS FOR TREATMENT WHO GRADE II AND III GLIOMAS LOCATED IN THE CENTRAL SULCUS AREA Maximilian I. Ruge1, Stefan Grau1, Manuel Fuetsch1, Philip Kickingereder1, Christina Hamisch1, Harald Treuer1, Juergen Voges2, and Volker Sturm1; 1University Clinics of Cologne, Cologne, Germany; 2University Clinics of Magdeburg, Magdeburg, Germany OBJECTIVE: Gliomas located in the central sulcus (CS) area still represent a major challenge for resective surgery due to a high risk of severe neurological deficits. Stereotactic brachytherapy (SBT) has been reported to be a minimally invasive, safe, and effective treatment alternative, especially for circumscribed tumors (, 4 cm) located in eloquent structures. In this study, we determined the value of SBT for treating gliomas infiltrating the CS. METHODS: For this retrospective analysis, we selected patients with circumscribed WHO grade II or III gliomas (, 4 cm in diameter) located within the CS who were being treated with SBT (125Iodine seeds; surface dose, 50 Gy; permanent implantation) at our institution from 1997 to 2010. Patients with de novo WHO grade III gliomas received adjuvant percutaneous radiotherapy (boost dose, 15-30 Gy). Early and late postoperative procedure-related complications and time-to-tumor progression (TTP) stratified by WHO grade were evaluated. For statistical analysis, Kaplan-Meier estimates and the log-rank test were used. RESULTS: Sixty patients (WHO grade II, 29; WHO grade III, 31) met the selection criteria. There was no procedure-related mortality. Thirty days after seed implantation, 3 of 60 patients (5%) had transient neurological deficits and 6 of 60 patients (10%) had temporarily increased seizure activity, all of which improved after adaptation of anticonvulsive medication. After a median of 66.6 months (range, 39-114 months), 3 patients (5.0%) developed spaceoccupying cysts requiring surgical intervention. Only 1 patient developed a mild permanent motoric deficit. The median TTP was 53.6 months for patients with WHO grade II gliomas and 26.2 months for patients with WHO grade III gliomas. CONCLUSION: Compared to neurosurgical resection, SBT, with low rates of transient complications, is a safe and minimally invasive treatment option for patients with circumscribed WHO grade II or III gliomas located in the CS. The achieved tumor control is comparable to results reported after resection in non-eloquent areas. ST-30. ANALYSIS OF LONG-TERM OUTCOMES OF GROSS TOTAL RESECTION VERSUS SUBTOTAL RESECTION FOLLOWED BY RADIOTHERAPY IN PRIMARY SPINAL EPENDYMOMAS Winward Choy, Andrew Yew, Marko Spasic, Daniel Nagasawa, Won Kim, and Isaac Yang; UCLA David Geffen School of Medicine, Los Angeles, CA INTRODUCTION: Spinal ependymomas are slow-growing neuroepithelial tumors that represent 60% of all primary spinal cord tumors. Although ideal, gross total resection (GTR) may not be safe or viable depending on tumor accessibility and proximity to sensitive structures. We analyzed the benefits of GTR followed by radiation therapy (GTR + RT) versus subtotal resection followed by RT (STR + RT) to determine whether the more conservative approach of STR + RT produces comparable long-term outcomes to GTR + RT. METHODS: To identify all reports of primary spinal ependymomas, a comprehensive PubMed search using the keywords spinal ependymoma radiotherapy and spinal ependymoma radiation therapy was performed. Exclusion criteria included treatment with only RT, follow-up , 5 years, or absent pertinent clinical data. Patients were divided into GTR + RT and STR + RT subgroups, and progression- free survival (PFS) and overall survival (OS) were aggregated and analyzed. Data that could not be disaggregated according to the 2 treatment groups were excluded. RESULTS: A total of 256 patients with primary spinal ependymomas receiving either GTR or STR with adjuvant RT met our inclusion criteria. For all patients, the average 5-year PFS, 10-year PFS, 5-year OS, and 10-year OS were 76.7%, 66.7%, 86.5%, and 72.5%, respectively. A total of 44 patients received GTR + RT while 212 patients received STR + RT. Patients receiving GTR + RT compared to those receiving STR + RT had better OS but no significant improvement in PFS, including 5-year PFS (85% vs 76.1 %, P .7344), 10-year PFS (75% vs 66.0%, P .4965), 5-year OS (96.5% vs 84.8%, P .0577), and 10-year OS (96.3% vs 70.1%, P .0009). CONCLUSION: Our data suggest that more aggressive resection with adjuvant RT is associated with improved OS. There was no significant difference in either the 5-year or 10-year PFS for a subtotal resection. Further studies will be necessary to elucidate the role of postoperative RT in local tumor control, PFS, and OS. ST-31. USE OF 5-AMINOLEVUNILIC ACID (ALA) FOR GLIOMA SURGERY Matthew R. Quigley, Joseph Hobbs, and Sanjay Bhatia; Allegheny General Hospital, Pittsburgh, PA INTRODUCTION: 5-aminolevunilic acid (ALA) has been reported to aid in the resection of glioblastoma (GBM). We sought to investigate its use for surgeries involving all grades of glioma. METHODS: A prospective clinical and radiographic review of all patients who were administered 5-ALA as an adjunct to their planned glioma resection, starting in December 2009 under an FDA Investigational New Drug. RESULTS: There were 40 patients enrolled and 25 were men (average age, 57). Pathology was 20 primary GBMs, 6 recurrent GBMs, 4 GBMs arising from previously known grade II tumors, 2 grade III tumors, 7 grade II tumors, and 1 metastatic adenocarcinoma. Only primary and recurrent GBM exhibited intra-operative fluorescence, whereas secondary GBM and lower-grade lesions did not. Multiple biopsies (1-10 per patient) of normal-appearing white matter adjacent to resected tumor were analyzed. Within fluorescing areas, 92 of 97 cases (94.8%) contained tumor on histologic review, whereas nonfluorescent white matter contained tumor in 24 of 43 cases (55.8%). Among the 26 GBMs that fluoresced, pre-operative review of the MR suggested a complete resection to be possible in 21; post-operative MR was free of tumor in 18 of 21 (85.7%). Fluorescing tumor routinely extended well beyond the boundaries of contrast enhancement as evidenced on neuronavigation, and in 23 of 26 cases (88.5%) tumor went to the ventricular wall. Morbidity associated with 5-ALA was only minor liver function elevations that plateaued at 1 week after administration and had normalized to , 2x normal by 1 month. CONCLUSIONS: Only primary GBM appears to exhibit intra-operative fluorescence following administration of 5-ALA. Virtually all tissue that exhibits fluorescence in the operative bed is tumor. Most GBMs will extend to the ventricular wall if the resection follows the track of fluorescing tissue. In our hands, the use of 5-ALA markedly improves the rate of gross total resection. ST-32. TRANSSPHENOIDAL RESECTION OF GROWTH HORMONE-SECRETING ADENOMAS: SURGICAL OUTCOME Zvi R. Cohen1, Ilan Shimon2, and Moshe Hadani1; 1Sheba Medical Center, Tel Ashomer, Israel; 2Rabin Medical Center, Petach-Tikva, Israel OBJECTIVE: Transsphenoidal surgery is the preferred treatment modality for growth hormone (GH)-secreting pituitary adenomas. In many series, the reported postoperative remission is based mainly on achievement of GH levels less than 2 ng/ml. Strict criteria for insulin-like growth factor I normalization and even lower GH levels (, 1 ng/ml) are suggested to define cure of acromegaly. We analyzed our postoperative results in a large cohort of patients with acromegaly. METHODS: 101 patients harboring GH-secreting adenomas (45 microadenomas and 56 macroadenomas) underwent transsphenoidal surgery between 2000 and 2009. Fifty were men (49.5%) and 51 were women (50.5%). The mean age was 47.6 years (range, 22-81 years). Ninety-seven patients were operated on for the first time, and 4 patients underwent reoperations because of previous surgical failure. Biochemical remission was defined as a repeated fasting or glucosesuppressed GH level of 2 ng/ml or less and a normal insulin-like growth factor I level. RESULTS: Remission was achieved in 70% of all patients after 1 operation, including 86 % of patients with microadenomas and 54% of patients with macroadenomas. Two patients with macroadenomas larger than 20 mm had postoperative active disease. Recurrence was rare (2 patients), and all failed surgical attempts could be detected during the immediate postoperative evaluation. None of the patients with previous surgical failure achieved remission following second operation. Seven patients (6.9 %) had spinal drainage postoperatively to prevent cerebrospinal fluid leaks, and 1 patient underwent surgical repacking. Transient diabetes insipidus was diagnosed and treated in 22 patients (21.7%), and 9 patients (8.9 %) had the syndrome of inappropriate antidiuretic hormone hypersecretion. CONCLUSION: Our series demonstrates the efficacy of transsphenoidal surgery for acromegalic patients with microadenomas and noninvasive macroadenomas. However, patients with large adenomas (. 20 mm) have a poor prognosis and require adjunctive medical or radiation therapy to control GH hypersecretion. ST-33. THE ROLE OF SURGERY IN THE COMBINED TREATMENT OF RECURRENT MALIGNANT GLIOMAS Carmine M. Carapella, Piero A. Oppido, Antonello Vidiri, Stefano Telera, Alfredo Pompili, Veronica Villani, Alessandra Fabi, and Andrea Pace; Regina Elena National Cancer Institute, Rome, Italy The relevance of surgical removal in the therapeutic strategy of malignant gliomas and the opportunity offered by unconventional treatments or second-line chemotherapeutic schedules have deeply modified the indications to surgical treatment of progressive and recurrent malignant gliomas. In the last 3 years, we have operated on 35 recurrent malignant gliomas (20 glioblastomas, 10 anaplastic oligodendrogliomas, and 5 anaplastic astrocytomas). At the moment of the initial diagnosis, the patients were submitted to first-line surgical treatment, radiotherapy, and chemotherapy; all the patients presented documented localized tumor progression or recurrence at least 6 months after the previous surgery and were evaluated as suitable for second-line treatments. In 10 patients, fluorescence-guided surgical removal was performed with the aim of better documenting active tumor tissue. In 21 cases, it was possible to realize an extended resection; in 14 cases, during the same procedure, intratumoral CT wafers were positioned into the resection cavity. The main side effects of second surgical treatment have been local, as cerebrospinal fluid fistula and wound dehiscence that complicated immediate postoperative period. The patients, who were followed in our department, with the relevant contributions of our home assistance service, have been submitted to second- and/or third-line chemotherapy; in 6 cases, highly focused stereotactic radiotherapy has also been performed. The present data tend to confirm the relevance of surgical treatment and subsequent combined treatments for recurrent malignant gliomas, providing for a median extension of overall survival of at least 9 months (3-16+ months in the present series), with a significant improvement of neurological status and prompt resolution of intracranial hypertension. Our clinical experience has been examined and the results in terms of clinical benefit, progression-free survival, and overall survival are presented, trying to design the patients setting where second surgical removal seems to yield more specific indication. Whether maximal surgical resection of glioblastoma improves patient survival has been controversial, as it is difficult to perform an unbiased assessment of extent of resection (EOR) independent of other patient-specific prognostic factors. Recently, glioblastoma has been sub-classified into 4 distinct molecular risk groups (RGs), which have been validated as prognostic biomarkers in the randomized clinical trial of temozolomide dosing in glioblastoma: the Radiation Therapy Oncology Group 0525 (RTOG-0525) trial. We sought to perform exploratory analyses examining gross total resection (GTR) versus sub-total resection (STR) within these RGs in RTOG-0525 patients. Across all randomized patients, n 354 had STR and n 450 had GTR as determined by neurosurgeon operative report. GTR was not significantly associated with survival across the overall study group. A total of 725 patients had sufficient tissue for determination of molecular RG. There were no significant differences in percentage of GTR between each of the 4 RGs (P 0.64). In exploratory subgroup analyses, GTR was associated with improved survival only for patients with tumors from RG4. Hazard ratios (95% confidence intervals) were 0.52 (0.08-2.07) for RG1 (n 28, 68% GTR), 1.74 (0.75-4.05) for RG2 (n 39, 56% GTR), 1.09 (0.84-1.42) for RG3 (n 284, 56% GTR), and 1.26 (1.01-1.56) for RG4 (n 374, 55% GTR). In univariate analysis within RG4, GTR was associated with a median survival of 14.6 months vs 12.7 months for STR (P 0.0352. In a Cox model adjusting for age, KPS, and neurologic function (NF), surgery remained an independent factor within RG4: GTR (P 0.0331), age (P 0.0014), KPS (P .3289), and NF (P 0.3804). There are important cautions in the interpretation of these data, including lack of MRI confirmation of EOR, and inclusion of a range of STR (from biopsy to near-total resection). However, these exploratory results raise the possibility that upfront characterization of tumor molecular profile may allow for personalized therapeutic strategies to improve outcomes for patients with glioblastoma. ST-35. SURGICAL RESECTION AS A PROGNOSTIC FACTOR IN HIGH-GRADE GLIOMAS TREATED WITH BEVACIZUMAB THERAPY Ashwatha Narayana, Saroj D. Kunnakkat, Erik Parker, Deborah Gruber, Michael Gruber, Edmond Knopp, David Zagzag, and John Golfinos; New York University Langone Medical Center, New York, NY INTRODUCTION: Extent of surgical resection is an important prognostic factor for newly diagnosed high-grade glioma (HGG), but its role remains controversial in recurrent tumors. In this analysis, we tested the impact of surgical resection as a prognosticator in the setting of bevacizumab therapy in both newly diagnosed and recurrent HGG. METHODS: From 2006-2010, 60 newly diagnosed HGG patients (51 with glioblastoma [GBM] and 9 with anaplastic astrocytoma [AA]) underwent surgical resection followed by involved field radiation therapy to 59.4 Gy in 33 fractions along with concomitant chemotherapy with temozolomide and bevacizumab at 10 mg/kg every 2 weeks followed by 6 cycles of adjuvant temozolomide with bevacizumab. During the same interval, 130 patients with recurrent HGG (92 with GBM and 38 with AA), of whom 35 underwent surgical re-resection, were treated with bevacizumab at 10 mg/kg every 2 weeks and CPT-11/carboplatin chemotherapy. Progression-free survival (PFS) and overall survival (OS) were assessed from the time of surgery for the analysis. RESULTS: The extent of surgical resection was gross total, subtotal, and biopsy alone in 39 (65%), 12 (20%), and 9 (15%) in newly diagnosed HGG and 12 (9.2%), 22 (16.9%), and 1 (0.8%) in recurrent HGG, respectively. The median PFS in the 3 groups was 13, 13, and 4 months in newly diagnosed HGG (P 0.024) and 4, 10, and 3 months in recurrent HGG (P 0.381), respectively. The median OS in the 3 groups was 24, 19, and 11 months in newly diagnosed HGG (P 0.113) and 8, 14, and 4 months in recurrent HGG (P 0.271), respectively. There was no difference between patients with AA and GBM. CONCLUSION: Patients with biopsy alone do poorly in the setting of anti-angiogenic therapy in both newly diagnosed and recurrent gliomas. However, patients with sub-total resection seem to benefit from anti-angiogenic therapy and do as well as those with gross total resection. ST-36. RETROSPECTIVE ANALYSIS OF OUTCOMES AFTER SURGICAL BIOPSY OR DEBULKING IN PATIENTS WITH BUTTERFLY GLIOBLASTOMAS Kristine Dziurzynski, David Blas-Boria, Dima Suki, Daniel Cahill, Sujit Prabhu, Vinay Puduvalli, and Nicholas Levine; The University of Texas MD Anderson Cancer Center, Houston, TX The role of surgical debulking in the management of patients with butterfly glioblastoma has not been determined. It is unknown if a survival or clinical performance advantage exists for patients with this diagnosis who will undergo radiation therapy and chemotherapy. A retrospective analysis was performed of records from 2000-2010 that identified 23 patients with butterfly glioblastoma. Patients were divided into 2 major treatment groups: biopsy or surgical debulking followed by radiation therapy and chemotherapy. Demographics, differences in survival, complication rates, and KPS among the groups were assessed. Median age was 59 years and 52% were women. Median preoperative KPS was 80 and median tumor volume was 37.8 cm3 (range, 0.8 to 132.3 cm3). Four patients underwent biopsy only prior to radiation therapy and chemotherapy; 11 underwent surgical debulking (6 with and 5 without adjuvant therapy); and 8 were treated with palliative measures only. No patient who did not undergo surgical debulking prior to radiation therapy and chemotherapy underwent urgent decompressive craniotomy during treatment. The postoperative complication rate was 22%. Overall median post-surgical survival of the whole group was 6.6 months (95% confidence interval [CI], 4.9-8.3). There was no significant difference in the survival of patients receiving biopsy versus debulking when surgery was followed by adjuvant therapy (hazard ratio [HR] for biopsy 1.1; 95% CI, 0.3-4.3; P 0.8); there was a strong trend (though not significant) for longer survival with debulking in the absence of adjuvant therapy (HR for biopsy 5.3; 95% CI, 0.6-48.6; P 0.1). Numbers were too small to assess outcomes among subgroups of patients with various preoperative characteristics. Among the 20 patients with available data, 65% had the same KPS postoperatively, 20% worsened, and 15% improved. The median reduction in enhancing volume from surgical resection was 100% (range, 25-100). ST-37. EXTENT OF REPEAT RESECTION IS PREDICTIVE OF SURVIVAL FOR RECURRENT GLIOBLASTOMA MULTIFORME Orin Bloch, Seunggu J. Han, Gurvinder Kaur, Manish K. Aghi, Michael W. McDermott, Mitchel S. Berger, and Andrew T. Parsa; University of California, San Francisco, San Francisco, CA INTRODUCTION: Initial extent of resection (EOR) of glioblastoma multiforme (GBM) has been shown to be an important prognostic factor for overall survival. Many GBM patients will undergo a second resection at the time of recurrence for histopathologic confirmation and cytoreduction. The implications of EOR at repeat craniotomy remain unclear. We have reviewed our experience with repeat resection for recurrent GBM to study the importance of EOR. METHODS: We retrospectively analyzed records of all patients undergoing craniotomy for resection of primary GBM at our institution from 2005-2009. Patients who underwent a second craniotomy for pathologically confirmed recurrent disease after primary therapy were included. Survival was compared between EOR groups using univariate and multivariate analysis. RESULTS: Repeat resection was performed in 114 patients, of whom 65(57%) had sub-total resection at initial craniotomy and 49(43%) had total resection. Among patients with initial sub-total resection, 32 (49%) had sub-total resection at recurrence with a median survival of 14.7 months (95% confidence interval [CI], 12.2-17.1 months) compared to 33 (51%) with complete resection and a median survival of 19.5 months (95% CI, 17.8-21.2 months), representing a significant difference (log rank, P 0.004). Among patients with initial total resection, 20 (41%) had sub-total resection at recurrence with a median survival of 18.3 months (95% CI, 17.4-19.2 months) compared to 29 (59%) with complete resection and a median survival of 19.7 months (95% CI, 16.9-22.5 months). This difference was not significant (log rank, P 0.31). A Cox proportional hazards model was constructed demonstrating age, KPS, and EOR at second resection as independent predictors of survival (hazard ratio 2.14, P 0.001). CONCLUSIONS: EOR at recurrence is an important predictor of overall survival in patients with initial sub-total resection but does not influence prognosis following initial complete resection. Patients with a total resection following initial sub-total resection have a median survival similar to patients with initial total resection, suggesting the importance of extensive repeat resection at the time of tumor progression. ST-38. EVOLVING PATTERNS OF CNS METASTATIC DISEASE IN BREAST CANCER Matthew R. Quigley, Olivia Fukui, Brandon Chew, and Sanjay Bhatia; Allegheny General Hospital, Pittsburgh, PA INTRODUCTION: Despite a marked improvement in overall survival among breast cancer patients over the last 2 decades, over 30% will eventually develop CNS disease, usually leading to a rapid demise. We wondered whether the combination of new therapies and diagnostics had altered the profile of CNS disease from breast cancer METHODS: A retrospective review of pathologic, radiographic, and clinical materials of all breast cancer patients with CNS involvement in a large tertiary center identified from 2 prospectively maintained databases from January 2003 to February 2010. RESULTS: There were 88 patients, all women, with age at initial diagnosis and CNS diagnosis of 51.6 and 56.9, respectively. Invasive ductal was the most common pathology (84%) and 52% were human epidermal growth factor receptor 2 (HER2) positive. Median time to CNS disease was 28.6 months, at which time 78% had metastatic disease elsewhere and 67% had active disease. CNS pathology was multiple metastases in 68%, solitary in 17%, and leptomeningeal disease (LMD) only in 15%. Median survival from original diagnosis was 50 months with estrogen receptor (ER) positive being the only significant positive factor (P .02). Median survival from the CNS diagnosis was 9.7 months and was negatively related to presenting with LMD (P .004; hazard ratio [HR] 3.1) and triple-negative disease (ER negative, progesterone receptor negative, and HER2 negative; P .02; HR 2.3). Those presenting with LMD or with triple-negative pathology survived only 3.1 months. Twenty-six percent of patients with metastatic disease went on to develop LMD, at which point survival was short (median, 1.8 months). HER2 positivity appeared protective against developing LMD (14.7% vs 42.3%; P .03). Lower extremity weakness was the most common sign of LMD (42%), followed by cranial neuropathies (33%). CONCLUSION: Despite advances in treating breast cancer, CNS involvement remains an ominous sign. Survival is adversely affected by presentation with LMD as well as triple-negative hormonal and HER2 status. Subsequent development of LMD is common and may present with such non-specific symptoms as back pain. ST-39. MANAGEMENT OF RECURRENT OR PROGRESSIVE CEREBRAL METASTASES FOLLOWING SURGICAL RESECTION OR STEREOTACTIC RADIOSURGERY John J. DePowell, Christopher Sanders-Taylor, Jessica Guarnaschelli, and Christopher McPherson; University of Cincinnati, Cincinnati, OH INTRODUCTION: Aggressive management of newly diagnosed cerebral metastases with resection and/or stereotactic radiosurgery (SRS) has led to improved outcomes. Despite the established role of these modalities in the treatment of newly diagnosed metastases, there is little consensus in the management following failed initial resection or radiosurgery. METHODS: A retrospective review from 2004-2010 was performed to identify patients who underwent resection or SRS of a cerebral metastasis with subsequent progression or recurrence followed by retreatment. A Kaplan-Meier analysis was performed to determine if the retreatment modality had a significant benefit in terms of local recurrence or progression as well as length of survival. RESULTS: During the study period, 892 patients were treated for cerebral metastasis with resection or SRS. Fifty-five subsequently developed progression or local recurrence. The median time to recurrence was 8 months, at which point 24 were retreated with SRS and 31 with resection. There were six (19%) recurrences following retreatment and 22 deaths (71%) in the surgical group. Nine (38%) of those retreated with SRS recurred/progressed and 19 (79%) died. Median time to progression or recurrence in the surgical resection group was 9.5 months compared to 6 months for the SRS group (P 0.06). Median length of survival following retreatment was 12 months for the surgery group compared to 7.5 months for the SRS group (P 0.37). CONLCLUSIONS: This retrospective review of patients with recurrent or progressive cerebral metastases following surgical resection or SRS found no statistically significant difference between retreatment modalities in regard to length of survival, although there was a trend toward longer time to a second recurrence following retreatment with surgery. This study is limited by the small study population and the small number of patients with recurrence or progression following retreatment. Further statistical analysis of subgroups may determine if one retreatment modality may be more effective for certain patients. An accumulating body of evidence is demonstrating the survival benefit of maximal safe resection in glioma surgery. Recent studies suggest that enhancing contrast between tumor and normal brain using fluorescent dyes improves the surgeons ability to identify and resect abnormal tissue. Current studies of fluorescence-guided resection (FGR) use wide-field, macroscopic imaging, but gliomas are known to exhibit microscopic invasion. We therefore sought to develop and employ a multi-modal imaging apparatus that permits both wide-field imaging and confocal microscopy to enable macroscopic scanning and microscopic resection at the tumorbrain interface. Fresh specimens from craniotomies for resection were obtained with institutional review board approval and examined ex vivo. We investigated 28 glioma specimens (17 grade IV, 4 grade III, and 7 grade II), and 8 specimens of normal brain (from temporal lobectomy for epilepsy). Tissue was stained in 0.2 mg/ml aqueous solution of methylene blue (MB). Multimodal wide-field and confocal images were acquired using a specially built microscope, capable of lateral resolution of 0.6 mm and axial resolution of 7 mm. Reflectance and fluorescence signals of MB were excited at 658 nm. Fluorescence emission and polarization were collected from 670 nm to 710 nm. After imaging, tissue was formalin-fixed and paraffin-embedded for standard neuropathologic evaluation. Four pathologists (2 with and 2 without subspecialty training in neuropathology) provided diagnoses based on the multimodal confocal images. The investigated tumor types exhibited distinctive characteristics in both reflectance and fluorescence imaging modes. Images showed distinct morphological features similar to standard histology. Pathologists were able to distinguish between high- and low-grade glioma and normal brain tissue in all cases and render diagnoses from the images in a manner comparable to hematoxylin & eosin slides. This study represents an important step toward intraoperative microscopic FGR, which will hopefully further improve extent of resection and attendant survival benefit. Little prospective data exist to address the relationship between extent of resection of newly diagnosed glioblastoma (GBM) and clinical outcome. The Radiation Therapy Oncology Group (RTOG) 0525 trial is the largest prospective, randomized trial completed to date in newly diagnosed GBM, randomized 833 patients. The surgeons estimate of the completeness of resection was available for 804 patients (post-operative images were not collected in this study). Kaplan-Meier, Cox proportional hazards, and logistical regression analysis were performed to analyze the relationship between completeness of resection (partial vs complete) and age, post-operative KPS, gender, MGMT methylation status, progression-free survival (PFS), and overall survival (OS). Three hundred fifty-four patients were reported to have a partial resection and 405 complete. There were no statistically significant associations between age, gender, or tumor methylation status and completeness of resection. When data were pooled, irrespective of adjuvant treatment arm, there was a trend toward improved OS with complete resection, but this was not statistically significant (16.6 vs 15.1 months; P 0.09; hazard ratio [HR] 0.87, 0.75- 1.02). No relationship was observed between completeness of resection and overall PFS. Furthermore, when the 2 adjuvant treatment arms were examined separately by completeness of resection, there were no differences in OS or PFS observed. There was a strong relationship between complete resection and higher post-operative KPS (90-100) (HR 1.66, 1.24-2.23; P 0.0007) and better neurological outcome (HR 1.44, 1.02-2.03; P 0.04). Furthermore, higher postoperative KPS was associated with improved OS (HR of death 0.63, 0.54-0.74; P , 0.0001). Overall, there was not a statistically significant relationship between surgeon-estimated completeness of resection and outcome. However, more complete resections were associated with improved KPS and neurological outcome. The follow-up study RTOG 0825 will allow for a comparison between volumetric measurement of completeness of resection and outcome. ST-42. OLIGODENDROGLIOMA ASSOCIATED WITH AN ACUTE SUBDURAL HEMATOMA: CASE REPORT AND REVIEW OF THE LITERATURE Joseph F. Megyesi, David Macdonald, Bill Wang, and Godefroy Hardy-St Pierre; University of Western Ontario, London, ON, Canada Acute subdural hematoma secondary to hemorrhage within a brain tumor is a rare occurrence. Case Report: After presenting with seizures, a 43-year-old man had a biopsy-proven oligodendroglioma diagnosed 11 years ago. He received focal cranial radiation. Neuroimaging had shown stability until 1 year ago. The left frontal-parietal tumor was showing gradual enlargement with a new area of enhancement. His seizures had become harder to control. Resection was not deemed to be safely feasible and he had started his first cycle of temozolomide chemotherapy. At the start of his second week of chemotherapy, while at work, he developed a severe left-sided headache with progressive impairment of speech and rightsided hemiplegia. He became unconscious and required intubation. Computed tomography of the head revealed a large left-sided acute subdural hematoma with shift of midline structures. There was evidence of intratumoral hemorrhage. He was taken urgently to the operating room and underwent evacuation of the acute subdural hematoma, evacuation of the intra-tumoral blood, and biopsy of the tumor wall in the vicinity of the hemorrhage. The biopsy showed oligodendroglioma with anaplastic features. He made a neurological recovery to his baseline. The literature was reviewed for cases of brain tumor causing acute subdural hematoma. This revealed a paucity of such cases. The literature review will be shown. Clinicians should consider acute subdural hematoma in patients with a brain tumor that present with acute neurologic deterioration. ST-43. TUMOR RESECTION IN THE HIGH-FIELD INTRAOPERATIVE MRI Jason M. Hoover, Stephan J. Goerss, Timothy J. Kaufmann, Fredric B. Meyer, and Ian F. Parney; Mayo Clinic, Rochester, MN High-field intra-operative MRI (iMRI) provides excellent guidance for brain tumor surgery. We report our initial experience and seek to define iMRIs safety and impact on resection. Data were collected from 75 adult brain tumor patients undergoing surgery in 1.5-Tesla iMRI from 2008-2010. Pre-operative, inter-dissection, and post-operative imaging was obtained. Seventy-three patients underwent resection and 2 underwent open biopsy. Images were registered to frameless stereotaxy system. Volumetric assessments were performed using software. Short-term outcomes at 3 months were assessed. Mean age was 42.4 years (range, 21-71 years) with 43 men and 32 women. Tumors were predominantly gliomas, located on the left, frontal in location, and non-enhancing. Motor or speech mapping was performed in 39 patients and awake craniotomy in 23. Temporary neurological deficits occurred in 25%, permanent deficits (both stroke) in 2.7%, infection in 2.7%, and cerebrospinal fluid leak in 1.3%. Pre-operative non-enhancing and enhancing tumor volumes were 43.4 mL and 33.4 mL, respectively. For non-enhancing tumors, between pre- and post-operative MRI, 80% of tumor was resected; between preand inter-dissection MRI, 67% of tumor was resected; and between interdissection and post-operative MRI, 41% more tumor was resected. For enhancing tumors, between pre- and post-operative MRI, 88% of tumor was resected; between pre- and inter-dissection MRI, 84% of tumor was resected; and between inter-dissection and post-operative MRI, only 20% more tumor was resected. Thus, the amount of tumor resected between interdissection and post-operative MRI was double for non-enhancing gliomas than enhancing gliomas. However, overall resection on post-operative MRI was higher for enhancing glioma than non-enhancing glioma at 88% vs 80%, respectively. High-field iMRI can be easily and safely combined with electrophysiological mapping for cases in eloquent cortex. Extent of resection was limited only by safety concerns due to tumor location. High-field iMRI had greater impact on extent of resection of non-enhancing than enhancing gliomas. ST-44. CRANIAL ANGIOSARCOMA: SUCCESSFUL MULTIDISCIPLINARY MANAGEMENT OF THIS RARE ENTITY Bharat Guthikonda, Jai Thakur, Imad Khan, Osama Ahmed, Cedric Shorter, Jon Wilson, James Welsh, Hugo Cuellar, and Majed Jeroudi; LSU Health Sciences Center Shreveport, Shreveport, LA Angiosarcomas are rare, highly malignant endothelial tumors. Central nervous system involvement with angiosarcoma has rarely been described involving the dura or the cranium. Cranial angiosarcoma is extremely rare and appears to have a particularly poor prognosis. We describe a multidisciplinary approach to managing cranial angiosarcoma that has led to a good neurological outcome and survival greater than 3 years in a pediatric patient. We present a 16-year-old male who was first noted to have a parietal subgaleal/cranial mass that was biopsied in 2008. Pathology was felt to be consistent with angiosarcoma. The patient subsequently underwent chemotherapy with vincristine. The patient did well until early 2010 when he suffered a right-sided intraparenchymal intratumoral hemorrhage. The patient recovered from this and subsequently underwent a second round of chemotherapy using vincristine, cyclophosphamide, and actinomycin. The tumor continued to progress despite this treatment and he developed extensive skull deformity with massive bilateral fronto-temporo-parietal angiosarcoma. At this point, surgical intervention, which had been previously thought to be too risky due to the extent and vascularity of the tumor, was reconsidered. Preoperative embolization of 90% of the blood supply of the mass was performed. Total removal of bilateral subgaleal and cranial tumor was then performed. The patient tolerated the extensive surgery well and was discharged on postoperative day 6. The patient is currently completing radiation therapy to the entire tumor bed. This will be followed by a third course of chemotherapy to which the patient is naive. He has clinically done well with no neurological deterioration. We highlight the successful management of cranial angiosarcoma in the pediatric population using a multidisciplinary approach. With the combined efforts of pediatric oncology, radiation oncology, interventional neuroradiology, and neurosurgery, longterm survival (. 3 years) is possible with this aggressive pathology.


This is a preview of a remote PDF: https://neuro-oncology.oxfordjournals.org/content/13/suppl_3/iii154.full.pdf

Lola B. Chambless, Scott L. Parker, Laila Hassam-Malani, Matthew J. McGirt, Reid C. Thompson, Tao Zhou, Xianghui Meng, Bainan Xu, Shaobo Wei, Xiaolei Chen, Philip C. De Witt Hamer, Santiago Gil Robles, Aeilko H. Zwinderman, Hugues Duffau, Mitchel S. Berger, Jose Daniel S. Ruiz Gonzalez, Olaya Vazquez Alberto, Herrera Mora Patricia, Kaisorn Chaichana, Courtney Pendleton, Lola Chambless, Jay Nathan, Joaquin Camara-Quintana, Gordon Li, Griffith Harsh, Reid Thompson, Michael Lim, Alfredo Quinones-Hinojosa, Mark E. Oppenlander, Andrew Wolf, Randall Porter, Peter Nakaji, Kris A. Smith, Robert F. Spetzler, Nader Sanai, Jeong Hoon Kim, Aaron J. Clark, Arman Jahangiri, Michael E. Sughrue, Michael W. McDermott, Manish K. Aghi, Clark Chen, Ekkehard Kasper, Peter Warnke, Chul-Kee Park, Se-Hoon Lee, Sang Woo Song, Jin Wook Kim, Tae Min Kim, Fumio Yamaguchi, Tomoko Omura, Hirotomo Ten, Yudo Ishii, Toyoyuki Kojima, Hiroshi Takahashi, Akira Teramoto, Erlick A. Pereira, James Livermore, Olaf Ansorge, Stana Bojanic, Xianghui Meng, Bainan Xu, Xiaolei Chen, Shaobo Wei, Tao Zhou, Huaiyu Tong, Xinguang Yu, Dingbiao Zhou, Yuanzheng Hou, Zhe Zhou, Jiashu Zhang, Andrew J. Fabiano, Nestor Rigual, Stephan Munich, Robert A. Fenstermaker, Xiaolei Chen, Xianghui Meng, Jiashu Zhang, Fei Wang, Yan Zhao, Bai-nan Xu, Eui Hyun Kim, Min Chul Oh, Eun Jig Lee, Sun Ho Kim, Young-Hoon Kim, Chae-Yong Kim, Yong Hwy Kim, Jung Ho Han, Chul-Kee Park, Seung-Ki Kim, Sun Ha Paek, Kyu-Chang Wang, Dong Gyu Kim, Hee-Won Jung, Xiaolei Chen, Xianghui Meng, Fei Wang, Yan Zhao, Bai-nan Xu, Dietmar Krex, Claudia Lindner, Tareq Juratli, Claudia Raue, Gabriele Schackert, Pablo A. Valdes, Anthony Kim, Frederic Leblond, Olga M. Conde, Brent T. Harris, Keith D. Paulsen, Brian C. Wilson, David W. Roberts, Dietmar Krex, Tareq Juratli, Claudia Lindner, Claudia Raue, Gabriele Schackert, Giuseppe Occhiogrosso, Pasquale Cascardi, Maria Blagia, Antonio De Tommasi, Nicholas Gelinas-Phaneuf, Nusrat Choudhury, Ahmed Al-Habib, Anne Cabral, Etienne Nadeau, Mora Vincent, Valerie Pazos, Patricia Debergue, Robert DiRaddo, Rolando F. Del Maestro, Nandita Guha-Thakurta, Sujit S. Prabhu, Michael Schulder, Salvatore Zavarella, Dominic Nardi, Sarah Schaffer, Maximilian I. Ruge, Stefan Grau, Manuel Fuetsch, Philip Kickingereder, Christina Hamisch, Harald Treuer, Juergen Voges, Volker Sturm, Winward Choy, Andrew Yew, Marko Spasic, Daniel Nagasawa, Won Kim, Isaac Yang, Matthew R. Quigley, Joseph Hobbs, Sanjay Bhatia, Zvi R. Cohen, Ilan Shimon, Moshe Hadani, Carmine M. Carapella, Piero A. Oppido, Antonello Vidiri, Stefano Telera, Alfredo Pompili, Veronica Villani, Alessandra Fabi, Andrea Pace, Daniel Cahill, Meihua Wang, Minhee Won, Kenneth Aldape, Rebecca Maywald, Monika Hegi, Minesh Mehta, Mark Gilbert, Erik Sulman, Michael Vogelbaum, Ashwatha Narayana, Saroj D. Kunnakkat, Erik Parker, Deborah Gruber, Michael Gruber, Edmond Knopp, David Zagzag, John Golfinos, Kristine Dziurzynski, David Blas-Boria, Dima Suki, Daniel Cahill, Sujit Prabhu, Vinay Puduvalli, Nicholas Levine, Orin Bloch, Seunggu J. Han, Gurvinder Kaur, Manish K. Aghi, Michael W. McDermott, Mitchel S. Berger, Andrew T. Parsa, Matthew R. Quigley, Olivia Fukui, Brandon Chew, Sanjay Bhatia, John J. DePowell, Christopher Sanders-Taylor, Jessica Guarnaschelli, Christopher McPherson, Sameer A. Sheth, Matija Snuderl, Churl-Su Kwon, Dennis Wirth, Anna Yaroslavsky, William T. Curry, Michael A. Vogelbaum, Meihua Wang, Costas G. Hadjipanayis, Minhee Won, Minesh P. Mehta, Mark R. Gilbert, Joseph F. Megyesi, David Macdonald, Bill Wang, Godefroy Hardy-St Pierre, Jason M. Hoover, Stephan J. Goerss, Timothy J. Kaufmann, Fredric B. Meyer, Ian F. Parney, Bharat Guthikonda, Jai Thakur, Imad Khan, Osama Ahmed, Cedric Shorter, Jon Wilson, James Welsh, Hugo Cuellar, Majed Jeroudi. SURGICAL THERAPIES, Neuro-Oncology, 2011, iii154-iii163, DOI: 10.1093/neuonc/nor164