by age, tumor site, and size, although the actual difference in median survival between biopsied and resected patients varied with some factors (especially age). CONCLUSION: This large, multivariate, population-based study confirmed the value of resection rather than biopsy as a prognostic factor for survival in glioblastoma. Patient- and tumor-related factors such as age and tumor size had little interaction with the difference in survival after resection in this cohort. 798 Accuracy of Image-guided Stereotactic Brain Biopsy in the Diagnosis of Glioma: Comparison of Biopsy and Open Resection Specimen Matthew J. McGirt, M.D., Graeme Woodworth, B.S., Amer Samdani, M.D., Ira Garonzik, M.D., Alessandro Olivi, M.D., Jon Weingart, M.D. INTRODUCTION: Tissue heterogeneity and rapid tumor progres- sion may decrease the accuracy and prognostic value of stereotactic brain biopsy in the diagnosis of gliomas. Correct tumor grading is therefore dependent on accuracy of biopsy needle placement. The accuracy of frameless and frame-based techniques in the diagnosis and grading of glioma remains unstudied. METHODS: The diagnoses of 21 astrocytic brain tumors were derived using image-guided stereotactic biopsy (12 frame-based, 9 frameless) and followed by open resection of the lesion 1.5 months (0.5–4 mo) later. The histological diagnoses yielded by biopsy were compared with subsequent histological diagnosis from open tumor resection. RESULTS: Stereotactic biopsy histology accurately represented the greater lesion at open resection a median of 45 days later in 16 patients (76%) and correctly guided therapy in 19 (91%). Biopsy accuracy of frameless versus frame-based stereotaxy was similar (89 versus 66%, P = 0.21). In 3 patients (14%), biopsy specimens were adequate to diagnose glioma; however, histology was insufficient for definitive tumor grading (grade III versus IV). Anaplastic oligodendroglioma (ODG) was undergraded as low-grade ODG in 1 patients (5%). Biopsy of new-onset GBM yielded necrosis/gliosis and was called nondiag- nostic in one patient. Tumors 50 cm 3 were eightfold less likely to accurately represent the grade of the entire lesion at resection com- pared with lesions 50 cm 3 (OR, 8.8; 95% CI, 0.9–100; P = 0.05). CONCLUSION: Both frameless and frame-based magnetic reso- nance imaging-guided stereotactic brain biopsy are safe and accu- rately represent the larger glioma mass sufficiently to guide subse- quent therapy. The diagnostic accuracy of frameless stereotactic brain biopsy is equivalent to the frame-based technique in the diagnosis and grading of glioma. Glioma volume 50 cm 3 may be a negative prog- nostic indicator of biopsy accuracy. 799 Independent Predictors of Morbidity after Image-guided Stereotactic Brain Biopsy: A Risk Assessment of 270 Consecutive Cases Matthew J. McGirt, M.D., Graeme Woodworth, B.S., James Frazier, M.D., Alex Coon, M.D., Ira Garonzik, M.D., Alessandro Olivi, M.D., Jon Weingart, M.D. INTRODUCTION: Image-guided stereotactic brain biopsy is asso- ciated with transient morbidity in 7 to 10%, permanent morbidity in 4 to 6%, and mortality in 0.5 to 1% of patients. To date, a critical analysis of risk factors predictive of enhanced operative risk in frame-based and frameless stereotactic brain biopsy has not been performed. METHODS: We reviewed the clinical and radiological records of 270 consecutive frame-based and frameless image-guided stereotactic brain biopsies. The associations between biopsy-related morbidity and clinical, operative, and radiological variables were assessed via multivariate logistic regression analysis. RESULTS: Thirty-six patients (13%) experienced streotactic biopsy- related morbidity; 23 (8%) transient, 13 (5%) permanent, 3 (1%) mor- tality. Biopsy site hematoma occurred in 25 patients (9%), and 10 (4%) were symptomatic. Diabetes mellitus (OR, 3.73; 95% CI, 1.37–10.17; P = 0.01), thalamic lesions (OR, 4.06; 95% CI, 1.63–10.11; P = 0.002), and basal ganglia lesions (OR, 3.29; 95% CI, 1.05–10.25; P = 0.040) were independent risk factors for morbidity. In diabetic patients, operative-day glucose 200 mg/dl had 100% PPV; glucose 200 mg/dl 95% NPV for biopsy-related morbidity. Pontine biopsy was not a risk factor for morbidity. Only 2 of 45 patients (1%) with prebiopsy epilepsy experience seizures postoperatively. More than one needle trajectory increased the incidence of deficits from 17% to 44% in deep lesions (basal ganglia, thalamus), P = 0.05, but was not associated with morbidity in cortex lesions. Number of biopsy specimens taken per trajectory was not associated with morbidity in all lesions. CONCLUSION: Basal ganglia lesions, thalamic lesions, and dia- betic patients were independent risk factors for biopsy-associated morbidity. Operative day hyperglycemia strongly predicted morbid- ity in the diabetic population. Epilepsy did not predispose to biopsy- associated seizure. For deep-seated lesions, increasing the number of biopsy samples along an established tract versus performing a second trajectory may minimize morbidity. Stereotactic biopsy is a safe pro- cedure with low permanent morbidity (5%) and should be performed cautiously for lesions of the thalamus and basal ganglia. Close peri- operative glucose control is mandatory. 800 Experience with Transient Spinal Cord Injury in Athletes Julian E. Bailes, M.D. INTRODUCTION: The phenomenon of transient spinal cord injury (TSCI) during athletic competition is one of the most difficult situa- tions that the neurosurgeon may encounter. METHODS: Thirty-five athletes with symptoms of TSCI were evalu- ated by athletic trainers, emergency medical service personnel, and team physicians. Diagnosis included physical and neurological examinations, radiographs with flexion extension studies, computed tomography, and magnetic resonance imaging (MRI). The majority of injuries occurred during football, but also during wrestling, baseball, gymnastics, and skiing. Mechanisms in football involved tackling, driving the helmet into an opponent, and being tackled. Symptoms included paralysis, weak- ness, or numbness of all extremities and hemiparesis and ranged from 15 minutes to 48 hours. No evidence of fracture/dislocation or ligamentous instability was seen, whereas 20 had spinal stenosis. Fourteen were allowed to return to contact sports without recurrent TSCI, with a mean follow-up of 40 months; others retired. RESULTS: Several mechanisms have been proposed to attempt to explain why this intriguing phenomenon with loss of function, with- out identifiable spinal structural deficit, may occur, often called con- cussion. Spinal stenosis was our most common finding. Recent expe- rience has led to the current emphasis on MRI assessment of ABSTRACTS OF OPEN PAPERS 486 | VOLUME 55 | NUMBER 2 | AUGUST 2004 www.neurosurgery-online.com Downloaded from https://academic.oup.com/neurosurgery/article-abstract/55/2/486/2744483 by guest on 28 November 2018