METHODS: A retrospective review was performed for all patients who underwent both cerebral angiography and cerebral biopsy for the workup of a presumptive diagnosis of CNSV at a single center from 2005 to 2016. The results were then subjected to statistical analyses. RESULTS: A total of 57 patients over the period of 2005 to 2016 underwent angiography for workup of presumed CNSV. Twenty patients exhibited angiograms suggestive of vasculitis and 28 patients underwent cerebral biopsy. Only one was positive. The negative predictive value (NPV) for angiography was 94.11% (confidence interval, 71.31-99.85). Specificity was 59.26%. Positive predictive value (PPV) was unable to be calculated based on these data, given that angiography did not accurately predict biopsy diagnosis. Of note, the patient with biopsy-proven vasculitis exhibited negative angiography. Only 3.7% of biopsies were positive for vasculitis. CONCLUSION: Cerebral angiography shows high NPV for CNSV. Angiography did not accurately predict diagnosis in a single case in this series. The prevalence of disease is too low to accurately ascertain PPV. Biopsy results are positive in a minority of cases; however, a large percentage of patients receive steroid treatment in the face of negative results. Cerebral biopsy may not be warranted in cases of positive angiography given the low likelihood of the disease. It may also not be indicated in cases of negative angiography given the high NPV. Surgical risks of cerebral biopsy can likely be avoided without significantly altering the course of treatment for many of these patients. 355 Endovascular Management of Cervical Carotid and Vertebral Artery Dissection: Indications, Techniques, and Outcomes From a 20-Year Experience Karam Moon, MD; Felipe Albuquerque, MD; Tyler Scott Cole, MD; Bradley A. Gross, MD; Cameron G. McDougall, MD INTRODUCTION: Endovascular intervention for cervical carotid artery and vertebral artery dissections (CAD and VAD) may be indicated in specific circumstances. METHODS: We reviewed a prospectively maintained database from January 1996 to January 2016 of extracranial dissections undergoing endovascular intervention. RESULTS: There were 116 patients, including 93 in the CAD cohort and 23 in the VAD cohort, with a mean age of 44.9 years (range 5-76) and mean postprocedure follow-up of 41.6 months (range 1-146). Interventions included stent placement (n ¼ 104), coil occlusion of parent artery (n ¼ 11), or stenting with contralateral vessel coil occlusion (n ¼ 1). The 2 cohorts were well matched in age, sex, dissection etiology, and admission/follow-up modified Rankin Scale (mRS) (P ¼ .362, .371, .175, .355, and .835, respectively). The CAD cohort was significantly more likely to undergo stent placement or have failed medical therapy (P , .001, P ¼ .004). The CAD cohort was also significantly more likely to undergo intervention for enlarging pseudoaneurysm or thromboembolic events (P ¼ .001, .004), whereas the VAD cohort was significantly more likely to undergo intervention for traumatic occlusions with recanalization (P , .001). Etiologies of dissection included spontaneous (n ¼ 67), traumatic (n ¼ 38), and iatrogenic (n ¼ 14), with traumatic dissections being associated with a poor admission mRS (mRS . 3) in the CAD cohort (P ¼ .014). Six (9.0%) patients of spontaneous etiology also reported recent chiropractic manipulation. The permanent morbidity/ mortality rate was 3.4%, including 2 deaths, with a stroke rate of only 0.9% over 4825 patient-years. At last follow-up, 31 of 93 (33.3%) CAD patients and 10 of 23 (43.5%) VAD patients disabled prior to intervention were nondisabled at last follow-up; no patients in either cohort were worsened. CONCLUSION: In a long-term experience, endovascular manage- ment of CAD and VAD is highly effective in specific indications, with an acceptable complication profile. CAD requiring intervention is more likely than VAD to have failed medical therapy, present with thrombo- embolic events and pseudoaneurysms, and undergo primary stent placement, whereas VAD is more likely to undergo treatment for traumatic occlusions with recanalization. 356 Microsurgical Anatomy of the Brainstem Safe Entry Zones: A Cadaveric Study With High-Resolution Magnetic Resonance Imaging and Fiber Tracking Debraj Mukherjee, MD, MPH; Veysel Antar, Bora Gurer, Ulas Cikla, MD; Gabriel Neves, Mehmet Ekici, Tomer Hananya, Aaron S. Field, Shahriar M. Salamat, Mustafa Kemal Baskaya, MD INTRODUCTION: Operative management of intrinsic brainstem lesions remains challenging despite advances in electrophysiological monitoring and neuroimaging. Surgical intervention in this region requires detailed knowledge of adjacent, critical white matter tracts and cranial nerve nuclei. Our aim was to systematically verify internal anatomy associated with each brainstem safety zone entry zone (BSEZ) using a cadaveric model supplemented with neuroimaging modalities com- monly used in preoperatively planning, namely high-resolution magnetic resonance imaging (MRI) and fiber tracking. METHODS: Twelve BSEZs were simulated in 8 formalin-fixed, cadaveric heads. Specimens then underwent radiological investigation including T2-weighted imaging and fiber tracking using 4.7 T MRI. The distance between simulated BSEZs and predefined, adjacent, critical structures was systemically recorded. RESULTS: Entry points and anatomic limits on the surface of the brainstem are described for each BSEZ, along with description of neurological sequelae if such limits are violated. With high-resolution imaging, we verified maximal depth and optimal angle of entry for each BSEZ. The relationship between BSEZs and adjacent, critical structures was quantified. Orbitozygomatic, suboccipital, retrosigmoid, retrolabyrinthine, and petrosectomy approaches were used to simulate BSEZs in the ventral, dorsal, and lateral brainstem. Critical structures most at risk for injury during BSEZ approach included the oculomotor nerve, trochlear nerve, red nucleus, medial lemniscus, medial longitudinal fasciculus, corticospinal tract, and hypoglossal nucleus. CONCLUSION: Once thought to be universally inoperable, select lesions of the brainstem may now be treated by experienced surgeons with adjunct instrumentation, imaging, neuromonitoring, and intricate knowledge of BSEZs. All approaches adhered to the 2-point rule while minimizing neural and vascular damage. In combination with cadaveric dissection, high-resolution MRI and fiber tracking allow the surgical team to develop a better understanding of the internal architecture of the brainstem, particularly as related to BSEZs. The careful study of such imaging may lead to more accurate and safe surgery through use of optimal surgical corridors. CNS ORAL PRESENTATIONS CLINICAL NEUROSURGERY VOLUME 63 | NUMBER 1 | AUGUST 2016 | 205 Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited