Evaluating the diagnostic accuracy of 3D contrast-enhanced magnetic resonance angiography versus digital subtraction angiography in spinal dural arteriovenous fistulas.

OBJECTIVE: Spinal dural arteriovenous fistulas (SDAVFs) often go undiagnosed, leading to irreversible spinal cord dysfunction. Although digital subtraction angiography (DSA) is the gold standard for diagnosing SDAVF, DSA is invasive and operator dependent, with associated risks. MR angiography (MRA) is a promising alternative. This study aimed to evaluate the performance of MRA as an equal alternative to DSA in investigating, diagnosing, and localizing SDAVF. METHODS: Prospectively collected data from a single neurosurgeon at a large tertiary academic center were searched for SDAVFs. Eligibility criteria included any patient with a surgically proven SDAVF in whom preoperative DSA, MRA, or both had been obtained. The eligible patients formed a consecutive series, in which they were divided into DSA and MRA groups. DSA and MRA were the index tests that were compared to the surgical SDAVF outcome, which was the reference standard. Accurate diagnosis was considered to have occurred when the imaging report matched the operative diagnosis to the correct spinal level. Comparisons used a two-sample t-test for continuous variables and Fisher-Freeman-Halton's exact test for categorical variables, with p < 0.05 specifying significance. Univariate, bivariate, and multivariate analyses were conducted to investigate group associations with DSA and MRA accuracy. Positive predictive value, sensitivity, and accuracy were calculated. RESULTS: A total of 27 patients with a mean age of 63 years underwent surgery for SDAVF. There were 19 male (70.4%) and 8 female (29.6%) patients, and the mean duration of symptoms at the time of surgery was 14 months (range 2-48 months). Seventeen patients (63%) presented with bowel or bladder incontinence. Bivariate analysis of the DSA and MRA groups further revealed no significant relationships between the characteristics and accuracy of SDAVF diagnosis. MRA was found to be more sensitive and accurate (100% and 73.3%) than DSA (85.7% and 69.2%), with a subanalysis of the patients with both preoperative MRA and DSA showing that MRA had a greater positive predictive value (78.6 vs 72.7), sensitivity (100 vs 72.7), and accuracy (78.6 vs 57.1) than DSA. CONCLUSIONS: In surgically proven cases of SDAVFs, the authors determined that MRA was more accurate than DSA for SDAVF diagnosis and localization to the corresponding vertebral level. Incomplete catheterization at each vertebral level may result in the failure of DSA to detect SDAVF.

Phase 1 Safety Trial of Autologous Human Schwann Cell Transplantation in Chronic Spinal Cord Injury.

A phase 1 open-label, non-randomized clinical trial was conducted to determine feasibility and safety of autologous human Schwann cell (ahSC) transplantation accompanied by rehabilitation in participants with chronic spinal cord injury (SCI). Magnetic resonance imaging (MRI) was used to screen eligible participants to estimate an individualized volume of cell suspension to be implanted. The trial incorporated standardized multi-modal rehabilitation before and after cell delivery. Participants underwent sural nerve harvest, and ahSCs were isolated and propagated in culture. The dose of culture-expanded ahSCs injected into the chronic spinal cord lesion of each individual followed a cavity-filling volume approach. Primary outcome measures for safety and trend-toward efficacy were assessed. Two participants with American Spinal Injury Association Impairment Scale (AIS) A and two participants with incomplete chronic SCI (AIS B, C) were each enrolled in cervical and thoracic SCI cohorts (n = 8 total). All participants completed the study per protocol, and no serious adverse events related to sural nerve harvest or ahSC transplantation were reported. Urinary tract infections and skin abrasions were the most common adverse events reported. One participant experienced a 4-point improvement in motor function, a 6-point improvement in sensory function, and a 1-level improvement in neurological level of injury. Follow-up MRI in the cervical (6 months) and thoracic (24 months) cohorts revealed a reduction in cyst volume after transplantation with reduced effect over time. This phase 1 trial demonstrated the feasibility and safety of ahSC transplantation combined with a multi-modal rehabilitation protocol for participants with chronic SCI.

The Association between Whole-Brain MR Spectroscopy and IDH Mutation Status in Gliomas.

BACKGROUND AND PURPOSE: Mutations in isocitrate dehydrogenase (IDH) have a direct effect on gliomagenesis. The purpose of this study is to quantify differences in brain metabolites due to IDH mutations. METHODS: Magnetic Resonance Spectroscopic Imaging (MRSI) was performed in 35 patients with gliomas of different grade and varied IDH mutation status. Volumes of interest (VOIs) for active tumor (tVOI), peritumoral area (pVOI), and contralateral normal-appearing white matter (cVOI) were created. Metabolite ratios of Choline (Cho) to both N-acetylaspartate (NAA) and Creatine (Cr) were estimated. Ratios of Glutamate/Glutamine complex (Glx) and myoinositol (mIno) to Cr were also quantified. General linear models (GLMs) were used to estimate the effects of IDH mutation on metabolite measures, with age, gender, and tumor grade used as covariates. RESULTS: GLM analysis showed that maximum Cho/NAA and Cho/Cr in the tVOI were significantly (P < .05) higher in IDH mutant lesions as compared to wild-type. In the pVOI, mean Cho/Cr was found to be significantly different among IDH mutant and wild-type gliomas. Mean Cho/NAA (P = .306) and Cho/Cr (P = .292) within the tVOI were not significantly different. Ratios of Glx/Cr and mIno/Cr in any region showed no significant differences between IDH mutant and wild-type gliomas. No significant differences in metabolite ratios were seen in the cVOI between IDH mutants and wild-types. CONCLUSION: IDH mutation's effect in gliomas show an increase in Cho in the tumor and perilesional regions as compared to wild-type lesions but do not show widespread changes across the brain.

Imaging characteristics of chronic spinal cord injury identified during screening for a cell transplantation clinical trial.

OBJECTIVEIn cell transplantation trials for spinal cord injury (SCI), quantifiable imaging criteria that serve as inclusion criteria are important in trial design. The authors' institutional experience has demonstrated an overall high rate of screen failures. The authors examined the causes for trial exclusion in a phase I, open-lab clinical trial examining the role of autologous Schwann cell intramedullary transplantation. Specifically, they reviewed the imaging characteristics in people with chronic SCI that excluded applicants from the trial, as this was a common cause of screening failures in their study.METHODSThe authors reviewed MRI records from 152 people with chronic (> 1 year) SCI who volunteered for intralesional Schwann cell transplantation but were deemed ineligible by prospectively defined criteria. Rostral-caudal injury lesion length was measured along the long axis of the spinal cord in the sagittal plane on T2-weighted MRI. Other lesion characteristics, specifically those pertaining to lesion cavity structure resulting in trial exclusion, were recorded.RESULTSImaging records from 152 potential participants with chronic SCI were reviewed, 42 with thoracic-level SCI and 110 with cervical-level SCI. Twenty-three individuals (55%) with thoracic SCI and 70 (64%) with cervical SCI were not enrolled in the trial based on imaging characteristics. For potential participants with thoracic injuries who did not meet the screening criteria for enrollment, the average rostral-caudal sagittal lesion length was 50 mm (SD 41 mm). In applicants with cervical injuries who did not meet the screening criteria for enrollment, the average sagittal lesion length was 34 mm (SD 21 mm).CONCLUSIONSWhile screening people with SCI for participation in a cell transplantation clinical trial, lesion length or volume can exclude potential subjects who appear appropriate candidates based on neurological eligibility criteria. In planning future cell-based therapy trials, the limitations incurred by lesion size should be considered early due to the screening burden and impact on candidate selection.

Value of diffusion kurtosis imaging in assessing low-grade gliomas.

BACKGROUND: Diffusion kurtosis imaging (DKI) measures have been shown to provide increased sensitivity relative to diffusion tensor imaging (DTI) in detecting pathologies. PURPOSE: To compare the sensitivity of DKI-derived kurtosis and diffusion maps for assessment of low-grade gliomas (LGG). STUDY TYPE: Prospective study. POPULATION: In all, 19 LGG patients and 26 healthy control subjects were recruited. FIELD STRENGTH/SEQUENCE: Echo-planar-imaging diffusion-weighted MR images (b-values = 0, 1000, and 2000 with 30 diffusion gradient directions) were acquired on a 3T scanner. ASSESSMENT: Maps for mean, axial, and radial diffusivity (MD, AD, and RD) and kurtosis (MK, AK, and RK), and fractional anisotropy (FA) were evaluated in the tumor, perilesional white matter, and contralateral normal-appearing white matter regions. STATISTICAL TESTING: General linear models (GLM), Cohen's d for effect size estimates, false discovery rate (FDR) for multiple corrections, Cochran Q-test. RESULTS: Pairwise differences were observed for all diffusion and kurtosis measures between the studied regions (FDR P < 0.001), except an FA map that failed to show significant differences between the lesion and perilesional white matter (FDR P = 0.373). Effect size analysis showed that kurtosis metrics were found to be 18.8% (RK, P = 0.144) to 29.1% (AK, P < 0.05) more sensitive in discriminating perilesional regions from the lesion than corresponding diffusion metrics, whereas AK provided a 25.0% (P < 0.05) increase in sensitivity in discriminating perilesional and contralateral white matter. RK was found to be the most sensitive to contralateral white matter differences between low-grade gliomas and controls, with MK and RK providing a significantly greater sensitivity of 587.2% (P < 0.001) and 320.7% (P < 0.001) than MD and RD, respectively. DATA CONCLUSION: Kurtosis maps showed increased sensitivity, as compared to counterpart diffusion maps, for evaluation of microstructural changes in gliomas with a 3-6-fold increment in assessing changes in contralateral white matter. LEVEL OF EVIDENCE: 2 Technical Efficacy: Stage 3 J. Magn. Reson. Imaging 2018;48:1551-1558.

Safety of Autologous Human Schwann Cell Transplantation in Subacute Thoracic Spinal Cord Injury.

The rationale for implantation of autologous human Schwann cells (SCs) in persons with subacute spinal cord injury (SCI) is based on evidence that transplanted SCs are neuroprotective, support local axonal plasticity, and are capable of myelinating axons. A Phase I clinical trial was conducted to evaluate the safety of autologous human SC transplantation into the injury epicenter of six subjects with subacute SCI. The trial was an open-label, unblinded, non-randomized, non-placebo controlled study with a dose escalation design and standard medical rehabilitation. Participants were paraplegics with neurologically complete, trauma-induced spinal lesions. Autologous SCs were cultured in vitro from a sural nerve harvested from each participant and injected into the epicenter of the spinal lesion. Outcome measures for safety were protocol compliance, feasibility, adverse events, stability of neurological level, absence of detectable mass lesion, and the emergence of clinically significant neuropathic pain or muscle spasticity no greater than expected for a natural course cohort. One year post-transplantation, there were no surgical, medical, or neurological complications to indicate that the timing or procedure for the cell transplantation was unsafe. There were no adverse events or serious adverse events related to the cell therapy. There was no evidence of additional spinal cord damage, mass lesion, or syrinx formation. We conclude that it is feasible to identify eligible candidates, appropriately obtain informed consent, perform a peripheral nerve harvest to obtain SCs within 5-30 days of injury, and perform an intra-spinal transplantation of highly purified autologous SCs within 4-7 weeks of injury.

Acute presentation of brachial plexus schwannoma secondary to infarction.

Schwannomas of the brachial plexus are rare and typically present as slowly growing masses. We describe a case of a 37-year-old female who presented with acute onset of severe left upper extremity pain. Magnetic resonance imaging (MRI) showed a 2.3 × 2.1 cm peripherally enhancing centrally cystic lesion in the left axilla, along the cords of the left brachial plexus, with significant surrounding edema and enhancement. The mass was surgically removed. Pathology was consistent with a schwannoma with infarction. The pain completely resolved immediately after surgery.

Epiglottic masses identified on CT imaging: A case report and review of the broad differential diagnosis.

Epiglottic masses may be cystic, granulomatous, infectious, benign or malignant neoplastic, or manifestations of a systemic disease. When large in size, the airway may become obstructed, and when accompanied by suspicious features such as cartilaginous invasion, extension to the pre-epiglottic or para-glottic spaces, or lymphadenopathy, the radiologist must consider malignancy as a primary differential diagnosis. However, when only benign features are identified, the differential diagnosis is broad. We present a 65-year-old female with an incidental 1 cm exophytic, pedunculated, papillomatous lesion on the laryngeal surface of the epiglottis discovered upon endoscopic evaluation for dyspepsia and heartburn. Because of her risk factors for malignancy, CT scan was requested and revealed only benign features. Subsequent excisional biopsy revealed a benign squamous papilloma; however, multiple additional differential considerations were entertained preoperatively.

Posterior spinal artery aneurysm rupture after 'Ecstasy' abuse.

Posterior spinal artery (PSA) aneurysms are a rare cause of subarachnoid hemorrhage (SAH). The commonly abused street drug 3,4-methylenedioxymethamphetamine (MDMA) or 'Ecstasy' has been linked to both systemic and neurological complications. A teenager presented with neck stiffness, headaches and nausea after ingesting 'Ecstasy'. A brain CT was negative for SAH but a CT angiogram suggested cerebral vasculitis. A lumbar puncture showed SAH but a cerebral angiogram was negative. After a spinal MR angiogram identified abnormalities on the dorsal surface of the cervical spinal cord, a spinal angiogram demonstrated a left PSA 2 mm fusiform aneurysm. The patient underwent surgery and the aneurysmal portion of the PSA was excised without postoperative neurological sequelae. 'Ecstasy' can lead to neurovascular inflammation, intracranial hemorrhage, SAH and potentially even de novo aneurysm formation and subsequent rupture. PSA aneurysms may be treated by endovascular proximal vessel occlusion or open surgical excision.

Sacral morel-lavallée lesion: a not-so-rare diagnosis.

Morel-Lavallée lesions are closed soft tissue degloving injuries with a propensity to become infected, arising in the lumbosacral region or even the scalp, common anatomical locations in neuroradiological studies. The radiologist must recognize this entity, its traumatic etiology, and treatment options. Our patient's Morel-Lavallée lesion was evaluated with ultrasound and MRI, demonstrating a predominantly hemorrhagic lesion successfully managed by aspiration.

Posterior spinal artery aneurysm rupture after 'Ecstasy' abuse.

Posterior spinal artery (PSA) aneurysms are a rare cause of subarachnoid hemorrhage (SAH). The commonly abused street drug 3,4-methylenedioxymethamphetamine (MDMA) or 'Ecstasy' has been linked to both systemic and neurological complications. A teenager presented with neck stiffness, headaches and nausea after ingesting 'Ecstasy'. A brain CT was negative for SAH but a CT angiogram suggested cerebral vasculitis. A lumbar puncture showed SAH but a cerebral angiogram was negative. After a spinal MR angiogram identified abnormalities on the dorsal surface of the cervical spinal cord, a spinal angiogram demonstrated a left PSA 2 mm fusiform aneurysm. The patient underwent surgery and the aneurysmal portion of the PSA was excised without postoperative neurological sequelae. 'Ecstasy' can lead to neurovascular inflammation, intracranial hemorrhage, SAH and potentially even de novo aneurysm formation and subsequent rupture. PSA aneurysms may be treated by endovascular proximal vessel occlusion or open surgical excision.

FLAIR distal hyperintense vessels as a marker of perfusion-diffusion mismatch in acute stroke.

BACKGROUND AND PURPOSE: Distal hyperintense vessels (DHV) on MRI FLAIR sequences in acute brain ischemia are thought to represent leptomeningeal collateral flow. We hypothesized that DHV are more common in acute stroke patients with perfusion-diffusion weighted mismatch (PDM) than in those without. METHODS: We performed a retrospective study of consecutive anterior circulation stroke patients who underwent multimodal MRI within 8 hours of onset. We correlated DHV occurrence with the presence or absence of PDM, and analyzed DHV correlates when angiography was available. RESULTS: Twenty-one patients with PDM and 28 without were included. On univariate analysis, there was no significant difference regarding demographic variables between the two groups, with the exception of a higher frequency of atrial fibrillation (33% vs. 7%; P = .02) and intravenous tissue plasminogen activator use (57% vs 25%; P = .03) in the PDM patients. The PDM group more commonly had DHV (85% vs 25%; P < .001). On multivariate analysis, DHV presence (odds ratio, 6.01; 95% confidence-interval, 1.08-33.29; P = .04) and vessel occlusion site (odds ratio, 3.17; 95% confidence-interval, 1.21-8.31; P = .01) were the only variables independently associated with PDM. Conventional angiography was useful correlating DHV presence and collateral flow in a subset of patients. CONCLUSIONS: DHV may be a surrogate marker for PDM in patients with hyperacute ischemic stroke.

Spinal cord ischemia following thoracotomy without epidural anesthesia.

PURPOSE: Paraplegia is an uncommon yet devastating complication following thoracotomy, usually caused by compression or ischemia of the spinal cord. Ischemia without compression may be a result of global ischemia, vascular injury and other causes. Epidural anesthesia has been implicated as a major cause. This report highlights the fact that perioperative cord ischemia and paraplegia may be unrelated to epidural intervention. CLINICAL FEATURES: A 71-yr-old woman was admitted for a left upper lobectomy for resection of a non-small cell carcinoma of the lung. The patient refused epidural catheter placement and underwent a left T5-6 thoracotomy under general anesthesia. During surgery, she was hemodynamically stable and good oxygen saturation was maintained. Several hours following surgery the patient complained of loss of sensation in her legs. Neurological examination disclosed a complete motor and sensory block at the T5-6 level. Magnetic resonance imaging (MRI) revealed spinal cord ischemia. The patient received iv steroid treatment, but remained paraplegic. Five months following the surgery there was only partial improvement in her motor symptoms. A follow-up MRI study was consistent with a diagnosis of spinal cord ischemia. CONCLUSION: In this case of paraplegia following thoracic surgery for lung resection, epidural anesthesia/analgesia was not used. The MRI demonstrated evidence of spinal cord ischemia, and no evidence of cord compression. This case highlights that etiologies other than epidural intervention, such as injury to the spinal segmental arteries during thoracotomy, should be considered as potential causes of cord ischemia and resultant paraplegia in this surgical population.

The brachial plexus: normal anatomy, pathology, and MR imaging.

The brachial plexus is the most technically and anatomically challenging area of the peripheral nervous system for diagnostic imaging. Marked improvements in spatial and contrast resolution of plexus images have resulted from the use of phased-array technology and newer MR pulse sequence designs. This article presents case material incorporating these improvements and discusses the primary factors that continue to limit MR image quality, such as inhomogenous fat suppression, motion artifacts, and small vessels that mimic or obscure plexus components, and potential solutions and imaging alternatives. Brachial plexus anatomy and its appearance on multiplanar MR images are reviewed. The morphologic features and MR signal characteristics that have been found useful in distinguishing between normal and abnormal plexus components,and in detecting neuropathic lesions, are addressed in the context of clinical indications for plexus imaging as follows: mass involving the plexus, traumatic injury, entrapment syndrome, posttreatment evaluation, and miscellaneous conditions.

Proton MR spectroscopy of gliomatosis cerebri: case report of elevated myoinositol with normal choline levels.

A 69-year-old woman presented with clinical and imaging findings suspicious for gliomatosis cerebri, later confirmed by biopsy (moderately cellular, infiltrating glioma). Single voxel proton MR spectroscopy (TE 20 and TE 135) and spectroscopic imaging (TE 135) performed at admission showed normal choline, decreased N-acetyl, and elevated myo-inositol levels relative to creatine. The primary conclusion is that in suspected cases of gliomatosis cerebri, myo-inositol/creatine and myo-inositol/N-acetyl should be determined because they may provide evidence of tumor, even though choline/creatine is normal. A corollary to this conclusion is that choline/creatine may be misleading if used to demarcate infiltrating glioma from edema.

MR angiography of the spine: update.

The role of MRA, as an adjunct to conventional MR imaging of the spine and spinal cord, is evolving. The older MRA methods that have been applied to spinal vascular imaging include 2D and 3D phase contrast techniques and a derivative of 3D time-of-flight techniques with data acquired for about several minutes after gadolinium contrast injection (standard 3D CE MRA). Newer 3D gradient-echo techniques, which allow the acquisition of each volume of data in tens of seconds as a contrast bolus traverses the region of interest (fast 3D CE MRA), offer the possibility of temporally resolving intradural arteries and veins. The appearance of normal and abnormal intradural vessels, primarily veins, on the standard 3D CE MRA method has been described for the thoracolumbar region. Normal intradural arteries have been more difficult to detect, although preliminary results with the fast 3D CE MRA method, are promising. Only by establishing the MRA appearance of normal arteries and veins, can one begin to define "abnormal" with greater confidence (presuming that the variability in the appearance of normal vessels is not so great as to preclude differentiation). In striving for this goal, MRA has already encountered competition from CT angiography. In the characterization of spinal vascular lesions, the value of MRA has been demonstrated most convincingly for dural AVF. This lesion is more accurately localized and more sensitively detected (by neuroradiolologists and others experienced in spine imaging) with combined MR imaging and standard 3D CE MRA than with MR imaging alone. Preliminary results suggest that sensitivity and specificity may be further improved if fast 3D CE MRA is combined with conventional MR imaging. Although less well documented, the value of MRA in characterizing other lesions, such as AVMs and vascular tumors, has been reported in recent publications. In the future, the role of MRA will depend on technical advances, such as parallel acquisition techniques and possibly implantable RF coils, which permit improved detection of, and differentiation between, intradural arteries and veins. With these improvements, MRA may play an expanded role in the characterization of spinal vascular abnormalities, encompassing trauma and degenerative spine disease and vascular malformations and tumors.

MR angiography of the spine and spinal cord.

OBJECTIVES: This review has three objectives: 1) to describe spinal vascular anatomy, focusing on thoracolumbar intradural vessels detectable by both magnetic resonance angiography (MRA) and digital subtraction x-ray angiography (DSA), 2) to compare the MRA techniques that have been used to detect the major intradural vessels, and 3) to illustrate the clinical application of these MRA techniques, especially their efficacy in characterizing spinal dural arteriovenous fistulae (AVF). METHODS: MRA is an adjunct to conventional magnetic resonance imaging. MRA is usually implemented as a three-dimensional (3D) contrast-enhanced (CE) gradient-echo technique, with two approaches to data acquisition: 1) "standard" 3D CE MRA, requiring approximately 10 minutes per 3D volume, and 2) "fast" (bolus/dynamic) 3D CE MRA, requiring approximately 0.5 to 2 minutes per 3D volume depending on k-space sampling schemes. Vessels are displayed on targeted maximum intensity projection images. RESULTS: Normal intradural vessels detected on standard CE MRA are primarily veins (medullary and median), whereas both arteries and veins are detected on fast CE MRA. Identification of arteries (artery of Adamkiewicz, anterior spinal artery) is limited, and their differentiation from veins can be incomplete. Intradural vessels in patients with dural fistulae have abnormal features on MRI (length of flow voids and postcontrast serpentine enhancement) and standard 3D CE MRA (length, tortuosity, and qualitative size of dominant perimedullary vessel), which differ significantly from those of normal vessels. Standard MRA added to a conventional MRI study significantly (P=0.016) increased the rate of detection of the spinal level of a dural fistula. The correct level +/- one vertebral segment was identified in 73% of true-positive patients. CONCLUSIONS: Application of spinal MRA requires knowledge of vascular anatomy, specifically the major intradural vessels, and careful implementation of 3D CE MRA techniques. The standard technique allows for more effective noninvasive screening for vascular lesions, particularly dural AVF, than magnetic resonance imaging alone. Preliminary results indicate that the fast technique may further improve characterization of normal and abnormal intradural vessels, especially if continued technical advances yield greater temporal resolution while maintaining adequate spatial resolution.

Detection of spinal dural arteriovenous fistulae with MR imaging and contrast-enhanced MR angiography: sensitivity, specificity, and prediction of vertebral level.

BACKGROUND AND PURPOSE: MR imaging and contrast-enhanced MR angiography have been used to detect evidence of spinal dural arteriovenous fistulae (AVF); however, the sensitivity and specificity of these techniques have not been shown. The purpose of this study was to establish the sensitivity, specificity, and accuracy of MR imaging alone compared with MR imaging plus MR angiography in determining whether dural AVF are present and to establish the accuracy of MR angiography in predicting fistula level. METHODS: Twenty patients with surgically proven dural AVF (diagnosed with radiographic digital subtraction angiography) and 11 control patients who had normal digital subtraction angiography findings underwent routine MR imaging plus 3D contrast-enhanced MR angiography of the spine. Images were reviewed in two stages (stage I, MR images only; stage II, MR images plus MR angiograms) by three neuroradiologists who were blinded to the final diagnoses. RESULTS: The sensitivity, specificity, and accuracy of the three reviewers in detecting the presence of fistulae ranged from 85% to 90%, from 82% to 100%, and from 87% to 90%, respectively, for stage I, compared with values of 80% to 100%, 82%, and 81% to 94%, respectively, for stage II. For each reviewer, there was no significant difference between the values for stages I and II; however, among the reviewers, one of the more experienced neuroradiologists had significantly greater sensitivity than a less experienced neuroradiologist for stage II. On average, the percentage of true positive results for which the correct fistula level was predicted increased from 15% for stage I to 50% for stage II, and the correct level +/- one level was predicted in 73% for stage II. MR evidence of increased intradural vascularity was significantly greater in patients with dural AVF. CONCLUSION: The addition of MR angiography to standard MR imaging of the spine may improve sensitivity in the detection of spinal dural fistulae. The principal benefit of MR angiography is in the improved localization of the vertebral level of the fistula, which potentially expedites the subsequent digital subtraction angiography study.