Transarterial Onyx Embolization of Intracranial Dural Fistulas: A Prospective Cohort, Systematic Review, and Meta-Analysis

Transarterial Onyx Embolization of Intracranial Dural Fistulas: A Prospective Cohort, Systematic... Abstract BACKGROUND Onyx is important embolic material in the endovascular treatment of intracranial dural arteriovenous fistula (DAVF). However, its impact on DAVF occlusion rates, morbidity, mortality, and complication rates is not fully examined. OBJECTIVE To improve understanding of safety and effectiveness profiles associated with transarterial endovascular treatment using Onyx for intracranial DAVF METHODS We analyzed data from our prospective clinical registry and conducted a systematic review of all previous transarterial embolization studies using Onyx published between January 2005 and December 2015 in MEDLINE and EMBASE. RESULTS In the prospective study, 41 transarterial procedures were performed in 33 consecutive patients harboring 36 DAVFs. Complete initial exclusion was obtained in 32 of 36 (88.9%) fistulas; 31 fistulas were followed up showing 4 (12.9%) recurrences. Procedure-related morbidity and mortality were 3% and 0%, respectively. The literature review identified 19 studies involving a total of 425 patients with 463 DAVFs. Meta-analysis, including our registry data, showed an initial complete occlusion rate of 82% (95% confidence interval [CI]: 74%, 88%; I2, 70.6%), and recurrence rate at midterm of 2% (95% CI: 0%, 5%; I2, 21.5%). Pooled postoperative neurological deficit, procedure-related morbidity, and mortality rates were 4% (95% CI: 2%, 6%; I2, 0%), 3% (95% CI: 1%, 5%; I2, 0%), and 0%, respectively. CONCLUSION This meta-analysis suggests that transarterial embolization with Onyx is a safe treatment modality for DAVFs. Although Onyx showed a low recurrence rate at midterm, the long-term risk is poorly addressed in our study and should warrant a longer follow-up. Endovascular treatment, Embolization, Meta-analysis, Dural fistula, Onyx, Outcome ABBREVIATIONS ABBREVIATIONS BCA butyl-cyanoacrylate CI confidence interval DAVF dural arteriovenous fistula DSA digital subtraction angiography EVT endovascular treatment mRS modified Rankin score Intracranial dural arteriovenous fistula (DAVF) accounts for 10% to 15% of intracranial vascular malformations and are treated by endovascular approach in the majority of cases.1 Ethylene vinyl alcohol/dimethyl sulfoxide polymer (Onyx, EV3, Irvine, California) is an important liquid embolic material in the endovascular treatment (EVT) of intracranial DAVFs. Onyx embolic material has been extensively used due to its nonadhesive properties, allowing prolonged injection times with good lesion penetration, and convenient control for the operator.2 Since the introduction of Onyx in 2005, several single-center studies have demonstrated acceptable rates of DAVF occlusion, morbidity, and mortality for patients treated with Onyx via transarterial approach.3-21 Improved understanding of safety and effectiveness profiles associated with Onyx treatment of intracranial DAVF is needed. Thus, we analyzed data from our prospective clinical registry and conducted a systematic review and meta-analysis of the literature with special interest in intracranial DAVF occlusion rates, procedure-related complication rates, and recurrence rates at follow-up for intracranial DAVFs treated with Onyx via a transarterial approach. METHODS Prospective Study Patients This study was approved by the local ethics committee and was found to conform to scientific principles and research ethics standards. An informed consent was obtained from each patient. This study was designed, conducted, and analyzed without industry assistance, and the article was written independently of industry or any other financial support. The population was nested within a longitudinal cohort of consecutive patients who were referred to our Neurologic University Hospital for EVT of intracranial DAVFs between January 2013 and December 2015. This prospectively maintained database was queried retrospectively to identify all consecutive patients matching the following inclusion criteria: (a) adult patients with intracranial DAVFs (b) who were treated by transarterial approach using Onyx (EV3) material. EVT and Initial Angiographic Results All procedures were performed under general anesthesia and full heparinization (60 IU/kg). Complete selective digital subtraction angiography (DSA) was performed before treatment. An Onyx-compatible microcatheter was coaxially positioned into one of major arterial feeder as close as possible to the fistula site. First, the dead space of the microcatheter was flushed with dimethyl-sulfoxide, followed by a slow injection of Onyx under subtraction fluoroscopy. If an Onyx reflux occurred, the injection was paused for 30 to 60 s. During Onyx progression into the shunt, it was continually and slowly injected. The removal of the microcatheter was done when the Onyx cast stopped advancing or if we had a complete occlusion of the shunt. A postoperative DSA was performed after the procedure to confirm the complete occlusion of the DAVF (Figure 1). Initial angiographic results of EVT were classified by a neuroradiologist not involved in the initial treatment (B.G. with 7 yr of experience in interventional neuroradiology) by using the Cognard classification.22 FIGURE 1. View largeDownload slide Demonstration of a case treated with transarterial embolization with Onyx (EV3). A and B, Pretreatment lateral and anterior angiograms after external carotid artery injection show a Cognard type II a+b DAVF centered along the left lateral sinus with arterial supply from branches of the middle meningeal and occipital arteries. There is an important cortical venous reflux, especially into the Labbé vein. C and D, The microcatheter is navigated through the parietal branch of the left middle meningeal artery near the fistula. E and F, The cast of Onyx is seen to extend from middle meningeal artery, occipital artery, fistula, and draining vein. G and H, Post-treatment lateral and anterior angiograms after primitive carotid artery injection show a complete occlusion of the fistula without cortical venous drainage. FIGURE 1. View largeDownload slide Demonstration of a case treated with transarterial embolization with Onyx (EV3). A and B, Pretreatment lateral and anterior angiograms after external carotid artery injection show a Cognard type II a+b DAVF centered along the left lateral sinus with arterial supply from branches of the middle meningeal and occipital arteries. There is an important cortical venous reflux, especially into the Labbé vein. C and D, The microcatheter is navigated through the parietal branch of the left middle meningeal artery near the fistula. E and F, The cast of Onyx is seen to extend from middle meningeal artery, occipital artery, fistula, and draining vein. G and H, Post-treatment lateral and anterior angiograms after primitive carotid artery injection show a complete occlusion of the fistula without cortical venous drainage. Systematic Standard Follow-up Protocol After discharge, the systematic follow-up included at least a clinical examination and DSA at 3 mo after EVT for ruptured DAVFs and at 6 mo for unruptured DAVFs. Clinical Follow-up Further, clinical follow-up data were collected during hospitalization for follow-up DSA or external consultation at 3 mo. Postoperative neurological complication was defined as any new neurological symptoms, including cranial nerve palsy. Procedure-related morbidity was defined as a permanent neurological deficit including cranial nerve palsy or change in modified Rankin Scale (mRS) score ≥ 1 at 3 mo after the procedure. Image Acquisition and Analysis Angiographic images were acquired in anteroposterior and lateral projections before and immediately after treatment. Angiographic images obtained immediately after EVT were compared with those obtained at angiographic follow-up. At follow-up, we considered a DAVF recurrence when an early venous opacification was observed at DAVF completely occluded regardless of how big it is. In addition, we also considered it as a recurrence when the type of DAVF was modified. Systematic Review We prepared this study in accordance with the Meta-analysis of Observational Studies in Epidemiology and Preferred Reporting Items for Systematic Reviews and Meta-analyses guidelines, including objectives and plans for collecting and analyzing the data.23,24 We performed a systematic review of the literature by 2 authors (U.S.G. and N.M.) using the keywords “dural arteriovenous fistula” and “onyx” and “cerebral” or “brain” or “intracranial” in MEDLINE and EMBASE databases. Inclusion criteria were the following: all types of study design published in the English language, ≥10 patients, from January 2005 to December 2015, where EVT was performed via a transarterial route and where Onyx was used as the embolic material. Exclusion criteria were the following: case report, studies with duplicate case series, studies with balloon-assisted technique. Baseline Characteristics of Patients and DAVFs Clinical data included sex, number of eligible patients, mean age of patients, and clinical presentation. DAVFs characteristics, procedure-related complications, morbidity and mortality, anatomic outcomes, and follow-up modality. All the data were reviewed and collected individually by 2 authors (S.G.U. and M.N.) and compared. Disagreements were discussed until a consensus was reached. Outcomes Only patients treated with Onyx alone were considered for the calculation of the postoperative complications, morbidity, and mortality rates after treatment and DAVF recurrence rates. End points included postoperative neurological deficit after EVT, including cranial nerve palsy, and anatomic results obtained at DSA or magnetic resonance angiography. Postoperative neurological complication was differentiated as ischemic or hemorrhagic. Individual Study Quality Assessment The included studies being uncontrolled, individual study quality was assessed using a checklist published by the National Institutes of Health for before–after (pre–post) studies with no control group.25 This 12-item checklist enables an assessment of uncontrolled studies and provides an overall quality rating. Quality assessment was performed independently by 2 authors (S.G.U. and M.N.). Statistical Analysis Systematic Review The 95% confidence intervals (CIs) of the estimates were built with the Wilson method. The estimate and the 95% CI of the mean percentage over all the studies were obtained for each outcome using a logistic mixed model with a random effect on the intercept in order to take into account the heterogeneity between the studies. In all analyses, inconsistency of findings throughout studies was assessed by using the P value and the I2 statistic. We searched for publication bias using Egger's test for small-study effects and presented funnel plots. The meta-analyses were carried out using the metaprop Stata command on Stata/SE 14.1 (Statacorp LP, College Station, Texas).26 RESULTS Prospective Study Population A total of 41 procedures were performed in 33 patients (21 men and 12 women, mean age, 57.4 yr) harboring 36 DAVFs (3 patients had 2 DAVFs). Table 1 provides baseline patient and DAVF characteristics. Initial clinical symptoms were 10 intracranial hemorrhages, 12 tinnitus, and 2 seizures, while the rest of the patients presented with incidentally discovered DAVFs. TABLE 1. Neurological Hospital Baseline Characteristics of Patients and Dural Fistulas Patients  Sex  Age  DAVF Classification22  Symptoms  Access artery  Initial complete occlusionb  Complication  Procedure-related morbidity  Recurrenceb  1  M  65  III  Nonspecific  MMA  Yes  No  0  NA  1a  M  65  IV  Hemorrhage  MMA  Yes  No  0  NA  2  F  39  II a  Tinnitus  MMA  Partial (I)  No  0  Recurrence (II b)  3  F  43  IV  Nonspecific  MMA  Yes  CNP (III)  0  NA  4  M  75  III  Hemorrhage  MMA  Partial (I)  No  0  Partial (I)  5  F  64  I  Tinnitus  MMA  Yes  No  0  NA  6  M  62  I  Tinnitus  MMA  Yes  No  0  NA  7  M  70  IV  Hemorrhage  APA  Yes  No  0  No  8  M  71  II b  Nonspecific  MMA  Yes  No  0  No  9  F  38  III  Hemorrhage  MMA  Yes  No  0  No  10  F  59  II b  Tinnitus  MMA  Yes  No  0  No  11  M  50  II b  Nonspecific  MMA  Yes  No  0  No  12  M  46  II a  Tinnitus  MMA  Yes  No  0  No  13  F  78  IV  Hemorrhage  MMA  Partial (IV)  No  0  Noc  14  M  66  I  Tinnitus  MMA  Yes  No  0  No  15  F  56  II b  Tinnitus  MMA  Partial (II a)  Ischemia  Yes  Partial (II a)  15a  F  56  II a+b  Tinnitus  MMA  Yes  CNP (VII)  0  Recurrence (II a)  16  F  45  III  Hemorrhage  PMA  Yes  No  0  No  17  M  57  II b  Nonspecific  MMA  Yes  No  0  Recurrence (II b)  17a  M  57  II a+b  Nonspecific  MMA  Yes  No  0  No  18  M  55  I  Tinnitus  MMA  Yes  Microcatheter fracture  0  No  19  M  54  IV  Seizure  MMA  Yes  No  0  No  20  F  47  II b  Tinnitus  MMA  Yes  No  0  No  21  F  42  I  Tinnitus  MMA  Yes  No  0  No  22  F  61  III  Hemorrhage  MMA  Yes  No  0  No  23  M  64  I  Tinnitus  MMA  Yes  No  0  Recurrence (I)  24  M  68  IV  Nonspecific  OA  Yes  No  0  No  25  M  44  II b  Nonspecific  SCA  Yes  No  0  No  26  M  65  IV  Nonspecific  OA  Yes  No  0  No  27  M  50  III  Seizure  MMA  Yes  No  0  No  28  M  81  II b  Hemorrhage  MMA  Yes  No  0  No  29  M  73  II a  Nonspecific  OccA  Yes  No  0  No  30  F  48  I  Tinnitus  MMA  Yes  No  0  No  31  M  51  III  Hemorrhage  MMA  Yes  No  0  No  32  M  55  III  Nonspecific  MMA  Yes  No  0  No  33  M  52  IV  Hemorrhage  MMA  Yes  No  0  No  Patients  Sex  Age  DAVF Classification22  Symptoms  Access artery  Initial complete occlusionb  Complication  Procedure-related morbidity  Recurrenceb  1  M  65  III  Nonspecific  MMA  Yes  No  0  NA  1a  M  65  IV  Hemorrhage  MMA  Yes  No  0  NA  2  F  39  II a  Tinnitus  MMA  Partial (I)  No  0  Recurrence (II b)  3  F  43  IV  Nonspecific  MMA  Yes  CNP (III)  0  NA  4  M  75  III  Hemorrhage  MMA  Partial (I)  No  0  Partial (I)  5  F  64  I  Tinnitus  MMA  Yes  No  0  NA  6  M  62  I  Tinnitus  MMA  Yes  No  0  NA  7  M  70  IV  Hemorrhage  APA  Yes  No  0  No  8  M  71  II b  Nonspecific  MMA  Yes  No  0  No  9  F  38  III  Hemorrhage  MMA  Yes  No  0  No  10  F  59  II b  Tinnitus  MMA  Yes  No  0  No  11  M  50  II b  Nonspecific  MMA  Yes  No  0  No  12  M  46  II a  Tinnitus  MMA  Yes  No  0  No  13  F  78  IV  Hemorrhage  MMA  Partial (IV)  No  0  Noc  14  M  66  I  Tinnitus  MMA  Yes  No  0  No  15  F  56  II b  Tinnitus  MMA  Partial (II a)  Ischemia  Yes  Partial (II a)  15a  F  56  II a+b  Tinnitus  MMA  Yes  CNP (VII)  0  Recurrence (II a)  16  F  45  III  Hemorrhage  PMA  Yes  No  0  No  17  M  57  II b  Nonspecific  MMA  Yes  No  0  Recurrence (II b)  17a  M  57  II a+b  Nonspecific  MMA  Yes  No  0  No  18  M  55  I  Tinnitus  MMA  Yes  Microcatheter fracture  0  No  19  M  54  IV  Seizure  MMA  Yes  No  0  No  20  F  47  II b  Tinnitus  MMA  Yes  No  0  No  21  F  42  I  Tinnitus  MMA  Yes  No  0  No  22  F  61  III  Hemorrhage  MMA  Yes  No  0  No  23  M  64  I  Tinnitus  MMA  Yes  No  0  Recurrence (I)  24  M  68  IV  Nonspecific  OA  Yes  No  0  No  25  M  44  II b  Nonspecific  SCA  Yes  No  0  No  26  M  65  IV  Nonspecific  OA  Yes  No  0  No  27  M  50  III  Seizure  MMA  Yes  No  0  No  28  M  81  II b  Hemorrhage  MMA  Yes  No  0  No  29  M  73  II a  Nonspecific  OccA  Yes  No  0  No  30  F  48  I  Tinnitus  MMA  Yes  No  0  No  31  M  51  III  Hemorrhage  MMA  Yes  No  0  No  32  M  55  III  Nonspecific  MMA  Yes  No  0  No  33  M  52  IV  Hemorrhage  MMA  Yes  No  0  No  Note: M, male; F, female; NA, nonassessed; MMA, middle meningeal artery; OA, ophtalmic artery; OccA, occipital artery; PMA, posterior meningeal artery; SCA, superior cerebellar artery; APA, ascending pharengeal artery; CNP, cranial nerve palsy. aPatients 1, 15, and 17 had 2 DAVFs. bData in parentheses is the type of DAVF according Cognard classification.22 cA complementary surgery was performed after initial endovascular treatment failure. View Large TABLE 1. Neurological Hospital Baseline Characteristics of Patients and Dural Fistulas Patients  Sex  Age  DAVF Classification22  Symptoms  Access artery  Initial complete occlusionb  Complication  Procedure-related morbidity  Recurrenceb  1  M  65  III  Nonspecific  MMA  Yes  No  0  NA  1a  M  65  IV  Hemorrhage  MMA  Yes  No  0  NA  2  F  39  II a  Tinnitus  MMA  Partial (I)  No  0  Recurrence (II b)  3  F  43  IV  Nonspecific  MMA  Yes  CNP (III)  0  NA  4  M  75  III  Hemorrhage  MMA  Partial (I)  No  0  Partial (I)  5  F  64  I  Tinnitus  MMA  Yes  No  0  NA  6  M  62  I  Tinnitus  MMA  Yes  No  0  NA  7  M  70  IV  Hemorrhage  APA  Yes  No  0  No  8  M  71  II b  Nonspecific  MMA  Yes  No  0  No  9  F  38  III  Hemorrhage  MMA  Yes  No  0  No  10  F  59  II b  Tinnitus  MMA  Yes  No  0  No  11  M  50  II b  Nonspecific  MMA  Yes  No  0  No  12  M  46  II a  Tinnitus  MMA  Yes  No  0  No  13  F  78  IV  Hemorrhage  MMA  Partial (IV)  No  0  Noc  14  M  66  I  Tinnitus  MMA  Yes  No  0  No  15  F  56  II b  Tinnitus  MMA  Partial (II a)  Ischemia  Yes  Partial (II a)  15a  F  56  II a+b  Tinnitus  MMA  Yes  CNP (VII)  0  Recurrence (II a)  16  F  45  III  Hemorrhage  PMA  Yes  No  0  No  17  M  57  II b  Nonspecific  MMA  Yes  No  0  Recurrence (II b)  17a  M  57  II a+b  Nonspecific  MMA  Yes  No  0  No  18  M  55  I  Tinnitus  MMA  Yes  Microcatheter fracture  0  No  19  M  54  IV  Seizure  MMA  Yes  No  0  No  20  F  47  II b  Tinnitus  MMA  Yes  No  0  No  21  F  42  I  Tinnitus  MMA  Yes  No  0  No  22  F  61  III  Hemorrhage  MMA  Yes  No  0  No  23  M  64  I  Tinnitus  MMA  Yes  No  0  Recurrence (I)  24  M  68  IV  Nonspecific  OA  Yes  No  0  No  25  M  44  II b  Nonspecific  SCA  Yes  No  0  No  26  M  65  IV  Nonspecific  OA  Yes  No  0  No  27  M  50  III  Seizure  MMA  Yes  No  0  No  28  M  81  II b  Hemorrhage  MMA  Yes  No  0  No  29  M  73  II a  Nonspecific  OccA  Yes  No  0  No  30  F  48  I  Tinnitus  MMA  Yes  No  0  No  31  M  51  III  Hemorrhage  MMA  Yes  No  0  No  32  M  55  III  Nonspecific  MMA  Yes  No  0  No  33  M  52  IV  Hemorrhage  MMA  Yes  No  0  No  Patients  Sex  Age  DAVF Classification22  Symptoms  Access artery  Initial complete occlusionb  Complication  Procedure-related morbidity  Recurrenceb  1  M  65  III  Nonspecific  MMA  Yes  No  0  NA  1a  M  65  IV  Hemorrhage  MMA  Yes  No  0  NA  2  F  39  II a  Tinnitus  MMA  Partial (I)  No  0  Recurrence (II b)  3  F  43  IV  Nonspecific  MMA  Yes  CNP (III)  0  NA  4  M  75  III  Hemorrhage  MMA  Partial (I)  No  0  Partial (I)  5  F  64  I  Tinnitus  MMA  Yes  No  0  NA  6  M  62  I  Tinnitus  MMA  Yes  No  0  NA  7  M  70  IV  Hemorrhage  APA  Yes  No  0  No  8  M  71  II b  Nonspecific  MMA  Yes  No  0  No  9  F  38  III  Hemorrhage  MMA  Yes  No  0  No  10  F  59  II b  Tinnitus  MMA  Yes  No  0  No  11  M  50  II b  Nonspecific  MMA  Yes  No  0  No  12  M  46  II a  Tinnitus  MMA  Yes  No  0  No  13  F  78  IV  Hemorrhage  MMA  Partial (IV)  No  0  Noc  14  M  66  I  Tinnitus  MMA  Yes  No  0  No  15  F  56  II b  Tinnitus  MMA  Partial (II a)  Ischemia  Yes  Partial (II a)  15a  F  56  II a+b  Tinnitus  MMA  Yes  CNP (VII)  0  Recurrence (II a)  16  F  45  III  Hemorrhage  PMA  Yes  No  0  No  17  M  57  II b  Nonspecific  MMA  Yes  No  0  Recurrence (II b)  17a  M  57  II a+b  Nonspecific  MMA  Yes  No  0  No  18  M  55  I  Tinnitus  MMA  Yes  Microcatheter fracture  0  No  19  M  54  IV  Seizure  MMA  Yes  No  0  No  20  F  47  II b  Tinnitus  MMA  Yes  No  0  No  21  F  42  I  Tinnitus  MMA  Yes  No  0  No  22  F  61  III  Hemorrhage  MMA  Yes  No  0  No  23  M  64  I  Tinnitus  MMA  Yes  No  0  Recurrence (I)  24  M  68  IV  Nonspecific  OA  Yes  No  0  No  25  M  44  II b  Nonspecific  SCA  Yes  No  0  No  26  M  65  IV  Nonspecific  OA  Yes  No  0  No  27  M  50  III  Seizure  MMA  Yes  No  0  No  28  M  81  II b  Hemorrhage  MMA  Yes  No  0  No  29  M  73  II a  Nonspecific  OccA  Yes  No  0  No  30  F  48  I  Tinnitus  MMA  Yes  No  0  No  31  M  51  III  Hemorrhage  MMA  Yes  No  0  No  32  M  55  III  Nonspecific  MMA  Yes  No  0  No  33  M  52  IV  Hemorrhage  MMA  Yes  No  0  No  Note: M, male; F, female; NA, nonassessed; MMA, middle meningeal artery; OA, ophtalmic artery; OccA, occipital artery; PMA, posterior meningeal artery; SCA, superior cerebellar artery; APA, ascending pharengeal artery; CNP, cranial nerve palsy. aPatients 1, 15, and 17 had 2 DAVFs. bData in parentheses is the type of DAVF according Cognard classification.22 cA complementary surgery was performed after initial endovascular treatment failure. View Large Initial Angiographic Results According to Cognard's classification, 7 cases were type I, 3 cases were type II a, 8 cases were type II b, 2 cases were type II a+b, 8 cases were type III, and 8 cases were type IV. Initial complete exclusion of the DAVFs was obtained in 32 of 36 (88.9%) DAVFs. DAVF obliteration was achieved in 2 procedures for 5 patients and in 1 procedure for the other cases. Of the 4 partial EVTs, the type of DAVF was initially 1 II a (patient 2), 1 II b (patient 15), 1 III (patient 4), and 1 IV (patient 13). After EVT, the type II b reverted to type II a, and the type III to a type I. For the type IV, the occlusion after EVT was partial because of high risk of facial nerve palsy due to important microcatheter reflux. Thus, considering these 4 patients, 2 procedures were partial due to the risk of nerve palsy or reversion into benign type. Angiographic Results at Follow-up Angiography was obtained in 29 patients harboring 31 DAVFs. A recurrence was observed in 4 DAVFs (12.9%). The 10 initially ruptured DAVF were totally occluded at follow-up. A retreatment was performed in 2 DAVFs (6.5%) due to the aggressive type (only via transarterial approach). For the patients with initial partial obliteration, 2 patients showed similar DAVF type (patients 4 and 15), 1 patient presented complete fistula exclusion at follow-up (patient 13; a complementary surgery was performed after EVT failure), and 1 patient presented a type II b (patient 2). Clinical Complications A total of 3 patients had postoperative neurological complication following EVT. Two patients presented a cranial nerve palsy: 1 had a third cranial nerve palsy after treatment of DAVF type IV located at the anterior temporal lobe with subsequent transitory diplopia and ophthalmoplegia, which completely resolved 3 mo later; and 1 patient had a facial nerve palsy that quickly improved during the first few days, but partially persisted at 3-mo follow-up after treatment of a right lateral sinus fistula type II a+b. After treatment of her second DAVF type II b located in the superior longitudinal sinus, the patient (patient 15) presented with a left hemiplegia due to venous infarction, persistent at follow-up (mRS of 2). Procedure-Related Morbidity and Mortality In our prospective cohort, the procedure-related morbidity was 3% (1/33 patients). There was no procedure-related mortality. One patient died due to a compressive hemorrhage in the posterior fossa due to a type III ruptured DAVF. The hematoma was located in the vermis, causing an obstructive hydrocephalus with subsequent ventricular shunt placement. Emergency embolization was performed without complication during procedure, but 1 mo later the patient died from complications of the intracranial hematoma. Systematic Review Of the 424 records that were identified in the initial search, 309 were screened after removal of duplicates, and 280 were excluded at title or abstract level. Twenty-nine were articles that were reviewed at full-text out of which 19 studies were selected for final analysis (Figure 2). At baseline, 425 patients harboring 463 DAVFs were included. Baseline patient and DAVF characteristics of the 19 included studies are shown in Table, Supplemental Digital Content 1. Outcomes of the 19 studies are shown in Table, Supplemental Digital Content 2. Table 2 summarizes the results of the meta-analysis of the safety and effectiveness of EVT with Onyx (EV3). FIGURE 2. View largeDownload slide Flowchart shows screening and selection of studies for meta-analysis. FIGURE 2. View largeDownload slide Flowchart shows screening and selection of studies for meta-analysis. TABLE 2. Meta-analysis of the Safety and Effectiveness after Endovascular Treatment with Onyx (EV3) Outcomes  Number of studies  Sample size  Pooled rates (95% CI)  Heterogeneity  P value  I2  Cranial nerve palsy (n = 13)  16  466a  2% (1%, 4%)  .96  0.0%  Postoperative neurological complications (n = 22)  16  466a  4% (2%, 6%)  .94  0.0%  Procedure-related morbidity (n = 17)  16  366b  3% (1%, 5%)  .53  0.0%  Procedure-related mortality (n = 0)  17  366b  0% (0%, 0%)  1  0.0%  Initial complete obliteration (n = 373)  19  463c  82% (74%, 88%)  <.05  70.6%  Recurrence (n = 5)  13  263c  2% (0%, 5%)  .23  21.5%  Outcomes  Number of studies  Sample size  Pooled rates (95% CI)  Heterogeneity  P value  I2  Cranial nerve palsy (n = 13)  16  466a  2% (1%, 4%)  .96  0.0%  Postoperative neurological complications (n = 22)  16  466a  4% (2%, 6%)  .94  0.0%  Procedure-related morbidity (n = 17)  16  366b  3% (1%, 5%)  .53  0.0%  Procedure-related mortality (n = 0)  17  366b  0% (0%, 0%)  1  0.0%  Initial complete obliteration (n = 373)  19  463c  82% (74%, 88%)  <.05  70.6%  Recurrence (n = 5)  13  263c  2% (0%, 5%)  .23  21.5%  aNumber of procedures. bNumber of patients. cNumber of DAVFs. View Large TABLE 2. Meta-analysis of the Safety and Effectiveness after Endovascular Treatment with Onyx (EV3) Outcomes  Number of studies  Sample size  Pooled rates (95% CI)  Heterogeneity  P value  I2  Cranial nerve palsy (n = 13)  16  466a  2% (1%, 4%)  .96  0.0%  Postoperative neurological complications (n = 22)  16  466a  4% (2%, 6%)  .94  0.0%  Procedure-related morbidity (n = 17)  16  366b  3% (1%, 5%)  .53  0.0%  Procedure-related mortality (n = 0)  17  366b  0% (0%, 0%)  1  0.0%  Initial complete obliteration (n = 373)  19  463c  82% (74%, 88%)  <.05  70.6%  Recurrence (n = 5)  13  263c  2% (0%, 5%)  .23  21.5%  Outcomes  Number of studies  Sample size  Pooled rates (95% CI)  Heterogeneity  P value  I2  Cranial nerve palsy (n = 13)  16  466a  2% (1%, 4%)  .96  0.0%  Postoperative neurological complications (n = 22)  16  466a  4% (2%, 6%)  .94  0.0%  Procedure-related morbidity (n = 17)  16  366b  3% (1%, 5%)  .53  0.0%  Procedure-related mortality (n = 0)  17  366b  0% (0%, 0%)  1  0.0%  Initial complete obliteration (n = 373)  19  463c  82% (74%, 88%)  <.05  70.6%  Recurrence (n = 5)  13  263c  2% (0%, 5%)  .23  21.5%  aNumber of procedures. bNumber of patients. cNumber of DAVFs. View Large Initial Angiographic Results The overall initial complete occlusion rate was 82% (95% CI, 74%, 88%; Figure 3). Analysis of the data suggested significant heterogeneity across studies (P < .05), the range of initial complete occlusion rate being 47% to 100%. FIGURE 3. View largeDownload slide Crude odds ratios. A, Initial complete occlusion. B, Recurrence at midterm. FIGURE 3. View largeDownload slide Crude odds ratios. A, Initial complete occlusion. B, Recurrence at midterm. Angiographic Results at Follow-up The recurrence rate was 2% (95% CI, 0%, 5%; Figure 3). Recurrence rate was reported at midterm (mean follow-up, 5 mo; range, 3-7.5 mo). Long-term DAVFs recurrences rates were reported in 2 studies comprising a small sample size of 45 patients. Clinical Complications The pooled postoperative neurological complications rate was 4% (95% CI: 2%, 6%; I2, 0%; Figure 4). The pooled rate of postoperative cranial nerve palsy was 2% (95% CI: 1%, 4%; I2, 0%; Figure 4). The rate of cerebral ischemic and hemorrhagic complications rates were 1% (95% CI: 0%, 2%, I2, 0%) and 0% (95% CI: 0%, I2, 0%), respectively (Figure, Supplemental Digital Content 3). FIGURE 4. View largeDownload slide Crude odds ratio. A, Postoperative neurological complications. B, Cranial nerve palsy. C, Procedure-related morbidity. FIGURE 4. View largeDownload slide Crude odds ratio. A, Postoperative neurological complications. B, Cranial nerve palsy. C, Procedure-related morbidity. Procedure-Related Morbidity and Mortality The pooled procedure-related morbidity rate was 3% (95% CI, 1%, 5%; I2, 0%; Figure 4) with no statistically significant heterogeneity across studies. There were no procedure-related deaths across the 17 studies. Quality Assessment Eighteen studies were retrospective whereas 1 was prospective. All were noncomparative. Studies had large heterogeneity in terms of methods for the assessment of outcomes (presence or not of an adjudication committee; presence or not of a centralized core laboratory; time of follow-up). Using the prespecified tool, the quality rating of studies was considered as fair or poor. The main limitations of studies were as follows: no prespecification of selection criteria for the study population; no justification of sample size; no independent assessment of outcome measures across all study participants. Consequently, the risk of bias was significant across studies. Furthermore, we identified potential publication bias on the rate of total complications (Figure, Supplemental Digital Content 3; Figure, Supplemental Digital Content 4 for funnel plots). DISCUSSION Based on 19 studies involving a total of 1425 patients with 463 DAVFs, our meta-analysis provides more representative results on DAVF occlusion rates, morbidity, mortality, and complication rates than any single study. The rate of recurrence at midterm follow-up was low after EVT using Onyx (EV3) of intracranial DAVFs (2%, 95% CI: 0%, 5%) in our meta-analysis including our registry data. Although this finding demonstrates the effectiveness of Onyx in the EVT of cranial DAFVs, we observed a nonnegligible rate of DAVF angiographic recurrence (12.9%) at 3to 6 mo in our prospective study. Results at follow-up of intracranial DAVFs beyond 1 yr after EVT are not well known. Longer follow-up periods were reported in 2 studies including only a small number of patients (45 patients). Chandra et al9 observed no recurrence of 28 patients at 28 mo mean follow-up, whereas Ambekar et al4 reported 3 (14.3%) recurrences of 21 patients at 14 mo mean follow-up.4 This underlines the importance of long-term follow-up for DAVFs, especially for initially ruptured ones with the risk of rebleeding. However, we used DSA, which is the gold standard modality for follow-up of Onyx-treated DAVFs in our study and the series included in the systematic review. To date, there are several embolic materials to treat DAVFs by endovascular approach including n-butyl-cyanoacrylate (n-BCA), Onyx, polyvinyl alcohol particles, and coils. To date, the best embolic material for cranial DAVFs is not well known and comparative studies with good methodological standard are warranted, which is not the objective in the present study. However, Rabinov et al20 compared the effectiveness of cranial DAVF EVT with Onyx vs n-BCA in single-center study. Although the sample size was limited (56 fistulas), the initial complete occlusion rate reported for Onyx was 82% vs 33.3% for n-BCA. A superior durability of the occlusion with Onyx on follow-up was observed. In a recent single-center series of 24 fistulas, the authors compared intracranial DAVFs embolization with Onyx vs n-BCA and coils, and they reported initial complete occlusion rate of 66% for Onyx vs 22% for n-BCA.27 A possible explanation is that DAVFs are complex and heterogenic lesions, with considerable anatomopathological diversity, influencing the difficulty of the access and EVT phase. Furthermore, heterogeneity in operator experience may be another explanation. Procedure-related morbidity and mortality rates were uniformly low across the studies with pooled rates of 3% (95% CI, 1%, 5%) and 0%, respectively. This meta-analysis, including our registry data, demonstrates the safety of EVT of intracranial DAVFs with Onyx, most morbidity events related to cranial nerve palsy (2%; 95% CI: 1%, 4%). EVT via transarterial approach of lesions close to the skull base, as cranial DAVFs, carries an elevated risk for ischemic nerve injury.1 As Onyx embolization technique usually utilizes some degree of Onyx reflux, adequate safety margins should be considered appropriately to minimize inadvertent Onyx migration to clinically important vascular branches. Limitations Our study had several limitations. First, the articles included in the systematic review contained a majority of retrospective studies with a limited number of patients; some included combined liquid embolic materials. Second, a small number of studies (2 of 20 studies, 45 of 465 patients, 19.6%) with long-term follow-up were eligible. Third, data presentation was not uniform among the source articles, especially for the procedure-related morbidity definition. Fourth, it is possible that some relevant studies were not taken into account in our systematic review. However, it is unlikely that this potential publication bias strongly distorted our findings because we found no evidence of such bias by examining the funnel plots. CONCLUSION EVT of intracranial DAVFs with Onyx (EV3) via transarterial approach is a safe treatment modality. Although Onyx showed a low recurrence rate at mean follow-up of 5 mo, the risk of long-term recurrence is poorly evaluated in our study, and should warrant a longer follow-up period, especially in ruptured cases and neuroaggressive ones. Disclosure The authors have no personal, financial, or institutional interest in any of the drugs, materials, or devices described in this article. REFERENCES 1. Miller TR, Gandhi D. Intracranial dural arteriovenous fistulae: clinical presentation and management strategies. Stroke . 2015; 46( 7): 2017- 2025. Google Scholar CrossRef Search ADS PubMed  2. Sadeh-Gonik U, Gory B, Riva R et al.   Ethylene vinyl alcohol copolymer (Onyx) embolization of cranial dural arteriovenous fistula via the ascending pharyngeal artery. Diagn Interv Imaging . 2016; 97( 6): 681- 685. Google Scholar CrossRef Search ADS PubMed  3. Abud TG, Nguyen A, Saint-Maurice JP et al.   The use of Onyx in different types of intracranial dural arteriovenous fistula. AJNR Am J Neuroradiol . 2011; 32( 11): 2185- 2191. Google Scholar CrossRef Search ADS PubMed  4. Ambekar S, Gaynor BG, Peterson EC, Elhammady MS. Long-term angiographic results of endovascularly “cured” intracranial dural arteriovenous fistulas. J Neurosurg . 2016; 124( 4): 1123- 7. Google Scholar CrossRef Search ADS PubMed  5. Cha KC, Yeon JY, Kim GH, Jeon P, Kim JS, Hong SC. Clinical and angiographic results of patients with dural arteriovenous fistula. J Clin Neurosci . 2013; 20( 4): 536- 542. Google Scholar CrossRef Search ADS PubMed  6. Chew J, Weill A, Guilbert F, Raymond J, Audet ME, Roy D. Arterial onyx embolisation of intracranial DAVFs with cortical venous drainage. Can J Neurol Sci . 2009; 36( 2): 168- 175. Google Scholar CrossRef Search ADS PubMed  7. Cognard C, Januel AC, Silva NA Jr, Tall P. Endovascular treatment of intracranial dural arteriovenous fistulas with cortical venous drainage: new management using Onyx. AJNR Am J Neuroradiol . 2008; 29( 2): 235- 241. Google Scholar CrossRef Search ADS PubMed  8. De Keukeleire K, Vanlangenhove P, Kalala Okito JP, Hallaert G, Van Roost D, Defreyne L. Transarterial embolization with Onyx for treatment of intracranial non-cavernous dural arteriovenous fistula with or without cortical venous reflux. J Neurointerv Surg . 2011; 3( 3): 224- 228. Google Scholar CrossRef Search ADS PubMed  9. Chandra RV, Leslie-Mazwi TM, Mehta BP et al.   Transarterial onyx embolization of cranial dural arteriovenous fistulas: Long-term follow-up. AJNR Am J Neuroradiol . 2014; 35( 9): 1793- 1797. Google Scholar CrossRef Search ADS PubMed  10. Ghobrial GM, Marchan E, Nair AK et al.   Dural arteriovenous fistulas: a review of the literature and a presentation of a single institution's experience. World Neurosurg . 2013; 80( 1-2): 94- 102. Google Scholar CrossRef Search ADS PubMed  11. Hu YC, Newman CB, Dashti SR, Albuquerque FC, McDougall CG. Cranial dural arteriovenous fistula: transarterial Onyx embolization experience and technical nuances. J Neurointerv Surg . 2011; 3( 1): 5- 13. Google Scholar CrossRef Search ADS PubMed  12. Huang Q, Xu Y, Hong B, Li Q, Zhao W, Liu J. Use of Onyx in the management of tentorial dural arteriovenous fistulae. Neurosurgery . 2009; 65( 2): 287- 293. Google Scholar CrossRef Search ADS PubMed  13. Long XA, Karuna T, Zhang X, Luo B, Duan CZ. Onyx 18 embolisation of dural arteriovenous fistula via arterial and venous pathways: preliminary experience and evaluation of the short-term outcomes. Br J Radiol . 2012; 85( 1016): 395- 403. Google Scholar CrossRef Search ADS PubMed  14. Luo CB, Chang FC, Mu-Huo Teng M et al.   Transarterial Onyx embolization of intracranial dural arteriovenous fistulas: a single center experience. J Chin Med Assoc . 2014; 77( 4): 184- 189. Google Scholar CrossRef Search ADS PubMed  15. Lv X, Jiang C, Zhang J, Li Y, Wu Z. Complications related to percutaneous transarterial embolization of intracranial dural arteriovenous fistulas in 40 patients. AJNR Am J Neuroradiol . 2009; 30( 3): 462- 468. Google Scholar CrossRef Search ADS PubMed  16. Macdonald JH, Millar JS, Barker CS. Endovascular treatment of cranial dural arteriovenous fistulae: a single-centre, 14-year experience and the impact of Onyx on local practise. Neuroradiology . 2010; 52( 5): 387- 395. Google Scholar CrossRef Search ADS PubMed  17. Maimon S, Nossek E, Strauss I, Blumenthal D, Frolov V, Ram Z. Transarterial treatment with Onyx of intracranial dural arteriovenous fistula with cortical drainage in 17 patients. AJNR Am J Neuroradiol . 2011; 32( 11): 2180- 2184. Google Scholar CrossRef Search ADS PubMed  18. Nogueira RG, Dabus G, Rabinov JD et al.   Preliminary experience with onyx embolization for the treatment of intracranial dural arteriovenous fistulas. AJNR Am J Neuroradiol . 2008; 29( 1): 91- 97. Google Scholar CrossRef Search ADS PubMed  19. Panagiotopoulos V, Möller-Hartmann W, Asgari S, Sandalcioglu IE, Forsting M, Wanke I. Onyx embolization as a first line treatment for intracranial dural arteriovenous fistulas with cortical venous reflux. Rofo . 2009; 181( 2): 129- 138. Google Scholar CrossRef Search ADS PubMed  20. Rabinov JD, Yoo AJ, Ogilvy CS, Carter BS, Hirsch JA. Onyx versus n-BCA for embolization of cranial dural arteriovenous fistulas. J Neurointerv Surg . 2013; 5( 4): 306- 310. Google Scholar CrossRef Search ADS PubMed  21. Saraf R, Shrivastava M, Siddhartha W, Limaye U. Evolution of endovascular management of intracranial dural arteriovenous fistulas: single center experience. Neurol India . 2010; 58( 1): 62- 68. Google Scholar CrossRef Search ADS PubMed  22. Cognard C, Gobin YP, Pierot L et al.   Cerebral dural arteriovenous fistulas: clinical and angiographic correlation with a revised classification of venous drainage. Radiology . 1995; 194( 3): 671- 680. Google Scholar CrossRef Search ADS PubMed  23. Stroup DF, Berlin JA, Morton SC et al.   Meta-analysis of observational studies in epidemi- ology: a proposal for reporting. Meta-analysis Of Observational Studies in Epidemiology (MOOSE) group. JAMA . 2000; 283( 15): 2008- 2012. Google Scholar CrossRef Search ADS PubMed  24. Moher D, Liberati A, Tetzlaff J, Group Altman DG; PRISMA. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Ann Intern Med . 2009; 151( 4): 264- 269. Google Scholar CrossRef Search ADS PubMed  25. U.S. department of Health and Human services-National Institutes of Health (NIH). Quality Assessment Tool for Before-After (Pre-Post) Studies With No Control Group . 2014 26. Nyaga VN, Arbyn M, Aerts M. Metaprop: a Stata command to perform meta-analysis of binomial data. Arch Public Health . 2014; 72( 1): 39. 27. Choo DM, Shankar JJ. Onyx versus nBCA and coils in the treatment of intracranial dural arteriovenous fistulas. Interv Neuroradiol . 2016; 22( 2): 212- 216. Google Scholar CrossRef Search ADS PubMed  Supplemental digital content is available for this article at www.neurosurgery-online.com. COMMENTS Definitive treatment of dural arteriovenous fistulas (dAVF) requires permanent occlusion of the draining vein. This can be accomplished with either surgery or with a variety of endovascular techniques. For intracranial dAVFs, endovascular treatment is appealing as a less-invasive alternative to open surgery, and because surgery for many intracranial dAVFs can be complex and challenging. This is the opposite of the situation for spinal dAVFs, for which surgery is usually straightforward and endovascular treatment can be technically challenging. We have come to regard endovascular treatment as first-line for most intracranial dAVFs and surgery first-line for most spinal dAVFs. For spinal dAVFs, this practice was supported by a recent systematic review, which reported initial definitive fistula occlusion in 96.6% of surgery patients and 72.2% of endovascular patients (P < .001).1 This study also found that recurrence was significantly more likely in endovascular patients. The authors of the present study present a comprehensive systematic review of endovascular treatment of intracranial dAVFs. The initial occlusion rate of 82%, a mid-term (mean follow-up 5 months) recurrence rate of 2%, and pooled adverse event rate of 4% seem favorable. This study is limited by the lack of surgery comparison group – which is somewhat understandable, given the paucity of intracranial dAVF surgical series, compared to spinal dAVFs – and also because of the possibility of publication bias. This report will serve as a contemporary summary of the risks and benefits of endovascular treatment of intracranial dAVFs for informed consent with patients prior to endovascular treatment. The usefulness of this study as an aid to clinical decision-making, however, is less clear. Because surgical studies were not included in the analysis, the reader-clinician must consider the endovascular results in this study in light of a sober estimate of the risks and benefits of surgery for intracranial dAVFs from personal experience and other published sources. For example, these reviewers regard surgery as first-line for most anterior fossa dAVFs. These cases are typically straightforward from a neurosurgical standpoint and endovascular treatment, because of the risk of reflux of embolic material into the ophthalmic artery and retinal ischemia, is essentially a circus trick.2 Furthermore, one cannot make any firm conclusion from this study regarding the efficacy of Onyx (EV3) versus n-BCA for the treatment of intracranial dAVFs. Onyx has become trendy in recent years, partly because of the apparent ease of use of Onyx compared to n-BCA. The use of a “non-adhesive” liquid embolic is perceived as being safer than “adhesive liquid embolics” to operators unaccustomed to n-BCA. Therefore, because Onyx has become more popular in recent years, concomitant improvements and evolution in embolization techniques in recent years can contribute to an appearance of Onxy superiority. We actually favor n-BCA for almost all dAVF embolization procedures, and have only rarely observed fistula recurrence after complete obliteration. Interestingly, the systematic review of spinal dAVFs found a nonsignificantly higher rate of recurrence in patients treated with Onyx compared to n-BCA (P = .13).1 Paul Foreman Mark Harrigan Birmingham, Alabama 1. Bakker NA, Uyttenboogaart M, Luijckx GJ, et al.   Recurrence rates after surgical or endovascular treatment of spinal dural arteriovenous fistulas: a meta-analysis. Neurosurgery  2015; 77: 137– 44; discussion 44. Google Scholar CrossRef Search ADS PubMed  2. Gross BA, Moon K, Kalani MY, et al.  . Clinical and anatomic insights from a series of ethmoidal dural arteriovenous fistulas at barrow neurological institute. World Neurosurg  2016; 93: 94– 9. Google Scholar CrossRef Search ADS PubMed  The present study demonstrates the effectiveness of transarterial embolization with Onyx (EV3) in the treatment of intracranial dAVF. The rate of recurrence is not surprising given the fact that dAVF represent a problem of the venous system, and the main limitation in the past with transarterial treatment has been reaching the fistula itself at the venous end. For this reason, a shorter-term follow-up (3 months vs 6 months) with DSA is indicated. Also not surprising is the better initial occlusion rate of Onyx vs n-BCA in the meta-analysis. The benefit of Onyx over n-BCA in the transarterial treatment of dAVF is that it is more controllable and the injection times can persist until the venous side is reached. The rate of cranial nerve palsies, although low, cannot be overemphasized. One should reserve the transarterial route for those dAVF in which the fistula itself is easily identified or venous access proves too difficult. Jay U. Howington Savannah, Georgia Copyright © 2017 by the Congress of Neurological Surgeons This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Neurosurgery Oxford University Press

Loading next page...
 
/lp/ou_press/transarterial-onyx-embolization-of-intracranial-dural-fistulas-a-vwyln8SKFA
Publisher
Congress of Neurological Surgeons
Copyright
Copyright © 2017 by the Congress of Neurological Surgeons
ISSN
0148-396X
eISSN
1524-4040
D.O.I.
10.1093/neuros/nyx309
Publisher site
See Article on Publisher Site

Abstract

Abstract BACKGROUND Onyx is important embolic material in the endovascular treatment of intracranial dural arteriovenous fistula (DAVF). However, its impact on DAVF occlusion rates, morbidity, mortality, and complication rates is not fully examined. OBJECTIVE To improve understanding of safety and effectiveness profiles associated with transarterial endovascular treatment using Onyx for intracranial DAVF METHODS We analyzed data from our prospective clinical registry and conducted a systematic review of all previous transarterial embolization studies using Onyx published between January 2005 and December 2015 in MEDLINE and EMBASE. RESULTS In the prospective study, 41 transarterial procedures were performed in 33 consecutive patients harboring 36 DAVFs. Complete initial exclusion was obtained in 32 of 36 (88.9%) fistulas; 31 fistulas were followed up showing 4 (12.9%) recurrences. Procedure-related morbidity and mortality were 3% and 0%, respectively. The literature review identified 19 studies involving a total of 425 patients with 463 DAVFs. Meta-analysis, including our registry data, showed an initial complete occlusion rate of 82% (95% confidence interval [CI]: 74%, 88%; I2, 70.6%), and recurrence rate at midterm of 2% (95% CI: 0%, 5%; I2, 21.5%). Pooled postoperative neurological deficit, procedure-related morbidity, and mortality rates were 4% (95% CI: 2%, 6%; I2, 0%), 3% (95% CI: 1%, 5%; I2, 0%), and 0%, respectively. CONCLUSION This meta-analysis suggests that transarterial embolization with Onyx is a safe treatment modality for DAVFs. Although Onyx showed a low recurrence rate at midterm, the long-term risk is poorly addressed in our study and should warrant a longer follow-up. Endovascular treatment, Embolization, Meta-analysis, Dural fistula, Onyx, Outcome ABBREVIATIONS ABBREVIATIONS BCA butyl-cyanoacrylate CI confidence interval DAVF dural arteriovenous fistula DSA digital subtraction angiography EVT endovascular treatment mRS modified Rankin score Intracranial dural arteriovenous fistula (DAVF) accounts for 10% to 15% of intracranial vascular malformations and are treated by endovascular approach in the majority of cases.1 Ethylene vinyl alcohol/dimethyl sulfoxide polymer (Onyx, EV3, Irvine, California) is an important liquid embolic material in the endovascular treatment (EVT) of intracranial DAVFs. Onyx embolic material has been extensively used due to its nonadhesive properties, allowing prolonged injection times with good lesion penetration, and convenient control for the operator.2 Since the introduction of Onyx in 2005, several single-center studies have demonstrated acceptable rates of DAVF occlusion, morbidity, and mortality for patients treated with Onyx via transarterial approach.3-21 Improved understanding of safety and effectiveness profiles associated with Onyx treatment of intracranial DAVF is needed. Thus, we analyzed data from our prospective clinical registry and conducted a systematic review and meta-analysis of the literature with special interest in intracranial DAVF occlusion rates, procedure-related complication rates, and recurrence rates at follow-up for intracranial DAVFs treated with Onyx via a transarterial approach. METHODS Prospective Study Patients This study was approved by the local ethics committee and was found to conform to scientific principles and research ethics standards. An informed consent was obtained from each patient. This study was designed, conducted, and analyzed without industry assistance, and the article was written independently of industry or any other financial support. The population was nested within a longitudinal cohort of consecutive patients who were referred to our Neurologic University Hospital for EVT of intracranial DAVFs between January 2013 and December 2015. This prospectively maintained database was queried retrospectively to identify all consecutive patients matching the following inclusion criteria: (a) adult patients with intracranial DAVFs (b) who were treated by transarterial approach using Onyx (EV3) material. EVT and Initial Angiographic Results All procedures were performed under general anesthesia and full heparinization (60 IU/kg). Complete selective digital subtraction angiography (DSA) was performed before treatment. An Onyx-compatible microcatheter was coaxially positioned into one of major arterial feeder as close as possible to the fistula site. First, the dead space of the microcatheter was flushed with dimethyl-sulfoxide, followed by a slow injection of Onyx under subtraction fluoroscopy. If an Onyx reflux occurred, the injection was paused for 30 to 60 s. During Onyx progression into the shunt, it was continually and slowly injected. The removal of the microcatheter was done when the Onyx cast stopped advancing or if we had a complete occlusion of the shunt. A postoperative DSA was performed after the procedure to confirm the complete occlusion of the DAVF (Figure 1). Initial angiographic results of EVT were classified by a neuroradiologist not involved in the initial treatment (B.G. with 7 yr of experience in interventional neuroradiology) by using the Cognard classification.22 FIGURE 1. View largeDownload slide Demonstration of a case treated with transarterial embolization with Onyx (EV3). A and B, Pretreatment lateral and anterior angiograms after external carotid artery injection show a Cognard type II a+b DAVF centered along the left lateral sinus with arterial supply from branches of the middle meningeal and occipital arteries. There is an important cortical venous reflux, especially into the Labbé vein. C and D, The microcatheter is navigated through the parietal branch of the left middle meningeal artery near the fistula. E and F, The cast of Onyx is seen to extend from middle meningeal artery, occipital artery, fistula, and draining vein. G and H, Post-treatment lateral and anterior angiograms after primitive carotid artery injection show a complete occlusion of the fistula without cortical venous drainage. FIGURE 1. View largeDownload slide Demonstration of a case treated with transarterial embolization with Onyx (EV3). A and B, Pretreatment lateral and anterior angiograms after external carotid artery injection show a Cognard type II a+b DAVF centered along the left lateral sinus with arterial supply from branches of the middle meningeal and occipital arteries. There is an important cortical venous reflux, especially into the Labbé vein. C and D, The microcatheter is navigated through the parietal branch of the left middle meningeal artery near the fistula. E and F, The cast of Onyx is seen to extend from middle meningeal artery, occipital artery, fistula, and draining vein. G and H, Post-treatment lateral and anterior angiograms after primitive carotid artery injection show a complete occlusion of the fistula without cortical venous drainage. Systematic Standard Follow-up Protocol After discharge, the systematic follow-up included at least a clinical examination and DSA at 3 mo after EVT for ruptured DAVFs and at 6 mo for unruptured DAVFs. Clinical Follow-up Further, clinical follow-up data were collected during hospitalization for follow-up DSA or external consultation at 3 mo. Postoperative neurological complication was defined as any new neurological symptoms, including cranial nerve palsy. Procedure-related morbidity was defined as a permanent neurological deficit including cranial nerve palsy or change in modified Rankin Scale (mRS) score ≥ 1 at 3 mo after the procedure. Image Acquisition and Analysis Angiographic images were acquired in anteroposterior and lateral projections before and immediately after treatment. Angiographic images obtained immediately after EVT were compared with those obtained at angiographic follow-up. At follow-up, we considered a DAVF recurrence when an early venous opacification was observed at DAVF completely occluded regardless of how big it is. In addition, we also considered it as a recurrence when the type of DAVF was modified. Systematic Review We prepared this study in accordance with the Meta-analysis of Observational Studies in Epidemiology and Preferred Reporting Items for Systematic Reviews and Meta-analyses guidelines, including objectives and plans for collecting and analyzing the data.23,24 We performed a systematic review of the literature by 2 authors (U.S.G. and N.M.) using the keywords “dural arteriovenous fistula” and “onyx” and “cerebral” or “brain” or “intracranial” in MEDLINE and EMBASE databases. Inclusion criteria were the following: all types of study design published in the English language, ≥10 patients, from January 2005 to December 2015, where EVT was performed via a transarterial route and where Onyx was used as the embolic material. Exclusion criteria were the following: case report, studies with duplicate case series, studies with balloon-assisted technique. Baseline Characteristics of Patients and DAVFs Clinical data included sex, number of eligible patients, mean age of patients, and clinical presentation. DAVFs characteristics, procedure-related complications, morbidity and mortality, anatomic outcomes, and follow-up modality. All the data were reviewed and collected individually by 2 authors (S.G.U. and M.N.) and compared. Disagreements were discussed until a consensus was reached. Outcomes Only patients treated with Onyx alone were considered for the calculation of the postoperative complications, morbidity, and mortality rates after treatment and DAVF recurrence rates. End points included postoperative neurological deficit after EVT, including cranial nerve palsy, and anatomic results obtained at DSA or magnetic resonance angiography. Postoperative neurological complication was differentiated as ischemic or hemorrhagic. Individual Study Quality Assessment The included studies being uncontrolled, individual study quality was assessed using a checklist published by the National Institutes of Health for before–after (pre–post) studies with no control group.25 This 12-item checklist enables an assessment of uncontrolled studies and provides an overall quality rating. Quality assessment was performed independently by 2 authors (S.G.U. and M.N.). Statistical Analysis Systematic Review The 95% confidence intervals (CIs) of the estimates were built with the Wilson method. The estimate and the 95% CI of the mean percentage over all the studies were obtained for each outcome using a logistic mixed model with a random effect on the intercept in order to take into account the heterogeneity between the studies. In all analyses, inconsistency of findings throughout studies was assessed by using the P value and the I2 statistic. We searched for publication bias using Egger's test for small-study effects and presented funnel plots. The meta-analyses were carried out using the metaprop Stata command on Stata/SE 14.1 (Statacorp LP, College Station, Texas).26 RESULTS Prospective Study Population A total of 41 procedures were performed in 33 patients (21 men and 12 women, mean age, 57.4 yr) harboring 36 DAVFs (3 patients had 2 DAVFs). Table 1 provides baseline patient and DAVF characteristics. Initial clinical symptoms were 10 intracranial hemorrhages, 12 tinnitus, and 2 seizures, while the rest of the patients presented with incidentally discovered DAVFs. TABLE 1. Neurological Hospital Baseline Characteristics of Patients and Dural Fistulas Patients  Sex  Age  DAVF Classification22  Symptoms  Access artery  Initial complete occlusionb  Complication  Procedure-related morbidity  Recurrenceb  1  M  65  III  Nonspecific  MMA  Yes  No  0  NA  1a  M  65  IV  Hemorrhage  MMA  Yes  No  0  NA  2  F  39  II a  Tinnitus  MMA  Partial (I)  No  0  Recurrence (II b)  3  F  43  IV  Nonspecific  MMA  Yes  CNP (III)  0  NA  4  M  75  III  Hemorrhage  MMA  Partial (I)  No  0  Partial (I)  5  F  64  I  Tinnitus  MMA  Yes  No  0  NA  6  M  62  I  Tinnitus  MMA  Yes  No  0  NA  7  M  70  IV  Hemorrhage  APA  Yes  No  0  No  8  M  71  II b  Nonspecific  MMA  Yes  No  0  No  9  F  38  III  Hemorrhage  MMA  Yes  No  0  No  10  F  59  II b  Tinnitus  MMA  Yes  No  0  No  11  M  50  II b  Nonspecific  MMA  Yes  No  0  No  12  M  46  II a  Tinnitus  MMA  Yes  No  0  No  13  F  78  IV  Hemorrhage  MMA  Partial (IV)  No  0  Noc  14  M  66  I  Tinnitus  MMA  Yes  No  0  No  15  F  56  II b  Tinnitus  MMA  Partial (II a)  Ischemia  Yes  Partial (II a)  15a  F  56  II a+b  Tinnitus  MMA  Yes  CNP (VII)  0  Recurrence (II a)  16  F  45  III  Hemorrhage  PMA  Yes  No  0  No  17  M  57  II b  Nonspecific  MMA  Yes  No  0  Recurrence (II b)  17a  M  57  II a+b  Nonspecific  MMA  Yes  No  0  No  18  M  55  I  Tinnitus  MMA  Yes  Microcatheter fracture  0  No  19  M  54  IV  Seizure  MMA  Yes  No  0  No  20  F  47  II b  Tinnitus  MMA  Yes  No  0  No  21  F  42  I  Tinnitus  MMA  Yes  No  0  No  22  F  61  III  Hemorrhage  MMA  Yes  No  0  No  23  M  64  I  Tinnitus  MMA  Yes  No  0  Recurrence (I)  24  M  68  IV  Nonspecific  OA  Yes  No  0  No  25  M  44  II b  Nonspecific  SCA  Yes  No  0  No  26  M  65  IV  Nonspecific  OA  Yes  No  0  No  27  M  50  III  Seizure  MMA  Yes  No  0  No  28  M  81  II b  Hemorrhage  MMA  Yes  No  0  No  29  M  73  II a  Nonspecific  OccA  Yes  No  0  No  30  F  48  I  Tinnitus  MMA  Yes  No  0  No  31  M  51  III  Hemorrhage  MMA  Yes  No  0  No  32  M  55  III  Nonspecific  MMA  Yes  No  0  No  33  M  52  IV  Hemorrhage  MMA  Yes  No  0  No  Patients  Sex  Age  DAVF Classification22  Symptoms  Access artery  Initial complete occlusionb  Complication  Procedure-related morbidity  Recurrenceb  1  M  65  III  Nonspecific  MMA  Yes  No  0  NA  1a  M  65  IV  Hemorrhage  MMA  Yes  No  0  NA  2  F  39  II a  Tinnitus  MMA  Partial (I)  No  0  Recurrence (II b)  3  F  43  IV  Nonspecific  MMA  Yes  CNP (III)  0  NA  4  M  75  III  Hemorrhage  MMA  Partial (I)  No  0  Partial (I)  5  F  64  I  Tinnitus  MMA  Yes  No  0  NA  6  M  62  I  Tinnitus  MMA  Yes  No  0  NA  7  M  70  IV  Hemorrhage  APA  Yes  No  0  No  8  M  71  II b  Nonspecific  MMA  Yes  No  0  No  9  F  38  III  Hemorrhage  MMA  Yes  No  0  No  10  F  59  II b  Tinnitus  MMA  Yes  No  0  No  11  M  50  II b  Nonspecific  MMA  Yes  No  0  No  12  M  46  II a  Tinnitus  MMA  Yes  No  0  No  13  F  78  IV  Hemorrhage  MMA  Partial (IV)  No  0  Noc  14  M  66  I  Tinnitus  MMA  Yes  No  0  No  15  F  56  II b  Tinnitus  MMA  Partial (II a)  Ischemia  Yes  Partial (II a)  15a  F  56  II a+b  Tinnitus  MMA  Yes  CNP (VII)  0  Recurrence (II a)  16  F  45  III  Hemorrhage  PMA  Yes  No  0  No  17  M  57  II b  Nonspecific  MMA  Yes  No  0  Recurrence (II b)  17a  M  57  II a+b  Nonspecific  MMA  Yes  No  0  No  18  M  55  I  Tinnitus  MMA  Yes  Microcatheter fracture  0  No  19  M  54  IV  Seizure  MMA  Yes  No  0  No  20  F  47  II b  Tinnitus  MMA  Yes  No  0  No  21  F  42  I  Tinnitus  MMA  Yes  No  0  No  22  F  61  III  Hemorrhage  MMA  Yes  No  0  No  23  M  64  I  Tinnitus  MMA  Yes  No  0  Recurrence (I)  24  M  68  IV  Nonspecific  OA  Yes  No  0  No  25  M  44  II b  Nonspecific  SCA  Yes  No  0  No  26  M  65  IV  Nonspecific  OA  Yes  No  0  No  27  M  50  III  Seizure  MMA  Yes  No  0  No  28  M  81  II b  Hemorrhage  MMA  Yes  No  0  No  29  M  73  II a  Nonspecific  OccA  Yes  No  0  No  30  F  48  I  Tinnitus  MMA  Yes  No  0  No  31  M  51  III  Hemorrhage  MMA  Yes  No  0  No  32  M  55  III  Nonspecific  MMA  Yes  No  0  No  33  M  52  IV  Hemorrhage  MMA  Yes  No  0  No  Note: M, male; F, female; NA, nonassessed; MMA, middle meningeal artery; OA, ophtalmic artery; OccA, occipital artery; PMA, posterior meningeal artery; SCA, superior cerebellar artery; APA, ascending pharengeal artery; CNP, cranial nerve palsy. aPatients 1, 15, and 17 had 2 DAVFs. bData in parentheses is the type of DAVF according Cognard classification.22 cA complementary surgery was performed after initial endovascular treatment failure. View Large TABLE 1. Neurological Hospital Baseline Characteristics of Patients and Dural Fistulas Patients  Sex  Age  DAVF Classification22  Symptoms  Access artery  Initial complete occlusionb  Complication  Procedure-related morbidity  Recurrenceb  1  M  65  III  Nonspecific  MMA  Yes  No  0  NA  1a  M  65  IV  Hemorrhage  MMA  Yes  No  0  NA  2  F  39  II a  Tinnitus  MMA  Partial (I)  No  0  Recurrence (II b)  3  F  43  IV  Nonspecific  MMA  Yes  CNP (III)  0  NA  4  M  75  III  Hemorrhage  MMA  Partial (I)  No  0  Partial (I)  5  F  64  I  Tinnitus  MMA  Yes  No  0  NA  6  M  62  I  Tinnitus  MMA  Yes  No  0  NA  7  M  70  IV  Hemorrhage  APA  Yes  No  0  No  8  M  71  II b  Nonspecific  MMA  Yes  No  0  No  9  F  38  III  Hemorrhage  MMA  Yes  No  0  No  10  F  59  II b  Tinnitus  MMA  Yes  No  0  No  11  M  50  II b  Nonspecific  MMA  Yes  No  0  No  12  M  46  II a  Tinnitus  MMA  Yes  No  0  No  13  F  78  IV  Hemorrhage  MMA  Partial (IV)  No  0  Noc  14  M  66  I  Tinnitus  MMA  Yes  No  0  No  15  F  56  II b  Tinnitus  MMA  Partial (II a)  Ischemia  Yes  Partial (II a)  15a  F  56  II a+b  Tinnitus  MMA  Yes  CNP (VII)  0  Recurrence (II a)  16  F  45  III  Hemorrhage  PMA  Yes  No  0  No  17  M  57  II b  Nonspecific  MMA  Yes  No  0  Recurrence (II b)  17a  M  57  II a+b  Nonspecific  MMA  Yes  No  0  No  18  M  55  I  Tinnitus  MMA  Yes  Microcatheter fracture  0  No  19  M  54  IV  Seizure  MMA  Yes  No  0  No  20  F  47  II b  Tinnitus  MMA  Yes  No  0  No  21  F  42  I  Tinnitus  MMA  Yes  No  0  No  22  F  61  III  Hemorrhage  MMA  Yes  No  0  No  23  M  64  I  Tinnitus  MMA  Yes  No  0  Recurrence (I)  24  M  68  IV  Nonspecific  OA  Yes  No  0  No  25  M  44  II b  Nonspecific  SCA  Yes  No  0  No  26  M  65  IV  Nonspecific  OA  Yes  No  0  No  27  M  50  III  Seizure  MMA  Yes  No  0  No  28  M  81  II b  Hemorrhage  MMA  Yes  No  0  No  29  M  73  II a  Nonspecific  OccA  Yes  No  0  No  30  F  48  I  Tinnitus  MMA  Yes  No  0  No  31  M  51  III  Hemorrhage  MMA  Yes  No  0  No  32  M  55  III  Nonspecific  MMA  Yes  No  0  No  33  M  52  IV  Hemorrhage  MMA  Yes  No  0  No  Patients  Sex  Age  DAVF Classification22  Symptoms  Access artery  Initial complete occlusionb  Complication  Procedure-related morbidity  Recurrenceb  1  M  65  III  Nonspecific  MMA  Yes  No  0  NA  1a  M  65  IV  Hemorrhage  MMA  Yes  No  0  NA  2  F  39  II a  Tinnitus  MMA  Partial (I)  No  0  Recurrence (II b)  3  F  43  IV  Nonspecific  MMA  Yes  CNP (III)  0  NA  4  M  75  III  Hemorrhage  MMA  Partial (I)  No  0  Partial (I)  5  F  64  I  Tinnitus  MMA  Yes  No  0  NA  6  M  62  I  Tinnitus  MMA  Yes  No  0  NA  7  M  70  IV  Hemorrhage  APA  Yes  No  0  No  8  M  71  II b  Nonspecific  MMA  Yes  No  0  No  9  F  38  III  Hemorrhage  MMA  Yes  No  0  No  10  F  59  II b  Tinnitus  MMA  Yes  No  0  No  11  M  50  II b  Nonspecific  MMA  Yes  No  0  No  12  M  46  II a  Tinnitus  MMA  Yes  No  0  No  13  F  78  IV  Hemorrhage  MMA  Partial (IV)  No  0  Noc  14  M  66  I  Tinnitus  MMA  Yes  No  0  No  15  F  56  II b  Tinnitus  MMA  Partial (II a)  Ischemia  Yes  Partial (II a)  15a  F  56  II a+b  Tinnitus  MMA  Yes  CNP (VII)  0  Recurrence (II a)  16  F  45  III  Hemorrhage  PMA  Yes  No  0  No  17  M  57  II b  Nonspecific  MMA  Yes  No  0  Recurrence (II b)  17a  M  57  II a+b  Nonspecific  MMA  Yes  No  0  No  18  M  55  I  Tinnitus  MMA  Yes  Microcatheter fracture  0  No  19  M  54  IV  Seizure  MMA  Yes  No  0  No  20  F  47  II b  Tinnitus  MMA  Yes  No  0  No  21  F  42  I  Tinnitus  MMA  Yes  No  0  No  22  F  61  III  Hemorrhage  MMA  Yes  No  0  No  23  M  64  I  Tinnitus  MMA  Yes  No  0  Recurrence (I)  24  M  68  IV  Nonspecific  OA  Yes  No  0  No  25  M  44  II b  Nonspecific  SCA  Yes  No  0  No  26  M  65  IV  Nonspecific  OA  Yes  No  0  No  27  M  50  III  Seizure  MMA  Yes  No  0  No  28  M  81  II b  Hemorrhage  MMA  Yes  No  0  No  29  M  73  II a  Nonspecific  OccA  Yes  No  0  No  30  F  48  I  Tinnitus  MMA  Yes  No  0  No  31  M  51  III  Hemorrhage  MMA  Yes  No  0  No  32  M  55  III  Nonspecific  MMA  Yes  No  0  No  33  M  52  IV  Hemorrhage  MMA  Yes  No  0  No  Note: M, male; F, female; NA, nonassessed; MMA, middle meningeal artery; OA, ophtalmic artery; OccA, occipital artery; PMA, posterior meningeal artery; SCA, superior cerebellar artery; APA, ascending pharengeal artery; CNP, cranial nerve palsy. aPatients 1, 15, and 17 had 2 DAVFs. bData in parentheses is the type of DAVF according Cognard classification.22 cA complementary surgery was performed after initial endovascular treatment failure. View Large Initial Angiographic Results According to Cognard's classification, 7 cases were type I, 3 cases were type II a, 8 cases were type II b, 2 cases were type II a+b, 8 cases were type III, and 8 cases were type IV. Initial complete exclusion of the DAVFs was obtained in 32 of 36 (88.9%) DAVFs. DAVF obliteration was achieved in 2 procedures for 5 patients and in 1 procedure for the other cases. Of the 4 partial EVTs, the type of DAVF was initially 1 II a (patient 2), 1 II b (patient 15), 1 III (patient 4), and 1 IV (patient 13). After EVT, the type II b reverted to type II a, and the type III to a type I. For the type IV, the occlusion after EVT was partial because of high risk of facial nerve palsy due to important microcatheter reflux. Thus, considering these 4 patients, 2 procedures were partial due to the risk of nerve palsy or reversion into benign type. Angiographic Results at Follow-up Angiography was obtained in 29 patients harboring 31 DAVFs. A recurrence was observed in 4 DAVFs (12.9%). The 10 initially ruptured DAVF were totally occluded at follow-up. A retreatment was performed in 2 DAVFs (6.5%) due to the aggressive type (only via transarterial approach). For the patients with initial partial obliteration, 2 patients showed similar DAVF type (patients 4 and 15), 1 patient presented complete fistula exclusion at follow-up (patient 13; a complementary surgery was performed after EVT failure), and 1 patient presented a type II b (patient 2). Clinical Complications A total of 3 patients had postoperative neurological complication following EVT. Two patients presented a cranial nerve palsy: 1 had a third cranial nerve palsy after treatment of DAVF type IV located at the anterior temporal lobe with subsequent transitory diplopia and ophthalmoplegia, which completely resolved 3 mo later; and 1 patient had a facial nerve palsy that quickly improved during the first few days, but partially persisted at 3-mo follow-up after treatment of a right lateral sinus fistula type II a+b. After treatment of her second DAVF type II b located in the superior longitudinal sinus, the patient (patient 15) presented with a left hemiplegia due to venous infarction, persistent at follow-up (mRS of 2). Procedure-Related Morbidity and Mortality In our prospective cohort, the procedure-related morbidity was 3% (1/33 patients). There was no procedure-related mortality. One patient died due to a compressive hemorrhage in the posterior fossa due to a type III ruptured DAVF. The hematoma was located in the vermis, causing an obstructive hydrocephalus with subsequent ventricular shunt placement. Emergency embolization was performed without complication during procedure, but 1 mo later the patient died from complications of the intracranial hematoma. Systematic Review Of the 424 records that were identified in the initial search, 309 were screened after removal of duplicates, and 280 were excluded at title or abstract level. Twenty-nine were articles that were reviewed at full-text out of which 19 studies were selected for final analysis (Figure 2). At baseline, 425 patients harboring 463 DAVFs were included. Baseline patient and DAVF characteristics of the 19 included studies are shown in Table, Supplemental Digital Content 1. Outcomes of the 19 studies are shown in Table, Supplemental Digital Content 2. Table 2 summarizes the results of the meta-analysis of the safety and effectiveness of EVT with Onyx (EV3). FIGURE 2. View largeDownload slide Flowchart shows screening and selection of studies for meta-analysis. FIGURE 2. View largeDownload slide Flowchart shows screening and selection of studies for meta-analysis. TABLE 2. Meta-analysis of the Safety and Effectiveness after Endovascular Treatment with Onyx (EV3) Outcomes  Number of studies  Sample size  Pooled rates (95% CI)  Heterogeneity  P value  I2  Cranial nerve palsy (n = 13)  16  466a  2% (1%, 4%)  .96  0.0%  Postoperative neurological complications (n = 22)  16  466a  4% (2%, 6%)  .94  0.0%  Procedure-related morbidity (n = 17)  16  366b  3% (1%, 5%)  .53  0.0%  Procedure-related mortality (n = 0)  17  366b  0% (0%, 0%)  1  0.0%  Initial complete obliteration (n = 373)  19  463c  82% (74%, 88%)  <.05  70.6%  Recurrence (n = 5)  13  263c  2% (0%, 5%)  .23  21.5%  Outcomes  Number of studies  Sample size  Pooled rates (95% CI)  Heterogeneity  P value  I2  Cranial nerve palsy (n = 13)  16  466a  2% (1%, 4%)  .96  0.0%  Postoperative neurological complications (n = 22)  16  466a  4% (2%, 6%)  .94  0.0%  Procedure-related morbidity (n = 17)  16  366b  3% (1%, 5%)  .53  0.0%  Procedure-related mortality (n = 0)  17  366b  0% (0%, 0%)  1  0.0%  Initial complete obliteration (n = 373)  19  463c  82% (74%, 88%)  <.05  70.6%  Recurrence (n = 5)  13  263c  2% (0%, 5%)  .23  21.5%  aNumber of procedures. bNumber of patients. cNumber of DAVFs. View Large TABLE 2. Meta-analysis of the Safety and Effectiveness after Endovascular Treatment with Onyx (EV3) Outcomes  Number of studies  Sample size  Pooled rates (95% CI)  Heterogeneity  P value  I2  Cranial nerve palsy (n = 13)  16  466a  2% (1%, 4%)  .96  0.0%  Postoperative neurological complications (n = 22)  16  466a  4% (2%, 6%)  .94  0.0%  Procedure-related morbidity (n = 17)  16  366b  3% (1%, 5%)  .53  0.0%  Procedure-related mortality (n = 0)  17  366b  0% (0%, 0%)  1  0.0%  Initial complete obliteration (n = 373)  19  463c  82% (74%, 88%)  <.05  70.6%  Recurrence (n = 5)  13  263c  2% (0%, 5%)  .23  21.5%  Outcomes  Number of studies  Sample size  Pooled rates (95% CI)  Heterogeneity  P value  I2  Cranial nerve palsy (n = 13)  16  466a  2% (1%, 4%)  .96  0.0%  Postoperative neurological complications (n = 22)  16  466a  4% (2%, 6%)  .94  0.0%  Procedure-related morbidity (n = 17)  16  366b  3% (1%, 5%)  .53  0.0%  Procedure-related mortality (n = 0)  17  366b  0% (0%, 0%)  1  0.0%  Initial complete obliteration (n = 373)  19  463c  82% (74%, 88%)  <.05  70.6%  Recurrence (n = 5)  13  263c  2% (0%, 5%)  .23  21.5%  aNumber of procedures. bNumber of patients. cNumber of DAVFs. View Large Initial Angiographic Results The overall initial complete occlusion rate was 82% (95% CI, 74%, 88%; Figure 3). Analysis of the data suggested significant heterogeneity across studies (P < .05), the range of initial complete occlusion rate being 47% to 100%. FIGURE 3. View largeDownload slide Crude odds ratios. A, Initial complete occlusion. B, Recurrence at midterm. FIGURE 3. View largeDownload slide Crude odds ratios. A, Initial complete occlusion. B, Recurrence at midterm. Angiographic Results at Follow-up The recurrence rate was 2% (95% CI, 0%, 5%; Figure 3). Recurrence rate was reported at midterm (mean follow-up, 5 mo; range, 3-7.5 mo). Long-term DAVFs recurrences rates were reported in 2 studies comprising a small sample size of 45 patients. Clinical Complications The pooled postoperative neurological complications rate was 4% (95% CI: 2%, 6%; I2, 0%; Figure 4). The pooled rate of postoperative cranial nerve palsy was 2% (95% CI: 1%, 4%; I2, 0%; Figure 4). The rate of cerebral ischemic and hemorrhagic complications rates were 1% (95% CI: 0%, 2%, I2, 0%) and 0% (95% CI: 0%, I2, 0%), respectively (Figure, Supplemental Digital Content 3). FIGURE 4. View largeDownload slide Crude odds ratio. A, Postoperative neurological complications. B, Cranial nerve palsy. C, Procedure-related morbidity. FIGURE 4. View largeDownload slide Crude odds ratio. A, Postoperative neurological complications. B, Cranial nerve palsy. C, Procedure-related morbidity. Procedure-Related Morbidity and Mortality The pooled procedure-related morbidity rate was 3% (95% CI, 1%, 5%; I2, 0%; Figure 4) with no statistically significant heterogeneity across studies. There were no procedure-related deaths across the 17 studies. Quality Assessment Eighteen studies were retrospective whereas 1 was prospective. All were noncomparative. Studies had large heterogeneity in terms of methods for the assessment of outcomes (presence or not of an adjudication committee; presence or not of a centralized core laboratory; time of follow-up). Using the prespecified tool, the quality rating of studies was considered as fair or poor. The main limitations of studies were as follows: no prespecification of selection criteria for the study population; no justification of sample size; no independent assessment of outcome measures across all study participants. Consequently, the risk of bias was significant across studies. Furthermore, we identified potential publication bias on the rate of total complications (Figure, Supplemental Digital Content 3; Figure, Supplemental Digital Content 4 for funnel plots). DISCUSSION Based on 19 studies involving a total of 1425 patients with 463 DAVFs, our meta-analysis provides more representative results on DAVF occlusion rates, morbidity, mortality, and complication rates than any single study. The rate of recurrence at midterm follow-up was low after EVT using Onyx (EV3) of intracranial DAVFs (2%, 95% CI: 0%, 5%) in our meta-analysis including our registry data. Although this finding demonstrates the effectiveness of Onyx in the EVT of cranial DAFVs, we observed a nonnegligible rate of DAVF angiographic recurrence (12.9%) at 3to 6 mo in our prospective study. Results at follow-up of intracranial DAVFs beyond 1 yr after EVT are not well known. Longer follow-up periods were reported in 2 studies including only a small number of patients (45 patients). Chandra et al9 observed no recurrence of 28 patients at 28 mo mean follow-up, whereas Ambekar et al4 reported 3 (14.3%) recurrences of 21 patients at 14 mo mean follow-up.4 This underlines the importance of long-term follow-up for DAVFs, especially for initially ruptured ones with the risk of rebleeding. However, we used DSA, which is the gold standard modality for follow-up of Onyx-treated DAVFs in our study and the series included in the systematic review. To date, there are several embolic materials to treat DAVFs by endovascular approach including n-butyl-cyanoacrylate (n-BCA), Onyx, polyvinyl alcohol particles, and coils. To date, the best embolic material for cranial DAVFs is not well known and comparative studies with good methodological standard are warranted, which is not the objective in the present study. However, Rabinov et al20 compared the effectiveness of cranial DAVF EVT with Onyx vs n-BCA in single-center study. Although the sample size was limited (56 fistulas), the initial complete occlusion rate reported for Onyx was 82% vs 33.3% for n-BCA. A superior durability of the occlusion with Onyx on follow-up was observed. In a recent single-center series of 24 fistulas, the authors compared intracranial DAVFs embolization with Onyx vs n-BCA and coils, and they reported initial complete occlusion rate of 66% for Onyx vs 22% for n-BCA.27 A possible explanation is that DAVFs are complex and heterogenic lesions, with considerable anatomopathological diversity, influencing the difficulty of the access and EVT phase. Furthermore, heterogeneity in operator experience may be another explanation. Procedure-related morbidity and mortality rates were uniformly low across the studies with pooled rates of 3% (95% CI, 1%, 5%) and 0%, respectively. This meta-analysis, including our registry data, demonstrates the safety of EVT of intracranial DAVFs with Onyx, most morbidity events related to cranial nerve palsy (2%; 95% CI: 1%, 4%). EVT via transarterial approach of lesions close to the skull base, as cranial DAVFs, carries an elevated risk for ischemic nerve injury.1 As Onyx embolization technique usually utilizes some degree of Onyx reflux, adequate safety margins should be considered appropriately to minimize inadvertent Onyx migration to clinically important vascular branches. Limitations Our study had several limitations. First, the articles included in the systematic review contained a majority of retrospective studies with a limited number of patients; some included combined liquid embolic materials. Second, a small number of studies (2 of 20 studies, 45 of 465 patients, 19.6%) with long-term follow-up were eligible. Third, data presentation was not uniform among the source articles, especially for the procedure-related morbidity definition. Fourth, it is possible that some relevant studies were not taken into account in our systematic review. However, it is unlikely that this potential publication bias strongly distorted our findings because we found no evidence of such bias by examining the funnel plots. CONCLUSION EVT of intracranial DAVFs with Onyx (EV3) via transarterial approach is a safe treatment modality. Although Onyx showed a low recurrence rate at mean follow-up of 5 mo, the risk of long-term recurrence is poorly evaluated in our study, and should warrant a longer follow-up period, especially in ruptured cases and neuroaggressive ones. Disclosure The authors have no personal, financial, or institutional interest in any of the drugs, materials, or devices described in this article. REFERENCES 1. Miller TR, Gandhi D. Intracranial dural arteriovenous fistulae: clinical presentation and management strategies. Stroke . 2015; 46( 7): 2017- 2025. Google Scholar CrossRef Search ADS PubMed  2. Sadeh-Gonik U, Gory B, Riva R et al.   Ethylene vinyl alcohol copolymer (Onyx) embolization of cranial dural arteriovenous fistula via the ascending pharyngeal artery. Diagn Interv Imaging . 2016; 97( 6): 681- 685. Google Scholar CrossRef Search ADS PubMed  3. Abud TG, Nguyen A, Saint-Maurice JP et al.   The use of Onyx in different types of intracranial dural arteriovenous fistula. AJNR Am J Neuroradiol . 2011; 32( 11): 2185- 2191. Google Scholar CrossRef Search ADS PubMed  4. Ambekar S, Gaynor BG, Peterson EC, Elhammady MS. Long-term angiographic results of endovascularly “cured” intracranial dural arteriovenous fistulas. J Neurosurg . 2016; 124( 4): 1123- 7. Google Scholar CrossRef Search ADS PubMed  5. Cha KC, Yeon JY, Kim GH, Jeon P, Kim JS, Hong SC. Clinical and angiographic results of patients with dural arteriovenous fistula. J Clin Neurosci . 2013; 20( 4): 536- 542. Google Scholar CrossRef Search ADS PubMed  6. Chew J, Weill A, Guilbert F, Raymond J, Audet ME, Roy D. Arterial onyx embolisation of intracranial DAVFs with cortical venous drainage. Can J Neurol Sci . 2009; 36( 2): 168- 175. Google Scholar CrossRef Search ADS PubMed  7. Cognard C, Januel AC, Silva NA Jr, Tall P. Endovascular treatment of intracranial dural arteriovenous fistulas with cortical venous drainage: new management using Onyx. AJNR Am J Neuroradiol . 2008; 29( 2): 235- 241. Google Scholar CrossRef Search ADS PubMed  8. De Keukeleire K, Vanlangenhove P, Kalala Okito JP, Hallaert G, Van Roost D, Defreyne L. Transarterial embolization with Onyx for treatment of intracranial non-cavernous dural arteriovenous fistula with or without cortical venous reflux. J Neurointerv Surg . 2011; 3( 3): 224- 228. Google Scholar CrossRef Search ADS PubMed  9. Chandra RV, Leslie-Mazwi TM, Mehta BP et al.   Transarterial onyx embolization of cranial dural arteriovenous fistulas: Long-term follow-up. AJNR Am J Neuroradiol . 2014; 35( 9): 1793- 1797. Google Scholar CrossRef Search ADS PubMed  10. Ghobrial GM, Marchan E, Nair AK et al.   Dural arteriovenous fistulas: a review of the literature and a presentation of a single institution's experience. World Neurosurg . 2013; 80( 1-2): 94- 102. Google Scholar CrossRef Search ADS PubMed  11. Hu YC, Newman CB, Dashti SR, Albuquerque FC, McDougall CG. Cranial dural arteriovenous fistula: transarterial Onyx embolization experience and technical nuances. J Neurointerv Surg . 2011; 3( 1): 5- 13. Google Scholar CrossRef Search ADS PubMed  12. Huang Q, Xu Y, Hong B, Li Q, Zhao W, Liu J. Use of Onyx in the management of tentorial dural arteriovenous fistulae. Neurosurgery . 2009; 65( 2): 287- 293. Google Scholar CrossRef Search ADS PubMed  13. Long XA, Karuna T, Zhang X, Luo B, Duan CZ. Onyx 18 embolisation of dural arteriovenous fistula via arterial and venous pathways: preliminary experience and evaluation of the short-term outcomes. Br J Radiol . 2012; 85( 1016): 395- 403. Google Scholar CrossRef Search ADS PubMed  14. Luo CB, Chang FC, Mu-Huo Teng M et al.   Transarterial Onyx embolization of intracranial dural arteriovenous fistulas: a single center experience. J Chin Med Assoc . 2014; 77( 4): 184- 189. Google Scholar CrossRef Search ADS PubMed  15. Lv X, Jiang C, Zhang J, Li Y, Wu Z. Complications related to percutaneous transarterial embolization of intracranial dural arteriovenous fistulas in 40 patients. AJNR Am J Neuroradiol . 2009; 30( 3): 462- 468. Google Scholar CrossRef Search ADS PubMed  16. Macdonald JH, Millar JS, Barker CS. Endovascular treatment of cranial dural arteriovenous fistulae: a single-centre, 14-year experience and the impact of Onyx on local practise. Neuroradiology . 2010; 52( 5): 387- 395. Google Scholar CrossRef Search ADS PubMed  17. Maimon S, Nossek E, Strauss I, Blumenthal D, Frolov V, Ram Z. Transarterial treatment with Onyx of intracranial dural arteriovenous fistula with cortical drainage in 17 patients. AJNR Am J Neuroradiol . 2011; 32( 11): 2180- 2184. Google Scholar CrossRef Search ADS PubMed  18. Nogueira RG, Dabus G, Rabinov JD et al.   Preliminary experience with onyx embolization for the treatment of intracranial dural arteriovenous fistulas. AJNR Am J Neuroradiol . 2008; 29( 1): 91- 97. Google Scholar CrossRef Search ADS PubMed  19. Panagiotopoulos V, Möller-Hartmann W, Asgari S, Sandalcioglu IE, Forsting M, Wanke I. Onyx embolization as a first line treatment for intracranial dural arteriovenous fistulas with cortical venous reflux. Rofo . 2009; 181( 2): 129- 138. Google Scholar CrossRef Search ADS PubMed  20. Rabinov JD, Yoo AJ, Ogilvy CS, Carter BS, Hirsch JA. Onyx versus n-BCA for embolization of cranial dural arteriovenous fistulas. J Neurointerv Surg . 2013; 5( 4): 306- 310. Google Scholar CrossRef Search ADS PubMed  21. Saraf R, Shrivastava M, Siddhartha W, Limaye U. Evolution of endovascular management of intracranial dural arteriovenous fistulas: single center experience. Neurol India . 2010; 58( 1): 62- 68. Google Scholar CrossRef Search ADS PubMed  22. Cognard C, Gobin YP, Pierot L et al.   Cerebral dural arteriovenous fistulas: clinical and angiographic correlation with a revised classification of venous drainage. Radiology . 1995; 194( 3): 671- 680. Google Scholar CrossRef Search ADS PubMed  23. Stroup DF, Berlin JA, Morton SC et al.   Meta-analysis of observational studies in epidemi- ology: a proposal for reporting. Meta-analysis Of Observational Studies in Epidemiology (MOOSE) group. JAMA . 2000; 283( 15): 2008- 2012. Google Scholar CrossRef Search ADS PubMed  24. Moher D, Liberati A, Tetzlaff J, Group Altman DG; PRISMA. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Ann Intern Med . 2009; 151( 4): 264- 269. Google Scholar CrossRef Search ADS PubMed  25. U.S. department of Health and Human services-National Institutes of Health (NIH). Quality Assessment Tool for Before-After (Pre-Post) Studies With No Control Group . 2014 26. Nyaga VN, Arbyn M, Aerts M. Metaprop: a Stata command to perform meta-analysis of binomial data. Arch Public Health . 2014; 72( 1): 39. 27. Choo DM, Shankar JJ. Onyx versus nBCA and coils in the treatment of intracranial dural arteriovenous fistulas. Interv Neuroradiol . 2016; 22( 2): 212- 216. Google Scholar CrossRef Search ADS PubMed  Supplemental digital content is available for this article at www.neurosurgery-online.com. COMMENTS Definitive treatment of dural arteriovenous fistulas (dAVF) requires permanent occlusion of the draining vein. This can be accomplished with either surgery or with a variety of endovascular techniques. For intracranial dAVFs, endovascular treatment is appealing as a less-invasive alternative to open surgery, and because surgery for many intracranial dAVFs can be complex and challenging. This is the opposite of the situation for spinal dAVFs, for which surgery is usually straightforward and endovascular treatment can be technically challenging. We have come to regard endovascular treatment as first-line for most intracranial dAVFs and surgery first-line for most spinal dAVFs. For spinal dAVFs, this practice was supported by a recent systematic review, which reported initial definitive fistula occlusion in 96.6% of surgery patients and 72.2% of endovascular patients (P < .001).1 This study also found that recurrence was significantly more likely in endovascular patients. The authors of the present study present a comprehensive systematic review of endovascular treatment of intracranial dAVFs. The initial occlusion rate of 82%, a mid-term (mean follow-up 5 months) recurrence rate of 2%, and pooled adverse event rate of 4% seem favorable. This study is limited by the lack of surgery comparison group – which is somewhat understandable, given the paucity of intracranial dAVF surgical series, compared to spinal dAVFs – and also because of the possibility of publication bias. This report will serve as a contemporary summary of the risks and benefits of endovascular treatment of intracranial dAVFs for informed consent with patients prior to endovascular treatment. The usefulness of this study as an aid to clinical decision-making, however, is less clear. Because surgical studies were not included in the analysis, the reader-clinician must consider the endovascular results in this study in light of a sober estimate of the risks and benefits of surgery for intracranial dAVFs from personal experience and other published sources. For example, these reviewers regard surgery as first-line for most anterior fossa dAVFs. These cases are typically straightforward from a neurosurgical standpoint and endovascular treatment, because of the risk of reflux of embolic material into the ophthalmic artery and retinal ischemia, is essentially a circus trick.2 Furthermore, one cannot make any firm conclusion from this study regarding the efficacy of Onyx (EV3) versus n-BCA for the treatment of intracranial dAVFs. Onyx has become trendy in recent years, partly because of the apparent ease of use of Onyx compared to n-BCA. The use of a “non-adhesive” liquid embolic is perceived as being safer than “adhesive liquid embolics” to operators unaccustomed to n-BCA. Therefore, because Onyx has become more popular in recent years, concomitant improvements and evolution in embolization techniques in recent years can contribute to an appearance of Onxy superiority. We actually favor n-BCA for almost all dAVF embolization procedures, and have only rarely observed fistula recurrence after complete obliteration. Interestingly, the systematic review of spinal dAVFs found a nonsignificantly higher rate of recurrence in patients treated with Onyx compared to n-BCA (P = .13).1 Paul Foreman Mark Harrigan Birmingham, Alabama 1. Bakker NA, Uyttenboogaart M, Luijckx GJ, et al.   Recurrence rates after surgical or endovascular treatment of spinal dural arteriovenous fistulas: a meta-analysis. Neurosurgery  2015; 77: 137– 44; discussion 44. Google Scholar CrossRef Search ADS PubMed  2. Gross BA, Moon K, Kalani MY, et al.  . Clinical and anatomic insights from a series of ethmoidal dural arteriovenous fistulas at barrow neurological institute. World Neurosurg  2016; 93: 94– 9. Google Scholar CrossRef Search ADS PubMed  The present study demonstrates the effectiveness of transarterial embolization with Onyx (EV3) in the treatment of intracranial dAVF. The rate of recurrence is not surprising given the fact that dAVF represent a problem of the venous system, and the main limitation in the past with transarterial treatment has been reaching the fistula itself at the venous end. For this reason, a shorter-term follow-up (3 months vs 6 months) with DSA is indicated. Also not surprising is the better initial occlusion rate of Onyx vs n-BCA in the meta-analysis. The benefit of Onyx over n-BCA in the transarterial treatment of dAVF is that it is more controllable and the injection times can persist until the venous side is reached. The rate of cranial nerve palsies, although low, cannot be overemphasized. One should reserve the transarterial route for those dAVF in which the fistula itself is easily identified or venous access proves too difficult. Jay U. Howington Savannah, Georgia Copyright © 2017 by the Congress of Neurological Surgeons This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

NeurosurgeryOxford University Press

Published: Jun 17, 2017

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off