Johannes Schramm, Karl Schaller, Jonas Esche and Azize Boström
The objective of this study was to review the outcomes after microsurgical resection of cerebral arteriovenous malformations (AVMs) from a consecutive single-surgeon series. Clinical and imaging data were analyzed to address the following questions concerning AVM treatment in the post-ARUBA (A Randomized Trial of Unruptured Brain Arteriovenous Malformations) era. 1) Are the patients who present with unruptured or ruptured AVMs doing better at long-term follow-up? 2) Is the differentiation between Ponce Class A (Spetzler-Martin Grade I and II) patients versus Ponce Class B and C patients (Spetzler-Martin Grade III and IV) meaningful and applicable to surgical practice? 3) How did the ARUBA-eligible patients of this surgical series compare with the results reported in ARUBA?
Two hundred eighty-eight patients with cerebral AVMs underwent microsurgical resection between 1983 and 2012 performed by the same surgeon (J.S.). This is a prospective case collection study that represents a consecutive series. The results are based on prospectively collected, early-outcome data that were supplemented by retrospectively collected, follow-up data for 94% of those cases. The analyzed data included the initial presentation, Spetzler-Martin grade, obliteration rates, surgical and neurological complications, and frequency of pretreatment with embolization or radiosurgery. The total cohort was compared using “small-AVM,” Spetzler-Martin Grade I and II, and ARUBA-eligible AVM subgroups.
The initial presentation was hemorrhage in 50.0% and seizures in 43.1% of patients. The series included 53 Spetzler-Martin Grade I (18.4%), 114 Spetzler-Martin Grade II (39.6%), 90 Spetzler-Martin Grade III (31.3%), 28 Spetzler-Martin Grade IV (9.7%), and 3 Spetzler-Martin Grade V (1.0%) AVMs. There were 144 unruptured and 104 ARUBA-eligible cases. Preembolization was used in 39 cases (13.5%). The occlusion rates for the total series and small AVM subgroup were 99% and 98.7%, respectively. The mean follow-up duration was 64 months. Early neurological deterioration was seen in 39.2% of patients, of which 12.2% had permanent and 5.6% had permanent significant deficits, and the mortality rate was 1.7% (n = 5). Outcome was better for patients with AVMs smaller than 3 cm (permanent deficit in 7.8% and permanent significant deficit in 3.2% of patients) and Ponce Class A status (permanent deficit in 7.8% and significant deficit in 3.2% of patients). Unruptured AVMs showed slightly higher new deficit rates (but 0 instances of mortality) among all cases, and in the small AVM and Ponce Class A subgroups. Unruptured Spetzler-Martin Grade I and II lesions had the best outcome (1.8% permanent significant deficit), and ARUBA-eligible Spetzler-Martin Grade I and II lesions had a slightly higher rate of permanent significant deficits (3.2%).
Microsurgery has a very high cure rate. Focusing microsurgical AVM resection on unruptured lesions smaller than 3 cm or on Spetzler-Martin Grade I and II lesions is a good strategy for minimizing long-term morbidity. Well-selected microsurgical cases lead to better outcomes than with multimodal interventions, as in the ARUBA treatment arm, or conservative treatment alone. Long-term prospective data collection is valuable.
Azize Boström, Timo Krings, Franz J. Hans, Johannes Schramm, Armin K. Thron and Joachim M. Gilsbach
Glomus-type spinal arteriovenous malformations (AVMs) are rare. In the literature only small series and anecdotal reports can be found, and there are no prospective series elucidating the natural course or the superiority of 1 treatment regimen over another (such as surgery versus embolization versus conservative treatment). Microsurgical treatment of spinal AVMs often seems difficult because many lesions are not anatomically suitable for primary microsurgical occlusion and are therefore treated with first-line neuroradiological interventions or not at all.
Between 1989 and 2005, 20 patients with glomus-type AVMs underwent microsurgical treatment at 2 major neurosurgical centers in Germany. The history of symptoms in these patients ranged from 2 days to 11 years. Four patients presented with subarachnoid hemorrhage, 2 with intramedullary hematoma, 4 with paresthesia or pain, and 10 with clinical signs of myelopathy. Seven patients underwent partial embolization prior to microsurgery. The authors only operated on AVMs accessible from a dorsal or dorsolateral approach. Neurological status was assessed with the McCormick classification scheme. Follow-up data were obtained from outpatient records. Three patients were interviewed over the telephone and 4 patients were not available for follow-up evaluation.
Surgery was performed via a laminectomy in 14 and hemilaminectomy in 6 patients. The microsurgical technique used consisted of retrograde dissection of the AVM from the venous side in most cases. Four (20%) of 20 patients showed worsening of neurological symptoms to a worse McCormick grade, probably caused by suspected venous stasis directly after surgery, however only 1 patient (5%) suffered permanent deterioration after surgery. In 14 patients postoperative angiography proved complete occlusion in 11 patients, including the presence of a remnant requiring a second operation with complete occlusion thereafter in 1 patient. In 3 patients occlusion was incomplete: a small residual AVM remained in 1 patient, and a discrete feeding vessel without a vein was evident in 2 patients.
Spinal cord AVMs are rare. If embolization is not possible, surgery may be indicated in selected cases. Spinal AVMs behave differently after incomplete occlusion either surgically or with embolization. A postoperative reduction in symptoms is frequent despite the presence of small remnants, and the risk of neurological deficits seems relatively low even in residual AVMs. Therefore, treatment need not necessarily aim at complete occlusion if that would be associated with an unacceptably high risk of neurological deficits.
Daniel Delev, Anna Pavlova, Alexander Grote, Azize Boström, Anke Höllig, Johannes Schramm, Rolf Fimmers, Johannes Oldenburg and Matthias Simon
Arteriovenous malformations (AVMs) of the brain are a frequent and important cause of intracranial hemorrhage in young adults. Little is known about the molecular-genetic pathomechanisms underlying AVM development. Genes of the NOTCH family control the normal development of vessels and proper arteriovenous specification. Transgenic mice with constitutive expression of active NOTCH4 frequently develop AVMs. Here, the authors report a genetic association study investigating possible associations between NOTCH4 gene polymorphisms and formation and clinical presentation of AVMs.
After PCR amplification and direct DNA sequencing or restriction digests, 10 single-nucleotide polymorphisms (SNPs) of the NOTCH4 gene were used for genotyping 153 AVM patients and 192 healthy controls (i.e., blood donors). Pertinent clinical data were available for 129 patients. Uni- and multivariate single-marker and explorative haplotype analyses were performed to identify potential genetic risk factors for AVM development and for hemorrhagic or epileptic presentation.
Eleven calculated haplotypes consisting of 3–4 SNPs (most of which were located in the epidermal growth factor–like domain of the NOTCH4 gene) were observed significantly more often among AVM patients than among controls. Univariate analysis indicated that rs443198_TT and rs915895_AA genotypes both were significantly associated with hemorrhage and that an rs1109771_GG genotype was associated with epilepsy. The association between rs443198_TT and AVM bleeding remained significant in the multivariate regression analysis.
The authors' results suggest NOTCH4 SNPs as possible genetic risk factors for the development and clinical presentation of AVMs and a role of NOTCH4 in the pathogenesis of this disease.
Matthias Simon, Daniel Franke, Michael Ludwig, Ales F. Aliashkevich, Gertraud Köster, Johannes Oldenburg, Azize Boström, Andreas Ziegler and Johannes Schramm
Important central nervous system (CNS) manifestations in patients with hereditary hemorrhagic telangiectasia (HHT) include arteriovenous malformations (AVMs) and dural arteriovenous fistulas (DAVFs). Hereditary hemorrhagic telangiectasia is caused by germline mutations of two genes: ENG (HHT Type 1) and ACVRL1 (HHT Type 2). The ENG gene variations have been associated with the formation of intracranial aneurysms. The authors studied whether sequence variations in ACVRL1 or ENG are associated with the development of clinically sporadic arteriovenous dysplasias and aneurysms of the CNS.
The coding sequence (in 44 patients with AVMs and 27 with aneurysms) and the 5′ end and the polyA site (in 53 patients with AVMs) of the ACVRL1 gene were analyzed for sequence variations using direct sequencing and single-strand conformational polymorphism analysis. One ENG and three ACVRL1 gene polymorphisms were genotyped using restriction enzyme–based analysis in 101 patients with sporadic AVMs and DAVFs of the CNS, 79 patients treated for intracranial aneurysms, and 202 control volunteers.
The authors identified a statistically significant association between the IVS3 −35A/T polymorphism in intron 3 of the ACVRL1 gene and the development of AVMs and DAVFs (p = 0.004; odds ratio [OR] 1.73; 95% confidence interval [CI] 1.19–2.51; after adjustments for age and sex), but not aneurysms (crude OR 0.82; 95% CI 0.55–1.18).
The results of this study link ACVRL1 (HHT Type 2 gene) to the formation of the clinically sporadic variants of vascular malformations of the CNS most commonly seen in patients with HHT, that is, AVMs and DAVFs.