The relationship of cortical folding and brain arteriovenous malformations
- Manish N. Shah†1Email author,
- Sarah E. Smith†2,
- Donna L. Dierker2,
- Joseph P. Herbert3,
- Timothy S. Coalson2,
- Brent S. Bruck4,
- Gregory J. Zipfel4,
- David C. Van Essen2 and
- Ralph G. DaceyJr.4
© The Author(s). 2016
Received: 25 October 2016
Accepted: 7 November 2016
Published: 22 November 2016
The pathogenesis of human intracranial arteriovenous malformations (AVMs) is not well understood; this study aims to quantitatively assess cortical folding in patients with these lesions.
Seven adult participants, 4 male and 3 female, with unruptured, surgically unresectable intracranial AVMs were prospectively enrolled in the study, with a mean age of 42.1 years and Spetzler-Martin grade range of II-IV. High-resolution brain MRI T1 and T2 sequences were obtained. After standard preprocessing, segmentation and registration techniques, three measures of cortical folding, the depth difference index (DDI), coordinate distance index (CDI) and gyrification index (GI)), were calculated for the affected and unaffected hemispheres of each subject as well as a healthy control subject set.
Of the three metrics, CDI, DDI and GI, used for cortical folding assessment, none demonstrated significant differences between the participants and previously studied healthy adults. There was a significant negative correlation between the DDI ratio between affected and unaffected hemispheres and AVM volume (correlation coefficient r = -0.74, p = 0.04).
This study is the first to quantitatively assess human brain cortical folding in the presence of intracranial AVMs and no significant differences between AVM-affected versus unaffected hemispheres were found in a small dataset. We suggest longitudinal, larger human MRI-based cortical folding studies to assess whether AVMs are congenital lesions of vascular development or de novo, dynamic lesions.
KeywordsAVM MRI Cortical folding de novo Congenital Brain development
Intracranial arteriovenous malformations (AVMs) are the direct communication of arteries to abnormally tortuous and dilated veins without an interposing capillary bed, often described as a tangle or “bag of worms” . The mechanism of AVM formation is not well understood. Unlike their pulmonary or abdominal counterparts, brain AVMs have long been thought to be congenital malformations situated in often eloquent, functional brain parenchyma. Due to this direct, high-pressure, high-flow connection, patients are subject to hemorrhage, seizures and strokes . If AVMs are truly congenital lesions of the intracranial vascular system, there is a poorly understood interaction between the processes of AVM formation and cortical folding. Although the large vessels in the intracranial vascular system are mostly formed and have perforated the cortex by weeks 8–10 in utero, the brain begins its intricate folding process at weeks 24–34 (reviewed in [3–5]). To date, no hypothesis or human study explains what happens in the 16 week interval between intracranial vasculature maturation and cortical folding initiation . There are no experimental animal AVM models within the brain parenchyma that expose either the mechanism of formation or the effect on cortical folding [7–10].
Cortical folding is a complex phenomenon that remains poorly understood. Our center has discovered cortical folding abnormalities in other diseases of the neurological system such as Williams’ syndrome  and autism spectrum disorder . Beginning at approximately 26 weeks gestation, cortical folding is thought to rely on a variety of factors including, but not limited to, gene expression, cortical growth, and tension from white matter fibers . In addition, adequate oxygenation is necessary for proper neurodevelopment and cortical folding. Maintenance of tissue perfusion and adequate oxygenation relies heavily on changes at the microvascular level in response to various physiologic cues. The neuronal vasculature in particular is exquisitely sensitive to such signals, which include CO2 and O2 levels, mechanical distention and compression, and changes in local neuronal activity . Dysregulation of the microvascular response to these factors results in an area of hypoperfusion. Brain tissue in an affected area is most at-risk for acute infarct or hypoperfusion when surrounding vasculature dilates in order to meet metabolic demand. Cortical steal syndrome occurs when increased resistance in the brain parenchyma relative to the surrounding normal vasculature causes a paradoxical drop in local perfusion pressure.
The extent to which AVMs induce this degree of local hypoxia is debated . Cerebral hypoperfusion has been demonstrated in some patients , but not in others . It appears that the main mechanism for maintenance of perfusion in brain surrounding AVMs in the latter group is neo-capillary formation . This phenomenon is presumably the result of some initial hypoxic state that induces local vasculogenesis. Hypoxia and cerebral ischemia can cause cortical thinning due to selective neuronal loss [18–20]. Fierstra et al. demonstrated that cortical thinning is also seen in patients with cortical steal syndrome . They proposed that repeated bouts of transient hypoperfusion associated with steal physiology, while insufficient to cause acute infarction, leads to selective neuronal loss over time.
It is likely that AVMs cause local tissue hypoxia at least initially and, as vasculogenesis precedes cortical folding by several months , AVMs could impact cortical development. The human brain has an overwhelming tendency to form complex cortical folds that are orderly in some respects but show a high degree of individual variability in most regions. If AVMs are true congenital lesions formed prior to cortical involution, their mechanical traction on adjacent brain parenchyma should substantially alter the process of cortical folding. This study aims to assess human cortical folding patterns with AVMs using advanced measurements of cortical shape and high-resolution MRI.
With institutional review board approval and informed consent, participants were prospectively enrolled in the study. A total of 7 otherwise healthy patients with surgically unresectable AVMs, incidentally found or presenting with seizures but without hemorrhage were enrolled. In addition, one healthy adult control subject was also enrolled, but was not used for the analysis. Some patients had recent stereotactic, focal irradiation to the AVM bed for treatment with Gamma Knife Radiosurgery, but there was no appreciable radiation effect on MRI.
Characterization of the Seven AVM Subjects. This table indicates the cohort demographics, AVM characteristics, presentation, treatment and follow-up. Yrs = years, F/U = follow-up, mo = months
Age (yrs) /Gender
All 8 subjects had sagittally-acquired 3D T1-weighted magnetization-prepared rapid gradient echo (MPRAGE) sequences and 3D T2-weighted sampling perfection with application optimized contrast using different angle evolutions (SPACE) on a Siemens 3 T TIM Trio MRI scanner. A generalized autocalibrating partially parallel acquisition (GRAPPA) factor of 2 was used for both scans with 50 % phase oversampling for the MPRAGE and no oversampling for the SPACE sequences, as previously described by Glasser and Van Essen .
Segmentation and registration
The T1w images were segmented using Freesurfer version 5.1.0. The resulting surfaces were registered to the fs_LR atlas and sulcal depth maps generated using previously described methods . Segmentations of the AVMs were manually traced using MRIcron .
Sulcal depth analysis
The PALS-B12.LR mean sulcal depth  was resampled to the fs_LR mesh using caret5 with an existing PALS-to-Conte69 deformation map . This was the average of the sulcal depth maps for twelve human subjects, all healthy, right-handed adults aged 18–24 (six female, six male). For each vertex within each hemisphere, the difference between the subject’s sulcal depth and the PALS-B12.LR mean sulcal depth was computed. The resulting depth difference reflects the degree to which cortical depth differs from the population mean at any given vertex. Figure 1 shows depth difference maps for each subject. For each hemisphere, a depth difference index (DDI) was computed by integrating across the surface (sum of the difference multiplied by a third of the area of the vertex’s tiles), but excluding vertices in the fs_LR medial wall, and then dividing by the surface area outside of the medial wall. These DDIs were computed for both affected and unaffected hemispheres, and then input to a paired t-test.
In some subjects, cortex near the AVM was abnormal enough to perturb cerebral hull generation, which in turn confounded sulcal depth computation. Inspection of all hemispheres revealed that the confounds affected cortex within 15 mm of the AVM, so we restricted our DDI computation to vertices farther than 15 mm from the AVM boundary. For the unaffected hemisphere, the same ROI was used, to exclude equivalent cortex from the measure on the unaffected hemisphere (left and right hemispheres are in register with one another in fs_LR standard mesh surfaces).
Coordinate distance analysis
The normative reference (RefSurf) was the mean midthickness surface from the Human Connectome Project (HCP) third release (Q1-2-3_Related196.L/R.midthickness.MSMSulc.164k_fs_LR.surf.gii), incorporating 196 healthy human young adults . Each subject’s midthickness surface (SubjectSurface) was normalized to MNI space by applying the talairach.xfm from the subject’s freesurfer directory.
For each vertex i within each hemisphere, the distance between RefSurf[i] and SubjectSurface[i] was computed. For each hemisphere, a weighted sum of distance was computed across the hemisphere (sum of the distance multiplied by the mean of the area of the vertex’s tiles), but excluding vertices in the fs_LR medial wall. Because this measure was not subject to the hull issues that limited DDI computation to vertices beyond 15 mm from the AVM, we computed it across two regions of interest: One local (within 15 mm of the AVM) and one remote ROI (further than 15 mm from the AVM). We hypothesized that differences in the affected hemisphere might be more pronounced near the AVM. This measure, the coordinate distance index (CDI), was computed by dividing the weighted sum of distance by the surface area of the associated region. CDIs were computed for both affected and unaffected hemispheres and then input to a paired t-test.
Gyrification index analysis
Gyrification index (GI) is defined as the ratio of cortical surface area to cerebral hull surface area [25, 27]. GI was computed for each subject, and then the GI for the affected and unaffected hemisphere was compared with a paired t test.
We assume and a one-tailed test with statistical significance, α = 0.05 and a study power of β = 0.80. For a moderate effect size (50 %), we would need 21 participants (G*Power 3.1, ).
Depth difference index (DDI), local (L) and remote (R) coordinate distance index (CDI and CDI), and gyrification index (GI) for Affected (A) and Unaffected (U) Hemispheres. Both DDI and CDI are vertex-wise differences (affected-unaffected) weighted by mean tile area and summed over vertices outside the medial wall (for DDI, vertices at least 20 mm from AVM)
L CDI (A)
L CDI (U)
R CDI (A)
R CDI (U)
This is the first morphometric study to quantitatively assess brain cortical folding differences in human subjects with brain arteriovenous malformations using Human Connectome Project (HCP) methodology [11, 22, 23]. We quantitatively assessed brain sulcal depth in 7 participants with AVMs compared to the mean depth of healthy young adults. We found no statistically significant differences between hemispheres containing the AVM and contralateral hemispheres in our sulcal depth analysis, coordinate distance analysis or gyrification index testing. However, the depth difference index (DDI) ratio (affected hemisphere/unaffected hemisphere) was larger in smaller AVMs and smaller in larger AVMs (both hemispheres had similar DDI). One hypothesis for why this may be occurring is that the larger AVMs are more likely to exert effects on the contralateral as well as ipsilateral hemisphere. However, as with distance from the AVM, the occurrence of significant outliers in this trend suggests that AVM volume is not the only factor influencing sulcal depth. The complexities of the interaction between AVM pathogenesis and cortical development are deserving of further study. Currently, the only mechanistic understanding of AVM pathogenesis comes from the observation of increased intracranial AVM prevalence in the main subtypes of an autosomal dominant syndrome with AVMs in various organs including the brain: Human Hereditary Telangiectasia (HHT1 and HHT2). HHT1 involves a mutation in endoglin (Eng) and HHT2 involves a mutation in actin-like kinase 1 (Alk1), both involved in the transforming growth factor beta (TGF-β) signaling cascade [29, 30]. As both Eng and Alk1 are expressed in endothelial cells, changes in their function and expression affect angiogenesis [31, 32]. A murine knockout of Alk1 needed angiogenic stimulation with VEGF for de novo formation of brain AVMs , potentially suggesting a “two-hit” model of AVM formation and explaining the relatively few congenital AVM cases. The vast majority of brain AVMs are present in the absence of HHT and there are no definitive human tissue studies implicating Alk1 in brain AVM formation [34, 35], leaving a major gap in the understanding of AVM pathogenesis.
Our results may support the de novo formation theory, as otherwise cortical folding differences would potentially be more evident. However, with only 7 patients at a single time instance, more longitudinal data is needed for such a conclusion. We originally hypothesized that sulcal depth differences would be confined to the cortex adjacent to the AVM. The AVM should provide local mechanical traction to folding and also a relative lack of perfusion caused by the absence of normal intervening capillaries. We were unable to define a spatial relationship of cortical folding differences caused by underlying AVMs, an important result that may suggest a decoupling of cortical folding and AVM pathogenesis. This lack of spatial relationship could be due to a much wider-spread pattern of cortical disruption, but it could also be that folding and AVM formation processes are independent.
Previous studies  have shown reorganization in eloquent areas overlying AVMs, which could have its basis in cortical folding but also in post-stroke cortical plasticity mechanisms that are also poorly understood . Additionally, it is possible that extreme sulcal depth differences predispose AVM patients to ipsilateral seizure onset during the initial presentation in nearly 40 % of AVM patients , as many epilepsy etiologies are structural . However, every healthy individual has a distinct sulcal depth pattern and our comparisons employ the population mean. Further testing of sulcal depth and functional changes could further elucidate this relationship, especially if performed longitudinally in our AVM patient cohort.
Our study might be simply underpowered to determine such a relationship with only six subjects with supratentorial AVMs. At our hypothesized moderate 50 % effect size, we would have needed 21 patients to have 80 % power to avoid a type II error. In addition to limiting the power of our study of sulcal depth differences, none of these patients presented with hemorrhage or focal neurological deficit. It is possible that these sequelae are in some way related to the minor folding differences we observed, but a larger study would be needed to determine what, if any, connection there is between the two.
This is the first study to quantitatively assess human developmental cortical folding in the presence of intracranial arteriovenous malformations. The study found no statistically significant cortical folding differences in the patient hemispheres with AVMs compared to their own contralateral hemispheres or compared to a previously obtained healthy, control dataset. However, longitudinal studies are recommended to definitively establish whether arteriovenous malformations are developmental, congenital lesions or dynamic, de novo entities.
DCVE is supported by grant National Institutes of Health (NIH) R01-MH-60974.
Availability of data and material
Please contact the authors for data requests.
MNS and SES are co-first authors and both equally designed the study, performed the analysis, analyzed the results and drafted the manuscript. DLD performed the analysis, analyzed results and helped draft the manuscript. JPH helped analyze results and draft the manuscript. TSC helped perform the analysis, analyze results and draft the manuscript. BSB helped collect data and draft the manuscript. GJZ helped analyzing results and drafting the manuscript. DCVE helped with study design, result analysis and manuscript drafting. RGD conceived and designed the study, analyzed results and helped draft the manuscript. All authors read and approved the final manuscript.
The authors declare that they have no competing interests.
Consent for publication
Ethics approval and consent to participate
This study was approved by the Washington University in St. Louis Institutional Review Board, IRB ID# 201107259. All participants willingly provided informed consent for the study.
DCVE is supported by grant National Institutes of Health (NIH) R01-MH-60974.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
- McCormick WF. The pathology of vascular (“arteriovenous”) malformations. J Neurosurg. 1966;24:807–16.View ArticlePubMedGoogle Scholar
- Moftakhar P, Hauptman JS, Malkasian D, Martin NA. Cerebral arteriovenous malformations. Part 2: physiology. Neurosurg Focus. 2009;26:E11.View ArticlePubMedGoogle Scholar
- Budday S, Steinmann P, Kuhl E. Physical biology of human brain development. Front Cell Neurosci. 2015;9:257.View ArticlePubMedPubMed CentralGoogle Scholar
- Marin-Padilla M. The human brain intracerebral microvascular system: development and structure. Front Neuroanat. 2012;6:38.View ArticlePubMedPubMed CentralGoogle Scholar
- Menshawi K, Mohr JP, Gutierrez J. A Functional Perspective on the Embryology and Anatomy of the Cerebral Blood Supply. J Stroke. 2015;17:144–58.View ArticlePubMedPubMed CentralGoogle Scholar
- Nelson Jr MD, Gonzalez-Gomez I, Gilles FH. Dyke Award. The search for human telencephalic ventriculofugal arteries. AJNR Am J Neuroradiol. 1991;12:215–22.PubMedGoogle Scholar
- Chaloupka JC, Vinuela F, Robert J, Duckwiler GR. An in vivo arteriovenous malformation model in swine: preliminary feasibility and natural history study. AJNR Am J Neuroradiol. 1994;15:945–50.PubMedGoogle Scholar
- Kashba SR, Patel NJ, Grace M, Lee VS, Raoufi-Rad N, Amal Raj JV, Duong TT, Stoodley M. Angiographic, hemodynamic, and histological changes in an animal model of brain arteriovenous malformations treated with Gamma Knife radiosurgery. J Neurosurg. 2015;123:954-60.
- Mut M, Oge K, Zorlu F, Undeger U, Erdem S, Ozcan OE. Effects of ionizing radiation on brain tissue surrounding arteriovenous malformations: an experimental study in a rat caroticojugular fistula model. Neurosurg Rev. 2004;27:121–7.View ArticlePubMedGoogle Scholar
- Qian Z, Climent S, Maynar M, Uson-Garallo J, Lima-Rodrigues MA, Calles C, Robertson H, Castaneda-Zuniga WR. A simplified arteriovenous malformation model in sheep: feasibility study. AJNR Am J Neuroradiol. 1999;20:765–70.PubMedGoogle Scholar
- Van Essen DC, Dierker D, Snyder AZ, Raichle ME, Reiss AL, Korenberg J. Symmetry of cortical folding abnormalities in Williams syndrome revealed by surface-based analyses. J Neurosci. 2006;26:5470–83.View ArticlePubMedGoogle Scholar
- Nordahl CW, Dierker D, Mostafavi I, Schumann CM, Rivera SM, Amaral DG, Van Essen DC. Cortical folding abnormalities in autism revealed by surface-based morphometry. J Neurosci. 2007;27:11725–35.View ArticlePubMedGoogle Scholar
- Dubois J, Benders M, Borradori-Tolsa C, Cachia A, Lazeyras F, Ha-Vinh Leuchter R, Sizonenko SV, Warfield SK, Mangin JF, Huppi PS. Primary cortical folding in the human newborn: an early marker of later functional development. Brain. 2008;131:2028–41.View ArticlePubMedPubMed CentralGoogle Scholar
- Hamel E. Perivascular nerves and the regulation of cerebrovascular tone. J Appl Physiol (1985). 2006;100:1059–64.View ArticleGoogle Scholar
- Homan RW, Devous Sr MD, Stokely EM, Bonte FJ. Quantification of intracerebral steal in patients with arteriovenous malformation. Arch Neurol. 1986;43:779–85.View ArticlePubMedGoogle Scholar
- Meyer B, Schaller C, Frenkel C, Schramm J. Physiological steal around AVMs of the brain is not equivalent to cortical ischemia. Neurol Res. 1998;20 Suppl 1:S13–7.View ArticlePubMedGoogle Scholar
- Meyer B, Schaller C, Frenkel C, Ebeling B, Schramm J. Distributions of local oxygen saturation and its response to changes of mean arterial blood pressure in the cerebral cortex adjacent to arteriovenous malformations. Stroke. 1999;30:2623–30.View ArticlePubMedGoogle Scholar
- Garcia JH, Lassen NA, Weiller C, Sperling B, Nakagawara J. Ischemic stroke and incomplete infarction. Stroke. 1996;27:761–5.View ArticlePubMedGoogle Scholar
- Saur D, Buchert R, Knab R, Weiller C, Rother J. Iomazenil-single-photon emission computed tomography reveals selective neuronal loss in magnetic resonance-defined mismatch areas. Stroke. 2006;37:2713–9.View ArticlePubMedGoogle Scholar
- Yamauchi H, Kudoh T, Kishibe Y, Iwasaki J, Kagawa S. Selective neuronal damage and borderzone infarction in carotid artery occlusive disease: a 11C-flumazenil PET study. J Nucl Med. 2005;46:1973–9.PubMedGoogle Scholar
- Fierstra J, Poublanc J, Han JS, Silver F, Tymianski M, Crawley AP, Fisher JA, Mikulis DJ. Steal physiology is spatially associated with cortical thinning. J Neurol Neurosurg Psychiatry. 2010;81:290–3.View ArticlePubMedGoogle Scholar
- Glasser MF, Van Essen DC. Mapping human cortical areas in vivo based on myelin content as revealed by T1- and T2-weighted MRI. J Neurosci. 2011;31:11597–616.View ArticlePubMedPubMed CentralGoogle Scholar
- Dierker DL, Feczko E, Pruett Jr JR, Petersen SE, Schlaggar BL, Constantino JN, Harwell JW, Coalson TS, Van Essen DC. Analysis of cortical shape in children with simplex autism. Cereb Cortex. 2015;25:1042–51.View ArticlePubMedGoogle Scholar
- Rorden C, Brett M. Stereotaxic display of brain lesions. Behav Neurol. 2000;12:191–200.View ArticlePubMedGoogle Scholar
- Van Essen DC. A Population-Average, Landmark- and Surface-based (PALS) atlas of human cerebral cortex. Neuroimage. 2005;28:635–62.View ArticlePubMedGoogle Scholar
- Van Essen DC, Smith SM, Barch DM, Behrens TE, Yacoub E, Ugurbil K, Consortium WU-MH. The WU-Minn Human Connectome Project: an overview. Neuroimage. 2013;80:62–79.View ArticlePubMedPubMed CentralGoogle Scholar
- Hill J, Dierker D, Neil J, Inder T, Knutsen A, Harwell J, Coalson T, Van Essen D. A surface-based analysis of hemispheric asymmetries and folding of cerebral cortex in term-born human infants. J Neurosci. 2010;30:2268–76.View ArticlePubMedPubMed CentralGoogle Scholar
- Faul F, Erdfelder E, Buchner A, Lang AG. Statistical power analyses using G*Power 3.1: tests for correlation and regression analyses. Behav Res Methods. 2009;41:1149–60.View ArticlePubMedGoogle Scholar
- Marchuk DA, Srinivasan S, Squire TL, Zawistowski JS. Vascular morphogenesis: tales of two syndromes. Hum Mol Genet. 2003;12 Spec No 1:R97–R112.View ArticlePubMedGoogle Scholar
- Kim H, Marchuk DA, Pawlikowska L, Chen Y, Su H, Yang GY, Young WL. Genetic considerations relevant to intracranial hemorrhage and brain arteriovenous malformations. Acta Neurochir Suppl. 2008;105:199–206.View ArticlePubMedPubMed CentralGoogle Scholar
- Cheifetz S, Bellon T, Cales C, Vera S, Bernabeu C, Massague J, Letarte M. Endoglin is a component of the transforming growth factor-beta receptor system in human endothelial cells. J Biol Chem. 1992;267:19027–30.PubMedGoogle Scholar
- Lamouille S, Mallet C, Feige JJ, Bailly S. Activin receptor-like kinase 1 is implicated in the maturation phase of angiogenesis. Blood. 2002;100:4495–501.View ArticlePubMedGoogle Scholar
- Chen W, Sun Z, Han Z, Jun K, Camus M, Wankhede M, Mao L, Arnold T, Young WL, Su H. De novo cerebrovascular malformation in the adult mouse after endothelial Alk1 deletion and angiogenic stimulation. Stroke. 2014;45:900–2.View ArticlePubMedPubMed CentralGoogle Scholar
- Hashimoto T, Wu Y, Lawton MT, Yang GY, Barbaro NM, Young WL. Coexpression of angiogenic factors in brain arteriovenous malformations. Neurosurgery. 2005;56:1058–65. discussion 1058-1065.PubMedGoogle Scholar
- Shenkar R, Elliott JP, Diener K, Gault J, Hu LJ, Cohrs RJ, Phang T, Hunter L, Breeze RE, Awad IA. Differential gene expression in human cerebrovascular malformations. Neurosurgery. 2003;52:465–77. discussion 477-468.View ArticlePubMedPubMed CentralGoogle Scholar
- Alkadhi H, Kollias SS, Crelier GR, Golay X, Hepp-Reymond MC, Valavanis A. Plasticity of the human motor cortex in patients with arteriovenous malformations: a functional MR imaging study. AJNR Am J Neuroradiol. 2000;21:1423–33.PubMedGoogle Scholar
- Liepert J, Miltner WH, Bauder H, Sommer M, Dettmers C, Taub E, Weiller C. Motor cortex plasticity during constraint-induced movement therapy in stroke patients. Neurosci Lett. 1998;250:5–8.View ArticlePubMedGoogle Scholar
- da Costa L, Wallace MC, Ter Brugge KG, O’Kelly C, Willinsky RA, Tymianski M. The natural history and predictive features of hemorrhage from brain arteriovenous malformations. Stroke. 2009;40:100–5.View ArticlePubMedGoogle Scholar
- Bhalla D, Godet B, Druet-Cabanac M, Preux PM. Etiologies of epilepsy: a comprehensive review. Expert Rev Neurother. 2011;11:861–76.View ArticlePubMedGoogle Scholar
- Kim H, Su H, Weinsheimer S, Pawlikowska L, Young WL. Brain arteriovenous malformation pathogenesis: a response-to-injury paradigm. Acta Neurochir Suppl. 2011;111:83–92.View ArticlePubMedPubMed CentralGoogle Scholar
- Lasjaunias P. A revised concept of the congenital nature of cerebral arteriovenous malformations. Interv Neuroradiol. 1997;3:275–81.PubMedGoogle Scholar
- Morales-Valero SF, Bortolotti C, Sturiale C, Lanzino G. Are parenchymal AVMs congenital lesions? Neurosurg Focus. 2014;37:E2.View ArticlePubMedGoogle Scholar
- Fierstra J, Maclean DB, Fisher JA, Han JS, Mandell DM, Conklin J, Poublanc J, Crawley AP, Regli L, Mikulis DJ, Tymianski M. Surgical revascularization reverses cerebral cortical thinning in patients with severe cerebrovascular steno-occlusive disease. Stroke. 2011;42:1631–7.View ArticlePubMedGoogle Scholar