Skip secondary menu

Biology of Vascular Malformations of the Brain Workshop

March 13-14, 2008
Embassy Suites at the Chevy Chase Pavilion, Washington, DC
Organizer: Dr. Eugene Golanov
Co-chairs: Drs. Issam Awad and William Young
Co-Sponsors: Angioma Alliance, HHT Foundation International, The Aneurysm
and AVM Foundation, Sturge-Weber Foundation, and UCB Pharma


Vascular malformations of the brain (VMBs) affect over 3 million Americans, and cause serious neurological disability or death in a significant proportion of individuals bearing them.  VMBs include arteriovenous malformations (AVMs), cerebral cavernous malformation (CCMs), and the brain AVMs of hereditary hemorrhagic telangiectasia (HHT).  Neurological symptoms associated with these lesions include hemorrhagic stroke due to lesion rupture, epilepsy, focal motor and sensory deficits, and cognitive impairment.  However, lesions visible by imaging can also remain clinically silent for an individual’s lifetime.  It is currently impossible to determine which individuals with lesions will go on to develop symptoms and which will not, nor do medical treatments exist to reduce the risk of lesion rupture.  The only interventions now available are surgical excision, endovascular embolisation, and/or stereotactic radiotherapy of selected lesions that have already become clinically problematic.

Most AVMs are sporadic, but a substantial proportion of cases of CCM and HHT are familial.  The genetic mutations responsible for certain subpopulations of CCM and HHT cases have now been discovered, along with a handful of genetic risk factors for AVMs.  The identification of these genes has provided the first insights into potential molecular mechanisms underlying VMBs, and has also enabled the development of the first animal models for these diseases.  Studies in these models, together with clues obtained elsewhere about the normal functions of VMB genes, increasingly support the idea that VMBs arise from defects in processes of vasculogenesis (de novo blood vessel formation), angiogenesis (blood vessel formation from preexisting vasculature), or vascular maintenance and remodeling.  Thus, further study of the pathogenic mechanisms occurring in VMBs will likely provide new insights into the basic biology of vascular development, as well as pointing the way to future therapies. 

Many of the ground breaking discoveries in VMB research, including realization of the critical interrelations between vascular and neural and glial cells, have occurred just in the past few years, and there has been relatively little opportunity so far for scientists studying different VMBs to interact with one another or with scientists studying the basic mechanisms of angiogenesis.  The goal of this workshop was to bring these researchers together to discuss common scientific themes and clinical challenges, and to identify critical questions and resources for future research.  The workshop focused on CCM and AVM, but there are conceptually similar issues that apply to related vascular anomalies of the brain including dural arteriovenous fistula (DAVF), capillary malformations, and the various mixed types that are sometimes encountered.


Vasculogenesis and Angiogenesis

Vasculogenesis of the cerebral vasculature occurs outside the brain with the formation of a perineural plexus, from which capillary sprouts penetrate the neural tube.  The subsequent elaboration of the cerebral vasculature occurs by angiogenesis.  Cerebral angiogenesis is tightly linked to the proliferation and growth of neurons and glia, and appears to be mediated at least in part by hypoxia-inducible transcription factors (HIFs).  HIFs up-regulate the production of vascular endothelial growth factor (VEGF), which in turn stimulates endothelial cell proliferation, migration, and survival.  Angiogenesis is down-regulated once the vascular bed is established (shortly after birth), but proceeds at a slower pace throughout life.  In animal models, for example, angiogenesis has been observed in the healthy adult brain in response to a variety of stimuli that increase neural activity, including exercise, exposure to enriched sensory environments, and certain hormones.  Brain angiogenesis can also undergo dramatic local up-regulation in response to disease states such as brain tumor, stroke, or trauma. 

VMBs are characterized by activated angiogenesis. Sporadic brain AVMs and those occurring in the context of HHT have the additional pathological feature of lacking capillary beds (resulting in direct arteriovenous shunts), while CCMs are characterized by dilated dysmorphic capillaries.  Thus, it seems likely that defects in vasculogenic and/or angiogenic processes underlie or contribute to all of these disease states.  However, the precise nature of these defects remains to be discovered.  In addition, it remains to be determined how VMB gene products normally interact with the cellular pathways controlled by key angiogenic regulators such as VEGF and the angiopoietins, and how their VMB gene mutations perturb those interactions.

Critical issues:

  • Define the extent of ongoing angiogenesis in normal adult brain
  • Understand to what extent the mechanisms that control developmental angiogenesis remain operative during adulthood, and what adult-specific mechanisms may come into play
  • More precisely define the angiogenic processes that occur in VMBs:  what angiogenic factors and receptors are up- or down-regulated, what cell types are involved?
  • Determine the mechanisms leading to capillary loss in brain AVMs.
  • Understand how the functional pathways mediated by VMB gene products (endoglin, Alk-1, Ccms) interact with those regulating angiogenesis and vascular patterning, stability, and maintenance
  • Understand how VMB genes interact with pathways regulating blood-brain barrier formation and vascular permeability

Perivascular microenvironment

Angiogenesis is actively regulated by surrounding tissue cells, and also by resident and circulating immune cells.  Angiogenic processes in the CNS are likely to differ from those in other organs, not only because of the specialized parenchymal cell types involved, but also because brain vascular endothelial cells themselves have unique properties.  Most prominent of these are (1) the presence of tight junctions and (2) the formation of specialized contacts with astrocyte endfeet.  These morphological features contribute to the special permeability properties of blood-brain barrier (BBB), and appear to be disrupted in CCMs. 

Formation and maintenance of the BBB requires cross-talk between the cells of the neurovascular unit (i.e., vascular endothelial cells, neurons, astrocytes, and pericytes), much of which appears to be mediated by cell adhesion molecules and molecules of the extracellular matrix.  Reciprocal communication between the cells of the neurovascular unit regulates not only BBB formation, but also angiogenesis and neurogenesis.  In stroke, for example, there is recruitment of new endothelial cells to ischemic tissue, and new vessels in turn attract neural precursors.  Hence, it is possible that defects in signaling between the cells of the neurovascular unit contribute to the development of VMBs, and also to their adverse impact on surrounding neural tissue. 

A fourth population of cells likely to be important in VMB pathogenesis is immune cells.  VMB lesions frequently show marked infiltrations of inflammatory cells and local up-regulation of inflammatory cytokines, some of which are powerful regulators of angiogenesis and vascular permeability.  In addition, polymorphisms in cytokine genes have been associated with increased risk of hemorrhage for sporadic brain AVMs.

Critical issues:

  • Characterize the state of the blood-brain barrier (BBB) in CCMs and the perinidal capillary beds of brain AVM.  In particular, assess:
    • Possible morphological alterations in tight junctions or astrocyte endfeet contacts with endothelium
    • Molecular changes in the neurovascular unit or surrounding extracellular matrix (for example, by proteomic analysis)
    • Regulation of tight junction formation and BBB permeability by VEGF and other angiogenic factors
  • Determine whether endothelial cell cross-talk with neurons and astrocytes is perturbed in VMBs
  • See if neuroblasts selectively localize to VMBs (as they do to new vessels formed after stroke)
  • More precisely characterize the types of inflammatory cells found around VMBs, what cytokines they produce, and the potential role of antigenic triggers in lesional inflammation
  • Determine whether VMBs have unique or abnormal inflammatory responses, including their responses to cytokines
  • Understand the effects of inflammatory cytokines on brain angiogenesis.

Basic biology of VMB gene products

Some clues are already available about the functions of genes whose mutation leads to CCM and HHT.  For example CCM1 (also known as KRIT-1) appears to be involved in the stabilization of integrin-mediated inter-endothelial cell junction formation.  CCM2 (or OSM) is involved in cellular responses to hyperosmotic stress, and appears to regulate both cytoskeletal dynamics and endothelial barrier properties.  CCM3 (or Programmed Cell Death 10, PDCD10) has anti-apoptotic and cell growth promoting activities in fibroblast and tumor cell lines. The two most prevalent HHT genes discovered so far, Endoglin and ALK-1, are both components of the TGFb super-family signaling system; ALK-1 polymorphisms have also been associated with hemorrhage risk in sporadic brain AVM.  All these genes doubtless have other biological functions in addition to the ones identified so far.  Moreover, in all cases the precise roles of these genes in brain vascular development and maintenance needs to be determined, and the ultimate question of which of their functions are disrupted in VMBs remains to be answered.

Critical issues for CCMs:

  • Understand how Ccm1, 2, and 3 control the stability of endothelial tight junctions, and which of their many binding partners participate in this process
  • Learn more about how the expression and activity of the Ccm proteins are regulated (e.g., by phosphorylation).
  • Determine how Ccm1, 2, and 3 cooperate with one another and with the integrins
  • Understand the dynamics of Ccm1, 2, and 3 interactions with each other and their other binding partners over time
  • Define the functions of nuclear Ccm1 and Ccm2
  • Understand the roles of Ccms in non-endothelial cells (i.e., neurons glia, immune cells)
  • Understand the role of the Ccms in vascular development and BBB formation.

Critical issues for Endoglin and Alk-1:

  • Further define the roles of endoglin and Alk-1 in vasculogenesis, angiogenesis, and vascular maintenance
  • Determine which TGFb super-family signaling pathways these receptors interact with:  i.e., what are their physiological ligands and downstream effectors?
  • Understand how endoglin and Alk-1 interact with vascular patterning and adhesion molecules
  • Determine how endoglin and Alk-1 expression are affected by hypoxia, hyperoxia, sheer stress, and inflammatory factors
  • Learn how endoglin or Alk-1 deficiency causes the loss of a capillary network and nidus formation that characterizes AVMs.

Pathobiology of VMBs

Much basic descriptive work still needs to be done to better understand when and in what form the lesions characteristic of different VMBs first appear during the lifespan in humans, and the paths by which they progress to rupture.  More extensive longitudinal imaging and clinical studies are needed, as well as more systematic pathological analyses.  It will also be important to develop animal models that more accurately recapitulate the human disease states, with regard not only to expression of appropriate histopathological features but also time course of lesion formation.  Much also remains to be learned about genetic risk factors for VMB, particularly with regard to genetic modifiers that may influence clinical outcome.  Finally, virtually nothing is known at present about the influence of environmental risk factors on VMB expression and progression.

Cellular mechanisms

  • Characterize the developmental time course of appearance of VMBs in humans.  Do they exist in utero and, if so, can they be detected?
  • Explore the role of environmental factors (e.g., pregnancy, hormones, vascular risk factors, anti-platelet and other drugs) in VMB appearance, progression, and clinical risk profile
  • Better understand the stages of formation of individual VMBs.  Are some static and some dynamic? 
  • Understand which cell types are primarily affected in different VMBs.
  • Determine if and how lesion pathogenesis may be different in sporadic versus inherited VMBs
  • Identify the processes either intrinsic or extrinsic to individual VMBs that cause them to proliferate and to induce clinical complications such as hemorrhage
  • Understand how hemodynamic and structural factors may affect lesion development and behavior
  • Determine if VMBs develop preferentially in certain topographic brain regions or in specific territories of the brain vasculature
  • Understand if and how inflammation modulates lesion behavior, and how this modulation may differ in different VMB types
  • Determine if and how the factors that trigger lesion genesis differ from those responsible for lesion proliferation and hemorrhage
  • Define which biological functions of the VMB gene products (CCMs, endoglin, and Alk1) are disrupted by mutations that cause VMBs
  • Define which cell types may be hypoxic or hyperoxic in and around different VMBs
  • Determine if altered flow in VMBs initiates inflammatory responses
  • Characterize microglial and mast cell behavior around VMBs
  • Evaluate the possibility of autoimmune responses to neural antigens exposed by open BBB
  • Define the mechanisms whereby secondary recruitment of feeders occurs in AVMs, and how new caverns form in CCMs
  • Understand whether the perinidal dilated capillary bed in AVMs is primary or secondary to the disease process.


  • Identify other genes involved in inherited VMBs, including low penetrance alleles
  • Evaluate the possibility of somatic mutations in sporadic VMBs
  • Identify additional genetic risk factors for sporadic VMBs
  • Understand how (1) type of VMB gene mutation and (2) genetic modifiers affect disease penetrance, age of onset, numbers of lesions, nature of symptoms, and other aspects of clinical behavior
  • Identify genetic modifiers that influence clinical outcomes, including possible protective as well as aggravating factors
  • Understand how inflammatory genotypes interact with VMB gene mutations
  • Identify endophenotypes that can be readily quantified for genetic studies (number, size, location of lesion?  Hemorrhaging or not?  Other clinical or MRI criteria?)
  • Identify genetic modifiers responsible for the organ-specific expression of HHT lesions, and the biological underpinning of this organ-specific of lesions caused by HHT1 vs. HHT2 mutations
  • Understand why CCM gene mutations seem to affect only the brain vasculature
  • Develop non-rodent models (zebrafish, etc.) to screen for modifiers of VMB genes.

Animal models

  • Develop animal models in which VMBs develop in heterozygous background, and in which all stages of VMB progression and behavior can be examined
  • Develop animal models with higher phenotypic penetrance and (for HHT and sporadic brain AVM) appropriate organ-specific expression
  • Explore the use of cell type-specific and temporally inducible gene constructs to develop new models
  • Develop non-rodent models (Drosophila, zebrafish, C. elegans).

Clinical Diagnosis, Management and Cure
Clinical management of VMBs is currently limited to (1) identification of patients at risk based on imaging and/or genetic tests, (2) neurological treatment of associated symptoms (such as headaches and seizures) and (3) surgical excision, radiotherapy, or in the case of AVM and HHT, endovascular therapy of lesions (often only after rupture or other serious related symptoms).  Clinicians currently cannot offer well-grounded individual prognoses or advice about risk factor management to patients with identified lesions, nor can they reliably predict which existing lesions are likely to rupture. Furthermore, only scant data exist comparing the natural history risk versus the potential benefit of interventional techniques, particularly for VBMs without hemorrhagic complications. Thus, much effort needs to be directed toward developing more refined diagnostic tools and prognostic measures, as well as more sophisticated techniques for lesion removal and methods for preventing de novo appearance of lesions and their progression to rupture.

  • Develop standardized methods and diagnostic criteria to quantify disease severity
  • Better understand the effects of vascular risk factors, anticoagulants, ASA, NSAIDs, vasoactive compounds and alcohol on VMB progression (e.g., through longitudinal natural history studies)
  • Foster longitudinal studies comparing the risks and benefits of brain VBM natural history versus interventional management regarding hemorrhage, seizures, headaches and long-term neurological outcome (e.g. through population-based surveys or controlled clinical trials).
  • Explore the possibility of measuring individual coagulation differences to predict outcomes
  • Determine if focused beam radiosurgery is useful for treating CCMs, and whether various doses of irradiation affect de novo lesion genesis
  • Identify targets for imaging and treatment, such as lesion-specific cell surface markers
  • Identify markers for VMBs that predict risk for hemorrhage and risk-stratification for treatment
  • Develop methods for imaging potential low flow states in brain regions surrounding VMBs
  • Develop animal models of VMBs suitable for testing treatments
  • Explore the possibility of using anti-angiogenic agents, anti-inflammatory agents, pro-coagulants, or MMP inhibitors as treatment for VMBs, including strategies for focused treatment delivery.

Resources and Administrative

Because patients with clinically manifest VMBs are relatively uncommon, multi-center collaborations will be absolutely critical to collecting sufficient patients for meaningful genetic and clinical studies.  Similarly, shared genetic and tissue banks and genetic and clinical databases will be key to well-powered genetic and natural history studies and clinical trials.  Finally, understanding the results of these analyses and using them to generate new research and treatment approaches will be enhanced by further improving communication between basic and clinical researchers studying different classes of VMBs, as well as their interactions with the angiogenesis research community.

Critical issues:

  • Increase communication of VMB research communities with each other and with basic scientists studying angiogenesis
  • Improve collaboration to develop large prospective patient cohorts for genetic and clinical studies, ideally in a setting of population-based surveys or controlled clinical trials
  • Develop a bank of human VMB lesion tissue for molecular and cellular studies
  • Establish large human cohorts to better characterize natural history and for genetic studies and future therapeutic trials
  • Develop a database of longitudinal imaging and other clinical data for genetic modifier studies
  • Explore use of existing funded clinical trials of VMB therapeutics such as TEAM (Trial of Endovascular Aneurysm Management) and ARUBA (A Randomized trial of Unruptured BAVM) to leverage existing infrastructures to collect human blood and tissue specimens for observational studies of VMBs
  • Optimize interactions between research community and organized Patient Support Organizations for both patient recruitment and patient educational purposes
  • Foster multidisciplinary efforts and coordinated projects.




Last updated February 24, 2010