Historically, the treatment of pediatric cerebrovascular disease, particularly structural lesions such as arteriovenous malformations (AVMs), vein of Galen malformations (VOGMs) and cavernous malformations (CMs), was largely predicated on surgical techniques that had remained nearly unchanged for decades. Although cures were possible, many outcomes were devastating, such as attempts to operate on VOGMs. However, the advent of several disparate technologies within a short window of time has led to a revolution in the management of these conditions. Advances in neuroimaging, endovascular access, and molecular biology - among others - have been critical to improving the outcomes of affected patients.

The pioneering work in cerebrovascular disease has inspired progress in other fields of medicine, but the converse is also true. One example is the complete restructuring of the classification of brain tumors from histology to mutational profiling. In neuro-oncology the use of molecular genetic markers has not only improved diagnostic and prognostic accuracy, but it has created opportunities for the development of targeted therapeutics. The profound success of this novel classification methodology, coupled with a surge of reports defining mutations in AVM, VOGM, and CM, has led to attempts to recapitulate this taxonomy schema in cerebrovascular neurosurgery [25]. In combination, the widespread array of advances in the treatment of pediatric cerebrovascular disease has fundamentally changed the field and this manuscript will highlight the impact of these innovations for AVM, VOGM, and CM.

The advent of the modern era of interventional radiology has ushered in a sea-change in the ability to visualize and treat AVMs and VOGMs. High-quality three-dimensional angiography can now be performed safely in the pediatric population, employing catheters and devices that are small enough for superselective access to individual pedicles in children [30,50]. The creation of detachable-tip and steerable microcatheters have been seminal to this progress, enabling interventional radiologists to reach crucial regions of vascular malformations for diagnostic and therapeutic access [13,27,32]. Complementing this newfound access is the ability to deliver agents capable of shutting down blood supply to the AVM/VOGM. These embolic agents include detachable metal coils (used commonly in the setting of short pedicle catheterization) and liquid embolic agents, n-butyl cyanoacrylate and ethylene-vinyl alcohol (EVOH or EVAL), with the well-known Onyx (EVOH dissolved in dimethyl sulfoxide), with tantalum added as a marker to make it radio-opaque) commonly used in North America, Europe, and Asia [14,39,60,61].

Endovascular approaches can assist with the treatment of AVMs through better visualization for operative or radiation planning, perioperative reduction of high-flow pedicles to make surgery safer or - in some cases, such as VOGMs - complete cure through embolization alone [1,55,62]. Innovation continues in this field, including transvenous access and embolization, flow modulation (in which cardiac output can be altered pharmacologically or via cardiac pacing to attempt safer delivery of embolic agents) and the use of single or multiple intravascular balloons to regulate flow during procedures [15,18,24,28,38,40]. These advances also come with risk, including injury to unaffected vessels during the procedure, off-target embolization, radiation exposure (especially important in smaller children/infants) and the risk of unexpected hemorrhage resulting from re-routing of blood flow post-embolization [4].

Stand-alone embolization as definitive treatment of highflow vascular lesions has increasingly been reported. While this approach is often successful for VOGM (typically requiring multiple rounds of embolization), or single-hole fistulae, it is more controversial with nidal AVMs [4,41,47,48]. Initial data suggests that attempts to cure complex nidal AVMs with embolization alone may actually worsen the bleeding risk and outcomes relative to natural history - and, as such, it is increasingly discouraged in many high-volume centers [20,22,48].

One of the most radical advances in the treatment of VOGM has been the recent report of percutaneous transuterine fetal cerebral embolization [36,37,51]. In utero, the fetus is largely protected from deleterious effects of VOGM shunting due to the presence of maternal blood supply, and the loss of this support at birth with the division of the umbilical cord results in immediate adverse effects to the newborn child. Treatment of the VOGM prior to birth obviates this dangerous transition and - in an initial case - has successfully achieved the goal of establishing normal cardiac and cerebral physiology [37]. While this technique will require further validation and refinement, it serves as an encouraging proof-of-principle to stimulate continued innovation in the field.

While neuro-oncology has led the way in translating benchtop discoveries in molecular biology into clinical practice with genetic classification of tumors and development of targeted therapeutics, cerebrovascular neurosurgery is now positioned as the next frontier most likely to reap the benefits of this approach [8,11,16]. Historically, neurosurgeons, neurologists and interventional radiologists have acted independently, characterizing diseases according to their own observations and practices. However, with the recent growth of multidisciplinary care teams in neurosurgery - arguably most common in pediatric cerebrovascular disease - the identification of shared phenotypes and radiographic features has accelerated the pace of translational research. In particular, cohorts of patients with structural vascular lesions - AVM, VOGM, and CM - have been subjected to intense study, providing insight about the genetic foundations underlying these conditions and laying the groundwork for a novel taxonomy to better illuminate the lineage and development of these malformations [25].

It is becoming increasingly apparent that many AVMs and VOGMs are vein-based lesions, arising in association with mutations affecting one of two canonical pathways - ephrins (EPHB4) and HHT (HHT1, also called ENG; and HHT2, also called ACVRL1) - that direct endothelial differentiation in development [6,16,25,67,68]. In contrast, while CMs also have dysfunctional endothelial cells, there is a distinct family of genes related to capillary formation that drive their growth, including the related CCM genes (CCM1, CCM2, and CCM3), with only limited overlap with the ephrin pathway [12,25,53]. Categorizing the malformations within these separate pathways and connecting specific mutations with sporadic and familial forms of these diseases has helped to improve diagnostic and prognostic capabilities of clinicians.

Familial forms of inherited AVMs - as are found in hereditary hemorrhagic telangiectasia (HHT) - often have germline mutations in ENG/HHT1 or ACVRL1/HHT2 [8,11,25,65]. However, many of the sporadic lesions - the most common in pediatric neurosurgical practice - arise from somatic mutations in the ephrin pathway, usually KRAS or BRAF; downstream drivers of cell invasion and proliferation [25]. Given that ephrin is the cell surface-based receptor upstream of a pathway that subsequently controls the downstream intracellular molecules RASA-1, KRAS, BRAF, and RAS, there is now basis for a biological connection between ephrin and RAS mutations in VOGM, RASA-1 mutations in AVM and AV fistulas and the aforementioned KRAS and BRAF mutated nidal AVMs - demonstrating how mutations in interrelated molecules in a specific pathway can manifest similar phenotypes [7,16,25,67,68].

Similar to AVM, CM has both familial and sporadic forms. The CCM genes - CCM1, CCM2, and CCM3 - can combine as a trimeric complex and can also work independently to affect the cellular pathways driven by MEKK3 and ROCK that are critical to cell-cell adhesion and proliferation [12,25,53]. Given the interaction between these molecules, there are similarities in presentation across mutational profiles, but the unique roles of each gene product helps to explain the clinical heterogeneity of disease seen by physicians and patients. For example, CCM3 has outsized impact on the regulation of MEKK3 through a pathway independent of the CCM trimeric complex. In patients with CCM3 mutations, this manifests clinically as more severe disease, with greater lesional burden and hemorrhages. Consequently, awareness of a CCM3 mutation in a patient provides the clinician with important and potentially actionable prognostic information, while also offering a specific molecular target for developing therapeutics. This clinical scenario presents an excellent example of how genetic characterization in a formal taxonomy can inform clinical practice.

In addition to the value of mutational profiles helping patients and families better understanding the course of disease and the potential role of screening other family members, knowing the underlying molecular driver of disease may help with creation of targeted drug therapies - such as the current trials with mTOR and KRAS inhibitors for AVM, and ROCK inhibitors for CCM [12,25,42,58]. The pharmacologic treatment of vascular lesions - including inhibition of growth, better control of existing malformations and possible prevention of recurrence after therapy - are exciting future areas of research in pediatric cerebrovascular neurosurgery.

Finally, there may be a role for future prognostic and theranostic tools derived from this taxonomic knowledge, such as the emerging field of non-invasive biomarkers [2,52,57]. Proof-of-concept data in clinical studies with these types of biomarkers support the premise that CMs and AVMs can be detected through urine sampling and that levels of biomarkers reliably change in response to treatment [8,53]. If validated in larger trials, non-invasive testing could potentially help to reduce cost and risk of CM, AVM, and VOGM care [2]. An example of case use might be to reduce the number of imaging studies needed (along with the concomitant fees and sedation often needed in children) while providing a complementary method of ascertaining treatment response after surgery or radiation. It is exciting to consider the use of biomarkers and targeted therapeutics within the context of an emerging molecular taxonomy as the next step in the evolution of care for pediatric cerebrovascular lesions [2,8,52,56,57].

Pediatric cerebrovascular lesions such as AVM, VOGM and CM remain among the most challenging conditions in neurosurgery. Collaboration across clinical disciplines coupled with dedicated translational research are vital to the development of better outcomes for children afflicted with these diseases. Consequently, this issue of the Journal of Korean Neurosurgical Society is vitally important as it creates a forum to share and promote knowledge critical to the advancement of this field.