Leveraging a Sturge-Weber Gene Discovery: An Agenda for Future Research

INTRODUCTION

The National Institutes of Health (NIH) sponsored a workshop on April 19–20^th in Bethesda, Maryland that convened a diverse group of clinical and translational researchers with a goal of discussing and agreeing upon a research agenda for the next few years. The need for this workshop was highlighted by the recent discovery of the somatic mosaic mutation in GNAQ which underlies both isolated port-wine birthmarks (PWBs) and Sturge-Weber syndrome (SWS).¹ This discovery has catapulted the field forward in ways that necessitated this workshop for the formation of new collaborations and the setting of research priorities for the optimal use of resources and focus. The Organizing Committee (see Appendix) envisioned a workshop focused on presenting the recent updates and gaps in our knowledge in a multi-disciplinary fashion and with an approach that would encourage the participation of attendees, trainees, and young investigators. The morning session of the first day was spent on a series of talks carefully selected to provide the participants with a brief review of the salient features of SWS and an overview of what is now known about the somatic mutation in GNAQ and the pathogenesis of SWS. By bringing together translational researchers and clinician researchers already involved with SWS and those with expertise in other biomedical fields related to these molecular pathways, the workshop enabled novel interactions and discussions around SWS.

In the afternoon of the first day, the participants attended breakout sessions in neurology, ophthalmology, or dermatology. Presentations were followed by a 90 minute discussion by attendees of the session, moderated by the chair with the goal of identifying several main priorities to bring to the group. The results of the breakout sessions were presented by the session chairs the following day and discussed by the entire group. The priorities identified by all three groups were noted and steps needed to address these research priorities were discussed. The workshop ended with a session on the steps needed to move clinical drug trials forward for the discovery of new and effective treatments for SWS. Email discussions, which followed the meeting and are summarized here, were centered on the four research priorities identified: clinical consensus, Clinical Trials Network, tissue banking, and animal and cell culture model development. Here we present a summary of the proceedings from this workshop and of the discussions that followed.

Sturge-Weber syndrome and GNAQ

SWS has long been suspected to result from a somatic mutation.^2;3 In 2013 a somatic nonsynonymous single-nucleotide variant (c.548G→A, p.Arg183Gln) in GNAQ was identified.¹ This R183Q mutation is associated with most of the Sturge-Weber syndrome tissue and isolated port-wine birthmark samples tested. The GNAQ gene codes for the protein Gαq, which is part of the trimeric G protein (Guanine Nucleotide Binding Protein) associated with a subset of the G protein coupled receptors (GPCRs). When activated by the GPCR ligand, Gαq binds GTP and releases GDP, dissociates from the trimeric protein complex, and activates downstream pathways. Hydrolysis of GTP to GDP and re-association of the trimeric G protein with the GPRC results in inactivation of these pathways.⁴ The R183Q mutation in GNAQ is predicted to result in a protein with impaired auto-hydrolysis of activated Gαq, and therefore impaired inactivation of Gαq. The current understanding and data suggests that the mutation results in hyper-activation of downstream pathways, which include RAS-MEK-ERK, HIPPO-YAP,⁵ and, indirectly, mTOR (Figure 1). Some evidence of this constitutive hyper-activation of downstream pathways has been demonstrated in cells transiently transfected with the R183Q mutation.¹ In uveal melanocytes, the R183Q and the Q209L mutation in GNAQ results in uveal melanoma.⁵

Figure 1.

Diagram of molecular pathways probably upregulated by the SWS/port-wine birthmark R183Q mutation in GNAQ. Evidence for this from cells transfected with the mutation^1;5 and from immunohistochemistry in human tissue.⁵⁴

This new knowledge and impetus for this workshop holds promise for targeted treatments aimed at blocking these over-activated pathways, and is focused on identifying the most pressing goals for SWS research.

FROM THE BREAKOUT SESSIONS

Neurology

The neurology breakout session focused on three main areas: (1) the clinical difficulties surrounding the diagnosis of brain involvement, (2) the need to identify the optimal windows for effective treatment, and (3) the practical application of the discovery of the somatic mutation in GNAQ to the treatment of the neurological involvement in SWS. Research has demonstrated that magnetic resonance imaging (MRI) with gadolinium contrast may be necessary to diagnose SWS and that post-contrast flair and susceptibility weighted imaging can aid in the sensitivity of this imaging.^6;7 Even so, MRI still lacks sensitivity in very young patients, and this remains a diagnostic issue. Electroencephalography (EEG) and quantitative EEG have both shown promise as biomarkers to aid in the timing of neuroimaging.⁸ However, these studies, as well as the precise timing of MRI for the optimal diagnosis of SWS brain involvement, require further investigation. The natural history of SWS is highly variable, and most published studies are either cross-sectional or short term longitudinal studies over a few years. Two factors, the extent of brain involvement and the age of seizure onset, have been identified as predictive of neurological outcome and epilepsy severity.^9;10 In addition, the presence of epilepsy and frequency of seizures have been correlated with greater risk for a variety of intellectual, behavioral, and mood concerns.¹¹ Additional studies are needed to better stratify patients into risk groups.

Furthermore, a better understanding of how the mutation results in SWS vascular malformations could aid in the development of new treatments. It would also help our understanding of the optimal timing of treatment. Biomarker development will require access to brain, skin, and eye tissue as well as to blood and urine. One topic area discussed and likely to be important to these efforts involves obtaining a better understanding of the role of inflammation and the immune system in SWS. Another point discussed was the need for an understanding of cell specific responses to the mutation, such as in endothelial cells, pericytes, nerves, etc. Questions surrounding the presymptomatic use of anticonvulsants¹² require further study, and both retrospective and prospective studies were advocated. Low-dose aspirin use remains controversial, with some centers frequently using it¹³ and other centers rarely utilizing it. In addition, the need for surgical centers to publish their long-term outcomes was discussed. All of these issues represent important treatment topics best addressed by a network of clinical research centers working together to achieve clarity.

While several novel treatment approaches have been suggested by the gene discovery, comparing their relative potential is difficult in the absence of animal models for preclinical testing. The need for at least two animal models for this purpose, as well as valid in vitro models for preclinical testing, was agreed to be urgent. Models are currently under development at multiple institutions, and so it is likely that successful models will be available soon. The need for additional tissue studies to better understand the pathology in SWS was also emphasized. An in-depth understanding of the pathology and the molecular histology is necessary in order to validate models and to generate new drug targets.

The ophthalmology discussion group also highlighted difficulties with diagnosis as an area in need of research. Protocols and screening schedules are needed to improve clarity in this area. While it was agreed that the initial eye examination should occur shortly after the identification of a PWB involving the periocular region, better evidence is needed regarding which children are actually at risk for glaucoma, who should be screened, and for how long. Newer imaging techniques, such as optical coherence tomography (OCT),¹⁴ offer some diagnostic utility with regards to choroidal hemangioma, but the timing for initial and periodic screening needs to be identified, as does the point at which screening can be stopped.

Biomarkers, which may include imaging techniques or serum/urine/ocular testing, could be helpful in making treatment decisions, including the best approach to the surgical and medical management of glaucoma. Biomarkers may also be helpful in the management of complications specific to glaucoma surgery in the setting of SWS, such as suprachoroidal hemorrhage. In particular, the role of episcleral venous pressure should be explored as a potential target for medical interventions.

The possibility of developing gene based treatment with a better understanding of the role of GNAQ in the ocular manifestations of SWS requires further investigation. Opportunities may be available to assist in the treatment of glaucoma and choroidal hemangioma in SWS, especially given the dramatic advances and successes of gene therapy in ocular diseases. The eye represents an ideal model for gene therapy. Intravitreal injection, for example, might be efficacious in the treatment of choroidal hemangioma. Although mTOR inhibitors have failed in the treatment of age related macular degeneration, this and other pharmacologic approaches to choroidal hemangioma and glaucoma are important to pursue. While efforts have been made to study the effect of skin laser on the eyeball, and the effect is empirically assumed to be nil, this issue should be further addressed^15;16 and will require collaboration between the two specialties (ophthalmology and dermatology). With respect specifically to choroidal hemangioma, the identification of specific modalities for specific patients in specific circumstances remains a challenge. The delivery systems for treatment and the possible use of metabolomics/proteomics are of particular interest to this area.

To make progress in these areas of SWS ophthalmology research, cell culture and animal models, tissue assay, tissue/intraocular fluid assays, and most importantly, collaborative networks and clinical trials are needed. All of these efforts may be helped by the development of a targeted registry that could perhaps be complemented with the efforts of the Sturge-Weber Foundation’s SWS Registry (https://swsregistry.patientcrossroads.org/), Pediatric Eye Disease Investigator Group (PEDIG),¹⁷ the Childhood Glaucoma Research Network (CGRN; http://www.gl-foundation.org/resources-for-professionals/meet-cgrn-doctors), Robison Harley, MD CGRN International Pediatric Glaucoma Registry (https://harleycgrn.patientcrossroads.org/), or other collaborative networks. Strong support was given for the Children’s Oncology Group (https://www.childrensoncologygroup.org/) model of funded expert consensus meetings that lead to publication. The other processes by which to establish clinical consensus protocols include the Delphi process¹⁸, the NIH model, and other types of expert panels. Such collaborative groups need to focus on identifying examination schedules, testing protocols, referral recommendations, and treatment protocols both for glaucoma and choroidal hemangioma.

Dermatology

The dermatology breakout session focused on the need for future clinical research to improve outcomes for facial PWB, which can range in extent, severity, and associated tissue hypertrophy. The importance of timing of treatment and shift to medical options was discussed as well as the need for collaboration with centers nationally and internationally. A priority for future research should be an emphasis on therapeutic outcomes from surgical, laser, and medical therapy to ablate and prevent progression of the cutaneous lesions. The treatment of the PWB in SWS has advanced significantly over the last several years with improvements in laser.¹⁹ Nevertheless, issues remain surrounding treatment resistant areas of the face, the poor response of birthmarks that have already undergone hypertrophy, and controversies surrounding the optimal approach to and use of sedation for laser treatments. There is some early data on medical intervention with mTOR inhibitors for the management of PWBs following laser treatments,²⁰ however, further work in this area is needed to determine the optimal use of topical or oral agents with laser treatment to prevent vascular re-growth.

Priority Group Recommendations

Four main research priorities were identified as common themes from the three breakout sessions: the need for clinical consensus guidelines, further work in clinical trial network development, recommendations for improvements in tissue banking for research purposes, and the urgent need for the creation of animal and cell culture models of SWS. Participants from the workshop volunteered to continue discussions begun at the workshop on these topics, and to provide more in depth summaries of research goals and resource requirements within each of these priority areas. Table 1 outlines the main points of these goals.

Table 1.

Summary of main research priorities and discussion points

Clinical Consensus	Consensus can be modeled after the Pediatric Oncology Cooperative Groups, which has used a consensus effort to improve care. Consensus would be developed using surveys and expert opinion. The first goal will be consensus on how pediatricians can screen at-risk patients and refer appropriately.
Clinical Trial Network Development	Multicenter collaboration is needed to create larger longitudinal, prospective studies. Goals to be addressed by clinical trial networks: Establishment of common diagnostic and clinical monitoring protocols and identification of treatment- responsive biomarkers. Development and validation of standardized neuroimaging protocols and imaging markers. Identification of the most immediate targets and potential drugs for clinical drug trials
Tissue Banking	Tissue should be linked to clinical information and test results when possible, and “virtual” bank of specimens and data would be available to registered users. Goals to achieve coordinated banking: Enlistment of key personnel at participating institutions to promote timely collection and banking Development of standardized instructions, banking consent forms, clinical data collection forms, and kits for all steps from collection and handling to shipping and storage. A Utilization Committee will help determine what is needed from the banks for any particular study.
Animal and Cell Culture Model Development	Different animal models will be needed to represent different aspects of SWS. The mutant gene that causes SWS must be expressed in a subset of cells in the animal model, such as via use of a conditional allele that can be induced by pharmacological means or targeted viral infection. Cells carrying the mutation could be injected into wild type mice. These cells could be developed by using cell cultures to grow samples taken from vascular lesions, or else by using retroviral/lentiviral expression constructs that introduce the mutation in the appropriate cell types before injection. Three-dimensional co-culture models to examine the tissue microenvironment of SWS lesions and endothelial cells lines that express the GNAQ mutation may be useful for disease modeling and drug testing. Zebrafish models could be created using morpholino-mediated gene knock-down or CRISPR-mediated gene editing in order to identify drug targets and screen drugs.

Clinical Consensus

The need for clinical consensus within the current treatment practice of SWS management and care was identified as a tremendous ongoing research necessity; these consensus recommendations can form a baseline against which to measure future diagnostic or treatment clinical trials. In the past, there have been no major consensus publications for SWS. A review of PubMed for “Sturge Weber” and “consensus” reveals no matches. While there was one small-group “recommendation” article (SWS: Recommendations for Surgery) published in 1994, and there have been numerous “How I Treat” articles published by individual experts in the fields, there remains significant variation in practice between sites and even among practitioners at the same site.

While there is a body of literature dedicated to the art of consensus decision-making, it has not yet been rigorously applied to SWS.²¹ In other medical fields such as pediatric oncology, the ability to achieve consensus on current standard of care has allowed the field to make dramatic improvements in outcomes for pediatric cancer and has significantly affected the quality of life of such patients, resulting in improved survival and decreased morbidity from treatment effects. Using the Pediatric Oncology Cooperative Groups (Pediatric Oncology Group and Children’s Cancer Group, which eventually merged to become the current Children’s Oncology Group) as a model, it should be possible to first establish consensus on current best practice and then improve upon it through cooperative multicenter trials.

The creation of a framework for achieving expert consensus through a consensus process will allow for the rapid generation of current recommendation papers. The first consensus effort planned will be a Guideline for Screening At-Risk Patients (those with facial PWB). This will aim to improve the general pediatrician’s recognition of the birthmark and encourage prompt referrals to the appropriate experts, such as an ophthalmologist, neurologist, and dermatologist/laser specialist. Additional recommendations targeted to specific specialties can be completed in a similar fashion. Table 2 outlines the recommended approach, which should be a practical way to take an identified clinical question from the point of a literature search, to review by a panel of experts, to a survey by a larger group of experts, and back to summarization and publication by the original group that targeted the consensus effort. The surveys will include 5–10 vetted questions addressing the main topic of the consensus effort, 1–2 additional questions proposed by the consensus committee that might allow for accompanying “Letter to the Editor” publication in another subspecialty’s journal, 1–2 questions to assess volume/expertise of responding physician, and an answer option will be available for each “consensus” question which states: “outside my area of expertise, would refer.” Further details on format of the survey responses and the analysis of these responses are seen in Figure 2.

Table 2.

Recommended approach to clinical consensus process

	Number of Participants	Participant Selection	Goals	Expected time frame
1	2–4	Volunteers (provide own topic of interest) OR Nominees (question consensus committee feels needs to be addressed)	-Identification of subject -Review of literature -Draft document to be shared -Draft questions for consensus assessment	3–6 months
2	10–15	Consensus Committee -To include representation from all subspecialties	-Review draft, particularly references to be sure it is complete/non-biased -Suggest additional references as appropriate -Approve/vet proposed consensus survey*	2–4 weeks (Conference call 1 month after initial distribution)
3	30–50	Sturge Weber Expert group -To include attendees from Spring SWS meeting PLUS -To include SWS Center of Excellence Directors (and subspecialty experts from those sites as appropriate to content/questions)	-Respond to electronic survey	1 month until survey closes
4	2–4+	Step 1 participants plus any additional experts recruited in steps 2–3	-Data analysis from survey -PUBLICATION	3 months to submission

Figure 2.

Recommended approach for surveying experts to arrive at clinical consensus by quantifying survey responses and level of consensus.

Clinical Trial Network

SWS is a rare disorder, and non-specialized centers see only a few SWS patients every year from their immediate referral area. The rarity of the syndrome poses several major challenges. These difficulties include lack of patient access to multidisciplinary expertise for the complex management of SWS-related issues, scatter of data on individual patients across multiple clinical centers, limited information flow among non-specialized centers for research collaboration, hampered collection of large enough data sets for more advanced data analysis, and slow single institution clinical drug trials recruitment. The Sturge-Weber Foundation (SWF) rightly recognized this gap to develop more experts in the field and to make it less stressful and financially less draining for families to seek local treatment. The Centers of Excellence across the country now total 18. Among 13 active PWB/SWS trials registered on clinicaltrials.gov (http://clinicaltrials.gov/), most single-institute trials have a low enrollment target (4–25 patients). These low enrollment targets are likely meant to accommodate the small number of patients with a rare disease such as SWS who would visit any individual center.

Role of Patient Advocacy

In recognition of some of these issues, the SWF, founded in 1987, developed a software based patient registry that currently holds data on more than 3,000 individuals with SWS, Klippel-Trenaunay, or isolated PWB. Several seminal research studies have relied on this database to identify patient subgroups for clinical studies that utilized, for example, clinical surveys or questionnaires for registered patients.^22;23 A severe limitation of such a voluntary patient-filled database is that the accuracy of the data entered is difficult to verify. The SWF established an Institutional Review Board (IRB) approved online registry (https://swsregistry.patientcrossroads.org/) to address this issue. The registry has 400 participants with clinical, surgical, medication, and family history data linked to uploaded medical reports and scans. An important step toward more coordinated patient management was the designation of The SWF Sturge-Weber Centers of Excellence (http://www.sturge-weber.org/sws-centers-of-excellence.html). These centers have experts in key medical fields (such as neurology, dermatology, ophthalmology) and can provide comprehensive multi-disciplinary care for SWS patients. Collection of patients from large regions facilitates enhanced expertise and fosters clinical research. In 1999, the SWF co-sponsored with the NIH the first workshop on consensus and research, which set the course for further collaborations. The SWF will continue to have an important role in facilitating future SWS patient-based research by encouraging subject participation and advocating for resources. The SWF announced that it would be funding two fellowships to engage young investigators and develop mentoring partnerships to continue the pace of discovery and care.

As another step towards progress in clinical and research collaboration, a standardized SWS registry and central clinical database has been established by a group of investigators to study SWS within the Brain Vascular Malformation Consortium (BVMC; Project 2: Innovative approaches to gauge progression of SWS, Clinical trial NCT01425944).^24;25 The BVMC is a multidisciplinary, inter-institutional group, one of 22 consortia in the Rare Diseases Clinical Research Network (RDCRN). The SWS BVMC registry has been created through the collaborative efforts of the lead clinical site at the Kennedy Krieger Institute, the SWF, and the recruiting SWF Centers of Excellence. Data are collected and entered in the central registry by co-investigators from each site. This effort includes collaboration with translational research labs at multiple centers. The breakthrough identification of the GNAQ somatic gene mutation in 2013 was one of the initial results of this multicenter collaboration.¹

The discovery of the GNAQ gene mutation and expected follow-up discoveries make it likely that future work will identify new drug targets for clinical trials. Such trials should include multiple centers, utilization of validated methods and common protocols, and a central database for analysis. Despite the promising initial steps outlined above, there are several areas where such multicenter collaboration needs to be established or expanded. Major areas needing development include defining predictors (markers) of poor neurological outcome (such as intractable seizures, focal weakness, intellectual disability, and behavior problems), and this has been a significant goal because of the very wide range in neurological outcome. A major impediment has been the lack of large, reliable, quantified longitudinal data. The BVMC trial has been collecting patient and family history as well as basic clinical data, both longitudinal and cross-sectional, since its inception in 2010. Biomarker development has also been a focus of this work with pilot biomarker studies completed and published, which have included the development of quantitative EEG^8;26;27, transcranial Doppler²⁸, medical rehabilitation²⁹ outcome measures, and urine biomarkers³⁰ seeking predictors of neurologic outcome and outcome measures for clinical trials. However, to date only some of these studies have been longitudinal and prospective (urine biomarkers ongoing and, to some extent, quantitative EEG), and more long-term outcome data is needed. Furthermore, these studies have all been single center and extension of these studies to other centers is a future goal.

Another prospective, longitudinal study (Longitudinal Neuroimaging in SWS, PI: C. Juhasz, Wayne State University; active since 2013) is focused primarily on imaging predictors of neuro-cognitive outcome. The initial results defined some imaging correlates in a cross-sectional setting and short-term (1 to 2 years) outcome data (Juhasz et al., 2007; Behen et al., 2011)^31;32; data on long-term outcome are lacking. Issues to resolve include the question of when (at what age) poor outcome becomes apparent and how early such outcome could be predicted. Clinical neuroimaging, particularly advanced MRI, has been undergoing rapid progress in the last decade. Some advanced MRI sequences (such as susceptibility weighted imaging, perfusion imaging, and diffusion weighted imaging) are increasingly but variably incorporated in routine clinical imaging protocols in different centers.²⁵ Other imaging techniques (e.g., resting state functional MRI, arterial spin labeling) are under intense clinical research. Progress in imaging research and multicenter imaging trials would require development, validation, and sharing of common imaging protocols that can be applied in most major imaging centers. Further development of the current database into a multicenter, longitudinal database could be a key step to provide the necessary data to resolve some of these and other important clinical issues outlined below. The most urgent reason for greater multicentered efforts is the development of clinical trials for SWS. While biomarker and longitudinal studies have been carried out with some success as single center studies, it is more difficult to recruit patient participation for drug trials. In order for drug trials to move beyond small phase I/II trials to randomized placebo controlled trials, multicentered studies are essential. mTOR inhibitors are currently receiving attention for treatment of the neurologic aspects of SWS (https://clinicaltrials.gov/ct2/show/NCT01997255?term=sturge-weber&rank=4); however multicentric efforts will be needed to establish this as a valid treatment option if initial pilot results of this treatment approach appear safe and helpful. Key dermatology topics include identifying markers that indicate risk of progression of cutaneous vascular malformations and soft tissue hypertrophy, and identifying candidate drugs for potential medical therapy for cutaneous lesions (alone or combining with pulsed dye laser [PDL]). A recent, phase II, randomized, double-blind, intra-individual placebo-controlled, multicenter clinical trial conducted in Spain in 23 patients with SWS and facial PWB suggested that the combination of PDL and rapamycin provided superior results as compared to other interventions (including PDL plus placebo or rapamycin alone).²⁰ Larger studies are needed, and trials in the future would benefit from a multicentric design. Diffuse choroidal hemangioma is treated by a number of approaches including photodynamic and radiotherapy, while local anti-VEGF therapy may also be an option based on some case reports.^33–35 Most of the published data, however, are based on case reports or small case series; therefore, evidence-based guidelines cannot be developed until results from larger, multicenter studies become available.

The multicenter collaboration within the BVMC and the SWF Centers of Excellence network nationally and internationally can be a starting point for participating institutes to work together. Such cooperation could help to establish common diagnostic and clinical monitoring protocols, develop robust biomarkers that are prognostic of outcome and responsive to treatment, and foster collaborations in future therapeutic trials. The National Biomarker Development Alliance has recently published reviews on the many issues that have hindered the development of biomarkers for many medical fields.³⁶ They highlighted several factors such as adequate sample sizes, high quality biospecimens and associated clinical data, consistent performance of measurement technologies and resulting test accuracy, high quality experimental design, data, and data analysis methods, and elimination of bias. Furthermore, they emphasized that a clinical biomarker must successfully inform a clinical decision that enhances patient care. Multicenter collaborations create both opportunities and challenges in this effort, but done correctly enable the development of validated biomarkers at centers that will use them in clinical drug trials and to enhance patient care. Nevertheless, to achieve timely progress, the search for better biomarkers should not prevent centers from developing new clinical trial protocols at the same time.

Based on the knowledge gaps listed in the previous section, several key areas could benefit from clinical network collaborations. These include development of standardized neuroimaging protocols for multicenter trials, as well as imaging marker development and validation. The protocols should be applicable for multiple 3 T MR systems (such as Siemens, GE, and Phillips) and should utilize objective, quantifiable parameters. These protocols will be used for patient stratification as well as treatment monitoring. In addition, integrated MR/PET scanners are increasingly available and should facilitate the exploration of multi-modal imaging (MRI plus molecular imaging) in future SWS clinical trials, e.g., in the setting of presurgical evaluation. Another goal should be the identification of the most immediate targets for clinical drug trials; due to relatively easy drug application and monitoring, treatments targeting PWB and soft tissue overgrowths both external and internal should have high priority. However, it should be noted that the goal in the skin (sustained obliteration of the abnormal vessels) may not be the same goal in the brain and eye where the aim is improved venous drainage and improved perfusion; positive results in the skin therefore may or may not translate to the eye and the brain. Centers should also develop consensus regarding drugs or other treatments that should have high priority for SWS clinical trials. To accelerate progress, drug repurposing should be explored while novel drugs are being developed. For neuro-cognitive comorbidities, priority should go to treatments targeting seizures, stroke-like episodes, and headaches³⁷, as well as those treatments with the potential to prevent or reverse cognitive impairment. We recommend pursuing the use of those participating Centers within the BVMC Network and utilizing the expanding SWF Centers of Excellence across North America, with the goal of eventually including centers in other countries and continents. Furthermore, using this model of current and expanding clinical network collaborations, multicenter clinical trials implementing trial designs optimized for rare diseases with limited sample sizes (discussed by Facey et al., 2014)³⁸ should be developed.

Tissue Banking

The tissue banking group concluded that a paucity of available tissue from SWS/PWB patients for research is a significant impediment to progress and can be alleviated by further development of a centralized tissue banking system. This need is particularly urgent due to lack of meaningful animal models of this disease, but the need will remain even as those imperfect models evolve. Investigators who have to date made important advances in understanding the underlying biology of SWS/PWB have of necessity relied upon limited “private” archives of often haphazardly collected and inconsistently preserved affected tissues and upon other NIH funded brain and tissue banks, where affected tissue is more consistently collected but also rarely paired with blood or unaffected adjacent tissue from the donating patients. Retrospective and prospective collection of paraffin-embedded blocks from well-characterized SWS/PWB patients from multiple collaborating institutions is a worthy current and future goal to support SWS/PWB research. At the same time, it is considered essential to tackle the more challenging goals of provision of quality snap-frozen tissues and, importantly, fresh tissue and various cell lines prepared from those tissues. SWS/PWB cell lines will become important once techniques for isolating and perpetuating cells with the mutation are established. For all banked tissues and derivative cell lines, standardized collection of relevant clinical information is considered of great added value that must not be neglected in prospective collections. Linkage of each banked tissue sample to any available imaging, electrophysiology, histology, and genomics studies, as well as clinical datasets, would be empowering.

In order to optimize tissue collection and distribution for research, the tissue banking group recommends funded utilization of the resources of one or more pre-existing, high-quality biobanks already affiliated with SWS investigators and active vascular anomalies centers. A closely coordinated “central” SWS/PWB bank would store SWS/PWB tissue from those facilities as well as tissue shipped in from other hospitals that do not have viable local banking options. Such a network of collaborating biobanks focused upon SWS/PWB tissue would together form a multi-site SWS/PWB bank with standardized collection, shipping, storage, and quality control protocols. In addition, the component primary physical banks of such a system, joined by other collaborating banks, could together create a “virtual” bank of specimens and associated clinical data available for use by a defined, hopefully large, group of registered users. The object of “registration” would be to facilitate thoughtful and fair stewardship of the bank’s limited stores, to be overseen by a multi-partisan utilization committee. We recommend NIH funding (or other non-partisan funding) for this biobank with the requirement that all quality research projects dealing with SWS and port-wine nevi have access to this tissue. Within such a centralized system it would be wise to use a system of universal identifiers for each sample/individual/cell line/etc. to retain important linkages. Once cell lines are established, these could also be distributed through established biobanks such as Corriel, and collaborations with already established national tissue banks could be considered as well to maximally leverage resources.

In order to achieve such a coordinated SWS/PWB banking mission, several goals must be met. A critical first goal is enlistment of key clinicians, surgeons, and pathologists at participating institutions to promote timely salvage and banking of excess tissue from medically necessary tissue resections. Inclusion criteria for each IRB-consented donor would include clinical documentation and verification of diagnosis and a system for assurance of standardized collection and storage of fixed, paraffin-embedded, and snap-frozen SWS/PWB tissue when that is an option. Banking of whole blood and, when available, “normal” adjacent tissue must be encouraged and facilitated. Development of a shared, multi-institutional IRB protocol would facilitate sample acquisition.

Central facilitation of the consenting and banking process will require development of standardized instructions and kits for collection, handling, portioning, shipping, and storage of tissue samples. Also required is development of standardized forms for collection of pertinent clinical information from each consented donor and input of that data into a centralized, web-accessed database cataloging all banked specimens from all participating institutions. Consenting would be facilitated by use of a standardized banking consent form and assignment of a unique SWS/PWB identifier – preferably pre-assigned prior to scheduled surgery – to each specimen, regardless of storage location.

Beyond the banking process per se, there is need for development of a process for requesting and choosing optimal samples for individual study protocols, and for determining appropriate aliquot sizes for any given study (to maximize availability of tissue from any given donor for multiple SWS studies). This will be predicated upon development of prioritization approaches and empowerment of a utilization committee to audit usage and help govern sustainable and fair distribution practices. Establishment of a dbGaP project (database of Genotypes and Phenotypes developed to archive and distribute the results of studies that have investigated the interaction of genotype and phenotype) should be considered to store SWS/PWB sequence data. Lastly, a cost/subsidy structure for infrastructural development, maintenance, and usage must be determined, and a system of outcomes/publication tracking must be developed.

Animal and Cell Culture Models

Animal and in vitro models provide crucial insight about gene function and provide preclinical models to identify and test targets for genetic disorders. Progress in neurogenetic diseases such as tuberous sclerosis complex (TSC), neurofibromatosis type 1 (NF1), and Rett syndrome (RTT) provides plausible roadmaps for the use of animal models for therapeutics development. In each of these diseases, different mouse models replicate different aspects of the disease. In TSC, for instance, some mouse models develop seizures,³⁹ which commonly affect patients, while others do not have any seizures but display autistic-like behaviors.⁴⁰ There is no single mouse model that replicates all the features of a human disease such as TSC or NF1.

In order to develop a mouse model for SWS, one may take a genetic or a cellular approach. Since SWS is caused by a somatic mosaic mutation, one will have to express the mutant gene in a subset of cells. There are several ways to achieve this. One may use an inducible conditional allele in a specified cell type and induce gene expression via pharmacological means. This often results in a mosaic expression pattern. Alternatively, one can induce expression of the gene via a viral expression by infecting only a subset of cells/tissues.⁴¹ All of these somatic mosaic approaches require some level of understanding of what cell types are affected in a disease.

Given the fact that the GNAQ mutation in SWS is thought to result in gain of function, not loss of function, one cannot simply “knockout” a gene. This creates an additional layer of complexity. Recently, sophisticated approaches have been developed to overcome this problem. The conditionals by inversion (COIN) method utilizes a conditional module inserted into the target gene in an orientation opposite to that of the gene’s direction of coding. Activation by Cre recombinase inverts the COIN module, resulting in knockout of the original gene sequence, and its replacement by the mutant form.⁴² This mutation still needs to be activated in a subset of cells in a temporally and spatially restricted manner. While less clinically germane models may certainly result in important insights, there are unique challenges present in the creation of a truly fit SWS model, which are shared with other disorders caused by somatic mutations. Such a model would be restricted to the face, eye and/or brain, and originate in the developing fetus. The fetal vasculature develops in tandem with the cortex and structure of the eye, and this developmental trajectory almost certainly impacts the later presentation of symptoms such as seizures and glaucoma.

One can also generate a mouse model by injecting cells carrying a mutation into a wild type mouse. There are two approaches to consider. The first is to isolate mutant cells from the vascular lesion and expand them in culture to sufficient quantities for injection into animals. The second is use of retroviral or lentiviral expression constructs to introduce the mutation into the appropriate cell type and inject these into animals. Both approaches have been successful for the study of vascular anomalies. Infantile hemangioma-derived stem cells rapidly form hemangioma-like blood vessels when implanted in immune-deficient mice.⁴³ Human endothelial cells engineered to express mutant Tie2 (L914F) form ectatic blood-filled channels highly reminiscent of human venous malformations after injection into immune-deficient mice.⁴⁴ Investigation of signal transduction pathways downstream of receptor tyrosine kinases, such as Tie2 and VEGF receptors, showed that Akt is a key signaling node that mediates the effects of angiogenic growth factors. Akt1 activation in endothelial cells is sufficient to induce pathological blood vessel formation that shares similar histologic features with capillary malformations in a genetically engineered mouse model.⁴⁵ Transiently transfected cells with the R183Q GNAQ mutation have already been published,¹ and efforts to generate and work with stably transfected cell lines are currently in progress in multiple labs.

It has been shown that the GNAQ p.R183Q mutation is enriched in endothelial cells in capillary malformations.⁴⁶ Therefore, similar approaches could be pursued. GNAQ mutant endothelial cells isolated from capillary malformation or SWS lesions could be implanted into immune-deficient mice, with the site of implantation an important aspect of the model. Combinations of mutant and wild-type cells should be tested to explore non-cell autonomous mechanisms. Three-dimensional models of lesional cells (such as endothelial cells) mixed with non-lesional cells (perhaps stromal cells) in spheroids or co-cultures on a collagen/fibronectin 3D platform would represent other options. Establishing relevant three-dimensional co-culture models to better recapitulate the tissue microenvironment of SWS lesions would be highly relevant for disease modeling and providing a good biological platform for drug testing. Endothelial cells engineered to express the GNAQ mutation could be developed and tested in parallel. Induced pluripotent stem cell (IPSC) development for SWS requires selection of cells with the R183Q GNAQ mutation from affected human skin or brain tissue.

Another animal model that has been very informative in neurobiology and in studies of vascular development is the zebrafish. There are reported lines available for studying different cell types separately. In the past, morpholino-mediated gene knock-down,⁴⁷ and more recently, CRISPR-mediated gene editing,⁴⁸ have provided useful genetic tools in zebrafish. Both of these approaches can also be used in mice.^49;50 One of the goals of developing an animal model of SWS is to be able to screen potential compounds. Zebrafish models provide a robust drug screening platform to identify targets. Mice also provide an excellent model for drug studies on the targets. The NF1 consortium is one example of a successful collaborative network that is using a number of models of NF1 for semi-high-through-put preclinical drug trials⁵¹. Medium- to high-throughput drug screening is perhaps best accomplished in vitro. Endothelial, HEK293, and other cells engineered to express the GNAQ mutation, as well as a gene reporter (fluorescent, ELISA or FRET based) for a downstream protein impacted by the mutation, can be developed to identify lead drugs for further preclinical screening and development in animal models. One aspect of preclinical studies that has been often neglected is the PK/PD determination.^52;53 These are important aspects of preclinical studies that will require attention as animal models are developed and used for preclinical screening of drugs.

CONCLUSIONS

SWS and PWB research has entered a new era, a genetic era, guided and inspired by the knowledge of the somatic mutation that causes these disorders. Clinical and translational researchers long committed to this field are now building new collaborations to tackle the next questions. Researchers who have steered clear of this area for lack of tractable direction and resources are now joining the endeavor. In this context, the recent SWS research workshop sought to guide future efforts in clinical and translational research by providing a map for the way forward. Patients and families impacted by SWS enjoy new hope from the discovery of the gene, and are counting on researchers to make efficient use of time and resources to take the genetic era of SWS forward into the era of targeted effective treatments.

Table 3.

Registered active clinical trials for Sturge-Weber syndrome (SWS) and Port-Wine Stain (PWS) on clinicaltrials.gov

Trial title	Trial Identifier	Institute	Period	Phase	N	Main Target	Age
Active trials - SWS
1. Treatment of Port-wine Mark in Sturge-Weber Syndrome Using Topical Timolol	NCT01533376	Wills Eye (USA)	2012–15	Phase I	10	PWS/SWS	2y-10y
2. Innovative Approaches to Gauge Progression of Sturge-Weber Syndrome	NCT01425944	Multicenter (USA)	2010–15	n/a	386	SWS	>1m
3. Adjunctive Everolimus (RAD 001) Therapy for Epilepsy in Children With SW	NCT01997255	Baylor (USA)	2014–16	Phase 2	25	SWS Epilepsy	2y-18y
4.Cannabidiol Expanded Access Study in Medically Refractory SWS	NCT02332655	Kennedy Krieger (USA)	2014–16	Phase1/2	10	SWS Epilepsy	1m-30y
Active trials - PWS
5. Monitoring the Response of PWS Birthmarks to Laser Therapy With Wide-field Functional Imaging Technologies	NCT01333553	UC Irvine/Beckman Laser Institute (USA)	2010–17	n/a	500	PWS	Any
6. Pathogenic Mechanisms of PWS and Repository of PWS Biopsy Samples	NCT02051101	UC Irvine/Beckman Laser Institute (USA)	2013–16	Phase 1	100	PWS biopsy samples	Any
7. Novel Treatment for PWS Birthmarks	NCT01924273	UC Irvine/Beckman Laser Institute (USA)	2013–16	Phase 1	24	PWS	>18y
8. Combined Bipolar Radiofrequency & PDL	NCT01775722	UC Irvine/Beckman Laser Institute (USA)	2012–18	Phase 1	30	PWS	>12y
9. A Randomized Trial to Study Combined PDL Rapamycin Treatment of PWS Birthmarks	NCT00830466	UC Irvine/Beckman Laser Institute (USA)	2008–17	Phase 1	40	PWS	>13y
10. Evaluate the PWS Birthmark Treatment Before and After PDL Treatment	NCT01774552	UC Irvine/Beckman Laser Institute (USA)	2012–18	Pilot	30	PWS	Any
11. Treatment of PWS Using PDL Erbium Yag Laser and Topical Sirolimus	NCT02214706	Erasmus Medical Center (Netherlands)	2014–16	Pilot	20	PWS	≥18y
Active but not recruiting – PWS
12. French National Cohort of Children With PWS (CONAPE)	NCT01364857	Multicenter (France)	2010–20	n/a	150	RASA1 polymorph	22y-12y
13. Treatment of Resistant PWS With Bosentan and PDL: a Pilot Study	NCT02317679	Centre Hospitalier Univ. de Nice (France)	2014–15	Phase 2	4	PWS	7–60y

PDL=Pulsed Dye Laser

List of Participants (in alphabetical order, * indicates Workshop Organizing Committee Members)

Abreu, Nicolas MD, NYU Langone Medical Center; Acosta, Maria MD, Children's National Medical Center; *Ball, Karen L., The Sturge-Weber Foundation; Berrocal, Audina MD, University of Miami Miller School of Medicine; Bischoff, Joyce PhD, Boston Children's Hospital; Brodie, James, GW Pharmaceuticals; Burkhart, Craig MD, University of North Carolina Hospitals; Cohen, Bernard MD, Johns Hopkins University School of Medicine; *Comi, Anne M. MD, Kennedy Krieger Institute; Dymerska, Gosia, Johns Hopkins University School of Medicine; Eckstein, David PhD, NIH Office of Rare Diseases Research; Enriquez-Algeciras, Mabel, University of Miami Miller School of Medicine; Ewen, Joshua MD, Kennedy Krieger Institute; Fisher, Brian, The Sturge-Weber Foundation; Freedman, Sharon MD, Duke University School of Medicine; Germain-Lee, Emily MD, Kennedy Krieger Institute; Geronemus, Roy MD, NYU Langone Medical Center; *Gold, Michael MD, Tennessee Clinical Research Center; Gopal-Srivastava, Rashmi MS, PhD, NIH Office of Rare Diseases Research; Hammill, Adrienne MD, PhD, Cincinnati Children’s Hospital Medical Center; Hebert, Adelaide MD, Memorial Hermann Texas Medical Center; Huang, Lan PhD, Boston Children's Hospital; Jampel, Henry MD, MHS, Johns Hopkins University School of Medicine; Juhász, Csaba MD, PhD, Wayne State University School of Medicine; Kaplan, Emma H., Kennedy Krieger Institute; Kaseka, Matsanga MD, St. Justine University Hospital Center; Kirkorian, Yasmine MD, Children's National Medical Center; Kossoff, Eric MD, Johns Hopkins University School of Medicine; *Levin, Alex V. MD, MHSc, FRCSC, Wills Eye Hospital; Lin, Doris MD, PhD, Johns Hopkins University School of Medicine; Lo, Warren MD, Nationwide Children’s Hospital; Loeb, Jeffrey MD, PhD, University of Illinois College of Medicine; Marathe, Kalyani MD, MPH, Children's National Medical Center; Marchuk, Doug PhD, Duke University School of Medicine; Mead, Alice JD, GW Pharmaceuticals; Mellis, Scott MD, PhD, Regeneron Pharmaceuticals; *Morris, Jill PhD, NIH National Institute of Neurological Disorders and Stroke; Murray, Timothy MD, MBA, FACS, Murray Ocular Oncology and Retina; North, Paula MD, PhD, Children's Hospital of Wisconsin; Phung, Thuy MD, PhD, Texas Children's Hospital; Pinto, Anna MD, PhD, Boston Children's Hospital; Puttgen, Kate MD, Johns Hopkins University School of Medicine; Ratner, Nancy PhD, Cincinnati Children’s Hospital Medical Center; Reeve, Jennifer MD, PhD, C. S. Mott Children’s Hospital; Roach, Steve MD, Nationwide Children's Hospital; *Sahin, Mustafa MD, PhD, Boston Children's Hospital; *Swindell, Charles PhD, The Sturge-Weber Foundation; Tseng, Hung PhD, NIH National Institute of Arthritis and Musculoskeletal and Skin Diseases; Tune, Miriya, Kennedy Krieger Institute; Wetzel-Strong, Sarah PhD, Duke University School of Medicine; White, Monica PharmD, Novartis Pharmaceuticals Whittemore, Vicky PhD, NIH National Institute of Neurological Disorders and Stroke