Every day, families receive the unthinkable news of a childhood cancer diagnosis, and a powerful alliance between the Children’s Hospital of Philadelphia’s Center for Data-Driven Discovery in Biomedicine (D3b) and Amazon Web Services (AWS) is delivering more than just hope. D3b recognizes that the key to unlocking new, life-saving treatments lies in breaking down the data silos that have long hampered progress. This vital partnership is dismantling these barriers with a revolutionary pediatric research model: the Real-time Analysis and Discovery in Integrated and Networked Technologies (RADIANT) project.

The Children’s Brain Tumor Network (CBTN), a D3b program, created a groundbreaking framework for sharing complex molecular, imaging, and clinical data across 35 leading pediatric research institutions worldwide. Dr. Rowland Illing, AWS’ Global Chief Medical Officer, states that “Data is the pediatric cancer and rare disease community’s superpower. Powered by AWS’ secure, scalable cloud infrastructure, RADIANT is the type of unprecedented near real-time data sharing mechanism that can get the right insights into the right patients, clinicians, and researchers when every second counts.”

Building upon that success, funding through the Agency for Research Projects in Health (ARPA-H), National Institutes of Health (NIH), and generous assistance from AWS have connected front-line clinicians and laboratory researchers. The goal: to leverage advanced artificial intelligence (AI) systems to more rapidly identify personalized treatment options for young patients fighting aggressive cancers and rare diseases.

RADIANT emerged as a direct result of this shared vision to accelerate discovery, enabling pediatric oncologists to instantly share critical patient insights with research scientists and other specialized care teams across institutional boundaries. It sets a new standard for precision medicine, directly benefiting children with medulloblastoma, diffuse intrinsic pontine glioma (DIPG), neuroblastoma, and other devastating pediatric conditions. For families facing these diagnoses, this means faster access to targeted therapies, reduced treatment toxicity, and ultimately, improved survival rates and quality of life.

The critical collaboration between D3b and AWS powers RADIANT to transform pediatric cancer research, turning valuable data into a tangible source of hope for children and families facing life’s most difficult challenges.

To learn more about this partnership and how AWS and D3b are driving discoveries faster than ever before, please visit here.

A New Model for Drug Development

This partnership marks a new and impactful model for pharmaceutical collaboration, where CBTN provides high-quality clinical data to inform the development of new drugs. The initial phase of the project, which involved data from 50 patients, was completed in April 2025, laying the groundwork for a larger effort and establishing a common language and workflow between the organizations. The second phase expanded the dataset to include an additional 150 subjects with BRAF-altered pLGG, further developing scalable workflows across CBTN partner institutions. The goal is to generate real-world insights that support scholarly research, clinical trials, and regulatory submissions.

This project exemplifies the dedication of the D³b teams, including the Biospecimen and Clinical Research Unit (BCRU) and the Advanced Data Applications and Platform Technologies (ADAPT) Unit. 

“The first dataset delivered by CBTN is an important foundation of a promising collaboration,” said Davy Chiodin, Chief Development Officer of Day One Biopharmaceuticals. 

This success highlights the caliber of work and commitment to data integrity within the D³b center.

Why This Matters for Pediatric Cancer Patients

The collaboration leverages the network’s ability to collect and integrate high-quality clinical, molecular, and imaging data with the consent of pediatric and young adult patients with brain tumors. This approach provides a rich, real-world data set that can help identify new therapeutic strategies without the need for time-consuming subject enrollment and protocol development.

Tatiana Patton, Director of Research Programs and Project Management at the D3b, emphasized that this partnership enables pharmaceutical companies to “shop” for specific genetic traits, allowing for the design of more precise clinical trials with a better chance of success. “This precision is critical because pediatric brain tumors can be rare, making it difficult to find effective therapies.” Tatiana shared. 

The real-world data and precision genetic information collected will help scientists better understand how these tumors behave and respond to treatment, leading to more targeted and effective therapies.

The Long-Term Impact

This groundbreaking collaboration isn’t just about a single project; it’s about building a sustainable system for innovation that will benefit future generations. Findings from this research will be shared with the broader research community through publications and presentations, ensuring this new knowledge contributes to the effort to find cures.

The partnership helps accelerate the discovery of new therapies by providing a comprehensive resource for BRAF-altered brain tumors. By analyzing a vast dataset, researchers can identify patterns and potential targets for treatment more quickly than ever before. This focused approach could be a game-changer for a disease where time is of the essence.

In addition to inspiring future research projects, this first collaborative project establishes a structured and scalable workflow for data curation and imaging review, laying the groundwork for potential future cancer breakthroughs and collaborations. This streamlined process can be replicated for other types of cancer, creating a ripple effect of efficiency and innovation across the medical field. By standardizing these procedures, the partnership helps ensure that future research can build on a solid, reliable foundation.

This partnership brings hope to families and is a powerful example of how the organization’s mission to accelerate cures for children with brain tumors is realized through trusted collaborations with business partners and robust data stewardship.

The Down Syndrome Research Group at Children’s Hospital of Philadelphia (CHOP) is embarking on this initiative, fueled by over $2 million in National Institute of Health (NIH) funding, enlisting data resources of the INvestigation of Co-occurring conditions across the Lifespan to Understand Down syndromE Data Coordinating Center (INCLUDE DCC), created and managed by D³b. This initiative aligns with the D³b mission to transform data into impact, empowering convergent research and accelerating translational research to improve patient outcomes, particularly in childhood cancers and rare diseases. The study leverages leading-edge data analysis to explore the intricate interplay of co-occurring conditions, social determinants of health, and the impact of ableism on individuals with Down syndrome.

D³b Starts a Data-Driven Reimagining

This project is a prime example of how the D³b infrastructure and expertise facilitate groundbreaking research. Medical data generated from this study will become a crucial asset, stored within and made accessible through INCLUDE DCC. This centralized, interactive data resource will empower researchers and clinicians to:

Uncover Hidden Patterns. By integrating diverse data streams—medical records, blood samples, endocrine, nutritional, sleep, and cognitive evaluations—the analytical tools developed by D³b enable scientists in the Down Syndrome Research Group to identify previously unseen relationships and patterns, thereby better managing co-occurring conditions. While, as a research center, D3b doesn’t treat conditions such as Down Syndrome or Alzheimer’s Disease, its ecosystem enables discoveries that lead to improved quality of life for patients.

Advance Personalized Medicine. With a comprehensive lifespan dataset, this collaboration aims to develop tailored interventions that address the unique needs of individuals with Down syndrome.

Bridge the Research and Practice Gap. The new Down syndrome project at CHOP has the potential to facilitate the translation of research findings into actionable clinical insights, enabling general pediatricians and specialists to make informed decisions.

A Community-Centric Approach to Data Collection

This initiative is not confined to a single institution. Project researchers are collaborating with experts across the country to ensure that data collected reflects diverse representation and maximizes impact.

Additionally, project partners understand that the success of the data collection initiative for this project depends on a strong partnership with the Down syndrome community. CHOP researchers have developed inclusive practices from accessible consent forms and websites informed by a community advisory panel, to gathering lived experiences through interviews and surveys to address ethical concerns and ableism. Tools like storyboards and behavioral supports help reduce anxiety, and community input has already led to at-home sleep studies that ease participation.

Data-Driven Discoveries for a Brighter Future

This NIH-funded study, the largest of its kind, promises to generate a wealth of data that will transform our understanding of Down syndrome. By harnessing the power of the D³b data infrastructure and fostering close collaboration with the community, this innovative team is poised to unlock groundbreaking discoveries that enhance the health and well-being of individuals with Down syndrome throughout their lifespan.

Director of Data Technology and Innovation at D3b, Dr. Allison Heath, explains, “We don’t have answers to many questions. We have hypotheses, but looking at these data over the lifespan is groundbreaking.” D3b is committed to making those groundbreaking discoveries a reality.

This research addresses the critical issue of data privacy in pediatric neuroimaging. Brain MRI scans contain identifiable facial information, making data sharing challenging due to privacy concerns. To tackle this, Familiar and her team developed a pediatric auto-defacing tool. This AI-driven solution automatically removes facial features from brain MRIs, enabling safer data sharing and advancing neuroscience research.  

This de-identification tool is an essential scientific breakthrough that provides a robust method for de-identifying pediatric brain images. The AI technology was meticulously developed using the Children’s Brain Tumor Network (CBTN) dataset and other control data to ensure accuracy and applicability across diverse pediatric cases and image types.  

Back to the Beginning

Ariana Familiar’s journey to this pivotal achievement is rooted in her extensive background in pediatric brain tumor research and data science. With a BA/BAS in psychology and philosophy from New York University and a PhD from the University of Pennsylvania, her expertise spans cognitive neuroscience, visual memory, and social behavior. Familiar’s passion for pediatric cancer research is evident in her leadership at D3b, where she spearheads efforts in multi-site data sharing, data aggregation, and the development of analytical methods for radiology data.  

“The intersection of healthcare and technology was super interesting to me,” Familiar notes, highlighting her motivation to focus on this critical area of research.  

The Science of Ethical Research

Developing the “pediatric-auto-defacer” underscores D3b’s commitment to ethical research practices. This work exemplifies the center’s dedication to advancing pediatric brain tumor research while upholding the highest ethical standards by prioritizing patient privacy and developing a tool that enables responsible data sharing.     

Impact on Pediatric Research

The development of this AI tool has profound implications for pediatric research. It promotes enhanced data sharing and discovery, accelerates the development of improved treatment options, and ultimately leads to better outcomes for children with cancer. By safeguarding children’s identities, this tool empowers researchers, clinicians, and families to contribute to and benefit from collaborative research efforts.  

The Future Begins with AI

This innovative AI tool and Familiar’s publication in the AJNR spotlight a significant leap forward in pediatric brain research. By prioritizing data privacy, this work underscores the commitment of CBTN and D3b to responsibly advancing the field and improving the lives of children and their families. We look forward to the future discoveries this advancement will enable and its positive impact on pediatric healthcare.

To read more about the pediatric-auto-defacer and what it means to pediatric research, visit the American Journal of Neuroradiology.

Dr. Anahita Fathi Kazerooni, a leading scientist at the Center for Data-Driven Discovery in Biomedicine (D³b), and her team members recently published “Multiparametric MRI along with machine learning predicts prognosis and treatment response in pediatric low-grade glioma” in the prestigious Nature Communications Journal. This study sheds new light on the complexities of low-grade gliomas and demonstrates the incredible impact of collaborative research at D3b.

Bridging the Gap Between Imaging and Molecular Data

Along with Dr. Kazerooni, Dr. Adam Kraya, Technical Director of Clinical and Translational Data Science at D3b, tackled a significant challenge: understanding the intricate relationship between imaging data (like MRIs) and molecular information (like gene expression) within these tumors. They developed a robust predictive model by ingeniously combining these different data types. This model can

• Identify high-risk patients by accurately predicting which children with low-grade gliomas are at a higher risk of their tumor progressing or returning.
• Stratify patients based on risk by categorizing patients into low, medium, and high-risk groups, enabling more personalized and targeted treatment plans.
• Continuously improve the model by constantly updating as new data becomes available, such as surgery results or further molecular analyses, providing an ever-evolving understanding of each child’s unique tumor.

Impact on Patient Care

This research potentially revolutionizes the care of children with low-grade gliomas. By identifying high-risk patients earlier, clinicians can:

• Adjust treatment plans, Explore whether there are more aggressive or innovative therapies, such as targeted therapies or immunotherapies, that may be more effective for these children.
• Improve long-term outcomes: By intervening earlier, clinicians can potentially prevent or delay tumor progression, leading to better functional and cognitive outcomes for these young patients.

Overcoming Challenges

Dr. Kazerooni and her team faced significant challenges in working with real-world clinical data, including variations in imaging protocols and missing data.

“One of the challenges we faced was that imaging wasn’t standardized across all patients,” explained Dr. Kazerooni. “This meant some images were of lower quality or even missing, which could have skewed our results.”

However, Dr. Kazerooni and her team overcame these obstacles by meticulously organizing and selecting data and developing innovative analytical methods.

The Future at D3b

The publication demonstrates the power of the collaborative approach and the invaluable data repository they’ve built. By integrating diverse data types, including imaging, genomics, and clinical information, D3b researchers are unlocking groundbreaking insights into the biology of pediatric brain tumors.

It also paves the way for future studies to refine these predictive models further and translate these findings into improved treatments for children battling brain tumors.

“We’re incredibly proud of Dr. Kazerooni and her groundbreaking work!” said Dr. Ali Nabavizadeh, D3b Director of Imaging. “This achievement exemplifies the dedication and innovation of our talented team of researchers. By fostering collaboration and leveraging the power of data, we’re poised to revolutionize the understanding and treatment of these devastating diseases.” 

This is just one example of the incredible work the fantastic team is currently pursuing. D3b is committed to pushing the boundaries of pediatric brain tumor research and improving the lives of children facing these challenges. Stay tuned for more innovative research in 2025.

Breaking Down Barriers for Faster Discoveries

Research on pediatric cancers and rare diseases often faces limitations due to fragmented data, siloed teams, and low incentives to share information. Critical information about patients is scattered across different hospitals and institutions. That’s where D3b’s “Real-time Analysis and Discovery in Integrated And Networked Technologies” (RADIANT) project steps in. RADIANT will create a secure, cloud-based, real-time data-sharing and analytics platform, alleviating this hurdle to successful discovery and clinical impact.

RADIANT aims to revolutionize the understanding and treatment of pediatric cancer and rare diseases by developing a new way to get this valuable data into the hands of the doctors and institutions facing the challenge of these deadly diseases. This innovative project leverages the power of data to accelerate research and advance precision medicine for children.

Collaboration is the Key to Success

This award is not only celebrating the tireless work of the brilliant minds at D3b and its partner organizations but also a first step in a project that will have a long-term impact on families worldwide. The center’s success stemmed from a powerful collaboration, a consistent facet of all the work done by these organizations. The project leverages the vast patient data collected through a network of over 35 healthcare systems and families within the Children’s Brain Tumor Network (CBTN).

The Pacific Neuro-Oncology Consortium (PNOC) is another key partner, working with D3b and CBTN to build a robust and integrated clinical trial research framework. Finally, the initiative boasts international collaborations with investigators at Centre Hospitalier Universitaire Sainte-Justine to enhance precision medicine platforms for data management and visualization. Tech giants like Amazon Web Services (AWS) provide cloud muscle for computation and analysis, while Mulesoft, Kno2, HealthShare Exchange and Flywheel.io offer crucial technological solutions for rapid data exchange and integration across hospitals. D³b’s partnership network is an influential collaborator in advancing pediatric brain tumor research, bringing us closer to life-changing cures.

RADIANT is a significant step forward in the fight against pediatric cancers and rare diseases. Dr. Jay Storm and Dr. Adam Resnick, D3b Co-Directors, and Dr. Allison Heath, D3b Director of Data Technology and Innovation, spearhead the RADIANT project, building upon D3B’s existing expertise in big data and collaborative research to create groundbreaking national networks that advance the care of patients.

What Does This Mean for Children Facing Cancer?

RADIANT has the potential to revolutionize pediatric cancer research in several ways, including through:

• Artificial Intelligence: Development of robust AI tools specifically designed for healthcare, enabling faster and more accurate diagnosis.
• Federated Learning: This allows local data integration into research ecosystems, leading to quicker treatment decisions for individual patients.
• National Infrastructure: A secure national data infrastructure will be created, facilitating interoperability between leading platforms and fostering collaboration across institutions.

RADIANT seeks to revolutionize research and precision medicine approaches for  rare diseases, particularly brain tumors in children, by enabling real-time discoveries and improving treatment outcomes. This project has the potential to unlock a new era of personalized care, tailoring treatment plans to each child’s unique needs and offering a more hopeful future for families facing this devastating illness.

The ARPA-H grant awarded to D3b marks a significant step forward for pediatric cancer research. By harnessing the power of data collaboration and cutting-edge technology, RADIANT holds the promise of better treatments, faster diagnoses, and, ultimately, a brighter future for children with cancer.

The Center for Data-Driven Discovery in Biomedicine (D³b) stands at the forefront of pediatric cancer and rare disease research. With a deep-rooted commitment to improving the lives of children, the multidisciplinary D3b team unites expertise from basic science, translational research, precision medicine, bioinformatics, and genomic research. By leveraging innovative technology and collaborating with top-tier institutions globally, D³b is revolutionizing how the world understands, diagnoses, and treats these complex conditions. And, through its robust cloud-based tools and open-science commitment, D³b empowers researchers to unlock the mysteries of pediatric diseases, accelerate clinical trials, and ultimately deliver better outcomes for children.

As part of its continued effort to change the landscape of pediatric disease research, D3b has been instrumental in developing a new tool, AutoGVP, that promises to significantly streamline the classification of potential disease-causing variants inherited from parental bloodlines. The development was a collaborative effort between D³b and the Center for Childhood Cancer Research (CCCR) at the Children’s Hospital of Philadelphia (CHOP), and the National Cancer Institute (NCI). This new tool introduces an automated approach to germline variant annotation and classification.

The Challenge of Gene Variant Classification

Identifying and classifying germline sequence variants associated with inherited diseases have been a cornerstone of biomedical research for decades. However, the sheer volume of variant data generated by modern sequencing technologies has posed significant challenges for researchers. Manual curation and classification of variants are time-consuming and prone to errors, hindering researchers’ abilities to uncover the genetic underpinnings of complex diseases.

The Solution: AutoGVP

Dr. Sharon Diskin, Associate Professor of Pediatrics at the University of Pennsylvania and CHOP CCCR, led the effort to develop AutoGVP in partnership with Dr. Jo Lynne Rokita at D³b (CHOP) and Dr. Jung Kim at the National Cancer Institute (NCI). This novel open-source tool can predict whether a genetic change passed down from parents will likely cause a disease or health problem. Included among the key features that appeal to AutoGVP users:

• Open-source and Dockerized: AutoGVP is freely available to the research community and is distributed as a Docker image (a packaged set of software including all the needed pieces, making it easy to use in various computing environments) and the CAVATICA research analytics application.   
• Clinically-Focused Classifications: The tool provides clinically relevant classifications for variants associated with childhood cancer and/or other diseases, aligning with current American College of Medical Genetics (ACMG) guidelines.
• Scalability: AutoGVP is not just a tool; it’s a powerful ally. Designed to handle large-scale variant datasets, it empowers researchers to annotate and analyze data from large-scale sequencing projects efficiently. With AutoGVP, researchers can tackle even the most complex research projects.
• Flexibility: The tool offers customizable options for users to tailor the classification process to their specific research needs.

AutoGVP has the potential to revolutionize biomedical research by providing researchers with a powerful tool for variant classification. By automating a previously time-consuming and error-prone process, AutoGVP can significantly accelerate the pace of discovery and improve the accuracy of genetic risk assessments, including:

• Genetic Risk Assessment: AutoGVP can identify high-risk individuals for inherited diseases, enabling early diagnosis, monitoring, and intervention.
• Drug Discovery: By understanding the genetic basis of diseases, researchers can develop more targeted and effective therapies.
• Precision Medicine: AutoGVP can help to personalize treatment plans based on an individual’s genetic makeup.

The development of AutoGVP represents a significant milestone in automating germline variant classification. By providing researchers with a reliable and efficient tool for variant annotation, AutoGVP undoubtedly accelerates scientific understanding of the genetic basis of human diseases and paves the way for new and innovative approaches to prevention and treatment.

Access AutoGVP

Researchers can access the AutoGVP open-source dockerized workflow, which is freely available on GitHub at https://github.com/diskin-lab-chop/AutoGVP.

Access the Cavatica application at https://cavatica.sbgenomics.com/public/apps/cavatica/apps-publisher/autogvp.

This article originally appeared on chop.edu.

Philadelphia October 9, 2024 – A multi-institutional program led by researchers Adam Resnick and Allison Heath at the Center for Data Driven Discovery in Biomedicine (D3B) at Children’s Hospital of Philadelphia (CHOP) has received up to $10 million Advanced Research Projects Agency for Health (APRA-H) Mission Office Innovative Solutions Openings award to research, validate and implement a new model for sharing and integrating data at a national level. This new model is designed to empower research on cancer and other rare diseases in real time, accelerating discoveries and transforming the quality of care for children affected by disease.

The project – Real-time Analysis and Discovery in Integrated And Networked Technologies (RADIANT) – builds upon years of pediatric research at D3B aimed at innovating health care and processing millions of points of data at a larger scale to discover underlying causes of complex cancers and diseases and making those findings rapidly available to researchers and clinicians across the country and the world. The RADIANT project is the first ARPA-H-funded project at CHOP.

“In the setting of pediatric brain tumors and other rare diseases, we often do not have a curative standard of care,” said Adam Resnick, PhD, research scientist in the Department of Biomedical and Health Informatics at CHOP and Co-Executive Director of D3B. “As a result, most patients actively participate in a research care model that includes participation in research protocols and clinical trials. Our vision for RADIANT is that clinicians and researchers are able to access real-time data about all patients who might share similarities across different data modalities with any one patient, harnessing cloud-based tools and computation to advance real-time decision making and diagnostics and expedite a patient’s access to potential precision medicines and personalized treatments or relevant clinical trials.”

Expanding on platform tools developed as part of the Gabriella Miller Kids First Data Resource Center and the INCLUDE (INvestigation of Co-occurring conditions across the Lifespan to Understand Down syndromE) Data Hub, the RADIANT project aims to advance new technologies and tools as part of a network architecture that advances the real-time integration of data from a participating patient’s electronic health records with genomic and imaging data. With CHOP as the coordination center of the initiative, the goal is to scale linkages across a network of hospitals throughout the nation, sharing critical data and information as well as providing resources for interpreting that data.

Researchers at CHOP’s D3B will be partnering with other institutions across more than 35 hospitals comprising the Children’s Brain Tumor Network (CBTN) and the Pediatric Neuro-Oncology Consortium (PNOC) to develop, deploy, and pilot this groundbreaking resource that advances the interoperability and use of data through platform-enabled standardization and emerging technologies that harness powerful new tools and computational approaches.

“D3B has been able to consistently accelerate discoveries in pediatric cancer, thanks to our focus on big data and the collaborative researchers around the world who utilize our platforms,” said Phillip “Jay” Storm, MD, Chief of the Division of Neurosurgery and Co-Director of the Neuroscience Center at CHOP and Co-Executive Director of D3B.data. “This award takes the advancements we’ve made in the field of pediatric brain tumors and expands them in exciting new directions,” “A core tenet of our mission has always been the marriage of research and clinical work, and when fully realized, the work accomplished through this award will transform how bench-to-bedside models of care can improve patients’ lives.”

While the initial goals of the RADIANT project will be to use these tools to enhance the precision-based care patients receive while under treatment by their doctor, advancing the research care framework will ultimately inform new models for clinical trial coordination and enhance patient access to care through such trials.

“We need to standardize and advance multiple data workflows across institutions in ways that create a unified, real-time view of a patient, combining clinical data workflows with research data from observational and clinical trials,” said Sabine Mueller MD, PhD, co-executive chair of CBTN and co-leader of PNOC. “Many pediatric brain tumor patients have no curative standard of care, it’s only through advancing new models of care empowered by these technologies that as clinicians we’ll be able to enhance options for our patients and their participation in clinical trials.”

In addition to partnering institutions across CBTN and PNOC, the RADIANT initiative brings together investigators from the Centre Hospitalier Universitaire Sainte-Justine and the HealthShare Exchange (HSX) as well as technology partners including Amazon Web Services, MuleSoft (from Salesforce), Kno2, Peyk, and Flywheel.

Founded in 2022, the ARPA-H is an agency within the Department of Health and Human Services with a mission of accelerating better health outcomes for everyone by supporting the development of high-impact solutions to society’s most challenging health problems. The agency aims to ensure access to solutions regardless of demographics or geography, leveraging diverse perspectives for sustained equity across their policy, practices and programs.

This research is supported by ARPA-H under agreement number 140D042490007.

About Children’s Hospital of Philadelphia:

A non-profit, charitable organization, Children’s Hospital of Philadelphia was founded in 1855 as the nation’s first pediatric hospital. Through its long-standing commitment to providing exceptional patient care, training new generations of pediatric healthcare professionals, and pioneering major research initiatives, the hospital has fostered many discoveries that have benefited children worldwide. Its pediatric research program is among the largest in the country. The institution has a well-established history of providing advanced pediatric care close to home through its CHOP Care Network, which includes more than 50 primary care practices, specialty care and surgical centers, urgent care centers, and community hospital alliances throughout Pennsylvania and New Jersey, as well as the Middleman Family Pavilion and its dedicated pediatric emergency department in King of Prussia. In addition, its unique family-centered care and public service programs have brought Children’s Hospital of Philadelphia recognition as a leading advocate for children and adolescents. For more information, visit https://www.chop.edu

The following news release was shared by Children’s Hospital of Philadelphia on May 31, 2023 – Available on chop.edu. 

Researchers from Children’s Hospital of Philadelphia (CHOP), the Alex’s Lemonade Stand Foundation Childhood Cancer Data Lab, the Children’s Brain Tumor Network (CBTN), the Pacific Pediatric Neuro-Oncology Consortium (PNOC), and more than 20 additional institutions have partnered to create a first-of-its-kind open-source, reproducible analysis platform for pediatric brain tumors. With the help of thousands of genomically sequenced samples, researchers have used this platform to identify initial findings about genetic variants associated with poorer outcomes that could help guide future diagnostic and therapeutic advances.

The details of the platform and those initial findings were published online today by the journal Cell Genomics.

Pediatric brain tumors are collectively the leading cause of cancer-related death in children in the United States. However, the severity of pediatric brain tumors varies wildly, with some having an almost universally fatal prognosis while others have relatively strong long-term survival rates, though all brain tumors negatively impact these children and their families to at least some degree. Limited access to tissue samples and patient-derived cell lines has been a significant barrier to understanding the differences between pediatric brain tumors at a molecular level. That long sought-after data could lead to better diagnostic techniques and potential targeted therapies that could treat these deadly tumors.

In 2011, CBTN and PNOC began extracting and preparing what has now become nearly 6,000 tumor samples with over 68,000 sub-samples. More than 1,000 of these tumors were sequenced to form the initial release of the Pediatric Brain Tumor Atlas (PTBA) in 2018, and data were made available without embargo so that researchers could study what variants might be driving certain types of brain tumors. With the help of the Alex’s Lemonade Stand Foundation Childhood Cancer Data Lab, the team of researchers was able to build an open-source version of this atlas, the Open Pediatric Brain Tumor Atlas (OpenPBTA), to analyze these data.

OpenPBTA is accessible to anyone conducting research who is looking for new therapeutic targets or finding new ways to translate research into clinical practice. At the time of this study, OpenPBTA contained genomic and clinical data from more than 1,000 pediatric brain tumors and 22 patient-derived cell lines from the CBTN and PNOC. The OpenPBTA provides an open, real-time framework to genomically characterize pediatric brain tumors. It is the first large-scale, collaborative, open analysis of genomic data and provides a cloud-based resource for researchers looking for more comprehensive data on pediatric brain tumors.

“While there have been many proponents of an open-source model for scientific research, nothing like this existed for pediatric cancer,” said Jo Lynne Rokita, PhD, a Supervisory Bioinformatics Scientist leading OpenPBTA at the Center for Data-Driven Discovery (D3B) at CHOP and one of the study’s corresponding authors. “We designed OpenPBTA so that anyone could access the data, contribute to its analysis, and/or use it in their own research.”

“Collaboration is key to accelerating new cure discovery. OpenPBTA made it possible for experts across the globe to come together and gain a deeper understanding of the leading cause of cancer-related death in children and young adults,” said Jay Scott, Co-Executive Director of Alex’s Lemonade Stand Foundation.

Jaclyn N. Taroni, PhD, another corresponding author on the study and Director of Alex’s Lemonade Stand Foundation Childhood Cancer Data Lab, said, “With our successful launch of OpenPBTA, we’re hoping the research community adapts this model to other pediatric cancers.”

OpenPBTA is already providing researchers with more insight into potential drivers of pediatric brain tumors. In this study, researchers found that the loss of the tumor suppressor gene TP53 is a significant marker for poor overall survival in fast-growing brain and spinal cord tumors called ependymomas and certain diffuse midline gliomas, and dysregulation of the gene was also implicated in hypermutant high-grade gliomas.

“Solving pediatric brain tumors cannot be accomplished by any one institution. The OpenPBTA model of shared, real-time collaboration supported by PNOC and CBTN has not only empowered new discoveries, but also innovative ways of performing the required science on behalf of accelerated, collaborative innovation for children affected by brain tumors,” said Sabine Mueller MD, PhD, MAS, Professor of Neurology, Neurosurgery and Pediatrics at the University of California, San Francisco, and the Lead of PNOC and executive co-chair of CBTN.

This study was supported by the Alex’s Lemonade Stand Foundation (ALSF) Childhood Cancer Data Lab, an ALSF Young Investigator Award, ALSF Catalyst Awards, an ALSF CCDL Postdoctoral Training Grant, Children’s Hospital of Philadelphia Division of Neurosurgery, the Australian Government, Department of Education, the St. Anna Kinderkrebsforschung, Austria, the Mildred Scheel Early Career Center Dresden P2, funded by the German Cancer Aid, and National Institutes of Health (NIH) Grants 3P30 CA016520-44S5, U2C HL138346-03, U24 CA220457-03, K12GM081259, R03-CA23036, and NIH Contract No. HHSN261200800001E. This project was also funded in part with Federal funds from the National Cancer Institute, National Institutes of Health, under Contract No. 75N91019D00024, Task Order No. 75N91020F00003. Additionally, this work was supported by the Intramural Research Program of the Division of Cancer Epidemiology and Genetics of the National Cancer Institute. 

Shapiro et al, “OpenPBTA: An Open Pediatric Brain Tumor Atlas.” Cell Genom. Online May 31, 2023. DOI: 10.1101/2022.09.13.507832.