Refining Tools, Techniques to Transform Pediatric Medicine

Researchers in the Center for Applied Genomics (CAG) at Children's Hospital of Philadelphia developed a tool, called spaVAE, designed to handle the unique challenges of spatial gene data. spaVAE uses advanced deep learning techniques to understand patterns in the data, especially how different data points relate to each other in a three-dimensional part of the tissue.

This novel processing tool helps simplify complex data from a variety of spatial transcriptiomics technologies (SRTs) that enable scientists to simultaneously profile gene expression and spatial location information in tissues. The new tool can create visual representations of the data, combine data from different experiments, predict gene activity in unmeasured areas of tissue, identify differences in gene activity, and find genes that vary across different spatial regions.

The researchers also adapted spaVAE to handle multiple types of data, including data captured from multi-omics sources — the genome, proteome, transcriptome, metabolome, and epigenome. Their findings, published in the Nature Methods, enable researchers to better identify and understand new therapeutic targets for optimized treatment strategies.

"While the data being generated by SRTs is important, by itself it is difficult to understand the big picture of how this information can be translated in a way that helps drive discovery," said Hakon Hakonarson, MD, PhD, Director of the CAG, and senior study author. "This tool helps cluster the data to identify relationships between different data points and data types better than any other method currently available."

Congenital heart disease (CHD), the most common birth defect in the United States, affects approximately 40,000 infants each year. While the mortality rate is currently below 10%, a higher rate of neurodevelopmental disabilities is apparent among school-age survivors of CHD.

Jennifer Lynch, MD, PhD, an attending physician in the Division of Cardiothoracic Anesthesiology and a physicist specializing in biomedical optics, aims to improve these outcomes with advanced neuromonitoring to detect periods of neurologic risk in CHD.

The Lynch Lab for Novel Biomedical Optics is trying to determine when and how this neurologic injury is occurring in these children. Their innovative approach revolves around building a novel device for research purposes that uses light to measure cerebral oxygen delivery and utilization. This device is not used clinically, but evidence of its clinical utility for children with CHD is mounting, and Dr. Lynch is at the forefront of this research.

"Individualizing care is important for the outcomes of our patients, and the use of this research device may help to improve neuroprotection by matching care to the individual needs of each child," Dr. Lynch said. "Our quest is to make a real impact in pediatric cardiac care, one optical innovation at a time."

Children's Hospital of Philadelphia researchers developed a new artificial intelligence-powered algorithm to help understand how different cells organize themselves into particular tissues and communicate with one another. This novel tool was tested on two types of cancer tissues to reveal how these "neighborhoods" of cells interact with one another to evade therapy.

Kai Tan, PhD, Director of the Center for Single Cell Biology, led a team of researchers using the algorithm to identify tissue cellular neighborhoods (TCNs) based on cell identities of a tissue sample and their spatial distributions, as well as patient clinical data that can help researchers better understand how these neighborhoods of cells are organized and are associated with certain clinical outcomes.

Tissue samples from breast and colorectal tumors, of which there is a large volume of available data, were used to train the algorithm — called CytoCommunity — to identify TCNs associated with high-risk disease subtypes. The algorithm revealed new fibroblast-enriched TCNs and granulocyte-enriched TCNs specific to high-risk breast cancer and colorectal cancer, respectively.

"Using data from childhood cancers such as leukemia, neuroblastoma, and high-grade gliomas, we hope to find tissue cellular neighborhoods that might be associated with responses to certain therapies and combine our findings with genetic data to help determine which genetic pathways may be involved at the cellular and molecular levels," Dr. Tan said.

Scientists from Children's Hospital of Philadelphia are creating and studying enteroids and colonoids, also called mini-intestines, to investigate the role of stem cells in the epithelial lining of the gastrointestinal tract as both a driver and potential treatment for inflammatory bowel disease (IBD).

Using an algorithm that determined a "prediction score," the researchers' findings validated enteroids and colonoids as valuable models for studying the intestinal lining of patients with IBD.

"You can take stem cells from anywhere in the gastrointestinal tract, supply them with the right nutrients, and they will start to grow into a three-dimensional structure," said Kathryn Hamilton, PhD, an investigator in the Division of Gastroenterology, Hepatology, and Nutrition at CHOP. "The technology has revolutionized the field of cell physiology."

The research team also aimed to answer the question of whether the patient-derived mini-intestines could accurately predict if a patient had disease in their upper intestinal tract by looking at tissue in the rectum. When a patient is not experiencing a flare-up, their lower intestine often appears "normal," and a diagnosis could be missed during certain testing procedures. Colonoscopies enable more accurate diagnosis but are invasive.

The study team's results, published in Gastro Hep Advances, suggest that artificial mini-intestines could be used as a noninvasive diagnostic tool to screen for IBD.

Dr. Hamilton, Co-director of the Gastrointestinal Epithelium Modeling (GEM) Program at CHOP, along with Amanda Muir, MD, was joined in this work by Tatiana Karakasheva, PhD, first author and GEM Associate Director. This is just one of many avenues of research at CHOP leveraging the potential of organoids in the diagnosis and treatment of gastrointestinal disease.

Researchers from Children's Hospital of Philadelphia and Stanford Medicine found that a protein called FOXO1 may hold a key to unlocking the benefits of chimeric antigen receptor (CAR) T-cell therapy for more patients. Clinical efficacy of CAR T-cell therapy is linked to the durability of the cells in the body, but persistence of these cells in vivo is a major limitation of the life-saving treatment. FOXO1 plays a critical role in regulating longevity and effectiveness of CAR T cells.

"Understanding what makes CAR T cells more likely to yield a durable response at a molecular level may help improve the design of CAR T-cell therapies and potentially benefit a wider range of patients," said Evan Weber, PhD, senior study author and a cell and gene therapy researcher with the Center for Childhood Cancer Research and the Raymond G. Perelman Center for Cellular and Molecular Therapeutics at CHOP.

Dr. Weber and the research team explored the role of the transcription factor, FOXO1, in enhancing memory and antitumor function of CAR T cells by examining the effects of removing it from cells. They found that the absence of FOXO1 resulted in T cells' inability to form a healthy memory cell and therefore become ineffective at tumor control. On the other hand, enhancing FOXO1 activity drove CAR T cells to be more memory-like, more persistent, and retain anti-tumor activity both in vitro and in vivo. The study was published in Nature.

The research team is now collaborating with labs at CHOP to analyze CAR T cells from patients with exceptional persistence to identify other proteins that could be leveraged to improve durability and therapeutic efficacy.

Artwork by Gerardo Sotillo

Genetic diseases that involve the central nervous system (CNS) often impact children before birth. Once a child is born, irreversible damage may have already been done. Given that many of these conditions result from a mutation in a single gene, there has been growing interest in using gene editing tools to correct these mutations before birth.

Researchers in the Center for Fetal Research at Children's Hospital of Philadelphia and Penn Engineering identified an ionizable lipid nanoparticle (LNP) that can deliver mRNA base editing tools to the brain. Their work demonstrated this LNP can moderate CNS disease in perinatal animal models. The findings, published in ACS Nano, open the door to mRNA therapies that could be delivered pre- or postnatally to treat genetic CNS diseases.

"This proof-of-concept study supports the safety and efficacy of LNPs for the delivery of mRNA-based therapies to the central nervous system," said co-senior author William Peranteau, MD, an attending surgeon in the Division of General, Thoracic and Fetal Surgery at CHOP and the Adzick-McCausland Distinguished Chair in Fetal and Pediatric Surgery. "Taken together, these experiments provide the foundation for additional translational studies."