Using blood proteomics to identify pediatric inflammatory bowel disease risk

For children with suspected inflammatory bowel disease (IBD), a chronic, relapsing condition of the gastrointestinal tract, getting the right diagnosis relies on a combination of clinical evaluation, imaging, endoscopy, and histopathology. Identifying reliable blood-based diagnostic tools remains an important unmet clinical need.

In a study published in eBioMedicine researchers at Beth Israel Deaconess Medical Center (BIDMC) and Mass General Brigham for Children used proteomics - the large-scale measurement of proteins in the blood - to identify distinct patterns of proteins that could accurately identify children with inflammatory bowel disease.

Our findings measuring expression of 1,300 proteins in blood samples demonstrate the utility of blood proteomics in identifying disease-specific biomarkers and new non-invasive tests for differential diagnosis in pediatric IBD as well as defining the underlying pathophysiological mechanisms. A validated blood-based diagnostic approach could help reduce diagnostic delays, minimize invasive procedures, and support earlier, more personalized treatment decisions for children with IBD."

Towia Libermann, PhD, co-senior author, director of the Genomics, Proteomics, Bioinformatics, and Systems Biology Center at BIDMC

"Non-invasive diagnostic tools can contribute to the diagnosis of inflammatory bowel disease but are not yet at the point of replacing standard clinical and pathologic data," said co-senior author Harland Winter, MD, Co-Director of the Center for Pediatric Inflammatory Bowel Disease at Mass General Brigham for Children. "Nevertheless, the data from these new proteomic platforms represent a much-needed advancement in the care of children with Crohn's disease and ulcerative colitis as we learn how best to not only treat diseases but also to personalize treatment."

Led by Libermann and Winter, the team analyzed blood samples from an initial group of 47 children with and without inflammatory bowel diseases. Using a proteomics platform capable of measuring more than 1,000 proteins at once, the team identified a distinct protein pattern in children with IBD. These patterns reflect small shifts in the quantities and combinations of many proteins - signals that are difficult to detect when each marker is considered on its own.

While individual protein markers have long been studied in IBD, this broader approach allowed the researchers to identify a high-level pattern and then narrow it down to a smaller, more practical set. Libermann, Winter and colleagues identified 95 proteins elevated in children with IBD, along with 70 that helped distinguish its two main forms, ulcerative colitis and Crohn's disease. The team further narrowed the field to eight proteins using machine learning that showed strong diagnostic performance in their initial analysis.

To make the approach more practical for potential clinical use, the researchers then reduced the signal to four proteins and validated them using conventional, clinically available tests in two larger groups of 295 and 105 children, respectively. The four-protein test identified IBD with high accuracy, in the 80- to 90-percent range. A separate four-protein test achieved more than 90 percent predictive performance in differentiating ulcerative colitis from Crohn's disease.

Beyond improved diagnostics, the newly identified protein patterns also offer insight into how inflammatory pathways are altered in children with inflammatory bowel diseases, helping explain differences between Crohn's disease and ulcerative colitis and may inform future efforts to develop targeted therapies and precision medicine approaches.

Co-authors included Mmeyeneabasi Omede, Jui-Yen Huang and Catherine Pursley of Mass General Brigham for Children; Hasan H. Out and Handan Can of University of Nebraska-Lincoln; Laurie B. Grossberg, Xuesong Gu and Simon T. Dillon of BIDMC; and Tina Lung Morhardt of Dartmouth Hitchcock Clinics, Manchester, Geisel School of Medicine.

This study was supported by philanthropic support from Martin Schlaff, The Diane and Dorothy Brooks Foundation, The Manessis Family, Thomas and Lynn Kuzma, and The Hasso Serrano Foundation.