By Mia Feldman

Graphic design by Yu-Wen Jan

Inflammatory bowel disease (IBD) is on the rise worldwide. In Canada, the number of cases is predicted to increase by approximately 1.3-fold by 2035.¹ IBD is used to describe two chronic intestinal diseases—Crohn’s disease (CD) and ulcerative colitis. Both are chronic immune mediated disorders of the intestine with a relapsing and remitting course.² Whereas ulcerative colitis affects the large intestine, CD can occur in any part of the gastrointestinal tract. Both result in a range of gastrointestinal symptoms, including abdominal pain, diarrhea, and rectal bleeding. In serious cases, individuals may ultimately have surgical resection of the intestine.²

Unfortunately, the cause of the disease remains unknown and there is no cure. This urgent medical mystery is what attracted Dr. Sun-Ho Lee to gastroenterology. Dr. Lee, who is also an IMS alum, practices as a Clinical Scientist at the Inflammatory Bowel Disease Centre at Mount Sinai Hospital. He was appointed to Assistant Professor at the University of Toronto in 2023. He focuses on translational IBD research, splitting his time between seeing patients and tackling a new approach to IBD treatment: precision medicine.

Precision Medicine in IBD

IBD is a very heterogenous disease, meaning its causes and clinical manifestations vary greatly from person to person.¹ Precision medicine, which predicts disease outcomes based on patient-specific variables, is conceptually a very useful tool in the treatment of IBD. For example, does having a parent with IBD increase your risk of having disease? Is there a microbial or genetic factor that predicts your risk of having disease complications? What about having a diet rich in refined sugars? Dr. Lee describes that ideally, precision medicine would be used for “stratifying a patient to accurately predict the course of disease or the treatment response to a specific medication,” ensuring the best outcome for each individual. 

To establish predictive models for IBD based on patient characteristics, significant amounts of data must be collected. As a clinician-scientist at the IBD Centre at Mount Sinai Hospital, Dr. Lee has unique access to patient samples, such as intestinal biopsies, as well as stool, urine, and blood samples. Data collected from these samples can be used to extrapolate information about the patient’s genetic susceptibility to IBD and microbiome composition, along with more clinical markers of gut inflammation and gut barrier function, for example. Collecting samples from patients at various time points reveals the chronology of disease progression. Scientists like Dr. Lee can also use this data to classify patients into defined groups based on their symptoms and how they progress. This process of data collection and classification is fundamental in the development of a predictive precision model.

The timing of this data collection is crucial. Dr. Lee explained that in most cases, once a patient is diagnosed with IBD after exhibiting the first symptoms, they have already undergone a series of treatments such as antibiotic courses, along with other medications. Therefore, studying a cohort of healthy at-risk participants before they become sick is necessary to capture a full and unbiased picture of IBD development and progression. 

The GEM Cohort: A New Type of Dataset

Dr. Lee is a lead researcher in the Genetic Environment and Microbial (GEM) Project. The GEM Project was founded in 2008 by Dr. Kenneth Croitoru—a leading physician in the Zane Cohen Centre for Digestive Diseases, and faculty member at the IMS. This international study follows the healthy first-degree relatives (such as siblings and children) of CD patients, collecting data (like blood and urine samples) from them until they themselves develop CD. Researchers compare baseline biomarkers of those that later develop CD to those that remain healthy, offering the opportunity to understand the pre-clinical phase of CD. This helps identify early markers and create predictive tools for development of CD. 

One of Dr. Lee’s earlier projects with GEM assessed participant intestinal barrier function. Dr. Lee and his colleagues used markers in participant urine to characterize the state of the intestinal barrier.³ They found that those who eventually developed Crohn’s showed signs of abnormal barrier function years before exhibiting symptoms or being diagnosed. This was the first study to prove abnormal barrier function is a predictor of developing CD. Without the baseline data from patients, collected long before any diagnosis, this finding wouldn’t have been possible. More recently, Dr. Lee and his colleagues published their findings on the gut microbiome and its implications in CD.⁴ By comparing the composition of each participant’s gut microbiome (pre-clinical versus post-diagnosis), they concluded that an altered microbiome could predict CD. Thus, Dr. Lee’s work provides the IBD world with two potential predictive methods for CD–assessing patient barrier function through the urine and examining their gut microbiomes.

Notably, all participants in the GEM project undergo genetic testing. Dr. Lee explained how researchers found little evidence that genetics predicted the likelihood of developing Crohn’s within the GEM cohort. Therefore, environmental factors, including diet and gut microbiome, likely contribute to elevated disease risk. Evidently, there is a seemingly endless number of variables that can contribute to, and markers that indicate, the development of IBD. 

Next Steps

Taking all patient variables into account is not an easy task. To integrate high dimensional data, Dr. Lee looks to new technology including machine learning (ML). ML algorithms draw conclusions from mass amounts of data. Dr. Lee and his colleagues have recently developed a ML model that integrates data derived from individual first-degree relatives of those with CD to predict their risk of developing the disease.⁵ For example, participant profiles can be created via data from urine samples, microbiota characteristics from stool samples, demographic traits, information about diet, and any other relevant information. Inputting these profiles into a ML model, then training and testing them with datasets, serves as a predictive tool for individual patient risk for CD. Dr. Lee’s current model is 80% accurate—showing promise for the future of this technology in clinical settings.

But Dr. Lee isn’t satisfied with 80% accuracy, he wants 100%. He believes there is a lot more work to be done. Dr. Lee is inspired by the researchers who have worked on type 1 diabetes (T1D) and rheumatoid arthritis. Their decades of work have resulted in a nuanced understanding of subtypes of disease and the diversity of pre-clinical phases before diagnosis. A thorough understanding of pre-clinical phases of T1D has enhanced disease screening and created avenues for early interventions to improve patient outcomes and to prevent or delay the development of disease. Dr. Lee hopes similar work will be done for IBD, especially with growing evidence that there is a prolonged pre-clinical phase in the disease. Additionally, with his newly approved grant from the National Institute of Diabetes, Digestive and Kidney Diseases IBD Genetic Consortium, he plans to further develop ML models integrating multi-omics to better stratify the risk of CD recurrence after surgical resection of the intestine. 

Dr. Lee’s work exemplifies the promise of precision medicine. The heterogeneous nature of the disease means there is no “one size fits all” solution for patients. Accounting for individual patient variables is therefore fundamental in optimizing treatment courses, ensuring the best outcomes and the most efficient healthcare systems. Though current predictive models are insufficient for clinical use, Dr. Lee and his colleagues are working to close the gap in knowledge with innovative solutions. His efforts bring us closer to a future where personalized care is the standard for IBD treatment.

References

1. Coward S, Benchimol E, Bernstein CN, et al. Forecasting the Incidence and Prevalence of Inflammatory Bowel Disease: A Canadian Nationwide Analysis. Am J Gastroenterol. 2024;119(8):1563-1570. 
2. John Hopkins Medicine. Inflammatory Bowel Disease (IBD); 2024 [cited 2024 November 15]. Available from: https://www.hopkinsmedicine.org/health/conditions-and-diseases/inflammatory-bowel-disease
3. Turpin W, Lee SH, Garay JAR, et al. Increased Intestinal Permeability Is Associated With Later Development of Crohn’s Disease. J. Gastroenterol. 2020;159(6):1995-2250. 
4. Garay JAR, Turpin W, Lee SH, et al. Gut Microbiome Composition Is Associated With Future Onset of Crohn’s Disease in Healthy First-Degree Relatives. J Gastroenterol. 2023;165(3):670-681
5. Lee SH, Turpin W, Espin-Garcia O, et al. Development and Validation of an Integrative Risk Score for Future Risk of Crohn’s Disease in Healthy First-Degree Relatives: A Multicenter Prospective Cohort Study. J Gastroenterol. 2024;S0016-5085(24):5401-5405