Introduction: Crohn's disease (CD) is a chronic immune-mediated inflammatory bowel disease that results in relapsing and remitting symptoms but progressive transmural bowel damage leading to significant morbidity. CD results from dysregulation of the immune system related to genetic and environmental factors. While the use of monoclonal antibodies targeting cytokines and adhesion molecules has been shown to improve outcomes in CD patients, their widespread use has been limited due to high costs as well as variable access. Here, we summarize the factors that have been shown to correlate with responsiveness to biologic agents for use in practice.
Areas covered: We summarize the current literature regarding factors that have been shown to influence patient response to various biologic agents including: patient-related factors (e.g. age, gender, weight smoking history); disease-specific factors (e.g. disease duration, location/extension, behavior/phenotype, severity); genetic markers; transcription factors, and the gut microbiome. Finally, we review the utility of prediction models and present data supporting the use of recently developed decision support tools.
Expert opinion: Clinical decision support tools developed by machine learning are currently available for the selection of biologic agents in CD patients. We expect these models to become an integral tool for clinicians in the treatment of CD in the coming years.
Keywords: Artificial intelligence; Crohn’s disease; biologic agent; clinical decision support tool; machine learning; monoclonal antibody; predictive models.