Intestinal Organoid-Based Mathematical Modeling Predicts Clinical Gastrointestinal Toxicity of Oral Oncology Drugs

Area of Science:

• Gastroenterology
• Oncology
• Pharmacology
• Computational Biology

Background:

• Gastrointestinal (GI) toxicity is a frequent and severe adverse event in antiproliferative cancer therapy, often necessitating treatment adjustments.
• Current preclinical methods lack robust quantitative assessment for GI toxicity, hindering effective drug development and patient management.

Purpose of the Study:

• To develop and validate a predictive model for GI toxicity of oral anticancer agents using human small intestinal organoids (hSIOs) and mathematical modeling.
• To integrate in vitro drug-exposure data with mathematical models of the intestinal epithelium for simulating toxicity.
• To establish a novel paradigm for GI safety assessment and clinical dose selection in oncology.

Main Methods:

• Utilized human small intestinal organoids (hSIOs) to quantify the in vitro exposure-toxicity relationship of oral antiproliferative drugs.
• Developed a mathematical model of the human intestinal epithelium incorporating hSIO-derived toxicity data.
• Simulated the impact of impaired crypt proliferation on epithelial dynamics, using enterocyte-free drug concentration as a surrogate for crypt exposure.

Main Results:

• The model accurately correlated enterocyte-specific drug exposure with clinical incidence and severity of diarrhea for oral anticancer drugs.
• Reliance on plasma drug concentration failed to predict intestinal injury for a subset of tested drugs.
• The approach successfully differentiated the GI toxicity profiles of CDK4/6 inhibitors (ribociclib vs. abemaciclib) and EGFR-TKIs, aligning with clinical observations of diarrhea.

Conclusions:

• A novel hSIO-based mathematical modeling approach enables accurate preclinical prediction of oral anticancer drug-induced GI toxicity, particularly diarrhea.
• This method offers a more reliable assessment of drug safety compared to plasma exposure-based models.
• The developed paradigm facilitates improved GI safety assessment and clinical dose selection, paving the way for personalized cancer therapy.