(M-109) Advancing Quantitative Systems Pharmacology Model for Inflammatory Bowel Disease for Clinical Efficacy Predictions in Ulcerative Colitis

Yue Fang, Ph.D. – Quantitative Systems Pharmacologist, Development China, Pfizer; Pei Luo, Ph.D. – Quantitative Systems Pharmacologist, Development China, Pfizer; Shinichi Tsuchiwata, Ph.D. – Pharmacometrician, Development Japan, Pfizer; Yamato Sano, M.S. – Clinical Pharmacologist, Development Japan, Pfizer; Bryce Johnson, M.S. – Senior Principal Scientist, Inflammation and Immunology Research Unit, Pfizer; Vivek Purohit, Ph.D. – Pharmacometrics Group Lead, Translational Clinical Sciences, Pfizer; Srividya Neelakantan, Ph.D. – Director, Translational Clinical Sciences, Pfizer; C.J. Musante, Ph.D. – Head of Pharmacometrics and Systems Pharmacology, Translational Clinical Sciences, Pfizer; Richard Allen, Ph.D. – QSP Group Lead, Translational Clinical Sciences, Pfizer

Objective: Inflammatory Bowel Disease (IBD) is a chronic autoimmune disease associated with gastrointestinal inflammation. While therapeutic options for the disease have expanded, patient response to these treatments can be highly variable. The aim of our work is to advance a Quantitative Systems Pharmacology (QSP) model of IBD such that it can predict efficacy at the level of relevant biomarkers with extension to clinical endpoints of remission for known mechanisms, and incorporate additional mechanisms to support the development of novel therapeutic approaches which simultaneously target multiple immunomodulatory pathways.

Methods: A previously published QSP model of IBD that accounts for immune cells and cytokine interaction [1,2] was expanded to account for epithelial damage in the presence of pro-inflammatory cytokine and a slower healing process. A virtual population for ulcerative colitis (UC) was developed and calibrated using published Phase 3 data for ustekinumab (anti-p40, a subunit of IL-12 and IL-23 cytokines) [3] and Phase 2 data for anti-TL1A (PF-06480605) [4]. A probabilistic connection was made between fecal calprotectin (FCP), a biomarker indicative of disease burden, with endoscopic score, a clinical endpoint associated with mucosal healing.

Results: We found that the epithelial healing mechanism is essential for capturing the observed slow FCP decline observed in clinical data, but not captured in the initial model. The simulated response of the virtual population covers the variability of CRP and FCP response during induction period observed in ustekinumab and anti-TL1A clinical data. By pooling and comparing data from several clinical trials, we established a quantitative relationship between FCP and endoscopic healing, with lower FCP scores more likely to correspond to endoscopic score 0-1 compared to those for higher scores. The model-generated predicted scores are in line with observed data in ustekinumab and anti-TL1A UC trials [3,4].

Conclusion: We extended a QSP model for IBD to integrate mechanisms for epithelial healing and for a novel target, TL1A under clinical development [4]. Previously, the model was partially calibrated to published clinical data for Chron’s Disease [2]; here we further refine and calibrate a virtual population for UC in response to two separate treatments. In the future, the model can be further developed to account for additional mechanisms and utilized to predict biomarker response and endoscopic score for novel IBD therapies.

Citations: 1. Rogers KV, et al. (2021) Clinical and Translational Science,14(1):239-48.
2. Rogers KV, et al. (2021) Clinical and Translational Science,14(1):249-59.
3. Sands BE, et al. (2019) New England Journal of Medicine, 381(13):1201-14.
4. Danese S, et al. (2022) Clinical Gastroenterology and Hepatology, 1;20(12):2858-67.