(T-106) Simulated CD8+ T Cell-Mediated Liver Injury During Ipilimumab Administration in a Simulated Population (SimPops®) Demonstrates Profiles Consistent with Observed Clinical Data
Tuesday, November 12, 2024
7:00 AM – 5:00 PM MST
Lara Clemens, PhD – Senior Scientist, Simulations Plus Inc; Michael Kelley, PhD – Scientist I, Simulations Plus Inc; James Beaudoin, PhD – Senior Scientist, Simulations Plus Inc; Christina Battista, PhD – Principal Scientist, Simulations Plus Inc; Scott Siler, PhD – Chief Scientific Officer, Simulations Plus Inc; Lisl Shoda, PhD – Associate VP and Director of Immunology, Simulations Plus Inc; Brett Howell, PhD – QSP Solutions President, Simulations Plus Inc; Kyunghee Yang, PhD – Senior Principal Scientist, Simulations Plus Inc
Senior Scientist Simulations Plus Inc Pittsboro, North Carolina, United States
Disclosure(s):
Conner I. Sandefur, PhD: No financial relationships to disclose
Objectives: Immune checkpoint inhibitors (ICIs) release the brakes on immune responses allowing immune-mediated tumor cell killing and have revolutionized treatment of various cancers. Many ICIs are also associated with immune related adverse events (irAEs), including liver injury1. Of patients treated with the ICI ipilimumab, 7% exhibit signs of liver injury, e.g., alanine aminotransferase (ALT) elevations2. Ipilimumab targets CTLA-4, an inhibitory cell surface protein expressed on activated T cells. One hypothesis for ipilimumab-mediated liver injury is that CTLA-4 inhibition is permissive for normally suppressed de novo T cell responses to liver antigens. We applied a quantitative systems toxicology (QST) model of biologics-mediated liver injury (BIOLOGXsym™) to investigate this hypothesis.
Methods: BIOLOGXsym represents liver parenchymal and nonparenchymal cell dynamics, such as innate and adaptive immune responses. The modeled adaptive immune response leverages CD8+ T cell responses specific for hepatocyte-expressed antigen as previously described3,4. A SimPops was developed (N>500) to capture variability in CD8+ T cell numbers, differentiation and killing rates, regulatory inhibition of CD8+ proliferation, antigen presentation, and T cell receptor avidity, including overrepresentation of susceptible patients. Variability in ipilimumab pharmacokinetics, patient anthropometric characteristics, and liver biochemistry were not included. Ipilimumab was modeled as inhibiting early regulatory signaling in the T cell activation cascade, leading to a net increase in T cell avidity for hepatocyte-expressed antigen and aberrant expansion of hepatocyte specific CD8+ T cells. Virtual patients (VPs) with peak ALT elevations consistent in timing and magnitude with to clinically reported ipilimumab liver injury were identified. Pairwise correlation plots were then used to establish CD8+ T cell response characteristics driving VP ALT dynamics.
Results: By design, simulated ipilimumab induced CD8+ T cell mediated injury in more patients than are observed clinically, 205/512 (40%). SimPops captured ALT elevation magnitudes and timings observed in the clinic5-8. Pairwise analysis demonstrated associations between lower avidity and delayed peak ALT and higher initial naïve CD8+ T cell numbers and higher peak ALT. Additionally, identified VPs showed dynamics that matched available clinical time series data for ALT levels in patients treated with ipilimumab.
Conclusions: The undeniable benefit of ICIs is tempered by the risk of irAEs, which are poorly understood and therefore difficult to avoid. These results provide an initial demonstration of how QST can be applied to explore mechanistic hypotheses and identify key drivers of ICI liver injury. Further, these results set the stage for collaborations to generate additional data to inform key parameters, confirm/refute assumptions, and improve our understanding of these important safety concerns. NIH-R44TR003535.
