Computational Modeling of Sodium Fluorescein Disposition in Ex Vivo, Machine-Perfused Livers with Ischemia-Reperfusion Injury

Rationale: Quantitation of bile formation machinery activity through biliary sodium fluorescein (SF) clearance is useful for assessing liver viability. Studies have shown that the bile-to-plasma SF ratio is inversely correlated with warm ischemia time (WIT) in rats, which is associated with ischemia-reperfusion injury (IRI)-induced intracellular internalization of the multidrug resistance protein 2 (MRP2) transporter (doi: 10.1152/ajpgi.00038.2022). Computational modeling studies by Monti et al. have demonstrated the sensitivity of biliary SF clearance to MRP2 transporter activity in vivo (doi: 10.48550/arXiv.2302.05511). In vivo experimentation is pertinent for establishing physiological variations occurring during IRI, but transplantation often requires a period of ex vivo machine perfusion (MP). The effects of total hepatectomy, WIT, cold ischemia time (CIT), and MP on biliary function remain unclear. We aim to develop a model for quantitative interpretation of biliary SF clearance kinetics in MP livers to estimate parameters descriptive of the dominant processes determining hepatic SF uptake kinetics in MP livers subjected to different periods of WIT. We hypothesize that increasing WIT will lead to a decrease in the maximal transport velocity (Vmax) parameter for MRP2, which is reflective of decreased biliary SF clearance and diminished liver viability. Methods: Using established protocols (doi: 10.3389/frtra.2023.1215182), livers (n = 3-5) were obtained from anesthetized, male, Sprague-Dawley rats and subjected to 0-, 10-, 20-, or 30-min WIT through ligation of the portal vein and hepatic artery. Livers were removed and subjected to 3 hours of CIT on ice. Livers were attached to a custom normothermic MP system, 0.4 mg/kg SF were added to the MP system’s perfusate-containing reservoir, and SF fluorescence measurements in perfusate and bile were obtained and converted to concentration measurements using standard curves. We modified the liver-centric model described in Monti et al. to account for MP by replacing the liver input functions with ordinary differential equations for SF and its conjugate SF glucuronide (SFG) kinetics in the MP system’s reservoir. The model also accounts for three liver regions (sinusoid, hepatocytes, bile) and major processes, including hepatic transporter kinetics and SF metabolism into SFG, which govern SF dynamics in different model regions. We fit the model solutions using a pseudo-Monte Carlo parameter estimation strategy to each of the datasets by allowing the Vmax parameters to vary, while others were fixed to physiologic values. Results: Time course data for IRI conditions showed a modest decrease in biliary SF with 10- and 20-min WIT and a marked change in biliary SF kinetics with 30-min WIT when compared to control. The MP model was fit to each dataset individually and the model was able to provide a good fit to each dataset. Comparison of Vmax parameters estimated from fitting showed a decrease in MRP2’s Vmax with increasing WIT relative to control. Sensitivity analysis supported our finding by demonstrating that MRP2’s Vmax­ was the most sensitive model parameter. Conclusions: We modified an existing in vivo biliary SF clearance model to account for MP livers and fit the new model to biliary SF clearance data for ex vivo MP livers with varying amounts of WIT. We found that the new model was able to fit biliary SF clearance data for ex vivo MP livers and that the Vmax for MRP2 decreased during MP with increasing WIT. Funded in part by 1F30AI179084-01 to C.M. and NSF DMS 2153387 to R.D. This is the full abstract presented at the American Physiology Summit 2024 meeting and is only available in HTML format. There are no additional versions or additional content available for this abstract. Physiology was not involved in the peer review process.