Developing a mathematical model for early detection of viral hepatitis induced liver dysfunction

Abstract

Acute liver failure (ALF) is characterized by a sudden loss of hepatic function within a few weeks without pre-existing cirrhosis or prior history of liver disease. ALF can be difficult to identify using non-invasive methods such as coagulation abnormality and encephalopathy alone as alternative causes may confound the ALF diagnostic. This often results in delayed treatment at the onset of the disease, and a liver transplant might be the only option for ALF patients with a limited success rate. Due to these variables, ALF treatment is highly subjective to the clinicians' expertise, with a very high morbidity and mortality rate. Therefore, a prognostic model can help identify patients who can be managed with medical treatments and those who need liver transplants, along with an optimum intervention time window. Currently, only statistical models are available to identify ALF, such as King's College Hospital criteria, Clichy criteria, and End-Stage Liver Disease score. While these models provide much-needed critical support to clinicians, these models have limited specificity and sensitivity toward the cause of ALF and liver disease progression. For example, among all the common causes, the viral hepatitis-induced liver disease remains a major cause of ALF in developing countries such as in India and Bangladesh. Therefore, a dynamic patient-specific model is needed to improve patient care by correlating liver dysfunction with various healthcare variables and interventions. In this work, a new physiological linked data-driven model is formulated that can track the progression of liver disease and failure due to viral hepatitis. The approach relies on integrating the dynamics of healthy hepatocytes with immune system response, which is a vital component in tracking the time-sensitive infectious disease progression. The model uses the patient-specific measured values of alanine transaminase and aspartate transaminase enzymes and the clinically calculated international normalized ratio value to estimate the sensitive model parameter. This patient-specific model is then used to determine the disease progression by predicting the number of damaged hepatocytes and correlating it with the indication criteria for liver transplantation. The model can reproduce the essential dynamics seen in the ALF patient data and can be adapted to incorporate the effect of different therapeutic interventions. Thus, providing a vital bridge between understanding and treating viral hepatitis-induced liver dysfunction. No conflicts of interest, financial or otherwise, are declared by the author. This is the full abstract presented at the American Physiology Summit 2023 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.