A semi-mechanistic model to predict the effects of liver cirrhosis on drug clearance

Abstract

Background and objective: Liver cirrhosis is characterized by a decrease in functional hepatocytes, lowered circulating levels of plasma proteins and alterations in blood flow due to the development of portacaval shunts. Depending on the interplay between these parameters and the characteristics of an administered drug, varying degrees of impaired systemic clearance and first-pass metabolism are anticipated. The Simcyp Population-based ADME Simulator has already been used successfully to incorporate genetic, physiological and demographic attributes of certain subgroups within healthy populations into in vitro-in vivo extrapolation (IVIVE) of xenobiotic clearance. The objective of this study was to extend population models to predict systemic and oral drug clearance in relation to the severity of liver cirrhosis.

Methods: Information on demographics, changes in hepatic blood flow, cytochrome P450 enzymes, liver size, plasma protein binding and renal function was incorporated into three separate population libraries. The latter corresponded to Child-Pugh scores A (mild), B (moderate) and C (severe) liver cirrhosis. These libraries, together with mechanistic IVIVE within the Simcyp Simulator, were used to predict the clearance of intravenous and oral midazolam, oral caffeine, intravenous and oral theophylline, intravenous and oral metoprolol, oral nifedipine, oral quinidine, oral diclofenac, oral sildenafil, and intravenous and oral omeprazole. The simulated patients matched the clinical studies as closely as possible with regard to demographics and Child-Pugh scores. Predicted clearance values in both healthy control and liver cirrhosis populations were compared with observed values, as were the fold increases in clearance values between these populations.

Results: There was good agreement (lack of statistically significant difference, two-tailed paired t-test) between observed and predicted clearance ratios, with the exception of those for two studies of intravenous omeprazole. Predicted clearance ratios were within 0.8- to 1.25-fold of observed ratios in 65% of cases (range 0.34- to 2.5-fold).

Conclusion: The various drugs that were studied showed different changes in clearance in relation to disease severity, and a 'one size fits all' solution does not exist without considering the multiple sources of the changes. Predictions of the effects of liver cirrhosis on drug clearance are of potential value in the design of clinical studies during drug development and, clinically, in the assessment of likely dosage adjustment.