Species differences in liver accumulation and metabolism
Image is modified from the graphical abstract and is showing the liver triphosphate levels across species at Human Equivalent Dose (µM)

Species differences in liver accumulation and metabolism of nucleotide prodrug are discussed in the April 2020 publication by Gilead team.

The difficulty in projecting human dose based on preclinical data is discussed based on sofosbuvir – a compound with complex first pass metabolism. All its phosphate metabolites (monophosphate, diphosphate and triphosphate) are not observed in the systemic circulation and are retained in hepatocytes. Hence, the prediction of human hepatic exposures of the active triphosphates of nucleoside/nucleotide prodrugs is challenging as plasma exposure purely reflects liver levels.

In the attempt to find the most suitable pharmacokinetic (PK) model to project human liver tissue levels for future nucleotide prodrugs, authors re-evaluated the following models in in vitro and in vivo setting:

  • Mouse models (C57BL/6 and/or FVB);
  • PXB-mouse® (cDNA-uPA/scid mouse with highly humanized liver);
  • Rat model (Sprague-Dawley rat);
  • Hamster (Golden Syrian hamster);
  • Dog (Beagle dog);
  • Monkey (Cynomolgus monkey).

Liver accumulation of triphosphate was investigated after oral administration of sofosbuvir. The intestinal absorption, which may be affected by species differences in efflux transport, is followed by exposure to plasma in portal vein and distribution into the liver. Uptake of the drug is anticipated to happen by passive diffusion, as this nucleotide prodrug is not a substrate for hepatic uptake transporters.

PXB-cells - Human Hepatocytes isolated from PXB-mouse
PXB-cells – Human Hepatocytes isolated from PXB-mouse

In vitro studies on intracellular formation of the active triphosphate have shown that PXB-cells® generated the same results as primary human hepatocytes (PHH). In addition, half-life (of 20 h) and Cmax shown high levels of triphosphate metabolites in both PXB-cells® and human hepatocytes. The half-lives of the intracellular triphosphate were generally consistent with the in vivo liver profiles; and on PXB-mouse® they were significantly longer than in non-humanized mouse.

The summary suggests that a humanized rodent model could be used to effectively model human liver loading.

For further details see Wang T. et al. “Species differences in liver accumulation and metabolism of nucleotide prodrug sofosbuvir” in Drug Metabolism and Pharmacokinetics (2020).