Desmopressin (1-deamino-8-D-arginine vasopressin, DDAVP) is commonly used as a nonreplacement therapy for mild von Willebrand disease (VWD) and hemophilia A. In humans, IV injection of 0.3 μg/kg of DDAVP induces a rapid 2 to 5-fold increase in plasma levels of both von Willebrand factor (VWF) and Factor VIII (FVIII) within 30–60 minutes, which is due to release from Wiebel-Palade bodies (WPBs) in endothelial cells. The stored FVIII may be synthesized by endothelial cells, which express FVIII in vitro. However, hepatoma cells can also express FVIII in vitro, and liver transplantation can correct hemophilia A. Thus, the liver may be the major site of production of FVIII in vivo, thus, an alternative explanation is that endothelial cells take up FVIII from blood and store it in WPBs with VWF, which can be released after DDAVP. DDAVP is effective in humans and dogs, but not in mice. In this study, we tested the effect of DDAVP on hemophilia A dogs after neonatal hepatic gene therapy with a retroviral vector (RV) expressing canine FVIII (cFVIII). With this gene therapy approach, canine hepatocytes express high levels of a reporter gene from an RV, but no expression is observed in endothelial cells. Thus, the major site of FVIII synthesis is the hepatocyte in this model. Our hypothesis is that if DDAVP increases FVIII levels in this dog model, it would indicate that the FVIII increase is due to uptake from blood by endothelial cells. Alternatively, if no increase in FVIII occurs after DDAVP stimulation, it would suggest that the increase in normal dogs is due to synthesis of FVIII by endothelial cells. An RV that contains the liver-specific human α1-antitrypsin promoter and the canine B-domain deleted FVIII cDNA was generated. RV was given IV to two hemophilia A dogs at 8x109 transducing units (TU)/kg at 3 days after birth. The whole blood clotting time (WBCT) and APTT time in both dogs have been normalized, and the plasma cFVIII COATEST activity has been maintained at 100–200% of normal for 11 months to date. DDAVP was injected IV at 0.5 μg/kg into RV-treated hemophilia A dogs at 7 months of age. Two separate doses of DDAVP were given with an interval of one week. The same dose of DDAVP was given to normal dogs as controls (N=4). In normal dogs, both VWF and FVIII levels increased 40% and 50% between 15 to 60 minutes after DDAVP, respectively. However, FVIII levels were not changed in RV-treated dogs, although VWF levels increased 150% or 60%. Thus, our data suggest that the normal FVIII increase after DDAVP administration is due to release of FVIII that is synthesized by endothelial cells. These data also demonstrate that DDAVP will not be effective at increasing FVIII activity in patients that receive liver-directed gene therapy and only achieve partial correction. Such patients would need to be treated with factor replacement if bleeding episodes occur.

Author notes

Corresponding author

Sign in via your Institution