Comparative pharmacokinetics of factor VIII and recombinant factor IX: for which coagulation factors should half-life change with age?


  • S. Björkman

    Corresponding author
    1. Department of Pharmaceutical Biosciences, Uppsala University, Uppsala, Sweden
    • Correspondence: Sven Björkman, Department of Pharmaceutical Biosciences, Uppsala University, Box 591, SE-751 24 Uppsala, Sweden.

      Tel.: +46 18 471 4995; fax: +46 18 471 4003;


    Search for more papers by this author
    • We regret to note that Prof. Sven Björkman has since passed away. Therefore, any correspondence about this article should be directed to Prof. Margareta Hammarlund-Udenaes at


The half-life of factor VIII (FVIII) increases with the age of the patient, while studies on recombinant factor IX (rFIX) and factor VIIa (rFVIIa) have not demonstrated corresponding age-related changes. The purpose of this analysis was to relate the changes in FVIII and rFIX pharmacokinetics (PK) with age to developmental changes in body size and fluid volumes and explain why the elimination half-life of FVIII, but not of rFIX, would change with age, and to consider how the findings could be applied prospectively to other coagulation factors. Published PK data for FVIII from 186 patients aged 1–74 years and for rFIX from 56 patients aged 4–56 years were used. The relationships of FVIII and rFIX clearance (CL) with body weight could be described by allometric expressions. Relative changes in CL with age or weight were similar for FVIII and rFIX. The age-related change in volume of distribution at steady state (Vss) of rFIX was parallel to the change in CL in the children while for FVIII the change was much less pronounced. Elimination half-life was clearly age-dependent for FVIII while only a very weak trend could be seen for rFIX. Limited data suggest that rFVIIa in this respect resembles rFIX, with parallel changes in CL and Vss producing insignificant change in half-life. To what extent the elimination half-life of a coagulation factor would show a correlation with age can in principle be predicted from the characteristics of its CL and distribution.