A Quantitative Systems Pharmacology (QSP) Model to Compare the Non-Clinical Biodistribution and Efficacy between Recombinant Factor IX (rIX) Therapies

Background and Aims: Replacement FIX therapy (rIX) is an effective treatment for hemophilia B even with undetectable levels in the blood ¹. However, the mechanistic reason for hemostasis with low plasma levels is not well understood. There is growing evidence that FIX interactions with one or multiple binding partners (BP), may play a significant role in the exposure and hemostatic efficacy of rIX ^2,3. The aim of this study is to explore this hypothesis by comparing the plasma PK, tissue biodistribution, and in vivo endpoints of different rIX variants using a mouse QSP model.

Method:  in vitro and in vivo FIX-KO mice studies and mathematical models were used to build a QSP model consisting of 8 tissue compartments , with each tissue divided into vascular, endothelial and interstitial spaces ^4,5. The model simulates endogenous mouse IgG (mIgG), mouse serum albumin (MSA), and rIX dynamics including key clearance and distribution mechanisms. Competition for the endothelial FcRn receptor between Fc, albumin, mIgG, and MSA is explicitly modeled ^6,7,8. The model was calibrated using mouse studies of radiolabeled rIX-Fc (Alprolix®), rIX-WT (BeneFIX®), and rIX-FP (Idelvion®). Tail-clip experiments following administration of rIX-Fc, rIX-WT, and rIX-FP were used to correlate the predicted exposures with the observed effects on bleeding time and total blood loss.

Results: Preliminary simulations proved that having at least one BP best explains the rapid distribution of rIX-Fc and rIX-WT into the tissues, and the long plasma T _1/2 of rIX-Fc and rIX-FP. Visual predictive checks of the full PBPK model showed good agreement with the PK in the tissues. The best fit was achieved using a specific arrangement of four distinct binding partners:

1. Shared BP (SBP) between all compounds (e.g. N-terminal binder) located within the vasculature with estimated K _D of 470/600/4100 nM, for rIX-WT/rIX-FP/rIX-Fc, respectively.

2. BP binding specific to rIX-WT (e.g. C-terminal binder) located in the interstitium of the tissue (varying densities) with estimated K _D of 23 nM

3. BP binding only for rIX-FP (e.g. albumin binder) located in both; the vasculature and interstitium of the tissue with estimated K _D 20/0.05 μM (vascular/interstitial)

4. BP binding only for rIX-Fc (e.g. Fc binder) located in the interstitium of tissue (varying densities) with estimated K _D 3 μM

The high degree of extravasation of rIX-Fc (and rIX-WT to a lesser degree) results in rapid distribution and sequestration in the tissues. The limited extravasation of rIX-FP and its high affinity to the SBP, results in increased recovery and a greater pool of bound rIX available in the tissue vasculature. Additionally, strong inverse correlation between the bound rIX in the vasculature and bleeding time/total blood loss suggests that the vascular pool plays a more significant role in FIX pharmacology, as compared to the pool in the extravascular space.

Conclusion: The mouse QSP model demonstrated that the plasma and tissue biodistribution of rIX-Fc, rIX-FP, and rIX-WT cannot be explained without a BP, and that it is plausible to assume that different binding partners, both intra- and extravascular, for different rFIX variants exist. The correlation between the levels of bound rIX and the coagulation endpoints suggests that the vascular bound rIX may be the pharmacologically active pool or reservoir for haemostasis. The extravasation and sequestration of rIX-WT and rIX-Fc into the tissues may explain the decreased vascular exposure, and hence, the reduced efficacy (increased bleeding time/total blood loss) at later time points. Although the exact identity of the BP's remains to be further elucidated, the model estimates of their affinity, density and location provide guidance for further experimental investigations. Expansion of the QSP model with additional data and coagulation kinetics will further our understanding of the role of BPs in rIX pharmacology.

References

¹Srivastava A et al (2013) Haemophilia 19(1), e1-47

²Feng D et al (2013) JTH, Vol. 11 (12), 2176-2178

³Cheung WF et al (1996) PNAS USA, 93(20), 11068-11073

⁴Li L et al (2014) AAPS Journal 16(5), 1097-1109

⁵Shah DK & Betts AM (2012) J Pharmacokinet Pharmacodyn 39(1), 67-86

⁶Chia J et al (2018) J Biol Chem 293(17), 6363-6373

⁷Andersen JT et al (2010) J Biol Chem 285(7), 4826-4836

⁸Andersen JT et al (2013) J Biol Chem 288(33), 24277-24285