A Pharmacokinetic/Pharmacodynamic Model of Recombinant Batroxobin
Jun Seok Cha(1,2), Do Hoon Keum(3,4), Choon Ok Kim(3), Min Soo Park(3,4,5), Dongwoo Chae(1,3,*)
(1) Department of Pharmacology, Yonsei University College of Medicine, Seoul, Korea. (2) Department of Pharmacology, Graduate School of Medical Science, Brain Korea 21 Plus Project, Yonsei University College of Medicine, Seoul, Korea. (3) Department of Clinical Pharmacology, Severance Hospital, Yonsei University College of Medicine, Seoul, South Korea. (4) Department of Pharmaceutical Medicine and Regulatory Sciences, College of Medicine and Pharmacy, Yonsei University, Incheon, South Korea (5) Department of Pediatrics, Yonsei University College of Medicine, Seoul, Korea. (*) To whom correspondence should be addressed.
Objectives: Batroxobin is a thrombin-like snake venom protein purified from the venom of Bothrops atrox moojeni[2]. It cleaves only the fibrinopeptide A of fibrinogen, unlike thrombin, which cleaves both fibrinopeptide A and fibrinopeptide B [5]. Paradoxically, batroxobin exhibits both hemostatic and anti-thrombotic effects[2,6]. This phenomenon occurs because a more soluble fibrin web is formed when only fibrinopeptide A is cleaved and because batroxobin does not activate factor XIII that stabilizes the fibrin web [1,3]. Recombinant Batroxobin, produced using the yeast Pichia pistoris system, has similar properties and activity as natural batroxobin [2,7]. Dynamical modeling may aid deciding the dosage regimen of batroxobin. We aimed to build a nonlinear mixed effect model describing batroxobin and fibrinogen dynamics using data extracted from publications and clinical trial data provided by Severance Hospital Clinical Trial Center.
Methods: Data for batroxobin concentration and fibrinogen concentration after single and multiple doses of native batroxobin was digitized from a publication (Sugai et al., 1986) [4]. The digitized data were subjected to model development and curve-fitting. The code was written in Python and executed using packages Numpy, Pandas, Scipy, Matplotlib. Clinical trial data was provided by Severance Hospital Clinical Trial Center, and the study was approved by the IRB. The trial was dose block-randomized, double-blinded, placebo-controlled, and single-dosing. Nonlinear mixed-effect modeling with SAEM algorithm was conducted on the clinical trial data using Monolix (2023R1), using the parameter estimates from the digitized data as typical vaules for Bayesian priors.
Results: A model supposing one-compartment distribution of batroxobin and fibrinogen explained the plasma concentrations from the extracted data well [4]. Estimated fibrinogen half-life of 4.0963 days conformed to previously reported values. The parameters estimated from the extracted data were used as priors to fit a mixed effects model to the clinical trial data. Only baseline fibrinogen level was assigned a random effect due to identifiability issues. The batroxobin clearance, half-life of fibrinogen, volume of distribution, baseline fibrinogen concentration, kon, and koff were each estimated to be 2.79L/h, 1.31 day, 1.8L, 244.84mg/dL, 3.81/h, 7504.99/h, which were physiologically plausible. Visual predictive check showed good concordance with observed data, although the model slightly under-predicted fibrinogen concentration after 12 hours.
Conclusions: A model of fibrinogen and batroxobin dynamics was successfully developed based on data extracted from a publication. Nonlinear mixed effects modeling was successfully carried out on clinical trial data from Severence Hospital Clinical Trial Center. Future work should examine the feasibility of building a covariate model incorporating the effect of the baseline lab test results and demographics on the random effect to suggest an optimal dosage regimen for batroxobin.
References:
[1] Braud, S., Bon, C., & Wisner, A. (2000). Snake venom proteins acting on hemostasis. Biochimie, 82(9-10), 851–859. https://doi.org/10.1016/s0300-9084(00)01178-0
[2] Choi, S. K., Kim, C. W., Kim, J. T., Seomun, Y., Park, M. S., & Kim, C. O. (2018). Coagulant Effect and Tolerability of Yeast-Produced Recombinant Batroxobin in Healthy Adult Subjects. Clinical drug investigation, 38(9), 829–835. https://doi.org/10.1007/s40261-018-0673-x
[3] Dempfle, C. E., Argiriou, S., Alesci, S., Kucher, K., Müller-Peltzer, H., Rübsamen, K., & Heene, D. L. (2001). Fibrin formation and proteolysis during ancrod treatment. Evidence for des-A-profibrin formation and thrombin independent factor XIII activity. Annals of the New York Academy of Sciences, 936, 210–214. https://doi.org/10.1111/j.1749-6632.2001.tb03507.x
[4] Sugai, K., Imamura, Y., Uechi, S., Namikata, S., Mihashi, S., Mori, N., Hada, M., Kurosoh, K., Ukita, M., & Fukutake, K. (1986). Metabolic fate of batroxobin in human. Yakugaku zasshi : Journal of the Pharmaceutical Society of Japan, 106(4), 335–342. https://doi.org/10.1248/yakushi1947.106.4_335
[5] Vu, T. T., Stafford, A. R., Leslie, B. A., Kim, P. Y., Fredenburgh, J. C., & Weitz, J. I. (2013). Batroxobin binds fibrin with higher affinity and promotes clot expansion to a greater extent than thrombin. The Journal of biological chemistry, 288(23), 16862–16871. https://doi.org/10.1074/jbc.M113.464750
[6] Wang, D. S., Hanamoto, M., Fang, F., Ohba, M., Ishii, M., Kimura, F., Higaki, E., & Senga, H. (2001). Defibrinogenating effect of batroxobin (Defibrase) in rats and inhibition of migration of human vascular smooth muscle cells by the plasma of batroxobin-treated rats in vitro. Atherosclerosis, 156(1), 73–80. https://doi.org/10.1016/s0021-9150(00)00628-6
[7] You, W. K., Choi, W. S., Koh, Y. S., Shin, H. C., Jang, Y., & Chung, K. H. (2004). Functional characterization of recombinant batroxobin, a snake venom thrombin-like enzyme, expressed from Pichia pastoris. FEBS letters, 571(1-3), 67–73. https://doi.org/10.1016/j.febslet.2004.06.060
