Biological Activities of Two Ultra-Low Molecular Weight Heparins, Semuloparin and Bemiparin, Reveal That a Common Potency Standard Is Inadequate

Ultra-low molecular weight heparins (U-LMWHs) are being developed to improve the safety and efficacy of antithrombotic therapy. Two of these agents, bemiparin (Rovi, Madrid, Spain) and semuloparin (Sanofi-Aventis, Paris, France), are produced by distinct methods. Bemiparin is produced by esterification of porcine mucosal heparin followed by alkaline hydrolysis and fractionation. Semuloparin is prepared by a highly selective depolymerization reaction using a phosphazene base which preserves the AT binding sequences from destruction. While both bemiparin and semuloparin are effective at reducing the incidence of post-surgical VTE, some biological differences have been observed in humans. The objective of this investigation was to determine whether a common standard could be used to define their potency.

Activities were compared using typical clinical coagulation assays and pharmacological assays required for potency assessment. Anticoagulant activity was assessed using aPTT and ACT assays. Anti-Xa and anti-IIa activities were determined using amidolytic assays. Constrained and unconstrained curves were determined for the anti-Xa and anti-IIa data to assess parallelism of the concentration-response curves. Thrombin generation was measured using a fluorometric substrate kinetic method (Technothrombin TGA assay; Technoclone GmbH, Vienna, Austria). Platelet function was assessed using platelet aggregometry and the serotonin release assay (SRA).

A significantly larger prolongation of the aPTT was observed with bemiparin at concentrations >1 μ g/ml. Differences in anticoagulant activity measured by the Heptest assay were not observed at concentrations below 2.5 μ g/ml. When supplemented to whole blood samples at higher concentrations, bemiparin and semuloparin are able to prolong the kaolin ACT. Anti-FXa activity for bemiparin and semuloparin was the same over the concentration range tested. Bemiparin produced more anti-FIIa activity at each concentration tested. Depletion of AT led to a complete loss of anti-FXa and anti-FIIa activities for both agents while depletion of heparin cofactor-II had minimal impact on anti-IIa activity. Parallelism assessment is used to identify similarity or difference of products in biological assays. The constrained and unconstrained models using anti-FXa concentration-response data were equivalent (p=0.422), indicating that bemiparin and semuloparin were equivalent (IC₅₀ ratio = 0.97, confidence interval= 0.95–0.98, CV = 0.7%). In contrast, the constrained and unconstrained models using anti-FIIa concentration-response data were not equivalent (p<0.001), indicating that the in vitro anti-FIIa activities of the two agents is not equivalent (IC₅₀ ratio = 1.25, confidence interval = 1.17–1.34, CV = 3.4%). At clinically relevant concentrations, bemiparin more effectively inhibited thrombin generation than semuloparin. Platelet aggregation in response to collagen, ADP and arachidonic acid was not affected by either agent at concentrations up to 10 μ g/ml. At a concentration of 5 μ g/ml, bemiparin was able to completely inhibit thrombin-induced aggregation (0.5 U/ml), whereas no effect was observed with semuloparin. Cross-reactivity of bemiparin and semuloparin with anti-PF4/heparin HIT antibodies was tested using platelet aggregation and the SRA. The level of maximal aggregation was comparable in the presence of bemiparin and semuloparin, however, the rate of aggregation was significantly slower in the presence of semuloparin compared to bemiparin. The typical bell-shaped concentration-response curve was observed with both drugs in the SRA. Although the peak release was comparable for the two drugs (65.7 ± 16.0% for semuloparin vs. 75.5 ± 6.6% for bemiparin), the peak serotonin release was observed at lower concentrations for bemiparin (1 μ g/ml) than for semuloparin (10 μ g/ml) suggesting a higher sensitivity of bemiparin to (stronger interaction with) the anti-PF4/heparin antibodies.

These data demonstrate that the molecular profiling of heparin-derived agents is insufficient to determine drug class. Additional structural characterization and multiple biologic activities relevant to clinical safety and efficacy need to be considered as well. As such, the use of a common reference standard for potency determination of ULMWH is not valid.

Rigal: Sanofi-Aventis: Employment. Bayol:Sanofi-Aventis: Employment. Viskov:Sanofi-Aventis: Employment.