Can the haemorrhagic component of heparin be identified? Or an attempt at clean thinking on a dirty drug.

Heparin consists of different classes of molecules. We distinguish below-critical-chain length heparin (BCLM, MW < 5,400), with only anti-factor Xa activity and above-critical-chain length material (ACLM, MW > 5,400) with both antithrombin and anti-factor Xa activity. In this article we introduce a division within the ACLM fraction, between extra large material (MW > 8,000) and ACLM-low (MW 5,400-8,000). Extra large material is abundantly present in unfractionated heparin but is rare in low-molecular-weight (LMW) heparins. We noted that injection of an LMW heparin causes 5- to 10-fold higher plasma levels of ACLM than injection of a clinically equivalent dose of unfractionated heparin (UFH) and proportionally higher inhibitions of the clotting system. So with LMW heparin one can afford higher levels of anticoagulation than with UFH at a lower risk of bleeding. We surmise that this is caused by the virtual absence of the (haemorrhagic) extra-large-molecular-weight fraction from LMW heparins. A laboratory artefact, i.e. the absence of Ca2+ in the anti-factor Xa tests, makes that heparin mixtures that lack extra large heparin molecules show a (spuriously) high ratio of anti-factor Xa activity over anti-thrombin activity. So the correlation between a high aXa/alla ratio and a favourable ratio of antithrombotic effect over bleeding is not necessarily caused by the presence of BCLM. In fact BCLM is a poor anticoagulant; in mixtures of ACLM and BCLM, ACLM causes by far the larger part of the anticoagulant effect. We surmise that the LMW fraction of ACLM is the active anticoagulant component in any heparin preparation and, isolated, would make a proper third-generation heparin.