Platelet activation by heparin

In this issue of Blood, Gao et al provide important insights into the proactivating effects of negatively charged anticoagulants on platelets.¹  They show outside-in signaling via the glycoprotein (GP) IIb-IIIa complex (integrin αIIbβ3) induced by heparin, low molecular weight heparin, and the therapeutic pentasaccharide fondaparinux.

The proactivating effects of heparin on platelets have been known for decades. Heparin can cause a moderate decrease in platelet count, which typically is seen at start of treatment with heparin in therapeutic doses. Concomitantly, markers of platelet activation are increased. Unfractionated heparin shows this effect to a greater extent than does low molecular weight heparin.²  In the 1970s and 1980s, this direct effect of heparin on platelets caused considerable confusion in the delineation of nonimmune (type I) and immune (type II) heparin-induced thrombocytopenia (HIT). Since the availability of sensitive assays for the detection of antibodies that cause immune HIT, differentiating both types has become much easier.³  Since then, however, immune HIT became the primary focus of research, and nonimmune interactions between heparin and platelets have been rather neglected.

In 1989, Chong and Ismael⁴  provided indirect evidence that the ability of heparin to potentiate platelet aggregation requires GPIIb-IIIa by showing that antibodies to this complex block the platelet-aggregating effect of heparin and that human platelets lacking this receptor could not be aggregated by heparin.

In the present study Gao and colleagues characterize several important steps in the mechanism of how negatively charged anticoagulants activate platelets.

Addition of heparin, low molecular weight heparin, or fondaparinux to platelets did not induce aggregation or granule secretion by itself, but potentiated the effect of low-dose adenosine diphosphate (ADP). That unfractionated heparin and fondaparinux had similar potentiating effects is an unexpected observation. While both molecules are strongly negatively charged, the small fondaparinux molecule is thought to bind poorly to proteins other than antithrombin, whereas the large unfractionated heparin molecules bind to many different proteins. Furthermore, in previous studies, the effects of low molecular weight heparins on platelets were much less pronounced compared with unfractionated heparin.⁵ 

Using a series of elegant experiments, Gao et al show direct binding of heparin to GPIIb-IIIa, triggering a signaling cascade that leads to activation of phosphatidyl-inositol-3 kinase. Interestingly, and potentially clinically relevant, platelets also bind to immobilized heparin via GPIIb-IIIa. This not only lowers the threshold for activation by other agonists like ADP, but induced platelet spreading involving a slightly different pathway involving phosphorylation of focal adhesion kinase (FAK).

This study has potential clinical implications far beyond the interesting aspects of platelet physiology. Negatively charged polysaccharide-based anticoagulants have been the mainstay for inhibiting the coagulation cascade in most acutely ill hospitalized patients. Thus, we may be inadvertently introducing an additional platelet-activating factor in nearly all treatment protocols. In this regard, it is of real interest that bivalirudin, a direct thrombin inhibitor used as an anticoagulant during percutaneous coronary interventions, showed in large clinical trials as having at least the same efficacy as combined treatment with heparin plus GPIIb-IIIa inhibition. In addition, many intravascular devices are coated with heparin based on the concept that this reduces activation of the clotting cascade. However, the present study challenges this theoretical concept. Surfaces exposing immobilized heparin likely induce mild activation of platelets. Activated platelets in turn expose phospholipids and thereby the catalytic surface for activating the clotting cascade. This needs to be explored by well-designed in vitro studies.

Currently, several nonheparin anticoagulants are entering clinical practice. Because of their completely different molecular structures, these drugs are very unlikely to interact with platelets in the same way as heparin, low molecular weight heparins, or fondaparinux do. It should be considered that, in patients treated with these novel drugs, platelets may have a comparatively higher threshold for activation. This may increase the bleeding risk if these new anticoagulants are combined with antiplatelets drugs.

Finally, the study of Gao et al also bridges into oncology. It is still an unresolved issue whether (and if so, how) low molecular weight heparins reduce disease progression and mortality in cancer patients. Platelets play a role in seeding of metastases, and heparin might interfere with this process. An alternative hypothesis could be that heparin also interacts with leukocytes or tumor cells, thereby inducing activating signaling that could modify the natural course of the disease.

We are in the middle of a paradigm shift of anticoagulant treatment. The current study is a perfect illustration about pleiotropic effects of heparin and heparin-like drugs beyond their anticoagulant capacity. Replacing heparin might be much more than just changing the anticoagulant.