HIT and run: heparin's unusual immune response

In this issue of Blood, Pötschke et al investigate the cellular underpinnings of the heparin-induced thrombocytopenia (HIT) immune response.¹ 

The immune response to heparin is peculiar. Unfractionated heparin (UFH) and its derivatives, the low-molecular weight heparins (LMWHs), are essential for initiating HIT, one of the most common and serious drug-dependent responses in hospitalized patients. In certain clinical settings, such as cardiac surgery, seroconversion is commonplace (∼ 50%),²  but clinical manifestations of disease are, fortunately, less frequent (1%-2%).³  Although the antibodies causing HIT recognize heparin complexed to an endogenous platelet protein, platelet factor 4 (PF4), heparin is not necessary to carry out the dire business of the antibodies⁴  as cell-surface glycosaminoglycans (GAGs; heparan sulfate or chondroitin sulfate) can readily substitute for the drug in the PF4/heparin antigenic complex.

Another peculiar aspect of the HIT immune response is its serologic transience. Given the abundance of antigenic material in the host in the form of PF4 and cell-surface glycosaminoglycans, one would naturally expect the PF4/heparin immune response, once it occurs, to be self-sustaining. Yet, thankfully, it is not.⁵  The immune transience of HIT is all the more striking, as recent studies indicate that bacterial infection and trauma may serve as “danger” signals to predispose individuals to become sensitized.^6,7  These unusual serologic features of HIT remain largely unexplained.

To understand why HIT is so atypical, Pötschke and colleagues undertook studies to place HIT in the context of known immune disease paradigms.¹  They pose 2 specific questions in this study: Is the PF4/heparin antibody response due to a dysregulated autoimmune response to self-antigens, PF4 and GAGs? Or is HIT caused by a heightened immune recall, due to polyclonal activation of memory B cells? To address these questions, the authors studied 166 patients undergoing cardiac surgery and compared the kinetics of the PF4/heparin immune response at various time points (baseline, days 3 and 10, and > 120 days after surgery) to antinuclear antibodies (ANAs) and recall responses to vaccine antigens, tetanus toxoid (TT), and diphtheria toxin (DT). Not surprisingly, the authors found that cardiac surgery is associated with a strong inflammatory response, as indicated by a mean C-reactive protein (CRP) level of approximately 165 mg/dL measured 3 days after surgery (reference value < 1 mg/dL). This inflammatory stimulus was associated with polyclonal activation of B cells, as indicated by an increase in mean total IgG (> 30% from baseline) as well as a gradual increase in IgG reactivity to TT and DT antigens. However, cardiac surgery did not breach tolerance; autoreactivity, as measured by ANA testing, was not affected by surgery. Consistent with previous studies, PF4/heparin seroconversions were maximal by day 10 and declined to baseline values by 120 days, thus showing that PF4/heparin antibodies conform to neither pattern of antibody responses. Surprisingly, there was no correlation of PF4/heparin seropositivity with increased CRP or higher IgG levels.

While the findings of Pötschke et al do not fully eliminate the role of inflammation and autoreactivity in HIT, they do provide compelling correlative in vivo data suggesting that the inflammatory milieu of cardiac bypass and autoreactive B cells have a minimal role in the etiology of the PF4/heparin immune response after cardiac surgery. If inflammation does not contribute to the HIT immune response, what can account for the extraordinary rates of seroconversion seen in cardiac surgery? One potential mechanism may be related to antigen load and/or antigenic features. It is recognized that cardiac surgery is associated with significant antigenic load, including a 10- to 25-fold increase in circulating levels of platelet granular proteins⁸  as well as increased levels of circulating heparin (300-400 IU/kg). In addition, there is emerging evidence that the unique structural features of the PF4/heparin antigenic complex are critical for immune induction. Studies from our murine immunization model have shown that biophysical aspects of the antigen, such as charge and size, contribute to the in vivo immunogenicity of PF4/heparin complexes.⁹  Parallel investigations of the immune response to protamine/heparin complexes in mice and humans suggest that the host immune response to PF4/heparin complexes may not be so unique after all, as antibodies to protamine/heparin complexes share a number of serologic features typically associated with HIT antibodies (platelet activation, heparin dependence, and serologic transience).¹⁰  These studies clearly point to conserved pathways of cellular activation by antigenic complexes that structurally resemble PF4/heparin. To what extent these cellular pathways involve other distinct B-cell compartments, cytokine regulatory networks, T cells, and/or all of the above, remains to be investigated in the wake of findings by Pötschke and colleagues.