Portal de Periódicos da CAPES

Heparin‐independent activation of platelets by heparin‐induced thrombocytopenia antibodies: a common occurrence

2008; Elsevier BV; Volume: 6; Issue: 1 Linguagem: Inglês

10.1111/j.1538-7836.2007.02815.x

1538-7933

Ines Socher, H. Kroll, S. Jorks, Sentot Santoso, Ulrich J. Sachs,

Platelet Disorders and Treatments

Heparin-induced thrombocytopenia (HIT) is a common prothrombotic complication of anticoagulant therapy with unfractionated heparin (UFH) and rarely with low molecular weight heparin (LMWH) [1]. It is caused by antibodies (abs) elicited by a conformational change in platelet factor 4 (PF4) which exposes neoepitopes resulting from its binding to heparin [2]. HIT antibodies bind to these PF4/heparin complexes and interact with platelet FcγIIa receptors, leading to platelet activation, thrombocytopenia and thrombosis. According to this pathomechanism, heparin must be present to induce ongoing platelet activation. However, it has been known for years that cessation of heparin alone is not sufficient to prevent further progression of HIT [3]. Moreover, patients with a so-called delayed-onset HIT, developing thrombosis or thrombocytopenia days or weeks after heparin cessation, have been observed [4-6]. In vivo studies performed in transgenic mice suggest that platelet activation in the absence of heparin is dependent on the density of PF4 on the platelet surface, where PF4 binding to the platelet is mediated by a chondroitin-like glycosaminoglycan on the cell surface [7, 8]. Accordingly, heparin may be unnecessary for HIT antibody binding. Indeed, Prechel and co-workers presented in vitro evidence that antibodies induced by heparin can elicit platelet activation in the absence of heparin in the serotonin release assay (SRA). These findings may indicate that the immune response of HIT resembles that of an autoimmune reaction. HIT sera are known to elicit platelet activation in the absence of heparin in the heparin-induced platelet activation (HIPA) assay. This spontaneous response in the HIPA assay might resemble a specific reaction of these antibodies with their (auto)antigens rather than an effect of heparin contamination as a result of residual heparin in the patient's sample as has been postulated previously [9, 10]. In the year 2005, 980 specimens were tested by HIPA assay at Giessen University Hospital. The HIPA assay was performed as described previously with minor modifications [11]. Briefly, each sample was studied with suspensions of washed platelets from four different platelet donors, using three different reaction conditions with each platelet suspension: buffer; 0.2 U mL−1 reviparin; and 100 U mL−1 UFH. Reactions were placed in microtiter wells containing spherical stir bars and stirred at approximately 500 rpm. Wells were examined optically at 5-min intervals for loss of turbidity (i.e. platelet aggregate formation). A shift from turbidity to transparency within 45 min indicated a positive result. In total, 103 sera tested positive. Of these, 51 (49.5%) caused activation of platelets without the addition of heparin. We chose 14 cases with adequate clinical data and sufficient material for further investigation. According to a clinical scoring system [12], six patients had an intermediate and eight had a high clinical score (Table 1). All patients had a typical onset of HIT symptoms. In the HIPA assay, all sera induced platelet activation in the presence of heparin (0.1 U mL−1) as well as in the presence of buffer only, but not in the presence of 100 U mL−1 heparin. A commercial ELISA (GTI, Brookfield, WI, USA) was positive in 11 but non-reactive in three patients. Preclearing of specimens of possible residual heparin by heparinase-I treatment (60 min at 37 °C) prevented platelet activation by HIT sera in the presence of buffer (but not 0.1 U mL−1 heparin) in two out of 14 sera (14%). In both samples, heparin was detectable by a chromogenic anti-FXa assay. The other 12 samples remained reactive after heparinase treatment (86%). In additional experiments, IgG fractions were prepared from all HIT sera employing MelonTM Gel Spin Purification kit (Pierce, Munich, Germany). In control experiments, heparin contamination up to 70 U mL−1 could be removed successfully using this technique. When IgG fractions from all HIT sera were subjected to HIPA testing, 12 samples remained reactive in the absence of heparin or in the presence of 0.1 U mL−1 heparin. Two IgG fractions were non-reactive in the absence of heparin; these two fractions were from the same sera that lost their reactivity after heparinase treatment. Results obtained in HIPA and ELISA using HIT sera or isolated IgG, respectively, are summarized in Table 1. Sera treated with heparinase gave similar results in the HIPA assay as IgG fractions (data not shown). Combining our findings obtained with heparinase treatment and IgG isolation, two types of HIT sera with reactivity in the HIPA without addition of heparin become evident: Type I sera are not contaminated with heparin (sera #1 to 12). These sera contain antibodies that have the physiological activity to activate platelets even in the absence of heparin. The HIPA results agree with those reported by Prechel et al. using SRA [9]. Type II sera represent those contaminated with heparin. Their activating capacity is removed by both heparinase treatment and IgG isolation (sera #13, 14). We conclude that the spontaneous response to HIT sera in the HIPA assay is rarely related to heparin contamination of the specimen. In the majority of cases, platelet activation in the 'buffer only' reaction of the HIPA can be attributed to the capability of HIT antibodies to induce platelet activation in the absence of heparin. Most likely, these antibodies react with epitopes exposed on PF4 that has bound to platelet membrane glycosaminoglycans [8, 13]. Notably, two sera tested negative in the PF4 ELISA both before and after preparation of the IgG fraction but were capable of activating platelets in the absence and presence of heparin (#3, 6). This could be because of differences in antibody affinity, specificity or both. The phenomenon has also been reported in a single sample by Prechel et al. [9]; if considered to represent epitope spreading, this would support the idea of HIT as an immune response with auto-immune features. An important question is whether the capability of HIT antibodies to cause platelet activation in the absence of heparin in vitro does have clinical implications. The patients described in this study did not display unusually pronounced thrombocytopenia, and all of our patients had a typical onset of HIT symptoms. HIT 'auto'-antibodies are an attractive explanation for pronounced thrombocytopenia or the delayed-onset type of HIT [14]. However, there is currently not sufficient data to substantiate this hypothesis. Our study supports the finding that HIT antibodies are heterogeneous, some of which are capable of activating platelets in the absence of heparin. These antibodies (HIT 'auto-'antibodies) are responsible for the majority of spontaneous responses to HIT sera in the HIPA assay. A prospective analysis will be necessary to address the question whether the occurrence of these antibodies as well as their persistence over time influences the course of HIT in patients. The authors state that they have no conflict of interest.