Platelet endothelial cell adhesion molecule-1 (PECAM-1) is a target glycoprotein in drug-induced thrombocytopenia

Drug-induced immune thrombocytopenia (DITP) is a serious complication of drug treatment. Previous studies demonstrated that most drug-dependent antibodies (DDAbs) react with the platelet membrane glycoprotein (GP) complexes IIb/IIIa and Ib/IX/V. We analyzed the sera from 5 patients who presented with DITP after intake of carbimazole. Notably, thrombocytopenia induced by carbimazole was relatively mild in comparison to patients with DITP induced by quinidine. The sera reacted with platelets in an immunoassay on addition of the drug. In immunoprecipitation experiments with biotin-labeled platelets and endothelial cells, reactivity with the platelet endothelial cell adhesion molecule-1 (PECAM-1, CD31) could be demonstrated, whereas neither GPIIb/IIIa nor GPIb/IX was precipitated in the presence of the drug. These results could be confirmed by GP-specific immunoassay (MAIPA) using monoclonal antibodies (mabs) against PECAM-1. In addition, the binding of DDAbs could be abolished by preincubation with soluble recombinant PECAM-1. Carbimazole-dependent antibodies showed similar reactivity with platelets carrying the Leu(125) and Val(125) PECAM-1 isoforms, indicating that this polymorphic structure, which is located in the first extracellular domain, is not responsible for the epitope formation. Binding studies with biotin-labeled mutants of PECAM-1 and analysis of sera with mabs against different epitopes on PECAM-1 in MAIPA assay suggested that carbimazole-dependent antibodies prominently bound to the second immunoglobulin homology domain of the molecule. Analysis of 20 sera from patients with quinidine-induced thrombocytopenia by MAIPA assay revealed evidence that DDAbs against PECAM-1 are involved in addition to anti-GPIb/IX and anti-GPIIb/IIIa. We conclude that PECAM-1 is an important target GP in DITP. (Blood. 2000;96:1409-1414)     

Patients with quinine-induced immune thrombocytopenia have both "drug-dependent" and "drug-specific" antibodies

Immune thrombocytopenia induced by quinine and many other drugs is caused by antibodies that bind to platelet membrane glycoproteins (GPs) only when the sensitizing drug is present in soluble form. In this disorder, drug promotes antibody binding to its target without linking covalently to either of the reacting macro-molecules by a mechanism that has not yet been defined. How drug provides the stimulus for production of such antibodies is also unknown. We studied 7 patients who experienced severe thrombocytopenia after ingestion of quinine. As expected, drug-dependent, platelet-reactive antibodies specific for GPIIb/IIIa or GPIb/IX were identified in each case. Unexpectedly, each of 6 patients with GPIIb/IIIa-specific antibodies was found to have a second antibody specific for drug alone that was not platelet reactive. Despite recognizing different targets, the 2 types of antibody were identical in requiring quinine or desmethoxy-quinine (cinchonidine) for reactivity and in failing to react with other structural analogues of quinine. On the basis of these findings and previous observations, a model is proposed to explain drug-dependent binding of antibodies to cellular targets. In addition to having implications for pathogenesis, drug-specific antibodies may provide a surrogate measure of drug sensitivity in patients with drug-induced immune cytopenia.     

Fine specificity of drug-dependent antibodies reactive with a restricted domain of platelet GPIIIA

Drug-induced immune thrombocytopenia is caused by drug-dependent antibodies (DDAbs) that bind tightly to platelet glycoproteins only when drug is present. How drugs mediate this interaction is not yet resolved. Several studies indicate that sites recognized by DDAbs tend to cluster in specific structural domains, suggesting they may recognize a limited number of distinct epitopes. To address this issue, we characterized the binding sites for 16 quinine-dependent antibodies thought on the basis of preliminary studies to be possibly specific for a single epitope on glycoprotein IIIa (GPIIIa). Fourteen of the antibodies reacted with a 29-kDa GPIIIa fragment comprising only the GPIIIa hybrid and plextrin-semaphorin-integrin homology domains. However, studies with mutant GPIIIa and the blocking monoclonal antibody AP3 showed that the 14 DDAbs recognize at least 6 and possibly more distinct, but overlapping, structures involving GPIIIa residues 50 to 66. The findings suggest that even antibodies specific for restricted domains on a target glycoprotein may each have a slightly different fine specificity; ie, "unique" epitopes recognized by DDAbs may be rare or nonexistent. The observations are consistent with a recently proposed model in which drug reacts noncovalently with both target protein and antibody to promote binding of an otherwise nonreactive immunoglobulin.     

Glycoprotein Ib/IX complex is the target in rifampicin-induced immune thrombocytopenia

Thrombocytopenia is a major adverse effect of several drug treatments. Rifampicin has been recognized as a cause of immune thrombocytopenia during intermittent high-dose therapy. We characterized the antibody of a patient who presented with purpura and thrombocytopenia during treatment of tuberculosis with rifampicin. Drug-dependent binding of the antibody to platelets was demonstrated by flow cytometry. In a glycoprotein-specific immunoassay, the binding epitope of the IgG antibody was found in the glycoprotein Ib/IX complex, using four different monoclonal antibodies (mAbs) against various epitopes on the GPIb/IX complex, as well as mAbs against GPIIb/IIIa, GPIa/IIa and GPIV. By immunoprecipitation of biotin-labelled platelets, reactivity of the antibody with GPIb/IX was found only in the presence of the drug. These findings clearly demonstrate that rifampicin induces the formation of drug-dependent antibodies capable of causing thrombocytopenia. The binding site of the rifampicin-dependent antibody, located in the GPIb/IX complex, seems to be a favoured target for antibodies induced by different drugs.     

Molecular characterization of quinine/quinidine drug-dependent antibody platelet interaction using monoclonal antibodies

Two murine monoclonal antibodies, FMC 25 and AN 51, directed against distinct epitopes on the glycoprotein Ib complex, have been used to further define the mechanism of quinine/quinidine drug-dependent antibody interaction with platelets. FMC 25, directed against an epitope on glycoprotein IX, had no effect on platelet aggregation induced by collagen or adenosine diphosphate and little, if any, effect on ristocetin-induced platelet agglutination. FMC 25 and its (Fab)2 fragment, however, were potent inhibitors of drug-dependent antibody- induced platelet aggregation and blocked binding of drug-dependent antibody to platelets as assessed by indirect platelet immunofluorescence. In contrast, AN 51, directed against an epitope on the alpha-subunit of glycoprotein Ib, blocked ristocetin-induced, factor VIII/von Willebrand factor (FVIII/vWF)-dependent platelet agglutination but not drug-dependent antibody-induced platelet aggregation or binding of drug-dependent antibody to platelets. Selective proteolytic removal of the majority of the alpha-subunit of glycoprotein Ib (glycocalicin) from platelets by treatment with calcium- dependent protease did not affect binding of drug-dependent antibody. In addition, a quinidine-dependent antiplatelet antibody immunoprecipitated glycoprotein Ib complex from normal platelets and the membrane-associated proteolytic remnant of the glycoprotein Ib complex from calcium-dependent protease-treated platelets. Preincubation of drug-dependent antibody with purified glycoprotein Ib complex inhibited subsequent binding of antibody to platelets, but the separated components, glycoprotein Ib and glycoprotein IX, were both ineffective, suggesting that the normal interaction between glycoprotein Ib and glycoprotein IX in the intact complex was necessary for drug-dependent antibody recognition. The functional response of platelets to drug-dependent antibody was not mediated by way of platelet Fc receptor, since aggregation of washed platelets by acetone- aggregated IgG was not inhibited by FMC 25 (Fab)2. FVIII/vWF was not required for drug-dependent antibody-induced platelet aggregation. The combined evidence is consistent with quinine/quinidine-dependent antibody-platelet interaction occurring by way of a FVIII/vWF- independent, Fc receptor-independent mechanism that probably involves binding of antibody to glycoprotein IX or the beta-subunit of glycoprotein Ib or both.     

Characterization of the binding domains on platelet glycoproteins Ib-IX and IIb/IIIa complexes for the quinine/quinidine-dependent antibodies.

Sera of 12 patients with quinine/quinidine-induced thrombocytopenia showed drug-dependent antibody binding to glycoprotein (GP) Ib-IX complex. The reaction with GPIb-IX complex of 11 of these 12 sera was strongly inhibited by the complex-specific monoclonal antibodies (MoAbs) AK1 and SZ1. The exception was a quinine-induced serum designated BU. The reaction of the six quinidine-induced sera was also partially blocked by an anti-GPIX MoAb, FMC25. Only 3 of the 12 patient sera showed drug-dependent antibody binding to GPIIb/IIIa, which was strongly inhibited by the anti-GPIIIa MoAb 22C4, and the anti-GPIIb alpha MoAb SZ22. With detergent-solubilized Serratia metalloprotease-treated platelets, quinine/quinidine-induced sera, except BU, immunoprecipitated a membrane-bound proteolytic fragment of GPIb-IX complex. In contrast, BU immunoprecipitated glycocalicin and a 40-Kd peptide tail fragment of GPIb alpha from the cell supernatant. Using purified GPIb-IX complex or its components as the target antigen, all the quinine-induced sera, except BU, immunoprecipitated GPIb-IX complex but failed to immunoprecipitate GPIb, GPIX, or the complex reformed from GPIb and GPIX. The quinidine-induced sera strongly immunoprecipitated purified GPIb-IX complex, weakly immunoprecipitated purified GPIX and the recombined complex, but did not immunoprecipitate purified GPIb. The combined data suggest that one quinine-dependent antibody (BU) recognizes an epitope in the peptide tail region of GPIb alpha and the other five quinine-dependent antibodies react with a complex-specific epitope on the membrane-associated region of GPIb-IX complex, whereas each of the six quinidine-induced sera contain two drug-dependent antibodies, one reactive with the GPIb-IX complex-specific epitope and the other reactive with GPIX. The binding domain(s) on GPIIb/IIIa for the quinine/quinidine-dependent antibodies appear to be sterically close to the epitopes for 22C4 and SZ22.     

Glycoprotein IIb/IIIa complex is the target in mirtazapine-induced immune thrombocytopenia

Mirtazapine (MW 265.36), a tetracyclic antidepressant of the piperazine-azapine group which augments central noradrenergic and serotonergic activity, is currently used as an oral antidepressant. We report a case of severe thrombocytopenia in a 66-year-old patient occurring after mirtazapine administration, suggesting an immune mechanism. This report documents the first case of mirtazapine-induced immune thrombocytopenia. The patient's serum was screened for drug-induced anti-platelet antibody with the chromium(51) (Cr(51)) platelet lysis technique. The drug-dependent antibody was characterized using flow cytometry, the monoclonal antibody immobilization of platelet antigens assay (MAIPA assay), and immunoprecipitation. By the Cr(51) platelet lysis technique, we obtained an equivocal result for the detection of mirtazapine-induced antibody. However, the patient's serum tested positive for mirtazapine-induced antibody by flow cytometry. The results showed that the binding ratio of 5.7 (mean fluorescence intensity) in the presence of the patient's serum and mirtazapine in a final concentration of 1.0 mmol/L was strongly positive. The antibody was found to bind the glycoprotein (GP) IIb/IIIa complex by MAIPA assay by using five different monoclonal antibodies against GP complexes Ib/IX, GPIIb/IIIa, or GPIa/IIa. Immunoprecipitation studies showed that the GPIIb/IIIa complex was precipitated by antibody in the presence, but not in the absence, of mirtazapine. These findings provide evidence that immune thrombocytopenia can be caused by sensitivity to the antidepressant mirtazapine. This is the first well-documented case of mirtazapine-induced immune thrombocytopenia.     

A site involving the "hybrid" and PSI homology domains of GPIIIa (beta 3-integrin subunit) is a common target for antibodies associated with quinine-induced immune thrombocytopenia

Drug-dependent antibodies (DDAbs) can cause the precipitous destruction of platelets if a patient is exposed to the drug for which the antibodies are specific. The molecular character of the epitopes recognized is poorly understood, and the mechanism by which drugs promote tight binding of these antibodies to platelet glycoproteins without linking covalently to protein or antibody is not yet known. We studied a group of quinine-dependent antibodies that react with human glycoprotein IIIa (GPIIIa; beta3-integrin subunit) but fail to recognize rat GPIIIa, despite close homology between the 2 proteins. By characterizing reactions of these antibodies with human/rat GPIIIa chimeras and selected GPIIIa mutants, we found that each of 3 quinine-dependent antibodies requires a 17-amino acid sequence in the newly recognized "hybrid" and PSI homology domains of GPIIIa for drug-dependent binding. Disulfide bonds are required to stabilize the target epitope. Monoclonal antibody AP3, which blocks the binding of these DDAbs to GPIIIa, was found to require a more limited stretch of the same peptide for its reaction with the glycoprotein. The findings suggest this region of GPIIIa may be a favored target for quinine-dependent antibodies and may provide a basis for further studies to elucidate the molecular basis of glycoprotein-drug-antibody interaction.     

Two distinct subgroups of tirofiban-induced thrombocytopenia exist due to drug dependent antibodies that cause platelet activation and increased ischaemic events

Tirofiban-associated thrombocytopenia is due to drug-dependent antibodies (DDAbs) directed against the GPIIb/IIIa complex which can bind after drug-induced conformational changes to the receptor complex. In such cases a higher incidence of myocardial infarction and mortality has been reported raising the possibility of platelet activation. We followed consecutive cases treated with tirofiban to determine the incidence of thrombocytopenia and confirmed that this was due to tirofiban-dependent antibodies. We then tested if these antibodies could cause platelet activation in vitro and correlated this with clinical outcome. In 871 treated patients, severe thrombocytopenia was observed in 11 cases, an incidence of 1.26%. Tirofiban dependent antibodies were confirmed in all cases using a flow cytometric assay. There were two distinct presentations of thrombocytopenia, one occurring acutely, and the second a delayed thrombocytopenia occurring after several days of tirofiban exposure and in keeping with a primary immune response. The effects of DDAbs on platelet activation was analysed by measuring P-selectin (CD62p) and annexin V, in the presence or absence of tirofiban, by flow cytometry. In addition, platelet activation was sought using the serotonin release assay. In six cases there was evidence of platelet activation and this was significantly associated with further coronary ischaemic events experienced at the time of acute thrombocytopenia. Tirofiban-induced thrombocytopenia due to DDAbs is a common occurrence and can lead to platelet activation and increased thrombotic events.     

Drug-dependent and non-drug-dependent antiplatelet antibody in drug-induced immunologic thrombocytopenic purpura

The mechanism of drug-dependent immunologic thrombocytopenic purpura (DITP) was investigated by studying the sera of four patients with classic DITP (two with quinidine-, one with acetaminophen-, and one with phenazopyridine-dependent antiplatelet antibody) using a solid-phase radioimmunoassay with 125I-staphylococcal protein A. Two forms of antiplatelet antibody could be demonstrated: one that required drug to bind to platelets and one that bound to platelets in the absence of drug. Drug-dependent antiplatelet antibody required the simultaneous addition of drug and the Fc domain of the drug-dependent IgG molecule for binding to platelets. It did not require serum complement or factor VIII-related antigen for binding to platelets. Drug-dependent binding of antibody to platelets was saturation-dependent. Non-drug-dependent antiplatelet antibody of two patients (one with quinidine-induced thrombocytopenia and the other with acetaminophen-induced thrombocytopenia) reacted with autologous platelets as well as with homologous platelets, indicating that they were autoantibodies. Both autoantibodies had disappeared when their sera were tested 23 and 138 days, respectively, after withdrawal of their initial positive sera. Non-drug-dependent antiplatelet antibody binding could be demonstrated with the F(ab')2 fragment of the purified IgG of the serum of the second patient with quinidine DITP, who did not have detectable alloantibodies against HLA. None of the four patients with non-drug-dependent antiplatelet antibody had a past or present history of autoimmune thrombocytopenic purpura.     

Characteristics of quinine- and quinidine-induced antibodies specific for platelet glycoproteins IIb and IIIa.

Recent studies have shown that antibodies characteristic of quinine- and quinidine-induced thrombocytopenia sometimes recognize the platelet membrane glycoprotein (GP) complex IIb/IIIa in addition to their well known target, GPIb/IX. We have investigated the frequency with which drug-induced antibodies bind to GPIIb/IIIa and the nature of their target epitopes. In studies of sera from 13 patients sensitive to quinidine or quinine, we found that 10 contained IgG antibodies specific for both GPIb/IX and GPIIb/IIIa, two reacted with GPIb/IX alone, and one reacted with GPIIb/IIIa alone. In all cases, the presence of drug was required for binding of IgG to target GPs. By immunoabsorption, we found that each of five polyspecific sera contained at least two different antibodies, one reactive with GPb/IX and the other with GPIIb/IIIa. Further studies with eight drug-dependent antibodies (DDAb) specific for GPIIb/IIIa showed that three recognized the GPIIb/IIIa complex only, one recognized GPIIb alone, and three recognized GPIIIa alone. The eighth serum appeared to bind to both GPIIIa alone and to an epitope determined by the GPIIb/IIIa complex. The three antibodies specific for GPIIIa alone also reacted with GPIIIa deglycosylated with endo-H, and with the major (61 Kd) fragment obtained by chymotryptic digestion of GPIIIa but failed to react with reduced GPIIIa. These findings demonstrate that, in drug-induced, immunologic thrombocytopenia, the anti-platelet immune response is typically directed against epitopes on both GPIb/IX and GPIIb/IIIa. The three DDAb we studied that were specific for GPIIIa alone recognize epitopes resistant to chymotrypsin and endo-H treatment that are dependent on intrachain disulfide bonding.     

Quinine- and quinidine platelet antibodies can react with GPIIb/IIIa

Quinine- and quinidine-dependent antiplatelet antibodies are generally believed to bind to the membrane glycoprotein complex, GPIb/IX. However, we and others have found that some drug-dependent antibodies bind to platelets from patients with the Bernard-Soulier syndrome which lack these glycoproteins. We therefore studied the reactions of a group of these antibodies with normal and Bernard-Soulier platelets and their membrane proteins. As indicated by rosette formation of the sensitized platelets around protein A-Sepharose beads, two quinine- and two quinidine-dependent antibodies reacted with both normal and Bernard-Soulier syndrome platelets at a high (300 microM) concentration of drug. At a pharmacologic drug concentration (10 microM), all four antibodies reacted with normal platelets but only the two quinine-induced antibodies reacted with Bernard-Soulier platelets. Immunoprecipitation studies with solubilized, tritium-labelled normal platelets, at both high and low drug concentrations, showed that each of the four antibodies precipitated proteins corresponding to GPIb and GPIX. Fainter bands corresponding to glycoproteins IIb and IIIa, which do not label well with tritium, were also detected. With radioiodinated normal platelets, it was found that each of the four antibodies was capable of precipitating GPIIb/IIIa, but only in the presence of drug. The four antibodies also promoted drug-dependent precipitation of GPIIb and GPIIIa from lysates of radioiodinated Bernard-Soulier platelets. The two quinine-dependent antibodies precipitated these glycoproteins at both high and low drug concentrations, while the quinidine-dependent antibodies reacted much more strongly at the higher drug concentration. Precipitation of GPIb/IX was not observed with BSS platelets. Absorption of a quinine-induced antibody with Bernard-Soulier platelets in the presence of drug eliminated its ability to precipitate GPIIb and GPIIIa. However, the absorbed antibody retained the ability to precipitate GPIb from solubilized normal platelets. Thus, at least two drug-dependent antibodies were present, one specific for GPIb/IX and the other for GPIIb/IIIa. These findings indicate that glycoproteins IIb and/or IIIa, in addition to the GPIb/IX complex, can serve as targets for drug-dependent antibodies in both intact and detergent-solubilized platelet preparations.     

Rifampicin-dependent antibodies bind a similar or identical epitope to glycoprotein IX-specific quinine-dependent antibodies

The drug-dependent antibody of a patient with rifampicin-induced thrombocytopenia was characterized using the antigen-capture enzyme-linked immunosorbent assay (MAIPA assay), flow cytometry, and immunoprecipitation. The antibody was found to bind glycoprotein (GP) Ib-IX but not GPIIb-IIIa because (1) it immunoprecipitated drug-dependently the former but not the latter glycoprotein complex and (2) the MAIPA assay showed strong rifampicin-dependent antibody binding when anti-GPIb-IX monoclonal antibodies (mAbs) (AK2 and FMC25) but not anti-GPIIb-IIIa mAbs (AP2, SZ21, and SZ22) were used to capture the antigen. The antibody binding site was further localized to the GPIX subunit of the GPIb-IX complex because flow cytometric analysis revealed drug-dependent antibody binding to L cells transfected with human GPIbbeta and GPIX complementary DNA (L betaIX cells) but not with human GPIbalpha and GPIbbeta complementary DNA (L alphabeta cells). Finally, in the MAIPA assay, the rifampicin-dependent antibody almost completely cross-blocked the binding of the anti-GPIX mAb (SZ1) to platelets. Similar cross-blocking of SZ1binding to platelets by the quinine-dependent antibodies was also observed. This finding not only confirms that the epitope of the rifampicin-dependent antibody is on GPIX but it is also identical to or located in close proximity to that of the quinine-dependent antibody and SZ1. Further characterization of the epitopes of these antibodies may have important implications for a general understanding of the mechanism of drug-induced thrombocytopenia. (Blood. 2000;95:1988-1992)     

Drug-induced Immune Thrombocytopenia

SUMMARY. Immune thrombocytopenia is a relatively common problem associated with the clinical usage of drugs. Drugs frequently implicated include quinine, quinidine, heparin, penicillins, cephalosporins, co-trimoxazole, gold and D-penicillamine. Bleeding including bruising and purpura is the usual clinical manifestation except in immune heparin-induced thrombocytopenia in which thrombosis occurs more frequently than bleeding. Cessation of the offending drug is the important step in the treatment but other measures may also be required such as platelet transfusion and steroid therapy for patients with clinical bleeding or antithrombotic therapy with warfarin and dextran or low molecular weight heparin/heparinoid for patients with heparin-induced thrombocytopenia and thrombosis. Idiosyncratic drug-induced thrombocytopenia is mediated by an antibody which binds to platelets only in the presence of the drug resulting in the clearance of sensitised platelets by the reticuloendothelial system. In quinine/quinidine-induced thrombocytopenia, the antibodies recognise drug-dependent epitopes on platelet membrane glycoproteins Ib-IX and/or glycoproteins IIb-IIIa. In immune heparin-induced thrombocytopenia the current data suggest a mechanism which probably involves the binding of heparin-antibody complexes to the platelet Fc receptors but the precise mechanism is yet to be fully characterised. The associated thrombosis in this condition is likely to be due to platelet activation and possibly endothelial cell damage induced by the heparin-related antibody.     

Platelet activation by CD9 monoclonal antibodies is mediated by the FCγII receptor

The function of the human cell surface CD9 antigen is not known, yet monoclonal antibodies (mAbs) of the IgG1 subclass in the CD9 cluster induce activation of platelets. Previously it had been shown that this activation pathway is comparable both in kinetics and extent to physiological agonists such as thrombin. Here it is demonstrated that activation with CD9 mAbs depends on interaction of the Fc part of the CD9 antibody molecule with Fc receptors on the platelet surface, since: (i) mAb directed against the Fc receptor totally blocked the platelet response to CD9 mAb; and (ii) F(ab')2 fragments of the CD9 mAb SYB-1 which bound to platelets, as demonstrated by flow cytometry, failed to activate them. Furthermore, platelet activation by CD9 mAb closely paralleled the activation caused by cross-linking Fc receptors when comparing: (i) kinetics and extent of aggregation; (ii) thromboxane synthesis; (iii) calcium flux; and (iv) the cytoplasmic alkalinization response. Thus it is concluded that CD9 antigen itself does not necessarily participate in stimulus-response coupling leading to platelet activation by CD9 mAbs, and that this activation can be entirely accounted for by the Fc receptor pathway mechanism. The results suggest a possible novel mechanism for platelet consumption in cases of immune thrombocytopenia.     

Acute thrombocytopenia after treatment with tirofiban or eptifibatide is associated with antibodies specific for ligand-occupied GPIIb/IIIa

Acute thrombocytopenia is a recognized complication of treatment with GPIIb/IIIa inhibitors whose cause is not yet known. We studied 9 patients who developed severe thrombocytopenia (platelets less than 25 x 10(9)/L) within several hours of treatment with the GPIIb/IIIa inhibitors tirofiban (4 patients) and eptifibatide (5 patients). In each patient, acute-phase serum contained a high titer (range, 1:80-1:20 000) IgG antibody that reacted with the glycoprotein IIb/IIIa complex only in the presence of the drug used in treatment. Four patients had been previously treated with the same drug, but 5 had no known prior exposure. Pretreatment serum samples from 2 of the latter patients contained drug-dependent antibodies similar to those identified after treatment. No tirofiban- or eptifibatide-dependent antibodies were found in any of 100 randomly selected healthy blood donors, and only 2 of 23 patients receiving tirofiban or eptifibatide who did not experience significant thrombocytopenia had extremely weak (titer, 1:2) tirofiban-dependent antibodies. In preliminary studies, evidence was obtained that the 9 antibodies recognize multiple target epitopes on GPIIb/IIIa complexed with the inhibitor to which the patient was sensitive, indicating that they cannot all be specific for the drug-binding site. The findings indicate that acute thrombocytopenia after the administration of tirofiban or eptifibatide can be caused by drug-dependent antibodies that are "naturally occurring" or are induced by prior exposure to drug. These antibodies may be human analogs of mouse monoclonal antibodies that recognize ligand-induced binding sites (LIBS) induced in the GPIIb/IIIa heterodimer when it reacts with a ligand-mimetic drug.     

Drug-induced Immune Thrombocytopenia

SUMMARY.

Immune thrombocytopenia is a relatively common problem associated with the clinical usage of drugs. Drugs frequently implicated include quinine, quinidine, heparin, penicillins, cephalosporins, co-trimoxazole, gold and D-penicillamine. Bleeding including bruising and purpura is the usual clinical manifestation except in immune heparin-induced thrombocytopenia in which thrombosis occurs more frequently than bleeding. Cessation of the offending drug is the important step in the treatment but other measures may also be required such as platelet transfusion and steroid therapy for patients with clinical bleeding or antithrombotic therapy with warfarin and dextran or low molecular weight heparin/heparinoid for patients with heparin-induced thrombocytopenia and thrombosis. Idiosyncratic drug-induced thrombocytopenia is mediated by an antibody which binds to platelets only in the presence of the drug resulting in the clearance of sensitised platelets by the reticuloendothelial system. In quinine/quinidine-induced thrombocytopenia, the antibodies recognise drug-dependent epitopes on platelet membrane glycoproteins Ib-IX and/or glycoproteins IIb-IIIa. In immune heparin-induced thrombocytopenia the current data suggest a mechanism which probably involves the binding of heparin-antibody complexes to the platelet Fc receptors but the precise mechanism is yet to be fully characterised. The associated thrombosis in this condition is likely to be due to platelet activation and possibly endothelial cell damage induced by the heparin-related antibody.     

Drug-mediated thrombocytopenia

Thrombocytopenia due to drug-dependent antibodies most frequently occurs with quinine/quinidine and with heparin. Considerable evidence has accumulated about the mechanism of action of quinine/quinidine-induced antibodies but less is known about the effect of heparin. Although there is controversy, it is likely that the action of quinine/quinidine-induced antibodies follows a loose association between drug and platelet with antibodies acting independently of the Fc receptor. There is strong evidence that the complex of glycoprotein Ib and glycoprotein IX, absent in the Bernard-Soulier syndrome, provides the binding site for quinine/quinidine-dependent antibodies. It also appears that the two glycoproteins must be present in complex form for antibody binding to occur. There is some heterogeneity of quinine/quinidine-dependent antibodies since there are reports of a proportion of patient antibodies reacting with other membrane determinants or acting independently of the drug. Heparin-induced thrombocytopenia may be the consequence of a direct effect, or a more serious condition associated with thrombosis may occur when heparin-dependent antibodies are formed. The mode of action of these antibodies and the nature of their antigenic determinants remain unclear. Recognition of heparin-associated thrombocytopenia is important so that serious bleeding or thrombotic sequelae can be forestalled.     

Analysis of GPIIb/IIIa receptor number by quantification of 7E3 binding to human platelets

A large number of glycoprotein (GP) IIb/IIIa receptors are present on the surface of platelets. Studies to define precisely the number of GPIIb/IIIa receptors using specific monoclonal antibodies (MoAbs) or fibrinogen binding have, however, yielded varying estimates of receptor number. To refine the quantitative estimation of GPIIb/IIIa receptors on resting platelets, we have used the MoAb 7E3, which has high affinity for GPIIb/IIIa. Quantitative binding studies were performed using radiolabeled conjugates of 7E3 IgG, as well as fragments and derivatives of 7E3. For platelets obtained from any single individual, the numbers of 7E3 F(ab')2 and IgG molecules bound per platelet were equivalent (approximately 40,000), whereas the number of Fab molecules bound per platelet was consistently approximately twofold higher (approximately 80,000). To investigate the basis of the quantitative disparity in binding of intact 7E3 and 7E3 F(ab')2 versus 7E3 Fab, we studied the binding of a newly constructed, bispecific (Fab')2 molecule containing only a single 7E3 combining site. Because this construct bound to the same extent as the Fab species, the larger size of the intact 7E3 and 7E3 F(ab')2 molecules could not explain the reduced number of molecules that bound per platelet compared to the Fab fragment. Rather, it appears that the valency of the antibody is the critical factor determining the number of antibody molecules bound per platelet. Thus, we conclude that the binding of 7E3 Fab corresponds most closely with surface GPIIb/IIIa number and that the number of GPIIb/IIIa receptors is approximately 80,000 per platelet.     

Identification of platelet glycoprotein IIb/IIIa as the major binding site for released platelet-von Willebrand factor [published erratum appears in Blood 1987 Jun;69(6):1789]

We studied the effects(s) of two monoclonal antibodies, 6D1 and 10E5 (directed against platelet glycoprotein Ib [GPIb] and the GPIIb/IIIa complex, respectively), and purified human plasma fibrinogen on the binding of released platelet-von Willebrand factor (vWf) to the platelet surface. Neither of the monoclonal antibodies nor fibrinogen had any effect on the amount of platelet-vWf expressed on unstimulated platelets or on the amount expressed on platelets stimulated in the absence of extracellular Ca++. However, the antibody directed against GPIIb/IIIa inhibited 72% of the thrombin-induced increase in the platelet-vWf bound to the platelet surface when platelets were stimulated in the presence of 5 mmol/L Ca++. The antibody against GPIb did not inhibit the surface expression of platelet-vWf on stimulated platelets in the presence of Ca++. Purified normal human fibrinogen inhibited the surface binding of platelet-vWf to thrombin-stimulated platelets to a degree similar to that observed with the monoclonal antibody directed against the GPIIb/IIIa complex. These data indicate that platelet-vWf released from platelets binds primarily to the GPIIb/IIIa complex at or near the plasma fibrinogen binding site.