Platelet-reactive Antibodies Research Articles

Detection and differentiation of platelet‐specific antibodies by flow cytometry: the bead‐mediated platelet assay

Platelet-reactive antibodies cause a number of clinical disorders. The detection and differentiation of these antibodies are prerequisites for the adequate treatment of these disorders. The bead-mediated platelet assay described here enables the detection and differentiation of platelet-bound antibodies by the use of flow cytometry. The bead-mediated platelet assay is based on the isolation of human platelet glycoproteins by using flow cytometric standardization beads after the incubation of typed platelets with human sera. The specificity and sensitivity of this assay were tested with five sera, each containing a known platelet-reactive antibody. The monoclonal antibody-specific immobilization of platelet antigens assay was used as a reference test. The bead-mediated platelet assay was able to determine the glycoprotein specificity of the antibody without cross-reactions in every case. In serial dilution tests, the bead-mediated platelet assay was able to detect the antibodies at higher dilutions than the monoclonal antibody-specific immobilization of platelet antigen assay. Total test time was 3.5 hours. The bead-mediated platelet assay is a fast and reliable method for the detection and differentiation of platelet-reactive antibodies.

Specificities of anti-platelet antibodies in multitransfused patients with haemato-oncological disorders.

The clinical condition and the formation of platelet-reactive antibodies influence the post-transfusion platelet increment. We analysed the specificities of platelet-reactive antibodies in 81 multitransfused patients with haemato-oncological diseases refractory to platelet transfusions, or prior to a scheduled stem cell transplantation. In 17 additional patients we prospectively determined the development of platelet-reactive antibodies at the time of chemotherapy in weekly intervals. Sera were tested by the monoclonal antibody-specific immobilization of platelet antigens (MAIPA)-technique for antiplatelet antibodies against HLA class I antigens, the human platelet-specific alloantigens (HPA)-1, -2, -3, -5, and the platelet membrane glycoproteins (GP) Ia/IIa, Ib/IX, IIb/IIIa (panreactive). Platelet reactive antibodies were found in 54% of blood samples. They were frequent in patients with a history of possible previous immunization, particularly if patients were refractory to platelet transfusions. Platelet-reactive HLA antibodies were the most common antibodies. Even patients without a known risk of primary immunization who received exclusively leucocyte-depleted blood products formed antibodies. By the MAIPA technique, even LCT-negative sera were found to contain platelet-reactive antibodies.

Biotin-avidin immobilization of platelet glycoproteins (BAIPG): a new capture assay for the detection of anti-platelet antibodies

Several ‘capture’ assays are currently employed to identify specific platelet antibodies, but all require the use of murine monoclonal antibodies (MoAbs) against the antigen of interest. We have developed a new antigen capture assay for the detection of platelet reactive antibodies, based on platelet surface sialoglycoprotein labelling with biotin hydrazide, and a following immobilization of the biotinylated platelet proteins to microtiter wells that had been coated with streptavidin. The resulting solid phase can then be used in a simple ELISA to detect serum and platelet associated antibodies. We describe here two versions of this biotin-avidin immobilization of platelet glycoproteins (BAIPG) assay. In BAIPG assay type I, the test sera are directly incubated in microtiter wells previously coated with streptavidin plus biotinylated platelet proteins. The BAIPG type II procedure involves the incubation of sera with biotinylated platelets before platelet solubilization, and, after platelet lysis, the immobilization of the immune complexes to streptavidin-coated wells. In both cases, the bound antibodies are determined by alkaline phosphatase conjugated anti-human IgG. Using BAIPG type I, positive results were obtained in 7 33 patients with idiopathic thrombocytopenic purpura (ITP), 1 10 patients with secondary immune thrombocytopenia (SIT) and 4 17 with non-immune thrombocytopenia (NIT). The BAIPG type II test was positive in 13 out of 33 patients with ITP, in six out of ten patients with SIT, and in three out of the 17 patients with NIT. A comparison between BAIPG and monoclonal antibody immobilization of platelet antigens (MAIPA) assays showed a high degree of correlation between the two methods. These results suggest that the BAIPG assay is a valuable new tool for the detection of anti-platelet antibodies.

Frozen Platelet Plates for Platelet Antibody Detection and Cross-Match

We performed 2 studies aimed at developing a frozen platelet panel suitable for platelet cross-matching. The stability of the most important platelet membrane glycoproteins and the reactivity of antigens of the human platelet antigen (HPA) and of the human leukocyte antigen (HLA) systems were evaluated with the platelet suspension immunofluorescence test (PSIFT) in a panel of platelets frozen in microplates with 6% dimethylsulfoxide. In study No. 1 we evaluated platelet reaction with a broad-spectrum weak anti-HLA and a potent anti-HPA-la antiserum and the expression of glycoproteins lb and Ilb/IIIa complex on platelet membrane before freezing and after 0.5,1,2, 3,4,5,6 and 12 months of storage at -80°C. In study No. 2 we examined platelet reactivity with anti-HPA-lb, -HPA-2a, -HPA-3a, -HLA-A2, -HLA-A3 of platelets stored frozen for 12 months in parallel with fresh platelets from the same donors. Study No. 1 showed that glycoprotein expression was stable and that the weak anti-HLA and the potent anti-HPA-la antibodies were clearly detected during 12 months at -80°C. Of the 35 paired PSIFT performed in study No. 2 with fresh and frozen/thawed platelets incubated with anti-HPA-lb, -HPA-2a, -HPA-3a -HLA-A2, -HLA-A3 antisera and AB serum, concordant reactions were obtained in all cases with the exception of 1 case of HLA-A3-positive platelets incubated with anti-HLA-A3 antiserum, that was reactive with frozen/thawed platelets but nonreactive with fresh platelets from the same donor. The discrepant finding obtained with fresh platelets from 1 donor could be due to the well-known variable and weak association of HLA antigens to platelet membrane. We conclude that frozen platelet plates can be stored and used for at least 12 months for detecting platelet-reactive antibodies in patients’ sera.

Detection of clonal platelet antibodies in immunologically-mediated thrombocytopenias: association with circulating clonal/oligoclonal B cells.

Aware that T and B cells in autoimmune thrombocytopenia are abnormal, including the existence of clonal B cell populations, we sought to characterize this clonal phenomenon in various immunological thrombocytopenias using platelet antibody light chain analysis, flow cytometry, Southern blot analysis, and PCR. Using a monoclonal antibody-antigen capture ELISA, we analysed sera from 21 of 26 patients with autoimmune, alloimmune, or drug-induced immunological thrombocytopenia for the light chain phenotypes of their platelet antibodies. Alloantibodies and drug-dependent antibodies from four and 14 patients, respectively, were found that expressed a predominant type of light chain, suggesting that these platelet-reactive antibodies were monoclonal or oligoclonal in nature. 14 of the 26 patients were available for light chain B cell phenotyping studies. Of these 14 patients, thrombocytopenia was due to autoimmunity in two, drug-induced immunity in four, and alloimmunity in eight. We detected clonal populations of B cells in all 14 patients by flow cytometry. Although six of these latter patients possessed platelet antibodies with clonal characteristics, light chain phenotypes of antibodies in five patients were opposite to those of their B cells. Eight of these patients were further examined for immunoglobulin gene rearrangement using Southern and/or polymerase chain reaction analysis. In all eight patients we detected clonal or oligoclonal B cell populations. Only two of these patients had malignancies (chronic lymphocytic leukaemia) that would be expected to have detectable clonal B cells, and thus the mechanism for clonal expansion in the other six patients did not appear to be related to an obvious neoplastic process. Prior to these studies, detection of clonal B cells in thrombocytopenic patients without known malignancies was limited to individuals with autoimmune thrombocytopenia, prompting the speculation that this particular autoimmune disorder arises from B cell dysregulation, rather than from expansion of specific autoantibody producing B cell clones. In contrast, the current studies provide evidence that clonal B cells are common to patients with any form of immunologically-mediated thrombocytopenia. Moreover, the majority of the platelet antibodies (86%) present in these disorders exhibited monoclonal characteristics in that there was an apparent restriction in light chain usage.

Antibodies to platelet glycoprotein V in polytransfused patients with haematological disease.

Polytransfused patients often develop platelet-reactive antibodies (PRAb). These give positive reactions in the platelet immunofluorescence test (PIFT) and may be either lymphocytotoxic (LCTAb) or platelet-specific antibodies (PSAb). The latter may be detected in the PIFT using chloroquine-treated platelets (Chl-PIFT) or by immunoblotting. Serial samples from 106 multiply transfused patients with bone marrow failure were screened by PIFT using a microplate method and flow-cytometric analysis. PSAb activity was confirmed by Chl-PIFT. In 45 (42%) of the patients studied PSAb were detected; 37 (35%) formed LCTAb and 19 (51%) had co-existent PSAb. Sera from 25 of 27 patients with a positive Chl-PIFT, retested by immunoblotting, recognised determinants of Mr 82-160 kD on whole platelets. A large group became sensitised to a component of Mr 105-115 kD reduced (99 kD non-reduced) with similar electrophoretic mobility to GPIIIa using a monoclonal anti-GPIIIa and two human polyclonal anti-HPA-1a sera; some also produced anti-GPIIb. The largest group recognised a determinant of Mr 80-83 kD, probably glycoprotein V (GPV). Three sera were immunoblotted against thrombin-treated platelets and the results confirmed GPV specificity.

Antibodies to Platelet Glycoprotein V in Polytransfused Patients with Haematological Disease

Polytransfused patients often develop platelet-reactive antibodies (PRAb). These give positive reactions in the platelet immunofluorescence test (PIFT) and may be either lymphocytotoxic (LCTAb) or platelet-specific antibodies (PSAb). The latter may be detected in the PIFT using chloroquine-treated platelets (Chl-PIFT) or by immunoblotting. Serial samples from 106 multiply transfused patients with bone marrow failure were screened by PIFT using a microplate method and flow-cytometric analysis. PSAb activity was confirmed by Chl-PIFT. In 45 (42%) of the patients studied PSAb were detected; 37 (35%) formed LCTAb and 19 (51%) had co-existent PSAb. Sera from 25 of 27 patients with a positive Chl-PIFT, retested by immunoblotting, recognised determinants of Mr 82-160 kD on whole platelets. A large group became sensitised to a component of Mr 105-115 kD reduced (99 kD non-reduced) with similar electrophoretic mobility to GPIIIa using a monoclonal anti-GPIIIa and two human polyclonal anti-HPA-la sera; some also produced anti-GPIIb. The largest group recognised a determinant of Mr 80-83 kD, probably glycoprotein V (GPV). Three sera were immunoblotted against thrombin-treated platelets and the results confirmed GPV specificity.

Carbamazepine (Tegretol)- Induced Thrombocytopenia

A case of carbamazepine-induced thrombocytopenia is reported. The patient, a 38-year-old female, developed thrombocytopenia 10 days after receiving 400 mg carbamazepine daily. The presence of an anti-IgG carbamazepine- dependent platelet-reactive antibody was documented. Using flow cytometry, the drug-induced antibody was detected at a drug concentration as low as 10^-6 M.

Immunoglobulin Classes and Subclasses of Platelet Antibodies in a Case of Post-Transfusion Purpura

A multiparous white woman developed severe thrombocytopenia following transfusion of red blood cells. Six ineffective platelet transfusions (a total of 42 random donor concentrates) were given from day 0 to day +6, high-dose steroids from day +1, progressively tapered until day +30, and a total of 150 g of intravenous immunoglobulins from day +2 to day +6. As platelet count had not increased significantly by day +8, four plasma exchange procedures, each consisting of 2,000 ml of plasma exchanged with 5% albumin solution, were performed on days +8, +10, +14 and +18. Platelet count was 5, 50, 100 and 234 x 10^9/l on days +8, +14, +26 and +32 (discharge), respectively. The patient's acute phase serum contained increased levels of platelet alloantibodies with anti-HPA-la (P1A1, Zwa) specificity and a titer of 3,200. IgGl, IgG2 and IgG3 subclasses of platelet-reactive antibodies in the patient’s serum were elevated, whereas IgG4, IgM and IgA were within the reference values. Levels of IgGl, IgG2 and IgG3 of antiplatelet antibodies showed a marked and parallel reduction during treatment, but were still above the reference values at the end of treatment and 1 year later, when the patient platelet count was normal. Although a failure of intravenous immunoglobulins cannot be proven in this case, plasma exchange seems to have contributed more than intravenous immunoglobulins to clinical remission.

Detection of drug‐dependent, platelet‐reactive antibodies by antigen‐ capture ELISA and flow cytometry

The effectiveness of flow cytometry in the detection of drug-dependent, platelet-reactive antibodies was investigated. In studies of seven sera known to contain quinine- or quinidine-dependent, platelet-reactive antibodies, flow cytometry was 5 to 10 times more sensitive in detecting drug-dependent antibodies (DDAbs) than the 51Cr release assay, antigen-capture enzyme-linked immunosorbent assay (ELISA), and indirect immunofluorescence microscopic assay. With flow cytometry, DDAbs could be detected at drug concentrations as low as 0.1 microM, or less than one-tenth the level required with other methods. Antigen-capture ELISA was not as sensitive as flow cytometry in DDAb detection, but it did allow identification of the DDAbs' target molecules. With this assay, five of the seven DDAbs recognized both the glycoprotein Ib/IX (GPIb/IX) and glycoprotein IIb/IIIa (GPIIb/IIIa) complexes, while the remaining two sera reacted only with GPIb/IX. Of 44 consecutive patients who developed thrombocytopenia while taking quinidine, DDAbs were detected by flow cytometry in 11 (25%), more than twice the number detected by other methods. In one patient who developed thrombocytopenia while taking trimethoprim/sulfamethoxazole, DDAbs could be detected only by flow cytometry. It can be concluded that flow cytometry is highly sensitive in detecting DDAbs and allows their detection at pharmacologic concentrations of the drug. Most quinidine-dependent antibodies recognize at least two different glycoprotein complexes in the platelet membrane.

Identification of the platelet glycoprotein IIb/IIIa complex as a target antigen in primary biliary cirrhosis-associated autoimmune thrombocytopenia: Evidence that platelet-reactive autoantibodies can also bind to the mitochondrial antigen M2

A 67-year-old woman with a 4-year history of primary biliary cirrhosis (PBC) unexpectedly developed autoimmune thrombocytopenia. The platelet-bound IgG antibody was eluted from the patient's platelets to determine the platelet target antigen. The autoantibodies were found to precipitate the platelet glycoprotein complex IIb/IIIa of autologous and allogeneic platelets. A further precipitate of 70 kDa was detectable under reducing conditions. In addition, platelet-reactive antibodies bound to the 70 kDa mitochondrial antigen M2. No cross-absorption studies were performed to confirm that a single antibody reacted with both antigens. Computer analysis of published peptide sequences of the mitochondrial protein and the platelet GPIIb/IIIa complex showed partial amino acid sequence homology suggesting the possibility of a common antibody binding site. These findings suggest a relationship between the development of autoimmune thrombocytopenia in PBC and the underlying liver disease.

In vitro studies with the platelet-reactive antibody 50H. 19 and its fragments

The platelet-reactive monoclonal antibody 50H.19, its F(ab′) 2 fragments, and 99m-technetium (99m-Tc)-labeled fragments used in thrombus imaging were evaluated for their ability to cause platelet aggregation. The intact antibody caused a dose-dependent platelet aggregation in either platelet-rich plasma (PRP) or defined buffer solutions. The F(ab′) 2 fragments did not cause platelet aggregation except at high calcium concentrations. Neither stannous-ion-treated antibody fragments nor 99m-Tc-labeled antibody fragments caused platelet aggregation. The antibody-induced platelet aggregation was completely inhibited by 8-bromo-cAMP, caffeine, theophylline diltiazem, and staurosporin; partially inhibited by EDTA, EGTA, cytochalasin B, colchicine, aspirin, or indomethacin. Treatment of platelets with the intact 50H.19 antibody resulted in phosphorylation of a 40K dalton platelet protein, similar to that caused by treatment with phorbol myristate acetate (PMA). Phosphorylization of the 40K protein was not observed after treatment with either the 50H.19 F(ab′) 2 fragments or the calcium ionophore A23187.

Development of lymphocytotoxic and platelet reactive antibodies: a prospective study in patients with acute leukaemia.

Lymphocytotoxic (LCT) and platelet reactive (PR) antibody (Ab) responses were serially determined in 49 patients with acute leukaemia. LCTAb were found in 20 patients and occurred in 13 patients with acute myeloid leukaemia and 7 patients with acute lymphoblastic leukaemia. Four differing patterns of LCTAb responses could be defined. Thirteen of 22 subjects showed marked reduction or loss of LCTAb. Indirect platelet immunofluorescence, measured by flow cytometry, provided the most convenient means of detecting PRAb which were found in 11 subjects and generally showed moderate or weak reactivity.

Effect of intravenous gammaglobulin on circulating and platelet-bound antibody in immune thrombocytopenia

Ten patients with idiopathic thrombocytopenic purpura (ITP) were studied before and following a rise in circulating platelets subsequent to infusions of intravenous gammaglobulin (400 mg/kg/day x 5 days). We quantitated the amount of circulating IgG capable of binding to normal donor platelets in vitro using an 125I-monoclonal anti-human IgG assay, as well as the amount of IgG associated with the patients' platelets before and following therapy. We found no evidence for a decrease in platelet-specific IgG antibodies in these patients undergoing an acute response to therapy. These data suggest that the short-term efficacy of intravenous gammaglobulin is due to effects other than a substantive reduction in platelet reactive antibodies, such as the alteration of IgG-coated platelet destruction.

Effect of Intravenous Gammaglobulin on Circulating and Platelet-Bound Antibody in Immune Thrombocytopenia

Abstract Ten patients with idiopathic thrombocytopenic purpura (ITP) were studied before and following a rise in circulating platelets subsequent to infusions of intravenous gammaglobulin (400 mg/kg/day x 5 days). We quantitated the amount of circulating IgG capable of binding to normal donor platelets in vitro using an 125I-monoclonal anti-human IgG assay, as well as the amount of IgG associated with the patients' platelets before and following therapy. We found no evidence for a decrease in platelet-specific IgG antibodies in these patients undergoing an acute response to therapy. These data suggest that the short-term efficacy of intravenous gammaglobulin is due to effects other than a substantive reduction in platelet reactive antibodies, such as the alteration of IgG-coated platelet destruction.

Quantitation of platelet antigens after chloroquine treatment

The effect of chloroquine treatment on the expression of various platelet antigens was studied. For this purpose, the binding of several human platelet-reactive antibodies (polyspecific HLA antibodies; platelet-specific PlA1, PlA2, Baka antibodies; platelet autoantibodies) and of murine monoclonal antibodies specific for epitopes on the glycoprotein (Gp) complexes IIb/IIIa or Ib/IX were quantitated by means of a competitive enzyme-linked immunoassay (CELIA) using fresh and stored chloroquine-treated platelets. We found that HLA antigens were substantially removed from platelets following chloroquine treatment (confirming earlier results) while platelet-specific antigens on the Gp complex IIb/IIIa were but little affected. Epitopes on the Gp complex Ib/IX did not show any alteration. Storage of chloroquine-treated panel platelets known to interfere with immunofluorescence had no effect on quantitative IgG determinations by CELIA. We conclude that chloroquine treatment of platelets is practicable for platelet antibody analysis, but exerts its effect unspecifically and requires cautious interpretation.

Development of Lymphocytotoxic and Platelet Reactive Antibodies: A Prospective Study in Patients with Acute Leukaemia

Lymphocytotoxic (LCT) and platelet reactive (PR) antibody (Ab) responses were serially determined in 49 patients with acute leukaemia. LCTAb were found in 20 patients and occurred in 13 patients with acute myeloid leukaemia and 7 patients with acute lymphoblastic leukaemia. Four differing patterns of LCTAb responses could be defined. Thirteen of 22 subjects showed marked reduction or loss of LCTAb. Indirect platelet immunofluorescence, measured by flow cytometry, provided the most convenient means of detecting PRAb which were found in 11 subjects and generally showed moderate or weak reactivity.

In vitro fixation of C3d and C5b-9 on platelets by human platelet reactive antibodies.

Most platelet-reactive autoantibodies and alloantibodies are not able to fix complement in vitro. However, exceptions have been found. These antibodies are usually characterized by the conventional platelet complement fixation test. A recently developed competitive enzyme immunoassay for quantitation of platelet-associated immunoglobulins and a modification thereof allowed the quantitative study of fixation of C3d and the membrane attack complex (C5b-9) on platelets by HLA antibodies, human platelet autoantibodies, and drug-dependent antibodies (ddab). The highest amounts of both complement products were fixed through ddabs, whereas autoantibodies only showed moderate complement fixation. This enzyme immunoassay is a valuable tool for the characterization of the complement-fixing properties of platelet-reactive antibodies.

Human anti‐PlEl antibody recognizes epitopes associated with the alpha subunit of platelet glycoprotein Ib

Together, a platelet-reactive antibody in the serum of a polytransfused patient (proband) and a platelet-reactive antibody in the serum of a mother of an infant with neonatal thrombocytopenia have served to establish the diallelic, platelet-specific alloantigen system, PlE. We now provide evidence that the platelet-specific antibody in the serum of the proband, anti-PlE1, recognizes epitopes associated with the alpha subunit of glycoprotein (GP) Ib. By 51Cr release, platelets from two of three patients with the Bernard-Soulier syndrome (BSS) responded sub-normally to anti-PlE1, and the apparently normal response of platelets from the last BSS patient was attributable to anti-HLA-A2 antibodies in the proband serum. These results suggested that the PlE1 antigen is associated with the GPIb complex (glycoproteins Ib kX) known to be absent from BSS platelets. This possibility was confirmed by ELISA using the purified GPIb complex or glycocalicin, the N-terminal fragment of GPIb alpha produced by proteolysis with endogenous platelet calpain, as solid-phase antigen. Anti-PlE1 antibody bound specifically to both the GPIb complex and glycocalicin. 3H-labelled platelet membrane glycoproteins with apparent molecular weights of 130k, 25k, and 21k (under reduced conditions) corresponding to GPIb alpha, GPIb beta, and GPIX were immunoprecipitated by anti-PlE1 plasma. Finally, at a titre of 1:16, anti-PlEl completely inhibited ristocetin-induced platelet agglutination, a property of platelets mediated by GPIb.

Relationship of HLA and platelet-reactive antibodies in alloimmunized patients refractory to platelet therapy.

Platelet crossmatching assays have been used to predict the outcome of platelet transfusions in alloimmunized patients by detecting antibodies against platelets. The transfusion failure of HLA-matched platelets predicted by platelet crossmatching may be related to HLA antibodies undetected by lymphocytotoxicity but detected by platelet immunoglobulin-binding assays or platelet-specific antibodies (both antibodies defined here as platelet-reactive antibodies). To differentiate platelet-reactive antibodies from lymphocytotoxic HLA antibodies, we used HLA characterized lymphocytes in parallel with platelets from individuals to form separate frozen panels. Sera from 10 allosensitized patients were studied in the lymphocyte panel by lymphocytotoxicity and in the platelet panel by enzyme-linked immunoassay (ELISA). By comparing pattern and percent wells reacting in each panel, lymphocytotoxic HLA antibodies and antibodies reactive with platelets in ELISA were detected separately. In all 10 allosensitized patients, platelet-associated antibodies were present and 7 had additional lymphocytotoxic HLA antibodies. Using this double parallel panel technique, we found platelet-reactive antibodies important in platelet alloimmunization, unrecognized by lymphocytotoxicity. These data indicate platelet-crossmatching be solely used in the selection of platelets for allosensitized patients.