Search

Two hundred and eight pregnancy sera were tested for the presence of antibodies specific for lymphocyte sub-populations by using the isolated B and T lymphocytes from the women's mating partners. This was done by the microlymphocytotoxicity and the indirect immunofluorescence techniques. Five sera (2.5%) reacted exclusively with B lymphocytes and sixty-three sera (30.2%) reacted with both B and T lymphocytes; none of the sera was specific for T cells. Several sera, reacting with both B and T lymphocytes, were absorbed with platelets and this procedure revealed nine additional antiseraa specific for B lymphocyte antigens. Specificity studies on a panel of forty-eight HLA-ABCD typed individuals indicated that most antisera possibly defined new B-cell antigens. Family studies established that the antigens defined by these antiser were coded for by genes in the Major Histocompatibility Complex.

Summary. The purpose of this study was to determine whether quantitative or qualitative factors are of major importance in the destruction of red cells sensitized with incomplete warm autoantibodies of subclass IgG1. To that end, the relative amount of IgG1 antibody present on the red cells of patients with autoantibodies of this subclass only, was measured by means of continuous flow cytofluorometry. This method appeared to give an idea of the amount of antibody on red cells and was reproducible. The intensity of the fluorescence of patient's red cells, measured after incubation with a FITC-labelled anti-IgG1, was compared with the presence or absence of signs of increased haemolysis in vivo and the cytotoxic activity of normal monocytes towards these red cells in vitro. It appeared that it was predominantly the amount of IgGl autoantibody that determined whether or not these antibodies induced haemolysis in vivo or cytotoxicity of monocytec in vitro. This was also true with methyldopa-induced IgG1 autoantibodies.

immunization in patients whohave been previously immunized by pregnan-cy or transfusion is not prevented. Neither canimmunization against platelet-specific al-loantigens be prevented by removing leuco-cytes from platelet concentrates, but platelet-specific antibodies seem to occur mainly inpatients who develop HLA class I antibodies.Not all patients with HLA antibodies are re-fractory. Often the antibodies only react witha small percentage of panel lymphocytes andthen most random donors are compatible.Furthermore, although there are individualdifferences, some HLA class I antigens areweakly expressed on platelets. Nevertheless,refractory alloimmunized patients in need ofplatelet transfusions will continue to occur.Opinions on the optimal management of suchpatients differ. Moreover, some simple and in-teresting new crossmatch techniques have re-cently been developed. For these reasons itseemed of interest to devote an InternationalForum to the subject.There seem to be two main approaches tothe management of refractory patients: (I) theselection of platelet concentrates based on theHLA class I phenotype and, or the specificityof the HLA (and platelet-specific) antibodies,and (II) the selection of platelet concentratesby crossmatching.To obtain more information on the sub-ject, the following questions were submittedto 15 experts in the field

In cases of Rh D alloimmunization, strong results in the antibody-dependent cell-mediated cytotoxicity (ADCC) assay (> 80% lysis as compared to that of the standard anti-D serum) are indicative of severe hemolytic disease to occur in the newborn (HDN). However, discrepant cases were found in which the maternal anti-D gave strong ADCC results and the newborns had no or only mild hemolysis. In the majority of these cases the mother had produced monocyte-reactive IgG alloantibodies, mostly with HLA-DR specificity. Such antibodies may be capable of blocking FcR-mediated functions of the fetal MPS, and it has been postulated that they inhibit destruction of anti-D-sensitized red cells. We here describe 2 families in which such discrepancies were noticed. In 1 family, in spite of ADCC results of > 80%, the Rh D-positive second child was born without signs of hemolysis. However, the Rh D-positive third child suffered from very severe hemolytic disease. The mother had produced monocyte-reactive HLA-DR antibodies in the second pregnancy which, however, did not react with the cells of the third child. In the other family, the severely Rh D-alloimmunized mother had lost her fourth child because of intrauterine death due to HDN. The Rh D-positive fifth child was born with only mild HDN and only in this pregnancy had the mother produced such antibodies. These 2 case histories give further evidence that maternal monocyte-reactive alloantibodies, in both these cases directed against HLA-DR antigens, can protect the Rh D-positive child against hemolytic disease in case of severe Rh D alloimmunization.

Salama et al. [1] noticed that in patients with autoimmune thrombocytopenia (AITP) treated with high-dose i.v. Ig, the red cells became weakly sensitized with IgG antibodies, present in the i.v. Ig preparation. They concluded that the effect of i.v. Ig on the platelet count is due to a competition between sensitized red cells and platelets sensitized with auto-antibodies, for Fc-receptors (FcR) on splenic macrophages. This observation was the basis of their treatment of patients with AITP with i.v. anti-D IgG. Since the first publication of Salama et al. [2] on the subject, many patients have been treated in several centres. There is controversy concerning the question whether the effect of i.v. anti-D IgG is indeed due to a competition between sensitized red cells and sensitized platelets for FcR and therefore to the RhD antibodies themselves, or to other factors, also present in the anti-D preparation. This question is interesting from a theoretical but also from a practical point of view. There is a general shortage of anti-D Ig for immunoprophylaxis in D-negative women. If not the anti-D but other factors are responsible for the effect in AITP, anti-D Ig should not be used to treat patients with this disease. These considerations led us to devote an International Forum to this subject. The questions listed below were sent to 18 experts of whom 7 have contributed to the Forum. Questions (1) Do you, or your colleagues, treat patients with AITP with i.v. anti-D Ig or have you done so in the past? (2) If so, what have the results of this treatment been? (3) If treatment with anti-D causes an increase in the platelet count (a) do you think that this effect is due to a competition between anti-D IgG sensitized red cells and auto-antibody-sensitized platelets for FcR on macrophages? If so, how do you explain the simultaneous, often severe, destruction of red cells by IgG warm auto-antibodies and of platelets by auto-antibodies in Evans’ syndrome? (b) do you think that any effect on the platelet count is not due to anti-D but, e.g, to other antibodies present in the anti-D preparation? (c) do you have experience with anti-D treatment of Rh-D-negative patients with AITP? (d) do you have experience with anti-D treatment of patients with IgA platelet autoantibodies?

IgG present in the circulation of newborn infants is predominantly of maternal origin. In cases of severe Rh D alloimmunization, the maternal anti-D anti-bodies may lead to Rh D-haemolytic disease of the newborn. However, an impaired placental transport of maternal IgG could be associated with the protection of the fetus/newborn against haemolytic disease as shown in this case report.

The extent to which maternal anti-Rh(D) antibodies support lysis of erythrocytes by monocytes in the antibody-dependent cell-mediated cytotoxicity (ADCC) assay is closely correlated with the severity of Rh(D) haemolytic disease of the newborn infant (HDN). However, in some cases HDN is much milder than would be predicted from the ADCC value. We postulated that maternal ADCC-blocking alloantibodies against paternal antigens on monocytes can protect these infants against severe haemolysis. We studied 13 severely Rh(D)-alloimmunised mothers whose infants showed unexpectedly mild HDN (group I) and 14 women with similar ADCC values but whose infants had severe HDN (group II). 7 group-I women had monocyte-reactive IgG alloantibodies that inhibited lysis by paternal monocytes in the ADCC. No such antibodies were found in group II (p less than 0.01). In 6 of the 7 serum samples with monocyte-reactive antibodies, the antibodies had HLA-DR specificity. Our findings suggest that Rh(D)-positive children of some severely Rh(D)-alloimmunised women may be protected from severe HDN by maternal non-HLA-class-I, IgG alloantibodies against paternal monocyte blood-group antigens. These antibodies may inhibit the mononuclear phagocyte system of the fetus.

We describe 2 patients with thrombocytopenia and autoantibodies in the blood against platelet membrane glycoprotein (GP) IIb/IIIa, in whom reactivity of the antibodies with their antigen was greatly diminished in the presence of sodium EDTA. One patient also showed a marked thrombocytopathy, suggesting a functional blocking of GP IIb/IIIa by these autoantibodies. These observations show that some immunologically important autoantibodies can easily be missed in routine serological investigations in which EDTA blood is used, obscuring the autoimmune nature of a thrombocytopenia and possibly of a thrombopathy as well.

The platelets of 11 patients with Glanzmann's thrombasthenia and their nearest family members were studied for the expression of the platelet- specific alloantigens of the Zw-, Ko- and Bak systems. The strength of the expression of the Zwa antigen was diminished on the platelets of 3 patients, and the antigen was absent from the platelets of the other 8. The platelets of none of the patients reacted with anti-Zwb serum. Therefore, Glanzmann's thrombasthenia is probably a “Zw-null disease.” The expression of the Zwa antigen on the platelets of all the relatives was normal, as indicated on the cytoflurograph. Investigations on the expression of the Koa antigen were complicated by agglutinations of the platelets from genetically Koa-negative thrombasthenic patients with the anti-Koa serum. The Kob antigen was normally expressed. The Baka antigen was absent from the platelets of all thrombasthenic patients and a relatively high percentage of the relatives. A close association between the Glanzmann gene and the Bak(a-) gene is assumed on statistical grounds. Thrombasthenic platelets showed no reaction with EDTA-dependent antibodies, which are reactive with all normal platelets. Owing to immunization by multiple blood and platelet transfusions, serum samples of most patients studied contained HLA antibodies and platelet-specific alloantibodies. However, antibodies directed against the Zw-antigen-bearing glycoproteins were detected in the serum of only one patient and, therefore, seem to be rare.

In patients with IgG incomplete non-complement binding warm autoantibodies, the subclass composition of the antibodies was studied in relation to the occurrence of increased haemolysis in vivo and the adherence of the patients red cells to peripheral blood monocytes (PBM) in vitro. The presence of IgG3 autoantibodies was almost always accompanied by haemolytic anaemia, but the presence of IgG1 autoantibodies only in some patients but not in others. IgG2 and IgG4 autoantibodies were not associated with increased red cell destruction. A relation identical to that between subclass composition and increased haemolysis was found between subclass composition and adherence of the patients erythrocytes to PBM and thus a strong correlation between positive adherence in vitro and increased red cell destruction in vivo. These results support an important role of adherence to mononuclear phagocytic cells in the destruction of red cells sensitized with non-complement binding IgG antibodies. Strong indications were found that IgG1 autoantibodies are of two kinds, only one of which causes adherence to phagocytes and thus increased red cell destruction.

Summary. By applying the platelet suspension immunofluorescence test (PSIFT), platelet-specific alloantibodies responsible for post-transfusion purpura were detected in eight patients within a period of 2 years. All patients were female and had previously received blood or had been pregnant. The platelet-specific alloantibodies had the specificity anti-Zwa in all the patients, who were all Zw(a -). In two patients the platelets were tested in the acute phase of the disease and found to be coated with IgG. In one patient an ether eluate was prepared from the platelets that reacted strongly with Zw(a +) platelets, but weakly with platelets from Zwb-homozygous individuals. The sera of these two patients, and of two others whose platelets were not directly tested, taken in the acute phase of the purpura, reacted strongly with Zwa-positive platelets. The four sera also reacted, however weakly, with Zwa-negative platelets, with autologous platelets taken during remission and with platelets from patients with Glanzmann's disease. It is postulated that Zwa-anti-Zwa complexes, present in the eluate and the sera, caused the reaction with Zwa-negative platelets and the patients’own platelets. Immunochemical characterization of the post-transfusion purpura antibodies showed that in all patients these were IgG, in two combined with IgM antibodies. Antibodies of the sub-class IgG1 were found in all patients, sometimes together with IgG3. In the indirect immunofluorescence test with anti-complement serum, the PTP antibodies in only four sera were able to fix complement. In only two of these sera were these complement-binding antibodies detectable in the 51Cr-lysis technique and then in a much lower titre than in the immunofluorescence technique.

Summary. Three techniques, based on the antiglobulin principle, used for the detection of autoantibodies against platelets, were compared; the antiglobulin consumption assay (QACA), the platelet radioactive antiglobulin test (PRAT) and the platelet suspension immunofluorescence test (PSIFT). Upon incubation of normal donor platelets with purified IgG, in concentrations higher than that in serum, an increased amount of platelet-associated IgG was demonstrated only in the QACA. Upon incubation with aggregated IgG, all three tests became positive, but the PSIFT only with high concentrations of aggregates. Binding of soluble C1q-binding immune complexes (IC), which consisted of tetanus toxoid and IgG antitetanus antibodies (TaT) to normal donor platelets, was only detectable in the QACA. However, a positive result was obtained in all three tests with platelets incubated with soluble DNA-IgG-antiDNA antibodies (DaD) IC. Fixation of the platelets with paraformaldehyde prevented the binding and the detection of the DaD-IC, but not of IgG, aggregated IgG or TaT-IC. Eluates from platelets incubated with aggregated IgG, TaT- or DaD-IC did not react with normal donor platelets in the three techniques, in contrast to eluates from platelets sensitized with platelet antibodies.

The purpose of this investigation was to test whether the generation of oxygen radicals is required for the cytotoxic activity of human monocytes. The cytotoxic activity of human monocytes was measured as the capacity to lyse, in vitro, human red cells sensitized with non-complement-binding IgG anti-D alloanti-bodies. Lysis was determined by measuring 51Cr release from the red cells. Studies were performed with monocytes from patients with chronic granulomatous disease (CGD). In contrast with normal neutrophils, phagocytosing neutrophils from these patients are unable to generate oxygen radicals, hydrogen peroxide, or singlet oxygen. We found that monocytes from CGD patients were similarly incapable of generating these reactive oxygen species during phagocytosis. This indicates that, like neutrophils, monocytes from CGD patients are defective in the generation of activated oxygen products. Despite this defect, monocytes from CGD patients displayed a normal cytotoxic activity towards sensitized red cells. In addition, neither scavengers of superoxide, hydrogen peroxide, hydroxyl radicals, or singlet oxygen, nor inhibitors of myeloperoxidase, had any effect on the cytotoxic activity of normal monocytes. Together, these results indicate that the cytotoxic activity of human monocytes towards sensitized red cells is not dependent on the generation of reactive oxygen derivatives.

The effect of high-dose intravenous gammaglobulin (IVG) therapy with a CLB preparation was studied in 42 patients: 8 patients had acute and 26 patients had chronic idiopathic thrombocytopenic purpura (ITP); 5 patients had thrombocytopenia accompanied by various diseases such as systemic lupus erythematosus, auto-immune haemolytic anaemia and neutropenia; 3 patients had hypoplastic anaemia and 1 patient had neutropenia and rheumatoid arthritis. After treatment, a rise in platelet count occurred in about 75% of the patients with ITP, although there was no sustained response in any of the patients. There was no correlation between the strength of platelet antibodies as detected by the direct immunofluorescence test before infusion and the pattern of response to the infusion. In most cases of ITP, no immune complexes, as measured by Clq-binding assay, were observed. Furthermore, we found no relationship between the amount of Clq-binding activity of patients' sera and the reaction pattern after infusion of IVG. Splenectomy of the patient had no influence on the outcome of IVG therapy.

A family is presented, in which the mother developed strong neutrophil granulocyte-specific antibodies with the specificity anti-NA1. The first and third children, both NA1-positive, suffered from severe transient neonatal neutropenia, but the second child, NA1-negative, did not. Anti-NA1 antibodies were detectable in the serum of one affected child and the antibody level was directly related to the extent of neutropenia. The incidence of granulocyte-specific antibodies in 198 pregnant women was determined and found to be low.

In 42 patients with autoimmune thrombocytopenia (AITP) and a positive direct platelet suspension immunofluorescence test (PSIFT), the antigenic specificity of the autoantibodies was studied. Because the autoantibodies were often not detectable in the serum and additional HLA antibodies may disturb the reaction pattern with the platelet panel, we used eluates prepared from the patients' platelets for this study. Thirty-five patients had antibodies equally reactive with normal platelets, irrespective of their antigenic make-up, but not with the platelets from two Glanzmann's disease patients. Absorption and elution experiments in two patients showed that his was probably not due to the presence of a combination of anti-Zwa and anti-Zwb antibodies. Thus, the majority of autoantibodies against platelets seems to be directed against antigenic determinants not present on Glanzmann's disease platelets, but perhaps located on the platelet-membrane glycoproteins IIb and/or IIIa. In ten patients, antibodies of no, or still unknown, specificity were detected. Three of these had additional antibodies not reactive with the platelets of the two Glanzmann patients.

Both 51Cr-survival studies with donor platelets and allogenic skin transplantations were performed in a patient with immune unresponsiveness to platelet antigens, i.e., HLA-and platelet-specific antigens. The patients, who suffered from hypoplastic anaemia, was successfully transfused with random donor platelets during 13 months. The serum of this patient contained only granulocyte-and mononuclear-cell-reactive antibodies, but no platelet-reactive antibodies. A nearly normal survival time of the donor platelets as well as a prolonged rejection time of a skin allograft of the same donor support the serological findings. The 51Cr-platelet survival time was not influenced by a leucocyte concentrate from the same donor, which was administered at the same time. Thus, in our patient, no increased platelet destruction could be induced via the so-called innocent bystander mechanism.

An antibody is described, produced by an Fy(a+b-) woman that at first reacted with her own cells, preferentially with Fy(b+) cells but not with Fy(a-b-) cells. Later, pure anti-Fyb specificity occurred. Elution and absorption studies revealed that the antibody recognized a determinant absent from Fy(a-b-) cells but present on Fy(a+b-) and Fy(a-b+) cells, much more strongly, however, on the latter, and that it was an autoantibody.

Cell electrophoresis was studied as a method for the detection of antibodies against platelets. Use was made of a Zeiss cytopherometer. After replacement of the original copper-copper sulfate electrodes by newly designed platinum electrodes, a good reproducibility of the method was obtained. Antiplatelet heteroantibodies and usually also antiplatelet isoantibodies were found to reduce the electrophoretic mobility of platelets. The influence of drug-induced antiplatelet antibodies on the electrophoretic mobility appeared to be dependent on the presence of complement in the incubation mixture, as was the case with some anti-red cell antibodies. From studies with sera from patients recovered or suffering from idiopathic thrombocytopenic purpura and with platelets from patients with thrombocytopenia it was concluded that cell electrophoresis is unsuitable for the demonstration of the thrombocytopenic factor.

By lowering the pH of the medium, allo- and autoantibodies bound to granulocytes and platelets could be eluted from these cells and their presence in the eluate could subsequently be detected with normal donor granulocytes and platelets, respectively, in the immunofluorescence test and/or agglutination test. IgG- as well as IgM- and IgA-containing antibodies were detectable in the eluates. Depending on the attraction mechanism between antigen and antibody, some antibodies were also eluted by dimethylsulphoxide at pH 9.5, ether and/or by heating at 56 degrees C. By differential elution of antibodies from donor platelets, we demonstrated the simultaneous presence of platelet-specific antibodies and HLA antibodies in the serum of a mother whose newborn suffered from transient neonatal thrombocytopenia.

Clinical data on 49 patients with chronic idiopathic neutropenia (CIN) and 42 patients with neutropenia secondary to a well-defined immunological disorder (SN) were collected and related to serological parameters. In 47% of the patients with CIN and 53% of those with SN, a positive direct immunofluorescence test was obtained with granulocytes from the patients. In the sera from the patients in the two groups, antibodies against donor granulocytes were detected by the indirect immunofluorescence test, the leucoagglutination test and/or the granulocytotoxicity test in 15%, 19% and 15%, respectively. The results of the above tests could not be correlated with any clinical or haematological parameter. Immune complexes in the serum were detected by the 125I-Clq-binding test in 29% of patients with CIN and in 58% of those with SN. The presence of serum immune complexes correlated well with the existence of a low neutrophil count, but not with the presence of recurrent infections, with bone-marrow abnormalities, or with positive reactions in other serological tests. The sera of eight out of 14 patients with CIN and seven out of 12 patients with SN had inhibitory activity for myeloid colony formation in vitro (CFU-GM). This CFU-GM inhibitory activity was correlated with the presence of recurrent infections and with hypoplasia of the myeloid compartment of the bone marrow, but not with positive reactions in other tests. We conclude that the 125I-Clq-binding test probably detects circulating immune complexes that induce a shift neutropenia, whereas serum activity inhibitory for CFU-GM possibly relates to clinically more serious forms of neutropenia. The significance of neutrophil-bound Ig and granulocyte-reactive antibodies in the serum is not clear.

Platelet immunofluorescence, together with other serologic tests on platelets, lymphocytes, and granulocytes, was used to investigate the sera of 38 mothers with newborns who suffered from thrombocytopenia. In sera of 33 mothers, platelet-specific IgG alloantibodies were demonstrable. Three sera also contained HLA antibodies, of which two were only detectable in the lymphocyte cytotoxicity test. Two other sera contained granulocyte-specific alloantibodies. In sera of 2 mothers, antibodies were found that reacted with all cell types in all tests. However, after further analysis, it became clear that platelet- specific alloantibodies were probably also present in these 2 sera. In 29 cases, the specificity of the platelet alloantibodies was anti-Zwa-- PlA1. One serum contained antibodies directed against a new antigen, Baka. This new antigen was defined after the investigation of the family and a small-scale population study. Two other sera had platelet antibodies with still undefined specificities. In all positive sera, IgG platelet alloantibodies were detected, and sometimes IgM antibodies were also present. The IgG antibodies were mostly of the IgG1 subclasses, but sometimes IgG3 and/or IgG4 was also found. In a few sera, only IgG3 antibodies were detected. In our series, we found no increased frequency of blood group ABO compatibility between mother and child, although it has been described by others and is well known to occur in rhesus alloimmunization. Of all the tests used, the platelet immunofluorescent test, especially the test on paraformaldehyde-fixed platelets in suspension, gave the best results in the detection of platelet antibodies in neonatal alloimmune thrombocytopenia.

IgG antibodies against blood group antigens A or B (anti-A/B) are able to sensitize erythrocytes for destruction in an antibody-dependent cell-mediated (ADCC) assay with monocytes as effector cells. The activity of maternal IgG anti-A/B in this test was compared with clinical signs of haemolytic disease of the newborn (HDN). When the ADCC was negative (less than 10% of the sensitized cells lysed), signs of increased red-cell destruction in the children were never observed. In three cases with a strongly positive ADCC (greater than 45% lysis), the children were severely affected and needed more than one exchange transfusion. In the cases with greater than 10% but less than 45% lysis in the ADCC, there was no clear correlation between the result of the ADCC and the degree of lysis in the newborn infants. In these cases, the degree of lysis of the red cells of the infant was shown to be strongly influenced by the number of A/B antigens per red cell. There was a direct correlation between the degree of lysis in the ADCC and the titre of IgG3 anti-A/B in the sera. There was comparable activity of maternal IgG anti-A/B in the ADCC test in the 32nd week of pregnancy and at the moment of delivery.

Antibodies specifically reacting with platelets only in the presence of EDTA, by the platelet immunofluorescence test, were found in the serum of 20 patients with pseudothrombocytopenia due to in vitro EDTA- dependent platelet agglutination. These antibodies reacted optimally at 0–4 degree C. In 19 patients, IgG antibodies were detected; in 8 patients, IgM or IgA antibodies were also found. In one patient, only IgM antibodies were found. In 14 patients, the IgG antibodies were IgG1, but IgG2, IgG3, and IgG4 antibodies were also seen in 7 patients. The reaction of platelets with the antibodies was detectable in the presence of Na2EDTA, the K, Ca, and Mg salts of EDTA, and K2EGTA. F(ab')2 or F(ab') fragments of the IgG antibodies reached as strongly as the intact antibodies, indicating that the reaction is dependent on the antibody-combining site. The EDTA-dependent antibodies did not show platelet-group specificity. However, platelets from patients with Glanzmann disease did not react with the antibodies.

Sera from 40 patients with febrile, nonhemolytic transfusion reactions were tested for the presence of alloantibodies using a number of techniques, including immunofluorescence tests on granulocytes, lymphocytes and platelets, a modified NIH lymphocytotoxicity test and the leukocyte agglutination test. Cells of at least 9 donors were used as target cells. Alloantibodies were detected in all sera. The frequency of the occurrence of antibodies was not much higher in sera obtained about 1 month after the transfusion reaction as in sera obtained within 4 days. Most of these antibodies were anti-HLA, but quite frequently platelet-specific antibodies were found, and sometimes these were the only antibodies detected. Granulocyte-specific antibodies were the least frequent. The nature of the antibodies was specified by their difference in reactivity with the cells of multiple donors, by applying panels of cells from typed donors and by absorption and elution experiments. It appeared that not only granulocyte-specific but also HLA- and perhaps platelet-specific antibodies may be responsible for a febrile transfusion reaction. We did not find that the occurrence of rigors, together with fever, was associated with particular serologic results.

A patient is described with immunoblastic non-Hodgkin's lymphoma and autoimmune haemolytic anaemia of the cold autoantibody type. The autoantibodies were monoclonal IgM-kappa cold haemagglutinins/haemolysins with blood-group specificity, anti-P. Red-cell autoantibodies directed against blood-group-P antigen have until now only been detected, as polyclonal IgG antibodies, in paroxysmal nocturnal haemoglobinuria.

Summary. The sera of 14 patients with quinine/quinidine-dependent thrombocytopenia were studied in the platelet suspension immunofluorescence test (PSIFT), the 51Cr-lysis assay and the complement fixation test (CFT). When anti-Ig or anti-IgG reagents were used, the PSIFT proved to be a little more sensitive than the 51Cr-lysis assay and far more sensitive than the CFT. With the PSIFT, it was possible to determine the immunoglobulin class of the antibodies, which in all sera was IgG of the subclass IgGl. In two sera there were also IgG3 antibodies, and in the sera of three patients additional IgM drug-dependent platelet antibodies were detected. With the anti-C3 reagent, however, lower titres were observed in the PSIFT than in the 51Cr-lysis assay, in contrast to the results with other complement-binding IgG platelet antibodies. The F(ab′), fragments of the antibodies did not react in the PSIFT. This indicates that the Fc part of the antibody is essential for the fixation on the platelet membrane. These findings support the theory that in this syndrome drug-antibody complexes are formed in the absence of platelets that selectively adhere to the Fc receptor on platelets. The platelets of patients with the Bernard-Soulier syndrome (BSS) reacted normally with the quinine/quinidine-dependent antibodies in some of the sera, but not with those in other sera in the PSIFT. However, lysis of BSS platelets was never observed. The results of absorption studies showed a reflection of the results in the PSIFT, indicating that the receptor for the quinine/quinidine-dependent antibodies is not absent in this disease but altered. Whether a reaction with BSS platelets will take place depends on characteristics of the antibodies.

Surface marker analysis with rosette tests and a large panel of xenoantisera and monoclonal antibodies was done on the malignant cells of 55 patients with acute myeloid leukemia (AML). The diagnosis was made on morphological and cytochemical grounds, and the leukemias were classified according to the quantified FAB criteria. The marker tests included the E- and EA-rosette test, immunofluorescence with rabbit- polyclonal antisera against human Ig, kappa, and lambda light chains, thymocytes, granulocytes, erythrocytes, platelets, lysozyme, (leukemic) myeloblasts, the common ALL antigen, SB cell-line cells (anti-Ia), and a mouse anti-Ia serum. The monoclonal mouse antibodies applied were anti-T-cell antibody (3A1), two anti-granulocyte-monocyte antibodies (OKM1 and B2.12), four antigranulocyte antibodies (MI/N1, UJ 308, B4.3, and B13.9), an antiplatelet antibody (C17.28), anti-HLA heavy chains (w6/32.HLK), anti-Ia antigen (OKI1), and OKT10. AML cells from many patients lacked the expression of myeloid markers, and we found that a correlation existed between the relative maturity of the leukemia subtype and the extent of positivity for these markers. Surface marker analysis discriminated poorly between the “myeloid” and “monocytoid” subtypes; OKT10 and the “T-cell marker” 3A1 were often expressed on AML cells. In two cases of AML, there was an unexpected expression of platelet antigens with the monoclonal antiplatelet antibody. One of them, classified as M1, was ultrastructurally a megakaryoblastic proliferation with a positive reaction for platelet peroxidase. Only with the help of computerized analysis, was it possible to prove a clear correlation between the surface marker profile and the FAB classification.

After a survey of the literature dealing with the demonstration of platelet autoantibodies by immunofluorescence techniques, the results are given of a study in which immunofluorescence microphotometry was used for this purpose. The serum of 58, the platelets of 34, and the megakaryocytes of. 2 patients -with thrombocytopenia were investigated. In 21 of 52 sera (40%) in which the presence of platelet autoantibodies could be expected, positive results were obtained that could not be due to isoantibodies, either because the patients had not been pregnant and had not received blood transfusions or because the reactivity of the serum with the patient's own platelets was demonstrated. The platelets of 28 patients with thrombocytopenia not due to a platelet defect or decreased thrombopoiesis were investigated. In the platelets of 15 (54%) of these, significant differences in fluorescence were'found with anti-immunoglobulin conjugate as well as with anti-IgG, -IgA, -IgM, or -complement reagents. It was concluded that in these patients in vivo sensitization of the platelets with autoantibody was demonstrated. In two patients an indication of the in vivo sensitization of the megakaryocytes was also obtained.

Summary An immunofluorescence test on paraformaldehyde-fixed platelets in suspension was used for the detection of antibodies on platelets and in the sera of patients with idiopathic or secondary thrombocytopenia. In idiopathic thrombocytopenia, a positive test was found on the patients' platelets in 55 out of 80 cases. The indirect immunofluorescence test with the patient's serum was positive in 28 out of 80 cases. In thrombocytopenia accompanied by other autoimmune diseases, the direct test was positive in 35 out of 39; the indirect test in 13 out of 39 patients. These results are similar to those obtained with other techniques recently described in the literature. By using FITC-labelled monospecific anti-immunoglobulin reagents, immunochemical characterization of the autoantibodies was possible. Of the 90 patients with a positive direct test, 80 were analysed in this way. Of these, platelet-bound IgG was detected in 76. This was accompanied by IgM in 17 and IgA in three. In four, only IgM was found. The IgG subclasses of platelet-bound IgG were studied in 48 patients. Mainly IgG1 was found (45), often together with IgG3 (22) or IgG4 (14), but rarely with IgG2 (1). Only a few patients had just IgG3 (2) or IgG4 (1) on their platelets. Platelet-bound autoantibodies could often be eluted (41 out of 86 cases) with ether and were demonstrable in the eluate with the indirect immunofluorescence test. This test was usually positive when the direct test was strongly positive (40 out of 44) and usually negative when it was weakly positive (1 out of 24) or negative (1 out of 18).

The immunochemical characteristics of granulocyte allo- and autoantibodies can easily be studied with indirect immunofluorescence techniques. The correlation of these properties with the serological behavior of the antibodies and the clinical condition in which they were found was studied. The investigations included the determination of immunoglobulin class, subclass, light-chain composition, as well as optimal temperature of activity. Complement fixation in vitro was also investigated. Striking differences were found between IgG and IgM antibodies in their serological properties, especially in the granulocytotoxicity and leukoagglutination tests at different temperatures. Differences in light-chain and IgG subclass compositions of the antibodies did not seem to play a role. In studying the distribution of granulocyte-specific antigens on normal granulocytes and granulocytes in all stages of development, it was found that, while NA1, NA2 and ND, are present in the same quantity on the cells of an individual, the NB1 antigen is unequally distributed, both over ripe granulocytes and granulocyte precursors.

Neutropenia may be caused by neutrophil autoantibodies. The detection of such antibodies has always been difficult. Recently, we developed a sensitive indirect immunofluorescence technique applicable to granulocytes which proved to be of value in the detection of granulocyte alloantibodies. We have now used this method to investigate the serum and cells of 29 patients with idiopathic or secondary neutropenia. In four patients neutrophil-antigen-specific autoantibodies were detected in the serum and the patient's own granulocytes showed direct fluorescence. Furthermore, differences in the immunoglobulin composition and the temperature of optimal activity of the autoantibodies were found. Direct fluorescence was also demonstrated with the granulocytes from five other neutropenic patients and with the granulocytes from five non-neutropenic patients mainly with infectious disease. However, no granulocyte antibodies could be eluted or shown to be present in the serum. This indicates that a positive direct granulocyte fluorescence test may not be considered as proof that autoantibodies are present.

The serum of a 39-year-old male under long-term platelet transfusion therapy for hypoplastic anemia with thrombocytopenia was investigated for the presence of leukocyte and platelet antibodies after the patients had received platelet concentrates from more than 700 random donors. The serological studies of his serum revealed: (1) the absence of platelet-reactive antibodies; (2) the absence of agglutinating or cytotoxic antibodies against leukocytes; (3) the presence of at least two granulocyte-specific antibodies, one with the specificity anti-NA2 and the other with an undefined specificity, both only detectable by indirect immunofluorescence, and (4) the presence of cytotoxic-negative fluorescence-positive peripheral-blood-mononuclear-cell-reactive antibodies, not directed against HLA-A, B or C antigens. The significance and implications of these findings are briefly discussed.

Anti-Zwa antibodies can induce a Glanzmann-like platelet dysfunction in normal Zwa-positive platelets. Although a normal shape change and release reaction was always recorded in response to adenosine diphosphate and collagen, the aggregation on these inducers could be completely inhibited by anti-Zwa antibodies. Analogous to the platelets from patients with Glanzmann's thrombasthenia, normal platelets sensitized with anti-Zwa did not associate with fibrinogen upon exposure to adenosine diphosphate. Our data indicate that the binding site for fibrinogen is closely associated with the Zwa-antigenic determinants on the platelet membrane glycoproteins and thus with glycoprotein IIIa which is known to carry Zwa.

Several variables that may affect accurate measurement of platelet-associated IgG (PA-IgG) were studied using a radioimmunoassay of the consumption type. The amount of PA-IgG of washed, unfixed normal donor platelets was 1.0 +/- 0.9 fg IgG/platelet (mean +/- 2 SD). Upon storage of washed platelets in a buffer containing EDTA, the amount decreased significantly to 0.2 +/- 0.2 fg IgG/platelet. Simultaneously, an increase in modal platelet volume was observed. Similar results were obtained when platelets were fixed with paraformaldehyde (PFA). We postulate that this decrease in PA-IgG is caused by the release of plasma IgG entrapped by the surface-connected canicular system of the platelet, when the platelets swell during storage in EDTA or fixation with PFA. This presence of varying amounts of entrapped plasma IgG may cause the wide discrepancies in PA-IgG found in normal donor platelets as well as platelets from ITP patients by other investigators. A good quantification of platelet-bound alloantibodies was possible with our assay when platelets were routinely fixed to diminish the amount of nonspecific PA-IgG. This was demonstrated with different anti-Zwa (= anti-PlA1), anti-Baka and anti-HLA sera. We also observed that fragments of platelets as well as fragments of cells of other types can cause aspecifically increased Pa-IgG values and can thus interfere with the proper measurement of platelet-bound antibodies in all kinds of immunoassays in general.

For the detection of auto-antibodies on the platelets and in the serum of patients with idiopathic thrombocytopenic purpura, three new techniques have recently been developed: the quantitative antiglobulin consumption assay (QACA), the platelet radioactive anti-globulin test (PRAT) and the platelet suspension immunofluorescence test (PSIFT). The results obtained by various investigators with these techniques differ considerably. We, therefore, studied the sensitivity of the three methods. This was done by testing platelet-reactive allo-antibodies (anti-Zwa, anti-HLA-A2) and auto-antibodies in titration. The results show that the PSIFT is the most sensitive technique, closely followed by the PRAT. The QACA was found to be much less sensitive than the other two methods. This suggests that a positive result in the QACA and a negative result in the PSIFT and/or PRAT cannot be attributed to the presence of platelet auto-antibodies.

Summary A severe haemolytic reaction after transfusion of apparently compatible blood was observed, without isoantibodies being detectable in the patient's serum. From survival studies with Wr tagged red cells could be concluded that the abnormal destruction of the transfused red cells is of immunological origin. Resume Une reaction hemolytique grave apres transfusion de sang apparemment compatible a ete observee, sans que l'on puisse mettre en evidence des iso-anticorps dans le serum du patient. Des etudev pratiquees avec des erythrocytes marques au 51Cr, on peut conclure que la destruction anormale des erythrocytes transfuses a une origine immunologique. Zusammenfassung EB wird eine schwere hamolytische Transfusionsreaktion beschrieben, bei welcher im Empfangerserum Isoantikorper nicht nachgewiesen werden konnten. Auf Grund von Uberlebenszeitbestimmungen mit 51Cr markierten Zellen wurde gescblossen, das der rasche Abbau der Spendersellen trotzdem auf eincm Immunmechanismus beruht.

Summary. Serum samples from 116 women taken at different times during and after the first pregnancy were tested for the presence of cytotoxic lcucocytc antibodies against the husbands' Ieucocytes. Antibodies were found in the serum of 15 women, 11 times during the pregnancy, and as early as at 24 weeks of gestation. All antibodies examined were IgG, even the earliest detectable antibodies. All sera contained antibodies against antigens common to granulocytes and lymphocytes; four sera also contained granulocytespecific antibodies and none lymphocyte-specific antibodies. No antibodies were demonstrable in any of the cord-blood and no granulo- or lympliopenia was seen in any of the children.

Summary Evidence was found that human lymphocytes and fibroblasts derived from human skin, sensitized by cytotoxic leucocyte iso-antibodies and complement adhere to the surface of indicator cells. The reaction mechanism showed the characteristics of the immune adherence (I.A.) phenomenon. Experiments made clear that in the I.A. assay the same antibody is detected as in the cytotoxic antibody test, based on dye exclusion. The sensitivity of the I.A. assay was shown to surpass slightly that of the cytotoxic antibody test.

Summary. Cytochemical and serological properties of the erythroid cells of a patient with a congenital dyserythropoietic anaemia, type III, are analysed. Especially striking was the great spread in DNA content per normoblast nucleus. The red cells of the patient were much more readily agglutinated by anti-i sera than normal cells. Absorption of anti-I and anti-i antibodies by the patient's red cells was increased, and there was increased lysis by these antibodies.

Summary The serological and immunochemical properties of antibodies with anti-I specificity found in the serum of a patient with high-titre cold auto-agglutinins are described. Survival studies with 51Cr tagged I and i red cells showed that i cells survived normally and that the I cells had a diminished survival time. The agglutinating and haemolysing antibodies were both β2M globulins. The nature of the positive reaction between the patient's red cells and various anti-human globulin reagents was studied. Anti-C' sera contain at least three different antibodies responsible for the agglutination of cold haemolytic anaemia cells, cells sensitized by naturally occurring incomplete cold antibodies and cells sensitized by Lewis antibodies. Resume Les proprietes serologiques et immunochimiques des anticorps de specificite anti-I trouves chez un patient presentant un titre eleve d'auto-agglutinines du type froid sont discutees. Les etudes de survie d'erythrocytes I et i marques au 51Cr ont demontre que les erythrocytes i survivaient normalement, tandis que les erythrocytes I avaient un temps de survie diminue, Les anticorps hemolysants et agglutinarits etaient tous deux des globulines β2M. La nature de la reaction positive entre les erythrocytes du patient et differents reactifs antiglobuline humaine a ete etudiee. Les serums anti-C' contiennent au moins trois anticorps differents respoilsables de l'agglutination des erythrocytes d'anemie hemolytique du type froid, des erythrocytes sensibilises par des anticorps froirls «naturrls» incomplets et des erythrocytes sensibilises par des anticorps anti-Lewis. Zusammenfassung Die serologischen und immunochemischen Eigenschaften der im Serum einer Patientin mit hochtitrigen Kalteautoagglutininen vorhandenen Antikorpern mit Anti-I-Spezifitat werden beschrieben. Die Uberlebenszeitbestimmung mit Cr-51-markierten I- und i-Zellen zeigte, das die letzteren normal uberlebten, wahrend die ersteren eine verkurzte uberlebenszeit aufwiesen. Die agglutinierenden wie auch die hamolysierenden Antikorper erwiesen sieh als β2-M-Globuline. Der Mechanismus der positiven Reaktion der Patientinnenerythrozyten mit verschiedenen Anti-Human-Globulinseren wurde untersucht. Anti-C'-Seren enthalten mindestens drei verschiedene Antikorper, welche bei der Agglutination van Erythrozyten van Patienten mit erworbener hamolytischer Anamie vom Kaltetypus beteiligt sind, Erythrozyten mit naturlichen inkompletten Kalteantikorpern sensibilisiert sowie Lewis-antikorpersensibilisierte Zellen.

The serum of 100 patients with myasthenia gravis and 441 of their first-degree relatives was studied for the presence of autoantibodies against several antigens. Antibodies to skeletal muscle were present in 22% of the patients and in 2% of the relatives. Both these frequencies were significantly higher than those in matched control subjects. Also, antinuclear antibodies were present more often both in the patients and in the relatives. Typing for HL-A antigens had shown a positive correlation between HL-A 8 and myasthenia gravis which was significantly higher in women than in men. Antibodies to skeletal muscle and thymomas were found to be much rarer in HL-A 8-positive patients than in HL-A 8-negative patients; HL-A 8-positive patients acquired the disease at an earlier age.HL-A 2-positive patients more often had thymomas and antibodies to skeletal muscle than HL-A 2-negative patients; HL-A 2-positive patients acquired myasthenia gravis at a later age.The fact that the clinical aspects of the HL-A 8-negative and HL-A 2-positive patients were different from those of the HL-A 8-positive and HL-A 2-negative patients justifies the hypothesis that there are two forms of myasthenia gravis.

The immunofluorescence test on paraformaldehyde-fixed cells was used for the detection of antibodies bound to the platelets and granulocytes and present in the sera of 24 patients with Evans syndrome and a further 29 patients with both idiopathic thrombocytopenia (ITP) and idiopathic neutropenia (INP), but without autoimmune haemolytic anaemia (AIHA). The direct immunofluorescence test on platelets and/or on granulocytes was positive in all patients with a cytopenia, but the sera of only 17 patients with the Evans syndrome and 15 of the other patients contained platelet- or granulocyte-specific autoantibodies. From absorption and elution experiments, it appeared that the autoantibodies were directed against antigens specific for the various peripheral blood cells, i.e. erythrocytes, platelets and granulocytes and that they were not cross-reacting.

There are 3 varieties of idiopathic thrombocytopenic purpura (ITP) in children: the acute, the chronic and the recurrent forms. We have tested the platelets from 80 children with these forms of ITP by an immunofluorescence test. Platelet-bound immunoglobulins were detected in 19/22 acute, 41/48 chronic and 5/10 recurrent cases of ITP, i.e. in 65 of a total of 80 cases. A significance differences between acute and chronic ITP was found in the observation that only IgM was bound to the platelets of 7/16 patients with acute ITP. In all cases of chronic ITP with a positive direct immunofluorescence test, IgG was present on the platelet membrane. It is assumed because some patients with chronic ITP do not have elavated platelet-bound immunoglobulins, and show refractoriness to immunosuppressive therapy, that in these patients the disorder is not of a humoral autoimmune nature.

In view of the importance of the species of animals used fat the preparation of specific heteroantisera against human blood cells, comparative immunization experiments Were carried out in goats, rabbit, chickens, guinea pigs, and rats, using lysates leukocytes, granulocytes, lymphocytes, monocytes, and thrombocytes. The antisera obtained were tested by the indirect immunofluorescence technique. The strongest antisera were obtained in goats, but these antisera required extensive absorptions to make them specific for the immunizing cell, whereas rabbit antisera needed only few absorptions. Antisera from both of these animals could be rendered specific for lymphocytes, granulocytes and platelets. In the guinea pig only immunization with lymphocytes and granulocytes resulted in useful antisera, which were all specific for granulocytes after appropriate absorptions. Antisera obtained from rats could be rendered specific for granulocytes and lymphocytes by absorption. Antibodies against platelets could not be detected in any of the rat antisera. Chickens, however, produced strong anti-platelet anti-bodies. A specific anti-granulocyte antiserum could also be obtained in these animals. Specific anti-monocyte antisera could not be prepared from any of the animal sera. Immunizations with leukocyte lysate resulted in granulocyte-specific antisera in most animals but not in chickens. All antisera produced in chickens were specific for platelets. It can be concluded from these results that animals of different species may react very differently to the antigens of some human blood cells.