The role of alloimmunization in platelet survival studies

HLA antigens on platelet membranes. In vitro and in vivo studies

In order to determine whether HLA-A,B antigens of platelets are integral membrane constituents or rather represent adsorbed plasma proteins, their presence in plasma and their adsorbability onto platelet membranes was studied by in vitro and in vivo experiments. The amount of HLA antigens was quantitated by inhibition of lymphocytotoxicity (LCT) and by enzyme-linked immunosorbent assay (ELISA) using operationally monospecific polyclonal HLA antibodies or murine HLA-specific monoclonal antibodies, respectively. We found that in 11 out of 13 HLA-A2 and in 9 out of 10 HLA-B13 experiments, platelets from antigen-negative donors pretreated with plasma from the same number of antigen-positive donors inhibited LCT to the same extent as platelets from antigen-positive donors. Nevertheless, the in vitro adsorbed HLA antigens onto antigen-negative platelets were, unlike those on antigen-positive platelets or in plasma, not reactive with monoclonal antibodies as quantitated by ELISA. Similarly, infusion of HLA-A2-negative platelets from single donors into 3 HLA-A2-positive, thrombocytopenic patients with bone marrow failure led to a good platelet increment, but did not convert the HLA type of donor platelets, neither at 2 h nor at 18 h posttransfusion. On the basis of these results, we conclude that soluble HLA antigens can be taken up by human platelets from plasma in small amounts. However, the major portion of HLA antigens appears to be integral membrane constituents.

Immune thrombocytopenic purpura: autoimmune or immune complex disease?

To assess the pathogenic role of circulating immune complexes (CIC) in idiopathic thrombocytopenic purpura (ITP), 39 patients with ITP were compared to 17 patients with other forms of thrombocytopenia (hypersplenism (N = 12), impaired thrombopoiesis (3), thrombocytopenia of unknown origin (2)) and six nonthrombocytopenic subjects. In all patients, platelet mean life span (MLS), platelet associated IgG (PAIgG), as well as circulating anti-platelet antibodies and C1q binding activities were determined. In most cases, immune complex solubilization capacity (ICSC) and immune complex precipitation inhibition capacity (ICPIC) of sera were also assessed. All patients with ITP had a reduced platelet MLS, but PAIgG was elevated in only 16 out of 24 patients with chronic ITP, in six out of 10 patients with acute ITP and in four out of five patients with secondary ITP. In the group of patients with thrombocytopenia due to splenomegaly, seven out of 12 patients had elevated PAIgG while the platelet MLS was only slightly reduced. Of the 39 patients with ITP only one with secondary ITP had C1q binding material in his serum, as opposed to six out of 12 thrombocytopenic patients with splenomegaly. Whereas only three patients with ITP had abnormal immune-complex modulating capacities, such deviations were found in seven out of 12 patients with thrombocytopenia due to splenomegaly. We conclude that our data render the role of CIC in the pathogenesis of ITP very questionable.

Platelet survival determined with 51Cr versus 111In

Optimal labelling conditions of human platelets with 111In-oxine were determined in vitro. Based on this optimized technique, platelet mean life span (MLS) and platelet sequestration site were comparatively evaluated in 79 patients with two labels, 51Cr (n = 26) and 111In (n = 53). Patients were subgrouped according to clinical criteria as autoimmune thrombocytopenic purpura (AITP) (group 1; n = 49), hypersplenism (2; n = 12), impaired thrombopoiesis (3; n = 3), unclassified thrombocytopenia (4; n = 6), and nonthrombocytopenic patients (5; n = 9). In patients with AITP and hypersplenism the mean values for the MLS determined either with 51Cr or with 111In were lowered but the difference was not statistically significant, neither for group 1 (18.6 h vs 17.3 h; P greater than 0.2) nor for group 2 (94.7 vs 122.3 h; P greater than 0.2). The correlation between MLS and platelet counts in patients with AITP was significant for both labels (P less than 0.001). The 15 min recovery tended to be higher with 111In in all groups, but the difference was significant (P less than 0.05) only for group 1. The sequestration sites were similar with both labels. We conclude that, contrary to previous reports, platelet survival studies yield similar results with both the 51Cr and 111In methods. Due to its distinct advantages 111In is the label of choice for investigation of platelet kinetics.

The value of platelet-associated IgG in predicting the efficacy of splenectomy in autoimmune thrombocytopenia

In a prospective study, the hypothesis of whether the quantitative determination of platelet-associated IgG (PAIgG) in patients with chronic autoimmune thrombocytopenic purpura (ATP) can predict the efficacy of splenectomy, was investigated. PAIgG levels were repeatedly determined in 16 patients with definite ATP pre- and postsplenectomy, and related to platelet counts and platelet mean life span. It was found that patients with an immediate remission after splenectomy tended to have lower PAIgG levels (less than 6%) than failures, but this difference was not statistically significant. We conclude that PAIgG is of limited value for the prediction of the efficacy of splenectomy in ATP.

Platelet associated IgG, platelet survival, and platelet sequestration in thrombocytopenic states

Platelet-associated IgG (PAIgG), platelet mean life span (MLS), and platelet sequestration sites were studied in 69 patients with immune (ITP) and presumed nonimmune thrombocytopenias (NTP). A shortened MLS was associated with elevated PAIgG (N = 46), and with normal PAIgG (N = 15). Four patients had a normal MLS, but elevated PAIgG, four patients were normal for both parameters. The highest PAIgG values occurred in ITP patients with a very short MLS. Nine NTP patients had also elevated PAIgG, but a normal or slightly shortened MLS. There was a significant double log correlation between PAIgG and MLS for ITP, but not for NTP patients. Judged from the coefficient of determination, only 10% of PAIgG were directly related to a shortened MLS. 70% of patients (N = 63) had exclusively splenic and 30% hepatosplenic sequestration. PAIgG was elevated in 29/44 patients with splenic (66%) and in 16/19 patients with hepatosplenic sequestration (84%). In ITP, PAIgG-positive cases were observed in 69% of splenic v 82% of hepatosplenic sequestration, while in NTP the corresponding figures were 6/11 and 2/2. No significant correlation between PAIgG and either sequestration type was demonstrable. We conclude that in immunologically mediated thrombocytopenia only a small portion of PAIgG accounts for a decreased MLS, and that the concentration of PAIgG per se does not determine the platelet sequestration type.

Kinetics and sites of destruction of 111Indium-oxine-labeled platelets in idiopathic thrombocytopenic purpura: a quantitative study

Kinetics and quantification of the sites of destruction of 111-Indium-oxine-labeled autologous platelets were investigated in eight patients with idiopathic thrombocytopenic purpura. The mean platelet count was 17 +/- 9 X 10(9)/liter; platelets were separated by differential centrifugation and labeled with 5.6 +/- 2.5 MBq 111In. Whole body and organ 111In-platelet distribution was quantitated with a scintillation camera and a computer-assisted imaging system acquisition matrix. Areas of interest were selected with the computer and organ 111In-radioactivity expressed as a percentage of whole body activity. Mean platelet survival was 49.5 +/- 29.6 hr and the survival curves were exponential. Equilibrium percentage organ 111In-radioactivity was (normal values in parentheses): spleen 33.7 +/- 8.8 (31.1 +/- 10.2); liver 16.1 +/- 9.5 (13.1 +/- 1.3); thorax 22.8 +/- 3.7 (28.2 +/- 5.6). Percentage organ 111In-activity at the time when labeled platelets had disappeared from the circulation was: spleen 44.5 +/- 16.4 (40 +/- 16); liver 16.0 +/- 11.5 (32.4 +/- 7.2); thorax 19.7 +/- 6.0 (17.7 +/- 10.3). Thorax activity corresponds to bone marrow radioactivity. Three patterns of platelet sequestration were evident. Three patients had mainly splenic sequestration, two mainly hepatic sequestration, and three diffuse reticuloendothelial system sequestration with a major component of platelets destroyed in the bone marrow. Splenectomy was performed in two patients. The pattern of 111In-platelet sequestration was not predictive of response of glucocorticoid therapy or indicative of the necessity for splenectomy. Quantitative 111In-labeled autologous platelet kinetic studies provide a new tool for the investigation of platelet disorders.U