Gestational thrombocytopenia and immune thrombocytopenias in pregnancy.

Appropriate management of thrombocytopenia in the pregnant patient is important for the well-being of both mother and fetus. The healthy-appearing mother with mild thrombocytopenia may have either gestational benign thrombocytopenia, which does not produce fetal thrombocytopenia, or immune-mediated thrombocytopenia, which can produce fetal thrombocytopenia. These two types of pregnancy-associated thrombocytopenias can be differentiated. Gestational benign thrombocytopenia is initially discovered during pregnancy, and in these patients a reliable test for antiplatelet antibody is usually negative. Conversely, patients with immune-mediated thrombocytopenia may have a history of thrombocytopenia before the pregnancy, and these patients usually have a detectable antiplatelet antibody. The pregnancy patient who presents with a normal platelet count and a history of neonatal alloimmune thrombocytopenia in a prior pregnancy or with a history of an infant of a close relative with NAT must be carefully monitored. Antiplatelet antibody assays performed on mother's and baby's blood will help determine if an antiplatelet antibody is present in maternal plasma, if the antibody reacts with the baby's platelets, and (with appropriate typing plasma) the antigenic specificity of the maternal and fetal platelets. In addition, antigenic typing of the father's platelets will help determine the risk of NAT in the current pregnancy. If a fetus is at risk for severe immune-mediated thrombocytopenia from either an autoantibody or an alloantibody, the fetal platelet count should be measured, if possible, from blood obtained by umbilical cord puncture. If the fetal platelet count is less than 50,000/microL or cannot be measured but is thought to have a high probability of being less than 50,000/microL, strong consideration should be given to a cesarean delivery.

p53 transactivity during in vitro osteoblast differentiation in a rat osteosarcoma cell line.

We previously demonstrated a correlation between wild-type p53 expression and appearance of osteoblastic-specific differentiation characteristics, as evidenced by basal osteocalcin gene expression in a mouse osteosarcoma tumor. The study reported here further explored the possibility of p53's having a distinct transcription-activating role in bone differentiation, in addition to its proposed role in G1 arrest and apoptosis. ROS17/2.3 osteoblastic osteosarcoma cells were stably transfected with a plasmid containing wild-type p53 binding sequences fused to the chloramphenicol acetyltransferase reporter gene. These cells were used to determine the transactivating role of p53 in regulation of osteocalcin gene expression. We chose two conditions under which osteocalcin expression is known to be upregulated: exposure of osteoblastic cells to differentiation-promoting medium and to vitamin D3. Exposure of the transfected cells to differentiation-promoting medium produced an increase in p53 transactivating activity correlating with the appearance of osteocalcin expression after about 1 wk. Vitamin D3 treatment resulted in upregulation of osteocalcin activity without a corresponding change in p53 transactivation activity or expression. In separate experiments, we tested whether changes in osteocalcin expression accompanied changes in p53 activity under conditions of downregulation of cell proliferation mediated by inhibition of DNA synthesis. Hydroxyurea treatment was used to inhibit DNA synthesis and produce growth arrest in osteoblastic cells. Inhibition of osteoblast cell proliferation was associated with a fourfold increase in p53 transactivating activity and a transient increase in osteocalcin steady-state expression. These results demonstrated a close relationship between p53 and osteocalcin and suggested a regulatory role for wild-type p53 in the control of basal osteocalcin gene expression in osteoblasts.

Staphylococcal protein A binding to canine IgG and IgM.

Staphylococcal protein A (SpA) binds with high affinity to immunoglobulins from many species, making it a useful reagent for immunoassays and immunoglobulin purification procedures. However, its use is limited by poor reactivity with some immunoglobulin subclasses including human IgG3 and murine IgG1. Reports of SpA's reactivity with canine immunoglobulins have been inconsistent. Because the most recent reports indicated a much greater reactivity of purified SpA with canine immunoglobulins than was suggested by our observations, we quantitatively reevaluated the binding of canine IgG and IgM to Cowan I strain SpA under the conditions of use in our laboratory. IgG and IgM were purified from pooled normal canine plasma by affinity chromatography with heavy chain specific polyclonal anti-IgG and anti-IgM antibodies. The purified IgG and IgM were assessed for SpA reactivity by affinity chromatography using a SpA-agarose column. The relative proportions of total chromatographed IgG or IgM in the flow-through (SpA-nonbindable) and eluate (SpA-bindable) fractions were determined by absorbance at 280 nm. The IgG and IgM in each immunoglobulin fraction were also nonspecifically adsorbed to microtitration plates and tested for reactivity with 125I-SpA using a solid phase immunoradiometric assay (IRMA). Approximately 18% of the affinity purified canine IgG and 33% of the affinity purified canine IgM did not bind to the SpA affinity column and were also unreactive with 125I-SpA using the IRMA. A second approach using a different polyclonal antibody to canine IgG yielded similar results: about 21% of the purified IgG was unreactive with SpA. Incomplete reactivity of SpA with canine IgG and IgM limits the usefulness of SpA in canine immunologic procedures.

Thrombin is a distal mediator of lipopolysaccharide-induced liver injury in the rat.

Previous results demonstrated that rats given Escherichia coli lipopolysaccharide (LPS; 4 mg/kg, i.v.) experience hepatocellular necrosis that begins within 4 hr and that prior treatment with anticoagulants (e.g., heparin) which target thrombin prevents the liver injury. In this study, hepatocellular injury, as marked by increased plasma alanine aminotransferase (ALT) activity and histologic changes, was prevented when heparin or hirudin was administered to rats shortly before the onset of injury. These results suggest that thrombin is a critical mediator that acts distally in the series of inflammatory events that culminates in hepatocellular damage. To explore further this hypothesis, livers isolated from rats 2 hr after LPS administration were perfused with various media. Perfusion of livers with medium comprising diluted blood from heparin-treated donors resulted in no release of ALT activity. By contrast, perfusion with similar medium anticoagulated with ancrod, which prevents clotting by depleting fibrinogen but does not inhibit thrombin, resulted in hepatocellular injury evidenced as a time-dependent appearance of ALT activity in the medium. Moreover, when livers from rats treated 2 hr previously with LPS were perfused with buffer to which thrombin had been added, injury resulted. No injury resulted when thrombin was omitted from the buffer or when livers from saline-treated rats were used. These results indicate that thrombin is a critical and distal mediator of LPS-induced liver damage and contributes to hepatocellular injury through a mechanism that is independent of clot formation. Furthermore, inflammatory events triggered by LPS exposure are a prerequisite for thrombin-induced injury.

Type I Glanzmann's thrombasthenia in a Great Pyrenees dog.

An 8-month-old female Great Pyrenees dog with chronic epistaxis and a history of gingival bleeding during shedding of deciduous teeth was evaluated for platelet function. Platelet morphology was normal at both the light and electron microscopic level. Platelet number and mean platelet volume were also normal. Platelet aggregation responses to adenosine diphosphate, collagen, platelet activating factor, and thrombin were markedly reduced, although shape change responses were normal. Clot retraction was markedly impaired. Monoclonal antibody (MoAb) Y2/51, a murine anti-human platelet beta 3 antibody that cross-reacts with canine platelet beta 3, and MoAb 5G11, a murine anti-dog platelet alpha IIb beta 3 antibody, bound minimally to affected dog platelets, as demonstrated by flow cytometry. Binding of MoAb Y2/51 was not detectable by immunoblot. MoAb CAP1, a murine anti-dog fibrinogen receptor-induced binding site antibody, failed to bind to affected dog platelets, as demonstrated by flow cytometry. A reduction in glycoproteins alpha IIb and beta 3 was demonstrated by two-dimensional protein electrophoresis. This is the first reported case of type I Glanzmann's thrombasthenia in the dog that closely resembles the clinical syndrome and the platelet morphology described in type I Glanzmann's thrombasthenia of human beings.

The thrombin inhibitor, hirudin, attenuates lipopolysaccharide-induced liver injury in the rat.

The administration of gram-negative bacterial lipopolysaccharide (LPS) to rats results in hepatic parenchymal cell injury within 6 hr. The coagulation system is critical to the pathogenesis, but previously reported results suggested that its critical role is independent of insoluble clot formation and that thrombin may be a key mediator of liver injury. To test the hypothesis that thrombin is involved in LPS-induced liver injury, animals were treated with the selective thrombin inhibitor, hirudin. The hirudin treatment regimen effectively inhibited thrombin, as evidenced by prolonged activated partial thromboplastin time and by maintenance of plasma fibrinogen concentrations in LPS-treated rats. Treatment with hirudin prevented LPS-induced liver injury, assessed by plasma alanine aminotransferase activity and histological evidence of hepatocellular necrosis. Previous studies have shown that LPS exposure results in the accumulation of neutrophils and platelets within the liver and that both of these cell types are critical for the development of LPS-induced liver injury. Hirudin attenuated in part the decrease in blood platelet concentration that accompanied LPS administration, but did not alter hepatic platelet or neutrophil accumulation. These results support the hypothesis that thrombin is required for hepatic injury from LPS exposure, but that it does not act by promoting the accumulation of platelets or neutrophils within the liver.

Biochemical and functional characterization of a new murine monoclonal antibody against human platelet glycoprotein IIIa.

A new murine monoclonal antibody, MDP-1, specific for human platelet glycoprotein IIIa has been produced and characterized. Following SDS-polyacrylamide gel electrophoresis, MDP-1 reacted with a 94kDa protein immobilized on a nitrocellulose membrane. Upon reduction, MDP-1 no longer bound to the 94kDa protein indicating an epitope requiring at least one disulfide bond. On crossed immunoelectrophoresis MDP-1 reacted to the same peak as the GP IIb-IIIa complex-specific antibody AP-2. After dissociation of the GP IIb-IIIa complex with EDTA, AP-2 showed no reactivity while MDP-1 bound to a new peak that was broader and anodal to the original GP IIb-IIIa peak, consistent with GP IIIa. MDP-1 inhibited ADP and thrombin induced aggregation. In addition, MDP-1 inhibited ADP induced release of ATP, but did not inhibit thrombin stimulated ATP release. Following chymotrypsin digestion, MDP-1 bound to a cleaved GP IIIa protein (nonreduced M, = 122 kDa) consistent with opening of the major disulfide loop. A second cleavage resulted in a 63 kDa species that reacted with MDP-1. Scatchard analysis revealed 22 000 molecules of MDP-1 bound per platelet, and indicated a type of binding consistent with positive cooperativity. The antibody bound equally well to stimulated and unstimulated platelets. MDP-1 binding was inhibited by a polyclonal anti-PI(A1) antibody, but bound to platelets from a PI(A1) negative individual indicating a binding site close to but not identical to the PI(A1) epitope. In addition, MDP-1 binding was not inhibited by Arg-Gly-Asp-Ser (RGDS) suggesting that it is not directed to the RGD binding site on GP IIIa.

Anabolic-androgenic steroid abuse in weight lifters: evidence for activation of the hemostatic system.

Anabolic-androgenic steroid abuse has recently been linked with acute vascular events in athletes. To date, the relationship between steroid abuse and the potential for cardiovascular disease has been considered almost exclusively in terms of lipid metabolism. However, recent reports of thrombosis in androgen abusing athletes with no evidence of atherosclerosis suggests the hypothesis that thrombosis risk in such athletes could be mediated through androgen induced abnormalities of coagulation. To determine if anabolic-androgenic steroid abuse in weight lifters is associated with an activation of the hemostatic system we studied forty-nine weight lifters recruited through advertisements. History of androgen use or abstinence was confirmed via urine assays. Plasma was assayed for clotting and fibrinolytic activity by measuring thrombin/antithrombin complexes (TAT), prothrombin fragment 1 + 1 (F1 + 2), and D-dimers (D-di); markers of the endothelial based fibrinolytic components were assayed by measuring tissue plasminogen activator antigen (t-PA Ag) and its inhibitor (PAI-1); finally, the activity of antithrombin III, protein C, and protein S were measured. Abnormally high concentrations of TAT complexes were noted in 16% of our confirmed steroid using weight lifters compared to 6% of our confirmed nonusers (P = .01). Steroid users also demonstrated abnormally high concentrations of F1 + 2 and D-dimers when compared to nonusers (44 vs. 24%, P < .001, and 9 vs. 0%, respectively). Non-steroid users were more likely to have elevated levels of t-PA Ag and PAI-1 than our steroid using weight lifters (both P < .001). The activities of antithrombin III and protein S were more likely to be higher in users compared to nonusers (22 vs. 6%, P = .005; 19 vs. 0%, respectively). Some anabolic-androgenic steroid using weight lifters have an accelerated activation of their hemostatic system as evidence by increased generation of both thrombin and plasmin. These changes could reflect a thrombotic diatheses that may contribute to vascular occlusion reported in young athletes using these drugs. The predictive value of these coagulation abnormalities in terms of risk of thrombosis to individual steroid using weight lifters or the population as a whole remains to be studied.

Morphologic investigations of the guinea pig model of iron overload.

We have developed a guinea pig model of iron overload toxicity. Animals were administered intraperitoneal iron dextran 3 times a week to achieve total body iron load of 0.25, 0.5, 1.0, 1.5, and 2.0 g Fe/kg body weight in less than 30 days. Quantitation of tissue iron levels with atomic absorption indicated increased iron deposition in liver and heart of the iron-loaded guinea pigs (p < 0.001). Additionally, the iron-loaded pigs demonstrated decreased nuclear magnetic resonance spectroscopy T1 relaxation times in both liver and heart (p < 0.001). Serum iron, total body iron capacity, and transferrin saturation values were also determined in guinea pigs treated with 0.25, 0.5, and 1.0 g Fe/kg body weight. Serum iron and total iron-binding capacity were significantly increased at 0.5 and 1.0 g Fe/kg; transferrin saturation was elevated at 0.25 and 1.0 g Fe/kg. kg. Histologic examination of liver, heart, and bone marrow as well as ultrastructural studies on liver and heart confirmed increased iron deposition in treated animals. At the low iron dose level of 0.5 g Fe/kg, liver iron particles were primarily confined to Kupffer cells with minimal hepatocellular localization. Increased hepatocellular iron deposition was observed with larger doses of loaded iron. Myocardial iron was most prominent in interstitial cells of the epicardium, endocardium, myocardium, and coronary adipose tissue. Ultrastructurally, the presence of iron particles in perinuclear, membrane-bound structures (consistent with lysosomes) was confirmed using x-ray microanalysis. These morphological studies suggest that in this animal model siderosis of hepatic mononuclear phagocyte and myocardial interstitial cells may be the initial lesions leading to further biochemical and functional abnormalities. Correlation between tissue iron measurements and both light and electron microscopic changes, presented in this report, serve to introduce the iron-loaded guinea pig as a model for the study of iron-induced tissue damage.

Iron-induced myocardial and hepatic lysosomal abnormalities in the guinea pig.

The guinea pig model of iron overload, described in the preceding article, was used to investigate the mechanism of excess iron toxicity in hepatic and cardiac tissues. Effects of iron overload on both lysosomal membrane fragility and membrane peroxidation were studied. The free activity of selected myocardial and hepatic lysosomal enzymes, in addition to serum activity, was measured in guinea pigs treated with iron dextran (0.25, 0.5, 1.0, and 2.0 g Fe/kg body weight); controls received dextran. Levels of malondialdehyde were also determined in whole homogenates of heart and liver in animals loaded with 0.5 and 1.5 g Fe/kg of iron dextran. Results indicated that the free activity of hepatic glucosaminidase (p < 0.05) and beta-glucuronidase (p < 0.05) were significantly elevated at all levels of iron loading; hepatic acid phosphatase was increased at all but the lowest iron dose. Similarly, increased serum glucosaminidase activity was observed (p < 0.01) at all dose levels. When compared to pooled controls, the free activity of myocardial glucosaminidase was also elevated (p < 0.05) at all levels of loading. However, myocardial acid phosphatase was increased only at the highest iron dose (p < 0.01). Increased malondialdehyde was measured at the high iron dose (1.5 g Fe/kg) in whole homogenates of both heart and liver (p < 0.01). We conclude that iron loading in this model profoundly alters the stability of hepatic and myocardial lysosomal membranes; furthermore, changes in serum glucosaminidase activity may be reflective of modified tissue lysosomal properties. Elevated levels of malondialdehyde in whole homogenates suggest that iron-mediated lipid peroxidation may be responsible in part for enhanced lysosomal membrane fragility.

A radioimmune microfilter plate assay for the detection of anti-granulocyte antibodies.

A radioimmune assay (RIA) for IgG anti-granulocyte antibodies (AGA) was developed which utilized 96 well microfilter plates, required less than 50 microliters plasma, and yielded reproducible autologous determinations on patients with absolute granulocyte counts as low as 300 cells/microliter. Granulocytes were isolated by a single density ficoll-hypaque gradient procedure with red blood cell (RBC) removal by lysis and washing. Anti-granulocyte IgG coating the cells was detected by the addition of 125I labeled anti-human IgG. Six patients with autoimmune granulocytopenia and three multiply transfused patients with febrile transfusion reactions believed to be secondary to allogeneic granulocyte antibodies had increased granulocyte IgG. Plasma containing anti-a and/or anti-B titers greater than or equal to 512 gave positive results when tested against ABO incompatible granulocytes. This technique was capable of detecting antibodies which were reactive in granulocyte agglutination and immunofluorescence tests as demonstrated with known anti-NA-1, anti-NB-1, and anti-Mart sera. Comparison with granulocyte agglutination test (GAT) using clinical criteria for immune granulocytopenia demonstrated concordance with 19 negative samples and 8 positive samples, 1 false negative and 3 false positives were observed using the agglutination method. The multimeric and monomeric IgG fractions of G200 purified Cohn Fraction II were nonreactive by both RIA and GAT methods. However, granulocyte activation by phorbol myristate acetate (PMA) produced false positives in 5 out of 5 cases by GAT but 0 out of 4 by RIA. This RIA assay is not effected by nonimmune granulocyte activation and is more reproducible when compared to existing techniques.

Dose-dependent platelet stimulation and inhibition induced by anti-PIA1 IgG.

The PIA1 antibody produces several clinically distinct and severe thrombocytopenias. Investigations have demonstrated divergent effects on platelet function; prior reports demonstrated inhibition, while a conflicting publication showed platelet activation. We have resolved this conflict using anti-PIA1 IgG produced by a patient with posttransfusion purpura. Relatively low concentrations stimulated platelet aggregation and release of adenosine triphosphate (ATP) whereas high concentrations inhibited platelet function, producing a thrombasthenia-like state. The number of molecules of platelet-associated IgG necessary to initiate aggregation and ATP release (2,086 +/- 556) or produce maximum aggregation (23,420 +/- 3,706) or complete inhibition (63,582 +/- 2654) were measured with a quantitative radiometric assay for bound anti-PIA1. Preincubation of platelets with high concentrations of PIA1 antibody inhibited platelet aggregation with 10 mumol/L adenosine diphosphate and blocked 125I-labeled fibrinogen platelet binding. Platelet activation with nonfibrinogen dependent agonist, 1 U/ml thrombin, was not inhibited by this high concentration of PIA1 IgG. In conclusion, anti-PIAI IgG produces (1) stimulation of platelet aggregation and ATP release that is initiated with 2000 molecules IgG per platelet and is associated with an increase of 125I-fibrinogen binding; (2) conversely, inhibition of platelet aggregation is observed with maximum antibody binding, 63,000 molecules IgG per platelet, and is mediated via a blockade of fibrinogen binding.

Precise quantitation of PAIgG: a new radiometric microtechnique.

We report the development of a radiometric assay for platelet-bound IgG that is both sensitive and quantitative. The assay utilized 96-well millititer plates incorporating a 0.2 microns filter membrane in the bottom. A 125I-labeled monoclonal antihuman IgG, as a secondary antibody, detected the platelet-bound human IgG. Since 5 x 10(6) platelets were used for each assay, tests for platelet-bound IgG can be performed on persons with severe thrombocytopenia. For the detection of circulating antiplatelet alloantibodies, as little as 10 microliters of platelet-free plasma per assay is required. Antiplatelet IgG was quantitated by using anti-PIA1 antibody that was purified with affinity and elution and DEAE chromatography. This purified antiplatelet antibody was labeled with 125I and was used to determine the binding ratio of secondary antibody to primary antibody. Under our standard conditions, this ratio was found to be stable at approximately 0.35 over the sensitivity range of the assay. The assay can detect approximately 200 molecules of human IgG per platelet (0.1 ng of secondary antibody bound per 5 x 10(6) platelets). It has a linear range from 0 to 7,000 molecules per platelet. Quantitation of anti-PIA1 binding for platelets stored for up to 6 months under refrigeration showed no change in number of PIA1 binding sites. Clinical studies showed that 18 of 19 ITP patients had an increased number of IgG molecules per platelet as did patients with malignancy and drug-induced immune thrombocytopenia. Patients who had received multiple platelet transfusions had antiplatelet antibody in their plasma. Normal amounts of PAIgG were observed in platelets and plasma of patients with nonimmune thrombocytopenia. The advantages of this method are that it is: 1) a more precise quantitation of PAIgG via direct measurement of binding ratio with PIA1 antibody; 2) performed with small amounts of platelets and plasma; 3) both sensitive and specific; and 4) reliably reproducible with both fresh and stored platelets.

Absent platelet aggregation with normal fibrinogen binding in basset hound hereditary thrombopathy.

Platelets from dogs with Basset Hound Hereditary Thrombopathy (BHT) initially displayed a thrombasthenia-like aggregation defect but have been shown to have normal amounts of platelet membrane glycoproteins IIb and IIIa (GPIIb-IIIa), and therefore are more accurately described as thrombopathic. The presence of normal quantities of GPIIb-IIIa, however, did not rule out the possibility of a functionally abnormal glycoprotein complex which would be unable to bind radio-labeled fibrinogen. Therefore, fibrinogen binding in BHT platelets was evaluated. Platelets from BHT and normal dogs were activated with 1 x 10(-5) M ADP in the presence of 125I-fibrinogen and the surface-bound radioactivity was quantitated. The amount of fibrinogen bound by BHT dog platelets was not significantly different than that bound by normal dog platelets. Platelets from dogs with BHT bound 30,282 +/- 3,133 and normal dog platelets bound 31,664 +/- 2,772 molecules of fibrinogen per platelet. The quantitatively normal GPIIb-IIIa complex binds fibrinogen in normal amounts and does not seem to represent the abnormality responsible for the aggregation defect in BHT platelets. Therefore, other factors central to normal platelet function and related to platelet aggregation must be considered.

Energy expenditure of autoperfusing heart-lung preparation.

The autoperfused heart-lung preparation was developed as a method for extending the acceptable donor-to-recipient interval in clinical heart-lung transplantation. Metabolic substrate enhancement has been shown to be necessary for the survival and homeostasis of the functioning preparation. To define basic metabolic requirements and to determine the resting energy expenditure of the working canine heart-lung preparation, two groups were studied. Ten canine heart-lung blocks were placed in a normothermic autoperfusion circuit. In Group 1 (n = 5), a hyperalimentation solution of balanced substrate was infused (15% dextrose, 4.25% amino acids, 8 meq magnesium sulfate, 30 IU/dl insulin, and 10% lipids). In Group 2 (n = 5), no substrate was given. The preparations were ventilated with a mixture of room air and 5% CO2 at a rate of 4 breaths/min to maintain physiological pH. Myocardial function was assessed by cardiac output determinations and mixed venous gases. Pulmonary function was assessed with arterial blood gases. The oxygen consumption (VO2) and carbon dioxide production (VCO2) were measured with a Metabolic Cart, and the resting energy expenditure was calculated. The mean survival time for Group 1 was 360 minutes, and all preparations were terminated electively. The mean survival time for Group 2 was 219 +/- 43 minutes (p less than 0.01) with congestive heart failure as the common terminal event. All parameters of cardiac function and blood gases remained within physiological limits without significant differences between groups. The resting energy expenditure, a measure of metabolic rate, was 2.5 +/- 0.3 kcal/hr in Group 1 and 1.0 +/- 0.2 in Group 2 at termination (mean +/- SD) (p less than 0.01).(ABSTRACT TRUNCATED AT 250 WORDS)

Platelet antibody: review of detection methods.

The driving force behind development of in vitro methods for platelet antibodies is identification of plasma factors causing platelet destruction. Early methods relied on measurement of platelet activation. Current methods are more specific and use a purified antibody against immunoglobulin or complement, which is usually labeled with 125I or tagged with an enzyme or fluorescein. Comparisons of quantitation of platelet-associated IgG show wide variability between different methods. The disparate results can be related both to differences in binding of secondary antibodies to immunoglobulin in solution compared to immunoglobulins attached to platelets and to the improper assumption that the binding ratio between the secondary detecting and primary antiplatelet antibody is one. Most assays can 1) identify neonatal isoimmune thrombocytopenia and posttransfusion purpura, 2) help to differentiate between immune and nonimmune thrombocytopenias, 3) help to sort out the offending drug when drug-induced thrombocytopenia is suspected, and 4) identify platelet alloantibodies and potential platelet donors via a cross match assay for refractory patients. However, the advantages of quantitative assays over qualitative methods with respect to predictions of patients clinical course and response to different treatments remain to be investigated.

Aluminum-induced anemia.

Although many questions still remain unanswered, it is clear that aluminum causes a microcytic hypoproliferative anemia and is one factor responsible for worsening anemia in patients with end-stage renal disease. Time course studies in a rat model have shown that the anemia is preceded by microcytosis; this has not yet been examined in man. The exact mechanism of aluminum-induced anemia is unknown, however it appears to involve inhibition of heme synthesis, either by inhibition of enzyme activity or interference with iron incorporation or utilization. The interrelationship between aluminum and iron, zinc, lead, or other metals in this anemia is also unknown, as are the effects of aluminum on erythroid colony forming units. The role of parathyroid hormone on aluminum-induced anemia has not been examined. Presently treatment of aluminum-induced anemia involves removal of the source of the aluminum, although recent studies with desferrioxamine show promise. It is unclear, however, exactly how desferrioxamine improves this anemia. It is clear, however, that aluminum in the dialysate can cause clinical problems including anemia, and that these problems can be substantially reduced if not eliminated by water treatment.

Complications of heparin administration in normal individuals.

The incidence, severity, and pathogenic mechanism of heparin-associated thrombocytopenia with both bovine and porcine heparin administration were studied in forty normal males randomized to one of four treatment groups: beef lung heparin #1, beef lung heparin #2, porcine gut heparin, and saline control. All of the subjects receiving heparin developed a reversible increase in serum transaminases (SGOT, SGPT). However, other measurements of liver function were normal. Thirty-three percent of these heparinized normals had decreased platelet counts. The incidence of platelet count decrease was similar for both bovine and porcine heparins, but 4 of the 20 normals receiving bovine heparin had platelet counts less than 150,000/microliters. Immune pathogenesis was investigated by analyzing plasma from the volunteers for both platelet antibody and immune complexes. None of the men had increased platelet-associated IgG. Among the ten subjects with decreased platelet counts, IgG immune complexes were detected in three and C1q in seven. The heparin-associated thrombocytopenia appears not to be mediated by a platelet antibody. More probably it reflects a direct effect of the heparin on platelets.

Microcytic anemia secondary to intraperitoneal aluminum in normal and uremic rats.

Dialysis patients exposed to high aluminum (Al) dialysate develop a microcytic anemia which is reversed by deionization (DI) of the dialysate. Because DI removes substances in addition to Al which are known to cause anemia, these experiments were undertaken to determine if Al causes anemia and if the anemia of uremia can be enhanced by Al. Four groups of rats were studied: sham control (A) N = 6; uremic control (B) N = 6; Al-loaded non-uremic (C) N = 7; and Al-loaded uremic (D) N = 5. Aluminum treatment was 1 mg Al intraperitoneally daily for 6 weeks. Uremic rats (B+D) were 1 5/6 nephrectomized; non-uremic (A+C) were sham-operated. Blood samples (200 microliter) were obtained prior to (C1) and weekly during treatment (T1 to T6) and analyzed by Coulter Counter. No significant difference in hemoglobin (Hb), hematocrit (Hct), or mean cell volume (MCV) was noted at C1 X At T3, MCV of Al-treated rats (C+D) was significantly less than sham control (A) (55.1 +/- 0.5 and 53.0 +/- 0.8 vs. 60.8 +/- 1.5 mu3, P less than 0.05). At T6, MCV, Hb, and Hct of Al-loaded uremic rats (D) (49 +/- 0.5 mu3; 11.8 +/- 0.5 g/dl; 25.1 +/- 2%) were significantly less than both A (58.6 +/- 1.3 mu3; 16.1 +/- 0.4 g/dl; 44.8 +/- 0.3%) and B (58.7 +/- 1.4 mu3; 13.8 +/- 0.4 g/dl; 33.6 +/- 0.5%) (P less than 0.05) and MCV, Hb, and Hct of Al-loaded non-uremic rats (C) (51.7 +/- 1.7 mu3; 12.6 +/- 0.3 g/dl; 29.4 +/- 1.5%) was significantly less than A (P less than 0.05).(ABSTRACT TRUNCATED AT 250 WORDS)