Michael Sobel

Characterization of platelet binding of heparins and other glycosaminoglycans

The binding of glycosaminoglycans to intact washed human platelets was studied. The platelet binding of a 3H-labeled unfractionated heparin was saturable and reached equilibrium in 10-15 minutes. Heparin binding was specific: a 50-fold molar excess of an equivalent unlabeled heparin displaced up to 90% of labeled heparin, while chondroitin sulfate A and hyaluronic acid minimally displaced the binding of labeled heparin. Low molecular weight heparin fragments showed intermediate efficacy in displacing the binding of unfractionated [3H]heparin. Dextran sulfate (Mr 8,000, sulfation 17%) was as potent as unfractionated heparin in displacing binding, while neutral dextrans were ineffective. We observed that platelet activation by the calcium ionophore A23187 increased heparin binding by 2 to 3-fold, principally by enhancement of binding capacity not binding affinity. This process of heparin binding to the platelet surface may mediate some of the reported effects of heparin on platelet behavior.

Surgical management of heparin-associated thrombocytopenia: Strategies in the treatment of venous and arterial thromboembolism

We report the vascular surgical strategies and results in 13 patients with heparin-associated thrombocytopenia and describe useful in vitro techniques for the evaluation of anticoagulant therapy. Thirteen of 40 patients with heparin-associated thrombocytopenia had 18 cardiovascular procedures done to save life or limb. Greenfield filters were placed in eight patients to prevent pulmonary embolism. Eight patients had 10 arterial procedures, with alternative anticoagulation that used dextran or warfarin in five cases. In three cases iloprost, a derivative of prostacyclin and a potent platelet inhibitor, was infused intraoperatively and heparin was given. Both the use of alternative anticoagulants and platelet suppression by iloprost were clinically effective strategies. The concurrent measurement of plasma levels of beta-thromboglobulin and fibrinopeptide A in two patients confirmed that both approaches can successfully prevent activation of platelets and plasma coagulation during arterial surgery. One operative death occurred; all vascular reconstructions remained patent at 3 to 6 months. In two patients who received heparin alone for arterial surgery, both procedures resulted in thrombosis and limb loss. When major venous thromboembolism is complicated by heparin-associated thrombocytopenia, insertion of a Greenfield vena cava filter should be considered if there is significant risk of pulmonary embolism. When necessary, arterial surgery is feasible in patients with heparin-associated thrombocytopenia if alternative anticoagulation or adequate suppression of platelet reactivity can be achieved.

Catheter-directed thrombolysis following vena cava filtration for severe deep venous thrombosis

Massive deep venous thrombosis with marked venous outflow obstruction can result in limb loss or end-organ injury. Systemically administered drugs may not reach thrombi in therapeutic concentrations and surgical and thrombolytic strategies carry a small but real risk of pulmonary embolus—similar to the risks with anticoagulation alone. We therefore developed a strategy in which catheter-directed thrombolysis was used to deliver high concentrations of a plasminogen activator directly to the thrombus combined with placement of a downstream Greenfield filter to protect patients from pulmonary embolus. From 1984 to 1993 six patients were treated with this regimen. All had severe symptoms of less than 4 days duration. On radiologic evaluation four patients had large iliofemoral and/or inferior vena cava thrombosis, one had subclavian/innominate vein thrombosis, and one had transplant renal vein/iliofemoral/inferior vena cava thrombosis. A Greenfield filter was first placed downstream prior to imbedding an infusion catheter in the greatest mass of thrombus for subsequent infusion of urokinase (n=4) or streptokinase (n=2). In four patients the catheter traversed the Greenfield filter. All patients were given bolus lytic therapy followed by maintenance infusions ranging in duration from 24 hours to 12 days. Five patients remained on heparin simultaneously. Clot lysis was achieved in all patients with hemodynamic, symptomatic, and arteriographic improvement. There were no deaths, pulmonary emboli, or complications of filter placement. One patient had minor bleeding at the puncture site and another had catheter-related infection. At follow-up ranging from 8 months to 9 years all patients are asymptomatic with patent venous systems confirmed by duplex ultrasound imaging. Thus catheter-directed thrombolysis combined with vena cava filtration offers a safe and effective alternative to simple anticoagulation, surgery, or systemic thrombolysis. This approach may warrant more frequent use in patients with severe and disabling thrombosis of major deep veins.

Dextran 40 reduces heparin-mediated platelet aggregation

Using platelet aggregometry, we investigated the in vitro efficacy of dextran, mean MW 40,000, in reducing the heparin-mediated platelet aggregation seen with the syndrome of heparin-associated thrombocytopenia (HAT). Six patients with the clinical syndrome of HAT were studied who had a plasma factor which induced aggregation of normal platelets in the presence of heparin. At a final concentration of 2% (20 mg/ml) Dextran 40 reduced heparin-stimulated aggregation by 60.5 +/- 36% (SD) in mixtures of normal platelet-rich plasma/HAT patient plasma. In contrast, the same dextran concentration reduced ADP-stimulated aggregation of normal platelet-rich plasma by only 18.6 +/- 12%. This selective inhibition of heparin-mediated aggregation by dextran was highly significant (P less than 0.005, Wilcoxon rank sum). Whereas protamine abolished heparins anticoagulant and platelet aggregating activity, dextran had no effect on heparins anticoagulant properties. These data show that Dextran 40 selectively blocks heparin-mediated aggregation in vitro, perhaps by blocking heparin binding to the platelet membrane. This study suggests that dextran may be a clinically useful adjunct in the treatment of HAT, and describes a simple method for testing the efficacy of drugs in the treatment of HAT.

Nonpenetrating vascular injury to the subclavian artery

Blunt subclavian artery injury has been uncommonly reported in the literature. Recent encounter with three such injuries prompted us to review our experience over the past 10 years uncovering only one additional case. These four cases and a review of pertinent literature form the basis for this article. Key clinical issues include a high index of suspicion in patients sustaining major blunt deceleration and rotational or direct injuries to the neck, thorax, and/or upper extremities. Prompt diagnosis remains obscured by the presence of severe associated injuries, the treatment of which requires prioritization. Arteriography is invaluable to elucidate injury because prompt vascular control is dictated by various approaches depending on the location. Expeditious surgical repair is indicated to prevent complications of hemorrhage, pseudoaneurysm, thromboembolism, and/or arteriovenous fistula. Long-term results appear to be good with major morbidity related to associated neurologic, soft tissue, and bony injuries.

Structural characterization of heparin's binding domain for human platelets

The structural features of heparin that are involved in binding to human platelets were investigated by a competitive binding approach. A range of heparin-derived glycosaminoglycans (GAGs) with relatively defined structure were prepared by different methods of depolymerization of pharmaceutical heparin, followed by fractionation according to molecular weight and net charge. Competitive binding to platelets was dependent on molecular weight but not on the net charge of the GAGs. The method for depolymerization significantly affected the binding activity of the resulting GAG. Heparinase I and nitrous acid depolymerization produced GAGs with lower binding affinity for platelets than those GAGs derived from the treatment with periodate followed by alkali. The IC20 (concentration producing 20% inhibition of binding) was 0.05 microM for unfractionated heparin, 0.11 microM for a periodate treated GAG, and 2 microM for comparably sized GAGs (M(r) approximately 6,000-8,000) derived by heparinase I or nitrous acid treatment. Thus, the disaccharide units GlcNSO3-6S--IdoA-2S or GlcNSO3--IdoA-2S [(2-deoxy-2-sulfoamido-6-O-sulfo-alpha-D-glycopyranosyl)-(1- 4)-O-(2-O-sulfo-alpha-L-idopyranosyluronic acid) or (2-deoxy-2-sulfoamido-alpha-D-glycopyranosyl)-(1-4)-O-(2-O-s ulfo-alpha-L-idopyranosyluronic acid)] may be crucial elements for binding to the platelet, because these are known to be preserved during periodate/alkali treatment, but readily decomposed by heparinase I and nitrous acid. Understanding this structural specificity for platelet binding may be useful for the development of heparins with high or low platelet reactivity.

Arterial-Ureteral Fistula: Case Study with Review of Published Reports

Arterial-ureteral fistula, a rare cause of gross hematuria, may be associated with life-threatening hemorrhage if not rapidly diagnosed and treated. Recently, a patient at the Hunter H. McGuire Veterans Administration Medical Center developed an arterial-ureteral fistula at the site of the confluence of the external iliac artery and a superior mesenteric artery bypass graft. Review of the worlds literature revealed 31 additional reported cases of arterial-ureteral fistulas [1—31]. This current case is only the second one reported in which the diagnosis was made with arteriography. Several common features of arterial-ureteral fistulas were present in this case: a history of ureteral obstruction and urinary tract infections, upper urinary tract disease, and previous vascular surgery. The condition is usually associated with either prior upper urinary tract instrumentation or vascular surgery, and an antecedent period of intermittent hematuria, followed by life-threatening hematuria, is common. A high index of suspicion and early surgical intervention are required for successful management. The major surgical challenges are to establish unobstructed urinary drainage and restore vascular continuity. Exclusion of prosthetic material from potentially infected areas is mandatory.

Symptomatic abdominal aortic aneurysms in long-term survivors of cardiac transplantation

Herein we report the only two long-term survivors of cardiac transplantation who underwent successful repair of symptomatic abdominal aortic aneurysms since the advent of cyclosporine therapy in 1983. Review of the worlds literature indicates that the only two recorded cases of repair of symptomatic abdominal aortic aneurysms after cardiac transplantation occurred before the use of cyclosporine. The presentation and clinical course of our patients recently treated are presented, and perioperative care and immunosuppressive management are outlined. As the number of long-term survivors after cardiac transplantation increases, the incidence of other atherosclerotic complications including abdominal aortic aneurysm is likely to become more common, requiring extended cardiovascular follow-up.

Influence of sulfation on platelet aggregation and activation with differentially sulfated hyaluronic acids

A number of sulfated hyaluronic acid derivatives (HyalS2.5, HyalS3, and HyalS4) were prepared by sulfation of the -OH groups present on hyaluronic acid and were generically termed HyalSx. The anticoagulant properties of this series of compounds has previously been shown to be good in terms of their whole blood clotting inhibition and factor Xa and thrombin inactivation. The purpose of the present study was to investigate whether the use of these compounds would be beneficial to patients who would normally be given heparin, and to perform some preliminary investigations into their effects on platelets. The three compounds were thus studied by investigating their ability to inhibit von Willebrand factor–dependent platelet agglutination in comparison with unfractionated heparin. Agglutination was determined turbidometrically after the addition of ristocetin to stirred formaldehyde-fixed platelets and was demonstrated to be dependent on the presence of sulfate groups on the polysaccharide chain and correlated with the degree of HyalSx sulfation. Interactions possibly important in low shear environments were investigated by measuring the pharmacological action of the HyalSx on spontaneous platelet activation and aggregate formation by flow cytometry. The data indicate that platelet activation is not correlated with the number of sulfate or hydroxyl groups on HyalSx, suggesting that activation occurs not via electrostatic interactions or H bonding, but via some other mechanism. A differentiation between low and high glycosaminoglycan sulfation densities is observed with respect to platelet aggregation, which is correlated with the number of sulfated groups per disaccharide unit. The ability of HyalSx to inhibit platelet aggregation induced by ADP and thrombin was measured by aggregometry. HyalS4 resisted thrombin stimulation to a similar extent as heparin. All Hyal derivatives, however, were better at inhibiting ADP-induced aggregation than was heparin. We conclude, therefore, that clinical use of HyalSx in place of heparin may be beneficial because ristocetin-dependent agglutination, and therefore resistance to platelet aggregation in high shear environments, in addition to resistance to stimulation by ADP, has been shown to be superior to heparin. Spontaneous platelet activation and aggregation are induced at an overall low level, even at high HyalSx concentrations, and are comparable with that of heparin.

Transfer of membrane proteins from human platelets to liposomal fraction by interaction with liposomes containing an artificial boundary lipid

The direct transfer of membrane proteins from human platelets to the liposomal fraction was examined, particularly in relation to platelet activation during the process. The incorporation of an artificial boundary lipid, 1,2-dimyristoylamido-1,2-deoxyphosphatidylcholine (DDPC), in the interacting liposome considerably enhanced the efficiency of the protein transfer. The transfer proceeded with neither significant activation nor lysis of the platelet, and the activation of the platelet with thrombin did not affect the amount of the transferred proteins. A wide range of platelet membrane proteins was transferred, and they were almost comparable to those in a sample prepared by glycerol lysis/centrifugation. In addition, they included the major surface glycoproteins GPIIb and GPIIIa without noticeable contamination of soluble cytosol proteins. The protein transfer method is a one-pot process and clearly more convenient than the conventional extract and reconstitute approach. These results strongly support the use of the transfer process, especially with DDPC, as an alternative to the conventional detergent-solubilization or the solvent-extraction methods for preparation of samples of platelet membrane proteins.