Karen Ramberg

Evaluation of arterial prostheses in a baboon ex vivo shunt: The effect of graft material and flow on platelet deposition

Surface thrombogenicity is recognized as an important factor in the failure of small caliber vascular prostheses. The baboon ex vivo shunt was developed to study small caliber grafts under controlled conditions at different flow rates. The shunt was created by percutaneous insertion of catheters into the baboon femoral artery and vein. Platelet-graft interactions were studied using autogenous indium 111 labeled platelets. Two graft materials were placed in series and exposed to blood flow for 2 1/2 hours at flow rates of 25 and 200 ml/min. At the end of this period, the grafts were removed for morphologic examination. Platelet adhesion to the grafts, especially with the less thrombogenic materials (PTFE and HUV), was found to be independent of flow rate. PTFE was found to be the least platelet-reactive material, HUV was intermediate, and knitted Dacron was the most thrombogenic surface. Platelet deposition on the flow surface was confirmed by light microscopy and scanning electron microscopy.

Increased in-vitro incubation time of endothelial cells on fibronectin-treated ePTFE increases cell retention in blood flow

Endothelial cell (EC) seeding is postulated as a mechanism of improving patency of small calibre vascular grafts. However, the majority of seeded cells are lost within hours following restoration of blood flow. We postulated that incubating EC in-vitro on a graft will improve adherence and resistance to the sheer stresses of pulsatile blood flow. Fibronectin-treated ePTFE (5 cm x 4 mm ID) seeded with Indium-111-labelled autologous canine EC (1.5 x 10(5) cells/cm2) were incubated for four different time periods; 90 min, 24 h, 72 h and 6 days. Incubated grafts were subjected to blood flow of 75 ml/min for 6 h, in a canine ex-vivo arteriovenous shunt circuit. EC retention during perfusion was studied by measuring gamma activity emitted by the grafts. Cell morphology of non-perfused control groups and perfused groups was compared using scanning electron microscopy (SEM). SEM of control grafts showed progressive EC spreading on the ePTFE surface for up to 72 h incubation. Gamma activity was significantly higher at 6 h perfusion in grafts incubated for 72 h (82 +/- 4%) and 24 h (63 +/- 6%) vs. 90 min (34 +/- 13%, p less than 0.05), and between grafts incubated for 72 h vs. 6 days (55 +/- 7%, p less than 0.05). Perfused grafts incubated for 72 h showed unaltered EC morphology on SEM, few cells remained on 90 min incubated grafts. We conclude that incubating EC on fibronectin-treated ePTFE for 72 h in-vitro after seeding improves cell retention during blood flow.

The effect of low molecular weight dextran on platelet deposition onto prosthetic materials.

Low molecular weight dextran (LMD) has recently been shown in a randomized, multicenter trial to improve early patency in difficult lower extremity vascular bypass. The question remained as to whether this effect was due to plasma volume expansion by LMD or to its effects on platelets and coagulation. To analyze this we have studied the effects of LMD on platelet-graft interactions using an ex vivo baboon shunt. Indium 111-labeled autologous platelet deposition on expanded polytetrafluoroethylene and knitted Dacron grafts was monitored at a flow rate of 25 ml/min for 21/2 hours. LMD was given by intravenous bolus at a dose of 5 ml/kg and was followed by a continuous infusion of 10 ml/hr. Flow through the shunt was started after the initial dextran bolus. Control studies were performed in a similar manner without dextran. An additional series of experiments was performed with a 5 to 10 ml/kg bolus of 5% human serum albumin to analyze the contribution of plasma volume expansion to platelet deposition in this model. The results revealed a significant (p less than 0.05) reduction in platelet deposition on both graft materials with the LMD infusion. No difference was seen between the control group and the albumin group. These studies provide direct in vivo evidence of the antiplatelet effect of LMD.

A new primate model for the study of intravenous thrombotic potential and its modification

Advances in venous reconstruction have been limited by inherent venous thrombogenicity and the absence of a suitable prosthetic material for use in the venous system. We have designed an in vivo experimental model to evaluate early blood-material interactions within the venous system and to quantitate drug efficacy in the alteration of platelet function and fibrin deposition in the baboon. An 8F catheter was placed percutaneously in the femoral vein of an adult male baboon. Indium 111-labeled autogenous platelets or iodine 125-labeled human fibrinogen was infused before the introduction, into the inferior vena cava, of a linear array of 5 x 15 mm alternating Dacron and polytetrafluoroethylene samples attached to a benzalkonium-heparin-treated guide wire. At 60 or 120 minutes the samples were removed and a 1 ml aliquot of blood was drawn. The materials and blood samples were counted in a gamma well counter, and the material counts were normalized to the circulating label present in the 1 ml blood sample. The experiment was repeated after pretreatment with heparin, aspirin, or dextran. Whole blood clotting times and bleeding times were monitored. The results showed decreased platelet and fibrin deposition on polytetrafluoroethylene when compared with Dacron in the venous system. Aspirin, heparin, and dextran were all found to decrease platelet and fibrin deposition onto intravenously placed graft material samples (p less than 0.05, Students t test). The data confirm the ability of the model to evaluate quantitatively anticoagulants, antiplatelet agents, and prospective graft materials for use in venous reconstructions.

Development and evaluation of a new polymeric material for Small caliber vascular prostheses

Polyethylene oxide (PEO), because of its low levels of protein and cellular adsorption, may provide a suitable coating for synthetic small caliber vascular prostheses. PEO/polysiloxane networks were synthesized via an acid-catalyzed epoxy/hydroxyl crosslinking reaction and used to produce conduits with a 4-mm internal diameter. Three networks with nominal PEO molecular weights of 2000, 8000, and 20,000 and all 65% PEO by weight were studied. Blood compatibility was assessed by measuring 111In-platelet and 125I-fibrinogen deposition in a baboon ex vivo shunt, over a 1-hr time period and at a flow rate of 50 ml/min. Differences in material performance were noted particularly after the initial 30-min blood contact period. Materials in the mid and high PEO molecular weight range (8000 and 20,000) had significantly lower levels of platelet adsorption than networks of low PEO molecular weight (2000) at 30 min (P less than 0.005) and 60 min (P less than 0.05). The lowest level of platelet deposition was noted on networks of high PEO molecular weight (20,000). During the observation period, platelet accumulation on this surface was less than one platelet per 1000 microns. Platelet deposition on Gore-Tex was two and three orders of magnitude greater than that on the high molecular weight PEO material at 30 and 60 min, respectively (P less than 0.001). Fibrinogen adsorption was also lower on materials of mid and high PEO molecular weights, when compared with low molecular weight networks (P less than 0.05) and Gore-Tex (P less than 0.05). Scanning electron micrographs confirmed these observations. Overall, platelet and fibrinogen depositions are low for PEO networks, particularly for materials of high PEO molecular weight. This latter observation may be related to increased surface molecular mobility and a relative enhancement of PEO content at the blood-material interface.

Granulocyte sequestration and early failure in the autoperfused heart-lung preparation☆

We investigated the role of pulmonary granulocyte sequestration in the development of early failure of the autoperfused working heart-lung preparation. A significant decline in the total circulating leukocyte count in 21 preparations at 60 minutes of perfusion (5.0 to 1.4 x 10(3)/microL; 28% of baseline; p less than 0.001) was observed. Differential cell counts in 14 of these preparations revealed a predominant decrease in granulocyte count (8.7% of baseline) and a moderate decline in lymphocyte count (46% of baseline). In study I, indium 111-labeled autologous granulocytes were injected intravenously into 10 adult New Zealand White rabbits. In group I (n = 5), an autoperfused working heart-lung preparation was harvested and perfused for 60 minutes. In group II (controls, n = 5), the heart-lung block was harvested following 60 minutes of in situ perfusion. Organ blocks were imaged before and after saline flush. There was a significant decline in granulocyte counts at 60 minutes of perfusion in group I versus no change in group II (I, 2.3 +/- 0.4 to 0.3 +/- 0.1; p less than 0.01; II, 1.7 +/- 0.2 to 2.3 +/- 0.5; not significant; x 10(3)/microL +/- standard error of the mean). Postflush lung activity was significantly increased in group I versus group II (I, 3,751 +/- 566; II, 1,867 +/- 532; p less than 0.05; counts +/- standard error of the mean). In study II, 15 autoperfused preparations were divided into two groups. Group I (n = 10) preparations were controls. Group II (n = 5) animals were depleted of leukocytes by pretreating with nitrogen mustard. Group I (controls) produced a bimodal survival distribution (Ia, 8.2 +/- 1.0; Ib, 26.4 +/- 2.0; hours +/- standard error of the mean). Group II survival was significantly longer than that of group Ia and similar to that of group Ib (II, 25.3 +/- 2.2; p less than 0.01 versus group Ia, not significant versus group Ib; hours +/- standard error of the mean). Hemodynamic profiles for group II closely paralleled those of group Ib. In conclusion, pulmonary sequestration of granulocytes occurs early in the autoperfused working heart-lung preparation (within 60 minutes of autoperfusion), and preoperative leukocyte depletion prolongs survival of the autoperfused working heart-lung preparation by eliminating the subset group Ia (short survivors) seen in untreated preparations.

Imaging of intrathoracic disease with indium 111-labeled leukocytes

Indium 111-labeled leukocyte scintigraphy provides an excellent noninvasive means for identifying and localizing infectious and inflammatory sites within the body. This article first compares gallium 67 citrate, the long-established radiopharmaceutical for infection imaging, to 111In-labeled leukocytes, particularly as applied to the chest. 111In-labeled leukocytes require time-consuming preparation, and the technical and practical aspects of cell labeling and imaging are reviewed in detail. The main focus of this article is the application of labeled leukocyte imaging to the evaluation of pleural, mediastinal, cardiovascular, and pulmonary disease.

Kinetics of Pulmonary Platelet Deposition and Clearance During Thrombin-Induced Microembolism in Rabbits

Using 111In-labeled autologous platelets, we studied the kinetics of pulmonary platelet deposition and clearance in relation to hemodynamic and structural events during thrombin-induced pulmonary microembolism in rabbits. Autologous platelets were radiolabeled and returned to animals prior to infusion of thrombin (100 units/kg over 15 min) (n = 20) or saline (n = 6). All animals were pretreated with tranexamic acid, an inhibitor of fibrinolysis. Thrombin-treated animals manifested progressive increases in mean pulmonary platelet activity, reaching a maximum of 38% above baseline (p less than .0001), whereas no change was observed in saline-treated controls. Animals that died during, or immediately following, thrombin infusion manifested significantly greater increases in pulmonary platelet uptake (mean 1.55 +/- 0.47 times baseline), compared to surviving animals (1.14 +/- 0.16; p less than .05 survivors vs. nonsurvivors). In surviving animals, following cessation of thrombin, pulmonary platelet activity cleared gradually, with a half-time of approximately 12 min. Thrombin reduced circulating platelet counts (p less than .001), increased mean pulmonary artery pressure (13 +/- 3 mm Hg to 18 +/- 6 mm Hg; p less than .0001), and reduced mean systemic arterial pressure (55 +/- 10 mm Hg to 44 +/- 7 mm Hg; p less than .001). The time courses of these events approximated that of thrombin-induced pulmonary platelet uptake. Furthermore, the increase in pulmonary artery pressure occurred predominantly in the group of animals in which the increase in pulmonary radiolabeled platelet activity exceeded the median value of 20%. Postmortem histology showed extensive pulmonary thrombus extending from small arterial to capillary levels in animals that died during, or immediately following, thrombin infusion, but not in surviving animals. Our findings suggest that platelet aggregation plays an important role in the pathogenesis of hemodynamic change following thrombin-induced pulmonary embolization.

Critical Evaluation of an in Vitro Model for Evaluation of Platelet Reactivity of Biomaterials

An in vitro model for measuring platelet reactivity to a variety of biomaterial candidates for vascular grafts is described. A model consisting of a standard area of test material exposed to freshly labeled In platelets in plasma was evaluated. The platelets were isolated from ACD anticoagulated blood and resuspended in ACD plasma. It has been previously demonstrated that platelets so treated circulate in the body and will deposit on biomaterials exposed to the blood in vivo. The in vitro test consisted of an incubation of the platelets and materials at 37°C for one hour. At the end of the incubation, the platelet rich plasma was removed and the materials washed and removed for gamma counting. Platelet reactivity was normalized as a percentage of the counts on the material to counts in an aliquot of the platelet-plasma incubation media. The maximum uptake of platelets occurred within one hour. Platelets from three species, human, baboon, and dog were tested. Platelet uptake by Dacron and PTFE were in the range of 30–40% and 1–5% respectively. This is in accord with the known reactivity of these two vascular graft materials in vivo. A second series of studies were conducted with physically and pharmacologically inactivated platelets and inert particles. Those studies suggest that the initial results do not represent a biologic event but may reflect the porosity of the materials. This emphasizes the necessity of adequately defining an in vitro model against known in vivo activity.