thoracotomy thrombi-hrt lung-machine-jtt-1996

Journal of Thrombosis and Thrombolysis 1996;3:195-208 �9 Kluwer Academic Publishers, Boston. Printed in the Netherlands.

Effect of Thoracotomy and Cardiopulmonary Bypass on Activated Platelet and Neutrophil Dynamics and Platelet Emboli in a Pig Model

Mrinal K. Dewanjee, 1 Victor Belinskiy, 2 Jamie F.W. Holland, 2 Mansoor Kapadvanjwala , 3 Shu Ming Wu, I~ Stana Novak, 2 Li-Chien Hsu, ~ Juan Sanchez, 2 Sumi t Dewanjee, I Aldo N. Sera f in i / Robert C. Duncan, 6 and George N. S fakianakis 1 Departments of 1Radiology, 2 Surgery, 3Biomedical Engineering, and 4Epidemiology, University of Miami, School of Medicine, Miami, Florida; ~Bentley Laboratories, Irvine, California

Abstract. The effects of thoracotomy and components of ex- t racorporeal circuits on dynamics of platelets and neutrophils were quantified with autologous I n - I l l - l a b e l e d platelets ( INPLT) and neutrophi ls (INN) dur ing cardiopulmonary bypass (CPB) operat ions in Yorkshire pigs. Cardiopulmo- nary bypass was carr ied out with a hollow-fiber oxygenator and an ar ter ia l filter in 48 pigs (30-35 kg; 12 unoperated controls for platelets and neutrophi ls ; 12 sham operated controls; 12 with 180 minutes of CPB with platelets and neutrophils ; 12 with 90 minutes of CPB and 90 minutes of reperfusion at 2.5-3.5 one/min. Pla te le ts and neutrophi ls were labeled witb I n - l l l t ropolone and were injected intravenously: platelets at 24 hours and neutrophi ls a t 15 minutes before CPB. All pigs were systemically hepar inized [activated coagulat ion t ime (ACT) >400 seconds]; CPB was ins t i tu ted with a roller pump, oxygenator (OX; Bent ley Univox, 1.8 m2), and ar ter ial filter (AF; 0.025 m 2) for dura t ions of 180 minutes and 90 minutes of bypass, followed by 90 minutes of reperfusion. The kinet ics and pooling of platelets and neutrophi ls were monitored by a Geiger probe. The adheren t th rombi and neutrophils in the OX, AF, viscera, and bra in were imaged with a gamma camera and were measured with an ion chamber and a gamma counter . The percenti le dis t r ibut ion of labeled platelets and neutrophi ls expressed as the mean - s tandard deviat ion of injected dose in eight groups was calculated and s tat is t ical analyses were performed (ANOVA and paired t- test). Sham operat ion alone increased platelet re tent ion in the lung, hear t , and brain significantly (p < 0.001) over tha t of unoperated pigs. Neutrophi l marg ina t ion to lung immedi- ately af ter inject ion was high; CPB and reperfusion altered the dis t r ibut ion in blood, viscera, and connective tissues. Dur ing CPB, an equi l ibr ium among single platelets, platelet thrombi, and emboli was reached in the blood, oxygenator, ar ter ia l filter, perfused organs, and tissues. After CPB, the pu lmonary neut rophi l re ten t ion increased significantly (p < 0.001). Reperfusion of 90 minutes following 90 minutes of CPB decreased the level of neutrophi ls and increased the level of platelets in the lung. Only a small amount of platelets and neutrophi ls was re ta ined in the oxygenator and ar ter ia l filter. Neutrophi l re ten t ion in the OX and AF was higher t han t ha t of platelets. The small amount of re ta ined neutrophi ls in the hear t , kidneys, and bra in suggested tha t cytokines,

r a the r t h a n marg ina ted neutrophi ls alone, may play a major role in inf lammatory insul t to these organs dur ing and after CPB. OX thrombi increased with the t ime of CPB; AF thrombus in both groups was almost similar. During CPB, AF func- t ioned minimal ly as a th rombus t rap with a small percent of re ta ined thrombi ; reperfusion post-CPB did not change the amount . Thoracotomy alone has a significant effect on platelet and neutrophi l kinetics, and on the subsequent effect of th rombus format ion, embolization, and neutrophi l margina- t ion in organs dur ing the CPB procedure.

Key Words. cardiopulmonary bypass, Yorkshire pigs, indium- 111, platelets, neutrophi ls , adheren t thrombi , platelet emboli, marg ina ted neutrophi ls

Primary and secondary factors affecting platelet and neutrophil dynamics and their margination during cardiopulmonary bypass (CPB) have not been evaluated previously. In previous studies, the parameters of activation of cellular elements of blood, coagulation, complements, and cytokine cascades have been evaluated. In the process of dissecting the effects of thoracotomy and CPB with radiolabeled platelets and neutrophils, we observed that thoracotomy alone significantly altered the localization of platelet emboli and the margination of neutrophils in the lung, brain, and heart. Cardiopulmonary bypass surgery induces damage in organs and tissues by ischemic insult with thromb] and emboli. Leukocytes add additional in-

Address for correspondence: Mrinal K. Dewanjee, Ph.D., Univer- sity of Miami, School of Medicine, Division of Nuclear Medicine (D- 57), Jackson Memorial Medical Center, Central Building, Room C266, P.O. Box 016960, Miami, FL 33101.

Received 28 September 1995, Accepted 18 October 1995

195

196 Dewanjee et al.

fiammatory insult by the release of cytokines and free radicals from activated neutrophils, monocytes, and lymphocytes.

Prosthetic devices with large surface areas (oxygenators and hemodialyzers) consume circulating platelets by adhesion of platelets, forming thrombi and emboli, and of neutrophils by margination and platelet/neutrophil disintegration due to the high shear rate [1-10]. In previous studies of the extracorporeal circulation of CPB and hemodialysis in the pig model, we carried out platelet balance studies with In-Ill-labeled autologous platelets [7-10,13] and pin- pointed the site of thrombus formation in hemodialyzers, oxygenators, and arterial filters.

Neutropenia during hemodialysis has been observed previously [24]. Neutrophil consumption by these processes in the intraluminal flow oxygenator and arterial filter was recently quantified with In-11 I - labeled neutrophils in our laboratory [28]. We also determined the comparative retention of neutrophils from beagle dogs, Yorkshire pigs, and healthy human volunteers using a flow loop consisting of a hemodialyzer and a roller pump, and demonstrated no difference in neutrophil reactivity to cellulose fibers in these three species [28]. The adhesive molecules on the membranes of neutrophils and monocytes (integrins: Mac-l, LFA-1, lectins) and endothelial cells (cell adhesion molecules, ICAM-1, VCAM-1, P-selectins, E- selectin: ELAM-1) result in the adhesion of rolling neutrophils to endothelial cells, rupture of endothelial junctions, and transendothelial migration in the inter- stitial space [17,28,29]. Activated neutrophils in inflammatory lesion possess a heterodimeric surface glycoprotein complex of CD11b/CD18 or Mac-1 (gp155/gp95), which binds the counter-receptor of the endothelial cell adhesion molecule, ICAM-1. They also interact with adsorbed proteins in hollow-fiber devices.

Dynamic interactions among adhesion receptors and counter-receptors expressed by leukocytes and endothelium control the process of margination and migration into the tissue space. The four critical stages of neutrophil-endothelial interactions are (1) tethering, (2) triggering, (3) gluing, and (4) transendothelial migration and depend on the cascade of multiple receptor-counter receptor interactions. The initial stage of tethering is mediated by selectins, with hy- brid receptors consisting of the NH 2 terminus of the lectin-binding domain, the epidermal growth factor (EGF)-like domain, and a variable number of short consensus repeats (SCR). E- and P-selectins are expressed after initial stimulation; within a few minutes of the initial signaling, the Weibel-Palade bodies containing P-selectin fuse with the membrane, but they are also internalized quickly within 30 minutes, when the E-selectin takes over the role of tethering. The synthesis of E-selectin is initiated by cytokine stimulation and peaks in 4-6 hours [17,26,29,30].

Over the last 15 years, we have been quantifying

the dynamics of platelets and neutrophils, and the level of thrombi and emboli in the critical organs, with bubble, membrane, and hollow-fiber oxygenators in the dog and pig models [6-14]. We also studied the effect of CPB sugery on the thrombogenicity of heart valve prostheses [14,20,21]. Our systematic cell- balance studies indicated that there is a significant reduction of thrombi and emboli formed in the components of the CPB circuit, mainly due to the improved design of the oxygenators and arterial filters. In spite of these improvements, the net retention of platelets and neutrophils by the lung, heart, brain, and lddneys decreased only marginally. The sham thoracotomy operation alone had a dramatic effect on the mobilization and retention of In-ill-labeled platelets and neutrophils by these organs. In this report, we describe our observations on the dominant effect of thoracotomy over the circulatory effect of blood in the extraluminal flow oxygenator on margination and retention of platelets and neutrophils during CPB and on reperfusion time in the pig model.

M a W r i a l s a n d M e t h o d s

The experiments reported herein were conducted according to the principles set forth in the Guide for the Care and Use of Laboratory Animals from the Institute of Laboratory Animal Resources, National Research Council (NIH 86-23, 1985). The details of blood collection; labeling of platelets and neutrophils with In-111 tropolone, CPB operations, static and dynamic measurements of adherent thrombi and neutrophils to the oxygenator, arterial filter, and platelet emboli, and marginating neutrophils in lung, brain, heart, and kidneys by imaging with a gamma camera, ionization chamber, gamma counter, and Geiger probe have been described previously [6,7,11,13,28]. A brief description of these procedures follows:

Labeling of autologous platelets with indium-lll tropolone Porcine platelets were labeled according to the method of Dewanjee et al. [11]. Sixty milliliters of whole blood were collected from anesthetized York- shire pigs in 10 ml of modified ACD anticoagulant solution (NIH-A, Fenwal Laboratory). The platelets were separated by differential centrifugation, were incubated with In-111 tropolone (600-700 ~Ci) for 20 minutes, and were washed by centrifugation. The In- 111-labeled platelets were measured with ionization chamber (Capintec CRC-5RH3) and were injected into the ear vein of anesthetized pigs. A platelet harvesting efficiency of typically 40-45% and a labeling efficiency of 80-95% were routinely obtained.

Labeling of autologous neutrophils with indium-Ill tropoione Porcine neutrophils were radiolabeled according to the method of Chowdhury et al. [12]. One hundred

Thoracotomy and CP Bypass on Thrombi/Emboli 197

milliters of venous blood were collected in two 60 ml syringes, each containing 10 ml of ACD anticoagulant (acid, citrate, dextrose; ACD, NIH-A; Baxter, Deer- field, IL); 25 ml of anticoagulated blood were transferred to each of four sterile 50 ml polypropylene round-bottom tubes, each containing 8 ml of Heta- starch (HESPAN, Du Pont Pharmaceuticals, Wilmin- gton, DE) from the vial. Blood was gently mixed with Volex, avoiding air bubble formation. Red blood cells (RBCs) were sedimented at a 45 ~ angle for 20-25 rain- utes in a sterile laminar flow hood. Supernatant containing white blood cells, platelets, and some RBCs was removed with a sterile pipette in a laminar flow hood. Ficoll-Hypaque (F-H) stock solutions were pre- pared by mixing 9% Ficoll (Pharmacia Fine Chemi- cals~ Uppsala, Sweden) and Hypaque (33.9%, Win- throp, New York) in sterile glass cylinders. The density was adjusted with a hydrometer, and finally the solution was sterilized by membrane filtration (0.22 ~m, NaIgene).

Five milliliters of leukocyte-rich plasma were transferred with a sterile polyethylene pipette to each of 12 sterile screw-capped tubes (Corning, 15 ml); then 3 ml of F-H solution (1.080 g/ml) and 3 ml of F-H solution (1.120 g/ml) were layered. They were spun at 2800 rpm (Jouan C412, Winchester, VA; 1650 g) for 12 minutes at 18~ The plasma and platelets (top layer) were withdrawn with a pipette and discarded. The neutrophil aliquots were pooled in a sterile (40 ml) centrifuge tube and were diluted with 20 ml of ACD saline to remove the EDTA and F-H residue. The centrifuge tubes were spun at 1300 rpm (Beck- man TJ6R, 400 g, 18~ for 6 minutes to settle the neutrophil pellet.

The supernatant was discarded using a sterile polyethylene tube; the plasma and platelets were withdrawn with a pipette and discarded. Seven hundred to 800 ~Ci of In-111 tropolone in 2 ml of ACD saline were added to the pellet and, after resuspension, were incubated for 25 minutes and washed twice with 4 ml of ACD saline by centrifugation at 400 g; 0.2 ml of aliquot was transferred for the CBS and differential (Coulter S +, Coulter, Miami, FL). The radioactivity was measured with a dose calibrator (Capintec CRC- 5RH3, Ramsey, N J). The suspension was injected intravenously with a butterfly needle into the ear vein of the anesthetized pig. A neutrophil harvesting effi~ ciency of typically 15-25% and a labeling efficiency of typically 60-70% were obtained by the above- mentioned procedures in our laboratory.

Cardiopulmonary bypass with a roller pump, a hollow-fiber oxygenator, and an arterial filter in Yorkshire pigs Twenty-four pigs in four groups (six on platelets, with 180 minutes CPB; six on platelets with 90 minutes CPB and 90 minutes reperfusion; six neutrophils with 180 minutes CPB; and six on neutrophils with 90 minutes CPB and 90 minutes reperfusion) were studied

(Figures 1A and 1B). Anesthesia was induced with ketamine (20 mg/kg, IM, Fort Dodge Laboratories, Ford Dodge, IA) and sodium pentobarbital (3-5 mg/ kg IV as needed, Abbott Laboratories, N. Chicago, IL) [15-18]. After endotracheal intubation, the pigs were connected to the ventilator. Carfentanyl was ad- ministered intravenously (bolus 20 ~g/kg, Abbott Laboratories, N. Chicago, IL) along with muscle re- laxants (pancuronium bromide, 0.1 mg/kg iv, Elkins- Sinn, Cherry Hill, N J). The extracorporeal system (volume, 2 1) consisted of polyvinyl chloride tubing, a venous reservoir (Bentley BMR 1900, Irvine, CA), a barely occlusive roller pump (American Optical), a membrane oxygenator (Bently Univox, 1.8 m2), and an arterial filter (Bentley AF-1025, 0.25 m2).

The whole extracorporeal system was flushed with C02 and was primed with lactated Ringer's solution. After endotracheal intubation and being connected to the ventilator (Harvard Apparatus, dual phase control ventilator, S. Natick, MA), arterial blood gases were monitored and the ventilator was adjusted to maintain normal pH, pOz, and p C Q values ( G e m - 6 plus, Mal- linckrodt Sensor System, Ann Arbor, MI). Aortic, arterial, and venous pressures were continuously monitored by transducers (TA400, Gould, Valley View, OH), along with the electrocardiogram and heart rate (Datascope 2001, Paramus, NJ). After a median sternotomy, an arterial cannula was inserted into the ascending aorta (16-18F) and a venous cannula (40F) was inserted into the right atrium via the atrial appendage and was retained by purse-string sutures. The arterial and venous lines were connected with components of the CPB [15-18]. Cardiopulmonary bypass was initiated and the left ventricle was vented via the left atrial appendage. The ascending aorta was clamped, and cold cardioplegia (4~ was infused via the aortic root.

Cardiopulmonary bypass was carried out in four groups of 24 pigs (six with the filter and six without the filter) for 180 and 90 minutes of CPB and for 90 minutes of reperfusion. Rectal temperature was monitored and maintained at 25-28~ Serial blood samples were withdrawn after endotracheal intubation, prior to initiation of CPB, at 1 hour of CPB and just prior to termination of CPB. During CPB, systemic hepa- rinization was maintained (3 mg/kg iv) and was monitored with repeated determinations of the activated coagulation time (ACT), every 20 minutes, main- taining it at four times normal (400-600 seconds, normal value 100-120 seconds; Hemochron 400, Tech- nidyne, Edison, NJ). Cross-clamp time was always maintained at 45 minutes for all animals. Cardioplegic solution (Plegisol, Abbott Laboratories, N. Chicago, IL) was infused initially through the aortic root at 15 ml/kg, and every 20 minutes an additional infusion was performed at 8 ml/kg. Hespan (500-1000 ml, Du Pont Pharmaceuticals, Wilmington, DE) was adminis- tered to maintain osmolarity.

A sham operation was carried out using a median

198 Dewanjee et al.

PROTOCOL OF THE STUDY OF PLATELET KINETICS AND DEPOSITION DURING CARDIOPULMONARY BYPASS

(CPB) IN A PIG MODEL

Blood collection, Injection (I.V.) 1111n CPB OFF " Imaging of Platelet labeling Platelets oxygenator, filter Reperfusion, sacrifice,

with 1111n-tropolone (400-550 ~Ci) CPB ON and quantification of quantification of Platelets Platelets

-180 mln -15 min 0 90 rain 180 min

A

TIME

PROTOCOL OF THE STUDY OF PLATELET KINETICS DURING CARDIOPULMONARY BYPASS

(CPB) IN A PIG MODEL

Blood collection, Platelet labeling

with 1111n-tropolone

Injection (I.V.) 111111- Plateleta

(400-550 pCi) CPB ON

CPB OFF Sacrifice, Imaging of

oxygenator, filter and quantification of

platelets

t t , , t t -180 rain -15 min 0 180 min

TIME

Fig. 1. Protocol of the study of platelet kinetics (A) during cardiopulmonary bypass (CPB; 3 hours at 2.5-3.5 1/min) and 90 minutes of CPB and 90 minutes of reperfusion (B) with In-ill-labeled autologous platelets in the pig model.

sternotomy in a similar fashion, and only the right pleural cavity was entered. The pigs were sacrificed after 180 minutes of sternotomy. At the end of CPB or reperfusion, each animal was sacrificed by exsan- guination. Both kidneys, as well as the lung, heart, brain, liver, and spleen were removed, preserved, sectioned, and inspected grossly for embolic material. The radioactivity in intact organs and tissue samples was measured with an ionization chamber and a gamma counter (Cobra II, Model 5003, Packard, Downers Grove, IL). The window of the gamma counter was adjusted to include the 171-, 245-, and 426-keV sum peaks of the indium-Il l radionuclide.

Biodistribution of i nd ium- I l l - labe led platelets and neutrophils in control and sham-operated pigs To differentiate between the role of thoracotomy and CPB, the biodistribution of radiolabeled platelets and

neutrophils was measured in four groups of 24 pigs. Two control groups of six unoperated pigs each were sacrificed after injection of In- I l l - labeled platelets (24 hours) and neutrophils (3 hours) and were desig- nated UNOPCPL and UNOPCN, respectively. Two additional sham-operated control groups of six pigs each were sacrificed after injection of In-I l l - labeled platelets (24 hours) and neutrophils (3 hours), and were called SHOPCPL and SHOPCN, respectively. All animals were heparinized and sacrificed with an overdose of barbiturate.

Dynamic measuremen t of radioactivity of indium--I l l platelets and neutrophils in the oxygenator and arterial filter during C P B The radioactivity in the oxygenator and arterial filter during CPB was measured with a lead-collimated Gei- ger probe detector (Ludlam Inc.). The background ra-


dioactivity was subtracted from the total radioactivity of adherent thrombus in oxygenator and trapped em- bolus in the arterial filter, and the net radioactivity was plotted with time of CPB.

Scintiphoto of platelet and neutrophil deposition with the gamma camera, and measurement of radioactivity with the ionization chamber and gamma-well counter At the end of CPB or reperfusion, the circuit was washed with 1 liter of sterile saline at a flow of 400-500 ml/min. The components of the arterial ill- te r and hollow-fiber oxygenator were removed from the circuit of the termination of CPB. The intact components were imaged with a gamma camera (Sie- mens, LFOV, Hoffman Estates, IL). The three- dimensional distribution of platelet and neutrophil deposition was also imaged with the single photon emission computed tomography (SPECT) technique. The oxygenator and arterial filter were placed on the head rest so that their center of rotation was coaxial with the central axis of the camera gantry. The triple-headed gamma camera (Triad 88, Trionix Research Laboratory, Twinsburg, OH) was fitted with a medium energy, parallel-hole collimator. The heads were rotated clockwise, with 30 stops for each head at a 4~ and with 60 seconds of data acquisition at each stop in a 256 • 256 matrix. Images were processed for transaxial, sagittal, and coronal distribution at a thickness of 0.5 cm.

The radioactivity in the components was measured in a large-bore ionization chamber (Capintec CRC- 5RH3). The samples of the fibers and filter membrane obtained from the periphery, center, top, middle, and bottom of the filter and oxygenator were weighed on a microbalance, and the radioactivity was determined with a gamma-well counter in the In - l l l channel. All the detectors were calibrated with standard sources of In-111 to convert microcuries into counts per minute (cpm) and to calculate the percent of the injected dose of In-111 platelets and neutrophils in the device, blood organs, and tissues.

Statistical analysis of the distribution of radiolabeled platelets and neutrophils in Yorkshire pigs The statistical model for each experiment (platelet thrombosis, neutrophil infiltration) is a factorial experiment with three factors of thoracotomy, 180 minutes of CPB, 90 minutes of CPB, and 90 minutes of reperfusion. Statistical analysis of the data was carried out by comparing organ distribution for labeled platelets and granulocytes, and their balance during and after CPB among three groups. Analysis of variance was performed for each study to assess the overall effects of thoracotomy and CFB on the distribution of platelets and neutrophils. Overall comparisons used a p value of <0.05 to indicate a significant difference.

R e s u l t s

We hypothesized that excess platelets and neutrophils in organs that were above the control unoperated ani- real levels were due to emboli and margination. Thora- cotomy alone activates several pathways, for example, clotting factor cascades, complement, and the extrinsic tissue factor pathway. In addition, whole organs and connective tissues in the body work as an arterial filter (Figure 4) in removing emboli (lung, liver, skeletal muscle) and microparticles (reticuloen- dothelial system). These organs and tissues also pro- vide a pool for the margination of activating neutrophils with sticky endothelial cells. In this study, we wanted to determine what additional activation and insult result from CPB that are over and above that induced by thoracotomy. The role of the redistribution of platelets and neutrophils due to CPB was estimated after subtracting the effect of thoracotomy.

Equilibria of circulating platelets with platelet microparticles and aggregates formed during CPB and their sequestration in Yorkshire pigs are shown in Figure 4. During CPB operations, platelet number and functions decrease [1-10]. The decline in platelet number is mainly due to hemodilution; adhesion to surfaces and the formation of adherent thrombus has decreased significantly (from 20% to 10% to 2% to 0.2%) from bubble to membrane to internal- to exter- nal-flow hollow-fiber oxygenators over the last 20 years due to improved designs and the use of compati- ble biomaterials ([5-10]. Activation of platelets was measured by the releasate (platelet factor 4, throm- bospondin, ~-thromboglobulin) from a-granules from activated platelets and was found to increase with the time of CPB [1-4]. A similar increase was observed for releasate (ADP, ATP, calcium and serotonin) from dense granules.

The receptor density for both ligands, fibrinogen and epinephrine, on the platelet membrane also decreases, and platelet fragments increase in the blood. Scanning and transmission electron micrographs illustrated the presence of a heterogenous population of platelets: intact activated platelets, partially and completely degranulated platelets, platelet ghosts, and an influx of larger platelets that enter into the circulation from the marrow pool. These platelet defects result in an increase in the postperfusion bleeding time and reduced platelet function and survival time [1,2].

Our quantitative platelet balance studies suggest that thoracotomy alone has a more profound effect on platelet balance than the effects of multiple passages of blood through the components of the CPB circuit. Thrombi in lung increased significantly from 5% to 18% (p < 0.001) and in heart increased from 0.3% to 0.6% due to thoracotomy alone; platelets were proba- bly activated by the tissue factor pathway. This increase in the lung and heart was followed by a concom- itant drop in the blood platelet level from 38% to 31% (p < 0.01). Scintiphotos of neutrophil deposition in the


lung, liver, and spleen of the unoperated and operated pig undergoing CPB are shown in Figure 2A and 2B, respectively. Neutrophil depletion from circulating blood is evident from the absence of radioactivity in the cardiac chambers. After CPB, the splenic neutrophil pool was minimal.

Although oxygenators have been perfected to a large extent, and the thrombus level has dropped significantly from 20% (bubble oxygenator) to 0.2-0.3% (hollow-fiber oxygenator), no major concurrent im- provement has occurred in the arterial filter. Neutro- phil retention was higher than platelet retention; the scintiphotos illustrated an inhomogenous distribution of platelets in the oxygenator and filter with the presence of adherent thrombi as hot spots on the OX and AF walls (Figure 3A). A similar distribution was also observed in the oxygenator (Figure 3C). The current filter uses a pleated Dacron (polyester hydrophobic surface) fiber network (porosity 25 ~m). The accor~ dion configuration with lateral (side-to-side) as op- posed to end-to-end blood flow results in a less favor- able hemorheology and thrombus formation both inside and outside of the filter, and the trapping of emboli in distal organs [3-5]. The capacity for thrombus formation when using this filter is high, and the capacity for thrombus retention is low. After 30 minutes of oxygenation, the thrombus level reached a steady state.

Our systematic studies with 1-125-1abeled plasma proteins (serum albumin, fibrinogen, ~-thrombin, ira- munoglobulin, and complement 5) demonstrated that plasma proteins are adsorbed in the hollow-fiber lumen instantaneously and reached a steady state within 5-10 minutes of circulation time [24,32]. The adsorbed adhesive proteins, for example, fibrinogen and von Willebrand factor, bind glycoprotein receptors of activated platelets to form platlet aggregates. The dynamic studies also showed that thrombus formation on the oxygenator and arterial filter reached a steady state due to thrombus formation and embolization. In addition, we observed that in vitro studies overestimate platelet deposition on the oxygenator by a factor of 10-15 (unpublished data).

D i s c u s s i o n

The oxygenator and dialyzer, having large surface areas, consume circulating platelets by platelet- thrombus formation and platelet fragmentation. Both processes occur inside or outside the narrow fibers by shear-induced platelet fragmentation. Our previous studies with flow cytometry showed a linear relation- ship between blood flow and platelet fragmentation

for both the roller pump and centrifugal pump [32]. The concentration of the platelet microparticle (PMP) increases with the flow rate. When dialysis was carried out in the presence of a hemodialyzer, the PMP density was much lower, suggesting that the formation of PMPs was shear dependent.

The large multimeric glycoprotein (1-20 • 10 ~ Da), von Willebrand factor (vWf), mediates platelet adhesion and thrombus formation at the site of vascular injury in a specific temporal sequence [1,2,32]. It serves as a carrier for factor VIII procoagulant factor by forming a VIII/vWf complex, vWf absorbed on the subendothelial or biomaterial surface or site of the wound interacts with the platelet glycoprotein GP Ib-IX-V complex, initiating platelet adhesion and activating the GF iIb-IIIa complex. The activated complex, in turn, promotes platelet spreading, irre- versible adhesion, and platelet aggregation. Both circulating vWf and vWf acutely released from 5- granules of activated platelets or endothelial cells are effective in platelet adhesion. Although this interaction was qualitatively evaluated in the everted aorta, platelet and protein deposition was never quantified in the hollow-fiber devices. Blood flow at a high shear rate outside the lumen of hollow fibers (internal diam- eter, 200 Ixm) may simulate blood flow in a porous matrix.

The adsorbed clotting factors (vWf and fibrinogen, fibronectin) on polypropylene in the oxygenator and Dacron polyester in the filter promote the adhesion of activated platelets via the exposed receptors. Calcula- tion of the shear rate in extraluminal flow is difficult due to the inhomogenous distribution of the fiber bun- dles. At a high shear rate, vWf reacts with both of the platelet-binding sites of GP Ib and GP IIb-IIIa. The GPIb interaction is independent of platelet activation, whereas IIb-IIIa binding occurs after platelet activation. The midsection of the disulfide loop at the A1 domain of vWf undergoes a conformational transi- tion, and this alteration increases the affinity for GP Ib. The cell adhesion segment of the tetrapeptide (Arg-Gly~Asp-Ser, with other amino acids substitut- ing for serine) also exists in the carboxy terminus (residue 1744-1747) of vWf. Antibodies specific to this peptide epitope completely inhibit vWf binding to GP IIb-IIIa and shear-induced platelet aggregation.

The emboli generated during CPB localize in the hippocampus, inducing short-term memory loss, and in other sensitive parts of the cortical brain, lung, hearL and kidneys, causing thromboembolic complications induced by ischemia and cell death [21,23,24]. The growth of platelet aggregates in the fiber lumen depends specifically on the shear rate and platelet releasate (ADP, serotonin, thromboxane A2, etc.), and

Fig. 2. A: Scintiphoto of neutrophil deposition in the lung, liver, and .spleen of an unoperated control pig at 3 hours postinjection of In-Ill-labeled neutrophils. B: Scintiphoto of neutrophil deposition in the lung, liver, and spleen of pig undergoing CPB at the termination of cardiopulmonary bypass (3 hours).

202 Dewanjee et al.

A


Fig. 3. A: Scintiphoto of plate deposition on a Univox oxygenator (first row, transverse; third row; sagittal; and fifth row; coronal) and arterial filter (second row; transverse; fourth row; sagittal; and sixth row; coronal) at 180 minutes of extraco~poreal circulation with the SPECT technique. Inhomogenous platelet deposition was observed in all segments of the oxygenator and arterial ill- ter. B, C: Photograph and scintiphoto of neutrophil deposition on the oxygenator (Univox) after 3 hours of extracorporeal circulation. The circuit was flushed at a flow rate of 400-500 ml/min with I 1 of sterile saline before imaging with the gamma camera.

I Cell-clusters] I Ar ter~-""~

,Oxy.q~nator~ ~ X ~ ~ '

[Circulating~~-I~ Platelet-fibrin~ ~ ~ 1 ~ [platelets ]Platel~et-~ ~ X lmicr~ I Platelet- I / ~

[neutrophils J ~ [PMP-neutrophil~

[Trapping of emboli l ~ A kung, liver~spleen I

idneys, heart, brain l Connective~[

~[RES 1 j~,~sequestration~

Fig. 4. Equilibria of circulating platelets with platelet microparticles and aggregates formed during cardiopulmonary bypass and their sequestration in Yorkshire pigs.

204 Dewanjee et al.

on the amount of a d h e r e n t p la te le t s decreases at a h igher flow. By redes ign ing the a r t e r i a l f i l ter to facili- t a t e the di lut ion of these a g g r e g a t i n g agents , th rom- bosis and subsequen t embolizat ion could be fu r the r reduced .

The r ad ioac t iv i ty moni to red wi th the Geiger p robe in organs and devices was normal ized by dividing the counts pe r minu te by the to ta l in jec ted rad ioac t iv i ty and was p lo t t ed aga ins t CPB t ime. The kinet ics of neu t roph i l t raff icking, marg ina t ion , and t r ansen- doethel ia l migra t ion we re moni to red using the Geiger probe. Lung and l iver r ad ioac t iv i ty in CPB groups of pigs r eached a s t e a d y s t a t e wi th in 30 minutes; in spleen, the neu t rophi l -bound rad ioac t iv i ty decl ined with neu t ropen ia . Due to continuous neut rophi l activa t ion b y the o x y g e n a t o r and filter, lung r e t en t ion did not decl ine wi th CPB t ime. Reper fus ion significantly dec reased the level of a d h e r e n t neut rophi l s in the lung. Rad ioac t iv i ty in p ig bra in was at a cons tant level dur ing CPB in all g roups and increased significantly from tha t in the unope ra t ed control group.

C o m p a r a t i v e neu t rophi l r e t en t ion exp res sed as the pe rcen t of the in jec ted dose (ID% or neu t roph i l /m 2) was e s t i m a t e d for the surfaces of the lung, oxygenator , and a r t e r i a l filter. Al though the r e t en t ion of neu- t rophi l s b y the oxygena to r and fi l ter was compara- t ive ly low (0.70-4.72 and 0.19-0.67, respec t ive ly) , the adso rbed cytokines and complements p robab ly pro- v ided a r eac t ive surface for continuous s t imula t ion of neut rophi l s and subsequen t marg ina t ion to the lungs and o the r o rgans and t issues . Dur ing CPB, the neu- t rophi l s we re cont inuously dep le ted from the s to rage pool in the spleen and blood, and were marg ina t ed to the lungs, l iver , ske le ta l muscle, bone mar row, and k idneys .

Separa t ion of neut rophi l s by cent r i fugat ion and o the r mechanical s teps of compact ion and resuspension m a y r e ve r s ib ly ac t iva te neut rophi l s dur ing radio- label ing procedures . Abou t 24%, 19%, and 4% of in- j ec t ed neut rophi l s m a r g i n a t e d to the lung, l iver, and ske le ta l muscle in the unope ra t e d control group; 19%, 25%, and 17% of in jec ted neut rophi l s ma rg ina t e d to the lung, l iver , and skele ta l muscle in the sham- ope ra t ed group. Thoraco tomy decreased ma rg ina t ed neut rophi l s to the lung from tha t of the unope ra t ed control group (p < 0.05). Abou t 0.37% and 0.65% of adhe ren t neut rophi l s were found in the oxygena to r and 0.02% and 0.43% were found in the a r te r ia l f i l ter a t 180 minu tes of CPB and reperfus ion, respec t ive ly . This neut rophi l d i s t r ibu t ion was dependen t on blood flow, the dura t ion of flow, and the type of oxygenator , and was s ignif icant ly a l t e r ed dur ing oxygenat ion with hollow-fiber in t ra lumina l and ex t ra lumina l flow oxygena tors . Since the surface a reas a re different , the %ID and neu t rophi l dens i ty were calculated pe r unit a r ea of the surfaces of the oxygena tor , a r t e r ia l filter, and lungs.

The confluence of s tudies of subse t s of leukocytes (neutrophi ls , p la te le t s , and T cells) and endothel ia l cells sugges t s t t ha t the four processes of (1) initial t e the r ing , (2) t r igge r ing , (3) s t rong adhesion, and (4) t r ansendo the l i a l migra t ion function at the level of the cel l-adhesion molecules. The ci rculat ing and margin- a ted neu t rophi l s in the lung of unope ra t ed pigs re ta in the i r homing function, and about 4% and 12% migra te to the lung and ske le ta l muscle in sham-opera ted pigs (Table 2).

Compar i son of the kinet ics of in te rac t ion of p lasma pro te ins and neu t rophi l s wi th b iomate r ia l s eva lua ted wi th 1-125-1abeled pro te ins and I n - I l l - l a b e l e d neu-

Table 1. Biodistribution (percent of injected dose) of indium-ill-labeled autologous platelets in Yorkshire pigs after 90 minutes cardiopulmona~g bypass, and 90 minutes reperfusion and 180 minutes bypass, with an oxygenator (Univox) and arterial filter (AF1025)

Organs and Unoperated control Sham-operated control CPB CPB-RP tissues (180 min, n = 6) (180 min, n = 6) (180 rain, n = 6) (90 rain CPB, 90 min rep., n = 6)

Lung 4.58 -+ 0.54 17.79 • 6.36 12.65 • 5.33 17.12 • 8.39 Liver 18.78 • 3.52 25.59 -- 2.12 28.96 • 5.42 28.96 • 5.42 Spleen 12.61 • 3.39 13.54 _+ 1.69 16.50 • 4.57 13.29 • 4.66 Kidneys 0.45 • 0.19 0.21 • 0.13 0.30 • 0.07 0.43 • 0.07 Heart 0.29 • 0.06 0.56 • 0.26 0.15 • 0.08 0.23 • 0.04 Muscle 4.88 -+ 0.78 4.35 • 0.55 5.13 • 4.50 7.58 • 2.96 Fat 4.0 • 0.51 2.64 § 0.91 1.22 -+ 0.65 1.41 • 0.69 Bone 2.41 _+ 1.12 2.45 +_ 1.76 1.88 • 0.87 1.78 • 0.84 Bone marrow 2.65 • 1.98 2.12 • 1.67 1.99 • 1.05 1.65 • 0.35 Skin 1.66 • 0.86 1.98 • 1.46 1.58 • 1.19 1.36 • 0.71 Brain 0.0004 • 0.0001 0.02 +_ 0.01 0.02 - 0.01 0.02 -+ 0.01 Blood 38.2 • 5.19 31.22 • 5.57 19.24 • 6.60 24.02 • 9.87 Oxygenator - - - - 0.16 • 0.07 0.29 • 0.26 Filter - - - - 0.04 • 0.01 0.04 • 0.02

The mean and standard deviation values of the percent of total In-ill-labeled platelet radioactivity retained in the viscera, hollow-fiber oxygenator (Univox), and arterial filter (AF 1025) in unoperated and sham-operated controls and in two CPB groups of Yorkshire pigs. CFB = cardiopulmonary bypass; RP - reperfusion 90 minutes.


Table 2. Biodistribution (percent of injected dose) of autologous indium-Ill-labeled neutrophils in four groups of Yorkshire pigs: Unoperated control, sham-operated control, 90 minutes of cardiopulmonary bypass and 90 minutes of reperfusion, and 180 minutes of cardiopulmonary bypass operation with oxygenator (Univox) and arterial filter (AF1025)

Organs and Unoperated control Sham-operated control CPB CPB-RP tissues (180 min, n = 6) (180 rain, n = 6) (180 rain, n = 6) (90 CPB, 90 rep., n = 6)

Lungs 23.66 _+ 5.70 18.49 ___ 4.86 48.31 • 3.53 31.38 • 4.78 Liver 18.94 • 5.27 24.69 --- 3.55 21.82 _ 11.11 18.62 • 11.91 Spleen 4.91 • 0.97 3.75 • 1.31 2.72 • 0.92 3.86 • 2.38 Kidneys 0.54 • 0.16 1.25 • 0.37 0.51 • 0.18 0.97 • 0.32 Heart 0.54 • 0.16 0.48 • 0.14 0.30 • 0.08 0.48 • 0.05 Muscle 3.65 -+ 1.06 16.95 • 2.92 4.46 • 2.59 4.80 • 3.14 Fat 0.54 -+ 0.11 2.21 ___ 1.36 0.70 • 0.13 1.58 • 1.14 Bone 1.31 _+ 0.71 0.45 • 0.24 0.82 • 0.49 0.98 • 0.64 Bone marrow 2.06 • 1.7 0.94 • 0.55 0.65 • 0.14 0.88 • 0.31 Brain 0.022 • 0.006 0.021 • 0.012 0.02 • 0.01 0.02 • 0.001 Blood 6.81 • 1.92 16.95 • 2.94 4.69 --- 1.96 4.07 • 2.52 Oxygenator - - - - 2.71 • 2.01 0.65 _+ 0.38 Filter - - - - 0.37 • 0.15 0.43 • 0.24

The mean and standard deviation values of the percent of total In-ill-labeled neutrophil radioactivity retained in the viscera, hollow-fiber oxygenator (Univox), and arterial filter (AF 1025) in unoperated and sham-operated controls and in two CPB groups of Yorkshire pigs. CPB = cardiopulmonary bypass; RP = reperfusion 90 minutes.

trophils sugges ted tha t the neutrophil interaction with the bi0materials is similar to tha t of plasma proteins (Figure 3C). Neutrophi l deposition is independent of species differences and shear rate. Neutrophil kinetics reached a s teady state almost instantaneously in the oxygenator ; on the o ther hand, they increased with a low slope in the arterial filter [28]. The radioac- t ivi ty level of t rapped and circulating neutrophils in A F during CPB was 2 -6 t imes that of the oxygenator , showing a higher re tent ive capacity in the screen mesh. Biodistribution depends on the time and route of injection of I n - I l l - l a b e l e d neutrophils. When the neutrophils were injected th rough the r ight atrium, 48 __ 4% and 22 ___ 11% of the injected neutrophils marginated to lungs. On the o ther hand, when they were injected th rough the left atrium, 17 _+ 5% and 30 -+ 7% of the injected neutrophils marginated to the liver, sugges t ing that these two organs in pigs have a similar capacity for neutrophil retention; neutrophil re tent ion depends on the first perfused and extract ing organ [28].

The large size, low deformabili ty, and receptor- mediated adhesion of neutrophils are responsible for microvascular obstruction and retent ion in organs. The distribution of platelets (Table 1) during CPB a n d after reperfusion was different from that of neutrophils (Table 2). About 5%, 19%, 5%, and 0.0004% of injected platelets marginated to the lung, liver, skeletal muscle, and brain in the unopera ted control group; 18%, 26%, 4%, and 0.02% of injected platelets marginated to the lung, liver, skeletal muscle, and brain in the sham-opera ted group. About 0.16% and 0.29% of adherent platelets were found in the oxygenator and 0.04% in the arterial filter at 180 minutes of CPB and

reperfusion, respectively. A scintiphoto of platelet distribution on segments of oxygenator and arterial filter is shown in F igure 3A. This platelet distribution was dependent on blood flow, the duration of flow, and the type of oxygena tor and was significantly al tered during oxygenat ion with hollow-fiber intralurninal and extraluminal flow oxygenators . Since the surface areas are different, the normalized values of %ID, platelet, and neutrophil densi ty were calculated per unit area of the surfaces of the oxygenator , arterial filter, and lung.

During CPB, neutrophils are activated by plasma proteins adsorbed to biomaterials in the CPB circuits. The results of neutrophil re tent ion on the oxygenator and arterial filter with radiolabeled neutrophils are shown in Tables 2 and 3. Our previous studies indicated tha t adhesion of neutrophils from the dog, pig, and human to biomaterials is similar, indicating tha t the porcine model provides clinically relevant results tha t are comparable with human patients during CPB [13,14]. The adherent neutrophil on the oxygenator and arterial filter, af ter saline wash, accounted for 1-3% of the total circulating neutrophils. The adherent neutrophil was measured before and after a saline rinse of the oxygena tor and arterial filter.

About 20% of the total neutrophils in the oxygenator and filter af ter a 3-hour run were lost after a rinse with 1000 ml of isotonic saline, suggest ing that 80% of the neutrophils were adherent to the hollow fibers and Dacron mesh. The calculated shear ra te for an intraluminal flow of 3000 ml/min is 1158/sec (fiber di- amete r -- 200 ~m and number of fibers = 55,000). Due to branching of the blood flow through multiple fibers, the Reynolds number is low, indicating the low

206 Dewanjee et al.

Table 3. Retention (percent of injected dose: ID %/m 2) of platelets and neutrophils and their surface density (• 108/m ~) in four groups of Yorishire pigs: unoperated control, sham-operated control, 90 minutes of cardiopulmonary bypass and 90 minutes of reperfusion, and 180 minutes of cardiopulmonary bypass operation with oxygenator (Univox) and arterial filter (AF1025)

Platelets Platelets Neutrophils Neutrophils Perfused system CPB 180 min CPB-RP CPB 180 rain CPB-RP

% ID/m 2 Lung 0.13 • 0.05 0.17 • 0.08 0.48 • 0.04 0.31 • 0.05

(0.09 • 0.04) a (0.24 --_ 0.05) a (0.18 -- 0.02) b (0.18 --_ 0.05) b

Oxygenator 0.09 ~- 0.04 0.16 • 0.14 1.51 • 1.12 0.35 • 0.21 filter 0.15 • 0.06 0.16 • 0.07 1.47 • 0.60 1.70 • 0.97

Density ( • 108/m 2) Lung 35 • 18 45 • 51 8 • 2 15 • 9

(22 • 8) ~ (14 • 4.7) ~ (30 • 11) b (3.2 • 1.3) b

Oxygenator 24 • 10 27 • 25 20 • 15 2 -+ 1.5 filter 40 • 15 28 • 13 21 • 3 3 • 1

The mean and standard deviation values of the percent of total In-Ill-labeled platelet and neutrophil radioactivity and their surface density retained in the oxygenator, arterial filter (Univox), and lungs in unoperated and sham-operated controls and m two CPB groups of York- shire pigs. RP = 90 minutes cardiopulmonary bypass and 90 minutes reperfusion. aUnoperated control; bsham operated control.

probability of turbulence in or around the fiber lumen. The neutrophil density was almost uniform in the fibers and was high at the bottom of the oxygenator.

Unlike the platelet kinetics and retention, neutrophil deposition was not dependent on blood flow, the species of animals, and the duration of exposure. The porosity of membranes, geometry of blood pathways, hemodynamic properties, and species differences do not significantly affect neutrophil adhesion to biomaterials. We measured platelet thrombi in oxygenators (bubble, membrane, and hollow fiber) in the dog and pig models [3-11]. The bubble oxygenators with the polyurethane defoamer used in the dog model accounted for 20-25% of adherent platelets in the 3-hour CPB; on the other hand, the intraluminal-flow, hollow- fiber oxygenator accounted for 2-4% of platelets in the pig model and for 1-2% in the arterial filter. Hepa- rin bonding to circuit components did not reduce the level of adherent platelets [10]. On the other hand, neutrophil retention in the intraluminal flow oxygenator and filter varied in the range of 1-2%. Neutrophil retention in the extraluminal flow oxygenator increased from 0.65% to 2.71% at 90 and 180 minutes of CPB. The extraluminal flow oxygenator (Univox) reduced the level of platelets and neutrophils significantly to 0.2-0.3% and 2-4%, respectively.

Neutrophil retention in the AF is lower, about 0.4-0.7% of the total neutrophils. Circulating platelets decreased significantly with CPB time; on the other hand, the circulating neutrophils remained at the low level of 4%. Reperfusion reduced the level of neutrophils and increased that of platelets to the lung (p < 0.05). A higher level of neutrophil margination to the skeletal muscle bed was observed in the sham-

operated group (p < 0.001). Further in vivo studies are in progress for quantification of neutrophil retention in the hollow-fiber oxygenator and of neutrophil trapping in different types of arterial filter to evaluate the overall detrimental effects of an inflammatory insult on ischemic organ damage induced by platelet era- boli during CPB [28].

After disruption of the normal architecture of a blood vessel, blood, with its attendent components of coagulation factors and circulating cellular elements, contacts tissues from which it is normally separated. Many tissues contain a specific protein t issue factor , inserted into the plasma membrane of their compo- nent cells [19]. Tissue factor always exists in an active state and forms a procoagulant enzyme complex with factor VII, initiating the proteolytic cascade, re- sulting in clot formation at the site of the damaged tissues. Most of the enzymes in blood require a cofac- tor to express their procoagulant activity. The cofactors, tissue factors (TF), and factors V and VI I I are localized to the plasma membrane. Factors V and VI I I undergo limited proteolysis; the product factors Va and VI I I a exhibit high affinity for cell membranes. The activated clotting factors--VIIa , Xa, IXa, and tissue factor--form tight tertiary complexes consisting of membrane-bound cofactors and enzymes.

Most previous studies focused on the activation and exhaustion of circulating cells and on activation of plasma proteins. Only with labeled platelets and neutrophils could the desired endpoints of the rate of transport and the absolute value of their fate in the blood, prostheses, organs, and connective tissues be determined. Barstad et al. recently demonstrated that tissue factor from monocytes mediates factor


V I I a - d e p e n d e n t t h r o m b u s format ion by an in vi t ro e x p e r i m e n t us ing a flow chamber [31]; t hey obse rved a 10-fold inc rease in a d h e r e n t t h rombus and fibrin de- posit ion. This level could be comple te ly blocked by an t i -TF ant ibody. The roles of p la te le ts , neut rophi l s , and c lot t ing factors a re i n t e r tw ined in this complex pa thobio logy dur ing CPB.

F u r t h e r s tudies a re in p r o g r e s s on the role of t i ssue factor and t u m o r necrosis fac tor b inding p ro te in ( T N F B P ) on p la te l e t and neut rophi l ac t iva t ion and the i r h igh- level r e t en t i on in the lung, hear t , and bra in a f te r thoraco tomy. In addit ion, we a re also eva lua t ing the role of t e m p e r a t u r e , an t icoagulan t (hepar in vs. r ecombinan t hirudin) and pumps (rol ler vs. centr i fugal pump) in o r d e r to opt imize CPB to reduce CPB- re l a t ed complications. By ident i fy ing the pro te ins in the s ignal ing p a t h w a y s [33] ac t iva t ing p la te le t s and neu t rophi l s dur ing CPB, we should be able to in ter - vene pharmacologica l ly to reduce the i r adver se effects.

A c k n o w l e d g m e n t s

The support of NHLBI (HL47201), a Shannon Award, the Bax- ter Healthcare Corporation, Department of Energy grant DE- FG05-88ER60728, and the Florida High Technology and Indus- try Council are gratefully acknowledged. The oxygenators and arterial filters were kindly provided by Dr. Li-Chien Hsu of Bentley Laboratories, Irvine, California. The authors appreci- ate the assistance of Ms. Sherry Armstrong, Mr. Jason Freed, and Mr. Joe Carvalho (CPI Inc.) with the roller pump; the tech- nical assistance of Mr. V. Srinivasan and Mr. D. De; and Mr. Bob Burke for producing the photographs.

R e f e r e n c e s

1. Kirklin JK, Kirklin JW. Cardiopulmonary bypass for cardiac surgery. In: Sabiston DC, Spencer FC, eds. Surgery of the Chest. Philadelphia: WB Saunders, 1990:1107-1125.

2. Woodman RC, Harker LA. Bleeding complications asso- ciated with cardiopulmonary bypass. Blood 1990;76:1680- 1697.

3. Reynolds LO, Mohammed SF, Solen KA, et al. Light scat- tering detection of microemboli in an extracorporeal LVAD bovine model. Trans ASAIO 1990;36:M518-M521.

4. Kurusz K. Arterial line filtration during cardiopulmonary bypass: history and controversy. Proc Am Acad Cardiovasc Pe~fus 1983;4:76-82.

5. Guidoin RG, Kenedi RM, Galleti P, et al. Thrombus formation and microaggregate removal during extracorporeal membrane oxygenation. J Biomed Mater Res 1979;13: 317-335.

6. Dewanjee MK, Tago M, Josa M, et al. Quantification of platelet retention in aoi"cocoronary femoral vein bypass graft in dogs treated with dipyridamole and aspirin. Circu- lation 1984;69:350-356.

7. Dewanjee MK, Vogel SR, Peterson KA, et al. Quantitation of platelet consumption on oxygenator and stabilization of platelet membrane with prostacyclin and ibuprofen during

cardiopulmonary bypass surgery in dogs. Trans ASAIO 1981;27:197-202.

8. Peterson KA, Dewanjee MK, Kaye MP. Fate of indium-ill- labeled platelets during cardiopulmonary bypass performed with membrane and bubble oxygenators. J Thoracic Cardiovasc Surg 1982;84:39-43.

9. Dewanjee MK, Palatianos GM, Kapadvanjwala M, et al. Increase of intraplatelet free calcium ion during extracorporeal circulation with a hollow fiber oxygenator, arterial ill- ter and dynamics of platelet thrombosis on oxygenator and filter in a pig model. Trans ASAIO 1990;36:M668-M671.

10. Palatianos GM, Dewanjee MK, Smith W, et al. Platelet preservation during cardiopulmonary bypass with Iloprost and heparin-coated surface. Trans ASAIO 1991;37:620-622.

11. Dewanjee MK, Rao SA, Didisheim P. Indium-ill tropolone, a new high affinity platelet label: Preparation and eval- uation of labeling parameters. J Nucl Med 1981;22:981-987.

12. Chowdhury S, Brown ML, Dewanjee MK, Forstrom LA. Labeled polymorphonuclear leukocytes: A comparison of methodology. Nucl Med Biol 1988;15:511-514.

13. Dewanjee MK, Kapadvanjwala M, Zilleruelo G, Sfakianakis GN, Sanchez A. Quantitation of comparative thrombogenicity of dog, pig and human platelets in a hemodialyzer. ASAIO J 1992;88-90.

14. Dewanjee MK. Cardiac and vascular imaging with labeled platelets and leukocytes. Semin Nucl Med 1984;14:154-187.

15. Haylor GDS, Hickey S, Bell G, PeiJM. Membrane oxygenators: Influence of design on performance. Perfusion 1995;9: 173-90.

16. Wickramsinghe SR, Semmens M J, Cussler EL. Mass trans- fer in various hollow-fiber geometries. J Membr Sci 1992; 69:235-250.

17. Anderson DC. The role of ~ integrins and intercellular adhesion molecule type 1 in inflammation. In: Granger DN, Schmid-Schonbein GW, eds. Physiology and Pathophysiol- ogy of Leukocyte Adhesion. New York: Oxford University Press, 1995:1-42.

18. Salzman EW, Merrill EW, Kent KC. Interaction of blood with artificial surfaces. In: Colman RW, Hirsh J, Marder V J, Salzman EW, eds. Hemostasis and Thrombosis: Basic Principles and Clinical Practice. Philadelphia: JB Lippin- cott, 1994:1469-1485.

19. Nemerson Y. The tissue factor pathway of blood coagulation. In: Colman RW, Hirsh J, Marder V J, Salzman FW, eds. Hemostasis and Thrombosis: Basic Principles and Clinical Practice. Philadelphia: JB Lippincott, 1994:81-93.

20. Dewanjee MK, Solis E, Mackey ST, et al. Quantification of regional platelet and calcium deposition on pericardial tissue valve prostheses in calves and effect of hydroxyethylene diphosphonate. J Thorac Cardiovasc Surg 1986;92:337-348.

21. Dewanjee MK, Fuster V, Rao SA, Forshaw PL, Kaye MP. Noninvasive radioisotopic technique for detection of platelet deposition on mitral valve prostheses and quantification of visceral microembolism in dogs. Mayo Clin Proc 1983;58: 307-314.

22. Schoephoerster RT, Oynes F, Nunez G, Kapadvanjwala M, Dewanj ee MK. Effects of local geometry and fluid dynamics on reg~ional platelet deposition on artificial surfaces. A,~e- rioscler Thromb 1993;13:1806-1813.

23. Dietrich WD, Dewanjee S, Prado R, Watson BD, Dewanjee MK. Transient platelet accumulation in the rat brain after common carotid artery thrombosis: An rain-labeled platelet study. Stroke 1993;24:1534-1540.

24. Dietrich WD, Alonso O, Busto R, Prado R, Dewanjee S, Dewanjee MK, Ginsberg MD. Widespread hemodynamic

208 Dewanjee et at.

depression and focal platelet accumulation after fluid per- cussion brain injury: A double label autoradiographic study in rats. J Cereb Blood Flow Metab, 1996;16:481-489.

25. Craddock PR, Fehr J. Brigham KL, Kronenberg RS, Jacob HS. Complement and leukocyte-mediated pulmonary dys- function in hemodialysis. N Engl J Med 1977;296:769-774.

26. Stahl N, Yancopoulos GD. The alphas, betas, and kinases of cytokine receptor complexes. Celt 1993;74:587-590.

27. Lasky LA. Combinatorial mediators of inflammation? Curr Biology 1993;3:366-368.

28. Dewanjee MK, Palatianos GN, Kapadvanjwala M, et al. Neutrophil dynamics and retention in lung, oxygenator and arterial filter during cardiopulmonary bypass (CPB) in pig model. ASAIO J 1994;40:M547-M553.

29. Rot A. Endothelial cell binding of NAP-1/IL-8; role in neutrophil. Imrnunol Today 1992;13:291-294.

30. Lutticken C, Wegenka UM, Yuan J, et al. Association of transcription factor APRF and protein kinase Jakl with the interleukin-6 signal transducer gp130. Science 1994;263: 89-92.

31. Barstad RM, Hamers MJAG, KieI~lf P, Westvik A-B, Sa- kariassen KS. Procoagulant human monocytes mediate tissue factor/factor VIIa-dependent platelet thrombus formation when exposed to flowing nonanticoagulated human blood. Atheroscteros Thrombos Vasc Biol 1995;15:11-16.

32. Dewanjee MK, Kapadvanjwala M, Mao WW, et al. Higher blood flow window of reduced thrombogenicity and ac- ceptible fragmentation in a hollow fiber hemodialyzer. ASAIO J 1993;39:M363-M367.

33. Clark EA, Brugge JA. Integrins and signal transduction pathways: The road taken. Science 1995;268:233-239.

thoracotomy thrombi-hrt lung-machine-jtt-1996

Health & Medicine