Ann Thorac Surg 2002;73:1101-1106
© 2002 The Society of Thoracic Surgeons
Original article: general thoracic
The effect of adhesion molecule blockade on pulmonary reperfusion injury
Adrian J. Levine, FRCSa,
Karen Parkesa,
Stephen J. Rooney, FRCSa,
Robert S. Bonser, FRCS*a
a Cardiothoracic Surgical Unit, Queen Elizabeth Hospital, Edgbaston, Birmingham, United Kingdom
* Address reprint requests to Dr Bonser, Cardiothoracic Surgical Unit, Queen Elizabeth Hospital, Edgbaston, Birmingham, B15 2TH, United Kingdom
e-mail: r.s.bonser{at}bham.ac.uk
Presented at the Poster Session of the Thirty-sixth Annual Meeting of The Society of Thoracic Surgeons, Fort Lauderdale, FL, Jan 31–Feb 2, 2000.
 |
Abstract
|
|---|
Background. Selectins are the molecules involved in the initial adhesion of the activated neutrophil on pulmonary endothelium. We investigated the efficacy of selectin blockade in a selective (monoclonal antibody RMP-1) and nonselective (Fucoidin) manner in pulmonary reperfusion injury.
Methods. Groups of six rat lungs were flushed with University of Wisconsin solution then stored at 4°C for 4 hours. They then underwent sanguinous reperfusion for 30 minutes during which functional measures (gas exchange, pulmonary artery pressure, and airway pressure) of lung performance were made. After reperfusion we estimated their capillary filtration coefficient (Kfc units g/cm water/minute/g wet lung tissue) using a gravimetric technique. Four groups were studied: group I had no reperfusion, group II had 30 minutes of reperfusion, group III had infusion of 20 mg/kg Fucoidin before reperfusion, and group IV had infusion of 20 µg/mL RMP-1 before reperfusion.
Results. Reperfusion injury was found between groups I and II by an increase in capillary filtration coefficient (1.048 ± 0.316 to 3.063 ± 0.466, p < 0.01). Groups III and IV had a significantly lower Kfc than group II (0.967 ± 0.134 and 1.205 ± 0.164, respectively, p < 0.01). There was no significant functional difference between groups II, III, and IV.
Conclusions. Reperfusion-induced hyperpermeability was ameliorated by selective (RMP-1) and nonselective (Fucoidin) selectin blockade.
|
Introduction
|
|---|
Primary graft failure is responsible for high early mortality rates after lung transplantation. It manifests as a mixed vascular and endothelial injury causing increased vascular resistance, pulmonary edema, and impaired gas exchange and represents the sequelae of reperfusion injury. The neutrophil plays a central role in the etiology of reperfusion injury in all organ systems. During reperfusion, activated circulating neutrophils adhere to endothelium. After adhesion neutrophil transmigration occurs followed by release of free radicals and enzymes responsible for further tissue injury. Selectins (P and E located on the endothelium, L on the neutrophil) are the adhesion molecules involved in the initial adhesion of the activated neutrophil onto endothelium. The importance of selectin-mediated adhesion has been shown in a number of studies [1–11], but many of those studies relied on the use of monoclonal antibodies with only marginal clinical relevance.
Fucoidin is a nontoxic oligosaccharide derived from seaweed; it blocks the function of the selectins and has been shown to reduce reperfusion injury [12–14]. Several studies of adhesion molecule blockade have investigated warm ischemic insults or long cold ischemic periods, and it is not known whether the benefits of such interventions are applicable to the shorter cold ischemic times (4 to 6 hours) seen in clinical practice [15]. Thus, in this study we investigated the efficacy of selectin blockade by a monoclonal antibody to P-selectin (RMP-1) and Fucoidin on pulmonary functional and endothelial markers after a 4-hour hypothermic ischemic injury.
|
Material and methods
|
|---|
Isolated rat lung grafts were reperfused in a circuit using a parabiotic animal [16]. During reperfusion functional measures of graft performance (gas exchange, airway pressure, and pulmonary artery pressure) were made, and at the end of the reperfusion period the capillary filtration coefficient [17] and wet:dry ratios were assessed by dessication (Pearce Edwards Freeze Drier EPTD3; Edwards, Sussex, UK).
Male Sprague-Dawley rats (Charles River Laboratories, Kent, United Kingdom) weighing 450 to 600 g were used and received humane care in compliance with the United Kingdom Government’s Animals (Scientific Procedures) Act of 1986. In all procedures terminal anesthesia was used. Halothane and oxygen inhalation and intraperitoneal ketamine and metatomidime were used for anesthetic induction and maintenance, respectively. Animal lungs and isolated lung grafts were ventilated using Harvard rodent ventilators (Harvard Apparatus, Kent, United Kingdom).
Graft harvest
After anesthesia, tracheostomy was performed with a 13-gauge venous cannula, and animals were ventilated with room air at a tidal volume of 10 mL/kg, a rate of 70 breaths per minute, and 2 cm of positive end-expiratory pressure. Median sternotomy was done followed by division of pleura and pericardium and retraction of the thymus. The aorta and pulmonary artery were then encircled with a ligature passed through the transverse sinus. Heparinization (1000 µg/kg) was performed through the inferior vena cava, which was then ligated to diminish venous return. An incision was made in the right ventricular outflow tract, and a spiggoted arteriotomy cannula was advanced into the pulmonary artery and secured with the transverse sinus ligature. The left ventricular apex was then amputated to allow venting. University of Wisconsin solution (DuPont Pharmaceuticals, Letchworth Garden City, United Kingdom) was then infused at a dose of 60 mL/kg at 4°C under 20 cm H2O pressure. Ventilation continued while the flush was completed, at which point the tracheostomy tube was clamped with the lungs fully inflated. The spigotted cannula was then changed for an appropriately primed and amputated triple-lumen catheter. The heart-lung block was then excised and stored in University of Wisconsin solution at 4°C for 4 hours. Control blocks then had a second amputated triple-lumen catheter placed in the left atrium through the ventriculotomy. Reperfusion blocks had this cannula inserted after their period of reperfusion.
Reperfusion
Support animals were anesthetized and ventilated and placed on a homeoiothermic warming blanket. After median sternotomy and pleural and pericardial division, the left superior vena cava was cannulated with an 18-gauge cannula. The right superior vena cava was cannulated with a 16-gauge cannula which was passed through the right atrium into the suprahepatic inferior vena cava. Deoxygenated blood drawn through the right superior vena cava cannula was delivered hydrostatically into the pulmonary artery of the donor block through the triple-lumen catheter. The blood then drained through the left ventriculotomy and was collected and returned by a peristaltic pump (Harvard Apparatus, Kent, United Kingdom) by way of the left superior vena cava cannula to the right atrium of the support animal (Fig 1).
The support animal and donor block were ventilated in the same manner as the donor animal. The extracorporeal circuit consisted of an inner blood carrying tube encased in an outer warming jacket with a counter-current 38°C water circulation circuit. The lung block was mounted in a plexiglass perfusion chamber with an outer warming jacket. The circuit was primed with heparinized blood from another animal and maintained free of air and bubbles.
Measurements
Graft effluent was sampled at 2, 5, 10, 15, 20, and 30 minutes for gas tension. The pulmonary artery pressure was measured through a limb of the triple-lumen catheter, as was the airway pressure through the tracheostomy tube.
Capillary filtration coefficient
The permeability of capillary walls allows the forces acting on either side to transport water across them. There are two hydrostatic and two osmotic pressures involved in this transfer. The two hydrostatic forces are the hydrostatic pressure in the capillary (Pc) and the hydrostatic pressure of the interstitial fluid (Pf). The two osmotic pressures are the plasma osmotic pressure (COPp) and the interstitial fluid oncotic pressure (COPt). However 
C, the difference between COPp and COPf, must also take into account the permeability of the cell membrane to proteins as delineated by the reflection coefficient 
d. Thus the true net oncotic pressure is the product of C and d. From these it is possible to derive the Starling-Landis equation correlating the flow rate of water (Jv) with the capillary filtration coefficient(Kfc) and the net pressure difference across the capillary
We modified the method of Drake and colleagues [17] to assess Kfc. This is a gravimetric technique in which Kfc is calculated from the plot of the rate of lung weight gain as a function of hydrostatic stress. Using a nonhemic but physiologic, iso-osmotic perfusate of L15 (Sigma Chemicals, United Kingdom) and fetal calf serum in the ratio of 9:1, a hydrostatic stress was applied to the isolated lung under test by increasing pulmonary perfusion pressure. On application of a known hydrostatic pressure, there is initially a short period of rapid weight gain attributable to recruitment and cardiac filling. Subsequently there is a slow weight gain phase; by extrapolating the slope of this rate of weight gain to time zero we can estimate the initial rate of fluid flux. The heart-lung block was retrieved from the storage area and the postcaval lobe identified, ligated, and removed for analysis of wet:dry ratio. The block was suspended (from a force transducer [Harvard Apparatus, Kent, United Kingdom]) with continuous weight monitoring by pen recorder (Harvard Apparatus, Kent, United Kingdom) through a 3-gauge stainless steel wire passed through the esophagus, and ventilation was maintained as previously described. The pulmonary artery and the left atrial cannulas were connected in turn. A pressure of 4 mm Hg was applied to the arterial cannula by increasing the L15 and fetal calf serum reservoir. The circuit was considered equilibrated when the left atrial pressure matched the pulmonary artery pressure thus equalling the pulmonary capillary pressure. At this point the arterial pressure was increased to 5 mm Hg and the weight gain trace recorded for 10 minutes. The arterial pressure was finally increased to 7 mm Hg and the trace recorded for a further 10 minutes. The blocks were then taken down and the parenchymal lung tissue excised for estimation of wet:dry ratio by freeze drying.
Experimental protocol
Four groups of lung blocks (n = 6) were used. Group I (control) underwent flush, excision, storage for 4 hours and measurement of filtration coefficient. Group II (normal reperfusion) underwent flush, excision, storage, and 30 minutes of reperfusion and then measurement of filtration coefficient. Group III (Fucoidin) underwent flush, excision, storage, and infusion Fucoidin (20 mg/kg body weight) (Sigma Chemical, Dorset, United Kingdom) through the pulmonary artery catheter before 30 minutes of reperfusion. After this they underwent estimation of their filtration coefficient. Group IV (RMP-1) underwent flush, excision, storage, and then infusion of of 20 µg/mL RMP-1 [18] before 30 minutes of reperfusion. After this they underwent estimation of their filtration coefficient.
Statistical analysis
Data are expressed as mean ± standard error of the mean. Group means of final measurements were compared by analysis of variance (Newman-Keuls). Analysis of repeated variables (functional data) was made by multivariate analysis of variance testing. A p value of less than 0.05 was considered significant. All analyses were carried out using SPSS software (SPSS Inc, Chicago, IL).
|
Results
|
|---|
Capillary filtration coefficient
There was a significant increase in capillary filtration coefficient between groups I and II (1.048 ± 0.316 to 3.063 ± 0.466, p < 0.01) demonstrating reperfusion-induced hyperpermeability. This effect was ameliorated by Fucoidin (0.967 ± 0.134, p < 0.01) and RMP-1 reperfusion (1.056 ± 0.173, p < 0.1) (Fig 2).
Functional measures
The functional data (Figs 3 and 4)
suggest that there was no significant difference with respect to functional measures between the normally reperfused, the Fucoidin reperfused, and the RMP-1 reperfused groups.
View larger version (20K):
[in this window]
[in a new window]
|
Fig 3. Functional changes associated with Fucoidin-assisted reperfusion. Values are shown as mean ± standard error of the mean. Dashed line = 30 minute reperfusion; solid line = Fucoidin. (AWP = airway pressure; PAP = pulmonary arterial pressure; PO2 = effluent oxygen tension.)
|
|
View larger version (17K):
[in this window]
[in a new window]
|
Fig 4. Functional changes associated with RMP-1-assisted reperfusion. Values are shown as mean ± standard error of the mean. Dashed line = 30 minute reperfusion; solid line = Mab P-selectin. (AWP = airway pressure; PAP = pulmonary arterial pressure; PO2 = effluent oxygen tension.)
|
|
Wet:dry ratio
There was no significant difference between the wet:dry ratios of group II (5.653 ± 0.124 [mean ± standard error of the mean]), group III (5.642 ± 0.189), and group IV (4.997 ± 0.479).
|
Comment
|
|---|
We found that endothelial permeability increased after 30 minutes of reperfusion after 4 hours of hypothermic preservation. Selectin blockade, both specific (RMP-1) and nonspecific (Fucoidin), offered protection against endothelial injury.
The neutrophil is important in the generation and modulation of reperfusion injury. Consequently, attempts to ameliorate neutrophil-related reperfusion injury have concentrated on leukocyte depletion and blocking neutrophil-endothelial adhesion. The classic model of neutrophil-endothelial interraction is that of a dynamic process made up of initial rolling (weak) adhesion followed by firm adhesion and finally diapedesis [19]. There is evidence that this simple model might not apply in the pulmonary microvasculature [20] because neutrophils are larger in diameter than many pulmonary microvessels, necessitating neutrophil deformation to allow transit. However, there is considerable evidence that much of the adhesive process is selectin dependent [19, 20] and occurs mostly in the pulmonary microvessels [21], although the most important selectin involved in this process is still debated [20].
The selectins are a family of carbohydrate-binding proteins found on vascular endothelium (P- and E-selectins), neutrohils (L), and platelets (P). All selectins have been shown to support neutrophil rolling under conditions of flow, but P-selectin seems to be the first expressed and the most important. Antiadhesion therapy aims to alter the reperfusion process by preventing the interactions between activated neutrophils and endothelium. Antiadhesion therapy can be considered to be specific (the use of monoclonal antibodies against specific adhesion molecules) or nonspecific (the use of soluble substances that saturate the adhesion molecules thereby preventing their interaction with their appropriate ligands). In the case of the selectins these ligands are high-affinity sialylated, fucosylated, and sulfated glycoproteins [22]. Fucoidin is a sulfated fucosylated polysaccharide derived from seaweed, which binds to both L- and P-selectin, and it has been found to block leukocyte rolling in a dose-dependent manner [23]. The actual mode of action is controversial. Handa and associates [24] suggested that sulfated glycans such as Fucoidin may interact with specific portions of the exocellular portion of the selectin molecules and thus induce conformational changes. Ley and colleagues [23] proposed that Fucoidin might inhibit neutrophil rolling through a nonspecific increase in charge density on the neutrophil surface.
Selective and nonselective selectin blockade has been examined in acute lung injury models associated with intestinal ischemia and reperfusion [1], cobra venom factor [5], and thermal skin injury [25], with varying effects. Selectin-mediated adhesion has also been shown to be important in pulmonary reperfusion injury. Naka and associates [8] have shown that homozygous mice lacking the P-selectin gene (P-selectin -/-) had reduced neutrophil infiltration, improved arterial oxygenation, and improved survival compared with P-selectin +/+ control mice after normothermic pulmonary reperfusion injury. Monoclonal antibodies against P-selectin uniformly have been shown to successfully ameliorate reperfusion injury. Moore and colleagues [3] showed that the increased microvascular permeability associated with a 45-minute period of warm ischemia and 30 minutes of reperfusion was attenuated by a P-selectin monoclonal antibody (PB1.3) with a coincident diminution in leucosequestration. Naka and associates [8] found that treatment with an anti-P-selectin monoclonal antibody in an isogenic rat lung transplant model (with a 6-hour cold ischemic time) was associated with reduced neutrophil sequstration and increased survival. Antibody against L-selectin (HRL3) showed no ability to attenuate the permeability change or leucosequestration [3]. In comparison L-selectin blockade in mesenteric postcapillary venules attenuated reperfusion-mediated permeability changes [4], and treatment with anti-E and anti-L-selectin (EL-246) improved survival in warm ischemic ovine pulmonary model [11].
Nonspecific selectin blockade can also ameliorate reperfusion injury. Brandt and colleagues [26], using a bolus of Sialyl-Lewis X on reperfusion, found reduced edema and infiltrate on x-ray in a isogenic rat lung transplant model. Reignier and associates [9], using a novel sulfated pentasaccharide (3'-sulfated Lewis[a]), investigated this further using a warm ischemia-reperfusion model and found reduced microvascular injury and decreased lung neutrophil accumulation. Sialyl-Lewis X analogue CY-1503 has also been shown to be protective in a cold ischemic canine lung transplantation model [10]; however, in that study CY-1503 was added to the preservation fluid (Eurocollins) and given on reperfusion. Fucoidin has been found to significantly affect neutrophil dynamics in the normal pulmonary circulation [21]. Intravital microscopy showed that Fucoidin reduced neutrophil rolling in pulmonary arterioles and venules, without affecting leucocyte adhesion, and also reduced the transit time through the alveolar capillaries. The reported differences in effective dose of Fucoidin to inhibit neutrophil rolling [23, 27] led us to select a dosage of 20 mg/kg. Our choice of single bolus application was influenced by reports of single bolus efficacy [21] in contrast to other infusion-based protocols [12]. The selection of RMP-1 dosage was based on knowledge of in vivo and in vitro properties of RMP-1. Although it was disappointing that neither RMP-1 nor Fucoidin conferred marked functional benefit, these effects might only become manifest as the period of reperfusion increases, a proposition not tested in this study. The approach of saturation blockade of adhesion molecules by ligand analogues could offer theoretical advantages over antiadhesive monoclonal antibody therapy in that sensitization to monoclonal antibodies can be avoided and nonselective blockade can be more efficacious in a multistage process. In summary, selective and nonselective selectin blockade caused a reduction in endothelial injury after cold ischemic preservation. The clinical potential of Fucoidin in ameliorating pulmonary reperfusion injury merits further investigation.
|
Acknowledgments
|
|---|
We acknowledge the generous contribution of Dr Alexander C. Issekutz (Dalhousie University, Nova Scotia, Canada) for supplying the RMP-1 and the help of Dr Peter Davies (University of Birmingham, UK) in the statistical analysis.
|
References
|
|---|
-
Carden D.L., Young J.A., Granger D.N. Pulmonary microvascular injury after intestinal ischemia-reperfusion: role of P-selectin. J Appl Physiol 1993;75:2529-2534.[Abstract/Free Full Text]
-
Gibbs S.A., Weiser M.R., Kobzik L., et al. P-selectin mediates intestinal ischemic injury by enhancing complement deposition. Surgery 1996;119:652-656.[Medline]
-
Moore T.M., Khimenko P., Adkins W.K., Miyasaka M., Taylor A.E. Adhesion molecules contribute to ischemia and reperfusion-induced injury in the isolated rat lung. J Appl Physiol 1995;78:2245-2252.[Abstract/Free Full Text]
-
Kurose I., Anderson D.C., Miyasaka M., et al. Molecular determinants of reperfusion-induced leukocyte adhesion and vascular protein leakage. Circ Res 1994;74:336-343.[Abstract/Free Full Text]
-
Mulligan M.S., Paulson J.C., De Frees S., et al. Protective effects of oligosaccharides in P-selectin-dependent lung injury. Nature 1993;364:149-151.[Medline]
-
Mulligan M.S., Polley M.J., Bayer R.J., et al. Neutrophil-dependent acute lung injury. Requirement for P-selectin (GMP-140). J Clin Investig 1992;90:1600-1607.
-
Mulligan M., Miyasaka M., Tamatami T., Jones M., Ward P. Requirements for L-selectin in neutrophil-mediated lung injury in rats. J Immunol 1994;xx:832-842.
-
Naka Y., Toda K., Kayano K., Oz M.C., Pinsky D.J. Failure to express the P-selectin gene or P-selectin blockade confers early pulmonary protection after lung ischemia or transplantation. Proc Nat Acad Sci USA 1997;94:757-761.[Abstract/Free Full Text]
-
Reignier J., Sellak H., Lemoine R., et al. Prevention of ischemia-reperfusion lung injury by sulfated Lewis(a) pentasaccharide. The Paris-Sud University Lung Transplantation Group. J Appl Physiol 1997;82:1058-1063.[Abstract/Free Full Text]
-
Schmid R.A., Yamashita M., Boasquevisque C.H., et al. Carbohydrate selectin inhibitor CY-1503 reduces neutrophil migration and reperfusion injury in canine pulmonary allografts. J Heart Lung Transplantation 1997;16:1054-1061.[Medline]
-
Steinberg J.B., Mao H.Z., Niles S.D., Jutila M.A., Kapelanski D.P. Survival in lung reperfusion injury is improved by an antibody that binds and inhibits L- and E-selectin. J Heart Lung Transplantation 1994;13:306-318.[Medline]
-
Miura T., Nelson D.P., Schermerhorn M.L., et al. Blockade of selectin-mediated leukocyte adhesion improves postischemic function in lamb hearts. Ann Thorac Surg 1996;62:1295-1300.[Abstract/Free Full Text]
-
Shimaoka M., Ikeda M., Iida T., et al. Fucoidin, a potent inhibitor of leukocyte rolling, prevents neutrophil influx into phorbol-ester-induced inflammatory sites in rabbit lungs. Am J Resp Crit Care Med 1996;153:307-311.[Abstract]
-
Gebb S.A., Graham J.A., Hanger C.C., et al. Sites of leukocyte sequestration in the pulmonary microcirculation. J Appl Physiol 1995;79:493-497.[Abstract/Free Full Text]
-
Hopkinson D.N., Bhabra M.S., Hooper T.L. Pulmonary graft preservation: a worldwide survey of current clinical practice. J Heart Lung Transplantation 1998;17:525-531.[Medline]
-
Hopkinson D.N., Odom N.J., Eloakley R.M., Au J., Hooper T.L. A technique for studying the function of an isolated, perfused and ventilated lung in a rat model. Lab Anim 1995;29:96-101.[Abstract/Free Full Text]
-
Drake R., Gaar K.A., Taylor A.E. Estimation of the filtration coefficient of pulmonary exchange vessels. Am J Physiol 1978;234:H266-H274.[Abstract/Free Full Text]
-
Walter U.M., Ayer L.M., Wolitzky B.A., et al. Characterization of a novel adhesion function blocking monoclonal antibody to rat/mouse P-selectin generated in the P-selectin-deficient mouse. Hybridoma 1997;16:249-257.[Medline]
-
Lefer A.M. Role of selectins in myocardial ischemia-reperfusion injury. Ann Thorac Surg 1995;60:773-777.[Abstract/Free Full Text]
-
Kubo H., Doyle N.A., Graham L., et al. L- and P-selectin and CD11/CD18 in intracapillary neutrophil sequestration in rabbit lungs. Am J Respir Crit Care Med 1999;159:267-274.[Abstract/Free Full Text]
-
Kuebler W.M., Kuhnle G.E., Groh J., Goetz A.E. Contribution of selectins to leucocyte sequestration in pulmonary microvessels by intravital microscopy in rabbits. J Physiol 1997;501:375-386.[Abstract/Free Full Text]
-
McEver R.P. Selectins [Review]. Curr Opin Immunol 1994;6:75-84.[Medline]
-
Ley K., Linnemann G., Meinen M., Stoolman L., Gaehtgens P. Fucoidin, but not yeast polyphosphomannan PPME, inhibits leukocyte rolling in venules of the rat mesentery. Blood 1993;81:177-185.[Abstract/Free Full Text]
-
Handa K., Neudelman E., Stroud M., Shizowa T., Hakomori S. Selectin GMP-140 (CD62; PADGEM) binds to sialosyl-Lea and sialosyl-Lex, sulfated glycans modulate this binding. Biochem Biophys Res Commun 1991;181:1223-1230.[Medline]
-
Mulligan M.S., Till G.O., Smith C.W., et al. Role of leukocyte adhesion molecules in lung and dermal vascular injury after thermal trauma of skin. Am J Pathol 1994;144:1008-1015.[Abstract]
-
Brandt M., Boeke K., Phillips M.L., Steinhoff G., Haverich A. Effect of oligosaccharides on rejection and reperfusion injury after lung transplantation. J Heart Lung Transplantation 1997;16:352-359.[Medline]
-
Kubes P., Jutila M., Payne D. Therapeutic potential of inhibiting leukocyte rolling in ischemia/reperfusion. J Clin Investig 1995;95:2510-2519.
This article has been cited by other articles:
|
|
|
|
 
H. Ichimura, K. Parthasarathi, A. C. Issekutz, and J. Bhattacharya
Pressure-induced leukocyte margination in lung postcapillary venules
Am J Physiol Lung Cell Mol Physiol,
September 1, 2005;
289(3):
L407 - L412.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
A. Kaminski, C. B. Pohl, C. Sponholz, N. Ma, C. Stamm, B. Vollmar, and G. Steinhoff
Up-Regulation of Endothelial Nitric Oxide Synthase Inhibits Pulmonary Leukocyte Migration Following Lung Ischemia-Reperfusion in Mice
Am. J. Pathol.,
June 1, 2004;
164(6):
2241 - 2249.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
