In beating heart coronary artery surgery, mechanical

Transcription

1 Coronary Endothelial Injury After Local Occlusion on the Human Beating Heart Herbert B. Hangler, MD, Kristian Pfaller, PhD, Herwig Antretter, MD, Otto E. Dapunt, MD, and Johannes O. Bonatti, MD Department of Cardiac Surgery and Institute of Anatomy and Histology, Leopold-Franzens-University, Innsbruck, Austria, and Städtische Kliniken, Oldenburg, Germany Background. Occlusion of coronary arteries during beating heart surgery bears the potential for mechanical trauma to the arterial wall with consequent endothelial injury. The aim of this study was to elucidate the effects of local occlusion on the beating heart in human coronary arteries. Methods. Coronary arteries of patients with dilated cardiomyopathy (n 7) or ischemic heart disease (n 10) undergoing heart transplantation were locally occluded after starting cardiopulmonary bypass. Immediately after excision of the diseased heart, the vessels were fixed. Unoccluded segments served as controls. Integrity of endothelial lining was observed with scanning electron microscopy. Results. Scanning electron microscopy revealed significantly more severe endothelial injury in the area of occlusion than in the adjacent, not manipulated control segments. In the region of local occlusion, plaque rupture was noted in three of 34 atherosclerotic vessel specimens, injury to side branches was evident in two of 44, and local microthrombus formation was evident in six of 44 samples. Conclusions. Local occlusion of human coronary arteries during beating heart coronary surgery may cause focal endothelial denudation, local microthrombosis, atherosclerotic plaque rupture, and injury to target vessel side branches. (Ann Thorac Surg 2001;71:122 7) 2001 by The Society of Thoracic Surgeons In beating heart coronary artery surgery, mechanical aids are required to achieve a clear and stabilized operative field enabling the surgeon to perform a technically appropriate coronary artery anastomoses. Therefore, in some instances, the target vessels are encircled with elastic silicone tapes or polypropylene sutures buttressed with pieces of silicone tubing and frapped with tourniquets to diminish native coronary blood flow. The sutures are usually placed deep into the myocardium to have a cushion of subjacent epimyocardial tissue protecting the posterior and lateral vessel wall against direct compression injury [1]. In contrast to these self-made snaring devices, commercially available tools such as the MyOcclude vessel occlusion device (Vascular Therapies, USSC, Elancourt, France) enable the surgeon to occlude coronary arteries with a standardized force and without the need to blindly underpass the target coronary artery. As shown in animal models, manipulation of coronary arteries during revascularization in off-pump surgery can lead to marked endothelial cell loss and local coronary dysfunction [2]. The aim of this investigation was to elucidate the effects of commonly used local coronary occlusion techniques on integrity of endothelial lining in human coronary arteries on the beating heart. Accepted for publication July 11, Address reprint requests to Dr Hangler, Department of Cardiac Surgery, Leopold-Franzens-University Innsbruck, Anichstrasse 35, 6020 Innsbruck, Austria; herbert.hangler@uibk.ac.at. Material and Methods Experimental Groups and Operative Technique Multiple coronary artery segments of patients undergoing orthotopic heart transplantation for dilated cardiomyopathy (group I, DCMP) (n 7) or ischemic coronary heart disease (group II, ICHD) (n 10) were occluded for a period of 15 minutes to simulate the average duration of a coronary anastomoses, immediately after starting cardiopulmonary bypass before the diseased recipient heart was explanted. The occlusion was carried out proximal as well as distal to the point where grafts are usually placed during coronary artery bypass grafting (middle third of left anterior descending coronary artery, right coronary artery before the Crux cordis, first to middle third of diagonal branch). The coronary arteries were occluded by encircling with either a 3/0 polypropylene suture, buttressed with a piece of silicone tubing (n 10), or a Teflon felt pledget (n 10) as abutment and lashed with a soft silicone tube. Coronary arteries were underpassed once with an elastic silicone loop (Ethiloop; Ethicon, Norderstedt, Germany) (n 10) passed through a tourniquet and frapped. Care was taken to have a pad of epimyocardial tissue included to avoid direct compression of the posterior and lateral coronary vessel wall. In addition, the MyOcclude vessel occlusion device (n 7) and Bulldog clamps ( ; Codman, Raynham, MA) The study was approved by the local human research committee, January 21, by The Society of Thoracic Surgeons /01/$20.00 Published by Elsevier Science Inc PII S (00)

2 Ann Thorac Surg HANGLER ET AL 2001;71:122 7 SNARE INJURY OF HUMAN CORONARIES 123 (n 7) were applied for local coronary occlusion. Uninstrumented segments adjacent to the occlusion site of the same coronary artery served as controls (n 27). After the occlusion period of 15 minutes, the aorta was crossclamped and the heart immediately excised. The period from end of occlusion to the start of pressure fixation of the coronary arteries was 9 3 minutes. Procurement of Tissue Instantly after excision of the diseased recipient heart, the coronary arteries were perfused for 10 minutes with 2.5% glutaraldehyde in 0.1 mol/l cacodylate buffer (ph 7.4) at a controlled pressure of 120 mm Hg through a perfusion cannula (P616; Stöckert, Munich, Germany) inserted into the left or right coronary ostium. In this way, the endothelium was fixed in situ at physiologic pressure before being further processed for scanning electron microscopy. Subsequently, the vessels were carefully dissected free from the adherent epicardial tissue in a no-touch technique under 12 magnification and cut transverse 5 mm apart from the occlusion area on each side. Control samples were taken from the adjacent not-instrumented areas of the same coronary artery. Preparation of the Coronary Arteries for Scanning Electron Microscopy (SEM) The coronary artery cylinders were cut longitudinally, pinned on cork plates, and postfixed in 1% Osmiumtetroxide (OsO 4 ), further dehydrated in a graded ethanol series, and subjected to critical-point drying (CPD 030; Bal Tec, Balzers, Lichtenstein). After drying, samples were mounted on specimen stubs using colloidal silver and coated with 15 nm gold (MED 020; Bal Tec). The entire endothelial surface of each specimen was examined with a Zeiss DSM 982 Gemini scanning electron microscope, operated at 5 kv. Histomorphology of the Endothelial Layer Was Classified Into Three Grades In grade I, the entire surface was covered by intact endothelial cells with a tight intercellular attachment (intact endothelial layer). In grade II, there were dehiscent intercellular junctions with isolated detachment of endothelial cells (minor endothelial injury). In grade III, there was an expanse of local endothelial denudation with the subendothelial tissue exposed (severe endothelial injury). Figure 1 depicts the three different categories of endothelial lining. Statistical Analysis All data were stored in a computerized database (MS Excel for Windows) and statistical assessments were performed using the SPSS for Windows statistical software package (SPSS Inc, Chicago, IL). Comparisons between groups were made using the 2 test or Fisher s exact test where appropriate. A p value less than 0.05 was considered to be significant. Continuous data are given as mean standard deviation. Results During preparation for SEM under 12 magnification in two of 44 manipulated coronary segments (DCMP and ICHD), injury to a septal branch of the left anterior descending coronary artery with perforation and a perivascular hematoma was evident. None of the control coronary artery segments exhibited endothelial damage greater than grade II. These minor changes of the coronary endothelial layer in atherosclerotic human coronary arteries have already been reported [3]. In contrast, 24 of 44 manipulated segments from both patient groups (DCMP and ICHD) showed grade III endothelial injury ( p 0.001). There was no significant difference in regard to grade III injury between instrumented coronary segments from patients with DCMP (14 of 23) and ICHD (10 of 21) (p NS). Table 1 lists the incidence of endothelial injury according to the five different occlusion methods applied. There was a trend towards a lesser occurrence of grade III injury when occlusion was performed with elastic silicone loops and the MyOcclude device. Occlusion of coronary arteries led to local micro-thrombus formation in six of 44 samples (Fig 2). In three specimens of patients with ICHD after local occlusion, rupture of an atherosclerotic plaque was encountered (Fig 3). Comment Local occluding maneuvers, during beating heart revascularization, are incriminated to induce lesions such as target coronary artery stenosis proximal or distal to the performed anastomosis [4 6] as well as septal myocardial infarction [7], septal branch right ventricular fistula [8], and distal embolization of atheromatous debris into the coronary circulation [9]. Integrity of the endothelial lining is an essential part in the equilibrium of the nonthrombogenic properties of the endothelial surface by inhibiting platelet function and coagulation [10]. When the subendothelial matrix is exposed to circulating blood elements, thrombosis may occur and result in early coronary artery or bypass graft closure. Higher postoperative platelet counts and a minor impaired platelet function, as well as a smaller decrease of circulating coagulation factors in off-pump coronary surgery than in on-pump surgery, with a procoagulant activity could boost local coronary artery thrombosis [11]. Therefore, platelet aggregation and thrombus formation in coronary artery areas denuded of endothelial cell coverage after external occlusion, as detected in this investigation, could be more at risk for thrombosis in off-pump than in on-pump surgery. We were surprised that in this series, despite full heparinization for cardiopulmonary bypass, six cases of local microthrombus formation were diagnosed. Because of this finding, we propose to routinely administer full heparin loading in off-pump procedures, keeping the activated coagulation time above 400 seconds with checks at 30- minute intervals. Other authors have suggested lower heparin doses [12]. At present, discussions are ongoing as

3 124 HANGLER ET AL Ann Thorac Surg SNARE INJURY OF HUMAN CORONARIES 2001;71:122 7 Fig 1. (A) Control specimen. Scanning electron micrograph depicts an overview of a coronary artery cut into halves longitudinally after pressure fixation with an intact endothelial surface. (B) Inset of A. Higher magnification with tight endothelial cell attachment completely covering the vascular surface, according to grade I classification. (C and D) Micrographs representing grade II classification of endothelial injury with dehiscent intercellular junctions and isolated detachment of endothelial cells (white arrow). (E) Expanse endothelial denudation with complete loss of endothelial cell coverage exposing the subendothelial tissue to the blood stream (white arrow). (F) Higher magnification of denuded area. to whether platelet-inhibiting drugs such as clopidogrel should be administered in the early postoperative period to counteract the procoagulant state in off-pump coronary artery bypass grafting. Perhaps a more vigorous postoperative anticoagulation and platelet-inhibiting pharmacotherapy could be effective. Adequate prospec-

4 Ann Thorac Surg HANGLER ET AL 2001;71:122 7 SNARE INJURY OF HUMAN CORONARIES 125 Table 1. Incidence of Endothelial Injury According to Occlusion Methods Applied Occlusion Method Classification of Endothelial Integrity Grade I/II Grade III Control (n 27) 27 0 DCMP (n 14) 14 0 ICHD (n 13) 13 0 Polypropylene suture: silicone abutment 3 7 (n 10) DCMP (n 8) 3 5 ICHD (n 2) 0 2 Polypropylene suture: pledget abutment 2 8 (n 10) DCMP (n 6) 1 5 ICHD (n 4) 1 3 Bulldog clamp (n 7) 3 4 DCMP (n 3) 1 2 ICHD (n 4) 2 2 Silicone loop (n 10) 7 3 DCMP (n 4) 3 1 ICHD (n 6) 4 2 MyOcclude (n 7) 5 2 DCMP (n 2) 1 1 ICHD (n 5) 4 1 Data are number of samples. DCMP dilated cardiomyopathy; disease. ICHD ischemic coronary heart tive randomized trials are needed to bring insight into these questions [13]. Furthermore, in the response to injury hypothesis [14], endothelial cell injury such as focal denudation or dysfunction of endothelium is considered to be a key event in the evolution of atherosclerosis by inducing growth factor secretion, and attachment of macrophages and monocytes. The loss of endothelial cell coverage because of external instrumentation, as demonstrated by our group, may be healed by regenerating endothelium. Nevertheless, regenerated endothelium is also dysfunctional, as these endothelial cells have lost some of their ability to release endothelium-dependent relaxing factors and are no longer able to prevent aggregating platelet-induced contraction [15]. Occurrence of vasospasm and atherosclerosis may be accelerated in these areas covered by regenerated endothelial cells. Up to now, experiments investigating the effects of coronary occlusion techniques have been preferably performed in normal pig coronary arteries [16] not burdened with preexisting atherosclerosis and chronic endothelial injury with vasomotor dysfunction [17] that could be aggravated by endothelial cell loss from mechanical manipulation. Perrault and colleagues found that intravascular devices create a significantly higher degree of functional damage compared with extravascular occlusion techniques. In regard to these animal experiments, one has to keep in mind the fact that there is already a difference in behavior between vessels of comparable size in the same individual, for instance, very little natural atherosclerotic disease in the internal thoracic arteries in contrast to a much higher degree in coronary arteries or the comparative rarity of naturally occurring atherosclerosis in many species [18]. Moreover, when snare sutures are placed in areas of coronary arteries with severe atherosclerotic disease and the circumferential tension on lipid-laden plaques exceeds its tensile strength, it possibly will rupture at its weakest point and expose atheromatous gruel, the most thrombogenic component of a plaque, to the blood stream. Acute coronary syndromes, depending on the extent of thrombus formation or embolization of atheromatous debris, may occur [19]. Another tool that facilitates constructing anastomosis in off-pump coronary surgery are intracoronary shunts that are carefully advanced into the proximal coronary Fig 2. (A) Scanning electron micrograph of an intraluminal local thrombus formation (arrow) in a coronary artery cylinder, cut open longitudinally in the region after local occlusion. (B) Inset of (A), higher magnification of local thrombus formation depicting red blood cells (*) trapped in a fibrin network based on the inner vascular wall.

5 126 HANGLER ET AL Ann Thorac Surg SNARE INJURY OF HUMAN CORONARIES 2001;71:122 7 Fig 3. (A) Scanning electron micrograph (overview) of the inner surface of a coronary artery: the local occlusion having been applied around the vessel in the region of an atherosclerotic plaque with disruption (arrows) of the vessel wall. (B) Inset of (A), higher magnification of the area with plaque cracks attributed to local occlusion with the snare-technique. lumen using a thumb forceps. Flow through the device affirms proper insertion and deairing. The opposite side of the shunt is then advanced into the distal part of the coronary artery so that at least partial blood flow is preserved and may prevent potential intraoperative ischemia with arrhythmias, ST segment elevations, or regional systolic dysfunction that occurs in as much as 40% of off-pump procedures [20]. On the other hand, introducing a device into the coronary lumen can be the cause of dissection or harm endothelial integrity with functional impairment of the coronary artery. Thus, intracoronary shunts could be a useful adjunct in beating heart revascularization. However, further research is required to investigate the impact of shunt insertion on coronary artery endothelial structure and function. In the current study, a trend towards a lesser incidence regarding denuding endothelial injury was found when coronary arteries were occluded using elastic silicone loops or the MyOcclude device. We therefore propose to use these tools rather than polypropylene sutures. Another advantage of the MyOcclude device is that there is no need to underpass the target coronary artery with the potential for septal branch injury of left anterior descending coronary artery. Limitations of the Study Limitations of our study are the differences in the techniques of occlusion between the two patient groups and the relatively small number of specimens with respect to the occluding methods. Conclusions From this study, we conclude that local occlusion of human coronary arteries by snaring sutures during beating heart coronary surgery can lead to injury of target coronary artery side branches, focal endothelial denudation, plaque rupture, and local micro-thrombus formation. Snaring of the target coronary artery distal to the arteriotomy should not be used. We recommend full heparin loading in off-pump coronary artery bypass procedures until there is sufficient data that lower heparin doses are safe and effective. We thank Karin Gutleben and Angelika Flörl for aid in preparation of the coronary artery samples for SEM. References 1. Calafiore AM, Di Giammarco G, Teodori G, et al. Left anterior descending coronary artery grafting via left anterior small thoracotomy without cardiopulmonary bypass. Ann Thorac Surg 1996;61: Perrault LP, Menasche P, Wassef M, et al. Endothelial effects of hemostatic devices for continuous cardioplegia or minimally invasive operations. Ann Thorac Surg 1996;62: Davies MJ, Woolf N, Rowles PM, Pepper J. Morphology of the endothelium over atherosclerotic plaques in human coronary arteries. Br Heart J 1988;60: Pfister AJ, Zaki SM, Garcia J, Mispireta LA, et al. Coronary artery bypass without cardiopulmonary bypass. Ann Thorac Surg 1992;54: Alessandrini F, Gaudino M, Glieca F, et al. Lesions of the target vessel during minimally invasive myocardial revascularization. Ann Thorac Surg 1997;64: Pagni S, Qaqish NK, Senior DG, Spence PA. Anastomotic complications in minimally invasive coronary bypass grafting. Ann Thorac Surg 1997;63:S Splittgerber FH, Minale C. Septal myocardial infarction: a complication of coronary artery stay sutures. Ann Thorac Surg 1993;56: Tanemoto KT, Kuroki K, Kanaoka Y, Murakami T. Septal branch right ventricular fistula: a complication in coronary artery snaring. Ann Thorac Surg 1999;68: Izzat MB, Yim AP, El Zufari MH. Snaring of a coronary artery causing distal atheroma embolization. Ann Thorac Surg 1998;66:

6 Ann Thorac Surg HANGLER ET AL 2001;71:122 7 SNARE INJURY OF HUMAN CORONARIES Boyle EM, Verrier ED, Spiess BD. The procoagulant response to injury. Ann Thorac Surg 1997;64:S Mariani MA, Gu YJ, Boonstra PW, Grandjean JG, van Oeveren W, Ebels T. Procoagulant activity after off-pump coronary operation: is the current anticoagulation adequate. Ann Thorac Surg 1999;67: Jansen EW, Grundemann PF, Borst C, et al. Less invasive off-pump CABG using a suction device for immobilization: the Octopus method. Eur J Cardiothorac Surg 1997;12: Ascione R, Lloyd CT, Gomes WJ, Caputo M, Bryan AJ, Angelini GD. Beating versus arrested heart revascularization: evaluation of myocardial function in a prospective randomized study. Eur J Cardiothorac Surg 1999;15: Ross R, Gloset JA. Atherosclerosis and arterial smooth muscle cell. Science 1973;180: Vanhoutte PM, Perrault LP, Vilaine JP. Endothelial dysfunction and vascular disease. In: Rubanyi GM, Dzau VJ, eds. The endothelium in clinical practice. New York: Marcel Decker, 1997: Chavanon O, Perrault LP, Menasche P, Carrier M, Vanhoutte PM. Endothelial effects of hemostatic devices for continuous cardioplegia or minimally invasive operations. Ann Thorac Surg 1999;63: Forstermann U, Mugge A, Alheid U, Haverich A, Frolich JC. Selective attenuation of endothelium-mediated vasodilation in atherosclerotic human coronary arteries. Circ Res 1988;62: Sims F. A comparison of structural features of the wall of coronary arteries from 10 different species. Pathology 1989; 21: Cheng G, Loree H, Kamm R, Fishbein M, Lee R. Distribution of circumferential stress in ruptured and stable atherosclerotic lesions. A structural analysis with histopathological correlation. Circulation 1993;87: Dapunt OE, Raji MR, Jeschkeit S, et al. Intracoronary shunt insertion prevents myocardial stunning in a juvenile porcine MIDCAB model absent of coronary artery disease. Eur J Cardiothorac Surg 1999;15: INVITED COMMENTARY Knowledge of the effects of coronary occlusion with different devices has become an important issue with the growing popularity of minimally invasive and beating heart off-pump coronary artery bypass (OPCAB) grafting surgery, which require local control of the target artery for stabilization and maintenance of a dry operative field. Optimal visualization is mandatory to maintain the high graft patency rates and long-term outcome expected from CABG surgery with cardioplegic arrest. Good visualization should obviously not be obtained at the detriment of local coronary injury, which can lead to occlusion or thrombus formation, as a worst case scenario, or iatrogenic coronary stenosis from intimal hyperplasia at a site of injury. Laboratory and clinical testing of devices used for ensuring hemostasis at the site of coronary occlusion is therefore mandatory to gain full understanding of the consequences of coronary artery manipulation at sites remote from the arteriotomy for anastomosis. Several groups, including our own, have chosen endothelial function studies as a sensitive marker of intimal injury that can predispose to spasm, thrombosis, and intimal hyperplasia. Others have used morphological evaluation of the endothelial coverage to assess vascular changes at the site of application of the devices. In this important study from Hangler and associates, the latter approach was used with scanning electron microscopy examination of the endothelial surface at the sites of local occlusion of coronary arteries on the beating heart under full heparinization and cardiopulmonary bypass with currently available devices. Although limitations inherent to the design of the experimental protocol, including an empty beating heart as opposed to a loaded one in OPCAB and use of full heparin dosage, may seamingly hamper the relevance of this work, this study raises several important concerns. First, currently used techniques may cause plaque fractures at the site of application, which are known from the cardiological literature to predispose to thrombosis. Considering the limited ability for surgeons to identify intraoperatively all but the most advanced plaques within the artery, plaque fracture remains a definitive threat during OPCAB. This emphasizes the need for minimal manipulation of the coronary artery distally to the anastomotic site to prevent compromise of the graft flow. Second, the observation of microthrombi formation at the site of local occlusion under full heparinization should spark greater effort in determining the optimal anticoagulation protocol including heparin dosage, optimal activated clotting time, reversal with protamine, and antiplatelet treatment in the perioperative period for OPCAB surgery. These studies are needed to ensure that progresses in surgical myocardial revascularization will be passed on safely to patients with coronary artery disease. Louis P. Perrault, MD, PhD Research Center Montreal Heart Institute 5000 Belanger St E Montreal, PQ HIT IC8 Canada lperrau@icm.umontreal.ca by The Society of Thoracic Surgeons /01/$20.00 Published by Elsevier Science Inc PII S (00)