Operations on the aortic arch involve sophisticated

Transcription

1 Surgical Management of Hemorrhage From Rupture of the Aortic Arch René Prêtre, MD, Nicolas Murith, MD, Dominique Delay, MD, and Tshibambula Kalonji, MD Cardiovascular Surgery, Department of Surgery, University Hospital Geneva, Geneva, Switzerland Background. Control of hemorrhage in patients with active bleeding from rupture of the aortic arch is difficult, because of the location of the bleeding and the impossibility of cross-clamping the aorta without interfering with cerebral perfusion. A precise and swift plan of management helped us salvage some patients and prompted us to review our experience. Methods. Six patients with active bleeding of the aortic arch in the mediastinum and pericardial cavity (5 patients) or left pleural cavity (1 patient), treated between 1992 and 1996, were reviewed. Bleeding was reduced by keeping the mediastinum under local tension (3 patients) or by applying compression on the bleeding site (2 patients), or both (1 patient) while circulatory support, retransfusion of aspirated blood, and hypothermia were established. The diseased aortic arch was replaced during deep hypothermic circulatory arrest, which ranged from 25 to 40 minutes. In 3 patients, the brain was further protected by retrograde (2 patients) or antegrade (1 patient) cerebral perfusion. Results. Hemorrhage from the aortic arch was controlled in all patients. Two patients died postoperatively, one of respiratory failure and the other of abdominal sepsis. Recovery of neurologic function was assessed and complete in all patients. The 4 survivors are well 8 to 49 months after operation. Conclusions. An approach relying on local tamponade to reduce bleeding, rapid establishment of circulatory support and hypothermia, retransfusion of aspirated blood, and swift repair of the aortic arch under circulatory arrest allows salvage of patients with active bleeding from an aortic arch rupture. (Ann Thorac Surg 1998;65:1291 5) 1998 by The Society of Thoracic Surgeons Operations on the aortic arch involve sophisticated techniques, such as deep hypothermic circulatory arrest and selective cerebral perfusion, to avoid ischemic injury to the brain and viscera when perfusion is interrupted. Although excellent results have been achieved by many groups [1 6], this segment of aorta still appears as one of the most difficult to operate on. Rupture of the aortic arch adds a further challenge because exsanguination must be prevented while systemic perfusion, especially perfusion of the arch arteries, must be ensured until systemic deep hypothermia is reached. Although rare, this situation may occur with various aortic pathologic processes and become rapidly desperate if a swift and efficient management is not applied. Reduction of bleeding by maintenance or application of local tamponade, immediate establishment of circulatory support and hypothermia to allow circulatory arrest, direct transfusion of aspirated blood, and swift repair of the aortic arch lesion are the keys to successful outcome. Our experience with this situation is reviewed and guidelines for proper treatment are emphasized. Accepted for publication Dec 13, Address reprint requests to Dr Prêtre, Clinic for Cardiovascular Surgery, University Hospital Zürich, Rämistrasse 100, CH-8091 Zürich, Switzerland. Patients and Methods Patients A retrospective review of 6 patients (all men with ages ranging from 60 to 86 years) treated in our institution between 1992 and 1996 for a rupture of the aortic arch that produced active hemorrhage was performed. Patients with an aortic arch lesion that produced only a contained local hematoma (without cardiac tamponade, massive hemothorax, or free hemorrhage) were not included, because the management and prognosis of these patients are similar to elective cases. Two patients with a rupture of the aorta at the junction of the ascending aorta and aortic arch were included because it was not possible to apply a cross-clamp between the aortic lesion and the arch arteries. Underlying aortic pathologic processes responsible for the rupture were a Stanford type A aortic dissection in 4 patients and a degenerative aneurysm and pseudoaneurysm in 1 patient each. The diagnosis of pericardial tamponade was established clinically and by transthoracic echocardiography in 3 patients. In these patients, transesophageal echocardiography established also the diagnosis of a type A aortic dissection (2 patients) and of an aneurysm of the aortic arch (1 patient). In the remaining 3 patients, the diagnosis of aortic dissection (2 patients) and of rupture of a distal aneurysm of the aortic arch (1 patient) was established by computed tomography. Rupture of the aortic arch produced hemorrhage in the mediastinum 1998 by The Society of Thoracic Surgeons /98/$19.00 Published by Elsevier Science Inc PII S (98)

2 1292 PRÊTRE ET AL Ann Thorac Surg AORTIC ARCH RUPTURE 1998;65: Fig 1. Reduction of free bleeding from the aortic arch is achieved by providing local tamponade with laparotomy pads. Blood is aspirated with cardiotomy suckers into the cardiopulmonary circuit and immediately retransfused. Full cardiopulmonary bypass is established by femoral vessel cannulation using a long venous cannula positioned in the right atria. and pericardial cavity in 5 patients, and in the mediastinum and left pleural cavity in 1 patient. Four patients had an active hemorrhage that occurred before chest opening and provoked shock by pericardial tamponade (3) or hemothorax (1). One patient with cardiac tamponade and a systolic blood pressure less than 60 mm Hg underwent subxyphoid drainage of the pericardium to restore blood pressure while cardiopulmonary bypass was set up and established. Two patients with cardiac tamponade and systolic blood pressure of 80 mm Hg did not undergo pericardial drainage because cardiopulmonary bypass could be established within minutes. The fourth patient had a massive left hemothorax caused by rupture of a pseudoaneurysm of the distal part of the aortic arch. He was hypotensive but responded to fluid and blood transfusion. The chest was opened after institution of cardiopulmonary bypass. Cross-clamping of the aortic arch was not possible. Bleeding was continuous although not massive. The hematoma surrounding the aneurysm was not disturbed and local compression by laparotomy pads was further applied while systemic cooling was performed. The remaining 2 patients had aortic dissection and were not in shock until a sudden, torrential hemorrhage occurred, which was triggered by spreading of the sternum. Surgical Technique Hemorrhage was reduced by maintaining or providing compression on the bleeding site while cardiopulmonary bypass was established through the femoral or iliac vessels and deep hypothermia was achieved. In 3 patients, the chest and pericardial cavity were kept closed until moderate hypothermia was achieved and cardiac fibrillation imminent. In the other patients, who had an opened chest and a free hemorrhage, bleeding was reduced by local compression with laparotomy pads while cardiopulmonary bypass and hypothermia were established. Blood was aspirated into the bypass circuit and immediately retransfused (Fig 1). Full circulatory support was achieved in all patients by inserting a long venous cannula in the right atria. In the patients with a tense pericardial effusion, the flow rate of the extracorporeal circulation was temporarily diminished close to arrest when the heart and electrocardiogram showed signs of imminent fibrillation. This allowed opening of the pericardium without being overflowed by blood, identification of the site of aortic rupture, and completing the equipment for cardiopulmonary bypass. If the site of bleeding precluded application of a clamp between the aortic rupture and the innominate artery (as was the case in our patients), the ascending aorta was cross-clamped proximal to the rupture, the aortic root was opened to allow complete decompression of the left ventricle, and a cannula was inserted in the coronary sinus for cardioplegic delivery. These maneuvers lasted less than 3 minutes. Cardiopulmonary bypass and cooling were then resumed and cardioplegia administered. Cooling was continued until core temperature reached 18 C to allow circulatory arrest. Blood gas was managed by the alphastat method. The patient in whom the surgical approach was through a thoracotomy had a continent aortic valve. A vent was inserted in the left ventricular apex to prevent ventricular distention during cardiac fibrillation and arrest. In this patient, the distal hemiarch and the proximal

3 Ann Thorac Surg PRÊTRE ET AL 1998;65: AORTIC ARCH RUPTURE 1293 Table 1. Clinical Characteristics, Surgical Treatment, and Outcome of Patients With Aortic Arch Rupture Age (y) Cause of Rupture Cardiac Tamponade a Surgical Approach Surgical Treatment Neurologic Outcome Outcome 86 Dissection No Sternotomy DHCA (30 min); retrograde cerebral perfusion; AAo 68 Dissection Yes Sternotomy DHCA (40 min); retrograde cerebral perfusion; AAo 75 Aneurysm Yes Sternotomy Subxiphoid pericardial drainage; DHCA (40 min); antegrade innominate artery perfusion; AAo 65 Pseudoaneurysm Yes Left thoracotomy DHCA (35 min); distal hemiarch proximal descending aorta 63 Dissection Yes Sternotomy DHCA (25 min); AAo 70 Dissection No Sternotomy DHCA (25 min); AAo Death D 31 of respiratory failure Death D 35 of abdominal sepsis Alive at 49 mo Alive at 24 mo Alive at 40 mo Alive at 30 mo AAo ascending aorta; D postoperative day; DHCA deep hypothermic circulatory arrest. a Cardiac tamponade was defined as a tense hemopericardium producing a systemic systolic blood pressure 80 mm Hg. half of the descending aorta were replaced by a prosthesis during a 35-minute hypothermic circulatory arrest. In the other patients, circulation was resumed immediately after repair or resection of the aortic arch. The resection (called resection in Table 1) took out various parts of the arch circumference (Fig 2), but included the site of rupture in all cases. The prosthesis was sutured superiorly on a tongue of aortic arch wall supporting the arch arteries and inferiorly on the origin of the descending aorta with a single row of sutures. This resection can be done quickly and was therefore chosen to reduce the duration of circulatory arrest. Duration of circulatory arrest ranged from 25 to 40 minutes. The brain was further selectively perfused with cold blood during circulatory arrest in 3 patients. Retrograde perfusion through the superior vena cava was performed in 2 patients (with flow rates of 350 and 450 ml/min) and antegrade perfusion through the innominate artery in 1 patient (with a flow of 600 ml/min). Once the aortic arch repair was performed, cardiopulmonary bypass was resumed, rewarming initiated, and the operation completed. This consisted of replacing the ascending aorta in the 5 patients in whom the surgical approach was by sternotomy. Fig 2. Types of aortic arch resection performed in 6 patients (left, 3 patients; middle, 2; right, 1) using an impervious synthetic graft. A tongue of aortic arch supporting the arch arteries allows quick implantation of these arteries on the prosthesis. Results Hemorrhage from the aortic arch rupture was controlled in all patients. Transfusion of homologous blood ranged from 3 to 15 units. Two patients died postoperatively (Table 1). One elderly patient, in whom we had planned to resect only the ascending aorta, had rupture of the aortic arch during spreading of the sternum. This unexpectedly magnified the operation and imposed resorting to hypothermic circulatory arrest. He awoke, was extubated, and sent to regular ward on the third postoperative day, but required intubation 2 days later for tachypnea and hypoxemia. Pneumonia developed, and he died later of respiratory failure. The other patient had an uneventful recovery after an ascending aorta and hemiarch resection. On the day of discharge, he suddenly vomited blood profusely. Emergent laparotomy was performed and a bleeding duodenal ulcer was oversewn.

4 1294 PRÊTRE ET AL Ann Thorac Surg AORTIC ARCH RUPTURE 1998;65: Acute necrotizing pancreatitis developed probably secondary to operative pancreatic injury. Despite multiple abdominal operations aimed at controlling a diffuse peritonitis, he died of sepsis with multiple organ failure 3 weeks later. Neurologic recovery was complete without a neurologic deficit in all 6 patients. Full awakening occurred within 48 hours in 5 patients. In the last patient, full awakening occurred later, when sedation was discontinued. This patient, who had chronic pulmonary disease and had undergone a left thoracotomy, acquired pneumonia and required intubation for 16 days. One patient was reoperated on for postoperative bleeding and another for drainage of a chronic pericardial effusion. Four patients survived and were discharged from 13 to 33 days after their operations. Follow-up was obtained by telephone contact with their private physician. They are all alive and well from 8 to 49 months after discharge, without another cardiac or aortic event. Two patients with aortic dissection have had follow-up of the aorta by magnetic resonance imaging, which did not show further enlargement of the aorta. Comment Rupture of the aortic arch may occur with aneurysms and pseudoaneurysms, with aortic dissection, or after trauma [7 10]. Because the aortic arch is surrounded by mediastinal tissue, its rupture may bear a better prognosis than rupture of other segments of the thoracic aorta: the ascending or descending aorta bleeds directly into the pericardial or pleural cavity and may produce immediate fatal cardiac tamponade or exsanguination. Bleeding from the aortic arch can reach the pericardial or pleural cavity by spreading along the aorta, but the ensuing cardiac tamponade or hemothorax may be less severe or evolve over a longer period allowing surgical intervention. Opening of the chest relieves tissue compression on the aortic rupture and may trigger massive bleeding. Hematomas surrounding an aortic tear are especially unstable when a tense pericardial effusion or a left hemothorax is already present. In this situation, when an aneurysm or an acute dissection of the aorta is present, it is certainly appropriate to open the chest after cardiopulmonary bypass has been established and moderate hypothermia achieved. If cardiac tamponade induces an insufficient cardiac output (primarily reflected by a low systemic blood pressure), a controlled drainage of the pericardium can be performed before cardiopulmonary bypass is established. A 1-cm opening of the pericardium, using a subxyphoid incision, allows relief of the tamponade. The opening can be occluded with a finger when acceptable blood pressure is restored and prevents excessive loss of blood. During cooling, a critical moment happens when the temperature is about to induce ventricular fibrillation because distention of the left ventricle may occur. In case of aortic valve insufficiency (as occurs in aortic dissection), the left ventricle must be vented and must be excluded from the systemic pressure by crossclamping the ascending aorta. Spreading of the sternum and opening of the pericardium are then necessary with the risk of disturbing a contained hemorrhage. If hemorrhage becomes profuse, a short period of low flow or circulatory arrest can be established to allow for crossclamping the ascending aorta, venting the left ventricle, and inserting a cannula in the coronary sinus for retrograde delivery of cardioplegia. When pump flow is resumed, free bleeding should be reduced by compressing the hemorrhagic site with laparotomy pads and by reapproximating the sternal edges, and blood should be aspirated in the bypass circuit and retransfused (Fig 1). Rupture of the aorta occurs most of the time on a diseased aortic wall and imposes its. The of the aortic arch can usually be performed within 45 minutes of deep hypothermic circulatory arrest [1, 11]. This period is regarded as safe if systemic and brain temperature are less than 18 C [6, 12, 13]. Cooling of the brain, however, may be inhomogeneous when compression of the aortic arch is applied because blood flow in the arch arteries may be reduced. Under these circumstances, we chose to limit the duration of circulatory arrest by resecting most of the diseased aortic wall except for a tongue of aorta that includes the arch arteries (Fig 2). This allowed a rapid insertion of a prosthesis with a single suture. This resection was also chosen in case of aneurysmal disease of the aortic arch. We believed that Teflon reinforcement of the suture line would induce local scarring that could prevent further dilatation of the remaining aortic wall. Selective cerebral perfusion enhances cerebral protection and prolongs the period of safe circulatory arrest. Antegrade perfusion of the brain by cannulation of the arch arteries has proved its effectiveness [13, 14] but presents the disadvantage of potentially damaging the wall of the carotid artery. Retrograde perfusion of the brain through the superior vena cava is simpler to establish [3 5], but the distribution of nutritive blood in the brain is uncertain. In patients such as ours, however, retrograde perfusion may have helped achieve a more adequate cerebral hypothermia. Until the mechanisms of action and the reliability of retrograde perfusion are better elucidated, the techniques that have extensively proved their value when circulatory arrest is used (ie, deep hypothermia and swift repair) [1, 2, 11, 12, 15] should not be altered, and retrograde perfusion should be used only as an adjunct to these techniques. References 1. Cooley D, Ott D, Frazier O, Walker W. Surgical treatment of aneurysms of the transverse aortic arch: experience with 25 patients using hypothermia technique. Ann Thorac Surg 1981;32: Crawford E, Coselli J. Replacement of the aortic arch. Semin Thorac Cardiovasc Surg 1991;3: Deeb G, Jenkins E, Bolling S, et al. Retrograde cerebral perfusion during hypothermic circulatory arrest reduces neurologic morbidity. J Thorac Cardiovasc Surg 1995;109: Ganzel B, Edmonds H, Pank J, Goldsmith L. Neurophysiologic monitoring to assure delivery of retrograde cerebral perfusion. J Thorac Cardiovasc Surg 1997;113:

5 Ann Thorac Surg PRÊTRE ET AL 1998;65: AORTIC ARCH RUPTURE Lytle B, McCarthy P, Meaney K, Stewart R, Cosgrove DR. Systemic hypothermia and circulatory arrest combined with arterial perfusion of the superior vena cava. Effective intraoperative cerebral protection. J Thorac Cardiovasc Surg 1995;109: O Connor J, Wilding T, Farmer P, Sher J, Ergin M, Griepp R. The protective effect of profound hypothermia on the canine central nervous system in surgery of the ascending aorta and the aortic arch. Ann Thorac Surg 1986;41: Elefteriades JA, Hartleroad J, Gusberg RJ, et al. Long-term experience with descending aortic dissection: the complication-specific approach. Ann Thorac Surg 1992;53: Bachet JE, Termignon JL, Dreyfus G, et al. Aortic dissection. Prevalence, cause, and results of late reoperations. J Thorac Cardiovasc Surg 1994;108: Prêtre R, von Segesser L. Aortic dissection. Lancet 1997;349: Von Segesser L, Genoni M, Künzli A, et al. Surgery for ruptured thoracic and thoraco-abdominal aortic aneurysms. Eur J Cardiothorac Surg 1996;10: Livesay J, Cooley D, Reul G, et al. Resection of aortic arch aneurysms: a comparison of hypothermic techniques in 60 patients. Ann Thorac Surg 1983;36: Svensson L, Crawford E, Hess K, et al. Deep hypothermia with circulatory arrest: determinants of stroke and early mortality in 656 patients. J Cardiovasc Surg 1993;106: Bachet J, Guilmet D, Goudot B, et al. Cold cerebroplegia. A new technique of cerebral protection during operations on the transverse aortic arch. J Thorac Cardiovasc Surg 1991; 102: Kazui T, Kimura N, Yamada O, Komatsu S. Surgical outcome of aortic arch aneurysms using selective cerebral perfusion. Ann Thorac Surg 1994;57: Ergin M, Galla J, Lansman S, et al. Hypothermic circulatory arrest in operations on the thoracic aorta: determinants of operative mortality and neurologic outcome. J Thorac Cardiovasc Surg 1994;107: REVIEW OF RECENT BOOKS Pediatric Cardiopulmonary Transplantation Edited by Kenneth L. Franco, MD Mount Kisco, NY, Futura, pp, illustrated, $89.00 ISBN: Reviewed by John M. Armitage, MD Pediatric Cardiopulmonary Transplantation, edited by Kenneth L. Franco, MD, is a comprehensive text with 20 chapters contributed to by an impressive list of 33 authors, many of whom are internationally recognized for their legion contributions to the field. Of necessity, a technology applied in children that depended upon and evolved from its application in adults leans heavily on past and ongoing experiences in its adult precursor. However, as any accomplished clinician in pediatrics or its subspecialties will attest, children are, indeed, not small adults. Chapter 3, Pathology of Heart and Lung Transplantation, and chapter 19, Pulmonary Retransplantation, rarely, if ever, mention the pediatric population, its experience, or its special considerations. It is not clear from chapter 19 if any of the 42 candidates or 13 ultimate recipients of pulmonary retransplantation were children. Other chapters on graft vasculopathy and heart-lung transplantation, as well as the transplant registry chapters, shed little new light on the pediatric patients. Also omitted, perhaps overshadowed by the biochemical and surgical heroics of this transplantation science, is a chapter devoted to the psychosocial aspects of this special group of patients. Suicide, noncompliance, growth, education, and procreation are alluded to in passing in some chapters, but probably warrant consideration in their own rights. Forgiving these minor transgressions, the remainder of the chapters, as well as the book as a whole, are outstanding, with some truly wonderful contributions both well written and excellently referenced. By far, David Kapelanski, MD, from San Diego writes the finest treatise on lung preservation I have ever read. It is truly a masterpiece to which all the other chapters aspire, and some come very close. Valluvan Jeevanandam s chapter on pediatric myocardial preservation is also excellent. The series of chapters on lung transplantation, 12 through 17, are each significant contributions, concise and informative. Chapter 18 on obliterative bronchiolitis, admitted by all authors to be the single greatest challenge in pediatric pulmonary transplantation, culminates these. Indeed, for us to continue to offer a heroic and costly therapy, the 3-year survival of which ranges from 40% to 60% in our best centers, we must do better to understand the process that costs the lives of many of these truly brave and pioneering children. Actuarial freedom from obliterative bronchiolitis in the pediatric cohort is 37% at 3 years! Another vitally important and well-written chapter is from the Children s Hospital of Pittsburgh, by Doctors Michaels and Green, on infectious complications (better termed infectious issues) in pediatric heart and lung transplantation. For children, many of the keys to success are anticipation and prevention of infection, which are well addressed here. A review of this text would not be complete without due recognition and appreciation of the extensive contributions, both in their contributed chapters and in their quality clinical work and exhaustive research, of the pediatric transplant groups at Loma Linda University Medical Center and Columbia University, New York. Fredericksburg, Virginia 1998 by The Society of Thoracic Surgeons Ann Thor Surg 1998;65: /98/$19.00 Published by Elsevier Science Inc PII S (98)