Since the first resection of the aortic arch performed by

Antegrade Cerebral Perfusion With Cold Blood: A 13-Year Experience Jean Bachet, MD, David Guilmet, MD, Bertrand Goudot, MD, Gilles D. Dreyfus, MD, Philippe Delentdecker, MD, Denis Brodaty, MD, and Claude Dubois, MD Service de Chirurgie Cardio-vasculaire, Hopital Foch - Universite Rene Descartes, Suresnes, France Background. In 1986 we introduced the technique of antegrade selective perfusion of the brain with cold blood during surgery of the aortic arch. Methods. Between January 1984 and March 1998, 171 patients (118 males and 53 females) aged 25 to 83 years (mean 56.5 17), underwent replacement of the transverse aortic arch with the aid of cold blood antegrade selective perfusion. One hundred twenty two patients (71.3%) with chronic lesions were operated on electively; 49 patients (28.6%) were operated on urgently for acute aortic dissection (42 patients) or for a ruptured chronic aneurysm (7 patients). Fifty-one patients (29.8%) had previously undergone a surgical procedure on the thoracic aorta. Mean duration of cardiopulmonary bypass was 121 minutes (range: 65 248); mean duration of cerebral perfusion was 60 minutes (range: 15 90), and mean duration of systemic circulatory arrest circuit was 32 minutes (range: 10 57). The electroencephalogram, routinely recorded, showed disappearance of electrical activity in a mean of 9 minutes (range: 3 16) initial return of electrical activity after a mean of 12 minutes (range: 1 35) and normalization in a mean time of 66 minutes. Results. All patients but 7 (4%) showed signs of normal awakening within 8 hours postoperatively. Six patients (3.5%) had fatal neurologic complications, and 16 patients (9.3%) had a non-fatal neurologic complications. Twenty-nine patients (16.9%) died during the postoperative hospital course. There was a significant difference between patients aged less than 60 years (9%) and patients older than 60 years (mortality rate 26.4%, p < 0.02). There was also a significant difference between patients undergoing an isolated replacement of the arch, and those in whom the replacement was extended to the descending aorta in whom mortality was 36.4% ( 2, p < 0.02). Lesion and gender had no significant influence on the outcome of the patients, nor had the duration of cardiopulmonary bypass, circulatory arrest, and cerebral perfusion. In particular, no correlation could be established between the duration of cerebral perfusion and the occurrence of neurologic complications. Conclusion. The clinical results obtained throughout this experience have demonstrated that selective antegrade cerebral perfusion with cold blood provides excellent protection during surgery of the transverse aortic arch. In addition, it avoids the use of deep hypothermia and prolonged cardiopulmonary bypass and does not limit the time allowed to perform the aortic repair. In our opinion it is the technique of choice, especially in frail patients or those requiring a long and difficult procedure. (Ann Thorac Surg 1999;67:1874 8) 1999 by The Society of Thoracic Surgeons Since the first resection of the aortic arch performed by one of us in 1966 [1], we have performed 212 replacements of the transverse aortic arch using three different methods of cerebral protection. From 1966 to 1977, we used antegrade carotid perfusion at moderate hypothermia through an arterial line derived from the main arterial circuit in 27 cases. From 1977 to 1984, in 34 cases, we used deep hypothermia and circulatory arrest. Since then, we have used an original technique [2] of antegrade carotid perfusion with cold blood (10 C) derived from the oxygenator through a separate heat exchanger in a total of 171 cases. For obvious reasons, including the number of patients at risk and the evolution in the medical and surgical management of the patients during these different time periods, these three techniques cannot be compared. The Presented at the Aortic Surgery Symposium VI, April 30 May 1, 1998, New York, NY. Address reprint requests to Dr Bachet, Hopital Foch, 40 rue Worth, 92150 Suresnes, France. present study reports on our experience with antegrade cold blood perfusion or so called cold blood cerebroplegia which, in our opinion, has proven to be more efficacious than the other methods hitherto employed. Material and Methods The technique of antegrade cold blood cerebral perfusion is aimed at cooling the brain independently of the rest of the body through selective perfusion of the brachiocephalic arteries with cold blood (6 to 12 C) in a manner analogous to the use of cold blood cardioplegia: the patient is maintained at moderate core hypothermia (25 to 28 C), cardiopulmonary bypass (CPB) is discontinued, and cerebral perfusion is carried out during the distal aortic repair. Perfusion Equipment A regular CPB circuit is modified by addition, beyond the oxygenator, of a heat exchanger usually dedicated to cold blood cardioplegia, and a roller pump (Fig 1). By means 1999 by The Society of Thoracic Surgeons 0003-4975/99/$20.00 Published by Elsevier Science Inc PII S0003-4975(99)00411-7

Ann Thorac Surg BACHET ET AL 1999;67:1874 8 CEREBRAL PROTECTION WITH COLD BLOOD 1875 Table 1. Operative Data Variable Mean Range Cardiopulmonary bypass (min) 135 43 60 248 Cerebral perfusion (min) 60 27 15 130 Circulatory arrest (min) 32 27 10 92 Rectal temperature ( C) 26 3 20 31 EEG disappearance (min) 9 6 3 16 EEG reappearance (min) 12 15 1 35 Fig 1. The perfusion circuit of the brain and the coronary arteries with cold blood. Ox oxygenator; P 1 roller pump for the cold blood perfusion; P 2 roller pump for the main circuit; E 1 extra heat exchanger for the cerebral circuit (10 12 C); E 2 heat exchanger of the main circuit (28 C). of the heat exchanger, blood derived from the oxygenator can be cooled to 6 to 12 C. A perfusion line distributes the cold blood to the brachiocephalic arteries and the coronary arteries through a quadrifurcated connector. Specially designed cannulas are available in several diameters to perfuse the carotid arteries. The surgical management and technique of antegrade cerebral perfusion with cold blood have previously been published [3, 4], so they will not be described again in this report, but it seems important to emphasize the following technical points. While the rectal temperature is lowered to 28 C, cannulae are inserted in the innominate and left carotid arteries, or in both carotid arteries, and held by means of adventitial 5-0 polypropylene purse-string sutures. The brachiocephalic arteries are then cross-clamped, and selective cold perfusion is initiated. When the electroencephalogram demonstrates total disappearance of activity, CPB is discontinued, and the aortic arch is opened. Myocardial protection is achieved by perfusing cold blood through selective cannulation of the coronary ostia, or of the ascending aorta if this segment is not being replaced. In addition, the pericardium is filled with iced slush. During circulatory arrest, selective perfusion of the cerebral and coronary arteries is maintained at a flow rate of 400 to 500 ml/min. Because of the shape and size of the coronary line and cannulas, a maximum of 150 ml/min is perfused into the coronary arteries. The perfusion flow of the brain is consequently 250 to 350 ml/min. The pressure in the perfusion line is maintained between 200 and 250 mm Hg, which corresponds to a pressure of 60 to 70 mm Hg in the carotid arteries. During cerebral perfusion and circulatory arrest, no adjuncts, such as barbiturates or steroids, are used to enhance cerebral protection. When the distal anastomosis is completed, the prosthesis is cross-clamped, CPB is resumed, and rewarming is started. The brachiocephalic vessels are then reimplanted into the prosthesis either in a single cuff or separately. The cephalic arteries are unclamped after careful removal of any air bubbles, and cerebral perfusion is discontinued. The prosthesis is then secured to the ascending aorta. If aortic valve replacement or a Bentall procedure are required, those procedures are generally carried out before replacement of the aortic arch. Patients Between January 1984 and March 1998, 171 patients (118 males and 53 females) underwent replacement of the transverse aortic arch with the aid of cold blood antegrade selective perfusion. Ages ranged from 25 to 83 years (mean 56.5 17). One hundred and twenty-two patients (71.3%) with chronic lesions were operated on electively; 49 patients (28.6%) were operated on urgently for acute aortic dissection necessitating replacement of the transverse arch (42 pts) or for a ruptured chronic aneurysm (7 pts). Fifty-one patients (29.8%) had previously undergone a surgical procedure on the thoracic aorta. In all patients but 8 (4.6%), the cephalic arteries were reimplanted through a single cuff. The left subclavian artery was ligated in 5 patients (2.9%). The extent of aortic replacement varied. Thirty-six patients (22.2%) underwent isolated replacement of the transverse arch; 33 patients (19.2%) underwent a replacement of the transverse arch associated either with a composite graft replacement of the aortic root (32), or a total replacement of the ascending aorta preserving the native valve (Yacoub s procedure, 1), and 76 patients

1876 BACHET ET AL Ann Thorac Surg CEREBRAL PROTECTION WITH COLD BLOOD 1999;67:1874 8 Table 2. Causes of Hospital Mortality Variable N % Neurologic complication 6 3.5 Multiorgan failure 6 3.5 Infection 6 3.5 Myocardial infarction 5 2.9 Distal aortic rupture 4 2.3 Respiratory failure 1 0.6 Bowel infarction 1 0.6 Total 29 16.9 (44.4%) underwent ascending aorta substitution, associated with an aortic valve replacement in 8 (4.6%). Twenty-four patients (14%) underwent replacement of the transverse arch associated with a more or less extended replacement of the descending aorta; in 10 of these patients, a left lateral thoracotomy was associated with the median sternotomy. Results Cardiopulmonary Bypass The duration of CPB, circulatory arrest, perfusion of the brachiocephalic and coronary arteries, and the level of rectal temperatures are indicated in Table 1. There was a significant difference in duration of CPB and coronary perfusion according to the lesion treated, and, consequently, the procedure performed. In contrast, the extent of aortic replacement had no influence on the duration of cephalic perfusion or circulatory arrest, or on the level of rectal and esophageal temperatures. Electroencephalogram The intraoperative electroencephalogram was recorded in all patients. During cooling, cerebral activity disappears completely in a mean of 9 minutes (3 to 16) after initiation of cold blood perfusion. During rewarming, the first electrical wave reappears in a mean of 12 minutes (1 to 35), and cerebral activity is restored to normal in a mean of 66 minutes. Mortality Twenty-nine patients (16.9 %) died. Their causes of death are indicated in Table 2. Neurologic Complications All patients but 7 (4%) showed signs of normal awakening within 8 hours postoperatively. Twenty-two patients (12.8%) experienced postoperative neurologic complications. Six patients (3.5%) had a fatal neurologic complication: the cause of coma and death was unknown in 2; death was probably related to embolism in another 2, and 2 had multiorgan failure and intractable low cardiac output. We observed 16 cases (9.3%) of non-fatal neurologic complications. There were 3 cases of paraplegia related to cord ischemia: two after closure of a critical entry site in chronic type B dissection, and one following ligation of the left subclavian artery. In 3 patients, the occurrence of a Brown-Sequard syndrome was related to the extension of an acute dissecting process. Two patients had transient hemianopia 9 and 11 days postoperatively, probably in relation to emboli. Lastly, 3 patients with preoperative hemiplegia were permanently disabled. Other Complications Seven patients had major respiratory failure, necessitating prolonged ventilation. Two patients experienced postoperative myocardial infarctions. One patient underwent a left hemicolectomy for colonic ischemia. Recurrent laryngeal nerve palsy was observed in 10 patients, and 4 patients had left phrenic nerve palsy. Eight patients had acute renal failure requiring hemodialysis; 3 patients suffered from mediastinitis, and one from peptic ulcer. Risk Factors Death was directly related to a neurologic complication in 6 patients. There was a significant difference in mortality between patients aged less than 60 (9%) and more than 60 (26.4%, p 0.02). Similarly, the type of replacement was a prominent risk factor. Although the performance of an ascending aortic or composite graft replacement had no adverse influence on mortality rate, there was a significant difference in mortality between patients undergoing isolated replacement of the arch and those in whom the replacement extended to the descending aorta, in whom mortality was 36.4% ( 2, p 0.02). Lesion and gender had no significant influence on the outcome of the patients, nor had the duration of CPB, circulatory arrest, or cerebral perfusion. In particular, no correlation could be established between the duration of cerebral perfusion and the occurrence of neurologic complications. Comments Deep hypothermia is the most usual method of cerebral protection during replacement of the aortic arch. This technique, however, only gives the surgeon a limited time to carry out the aortic repair. It also requires that cardiopulmonary bypass be prolonged to rewarm the patient, which may cause some complications [5, 6]. In 1992, Ueda and associates proposed the use of retrograde cerebral perfusion through the superior vena cava associated with deep hypothermia as a method of brain protection during aortic arch exclusion [7]. This technique has rapidly gained wide acceptance and is presently used routinely in many centers. Selective carotid perfusion can also be used, but when such perfusion is derived from the main arterial line, the aorta must be cross-clamped to perform the aortic repair. In addition, there is some uncertainty as to what constitutes adequate cerebral perfusion at normal temperature or with moderate hypothermia. The rationale for perfusing the cerebral arteries with blood cooled to 6 to 12 C was to reconcile the advantages of both methods whilst avoiding their disadvantages: in particular, it allows one to

Ann Thorac Surg BACHET ET AL 1999;67:1874 8 CEREBRAL PROTECTION WITH COLD BLOOD 1877 shorten the duration of CPB significantly, but still carry out the distal anastomoses in the open manner. Autoregulaton of cerebral blood flow is an important factor in maintenance of a stable environment in the brain. Several reports [8, 9] have suggested that autoregulation of cerebral blood flow is maintained in deep hypothermic situations (20 C) and prevents cerebral ischemia or hypoperfusion for perfusion pressures ranging from 30 to 100 mm Hg. If perfusion flow rate is reduced, total brain flow decreases, but even at the lowest perfusion flow rates all areas of the brain appear to remain perfused. Normally, individual cerebral structures are perfused in proportion to their metabolic demands. At 20 C, oxygen consumption of the brain is reduced to a fraction of what is required at normothermia as a result of the effect of hypothermia on metabolic processes, and it is suppressed even further at lower temperatures ( 15 C). At this level of hypothermia, the dissociation curve of hemoglobin is displaced toward the left, and oxygen delivered to the tissues is mainly transported in the dissolved form. The main advantage of using oxygenated blood consists therefore in the buffer capacity of the imidazole nucleus of the hemoglobin molecule. These factors constitute the basic reasons for the beneficial effect of continuously perfusing the brain at a very low temperature: if cerebral flow rate and pressure are maintained in a range that permits autoregulation, protection against ischemia or hyperperfusion should be provided. Since there are presently three main methods of brain protection used during aortic arch exclusion, it seems relevant to ask which one is best, and, in particular, whether antegrade selective perfusion is superior to retrograde cerebral perfusion in conjunction with deep hypothermia. The efficiency of retrograde cerebral perfusion remains controversial. Deep hypothermia alone, provided it is carried out properly, is efficient in protecting the central nervous system when circulatory arrest lasts less than 45 minutes [6]; because in most reported experiences with retrograde perfusion the duration of circulatory arrest does not exceed this limit, it is not clear whether the brain is protected by profound hypothermia or by retrograde perfusion or, as we believe, by a combination of the two. The presence of valves in the human jugular system has also been unequivocally demonstrated, [10], and may be a major impediment to proper distribution of perfusate to the brain. Recently, DeBrux and coworkers have demonstrated in cadavers that most of the liquid perfused retrograde into the superior vena cava irrigates the perivertebral venous system through the azygos vein [11]. Similarly, Midulla and coworkers demonstrated in pigs that aortic recovery was less than 5% of total retrograde inflow [12]. Boeckxstaens and Flameng demonstrated that retrograde perfusion does not perfuse the brain in baboons [13], but Pagano and Bonser, using 99 Tmc-HMPAO as a marker, have elegantly shown during operative procedures that retrograde cerebral perfusion does reach the cerebrum in humans, and results in a homogenous cerebral distribution [14]. In our opinion, it is presently impossible to conclude for or against this method. It is possible that retrograde cerebral perfusion, by ensuring a better distribution of cold to the brain, enhances the protective effect of profound hypothermia, and therefore allows longer times of circulatory arrest. In contrast, our technique of cold blood selective cerebral perfusion has several important sources of experimental support. In an article published in 1991 by Swain and associates [15], there is a striking difference in favor of selective antegrade deep hypothermic perfusion at moderate flow, during which no alterations of any of the energetic components of cerebral tissues are seen. More recently, Ye and colleagues [16] concluded that retrograde perfusion provides some brain protection but does not avert moderately severe neuronal alteration, and that only antegrade cerebral perfusion prevents ischemic damage to the brain during prolonged circulatory arrest. Similarly, Sakurada and coworkers carried out a comparative study of the three techniques [17], and concluded that retrograde perfusion had some advantage for cerebral protection compared with deep hypothermic circulatory arrest, but could not supply sufficient blood flow to maintain brain function, and that selective antegrade perfusion was the safest method of brain protection during arch exclusion. Although the present study was not intended to compare different techniques of cerebral protection during aortic arch replacement, these experimental results, together with our own and others [18] clinical experience, suggest that the safety provided by selective antegrade perfusion of the brain with cold blood during an unlimited time of arch exclusion makes this technique the method of choice. This is particularly true in frail and aged patients or in patients requiring long and difficult operations. References 1. Guilmet D, Scetbon V, Ricordeau G, et al. Un cas d anévrysme de la totalité de la crosse aortique traité avec succès par résection greffe. Mémoires de l Académie de Chirurgie 1966; 92:479 87. 2. Guilmet D, Roux PM, Bachet J, et al. Nouvelle technique de protection cérébrale: chirurgie de la crosse aortique. Presse Med 1986;15:1096 8. 3. 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; 102:85 94. 4. Bachet J, Goudot B, Dreyfus G, et al. Antegrade selective cerebral perfusion with cold blood during aortic arch surgery. J Cardiac Surg 1997;12:193 200. 5. Griepp RB, Stinson EB, Hollingsworth JF, Buehler D. Prosthetic replacement of the aortic arch. J Thorac Cardiovasc Surg 1975;70:1051 63. 6. Griepp RB, Ergin A, McCullough JN, et al. Use of hypothermic circulatory arrest for cerebral protection during aortic surgery. J Card Surgery 1997;12:312 21. 7. Ueda Y, Miki S, Kusuhara K, et al. Deep hypothermic systemic circulatory arrest and continuous retrograde cerebral perfusion for surgery of aortic arch aneurysm. Eur J Cardiothorac Surg 1992;6:36 42. 8. Tanaka J, Shiki J, Asai T, et al. Cerebral autoregulation

1878 BACHET ET AL Ann Thorac Surg CEREBRAL PROTECTION WITH COLD BLOOD 1999;67:1874 8 during deep hypothermic non pulsatile cardiopulmonary bypass with selective cerebral perfusion in dogs. J Thorac Cardiovasc Surg 1988;95:124 32. 9. Fox LS, Blackstone EH, Kirklin JW, et al. Relationship of brain blood flow and oxygen consumption to perfusion flow rates during profoundly hypothermic cardiopulmonary bypass. J Thorac Cardiovasc Surg 1984;87:658 64. 10. Dresser LP, Mckinney WM. Anatomic and pathophysiologic studies of the human internal jugular valves. Am J Surg 1987; 154:220 4. 11. De Brux JL, Subayi JB, Pegis JD, Pillet J. Retrograde cerebral perfusion: Anatomic study of the distribution of blood to the brain. Ann Thorac Surg 1995;60:1294 8. 12. Midulla PS. Retrograde cerebral perfusion does not protect the brain in non-human primates. Discussion. Ann Thorac Surg 1995;60:327 8. 13. Boeckxstaens CJ, Flameng WJ. Retrograde cerebral perfusion does not protect the brain in non-human primates. Ann Thorac Surg 1995;60:319 28. 14. Pagano D, Bovin M, Faroqui MH, et al. Retrograde perfusion through the superior vena cava perfuses the brain in human beings. J Thorac Cardiovasc Surg 1996;11:270 2. 15. Swain JA, McDonald TJ, Griffith PK, et al. Low-flow hypothermic cardiopulmonary bypass protects the brain. J Thorac Cardiovasc Surg 1991;102:76 84. 16. Ye J, Yang J, Del Bigio MR, et al. Neuronal damage after hypothermic circulatory arrest and retrograde cerebral perfusion in the pig. Ann Thorac Surg 1996;61:1316 22. 17. Sakurada T, Kazui T, Tanaka H, Komatsu S. Comparative experimental study of cerebral protection during aortic arch reconstruction. Ann Thorac Surg 1996;61:1348 54. 18. Kazui T, Kimura N, Yamada O, Komatsu S. Surgical outcome of aortic arch aneurysm using selective cerebral perfusion. Ann Thorac Surg 1994;57:904 11.