Influence of Perioperative Hemodynamics on Spinal Cord Ischemia in Thoracoabdominal Aortic Repair

Transcription

1 Influence of Perioperative Hemodynamics on Spinal Cord Ischemia in Thoracoabdominal Aortic Repair Yujiro Kawanishi, MD, Kenji Okada, MD, Masamichi Matsumori, MD, Hiroshi Tanaka, MD, Teruo Yamashita, MD, Keitaro Nakagiri, MD, and Yutaka Okita, MD Department of Cardiovascular, Thoracic, and Pediatric Surgery, Kobe University Graduate School of Medicine, Kobe City, Hyogo, Japan Background. The purpose of this study is to investigate the influence of perioperative circulation on spinal cord during the repair of descending thoracic or thoracoabdominal aortic aneurysms. Methods. From October 1999, 92 patients (aged years; 65 men) underwent the repair of descending thoracic (n 30) or thoracoabdominal aortic aneurysm (Crawford I, 9; II, 14; III, 35; IV, 4). We measured the time duration of hypotension, defined as follows, and evaluated the relationship between the incidence of paraplegia and each duration: T1, systolic arterial pressure less than 80 mm Hg, or mean pressure less than 60 mm Hg during aortic cross-clamping; T2, distal aortic pressure less than 60 mm Hg during aortic cross-clamping; T3, systolic arterial pressure less than 80 mm Hg after coming off bypass; T4, systolic arterial pressure less than 80 mm Hg in the intensive care unit. Results. Hospital mortality was 8% (7 patients). Neurologic deficits occurred in 10 patients (10.9%). The T1 and T2 periods showed no difference between paraplegia cases (group P) and normal cases (group N). The T3 periods in both groups were and , and the T4 periods were and , respectively. The T3 and T4 periods in group P were significantly longer than in group N (p < ). Multivariate analysis demonstrated that T3 was an independent risk factor for paraplegia. When divided according to body temperature, the T2 period under mild hypothermia was significantly longer in group P than in group N, as well as the T3 and T4 periods. Conclusions. Perioperative hemodynamics stability is of vital importance for spinal cord protection during thoracoabdominal aortic surgery. In particular, the duration of hypotension after coming off bypass was an independent risk factor for paraplegia. (Ann Thorac Surg 2007;84:488 92) 2007 by The Society of Thoracic Surgeons Paraplegia remains a devastating complication of thoracoabdominal aortic surgery. Hypoperfusion of the spinal cord during aortic cross-clamping (ACC) is one cause of paraplegia after the repair of thoracoabdominal aortic aneurysms. The spinal cord obtains a major blood supply from the segmental radicular arteries, which provide significant flow into the anterior spinal artery [1, 2]. The ischemic insult to the spinal cord during aortic cross-clamping is altered depending on perfusion through the collateral circulation. One important factor that influences collateral circulation is proximal arterial perfusion pressure as well as distal aortic pressure. Postoperative hypotension has been reported as a risk factor of delayed paraplegia after both open surgery and endovascular graft repair of thoracic or thoracoabdominal aortic aneurysm [3 6]. We hypothesized that systemic hypotension during the operation could be a risk factor of paraplegia, although it has not been previously reported, and evaluated the Accepted for publication Feb 28, Address correspondence to Dr Okita, Kusunoki-Cho, Chuo-Ku, Kobe City, Hyogo, , Japan; yokita@med.kobe-u.ac.jp. influence of hemodynamics stability on spinal cord ischemia during or after thaoracoabdominal aortic aneurysm repair. Material and Methods Patients This study was approved by the Institutional Review Board, and the need for individual consent was waived. Between December 23, 1999, and October 27, 2005, 101 patients underwent the repair of thaoracoabdominal aortic aneurysm or descending thoracic aortic aneurysm. Of these, 92 patients who had neurologic evaluation postoperatively were enrolled in this study. Patients ages ranged from 18 to 83 years (mean, 66 13). There were 65 men (71%) and 27 women (29%). Thirty patients had descending thoracic aortic aneurysm. Using the Crawford classification, 9 patients had type I thaoracoabdominal aortic aneurysm, 14 had type II, 35 had type III, and 4 had type IV. Aortic dissection was present in 40 patients (43%) and nondissecting aneurysm in 52 (57%). Six patients (7%) had Marfan s syndrome and 19 (21%) had 2007 by The Society of Thoracic Surgeons /07/$32.00 Published by Elsevier Inc doi: /j.athoracsur

2 Ann Thorac Surg KAWANISHI ET AL 2007;84: INFLUENCE OF HEMODYNAMICS ON SPINAL CORD 489 Table 1. Preoperative Characteristics Characteristic N (%) Age Sex (male/female) 65/27 Aneurysm etiology and extent Degenerative disease 52 (57%) Dissection 40 (43%) Type I 9 (10%) Type II 14 (15%) Type III 35 (38%) Type IV 4 (4%) Descending 30 (33%) Preoperative variables Hypertension 74 (80%) Chronic obstructive pulmonary disease 13 (14%) Cerebrovascular disease 21 (23%) Diabetes mellitus 12 (13%) Renal insufficiency a 12 (13%) Rupture 19 (21%) a Serum creatinine 2.0. ruptured aneurysms. The characteristics of the patients are shown in Table 1. Table 2. Periods of Hypotension in Patients With and Without Paraplegia N T1 (min) T2 (min) T3 (min) T4 (min) Paraplegia ( ) a a Paraplegia ( ) a p compared with group N (Mann-Whitney U test). Surgical Technique A cerebrospinal fluid drainage catheter was routinely placed into the subarachnoid space on the day before operation. Cerebrospinal fluid was allowed to drain freely whenever cerebrospinal fluid pressure exceeded 10 mm Hg. Transcranial motor evoked potential was applied to monitor spinal cord ischemia during the operation. The thoracoabdominal aorta was exposed with left thoracotomy, dissection of the retroperitoneal space, and division of the costal cartilage and diaphragm. Cardiopulmonary bypass was established by means of cannulation of the left femoral artery and the right femoral vein, after the administration of heparin (2.0 ml/kg). Mild hypothermia (31 C to 34 C, rectal temperature) was also used routinely to minimize ischemic complications. If cross-clamping of the proximal aorta seemed impossible, deep hypothermia and circulatory arrest were applied. The aorta was anastomosed to a gelatin-impregnated knitted Dacron graft (C.R. Bard, Haverhill, Pennsylvania) with side branches presewn to reattach segmental and visceral arteries. The aorta was sequentially clamped, and the patent intercostal and lumbar arteries at the Th8 to L2 level were reimplanted and immediately reperfused. Significant back-bleeding from segmental arteries was managed by inserting balloon catheters or by external clamping the segmental arteries to reduce the stealing effect from the anterior spinal artery. In cases of type II, III, and IV thaoracoabdominal aortic aneurysm, visceral artery perfusion was routinely performed with selective cannulae. Distal anastomosis of the graft and reconstruction of the visceral arteries were completed during rewarming. Management of Hemodynamics Proximal and distal arterial pressures were obtained from right radial and right dorsalis pedis arterial catheter, respectively, and mean arterial pressure was managed to between 60 and 100 mm Hg. During crossclamping, blood pressure was mainly controlled by adjusting the bypass flow of the centrifugal pump. For hypotension, arterial blood pressure augmentation was achieved by the administration of vasopressor agents, volume expansion therapy including blood transfusion, and increased bypass flow. Table 3. Multiple Regression Model for Preoperative and Operative Predictors of Neurologic Deficit Variable Coefficient Odds Ratio 95% CI p Value Fig 1. Duration of hypotension. T1 systolic arterial pressure (SAP) less than 80 mm Hg or mean pressure less than 60 mm Hg during aortic cross-clamping (ACC); T2 distal aortic perfusion pressure less than 60 mm Hg during aortic cross-clamping; T3 SAP less than 80 mm Hg after coming off bypass; T4 SAP less than 80 mm Hg in intensive care unit (ICU). (BP blood pressure.) T Prior aortic repair Diabetes mellitus Age T Cerebrovascular disease CI confidence interval.

3 490 KAWANISHI ET AL Ann Thorac Surg INFLUENCE OF HEMODYNAMICS ON SPINAL CORD 2007;84: Table 4. Incidence of Neurologic Deficit and Periods of Hypotension in Patients Under Mild and Deep Hypothermia N Deficit T1 (min) T2 (min) T3 (min) T4 (min) Mild hypothermia 72 8 (11.1%) Deep hypothermia 20 2 (10%) 44 9 a a a a p compared with mild hypothermia (Mann-Whitney U test). Collection of Hemodynamic Data Hemodynamic data for all patients were obtained from anesthesia records, extracorporeal circulation records, and intensive care unit charts, retrospectively. We measured the duration of hypotension, defined as follows (Fig 1): T1 systolic arterial pressure in the radial artery is below 80 mm Hg, or mean arterial pressure is below 60 mm Hg during aortic cross-clamping; T2 distal aortic perfusion pressure is below 60 mm Hg during aortic cross-clamping; T3 systolic arterial pressure is below 80 mm Hg after coming off bypass; T4 systolic arterial pressure is below 80 mm Hg in the intensive care unit. In anesthesia records, arterial pressure was recorded every 5 minutes and every 10 minutes on extracorporeal circulation records and intensive care unit charts. When arterial pressure was below the criterion pressure at one time, the duration was counted as 5 minutes and 10 minutes, respectively. Statistical Analysis Data were processed using Stat View J-5.0 software (SAS Institute, Cary, North Carolina). All values are expressed as the mean SE. Statistical analysis was performed with the Mann-Whitney U test to compare the durations between patients with and without postoperative neurologic deficits or between mild and deep hypothermia patients. Unpaired t test was performed only when patients under deep hypothermia were analyzed. Logistic regression multivariate analysis was performed to evaluate the risk factors of spinal cord ischemia. Differences were considered statistically significant at p less than Results Clinical Results The overall 30-day and in-hospital mortality rates were 3.3% (3 patients) and 7.6% (7 patients), respectively. The causes of death were respiratory failure, gastrointestinal bleeding, and sepsis in 2 patients, and mesenteric ischemia in 1. The overall incidence of paraplegia or paraparesis was 10.9% (10 patients). Of the 10 patients with neurologic deficits, 6 had paraplegia, 3 had paraparesis, and 1 had monoplegia. Five patients (5.4%) were returned to the operating room due to postoperative bleeding. The rates of renal failure, pulmonary complication, and stroke were 7.6% (7 patients), 9.8% (9 patients), and 2.2% (2 patients), respectively. Six patients (6.5%) were successfully treated by catheter interventions because of graft occlusion, with which visceral, renal, and subclavian arteries were reattached. Evaluation of Hemodynamic Stability Each period of hypotension in patients with postoperative neurologic deficits (group P) and patients without deficits (group N) is shown in Table 2. The T1 periods in group P and group N were 2 2 and 12 3, and the T2 periods in both groups were 14 9 and 11 4, respectively. The T1 and T2 periods showed no difference between group P and group N. The T3 periods in both groups were and , and the T4 periods were and , respectively. The T3 and T4 periods in group P were significantly longer than those in group N (p ). Multivariate logistic regression analysis identified T3 periods (odds ratio [OR] 1.058; 95% confidence interval [CI]: to 1.114; p 0.03), prior aortic surgery (OR 3.751; 95% CI: to 12.02; p 0.05), and diabetes mellitus (OR 5.488; 95% CI: to 18.89; p 0.03) as significant independent risk factors for spinal cord ischemia (Table 3). When divided based on the mode of hypothermia, T1, T2, and T3 periods in patients who underwent operations under deep hypothermia were significantly longer than in patients under mild hypothermia (44 9, 36 14, and versus , , and , respectively; p ; Table 4). The incidences of neurologic deficits were 11% (8 of 72) in patients under mild hypothermia and 10% (2 of 20) in patients under deep hypothermia (not significant). Among patients under mild hypothermia, T2 was significantly longer in group P than in group N as well as in the T3 and T4 periods (p 0.03, 0.003, 0.003, respectively; Table 5). On the other hand, T2 was not associated with postoperative paraplegia among patients under deep hypothermia. Comment The causes of paraplegia after the repair of thoracic or thoracoabdominal aortic aneurysms were believed to be spinal cord ischemia from hypotension during aortic cross-clamping, the interruption of segmental arteries, thrombosis or embolism, and postoperative hypotension or hypoxia. Midthoracic and thoracolumbar regions of the spinal cord receive their major blood supply from the segmental radicular arteries, which provide significant flow into the anterior spinal artery [1, 2]. During aortic cross-clamping, the spinal cord blood supply depends on spinal cord perfusion through the collateral circulation, such as vertebral arteries, iliolumbar arteries and other segmental arteries. To maximize the collateral circulation, the maintenance of distal aortic perfusion, cerebrospinal fluid drainage, segmental aortic clamping, and reimplantation of intercostal or lumbar arteries have been applied [7 10].

4 Ann Thorac Surg KAWANISHI ET AL 2007;84: INFLUENCE OF HEMODYNAMICS ON SPINAL CORD 491 Table 5. Periods of Hypotension in Patients With and Without Paraplegia Under Each Mode of Hypothermia Paraplegia N T1 (min) T2 (min) T3 (min) T4 (min) Mild hypothermia ( ) a b b ( ) Deep hypothermia ( ) c 60 d Distal aortic perfusion enables minimization of the ischemic time of distal organs, including the spinal cord, through the lumbar or iliolumbar arteries, and the usefulness of this adjunct has been previously reported [7, 8]. Proximal arterial perfusion pressure is one important factor that influences collateral circulation as well as distal aortic pressure, and some authors reported its importance in an experimental model [11 13]. Lu and colleagues [11] reported the relationship between proximal arterial pressure and spinal cord blood flow during aortic cross-clamping in rats, and concluded that thoracic aortic occlusion with systemic hypovolemic hypotension induced more profound and lasting spinal cord hypoperfusion, which resulted in severe ischemic injury of the spinal cord compared with normal arterial pressure. Taira and coworkers [12] reported a significant proportional correlation between the magnitude of the reduction of proximal arterial pressure and spinal cord blood flow during aortic cross-clamping; moreover, the higher the proximal aortic pressure, the higher the ischemic tolerance. These results implicate the importance of systemic arterial pressure, which affects collateral circulation, to the spinal cord during aortic cross-clamping. Although it has not been reported that intraoperative hypotension was an independent risk factor in clinical study, Acher and coworkers [14] reported that a significant drop in the cardiac index during cross-clamping was a risk factor for paraplegia. This result may reflect the important contribution of cardiac function to collateralized circulation of the spinal cord. On the other hand, many authors have recognized postoperative hypotension as a risk factor of delayed paraplegia after both open surgery and endovascular graft repair of thoracic or thoracoabdominal aortic aneurysm [3 6], although preoperative and operative predictors did not include hypotension [15]. Azizzadeh and coworkers [3] reported that mean arterial pressure less than 60 mm Hg and cerebrospinal fluid drain complications increased the incidence of delayed paraplegia after thoracoabdominal aortic open repairs. Chiesa and colleagues [5] reported perioperative hypotension (mean arterial pressure 70 mm Hg) as a significant risk factor of spinal cord ischemia after stent-graft repair of the thoracic aorta. Arterial pressure augmentation and the reduction of cerebrospinal fluid pressure, which restore spinal cord perfusion, have been reported to be effective for delayed paraplegia [4 6, 16]. In our study, systemic hypotension after coming off bypass was revealed to have adverse effects on spinal ( ) a p 0.03 compared with patients without paraplegia; b p compared without patients with paraplegia (mann-whitney U test); c p 0.02 compared with patients without paraplegia; d p 0.05 compared with patients without paraplegia (unpaired t test). cord perfusion. The periods of hypotension after coming off bypass and in the intensive care unit (T3 and T4) were longer in patients with postoperative paraplegia than in those with normal spinal function. In particular, T3 was an independent predictor of spinal cord ischemia. This is not because hypotension after coming off bypass deteriorated spinal perfusion most, but because some paraplegia patients suffered serious bleeding or severe hypoxia after coming off bypass, which might have impaired the spinal cord. Actually, in 3 of these patients, motor evoked potentials demonstrated significant ischemic changes in accordance with hemodynamic instability (Fig 2). The periods of hypotension during cross-clamping (T1 and T2) were not associated with postoperative paraplegia. This result does not imply that proximal or distal arterial pressure during aortic cross-clamping is not important for spinal cord perfusion. In this study, it is difficult to evaluate the influence of hypotension during aortic cross-clamping because we can maintain sufficient proximal and distal arterial pressure in most patients with ease owing to extracorporeal circulation. When surgery was performed under deep hypothermia, the duration of hypotension during aortic crossclamping and after coming off bypass (T1, T2, and T3) Fig 2. Relation between arterial blood pressure and motor evokedpotential (MEP) monitoring in a patient with postoperative paraplegia. Motor evoked-potential change is shown below the graph. Motor evoked-potential amplitude recovered after transient ischemic change and disappeared in accordance with hemodynamic instability owing to serious bleeding after coming off bypass. The gray bar represents time duration of hypotension. (ACC aortic cross-clamping; BP blood pressure.)

5 492 KAWANISHI ET AL Ann Thorac Surg INFLUENCE OF HEMODYNAMICS ON SPINAL CORD 2007;84: was longer than under mild hypothermia, being affected by deep hypothermia and circulatory arrest; however, the incidence of neurologic deficits did not differ in spite of the mode of hypothermia, and the periods of hypotension during aortic cross-clamping (T1 and T2) were not associated with neurologic deficits (Tables 4, 5). This result might reflect the neuroprotective effect of deep hypothermia [17]. The duration of distal arterial hypotension (T2) was longer in group P than in group N only when the operation was performed under mild hypothermia (Table 5). Distal aortic perfusion might be more important, especially in mild hypothermia, under which the protective effect of spinal cord is less than under deep hypothermia, although mild hypothermia is also considered to have this effect [17 19]. In conclusion, perioperative hypotension was associated with spinal cord ischemia and the duration of hypotension after coming off bypass was particularly an independent risk factor for postoperative neurologic deficit. Although proximal and distal arterial pressure during aortic cross-clamping was not revealed to be a significant risk factor for paraplegia, maintaining perioperative hemodynamics was considered of vital importance for spinal cord protection during thoracoabdominal aortic surgery. References 1. Gray H. Anatomy of the human Body. 29th ed. Philadelphia: Lea & Febiger, 1973: Svensson LG, Rickards E, Coull A, Rogers G, Fimmel C, Hinder RA. Relationship of spinal cord blood flow to vascular anatomy during thoracic aortic cross-clamping and shunting. J Thorac Cardiovasc Surg 1986;91: Azizzadeh A, Huynh TT, Miller CC, et al. Postoperative risk factors for delayed neurological deficit after thoracic and thoracoabdominal aortic aneurysm repair: a case-control study. J Vasc Surg 2003;37: Maniar HS, Sundt TM, Prasad SM, et al. Delayed paraplegia after thoracic and thoracoabdominal aneurysm repair: a continuing risk. Ann Thorac Surg 2003;75: Chiesa R, Melissano G, Marrocco-Trischitta MM, Civilini E, Setacci F. Spinal cord ischemia after elective stent-graft repair of the thoracic aorta. J Vasc Surg 2005;42: Cheung AT, Pochettino A, McGarvey ML, et al. Strategies to manage paraplegia risk after endovascular stant repair of descending thoracic aortic aneurysms. Ann Thorac Surg 2005;80: Schepens MA, Vermeulen FE, Morshuis WJ, et al. Impact of left heart bypass on the results of thoracoabdominal aortic aneurysm repair. Ann Thorac Surg 1999;67: Coselli JS, LeMaire SA. Left heart bypass reduces paraplegia rates after thoracoabdominal aortic aneurysm repair. Ann Thorac Surg 1999;67: Coselli JS, Le Maire SA, Koksoy C, Schmittling ZC, Curling PE. Cerebrospinal fluid drainage reduces paraplegia after thoracoabdominal aortic aneurysm repair: results of a randomized clinical trial. J Vasc Surg 2002;35: Safi HJ, Miller CC, Carr C, Iliopoulos DC, Dorsay DA, Baldwin JC. Importance of intercostal artery reattachment during thoracoabdominal aortic aneurysm repair. J Vasc Surg 1998;27: Lu K, Liang CL, Chen HJ, et al. Injury severity and cell death mechanisms: effects of concomitant hypovolemic hypoperfusion on spinal cord ischemia-reperfusion in rats. Exp Neurol 2004;185: Taira Y, Marsala M. Effect of proximal arterial perfusion pressure on function, spinal cord blood flow, and histopathologic changes after increasing intervals of aortic occlusion in the rat. Stroke 1996;27: Toung TJ, Chang Y, Williams M, Crain BJ, Traystman RJ, Bhardwaj A. Experimental spinal cord ischemia: model characterization and improved outcome with arterial hypertension. Crit Care Med 2004;32: Acher CW, Wynn MM, Hoch JR, Kranner PW. Cardiac function is a risk factor for paralysis in thoracoabdominal aortic replacement. J Vasc Surg 1998;27: Estrena AL, Miller CC, Huynh TT, et al. Preoperative and operative predictors of delayed neurologic deficit following repair of thoracoabdominal aortic aneurysm. J Thorac Cardiovasc Surg 2003;126: Cheung AT, Weiss SJ, McGarvey ML, et al. Interventions for reversing delayed-onset postoperative paraplegia after thoracic aortic reconstruction. Ann Thorac Surg 2002;74: Svensson LG, Khitin L, Nadolny EM, Kimmel WA. Systemic temperature and paralysis after thoracoabdominal and descending aortic operations. Arch Surg 2003;138: Michenfelder JD, Milde JH. The effect of profound levels of hypothermia (below 14 degrees Celsius) on canine cerebral metabolism. J Cereb Blood Flow Metab 1992;12: Kakinohara M, Taira Y, Marsala M. The effect of graded postischemic spinal cord hypothermia on neurologic outcome and histopathology after transient spinal ischemia in rat. Anesthesiology 1999;90: