- Email: [email protected]
Neurologic outcome after thoracic and thoracoabdominal aortic aneurysm repair
Neurologic Outcome After Thoracic and Thoracoabdominal Aortic Aneurysm Repair Anthony L. Estrera, MD, Charles C. Miller III, PhD, Tam T. T. Huynh, MD, Eyal Porat, MD, and Hazim J. Safi, MD Department of Cardiothoracic and Vascular Surgery, The University of Texas at Houston Medical School, Memorial Hermann Hospital, Houston, Texas
Background. Neurologic deficit (paraparesis and paraplegia) after repair of the thoracic and thoracoabdominal aorta remains a devastating complication. The purpose of this study was to determine the effect of cerebrospinal fluid drainage and distal aortic perfusion upon neurologic outcome during repair of thoracic and thoracoabdominal aortic aneurysm (TAAA) repair. Methods. Between February 1991 and March 2000, we performed 654 repairs of the thoracic and thoracoabdominal aorta. The median age was 67 years and 420 (64%) patients were male. Forty-five cases (6.9%) were performed emergently. Distribution of TAAA was the following: extent I, 164 (25%); extent II, 165 (25%); extent III, 61 (9%); extent IV, 95 (15%); extent V, 23 (3.5%); and descending thoracic, 147 (22%). The adjuncts cerebrospinal fluid drainage and distal aortic perfusion were used in 428 cases (65%). Results. Thirty-day mortality was 14% (94 of 654). The in-hospital mortality was 16% (106 of 654). Early neurologic deficits occurred in 33 patients (5.0%). Overall, 14 of
428 (3.3%) neurologic deficits were observed in the adjunct group, and 19 of 226 (8.4%) in the nonadjunct group (p ⴝ 0.004). When the adjuncts were used during extent II repair, the incidence was 10 of 129 (7.8%) compared with 11 of 36 (30.6%) in the nonadjunct group (p < 0.001). Multivariate analysis demonstrated that risk factors for neurologic deficit were cerebrovascular disease and extent of TAAA (II and III) (p < 0.05). Conclusions. The combined adjuncts of distal aortic perfusion and cerebrospinal fluid drainage demonstrated improved neurologic outcome with repair of thoracic and TAAAs. In extent II aneurysms, adjuncts continue to make a considerable difference in the outcome and to provide significant protection against spinal cord morbidity. Future research should focus on spinal cord protection in patients with high-risk extent II aneurysms.
R
or steroids. These adjuncts have had varying degrees of success in preventing neurologic deficit [3–9]. Despite these advances, neurologic deficit still remains a significant obstacle in the treatment of aortic aneurysms. The purpose of this study was to review our experience with repair of thoracic and thoracoabdominal aortic aneurysms and determine the effect of the combined adjuncts of distal aortic perfusion and cerebrospinal fluid drainage on neurologic outcome.
epair of thoracic and thoracoabdominal aortic aneurysms (TAAA) remains a surgical challenge. The need for spinal cord protection during repair of these aneurysms is of paramount importance because the development of neurologic deficits (ie, paraplegia and paralysis) continues to have devastating consequences. In the era of the “cross-clamp and sew” technique, the incidence of neurologic deficits correlated with the extent of the aneurysm (Crawford I-IV) and aortic ischemic time [1, 2]. However, in recent years, advances in surgical technique and new adjuncts that increase the ischemic tolerance of the spinal cord have improved results. Many methods have been devised to protect the spinal cord during aortic surgery. These include active and passive distal aortic perfusion, total cardiopulmonary bypass, profound hypothermic circulatory arrest, direct spinal cord cooling, cerebrospinal fluid drainage, monitoring of somatosensory and motor-evoked potentials, and pharmacologic methods using papaverine, nalaxone, Presented at the Thirty-seventh Annual Meeting of The Society of Thoracic Surgeons, New Orleans, LA, Jan 29 –31, 2001. Address reprint requests to Dr Safi, Department of Cardiothoracic and Vascular Surgery, The University of Texas at Houston Medical School, UTH Medical Center, 6410 Fannin St, Suite 450, Houston, TX 77030; e-mail: [email protected].
© 2001 by The Society of Thoracic Surgeons Published by Elsevier Science Inc
(Ann Thorac Surg 2001;72:1225–31) © 2001 by The Society of Thoracic Surgeons
Material and Methods Between February 1991 and March 2000, we performed 685 repairs of either thoracic or thoracoabdominal aortic aneurysms. Nineteen cases utilized hypothermic circulatory arrest for inability to place a clamp and were excluded from the study. In addition, 12 intraoperative deaths occurred and were excluded from the study group. Thus, 654 cases formed the basis of this study. Patient characteristics at the time of repair are listed in Table 1. There were 420 men (64%) and 234 women (36%). Patient age ranged from 8 to 88 years (mean 67 years). Forty-five cases (6.9%) were performed emergently. One hundred and sixty-four (25.1%) involved extent I; 165 0003-4975/01/$20.00 PII S0003-4975(01)02971-X
1226
ESTRERA ET AL NEUROLOGIC OUTCOME AFTER AORTIC ANEURYSM REPAIR
Ann Thorac Surg 2001;72:1225–31
Table 1. Patient Characteristics Variable All patients Age (years) 8 –58 59 – 66 67–72 73– 88 Female Male COPD No COPD Hypertension No HTN Cerebrovascular disease No Dz Acute dissection No acute dissection Chronic dissection No chronic dissection TAAAI No TAAAI TAAAII No TAAAII TAAAIII No TAAAIII TAAAIV No TAAAIV TAAAV No TAAAV Descending No descending Emergent Elective ICS reattach No reattach Clamp time (minutes) 7–27 28 –39 40 –55 56 –154 Adjunct Otherwise
No. Patients (%)
No. Neurologic Deficit (%)
Odds Ratioa
95% CIb
pc
654 (100)
33 (5.0)
162 (24.8) 146 (22.3) 174 (26.6) 172 (26.3) 234 (35.8) 420 (64.2) 219 (33.5) 435 (66.5) 502 (76.8) 152 (23.2) 79 (12.1) 575 (87.9) 31 (4.7) 623 (95.3) 171 (26.2) 483 (73.8) 164 (25.1) 490 (74.9) 165 (25.2) 489 (74.8) 61 (9.3) 593 (90.7) 95 (14.5) 559 (85.5) 23 (3.5) 631 (96.5) 146 (22.5) 507 (77.5) 45 (6.9) 609 (93.1) 332 (51%) 310 (48%)
5 (3.1) 8 (5.4) 11 (6.2) 9 (5.2) 8 (3.4) 25 (6.0) 14 (6.4) 19 (4.4) 26 (5.2) 7 (4.6) 9 (11.4) 24 (4.2) 4 (12.9) 29 (4.7) 7 (4.1) 26 (5.4) 1 (0.6) 32 (6.5) 21 (12.7) 12 (2.5) 6 (9.8) 27 (4.6) 1 (1.1) 32 (5.7) 0 (0.0) 33 (5.2) 4 (2.8) 29 (5.7) 5 (11.1) 28 (4.6) 15 (4.5) 18 (5.8)
1.03
0.99 –1.06
0.58 (0.12)
0.56 1 1.50 1 1.13 1 2.95 1 3.03 1 0.75 1 0.09 1 5.80 1 2.29 1 0.18 1 0.38d 1 0.46 1 1.7 1 0.77 1
0.25–1.26
0.16
0.74 –3.04
0.27
0.48 –2.66
0.78
1.32– 6.61
0.006
1.00 –9.25
0.041
0.32–1.76
0.51
0.01– 0.65
0.003
2.78 –12.07
0.001
0.91–5.78
0.08
0.02–1.30
0.06
0.02– 6.39
0.26
0.16 –1.33
0.14
0.21–13.7
0.17
0.38 –1.55
0.46
157 (24.0) 169 (25.8) 164 (25.1) 164 (25.1) 428 (65.4) 226 (34.5)
1 (0.6) 3 (1.8) 11 (6.7) 18 (11.0) 14 (3.3) 19 (8.4)
1.03
0.02–1.02
0.01 (0.0001)
0.37 1
0.18 – 0.75
0.004
a
For dichotomous variables, the odds ratio represents a test against a reference category whose referent odds ratio is equal to 1. For continuous data, the odds ratio refers to the increase in odds associated with a one-unit increase in the variable value. Although continuous data are presented in quartiles, b the odds ratios are against the continuous variable. 95% CI ⫽ 95% confidence interval. This reflects the units against which its companion odds ratio c is computed. Confidence intervals are test based. p ⫽ probability of Type I statistical error (common p value). Values without parentheses are 2 d Pearson probabilities. Probability values in parentheses are univariate logistic regression likelihood ratio p values. Odds ratio computed using the logit method with a zero-cell correction. COPD ⫽ chronic obstructive pulmonary disease; nal aortic aneurysm.
Dz ⫽ disease;
HTN ⫽ hypertension;
(25.2%) extent II; 61 (9.3%) extent III; 95 (14.5%) extent IV; 23 (3.5%) extent V; and 146 (22.5%) descending thoracic. The adjuncts cerebrospinal fluid drainage and distal aortic perfusion were used in 428 (65%) cases. This was compared with the group of 226 (35%) cases who did not have the combined adjuncts applied, and were operated
ICS ⫽ intercostal space;
TAAA ⫽ thoracoabdomi-
upon utilizing either the simple clamp-sew, or distal aortic perfusion only. The details of our technique have been described previously and are reviewed here. The patient is anesthetized and intubated using a double-lumen endotracheal tube. An arterial line and a pulmonary artery
Ann Thorac Surg 2001;72:1225–31
catheter monitors patient hemodynamics through central venous access. A cerebrospinal fluid (CSF) catheter placed in the third or fourth lumbar space provides CSF drainage and monitoring of CSF pressure. The CSF pressure is maintained at less than 10 mm Hg throughout the procedure. The patient is positioned in the right lateral decubitus position. The incision is tailored to complement the extent of the aneurysm [10]. The diaphragm is preserved exposing only the aortic hiatus. The pericardium is opened posterior to the phrenic nerve to expose the left atrium. The patient is anticoagulated with 1 mg/kg heparin. The left atrium is cannulated through the left pulmonary vein or the left atrial appendage. A BioMedicus (Minneapolis, MN) pump with an in-line heat exchanger is attached to this cannula and the arterial inflow is established through the left common femoral artery, or the descending thoracic aorta, if the femoral artery is not accessible. For the proximal anastomosis of the descending thoracic, extent I, or II thoracoabdominal aneurysms, the inclusion technique is not used and complete transection of the aorta is undertaken to separate the aorta from the underlying esophagus [1]. Distal aortic perfusion is begun as the cross-clamp is applied. For extensive aneurysms, sequential clamping is used, and after completion of the proximal anastomosis, the mid-descending aortic clamp is moved distally to the infrarenal aorta to accommodate intercostal and visceral reattachment. Reattachment of patent, lower intercostal arteries (T8 to T12) is performed routinely except in cases of occluded arteries, heavily calcified or atheromatous aorta, or when technically not feasible. With completion of intercostal reattachment, the abdominal aorta, above the distal clamp, is opened and the visceral vessels are identified. The celiac, superior mesenteric, and renal arteries are perfused using no. 9 Pruitt catheters (Cryolife, St. Petersburg, FL) (Fig 1). The cold perfusate (blood at 4°C) that is delivered to the viscera is dependent on proximal aortic pressure and maintained between 300 and 600 mL/min. Renal temperature is directly monitored and maintained at approximately 15°C. Because cold visceral perfusion may cause hypothermia, core body temperature is kept between 32°C and 33°C by warming the lower circulation (ie, lower extremities). If a distal clamp cannot be placed due to technical reasons and distal circulatory warming cannot be performed, cold visceral perfusion is avoided to prevent core body cooling that may result in cardiac dysrhythmias. Before completion of the distal anastomosis, the graft is flushed proximally and the aorta distally. The patient is weaned from partial bypass once the rectal temperature has reached 36°C. Protamine is administered and the atrial and femoral cannulae are removed. Postoperatively, the mean arterial pressure is maintained between 80 and 100 mm Hg. CSF is drained to maintain a CSF pressure of less than 10 mm Hg for 3 days. If a delayed neurologic deficit appears after removal of the drain, the CSF drain is immediately rein-
ESTRERA ET AL NEUROLOGIC OUTCOME AFTER AORTIC ANEURYSM REPAIR
1227
Fig 1. Distal aortic perfusion with selective visceral perfusion for repair of extent II thoracoabdominal aortic aneurysm.
serted to decrease the CSF pressure. This may lead to the resolution of the deficit [11].
Outcome Variables All preoperative, intraoperative, and postoperative data were collected prospectively over a 9-year period and entered into a database. Analysis was retrospective. Preoperative characteristics analyzed in this study and listed in Table 1 included age, gender, hypertension, cerebrovascular disease, chronic obstructive pulmonary disease, acute dissection, chronic dissection, and aneurysm extent. Intraoperative data included aortic cross-clamp time and the combined adjuncts (cerebrospinal fluid drainage and distal aortic perfusion) used. Thoracoabdominal aortic aneurysms were classified according to the modified Crawford classification (Fig 2). Operative mortality was defined as death occurring within 30 days of surgery. In-hospital mortality was defined as death occurring during hospitalization. Early neurologic deficit was defined as either paraplegia or paraparesis observed upon the patient awakening from anesthesia, regardless of severity. Any neurologic deficit that occurred after a period of observed normal neurologic function was considered delayed neurologic deficit. Patients with stroke were identified by a thorough neurologic examination and computed tomographic scan of the head. Aneurysms with dissection were considered acute if surgery was performed in less than 14 days from
1228
ESTRERA ET AL NEUROLOGIC OUTCOME AFTER AORTIC ANEURYSM REPAIR
Fig 2. Classification of thoracoabdominal aortic aneurysms (modified Crawford classification).
the onset of pain, and chronic if after 14 days. Cerebrovascular disease is defined as any previous history of cerebrovascular accident, or having undergone intervention for carotid artery disease. Emergent refers to cases that were performed for rupture, whether contained or free-rupture. Data were collected from chart reviews done by a trained nurse abstractor, and were entered into a dedicated Microsoft Access database. Data were exported to SAS for analysis, and all computations were performed using SAS version 6.12 running under Windows 2000. Univariate categorical data were analyzed using contingency table analyses. For 2 ⫻ 2 tables, common odds ratios with test-based confidence intervals were computed, and 2 statistics are reported for hypothesis tests. Where a zero cell was present in a table, odds ratio was computed using a zero-cell correction and the confidence interval was computed using the logit method. For tables greater than 2 ⫻ 2, 2 statistics were computed for hypothesis tests, and univariate logistic regression estimates were also computed keeping data in their native continuous distribution. Values of p and confidence intervals for continuous data are based on maximumlikelihood estimates. The null hypothesis was rejected at p less than 0.05.
Ann Thorac Surg 2001;72:1225–31
When the combined adjuncts of cerebrospinal fluid drainage and distal aortic perfusion were used during repair, the incidence of neurologic deficit for all aneurysm types was 3.3% (14 of 428), compared with 8.4% (19 of 226) when the combined adjuncts were not used ( p ⫽ 0.004). For extent II TAAA, the incidence of neurologic deficit was 7.8% (10 of 129) with adjuncts, versus 30.6% (11 of 36) without adjuncts ( p ⬍ 0.001). Logistic regression analysis was performed comparing the risk of neurologic deficit with aortic ischemic time, and demonstrated that the use of the combined adjuncts was protective. The greatest reduction in the incidence of neurologic deficit was observed in the extent II TAAA group ( p ⫽ 0.003) (Fig 3). Multivariate analysis demonstrated that the independent predictors of neurologic deficit were history of cerebrovascular disease ( p ⫽ 0.0019), extent II TAAA ( p ⫽ 0.0001), extent III TAAA ( p ⫽ 0.0002), and acute dissection ( p ⫽ 0.049). The combination of the adjuncts cerebrospinal fluid drainage and distal aortic perfusion were protective against neurologic deficit ( p ⫽ 0.0001) (Table 2). The incidence of delayed neurologic deficit was 2.4% (16 of 654). Cerebrospinal fluid drainage was utilized in 14 of these patients. The incidence of improved neurologic status was 56% (9 of 16).
Comment Despite improvements in the perioperative care of patients with thoracic and thoracoabdominal aortic aneurysms, the morbidity and mortality of surgical repair remain significant. Neurologic deficits (ie, paraplegia and paraparesis) continue to be devastating consequences, and thus, much attention has focused on prevention.
Results The 30-day mortality for repair of thoracic and thoracoabdominal aortic aneurysms was 14% (94 of 654). The 30-day mortality for elective and emergent repair was 13.7% (85 of 620) and 26.5% (9 of 34), respectively. The in-hospital mortality was 16% (106 of 654). The incidence of neurologic deficit for all evaluable patients was 5.0% (33 of 654). The distribution of patients suffering neurologic deficit based on extent of aneurysm is listed in Table 1. Univariate analysis revealed that preoperative factors that were predictive of neurologic deficits were history of cerebrovascular disease ( p ⫽ 0.006), extent III, and extent II TAAA ( p ⫽ 0.001). For operative factors, the aortic ischemic time was an independent predictor of neurologic deficit ( p ⫽ 0.01).
Fig 3. Logistic regression curve: risk of neurologic deficit versus aortic cross-clamp time; adjunct versus no adjunct for extent II thoracoabdominal aortic aneurysm.
Ann Thorac Surg 2001;72:1225–31
ESTRERA ET AL NEUROLOGIC OUTCOME AFTER AORTIC ANEURYSM REPAIR
Table 2. Multiple Logistic Regression Model
Variable Intercept Cerebrovascular disease TAAAII TAAAIII Adjunct
Parameter Estimate
Adjusted Odds Ratio
95% CI
p
⫺3.9176 1.4320
4.19
1.70 –10.32
0.0019
2.8275 2.3403 ⫺1.6296
16.90 10.39 0.20
6.23– 45.89 3.07–35.13 0.09 – 0.44
0.0001 0.0002 0.0001
Adjunct ⫽ combined distal aortic perfusion and cerebrospinal fluid drainage; TAAA ⫽ thoracoabdominal aortic aneurysm.
Many spinal cord protective strategies have been developed with varying success, and it has been with these adjuncts that significant improvements in neurologic outcome have been attained. Previous studies analyzing cerebrospinal fluid drainage and distal aortic perfusion individually did not demonstrate a dramatic benefit in the prevention of neurologic deficit [12, 13]. The limitations of these studies have been cited previously [3, 5]. More recently, a review of cerebrospinal fluid drainage during thoracoabdominal aortic aneurysm repair demonstrated that the benefit might not be substantiated [14]. However, since those earlier studies that found no advantages in the use of cerebrospinal fluid drainage or distal aortic perfusion, we have demonstrated the importance of the combination of these adjuncts upon neurologic outcome [3, 15]. Moreover, recent studies have shown distal aortic perfusion and cerebrospinal fluid drainage to be effective in reducing neurologic deficit when these adjuncts are applied separately [5, 16, 17]. Thus, it is our continued approach to use these adjuncts in combination in order to increase the tolerance of the spinal cord to ischemia. Improving “ischemic tolerance” of the spinal cord theoretically occurs by reversing the decreased distal aortic pressure and increased cerebrospinal fluid pressure during aortic cross-clamping [15]. A decrease in distal aortic pressure leads to a decrease in spinal artery perfusion pressure. A concomitant rise in CSF pressure from aortic cross-clamping may also lead to decreased spinal cord perfusion resulting in spinal cord ischemia. The combination of distal aortic perfusion and cerebrospinal fluid drainage may augment the spinal cord perfusion and provide spinal cord ischemic tolerance. In previous reports, the extent of the thoracoabdominal aortic aneurysm has been associated with neurologic deficits, implicating both extent I and II as having the highest risk [1, 2]. In this study, extent II TAAA was significantly associated with neurologic deficits ( p ⫽ 0.001). However, the incidence of neurologic deficits for extent I TAAAs was 0.6% (1 of 164) and was not significantly associated with neurologic outcome. A recent contemporary report has also reported a low incidence of
1229
neurologic deficit with extent I TAAA [18]. With the use of the combined adjuncts, the beneficial effect of preventing neurologic deficit with non-high-risk TAAAs may become statistically less dramatic. Thus, with use of the combined adjuncts, extent I TAAAs may be similar to descending thoracic, extent IV, or extent V TAAAs in regard to neurologic deficit. Extent III TAAA, however, was shown to be significantly associated with neurologic deficit ( p ⫽ 0.0002). In reclassifying TAAAs, many of the extent V TAAAs, which have a low risk for neurologic deficit, were previously classified as extent III. Thus, the reclassification and reduction of cases in the extent III TAAA group may be responsible for this extent becoming associated with neurologic deficit. The use of the combined adjunct of cerebrospinal fluid drainage and distal aortic perfusion was demonstrated to improve neurologic outcome for the entire group of thoracic and thoracoabdominal aortic aneurysm repairs ( p ⫽ 0.04). Though a reduction in the incidence of neurologic deficits was observed with each extent of aneurysm when the combined adjunct was used, the greatest reduction was noted for extent II thoracoabdominal aortic aneurysms (30.6% neurologic deficit without adjunct vs 7.8% with adjunct, p ⬍ 0.001). A history of cerebrovascular disease was demonstrated to be an independent risk factor for neurologic deficit in this study. The significance of this is uncertain at this time, and its relevance is under further investigation. Moreover, intercostal artery reattachment has previously been demonstrated to be beneficial in regard to neurologic deficit [19]. Although we emphasize the importance of intercostal artery reattachment, its significance to spinal cord protection was inconclusive here ( p ⫽ 0.46). The limitation of this study is that it is retrospective. In addition, our current strategy for providing ischemic tolerance to the spinal cord does include intercostal artery reattachment, mild heparinization, and moderate hypothermia, but these factors were not controlled throughout the length of the study period. The study was not randomized because it represented an evolution in technique since 1991. With modern techniques, early neurologic deficits have fallen to below 5% in all TAAAs except extent II TAAA. Consequently, the beneficial effect of adjuncts is most dramatic in patients with extent II aneurysms. In extent II aneurysms, adjuncts continue to make a considerable difference in outcome and to provide “ischemic tolerance” and significant protection against spinal cord morbidity. Though this study supports the combined use of distal aortic perfusion and cerebrospinal fluid drainage for the prevention of neurologic deficits, it must be reiterated that the cause of neurologic deficits after repair of descending and thoracoabdominal aortic aneurysms is multifactorial. Future research should focus on spinal cord protection in patients with high-risk extent II TAAAs.
1230
ESTRERA ET AL NEUROLOGIC OUTCOME AFTER AORTIC ANEURYSM REPAIR
Ann Thorac Surg 2001;72:1225–31
References 1. Svensson LG, Crawford ES, Hess KR, Coselli JS, Safi HJ. Experience with 1509 patients undergoing thoracoabdominal aortic operations. J Vasc Surg 1993;17:357–70. 2. Crawford ES, Crawford JL, Safi HJ, et al. Thoracoabdominal aortic aneurysms: preoperative and intraoperative factors determining immediate and long-term results of operations in 605 patients. J Vasc Surg 1986;3:389 – 404. 3. Safi HJ, Hess KR, Randel M, et al. Cerebrospinal fluid drainage and distal aortic perfusion: reducing neurologic complications in repair of thoracoabdominal aortic aneurysm types I and II. J Vasc Surg 1996;23:223–9. 4. Cambria RP, Davison JK, Zannetti S, et al. Clinical experience with epidural cooling for spinal cord protection during thoracic and thoracoabdominal aneurysm repair. J Vasc Surg 1997;25:234– 43. 5. Svensson LG, Hess KR, D’Agostino RS, et al. Reduction of neurologic injury after high-risk thoracoabdominal aortic operation. Ann Thorac Surg 1998;66:132– 8. 6. McCullough JL, Hollier LH, Nugent M. Paraplegia after thoracic aortic occlusion: influence of cerebrospinal fluid drainage. Experimental and early clinical results. J Vasc Surg 1988;7:153– 60. 7. Acher CW, Wynn MM, Hoch JR, Popic P, Archibald J, Turnipseed WD. Combined use of cerebral spinal fluid drainage and naloxone reduces the risk of paraplegia in thoracoabdominal aneurysm repair. J Vasc Surg 1994;19: 236– 48. 8. de Haan P, Kalkman CJ, de Mol BA, Ubags LH, Veldman DJ, Jacobs MJ. Efficacy of transcranial motor-evoked myogenic potentials to detect spinal cord ischemia during operations for thoracoabdominal aneurysms. J Thorac Cardiovasc Surg 1997;113:87–101. 9. Jacobs MJ, Meylaerts SA, de Haan P, de Mol BA, Kalkman CJ. Strategies to prevent neurologic deficit based on motor-
10. 11. 12.
13.
14.
15.
16. 17. 18. 19.
evoked potentials in type I and II thoracoabdominal aortic aneurysm repair. J Vasc Surg 1999;29:48–59. Safi HJ. How I do it: thoracoabdominal aortic aneurysm graft replacement. Cardiovasc Surg 1999;7:607–13. Huynh TT, Miller CC, 3rd, Safi HJ. Delayed onset of neurologic deficit: significance and management. Semin Vasc Surg 2000;13:340– 4. Crawford ES, Svensson LG, Hess KR, et al. A prospective randomized study of cerebrospinal fluid drainage to prevent paraplegia after high-risk surgery on the thoracoabdominal aorta. J Vasc Surg 1991;13:36– 46. Svensson LG, Crawford ES, Hess KR, Coselli JS, Safi HJ. Variables predictive of outcome in 832 patients undergoing repairs of the descending thoracic aorta. Chest 1993;104: 1248–53. Ling E, Arellano R. Systematic overview of the evidence supporting the use of cerebrospinal fluid drainage in thoracoabdominal aneurysm surgery for prevention of paraplegia. Anesthesiology 2000;93:1115–22. Safi HJ, Campbell MP, Miller CC, 3rd, et al. Cerebral spinal fluid drainage and distal aortic perfusion decrease the incidence of neurological deficit: the results of 343 descending and thoracoabdominal aortic aneurysm repairs. Eur J Vasc Endovasc Surg 1997;14:118–24. Coselli JS, LeMaire SA. Left heart bypass reduces paraplegia rates after thoracoabdominal aortic aneurysm repair. Ann Thorac Surg 1999;67:1931– 8. 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;67:1963– 80. Coselli JS, LeMaire SA, Miller CC, 3rd, et al. Mortality and paraplegia after thoracoabdominal aortic aneurysm repair: a risk factor analysis. Ann Thorac Surg 2000;69:409–14. Safi HJ, Miller CC, 3rd, Carr C, Iliopoulos DC, Dorsay DA, Baldwin JC. Importance of intercostal artery reattachment during thoracoabdominal aortic aneurysm repair. J Vasc Surg 1998;27:58– 68.
DISCUSSION DR LARS SVENSSON (Burlington, MA): I am grateful to the Society for asking me to comment on this excellent presentation by Dr Estrera from Dr Safi’s group. The authors kindly provided me with the paper, which is well written and referenced, presenting their latest thoughts on the prevention of paraplegia and paraparesis. The authors have added to the literature confirming that both distal aortic perfusion and CSF drainage decreases the incidence of postoperative paraplegia/ paraparesis. There are several papers, including prospective randomized studies, showing the validity of using distal aortic perfusion and also CSF drainage. The most recent paper to be published by Dr Coselli in patients with Type I and Type II thoracoabdominal aneurysms also showed a protective effect. In 1998, we reported a prospective randomized study of Type I and Type II thoracoabdominal aneurysm repair with the addition of active cooling by atrial-femoral bypass. In logistic regression comparison of the aortic cross-clamp versus the incidence of neurologic injury, there was a significant protective effect with CSF drainage and combined intrathecal papaverine and also with active cooling on atrio-femoral bypass. The greatest protective effect, with very marked shifting of the curve to the right, was with the combination of all these protective measures (CSF drainage, intrathecal papaverine, and active cooling). Thus, there is good evidence, including the study presented today, that CSF drainage with atrio-femoral bypass is protective and, in addition, based on our data, it would appear that active cooling prior to aortic cross-clamping is also further protective.
It should also be noted that CSF drainage was first advocated by Drs Blaisdell and Cooley nearly 40 years ago and only now has received widespread acceptance. I have several questions for the authors. First, the authors do not state in the paper to what temperature they allow their patients to cool before aortic cross-clamping, if at all. What was their systemic temperature policy during the study and has it changed over time? Do they have any data as to what the systemic temperature was when they cross-clamped the aorta and its influence on outcome? They do not report examining the influence of date of operation on results. It would be interesting to know how this influenced their results, since presumably the patients done in the earlier part of the study did not necessarily all have atriofemoral bypass with CSF drainage. This may have affected the improved results with the combination of treatments. Did all the patients in the adjunct group have both CSF drainage and atrio-femoral bypass? If the patients only had CSF drainage or atrio-femoral bypass alone, were there benefits found in comparison with patients who had neither? In their study, the Type III thoracoabdominal aneurysms had a higher than expected rate of paraplegia and paraparesis. Clearly, this is influenced by their reclassification of some of their patients in their new classification of Type V, however, was atrio-femoral bypass used less frequently in these patients as has been advocated by some authors? In this study, the authors drained CSF when the pressure exceeded 10 cm of mercury. Is there any reason why the authors
Ann Thorac Surg 2001;72:1225–31
do not allow free gravity drainage during the period of aortic cross-clamping, a method that we have used? Do the authors have any data on the amount of CSF drained during the period of aortic cross-clamping and its influence on neurologic injury? In the multivariate analysis, the authors found for the first time that cerebrovascular disease was a risk factor for postoperative paraplegia and paraparesis, namely 11.4% versus 4.2%. Can the authors comment on why they think this may have been a factor in their study that has not been reported previously? Finally, the authors have advocated CSF drainage for delayed paraplegia after surgery. It is our impression that this is an increasingly vexing problem as we are more effective in preventing initial paraplegia and paraparesis. It would be interesting to know what the authors’ results have been with delayed insertion or repeat insertion of an intrathecal catheter for CSF drainage. The authors are to be congratulated on a fine paper and their excellent results, and confirming the trend and the decreased incidence of paraplegia and paraparesis after both descending and thoracoabdominal aneurysm repairs, particularly with the use of the adjuncts of distal aortic perfusion and CSF drainage. Thank you. DR ESTRERA: I would like to thank Dr Svensson for all his kind comments and the very intriguing questions. In reference to systemic temperature, we do not actively cool prior to aortic cross-clamping. We use passive hypothermia to about 32°C. We do actively cool the viscera to approximately 15°C, as with an extent II thoracoabdominal aortic aneurysm, once the aortic cross-clamps are applied. The benefit of using distal aortic perfusion with an in-line heat exchanger is related to the prevention of dysrhythmias that may occur with visceral cooling. In reference to the systemic temperature before clamping, we have gathered these data, but we have not analyzed it. In reference to the subgroupings of the adjunct group, all patients in the combined adjunct group used both distal aortic perfusion and cerebrospinal fluid drainage. The noncombined group had only distal aortic perfusion or the just clamp and sew. There were no cases of cerebrospinal fluid drainage alone with thoracoabdominal aortic aneurysm repair. Because the event rate was relatively low, the benefit of just distal aortic perfusion was not significant. In reference to the extent III thoracoabdominal aortic aneurysm, it was previously determined that the extent III group had a low probability of neurologic deficit. However, since our reclassification in 1997 adding the extent V, some of the extent III aneurysms were now reclassified as extent V. Because the extent
ESTRERA ET AL NEUROLOGIC OUTCOME AFTER AORTIC ANEURYSM REPAIR
1231
V aneurysms appear to have a similar neurologic deficit rate as the extent IV aneurysms, this may have caused the extent III aneurysms to become a risk factor for neurologic deficit. Why did Dr Safi reclassify these aneurysms? The extent V group was conceived to include those aneurysms that did not exactly fit the criteria of the extent I or the extent III aneurysm. We did not use the combined adjuncts less frequently with the extent III aneurysms. In reference to cerebrospinal fluid drainage, we used open drainage of the cerebrospinal fluid only when the pressure exceeded 10 mm Hg during the case. Data on the exact amount of cerebrospinal fluid have been collected, but we have not analyzed this as to its relation to neurologic outcome. In reference to the history of cerebrovascular disease as a significant factor for neurologic deficit, we are currently looking at this finding, and its significance is uncertain. Finally, we have had 15 cases of delayed paraplegia. We feel that this is different from immediate paraplegia because the events related to this complication appear to be different from those that occur intraoperatively. Our analysis of these cases has revealed blood pressure instability (ie, low blood pressure) as a significant factor. We do advocate the use of cerebrospinal fluid drainage when this complication does occur, and our success rate has been around 50%. Success appears to be dependent upon early identification and reinsertion of the drain. DR C. E. ANAGNOSTOPOULOS (New York, NY and Ioannina, Greece): Did the incidence of reattachment of intercostal arteries T8 to T12 vary in the extent II from the other groups, and have you tried Griepp’s method of successive ligation instead of reattachment for ischemic preconditioning? DR ESTRERA: We have not utilized ligation for ischemic preconditioning. Reattachment of the intercostal arteries did vary depending on patency of T8 to T12. It is our policy to reattach whenever possible. What is important to understand is this study is a retrospective analysis, and it is an evolution in Dr Safi’s technique over the past 9 years. As you have seen, we have tried to provide a method that allows ischemic tolerance to the spinal cord, and there are many methods to accomplish this. Factors that cause paraplegia are multifactorial, and there is not just one method to prevent this. What we have shown here is the benefit of our technique, which is the use of cerebrospinal fluid drainage and distal aortic perfusion.
Background. Neurologic deficit (paraparesis and paraplegia) after repair of the thoracic and thoracoabdominal aorta remains a devastating complication. The purpose of this study was to determine the effect of cerebrospinal fluid drainage and distal aortic perfusion upon neurologic outcome during repair of thoracic and thoracoabdominal aortic aneurysm (TAAA) repair. Methods. Between February 1991 and March 2000, we performed 654 repairs of the thoracic and thoracoabdominal aorta. The median age was 67 years and 420 (64%) patients were male. Forty-five cases (6.9%) were performed emergently. Distribution of TAAA was the following: extent I, 164 (25%); extent II, 165 (25%); extent III, 61 (9%); extent IV, 95 (15%); extent V, 23 (3.5%); and descending thoracic, 147 (22%). The adjuncts cerebrospinal fluid drainage and distal aortic perfusion were used in 428 cases (65%). Results. Thirty-day mortality was 14% (94 of 654). The in-hospital mortality was 16% (106 of 654). Early neurologic deficits occurred in 33 patients (5.0%). Overall, 14 of
428 (3.3%) neurologic deficits were observed in the adjunct group, and 19 of 226 (8.4%) in the nonadjunct group (p ⴝ 0.004). When the adjuncts were used during extent II repair, the incidence was 10 of 129 (7.8%) compared with 11 of 36 (30.6%) in the nonadjunct group (p < 0.001). Multivariate analysis demonstrated that risk factors for neurologic deficit were cerebrovascular disease and extent of TAAA (II and III) (p < 0.05). Conclusions. The combined adjuncts of distal aortic perfusion and cerebrospinal fluid drainage demonstrated improved neurologic outcome with repair of thoracic and TAAAs. In extent II aneurysms, adjuncts continue to make a considerable difference in the outcome and to provide significant protection against spinal cord morbidity. Future research should focus on spinal cord protection in patients with high-risk extent II aneurysms.
R
or steroids. These adjuncts have had varying degrees of success in preventing neurologic deficit [3–9]. Despite these advances, neurologic deficit still remains a significant obstacle in the treatment of aortic aneurysms. The purpose of this study was to review our experience with repair of thoracic and thoracoabdominal aortic aneurysms and determine the effect of the combined adjuncts of distal aortic perfusion and cerebrospinal fluid drainage on neurologic outcome.
epair of thoracic and thoracoabdominal aortic aneurysms (TAAA) remains a surgical challenge. The need for spinal cord protection during repair of these aneurysms is of paramount importance because the development of neurologic deficits (ie, paraplegia and paralysis) continues to have devastating consequences. In the era of the “cross-clamp and sew” technique, the incidence of neurologic deficits correlated with the extent of the aneurysm (Crawford I-IV) and aortic ischemic time [1, 2]. However, in recent years, advances in surgical technique and new adjuncts that increase the ischemic tolerance of the spinal cord have improved results. Many methods have been devised to protect the spinal cord during aortic surgery. These include active and passive distal aortic perfusion, total cardiopulmonary bypass, profound hypothermic circulatory arrest, direct spinal cord cooling, cerebrospinal fluid drainage, monitoring of somatosensory and motor-evoked potentials, and pharmacologic methods using papaverine, nalaxone, Presented at the Thirty-seventh Annual Meeting of The Society of Thoracic Surgeons, New Orleans, LA, Jan 29 –31, 2001. Address reprint requests to Dr Safi, Department of Cardiothoracic and Vascular Surgery, The University of Texas at Houston Medical School, UTH Medical Center, 6410 Fannin St, Suite 450, Houston, TX 77030; e-mail: [email protected].
© 2001 by The Society of Thoracic Surgeons Published by Elsevier Science Inc
(Ann Thorac Surg 2001;72:1225–31) © 2001 by The Society of Thoracic Surgeons
Material and Methods Between February 1991 and March 2000, we performed 685 repairs of either thoracic or thoracoabdominal aortic aneurysms. Nineteen cases utilized hypothermic circulatory arrest for inability to place a clamp and were excluded from the study. In addition, 12 intraoperative deaths occurred and were excluded from the study group. Thus, 654 cases formed the basis of this study. Patient characteristics at the time of repair are listed in Table 1. There were 420 men (64%) and 234 women (36%). Patient age ranged from 8 to 88 years (mean 67 years). Forty-five cases (6.9%) were performed emergently. One hundred and sixty-four (25.1%) involved extent I; 165 0003-4975/01/$20.00 PII S0003-4975(01)02971-X
1226
ESTRERA ET AL NEUROLOGIC OUTCOME AFTER AORTIC ANEURYSM REPAIR
Ann Thorac Surg 2001;72:1225–31
Table 1. Patient Characteristics Variable All patients Age (years) 8 –58 59 – 66 67–72 73– 88 Female Male COPD No COPD Hypertension No HTN Cerebrovascular disease No Dz Acute dissection No acute dissection Chronic dissection No chronic dissection TAAAI No TAAAI TAAAII No TAAAII TAAAIII No TAAAIII TAAAIV No TAAAIV TAAAV No TAAAV Descending No descending Emergent Elective ICS reattach No reattach Clamp time (minutes) 7–27 28 –39 40 –55 56 –154 Adjunct Otherwise
No. Patients (%)
No. Neurologic Deficit (%)
Odds Ratioa
95% CIb
pc
654 (100)
33 (5.0)
162 (24.8) 146 (22.3) 174 (26.6) 172 (26.3) 234 (35.8) 420 (64.2) 219 (33.5) 435 (66.5) 502 (76.8) 152 (23.2) 79 (12.1) 575 (87.9) 31 (4.7) 623 (95.3) 171 (26.2) 483 (73.8) 164 (25.1) 490 (74.9) 165 (25.2) 489 (74.8) 61 (9.3) 593 (90.7) 95 (14.5) 559 (85.5) 23 (3.5) 631 (96.5) 146 (22.5) 507 (77.5) 45 (6.9) 609 (93.1) 332 (51%) 310 (48%)
5 (3.1) 8 (5.4) 11 (6.2) 9 (5.2) 8 (3.4) 25 (6.0) 14 (6.4) 19 (4.4) 26 (5.2) 7 (4.6) 9 (11.4) 24 (4.2) 4 (12.9) 29 (4.7) 7 (4.1) 26 (5.4) 1 (0.6) 32 (6.5) 21 (12.7) 12 (2.5) 6 (9.8) 27 (4.6) 1 (1.1) 32 (5.7) 0 (0.0) 33 (5.2) 4 (2.8) 29 (5.7) 5 (11.1) 28 (4.6) 15 (4.5) 18 (5.8)
1.03
0.99 –1.06
0.58 (0.12)
0.56 1 1.50 1 1.13 1 2.95 1 3.03 1 0.75 1 0.09 1 5.80 1 2.29 1 0.18 1 0.38d 1 0.46 1 1.7 1 0.77 1
0.25–1.26
0.16
0.74 –3.04
0.27
0.48 –2.66
0.78
1.32– 6.61
0.006
1.00 –9.25
0.041
0.32–1.76
0.51
0.01– 0.65
0.003
2.78 –12.07
0.001
0.91–5.78
0.08
0.02–1.30
0.06
0.02– 6.39
0.26
0.16 –1.33
0.14
0.21–13.7
0.17
0.38 –1.55
0.46
157 (24.0) 169 (25.8) 164 (25.1) 164 (25.1) 428 (65.4) 226 (34.5)
1 (0.6) 3 (1.8) 11 (6.7) 18 (11.0) 14 (3.3) 19 (8.4)
1.03
0.02–1.02
0.01 (0.0001)
0.37 1
0.18 – 0.75
0.004
a
For dichotomous variables, the odds ratio represents a test against a reference category whose referent odds ratio is equal to 1. For continuous data, the odds ratio refers to the increase in odds associated with a one-unit increase in the variable value. Although continuous data are presented in quartiles, b the odds ratios are against the continuous variable. 95% CI ⫽ 95% confidence interval. This reflects the units against which its companion odds ratio c is computed. Confidence intervals are test based. p ⫽ probability of Type I statistical error (common p value). Values without parentheses are 2 d Pearson probabilities. Probability values in parentheses are univariate logistic regression likelihood ratio p values. Odds ratio computed using the logit method with a zero-cell correction. COPD ⫽ chronic obstructive pulmonary disease; nal aortic aneurysm.
Dz ⫽ disease;
HTN ⫽ hypertension;
(25.2%) extent II; 61 (9.3%) extent III; 95 (14.5%) extent IV; 23 (3.5%) extent V; and 146 (22.5%) descending thoracic. The adjuncts cerebrospinal fluid drainage and distal aortic perfusion were used in 428 (65%) cases. This was compared with the group of 226 (35%) cases who did not have the combined adjuncts applied, and were operated
ICS ⫽ intercostal space;
TAAA ⫽ thoracoabdomi-
upon utilizing either the simple clamp-sew, or distal aortic perfusion only. The details of our technique have been described previously and are reviewed here. The patient is anesthetized and intubated using a double-lumen endotracheal tube. An arterial line and a pulmonary artery
Ann Thorac Surg 2001;72:1225–31
catheter monitors patient hemodynamics through central venous access. A cerebrospinal fluid (CSF) catheter placed in the third or fourth lumbar space provides CSF drainage and monitoring of CSF pressure. The CSF pressure is maintained at less than 10 mm Hg throughout the procedure. The patient is positioned in the right lateral decubitus position. The incision is tailored to complement the extent of the aneurysm [10]. The diaphragm is preserved exposing only the aortic hiatus. The pericardium is opened posterior to the phrenic nerve to expose the left atrium. The patient is anticoagulated with 1 mg/kg heparin. The left atrium is cannulated through the left pulmonary vein or the left atrial appendage. A BioMedicus (Minneapolis, MN) pump with an in-line heat exchanger is attached to this cannula and the arterial inflow is established through the left common femoral artery, or the descending thoracic aorta, if the femoral artery is not accessible. For the proximal anastomosis of the descending thoracic, extent I, or II thoracoabdominal aneurysms, the inclusion technique is not used and complete transection of the aorta is undertaken to separate the aorta from the underlying esophagus [1]. Distal aortic perfusion is begun as the cross-clamp is applied. For extensive aneurysms, sequential clamping is used, and after completion of the proximal anastomosis, the mid-descending aortic clamp is moved distally to the infrarenal aorta to accommodate intercostal and visceral reattachment. Reattachment of patent, lower intercostal arteries (T8 to T12) is performed routinely except in cases of occluded arteries, heavily calcified or atheromatous aorta, or when technically not feasible. With completion of intercostal reattachment, the abdominal aorta, above the distal clamp, is opened and the visceral vessels are identified. The celiac, superior mesenteric, and renal arteries are perfused using no. 9 Pruitt catheters (Cryolife, St. Petersburg, FL) (Fig 1). The cold perfusate (blood at 4°C) that is delivered to the viscera is dependent on proximal aortic pressure and maintained between 300 and 600 mL/min. Renal temperature is directly monitored and maintained at approximately 15°C. Because cold visceral perfusion may cause hypothermia, core body temperature is kept between 32°C and 33°C by warming the lower circulation (ie, lower extremities). If a distal clamp cannot be placed due to technical reasons and distal circulatory warming cannot be performed, cold visceral perfusion is avoided to prevent core body cooling that may result in cardiac dysrhythmias. Before completion of the distal anastomosis, the graft is flushed proximally and the aorta distally. The patient is weaned from partial bypass once the rectal temperature has reached 36°C. Protamine is administered and the atrial and femoral cannulae are removed. Postoperatively, the mean arterial pressure is maintained between 80 and 100 mm Hg. CSF is drained to maintain a CSF pressure of less than 10 mm Hg for 3 days. If a delayed neurologic deficit appears after removal of the drain, the CSF drain is immediately rein-
ESTRERA ET AL NEUROLOGIC OUTCOME AFTER AORTIC ANEURYSM REPAIR
1227
Fig 1. Distal aortic perfusion with selective visceral perfusion for repair of extent II thoracoabdominal aortic aneurysm.
serted to decrease the CSF pressure. This may lead to the resolution of the deficit [11].
Outcome Variables All preoperative, intraoperative, and postoperative data were collected prospectively over a 9-year period and entered into a database. Analysis was retrospective. Preoperative characteristics analyzed in this study and listed in Table 1 included age, gender, hypertension, cerebrovascular disease, chronic obstructive pulmonary disease, acute dissection, chronic dissection, and aneurysm extent. Intraoperative data included aortic cross-clamp time and the combined adjuncts (cerebrospinal fluid drainage and distal aortic perfusion) used. Thoracoabdominal aortic aneurysms were classified according to the modified Crawford classification (Fig 2). Operative mortality was defined as death occurring within 30 days of surgery. In-hospital mortality was defined as death occurring during hospitalization. Early neurologic deficit was defined as either paraplegia or paraparesis observed upon the patient awakening from anesthesia, regardless of severity. Any neurologic deficit that occurred after a period of observed normal neurologic function was considered delayed neurologic deficit. Patients with stroke were identified by a thorough neurologic examination and computed tomographic scan of the head. Aneurysms with dissection were considered acute if surgery was performed in less than 14 days from
1228
ESTRERA ET AL NEUROLOGIC OUTCOME AFTER AORTIC ANEURYSM REPAIR
Fig 2. Classification of thoracoabdominal aortic aneurysms (modified Crawford classification).
the onset of pain, and chronic if after 14 days. Cerebrovascular disease is defined as any previous history of cerebrovascular accident, or having undergone intervention for carotid artery disease. Emergent refers to cases that were performed for rupture, whether contained or free-rupture. Data were collected from chart reviews done by a trained nurse abstractor, and were entered into a dedicated Microsoft Access database. Data were exported to SAS for analysis, and all computations were performed using SAS version 6.12 running under Windows 2000. Univariate categorical data were analyzed using contingency table analyses. For 2 ⫻ 2 tables, common odds ratios with test-based confidence intervals were computed, and 2 statistics are reported for hypothesis tests. Where a zero cell was present in a table, odds ratio was computed using a zero-cell correction and the confidence interval was computed using the logit method. For tables greater than 2 ⫻ 2, 2 statistics were computed for hypothesis tests, and univariate logistic regression estimates were also computed keeping data in their native continuous distribution. Values of p and confidence intervals for continuous data are based on maximumlikelihood estimates. The null hypothesis was rejected at p less than 0.05.
Ann Thorac Surg 2001;72:1225–31
When the combined adjuncts of cerebrospinal fluid drainage and distal aortic perfusion were used during repair, the incidence of neurologic deficit for all aneurysm types was 3.3% (14 of 428), compared with 8.4% (19 of 226) when the combined adjuncts were not used ( p ⫽ 0.004). For extent II TAAA, the incidence of neurologic deficit was 7.8% (10 of 129) with adjuncts, versus 30.6% (11 of 36) without adjuncts ( p ⬍ 0.001). Logistic regression analysis was performed comparing the risk of neurologic deficit with aortic ischemic time, and demonstrated that the use of the combined adjuncts was protective. The greatest reduction in the incidence of neurologic deficit was observed in the extent II TAAA group ( p ⫽ 0.003) (Fig 3). Multivariate analysis demonstrated that the independent predictors of neurologic deficit were history of cerebrovascular disease ( p ⫽ 0.0019), extent II TAAA ( p ⫽ 0.0001), extent III TAAA ( p ⫽ 0.0002), and acute dissection ( p ⫽ 0.049). The combination of the adjuncts cerebrospinal fluid drainage and distal aortic perfusion were protective against neurologic deficit ( p ⫽ 0.0001) (Table 2). The incidence of delayed neurologic deficit was 2.4% (16 of 654). Cerebrospinal fluid drainage was utilized in 14 of these patients. The incidence of improved neurologic status was 56% (9 of 16).
Comment Despite improvements in the perioperative care of patients with thoracic and thoracoabdominal aortic aneurysms, the morbidity and mortality of surgical repair remain significant. Neurologic deficits (ie, paraplegia and paraparesis) continue to be devastating consequences, and thus, much attention has focused on prevention.
Results The 30-day mortality for repair of thoracic and thoracoabdominal aortic aneurysms was 14% (94 of 654). The 30-day mortality for elective and emergent repair was 13.7% (85 of 620) and 26.5% (9 of 34), respectively. The in-hospital mortality was 16% (106 of 654). The incidence of neurologic deficit for all evaluable patients was 5.0% (33 of 654). The distribution of patients suffering neurologic deficit based on extent of aneurysm is listed in Table 1. Univariate analysis revealed that preoperative factors that were predictive of neurologic deficits were history of cerebrovascular disease ( p ⫽ 0.006), extent III, and extent II TAAA ( p ⫽ 0.001). For operative factors, the aortic ischemic time was an independent predictor of neurologic deficit ( p ⫽ 0.01).
Fig 3. Logistic regression curve: risk of neurologic deficit versus aortic cross-clamp time; adjunct versus no adjunct for extent II thoracoabdominal aortic aneurysm.
Ann Thorac Surg 2001;72:1225–31
ESTRERA ET AL NEUROLOGIC OUTCOME AFTER AORTIC ANEURYSM REPAIR
Table 2. Multiple Logistic Regression Model
Variable Intercept Cerebrovascular disease TAAAII TAAAIII Adjunct
Parameter Estimate
Adjusted Odds Ratio
95% CI
p
⫺3.9176 1.4320
4.19
1.70 –10.32
0.0019
2.8275 2.3403 ⫺1.6296
16.90 10.39 0.20
6.23– 45.89 3.07–35.13 0.09 – 0.44
0.0001 0.0002 0.0001
Adjunct ⫽ combined distal aortic perfusion and cerebrospinal fluid drainage; TAAA ⫽ thoracoabdominal aortic aneurysm.
Many spinal cord protective strategies have been developed with varying success, and it has been with these adjuncts that significant improvements in neurologic outcome have been attained. Previous studies analyzing cerebrospinal fluid drainage and distal aortic perfusion individually did not demonstrate a dramatic benefit in the prevention of neurologic deficit [12, 13]. The limitations of these studies have been cited previously [3, 5]. More recently, a review of cerebrospinal fluid drainage during thoracoabdominal aortic aneurysm repair demonstrated that the benefit might not be substantiated [14]. However, since those earlier studies that found no advantages in the use of cerebrospinal fluid drainage or distal aortic perfusion, we have demonstrated the importance of the combination of these adjuncts upon neurologic outcome [3, 15]. Moreover, recent studies have shown distal aortic perfusion and cerebrospinal fluid drainage to be effective in reducing neurologic deficit when these adjuncts are applied separately [5, 16, 17]. Thus, it is our continued approach to use these adjuncts in combination in order to increase the tolerance of the spinal cord to ischemia. Improving “ischemic tolerance” of the spinal cord theoretically occurs by reversing the decreased distal aortic pressure and increased cerebrospinal fluid pressure during aortic cross-clamping [15]. A decrease in distal aortic pressure leads to a decrease in spinal artery perfusion pressure. A concomitant rise in CSF pressure from aortic cross-clamping may also lead to decreased spinal cord perfusion resulting in spinal cord ischemia. The combination of distal aortic perfusion and cerebrospinal fluid drainage may augment the spinal cord perfusion and provide spinal cord ischemic tolerance. In previous reports, the extent of the thoracoabdominal aortic aneurysm has been associated with neurologic deficits, implicating both extent I and II as having the highest risk [1, 2]. In this study, extent II TAAA was significantly associated with neurologic deficits ( p ⫽ 0.001). However, the incidence of neurologic deficits for extent I TAAAs was 0.6% (1 of 164) and was not significantly associated with neurologic outcome. A recent contemporary report has also reported a low incidence of
1229
neurologic deficit with extent I TAAA [18]. With the use of the combined adjuncts, the beneficial effect of preventing neurologic deficit with non-high-risk TAAAs may become statistically less dramatic. Thus, with use of the combined adjuncts, extent I TAAAs may be similar to descending thoracic, extent IV, or extent V TAAAs in regard to neurologic deficit. Extent III TAAA, however, was shown to be significantly associated with neurologic deficit ( p ⫽ 0.0002). In reclassifying TAAAs, many of the extent V TAAAs, which have a low risk for neurologic deficit, were previously classified as extent III. Thus, the reclassification and reduction of cases in the extent III TAAA group may be responsible for this extent becoming associated with neurologic deficit. The use of the combined adjunct of cerebrospinal fluid drainage and distal aortic perfusion was demonstrated to improve neurologic outcome for the entire group of thoracic and thoracoabdominal aortic aneurysm repairs ( p ⫽ 0.04). Though a reduction in the incidence of neurologic deficits was observed with each extent of aneurysm when the combined adjunct was used, the greatest reduction was noted for extent II thoracoabdominal aortic aneurysms (30.6% neurologic deficit without adjunct vs 7.8% with adjunct, p ⬍ 0.001). A history of cerebrovascular disease was demonstrated to be an independent risk factor for neurologic deficit in this study. The significance of this is uncertain at this time, and its relevance is under further investigation. Moreover, intercostal artery reattachment has previously been demonstrated to be beneficial in regard to neurologic deficit [19]. Although we emphasize the importance of intercostal artery reattachment, its significance to spinal cord protection was inconclusive here ( p ⫽ 0.46). The limitation of this study is that it is retrospective. In addition, our current strategy for providing ischemic tolerance to the spinal cord does include intercostal artery reattachment, mild heparinization, and moderate hypothermia, but these factors were not controlled throughout the length of the study period. The study was not randomized because it represented an evolution in technique since 1991. With modern techniques, early neurologic deficits have fallen to below 5% in all TAAAs except extent II TAAA. Consequently, the beneficial effect of adjuncts is most dramatic in patients with extent II aneurysms. In extent II aneurysms, adjuncts continue to make a considerable difference in outcome and to provide “ischemic tolerance” and significant protection against spinal cord morbidity. Though this study supports the combined use of distal aortic perfusion and cerebrospinal fluid drainage for the prevention of neurologic deficits, it must be reiterated that the cause of neurologic deficits after repair of descending and thoracoabdominal aortic aneurysms is multifactorial. Future research should focus on spinal cord protection in patients with high-risk extent II TAAAs.
1230
ESTRERA ET AL NEUROLOGIC OUTCOME AFTER AORTIC ANEURYSM REPAIR
Ann Thorac Surg 2001;72:1225–31
References 1. Svensson LG, Crawford ES, Hess KR, Coselli JS, Safi HJ. Experience with 1509 patients undergoing thoracoabdominal aortic operations. J Vasc Surg 1993;17:357–70. 2. Crawford ES, Crawford JL, Safi HJ, et al. Thoracoabdominal aortic aneurysms: preoperative and intraoperative factors determining immediate and long-term results of operations in 605 patients. J Vasc Surg 1986;3:389 – 404. 3. Safi HJ, Hess KR, Randel M, et al. Cerebrospinal fluid drainage and distal aortic perfusion: reducing neurologic complications in repair of thoracoabdominal aortic aneurysm types I and II. J Vasc Surg 1996;23:223–9. 4. Cambria RP, Davison JK, Zannetti S, et al. Clinical experience with epidural cooling for spinal cord protection during thoracic and thoracoabdominal aneurysm repair. J Vasc Surg 1997;25:234– 43. 5. Svensson LG, Hess KR, D’Agostino RS, et al. Reduction of neurologic injury after high-risk thoracoabdominal aortic operation. Ann Thorac Surg 1998;66:132– 8. 6. McCullough JL, Hollier LH, Nugent M. Paraplegia after thoracic aortic occlusion: influence of cerebrospinal fluid drainage. Experimental and early clinical results. J Vasc Surg 1988;7:153– 60. 7. Acher CW, Wynn MM, Hoch JR, Popic P, Archibald J, Turnipseed WD. Combined use of cerebral spinal fluid drainage and naloxone reduces the risk of paraplegia in thoracoabdominal aneurysm repair. J Vasc Surg 1994;19: 236– 48. 8. de Haan P, Kalkman CJ, de Mol BA, Ubags LH, Veldman DJ, Jacobs MJ. Efficacy of transcranial motor-evoked myogenic potentials to detect spinal cord ischemia during operations for thoracoabdominal aneurysms. J Thorac Cardiovasc Surg 1997;113:87–101. 9. Jacobs MJ, Meylaerts SA, de Haan P, de Mol BA, Kalkman CJ. Strategies to prevent neurologic deficit based on motor-
10. 11. 12.
13.
14.
15.
16. 17. 18. 19.
evoked potentials in type I and II thoracoabdominal aortic aneurysm repair. J Vasc Surg 1999;29:48–59. Safi HJ. How I do it: thoracoabdominal aortic aneurysm graft replacement. Cardiovasc Surg 1999;7:607–13. Huynh TT, Miller CC, 3rd, Safi HJ. Delayed onset of neurologic deficit: significance and management. Semin Vasc Surg 2000;13:340– 4. Crawford ES, Svensson LG, Hess KR, et al. A prospective randomized study of cerebrospinal fluid drainage to prevent paraplegia after high-risk surgery on the thoracoabdominal aorta. J Vasc Surg 1991;13:36– 46. Svensson LG, Crawford ES, Hess KR, Coselli JS, Safi HJ. Variables predictive of outcome in 832 patients undergoing repairs of the descending thoracic aorta. Chest 1993;104: 1248–53. Ling E, Arellano R. Systematic overview of the evidence supporting the use of cerebrospinal fluid drainage in thoracoabdominal aneurysm surgery for prevention of paraplegia. Anesthesiology 2000;93:1115–22. Safi HJ, Campbell MP, Miller CC, 3rd, et al. Cerebral spinal fluid drainage and distal aortic perfusion decrease the incidence of neurological deficit: the results of 343 descending and thoracoabdominal aortic aneurysm repairs. Eur J Vasc Endovasc Surg 1997;14:118–24. Coselli JS, LeMaire SA. Left heart bypass reduces paraplegia rates after thoracoabdominal aortic aneurysm repair. Ann Thorac Surg 1999;67:1931– 8. 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;67:1963– 80. Coselli JS, LeMaire SA, Miller CC, 3rd, et al. Mortality and paraplegia after thoracoabdominal aortic aneurysm repair: a risk factor analysis. Ann Thorac Surg 2000;69:409–14. Safi HJ, Miller CC, 3rd, Carr C, Iliopoulos DC, Dorsay DA, Baldwin JC. Importance of intercostal artery reattachment during thoracoabdominal aortic aneurysm repair. J Vasc Surg 1998;27:58– 68.
DISCUSSION DR LARS SVENSSON (Burlington, MA): I am grateful to the Society for asking me to comment on this excellent presentation by Dr Estrera from Dr Safi’s group. The authors kindly provided me with the paper, which is well written and referenced, presenting their latest thoughts on the prevention of paraplegia and paraparesis. The authors have added to the literature confirming that both distal aortic perfusion and CSF drainage decreases the incidence of postoperative paraplegia/ paraparesis. There are several papers, including prospective randomized studies, showing the validity of using distal aortic perfusion and also CSF drainage. The most recent paper to be published by Dr Coselli in patients with Type I and Type II thoracoabdominal aneurysms also showed a protective effect. In 1998, we reported a prospective randomized study of Type I and Type II thoracoabdominal aneurysm repair with the addition of active cooling by atrial-femoral bypass. In logistic regression comparison of the aortic cross-clamp versus the incidence of neurologic injury, there was a significant protective effect with CSF drainage and combined intrathecal papaverine and also with active cooling on atrio-femoral bypass. The greatest protective effect, with very marked shifting of the curve to the right, was with the combination of all these protective measures (CSF drainage, intrathecal papaverine, and active cooling). Thus, there is good evidence, including the study presented today, that CSF drainage with atrio-femoral bypass is protective and, in addition, based on our data, it would appear that active cooling prior to aortic cross-clamping is also further protective.
It should also be noted that CSF drainage was first advocated by Drs Blaisdell and Cooley nearly 40 years ago and only now has received widespread acceptance. I have several questions for the authors. First, the authors do not state in the paper to what temperature they allow their patients to cool before aortic cross-clamping, if at all. What was their systemic temperature policy during the study and has it changed over time? Do they have any data as to what the systemic temperature was when they cross-clamped the aorta and its influence on outcome? They do not report examining the influence of date of operation on results. It would be interesting to know how this influenced their results, since presumably the patients done in the earlier part of the study did not necessarily all have atriofemoral bypass with CSF drainage. This may have affected the improved results with the combination of treatments. Did all the patients in the adjunct group have both CSF drainage and atrio-femoral bypass? If the patients only had CSF drainage or atrio-femoral bypass alone, were there benefits found in comparison with patients who had neither? In their study, the Type III thoracoabdominal aneurysms had a higher than expected rate of paraplegia and paraparesis. Clearly, this is influenced by their reclassification of some of their patients in their new classification of Type V, however, was atrio-femoral bypass used less frequently in these patients as has been advocated by some authors? In this study, the authors drained CSF when the pressure exceeded 10 cm of mercury. Is there any reason why the authors
Ann Thorac Surg 2001;72:1225–31
do not allow free gravity drainage during the period of aortic cross-clamping, a method that we have used? Do the authors have any data on the amount of CSF drained during the period of aortic cross-clamping and its influence on neurologic injury? In the multivariate analysis, the authors found for the first time that cerebrovascular disease was a risk factor for postoperative paraplegia and paraparesis, namely 11.4% versus 4.2%. Can the authors comment on why they think this may have been a factor in their study that has not been reported previously? Finally, the authors have advocated CSF drainage for delayed paraplegia after surgery. It is our impression that this is an increasingly vexing problem as we are more effective in preventing initial paraplegia and paraparesis. It would be interesting to know what the authors’ results have been with delayed insertion or repeat insertion of an intrathecal catheter for CSF drainage. The authors are to be congratulated on a fine paper and their excellent results, and confirming the trend and the decreased incidence of paraplegia and paraparesis after both descending and thoracoabdominal aneurysm repairs, particularly with the use of the adjuncts of distal aortic perfusion and CSF drainage. Thank you. DR ESTRERA: I would like to thank Dr Svensson for all his kind comments and the very intriguing questions. In reference to systemic temperature, we do not actively cool prior to aortic cross-clamping. We use passive hypothermia to about 32°C. We do actively cool the viscera to approximately 15°C, as with an extent II thoracoabdominal aortic aneurysm, once the aortic cross-clamps are applied. The benefit of using distal aortic perfusion with an in-line heat exchanger is related to the prevention of dysrhythmias that may occur with visceral cooling. In reference to the systemic temperature before clamping, we have gathered these data, but we have not analyzed it. In reference to the subgroupings of the adjunct group, all patients in the combined adjunct group used both distal aortic perfusion and cerebrospinal fluid drainage. The noncombined group had only distal aortic perfusion or the just clamp and sew. There were no cases of cerebrospinal fluid drainage alone with thoracoabdominal aortic aneurysm repair. Because the event rate was relatively low, the benefit of just distal aortic perfusion was not significant. In reference to the extent III thoracoabdominal aortic aneurysm, it was previously determined that the extent III group had a low probability of neurologic deficit. However, since our reclassification in 1997 adding the extent V, some of the extent III aneurysms were now reclassified as extent V. Because the extent
ESTRERA ET AL NEUROLOGIC OUTCOME AFTER AORTIC ANEURYSM REPAIR
1231
V aneurysms appear to have a similar neurologic deficit rate as the extent IV aneurysms, this may have caused the extent III aneurysms to become a risk factor for neurologic deficit. Why did Dr Safi reclassify these aneurysms? The extent V group was conceived to include those aneurysms that did not exactly fit the criteria of the extent I or the extent III aneurysm. We did not use the combined adjuncts less frequently with the extent III aneurysms. In reference to cerebrospinal fluid drainage, we used open drainage of the cerebrospinal fluid only when the pressure exceeded 10 mm Hg during the case. Data on the exact amount of cerebrospinal fluid have been collected, but we have not analyzed this as to its relation to neurologic outcome. In reference to the history of cerebrovascular disease as a significant factor for neurologic deficit, we are currently looking at this finding, and its significance is uncertain. Finally, we have had 15 cases of delayed paraplegia. We feel that this is different from immediate paraplegia because the events related to this complication appear to be different from those that occur intraoperatively. Our analysis of these cases has revealed blood pressure instability (ie, low blood pressure) as a significant factor. We do advocate the use of cerebrospinal fluid drainage when this complication does occur, and our success rate has been around 50%. Success appears to be dependent upon early identification and reinsertion of the drain. DR C. E. ANAGNOSTOPOULOS (New York, NY and Ioannina, Greece): Did the incidence of reattachment of intercostal arteries T8 to T12 vary in the extent II from the other groups, and have you tried Griepp’s method of successive ligation instead of reattachment for ischemic preconditioning? DR ESTRERA: We have not utilized ligation for ischemic preconditioning. Reattachment of the intercostal arteries did vary depending on patency of T8 to T12. It is our policy to reattach whenever possible. What is important to understand is this study is a retrospective analysis, and it is an evolution in Dr Safi’s technique over the past 9 years. As you have seen, we have tried to provide a method that allows ischemic tolerance to the spinal cord, and there are many methods to accomplish this. Factors that cause paraplegia are multifactorial, and there is not just one method to prevent this. What we have shown here is the benefit of our technique, which is the use of cerebrospinal fluid drainage and distal aortic perfusion.