Mortality and Paraplegia After Thoracoabdominal Aortic Aneurysm Repair: A Risk Factor Analysis

Transcription

1 ORIGINAL ARTICLES: CARDIOVASCULAR Mortality and Paraplegia After Thoracoabdominal Aortic Aneurysm Repair: A Risk Factor Analysis Joseph S. Coselli, MD, Scott A. LeMaire, MD, Charles C. Miller III, PhD, Zachary C. Schmittling, MD, Cüneyt Köksoy, MD, José Pagan, MD, and Patrick E. Curling, MD Baylor College of Medicine, The Methodist Hospital, Houston, Texas Background. Recent recommendations regarding thoracoabdominal aortic aneurysm (TAAA) management have emphasized individualized treatment based on balancing a patient s calculated risk of rupture with their anticipated risk of postoperative death or paraplegia. The purpose of this study was to enhance this risk-benefit decision by providing contemporary results and determining which preoperative risk factors currently predict mortality and paraplegia after TAAA surgery. Methods. Risk factor analyses based on data regarding 1,220 consecutive patients undergoing TAAA repair from 1986 through 1998 were performed using multiple logistic regression with step-wise model selection. Results. The 30-day mortality rate was 4.8% (58 of 1,220) and the incidence of paraplegia was 4.6% (56 of 1,206). For elective cases, predictors of operative mortality included renal insufficiency (p ), increasing age (p ), symptomatic aneurysms (p ), and extent II aneurysms (p ). Extent II aneurysms (p ) and diabetes (p ) were predictors of paraplegia. Conclusions. These risk models may assist in decisions regarding elective TAAA operations. For patients who are acceptable candidates, contemporary surgical management provides favorable results. (Ann Thorac Surg 2000;69:409 14) 2000 by The Society of Thoracic Surgeons The more detailed consideration of various risk factors associated with rupture must be balanced against a similarly nuanced consideration of the risk of operation, including not only the risk of death but also of paraplegia. T. Juvonen and associates, 1997 J. Maxwell Chamberlain Memorial Paper [1] The foundation for all patient management decisions involves determining whether the risk related to a disease s natural history outweighs the risk of its treatment. In keeping with this tenet, Juvonen and associates [1] recently emphasized that the decision to proceed with thoracoabdominal aortic aneurysm (TAAA) repair must be based on each individual patient s risk of rupture without operation versus their risk of death or paraplegia with operation. To facilitate the first half of this fundamental risk-versus-benefit analysis, the authors developed a predictive model that estimates a patient s risk of rupture in 1 year if the aneurysm is not repaired. The other half of the assessment, however, requires the development of a separate model. The purpose of our analysis of contemporary results was to determine which preoperative risk factors currently predict early mortality and paraplegia after elective TAAA repair in order to Presented at the Thirty-fifth Annual Meeting of The Society of Thoracic Surgeons, San Antonio, TX, Jan 25 27, Address reprint requests to Dr Coselli, 6560 Fannin, #1100, Houston, TX 77030; jcoselli@bcm.tmc.edu. enhance the risk-benefit decision-making process during management of individual patients. Patients and Methods Patients Between January 1986 and December 1998, 1,532 consecutive patients underwent graft repair of descending thoracic or TAAAs. Of these, 1,220 patients had TAAA repairs; the patients with aneurysms limited to the descending thoracic aorta are not considered further in this study. The patient characteristics at the time of TAAA repair are detailed in Table 1. There were 721 men (59.1%) and 499 women (40.9%). Patient ages ranged from 18 to 88 years (mean 65.7 years, median 68 years). Extensive TAAAs (Crawford extents I and II) were present in 65.1% of patients. Surgical Technique All operations were performed by the senior author (J.S.C.); the details of the surgical technique have been recently described elsewhere [2 4]. With regard to spinal cord protection, a combination of moderate heparinization, permissive mild hypothermia, sequential aortic clamping, and aggressive reattachment of critical intercostal arteries (T8 to L1) was used consistently throughout the entire series. Left heart bypass was used in 381 (31.2%) cases; no blood reservoir, heat exchanger, or oxygenator was incorporated in the bypass circuit. A detailed technical report describing our methods for 2000 by The Society of Thoracic Surgeons /00/$20.00 Published by Elsevier Science Inc PII S (99)

2 410 COSELLI ET AL Ann Thorac Surg RISK ANALYSIS FOR TAAA REPAIR 2000;69: Table 1. Preoperative Characteristics of 1,220 Patients Undergoing Thoracoabdominal Aortic Aneurysm Repair Characteristic No. of Patients (%) Crawford extent Extent I 423 (34.7) Extent II 371 (30.4) Extent III 201 (16.5) Extent IV 225 (18.4) Acute dissection 46 (3.8) Chronic dissection 272 (22.3) Marfan syndrome 72 (5.9) Symptomatic aneurysms 855 (70.1) Acute presentation 112 (9.2) Rupture 76 (6.2) Preoperative paraplegia or paraparesis 16 (1.3) Concurrent aneurysm 224 (18.4) Prior aneurysm repair 502 (41.2) Prior thoracic aortic aneurysm repair 281 (23.0) Diabetes 69 (5.7) Hypertension 940 (77.1) Coronary artery disease 435 (35.7) Prior coronary artery bypass or angioplasty 202 (16.6) Cerebrovascular disease 135 (11.1) Renal arterial occlusive disease 312 (25.6) Renal insufficiency 151 (12.4) Chronic obstructive lung disease 491 (40.3) Peptic ulcer disease 83 (6.8) using left heart bypass during TAAA repair was recently published [5]. Spinal evoked potentials were not monitored and cerebral spinal fluid drainage was not used routinely. Study Variables and Definitions All preoperative, intraoperative, and postoperative data were gathered prospectively over the 13-year period and entered into a database. In addition to patient age and gender, the preoperative characteristics analyzed are listed in Table 1. The aneurysms were classified based on extent as defined by Crawford and associates [6]. Aneurysms associated with aortic dissection were considered acute if surgery was performed within 14 days of the onset of pain; after 14 days, dissection was considered chronic. Patients were considered symptomatic when any symptom (acute or chronic, severe or mild) related to the aneurysm was present, including pain, hoarseness, dysphagia, etc. Patients with acute presentations were defined as those with acute pain, rupture, contained rupture, and complicated acute dissection [7]. Renal insufficiency was defined as serum creatinine exceeding 3.0 mg/dl or need for hemodialysis. All patients with a history of diabetes, regardless of the duration of disease or the current need for antidiabetic agents, were included in the diabetes category. Operative mortality was defined as death occurring within 30 days or within the initial postoperative hospitalization [8]. All patients with postoperative neurologic deficits involving the lower extremities were included in the paraplegia category, regardless of whether the deficit was weakness (paraparesis) or paralysis, immediate or delayed, or transient or permanent. This included patients with unilateral lower extremity deficits, unless an associated deficit involving the ipsilateral upper extremity (indicating a stroke) was present. Renal failure was defined as an increase in serum creatinine to greater than 3.0 mg/dl (if normal preoperatively) or the need to initiate hemodialysis. Pulmonary complications were defined as ventilator support exceeding 48 hours, reintubation, adult respiratory distress syndrome, atelectasis requiring bronchoscopy, chylothorax, prolonged air leak, pleural effusion requiring drainage, pneumonia, or pneumothorax requiring evacuation. Statistical Analyses The statistical analysis was performed using the SAS (release 6.10; SAS Institute, Inc, Cary, NC) and SPSS (release 6.1.3; SPSS, Inc, Chicago, IL) systems for Windows. Risk factors were evaluated for association with operative mortality or paraplegia using univariate analyses: categorical variables were analyzed using the 2 or Fisher s exact test and continuous data were analyzed using Student s t test. Risk factors that emerged with significance levels below 0.25 were analyzed via multiple logistic regression with step-wise model selection. Associations with outcomes were considered statistically significant when p values less than Actuarial survival was estimated using the Kaplan-Meier method. In creating the model for predicting an individual patient s risk of death or paraplegia, only factors relevant to the decision making process (ie, information available at the time of preoperative evaluation) were entered into the analysis. Furthermore, because almost all patients with acute presentations undergo emergency operation, the need for a detailed decision analysis is essentially limited to elective cases; the 112 patients with acute presentations, therefore, were excluded in this portion of the analysis. Results Overall Operative Morbidity and Mortality Four patients (0.3%) died in the operating room. The early operative mortality rate was 7.3% (89 patients), which included day deaths (4.8%) and 87 inhospital deaths (7.1%). The incidence of paraplegia was 4.6% (56 of 1,206; excludes the 10 patients with preoperative paraplegia and 4 patients who died during operation); of these 56 patients, 33 (58.9%) had paraparesis. Renal failure developed in 11.1% of patients (133 of 1,198; excludes 18 patients receiving preoperative hemodialysis and 4 patients who died during operation); 72 of these 133 patients (54.1%) required hemodialysis. Pulmonary complications occurred in 415 patients (34.1%). Twenty-seven patients (2.2%) required reoperation for postoperative bleeding. Stratified results based on aneurysm extent are listed in Table 2. Actuarial survival rates were 88.9% at 1 year, 87.8% at 2 years, 81.6% at 4 years, and 63.7% at 6 years (Fig 1).

3 Ann Thorac Surg COSELLI ET AL 2000;69: RISK ANALYSIS FOR TAAA REPAIR 411 Table 2. Results of Thoracoabdominal Aortic Aneurysm Repair in 1,220 Consecutive Patients Extent No. of Patients 30-Day Survival In-Hospital Survival Risk Analysis for Operative Mortality After Elective Repair This analysis of risk factors predictive of operative mortality in individual patients was based on preoperative variables in the 1,108 patients undergoing elective operation. In this group of patients, the operative mortality rate was 6.3% (70 patients). Univariate analysis revealed the following factors to be associated with operative death (Table 3): increasing age, extent II aneurysm, symptomatic aneurysms, renal arterial occlusive disease, renal insufficiency, and hemodialysis. Based on the significant risk factors determined by multivariable analysis (Table 4), the probability of a patient dying after TAAA repair is predicted by: risk odds/ 1 odds Paraplegia a I 423 (34.7%) 402 (95.0%) 392 (92.7%) 16 (3.9%) II 371 (30.4%) 350 (94.3%) 336 (90.6%) 30 (8.2%) III 201 (16.5%) 192 (95.5%) 192 (95.5%) 7 (3.5%) IV 225 (18.4%) 218 (96.9%) 213 (94.7%) 3 (1.3%) Total 1220 (100%) 1162 (95.2%) 1133 (92.9%) 56 (4.6%) a Excludes 4 patients who died during operation and 10 patients with preoperative paraplegia. where odds exp [(age ) (C ) (symptoms ) (renal ) ]; age patient age in years, C 2 1 for patients with an extent II aneurysm and 0 for patients with an extent I, III, or IV aneurysm; symptoms 1 or 0, respectively, for patients with or without symptoms related to the aneurysm, and renal 1 or 0, respectively, for patients with or without renal insufficiency. Risk Analysis for Paraplegia After Elective Repair The incidence of paraplegia after elective TAAA repair was 3.6% (40/1,099; excludes intraoperative deaths and Table 3. Results (p Values) of Univariate Analyses Regarding Factors Associated With Operative Mortality and Paraplegia After Thoracoabdominal Aortic Aneurysm Repair Variable Operative Mortality Paraplegia Patient age Patient gender Chronic dissection Nondissection Extent I Extent II Extent III Extent IV a Marfan syndrome Symptomatic Prior aneurysm repair a Prior thoracic aneurysm repair a Concurrent aneurysm Hypertension Diabetes mellitus Coronary artery disease Prior CAB or coronary angioplasty Renal arterial occlusive disease Renal insufficiency Hemodialysis Cerebrovascular disease Peptic ulcer disease Chronic obstructive pulmonary disease Preoperative paraplegia a Factor associated with decreased risk. CAB coronary artery bypass. patients with preoperative paraplegia). Age and extent II TAAA were the only risk factors associated with paraplegia in the univariate analysis (Table 3). Extent IV aneurysms and prior thoracic aneurysm repair were associated with decreased risk. The results of the multivariable analysis are listed in Table 4. The probability of a patient developing paraplegia or paraparesis after TAAA repair is predicted by: risk odds/ 1 odds where odds exp [(diabetes ) (C ) ], diabetes 1 or 0, respectively, for patients with or without diabetes, and C 2 1 for patients with an extent II TAAA and 0 for patients with an extent I, III, or IV TAAA. Fig 1. Kaplan-Meier curve demonstrating actuarial survival after TAAA repair in 1,220 consecutive patients. Comment The recent analysis regarding the risk of rupture for patients with thoracic aortic aneurysms by Juvonen and associates [1] was the primary impetus for our analysis concerning elective cases. Unsatisfied with the usual method of assessing the risk of rupture ( big aneurysm vs small aneurysm [Fig 2A]) the Mount Sinai group performed a multivariable analysis that included data from computer-generated three-dimensional computed tomographic reconstructions of the thoracoabdominal

4 412 COSELLI ET AL Ann Thorac Surg RISK ANALYSIS FOR TAAA REPAIR 2000;69: Table 4. Results of Multivariable Analyses for the Patient Risk Models for Operative Mortality and Paraplegia Risk Model Variable Parameter Estimate p Value Odds Ratio 95% Confidence Interval Operative mortality Intercept Age Extent II Symptomatic Renal insufficiency Paraplegia Intercept Extent II Diabetes aorta. The resulting formula determines the probability of rupture within 1 year based on patient age, the presence of pain and chronic obstructive pulmonary disease, and the maximum true diameters of the descending thoracic and abdominal aortic segments (Fig 2B). The risk analysis of our series was undertaken to balance the Mount Sinai model for risk of rupture with a complementary model that predicts operative risk based on contemporary results (Fig 2C). Using the risk formulas presented above, the probabilities of early death and paraplegia (Table 5) after TAAA surgery can be directly calculated for an individual patient. Alternatively, risk stratification curves can be used to rapidly obtain an estimation of risk for any given patient (Fig 3) [1]. For example, using Figure 3 and Table 5, a 75-year-old nondiabetic patient with renal insufficiency and an asymptomatic extent II TAAA would have an 18% risk of operative death and a 5.9% risk of paraplegia. By comparing the calculated risks of death or paraplegia after operation with the calculated risk of rupture without surgery, decisions regarding treatment can be supported with objective data. Increasing age and preoperative renal insufficiency have remained major risk factors for early mortality throughout the history of TAAA repair. Both were among the predictive variables determined by Svensson and associates [9] in their multivariable analysis of Crawford s complete experience with TAAA surgery in 1,509 patients treated between 1960 and The recent report by Acher and associates [10] confirms that, along with acute presentation, age and elevated creatinine levels remain important predictors of early death. The presence of symptoms related to the aneurysm was also an important predictor of operative mortality. This occurred despite excluding all patients with acute presentations in the analysis, leaving only those patients who had chronic or mild symptoms that were not con- Table 5. Predicted Risk for Postoperative Paraplegia After Thoracoabdominal Aortic Aneurysm Repair Based on the Presence of Diabetes and the Extent of Repair Fig 2. Demonstration of the risk versus benefit balance that is central to decisions regarding elective surgical repair of thoracoabdominal aortic aneurysms. Estimation of the risk of rupture was previously based solely on aneurysm size (A). The recently developed Mount Sinai formula provides a means for calculating the probability of rupture (B). New predictive formulas were developed to estimate the risks of postoperative mortality and paraplegia based on contemporary results (C). Modified from Juvonen and associates [1]. Patient Characteristics Probability of Postoperative Paraplegia Nondiabetic Extent I, III, or IV aneurysm 2.25% Nondiabetic Extent II aneurysm 5.86% Diabetic Extent I, III, or IV aneurysm 6.08% Diabetic Extent II aneurysm 14.87%

5 Ann Thorac Surg COSELLI ET AL 2000;69: RISK ANALYSIS FOR TAAA REPAIR 413 Fig 3. Risk stratification curves (based on particular combinations of risk factors) to allow rapid estimation of the risk for operative mortality after elective repair of TAAAs. The probability of operative mortality is determined using the first set of curves (A) for patients with extent II TAAAs and the second set (B) for patients with extent I, III, or IV aneurysms. For patients without renal insufficiency and without symptoms, the age curve is used. The subsequent curves are used for those patients with symptoms, renal insufficiency, or both risk factors, respectively. sidered signs of impending rupture. Similarly, Juvonen and associates [1] documented an increased risk of rupture in patients who had pain that experienced surgeons had characterized as being unrelated to the aneurysm. Along the continuum between truly asymptomatic aneurysms and ruptured aneurysm, the appearance of even mild symptoms seems to represent progression into a subacute phase that carries both an increased risk of rupture and an increased perioperative mortality rate. Therefore, the development of any symptoms, no matter how mild or uncharacteristic, in a patient with a TAAA demands immediate evaluation; the aneurysm must be considered the cause until proven otherwise. If the source of the problem remains unexplained, aneurysm repair should be considered. With regard to paraplegia, extent II aneurysms remain a major risk factor [9, 10]. This high-risk group of patients has benefitted the most from evolving refinements in operative technique and innovations in spinal cord protection. We have recently reported that the use of left heart bypass in patients with extent II TAAAs has reduced the incidence of paraplegia from 13.1% to 4.8% ( p 0.007) [11]. The emergence of diabetes as a predictor of paraplegia was unexpected, and the strength of this association was particularly surprising: the risk of paraplegia in diabetic patients with less extensive aneurysms was similar to that in nondiabetic patients with extent II aneurysms (Table 5). The damaging effects of diabetes on small peripheral arteries, the coronary arteries, and the retinal capillaries suggest that similar insults to the spinal cord s blood supply may impair its ability to tolerate periods of ischemia. Our previous analysis of 660 patients revealed that chronic dissection was not a risk factor for either early mortality or paraplegia after TAAA repair [12]. The current study confirms these findings: chronic dissection was not associated with death or paraplegia. Our models, therefore, can be applied in cases both with and without dissection. In contrast, the formula of Juvonen and associates is only applicable to patients without dissection, because patients with chronic distal dissection were excluded from the analysis [1]. Classically, TAAAs with dissection have been characterized as being more prone to rupture than those without dissection [13]. A prospective risk analysis focusing on the natural history of TAAAs with dissection would be a beneficial counterpart to the new operative risk models. In conclusion, contemporary surgical management of TAAAs provides favorable results for patients who are acceptable candidates. When balanced with models predicting the probability of aneurysm rupture, the operative risk models presented above may assist in decisions regarding elective aortic repair. The predictive accuracy of these formulas, however, will require validation through prospective evaluations. We gratefully acknowledge Autumn Jamison for providing database management, statistical analysis, and invaluable assistance with manuscript preparation. References 1. Juvonen T, Ergin MA, Galla JD, et al. Prospective study of the natural history of thoracic aortic aneurysms. Ann Thorac Surg 1997;63: Coselli JS. Thoracoabdominal aortic aneurysms. In: Yao JST, Pearce WH, eds. Techniques in vascular and endovascular surgery. Stamford: Appleton & Lange, 1998: Coselli JS. Surgical technique, preoperative and intraoperative management of thoracoabdominal aortic aneurysm. In: Yao JST, Pearce WH, eds. Arterial surgery: management of challenging problems. Stamford: Appleton & Lange, 1996: Coselli JS. Thoracoabdominal aortic aneurysm. In: Rutherford RB, ed. Vascular surgery, 4th ed. Philadelphia: WB Saunders, 1994: Coselli JS, LeMaire SA, Ledesma DF, Ohtsubo S, Tayama E, Nosé Y. Initial experience with the Nikkiso centrifugal pump during thoracoabdominal aortic aneurysm repair. J Vasc Surg 1998;27:

6 414 COSELLI ET AL Ann Thorac Surg RISK ANALYSIS FOR TAAA REPAIR 2000;69: Crawford ES, Crawford JL, Safi HJ, et al. Thoracoabdominal aortic aneurysms: preoperative and intraoperative factors determining immediate and long-term results of operation in 605 patients. J Vasc Surg 1986;3: 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 aortic aneurysm repair. J Vasc Surg 1994; 19: Guidelines for data reporting and nomenclature for The Annals of Thoracic Surgery. Ann Thorac Surg 1988;46: Svensson LG, Crawford ES, Hess KR, Coselli JS, Safi HJ. Experience with 1509 patients undergoing thoracoabdominal aortic operations. J Vasc Surg 1993;17: 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: Coselli JS, LeMaire SA. Left heart bypass reduces paraplegia rates following thoracoabdominal aortic aneurysm repair. Ann Thorac Surg 1999;67: Coselli JS, LeMaire SA, Poli de Figueiredo L, Kirby RP. Paraplegia after thoracoabdominal aortic aneurysm repair: is dissection a risk factor? Ann Thorac Surg 1997;63: Coselli JS, Poli de Figueiredo LF. Natural history of descending and thoracoabdominal aortic aneurysms. J Card Surg 1997;12: DISCUSSION DR RANDALL B. GRIEPP (New York, NY): I would like to thank the Society for the opportunity to discuss this paper, and Dr Coselli for sending me his manuscript. This is an important paper. It is by far the largest contemporary series of surgical treatment at thoracoabdominal aneurysms. The careful tabulation of the results by Coselli and associates, plus modern statistical techniques, allow one, as they have pointed out, to now calculate risk factors for individual patients facing operation. This is our equation for risk of rupture of thoracic and thoracoabdominal aneurysms that Dr. Coselli has alluded to, and this is the way we utilize this equation. It is very easy to put it into a PC so you can just plug in patient data and read out a risk of rupture. Dr Coselli s formulae, which are contained in the body of the manuscript, allow one to do exactly the same thing for operative risk. Thus, one can calculate rupture risk and operative risk for an individual patient. It should be pointed out, however, though that those are Coselli and associates results. It is probably not quite fair for all of us to tell our patients that we may accomplish the same thing. However, with a group analysis, one can enter one s own patients in, see how one s results compare with Coselli and associates, and then apply a correction factor to arrive at the most honest estimate possible of the operative risk that your patient faces. I would like to suggest one additional factor, however, and that is that we begin thinking a little bit about risk from the patient s standpoint. When I sit down with a patient to discuss risk of operation, I find that he or she is usually not too interested in all of the horrible things that can happen, but rather the probability of a good outcome. That is: what is the chance of walking out of here in as good a shape as I came in? I would encourage Coselli and associates to utilize the data to give us those sorts of numbers for risk from the patient s standpoint. The fields of paraplegia, death, and, to some extent, renal failure, overlap. It is of little interest to a patient if he dies in 5 days whether he has been paraplegic or anuric during that time. He is interested in the chances of getting out this alive and in good shape. I would also like to take this opportunity, for my own benefit and perhaps for some other members of the Society, to ask Dr Coselli a few questions regarding the techniques that led to these wonderful results. When is hypothermic circulatory arrest appropriate in the resection of these aneurysms? If feasible, is distal perfusion always advisable, or are there instances in which it can be dispensed with? When is visceral perfusion important in these resections? And finally, I would like to bring up the question of paraplegia. Coselli and associates rate here for the type II aneurysms is 8.2%; our reported rate in a group of aneurysms with similar extent is 10%. We almost never put in intercostal arteries; they almost always do. What insights do they have to further reduce this 8%to 10% paraplegia rate. I enjoyed this paper very much, and I salute Coselli and associates for the unique accomplishment that this series represents. DR COSELLI: I would like to thank Dr Griepp for his very kind comments. We agree entirely that for an individual surgeon, it is important to evaluate and analyze your own data and results. Furthermore, there is the potential for the adjustment of risk coefficients for an individual practice or institution. I also agree that it means very little to an individual patient, who dies 5 days postoperatively, whether or not that patient was paraplegic or anuric. However, for the analysis here, in the derivation of our formulas, we included all incidence of paraplegia including those with early mortality. This is justified, I believe, by virtue of providing a more accurate estimation. If, for example, we had excluded paraplegia in patients with early mortality, it would underestimate the overall incidence of paraplegia and possibly downplay operative risk. We use hypothermic circulatory arrest in those patients in whom proximal cross-clamping cannot be achieved safely for anatomic reasons. We have not used circulatory arrest specifically for purposes of visceral and spinal cord protection. I use distal aortic perfusion almost exclusively in patients with extent I and II aneurysms. I have not used left heart bypass or distal aortic perfusion routinely in patients of lesser extent, ie, extent III and IV. In patients with extent I or II aneurysms in whom left heart bypass is employed and the origins of the visceral and renal vessels are exposed, we have supplemented the technique with direct visceral perfusion using catheters placed within the origin of these branch vessels. There has been an overall general trend towards a reduction in the incidence of postoperative paraplegia after thoracoabdominal aortic aneurysm surgery across the board, including Dr Griepp s group at Mt. Sinai, our group, and others. It would be overly simplistic to represent this overall reduction in operative risk, with regards to neurological deficits, entirely to a single aspect such as the reattachment of intercostal arteries. Along with this overall trend, there has been one of increasing expertise and experience by a large number of surgeons. Consequently, any single specific aspect of technique and surgical approach assumes lesser importance. It has been our approach to place emphasis on the importance of reattachment of intercostal arteries. In a given patient, it is difficult to determine which intercostal arteries are the ones that are most important; consequently, I have been rather aggressive in reattaching as many as possible in the region from T7 to L1. Again, I would like to thank the Society for the opportunity to present this material.