Indications for operating on acute dissection of the

Transcription

1 Outcomes and Survival in Surgical Treatment of Descending Thoracic Aorta With Acute Dissection John Bozinovski, MD, and Joseph S. Coselli, MD Cardiovascular Surgery Service, The Texas Heart Institute at St. Luke s Episcopal Hospital, Division of Cardiothoracic Surgery, Michael E. DeBakey Department of Surgery, Baylor College of Medicine, Houston, Texas Background. Thoracic aortic replacement for acute De- Bakey type III aortic dissection is associated with significant morbidity and mortality. We report the outcomes of 76 consecutive patients who underwent surgical repair of the descending thoracic aorta or the thoracoabdominal aorta for acute dissection. Methods. During a 16-year period (1989 to 2004), we identified 76 patients who underwent surgery for acute type III aortic dissection. The average patient age was years (range, 36 to 84), and 55 patients (72.4%) were male. Surgical adjuncts included hypothermic circulatory arrest (8 patients), left heart bypass (15 patients), and cerebrospinal fluid drainage (5 patients). The mean aortic clamp time was minutes. Rupture was present in 17 patients (22.4%). Results. There was 1 intraoperative death. Operative mortality was 22.4% (17 patients), including 11 patients (14.5%) who died within 30 days of operation. Five patients (6.6%) had paraplegia, and 15 patients (19.7%) required hemodialysis, 7 temporarily. Cardiac complications occurred in 33 patients (43.4%), 2 patients (2.6%) were returned to the operating room for bleeding, and 10 patients (13.6%) required tracheostomy. The mean hospital stay was days. Rupture was not associated with increased risk of postoperative complications or operative mortality. Conclusions. In selected patients with emergent indications, operative intervention with open replacement of the descending thoracic aorta or thoracoabdominal aorta for acute dissection repair can be carried out with respectable mortality, morbidity, and survival rates. (Ann Thorac Surg 2008;85:965 71) 2008 by The Society of Thoracic Surgeons Indications for operating on acute dissection of the ascending aorta (acute DeBakey types I and II dissection) are straightforward. Without operation, acute dissection in this region of the aorta carries a high mortality rate [1], and surgery is mandated in nearly all such cases. Knowing when and whether to operate on an acute DeBakey type III dissection, which involves the descending aorta, is more difficult as the mortality associated with acute type III dissection treated nonoperatively [2 10] is considerably less than that of acute types I and II. As such, nonoperative management plays a greater role in the treatment of acute type III dissections. Twenty to forty years ago, in-hospital mortality rates for patients with acute type III aortic dissections receiving either nonoperative or operative treatment were high, ranging from 18% to 32% [2 5] and 24% to 49% [6 9] of patients respectively. Thus, when reviewing survival outcomes for patients with acute type III aortic dissections, no difference was shown between surgical or medical treatment over a 10-year period [10]. However, recent reports now show improved outcomes for medically managed, nonoperative acute type III dissection Accepted for publication Nov 2, Presented at the Fifty-third Annual Meeting of the Southern Thoracic Surgical Association, Tucson, AZ, Nov 8 11, Address correspondence to Dr Coselli, Baylor College of Medicine, One Baylor Plaza, BCM 390, Houston, TX 77030; jcoselli@bcm.edu. patients with mortality rates between 6.5% and 17.6% [11 15]. Surgical treatment of acute type III aortic dissection is usually reserved for patients who have symptoms of malperfusion, significant aortic dilation, refractory hypertension, refractory pain, or rupture, and are therefore generally not expected to fare as well as their medically managed cohort. We report the outcomes of 76 consecutive patients who underwent surgical repair of the descending thoracic aorta (DTA) or thoracoabdominal aorta (TAA) for acute DeBakey type III dissection. Patients and Methods Institutional Review Board protocol H for the collection of data for this study was approved and in place at the time of this study. Patients During a 16-year period (1989 to 2004), we identified 76 patients who underwent surgery for acute type III aortic dissection. The average patient age was years (range, 36 to 84) and 55 patients (72.4%) were male. Preoperative demographic data are shown in Table 1. Most of the patients had a history of hypertension (n 56; 73.7%) or smoked (n 53; 69.7%) whereas a sizeable minority had a history of coronary artery disease (n 17; 22.4%) or chronic obstructive pulmonary disease (n 20; 26.3%). A large number of patients had coexisting de by The Society of Thoracic Surgeons /08/$34.00 Published by Elsevier Inc doi: /j.athoracsur

2 966 BOZINOVSKI AND COSELLI Ann Thorac Surg ACUTE DISTAL AORTIC DISSECTION OUTCOMES 2008;85: Table 1. Preoperative Demographics of 76 Patients Operated on for Acute DeBakey Type III Aortic Dissection n % Smoking Hypercholesterolemia Dyslipidemia Coronary artery disease Congestive heart failure 0 0 Hypertension Diabetes mellitus type I 0 0 Diabetes mellitus type II Renal insufficiency Renal failure Chronic obstructive pulmonary disease Cerebrovascular accident scending thoracic, thoracoabdominal, or abdominal aortic aneurysms superimposed on the acute dissection (n 73; 96%). The chief complaint was most commonly chest pain, in 52.6% of patients. Operations were performed emergently, within 24 hours of presentation to our center, in 65 patients (85.6%) and urgently in the remaining 11 patients (14.4%). Seventeen patients (22.4%) experienced rupture before operation, including 1 case that occurred intraoperatively. The authors have in the past consistently managed DeBakey Type III (Stanford B) aortic dissections nonoperatively with medical management as the primary therapeutic option. Operation was offered to patients identified as having rupture or threatened rupture. The latter includes a total aortic diameter of 5 cm or greater in a previously normal aorta, hemorrhagic plural effusion that could not be separated from frank rupture, acute dissection superimposed upon a previously aneurysmal (6 cm) descending thoracic aorta, and unrelenting pain despite adequate blood pressure management. Only patients with aortic resection and replacement were included in this review. Patients who had preexisting thoracoabdominal ectasia or aneurysm, who then suffered superimposed acute dissection with the abdominal segment 4.5 cm or greater, underwent thoracoabdominal versus descending thoracic replacement providing they were hemodynamically stable. Operative Technique Details of our surgical technique have been described elsewhere [16, 17]. Replacement of the aorta was limited to the descending thoracic aorta in 57 patients (75%), and extended to include the thoracoabdominal aorta in 19 patients (25%). No patient had isolated replacement of the abdominal aorta for treatment of the acute dissection. All patients underwent left posterolateral thoracotomy, and often with extension of the incision across the costal margin. The majority of operations (69.7%) were performed without cardiopulmonary bypass, using the clamp-and-sew technique. Left heart bypass (LHB) was used in 15 patients (19.7%). For most patients, moderate systemic heparinization (1 mg/kg) with protamine reversal after completion of the final anastomosis was used. When used, LHB employed a centrifugal pump that drained the left atrium, through the left inferior pulmonary vein, and returned blood to the distal descending thoracic aorta or one of the femoral arteries. No oxygenator, cardiotomy reservoir, or heat exchanger was used for left heart bypass cases. Cerebrospinal fluid drainage through an intrathecal catheter was used in 5 patients (6.6%). Hypothermic circulatory arrest (HCA) was used in 8 patients (10.5%), in which case full heparinization therapy (4 mg/kg) and protamine reversal were utilized. In HCA cases, we did not use cerebrospinal fluid drainage. Hypothermic circulatory arrest employed a centrifugal pump that drained both the right atrium, through a long venous catheter inserted in the femoral vein with the tip lying in the right atrium, and the left atrium, through the left inferior pulmonary vein. Once the proximal anastomosis was completed, blood was returned to the patient through either the distal descending thoracic aorta or a femoral artery, and a side branch in the graft. The HCA circuit included an oxygenator, heat exchanger, and cardiotomy reservoir. Patients were actively cooled to electroencephalographic silence (which usually corresponds to a temperature between 16 C and 18 C), given pentothal intravenously, and 5 minutes later were partially exsanguinated before discontinuation of cardiopulmonary bypass. At this point, the head was packed in ice. After clamping and opening the aorta, a decision regarding management of intercostal arteries was made. Larger intercostal arteries that were not aggressively back-bleeding were reattached to the graft whenever possible. Smaller, aggressively bleeding intercostal arteries were sequentially oversewn. All reattached intercostal arteries were done with a lateral patch of aorta. In patients with prior abdominal aortic replacement, an aggressive approach toward intercostal artery reattachment was taken, which frequently necessitated endarectomy of atherosclerotic intima. Proximal and distal anastomoses were performed using 3-0 or 4-0 polypropylene suture as a running stitch and often reinforced using interrupted 3-0 or 4-0 polypropylene suture. When the repair did not completely remove the entire length of dissected aorta, the septal flap was tacked to the distal aortic wall to reperfuse the true lumen. The mean total aortic clamp time for the procedures was minutes and resulted in minutes of intercostal ischemia and minutes of visceral ischemia. When HCA was used, the mean total aortic clamp time was minutes, and overall circulatory arrest was minutes. Circulatory arrest lasted minutes. As expected in these procedures, bleeding after completion of the repair was difficult to control. Blood products per patient included a mean of units of packed red blood cells, units of fresh frozen plasma, units of platelets, and units of cryoprecipitate. Surviving patients, who were identified as hoarse after postoperative physical examination, were additionally

3 Ann Thorac Surg BOZINOVSKI AND COSELLI 2008;85: ACUTE DISTAL AORTIC DISSECTION OUTCOMES Table 2. Operative Mortality for 76 Patients Operated on for Acute DeBakey Type III Aortic Dissection n % Intraoperative mortality Thirty-day mortality In-hospital mortality Total operative mortality evaluated by an otolaryngologist to confirm vocal cord paralysis by direct examination. Study Variables and Definitions Data were retrieved from a prospectively maintained database. Acute aortic dissection was defined as that occurring within 14 days of onset of chest, back, or abdominal pain. Operative mortality included all deaths occurring within 30 days and all deaths occurring during the initial hospitalization. Stroke was defined as any new clinically or radiographically evident brain injury present after operation, including focal and global deficits, as well as transient and permanent deficits. Paraplegia was defined as a deficit of the lower extremities. This included weakness (paraparesis) or complete loss of motor function (paralysis). Renal failure was defined as serum creatinine exceeding 3.0 mg/dl or the need for hemodialysis. Temporary renal failure was defined as renal failure, as described above, that resolved by the time of discharge from hospital. Cardiac complications included myocardial infarction, atrial or ventricular dysrrhythmia requiring treatment, pericardial effusion requiring drainage, or cardiac failure requiring inotropic support longer than 48 hours, and the use of intra-aortic balloon counterpulsation or mechanical ventricular support. Statistical Analysis Data were analyzed using the SPSS statistical software program (SPSS, Chicago, Illinois). Continuous variables are presented as mean 1 SD. Categorical variables were compared using Fisher s exact tests. Table 3. Postoperative Complications in 76 Patients Operated on for Acute DeBakey Type III Aortic Dissection n % Table 4. Comparison of Rupture and Nonrupture With Use of Intraoperative Adjuncts, Postoperative Complications, and Operative Mortality Results Rupture n 17 Nonrupture n p Value Cerebrospinal fluid drainage 0 (0) 5 (8.5) catheter used [n (%)] Left heart bypass used 3 (17.6) 12 (20.3) [n (%)] Hypothermic circulatory 0 (0) 8 (13.6) arrest used [n (%)] Reoperation for bleeding 1 (5.9) 1 (1.7) Paraplegia 1 (5.9) 4 (6.8) Renal failure 4 (23.5) 11 (18.6) Cardiac complications 6 (35.3) 19 (32.2) Vocal cord paralysis 4 (23.5) 26 (44.1) Tracheostomy 3 (17.6) 7 (11.9) Stroke 0 (0) 5 (8.5) Operative mortality 4 (23.5) 13 (22.0) One patient (1.3%) died intraoperatively, 11 patients (14.5%) died within 30 days, and another 6 died inhospital beyond 30 days, resulting in an operative mortality of 17 patients (22.4%; Table 2). Two of these deaths occurred at an outside hospital after patients were transferred and the cause of death is unknown. Eight of these deaths were secondary to cardiac arrest, 6 were a result of multiple system organ failure, and 1 patient died intraoperatively during a second operation within the same initial hospitalization to address an acute retrograde dissection. This patient developed symptoms of cardiac ischemia as his dissection retrogradely progressed into his left main coronary artery. Emergency surgery was required to replace his aortic valve, aortic root, and ascending aorta, and a saphenous vein graft was brought to the left anterior descending coronary artery. Unfortunately, he could not be weaned from the cardiopulmonary bypass circuit and expired. Postoperative complications are listed in Table 3. Two patients (2.6%) were reoperated on for postoperative bleeding. Five patients (6.6%) developed paraplegia, all which were permanent and immediately occurring postoperative deficits, and 5 patients (6.6%) suffered a stroke. There were 15 patients (19.7%) who developed renal failure postoperatively. Renal failure was temporary in 7 Reoperation for bleeding Paraparesis 0 0 Paraplegia Renal failure Temporary Permanent Cardiac complications Stroke Left vocal cord paralysis Tracheostomy Table 5. Comparison of Patients With Rupture and Nonrupture to Extent of Aortic Replacement Replacement of the descending thoracic aorta [n%)] Replacement of the thoracoabdominal aorta [n (%)] Rupture n 17 Nonrupture n 59 9 (53.0) 48 (81.4) p Value 8 (47.1) 11 (18.6) 0.026

4 968 BOZINOVSKI AND COSELLI Ann Thorac Surg ACUTE DISTAL AORTIC DISSECTION OUTCOMES 2008;85: Table 6. Operative Mortality Comparisons for Urgency of Operation, Extent of Aortic Replacement, Intraoperative Adjuncts, Postoperative Complications, and Rupture Factor Operative Mortality Ratio Percent Relative Risk p Value Urgent surgery 4/ Emergency surgery 13/ Descending thoracic aorta replacement 12/ Thoracoabdominal aorta replacement 5/ Cerebrospinal fluid drain used 1/ Cerebrospinal fluid drain not used 16/ Left heart bypass used 1/ Left heart bypass not used 16/ Hypothermic circulatory arrest used 3/ Hypothermic circulatory arrest not used 14/ Clamp-and-sew technique used 13/ Left heart bypass or hypothermic circulatory arrest used 4/ Reoperation for bleeding present 1/ Reoperation for bleeding absent 16/ Postoperative paraplegia present 1/ Postoperative paraplegia absent 16/ Postoperative renal failure present 5/ Postoperative renal failure absent 12/ Postoperative cardiac complication present 12/ Postoperative cardiac complication absent 4/ Postoperative vocal cord paralysis present 7/ Postoperative vocal cord paralysis absent 10/ Postoperative tracheostomy placed 4/ Postoperative tracheostomy not placed 13/ Postoperative stroke present 2/ Postoperative stroke absent 15/ Preoperative aortic rupture present 4/ Preoperative aortic rupture absent 13/ Only postoperative cardiac complication was a significant postoperative risk factor for operative mortality. patients (9.2%) and permanent in the other 8 patients (10.5%). Paralysis of the left vocal cord occurred in 30 patients (39.5%), and 10 patients (13.2%) required tracheostomy. Cardiac complications occurred in 33 patients (43.4%), most of which were atrial dysrhythmias. There were no preoperative patient characteristics significantly associated with risk of operative mortality. Age was the only preoperative patient characteristic shown to be significantly associated with the risk of rupture of the aorta. Patients with rupture were older than those without rupture ( years versus years; p 0.05). Looking at our subset of 17 patients with a preoperative aortic rupture, the impact of rupture on use of intraoperative adjuncts was assessed (Table 4). As expected by the expedience required, none of these 17 patients had an intrathecal catheter placed preoperatively; however, this was not a statistically significant difference from the 5 nonruptured patients who received intrathecal catheters. Left heart bypass was utilized in the repair of 3 of 17 patients (17.6%) with rupture compared with 12 of 59 patients (20.3%) without rupture. Hypothermic circulatory arrest was not used in any of the rupture patients compared with 8 of 59 (13.6%) of the nonrupture patients. Neither LHB nor HCA use differed significantly between the two groups. Rupture of the aorta was more likely to result in replacement of the thoracoabdominal aorta as opposed to replacement of the descending thoracic aorta alone (Table 5). Eight of 17 (47.0%) rupture patients had replacement of the thoracoabdominal aorta compared with 11 of the 59 nonrupture patients (18.6%). Rupture was not significantly associated with an increased risk of any postoperative complication analyzed in the present study. Interestingly, the operative mortality in the group experiencing rupture was 23.5% (4 of 17 patients), compared to 22.0% (13 of 59 patients) operative mortality in the nonrupture group. Patients who experienced operative mortality were no more likely to have required an emergency operation. Except for postoperative cardiac complications, none of the other postoperative complications was significantly associated with a risk of operative death (Table 6). Comment This study is a descriptive measure of our experience with the surgical treatment of acute type III aortic dissec-

5 Ann Thorac Surg BOZINOVSKI AND COSELLI 2008;85: ACUTE DISTAL AORTIC DISSECTION OUTCOMES tion. As no comparison was made to a cohort of patients managed nonsurgically, a statement regarding the better approach to treatment would not be supported in either direction by the present study. The decision to operate on these patients is based on mortality rates for operative and nonoperative treatment described in the literature, taking into account risk stratification based on presenting symptoms. Current guidelines for the operative treatment of acute type III aortic dissection include rupture of the aorta, organ or limb malperfusion, intractable pain, uncontrollable hypertension, and expansion of the aorta. Our approach to the treatment of acute type III aortic dissection follows these guidelines. Most of the patients in the study had several indications for operation. Sixteen patients had rupture of the aorta before arrival in the operating room, and nearly all (n 73; 96%) of the patients in this study had an associated aneurysm of the aorta. Typically, we decrease the diameter threshold for surgical intervention whenever aneurysm is accompanied by dissection. Pain, which is suggestive of progressive dissection, impending rupture, or ischemia, was also present in the vast majority of these patients (n 71; 93%). This study is limited by the small numbers of patients enrolled and therefore is insufficiently powered to identify subtle differences in outcomes between groups of patients. Additionally, we were unable to identify any risk factor for operative mortality in this set of patients presenting with acute type III aortic dissection. However, rupture occurred in a substantial number of the patients, illustrating some interesting trends. Patients with rupture were more likely to undergo repair of the thoracoabdominal aorta than repair of the descending thoracic aorta. The explanation for this trend is most likely the location of the rupture, which unfortunately was not captured in the database. However, we do know that ruptures tend to occur in regions where the aorta either branches or is anchored, such as the portion of the aorta crossing the diaphragm. As a result, the repair is dictated by the location of rupture, which would explain our finding of increased thoracoabdominal aortic replacements in patients with rupture. It would be reasonable to expect that patients with aortic rupture would have a worse outcome than the other patients; however, this was not the case. Patients with and without rupture had similar risks of postoperative complications and operative mortality. A clampand-sew technique was used in 82% of the rupture patients and was used in 66% of the nonrupture patients. Despite this, or perhaps because of this, mortality in the rupture group was no different than that of the nonrupture group and should continue to be included as an acceptable approach to these patients. For patients presenting with urgent or emergent surgical indications, we generally do not perform repairs that are more extensive than absolutely necessary. They are operated on expeditiously; there is less time to implement adjunctive measures that might otherwise be used in an elective procedure. If dissection extends into the aortic arch, hypothermic circulatory arrest is typically used. For non-hca cases, the decision whether or not to use LHB is made. We usually employ left heart bypass for repairs of acute type III dissection that involve a segment of the proximal descending thoracic aorta extending below the level of T6/7, but this is not an absolute guideline. Although the sample size was small in this study, left heart bypass did not confer a protective effect on the incidence of paraplegia or operative mortality. After reattaching important intercostal, lumbar, and visceral branch vessels, the aorta is opened until a normal caliber aorta is encountered distally, and an open aortic anastomosis is fashioned. Occasionally, the aorta is still dissected at this point, and the septal flap is tacked to the aortic outer wall before fashioning the distal anastomosis to restore true lumen flow. This is a departure from our approach in chronic dissections, where occasionally a wedge of the dissection flap is resected to enable flow into both true and false lumens. Acute dissection of the descending thoracic aorta remains a catastrophic cardiovascular condition. Newer diagnostic techniques and management strategies have demonstrated the potential to improve patient outcomes and reduce harmful consequences of major surgery. During the period of this review, endovascular stentgrafting covered by the descending thoracic aorta was available only to patients participating in clinical trials and rarely, if ever, available to those patients with dissection. Since the conclusion of this patient series in 2004, we have endorsed the use of stent-grafting for select acute and chronic aortic pathologies and continue to explore its specific use in the short and long term for acute DeBakey type III aortic dissections [18 22]. Many of our patients are referred from out of state and follow-up data is difficult to obtain. Generally, we recommend a computed tomography scan before discharge from hospital, at 3 months, 6 months, and 1 year, and then yearly thereafter for continued surveillance of the aorta. References Anagnostopoulos CE, Prabhakar MJ, Kittle CF. Aortic dissections and dissecting aneurysms. Am J Cardiol 1972;30: Doroghazi RM, Slater EE, DeSanctis RW, Buckley MJ, Austen WG, Rosenthal S. Long term survival of patients with treated aortic dissection. J Am Coll Cardiol 1984;3: Glower DD, Fann JI, Speier RH, et al. Comparison of medical and surgical therapy for uncomplicated descending aortic dissection. Circulation 1990;82(Suppl):IV Appelbaum A, Karp RB, Kirklin JW. Ascending versus descending aortic dissections. Ann Surg 1976;183: Pate JW, Richardson RL, Eastridge CE. Acute aortic dissection. Am Surg 1976;42: Crawford ES, Svensson LG, Coselli JS, Safi HJ, Hess KR. Aortic dissection and dissecting aortic aneurysms. Ann Surg 1988;208: Miller DC, Mitchell RS, Oyer PE, Stinson EB, Jamieson SW, Shumway NE. Independent determinants of operative mortality for patients with aortic dissections. Circulation 1984; 70:I Miller DC, Stinson EB, Oyer PE, et al. Operative treatment of aortic dissections. Experience with 125 patients over a sixteen-year period. J Thorac Cardiovasc Surg 1979;78:

6 970 BOZINOVSKI AND COSELLI Ann Thorac Surg ACUTE DISTAL AORTIC DISSECTION OUTCOMES 2008;85: Reul GJ, Cooley DA, Hallman GL, Reddy SB, Kyger ER III, Wukasch DC. Dissecting aneurysm of the descending aorta. Improved surgical results in 91 patients. Arch Surg 1975;110: Glower DD, Speier RH, White WD, Smith LR, Rankin JS, Wolfe WG. Management and long-term outcome of aortic dissection. Ann Surg 1991;214: Gysi J, Schaffner T, Mohacsi P, Aeschbacher B, Althaus U, Carrel T. Early and late outcome of operated and nonoperated acute dissection of the descending thoracic aorta. Eur J Cardiothor Surg 1997;11: Rosenborough G, Burke J, Sperry J, Perler B, Parra J, Williams GM. Twenty-year experience with acute distal thoracic aortic dissections. J Vasc Surg 2004;40: Estrera AL, Miller CC, Safi HJ, et al. Outcomes of medical management of acute type B aortic dissection. Circulation 2006;114(Suppl 1): Elefteriades JA, Lovoulos C, Coady MA, Tellides G, Kopf GS, Rizzo JA. Management of descending aortic dissection. Ann Thorac Surg 1999;67: Masuda Y, Yamada Z, Morooka N, Watanabe S, Inagaki Y. Prognosis of patients with medically treated aortic dissection. Circulation 1991;84(Suppl):III Coselli JS, Plestis KA, LaFrancesca S, Cohen S. Results of contemporary surgical treatment of descending thoracic aortic aneurysms: experience with 198 patients. Ann Vasc Surg 1996;10: Coselli JS, LeMaire SA. Descending and thoracoabdominal aneurysm. In: Cohn LH, ed. Cardiac surgery in the adult. 3rd ed. New York: McGraw-Hill Companies; 2008: Trimarchi S, Nienaber CA, Rampoldi V, et al, on behalf of the IRAD Investigators. Role and results of surgery in acute type B aortic dissection: insights from the International Registry of Acute Aortic Dissection. Circulation 2006; 114(Suppl 1):I Suzuki T, Mehta RH, Ince H, et al. Clinical profiles and outcomes of acute type B aortic dissection in the current era: lessons from the International Registry of Aortic Dissection (IRAD). Circulation 2003;108(Suppl 1):II Duebener LF, Lorenzen P, Richardt G, et al. Emergency endovascular stent-grafting for life-threatening acute type B dissections. Ann Thorac Surg 2004;78: Lin PH, El Sayad HF, Kougias P, Zhou W, LeMaire SA, Coselli JS. Endovascular repair of thoracic aortic disease: overview of current devices and clinical results. Vascular 2007;15: Bakaeen FG, Coselli JS, LeMaire SA, Huh J. Continued aortic aneurysmal expansion after thoracic endovascular stent grafting. Ann Thorac Surg 2007;84: DISCUSSION DR CURT TRIBBLE (Gainesville, FL): Doctor Coselli and his colleague had asked Dr Tom Martin to comment on this paper. I suspect he thought Tom would prove to be a kindred spirit, and I bet that is true, given that they were both protégés of Dr Crawford. Unfortunately, Tom could not be here and he asked me to stand in for him, a tough job in my opinion. I myself have a slightly different mindset about the management of type B dissections developed during more than 25 years of dealing with such cases at the University of Virginia, where we were more conservative about these cases than you all seem to have been. Tom and my new Florida colleagues have a practice similar to yours. They did 500 aortic cases at Gainesville last year, encompassing all aspects of aortic surgery, including dissections such as those you report on here. I know we are short on time, and I have three questions I d like to ask you, but I will do them one at a time in case we don t have time to deal with them all. You assert that we can operate on type B dissections with a reasonable margin of safety. I think you ll have to admit, when you add up the numbers in your charts in your manuscript, that fully two thirds of your patients had major complications when operated on. So these data do force the question: obviously you can operate on them, but should you? At UVA, we had a much more conservative approach to these cases and managed them medically in most cases, and we certainly didn t have anywhere near a 25% mortality with medical management of these patients. They were all admitted to our cardiothoracic service. We managed them all initially in our intensive care unit. In fact, to tell you the truth, I don t recall a patient dying with medical management. Obviously, in cases when the aorta has leaked, you have to address that surgically. So the question I have for you and I could not discern this from looking over the paper, is who do you turn down? Are some of the patients who are referred to you managed medically? To be even more specific, one of the indications in your manuscript for operating on patients was aortic size, but you didn t mention what size aorta you would designate as requiring an operation in the setting of an acute type B dissection. DR BOZINOVSKI: To begin with, the vast majority of our patients are going to be treated medically when they come to us with an acute type III aortic dissection. We still follow the same principles. They need to have an indication for surgery, and those indications are evidence of rupture, evidence of malperfusion of the end organs, inability to control their pain or blood pressure with medical management, or if their aorta is at a certain size. Each center has to come to some sort of realization of what their experience with this is and what the absolute numbers should be. For dissection patients, we generally lower the threshold diameter about half a centimeter versus our nondissection patients. In other words, if there is no dissection there, then we operate on distal aortas that are dilated to about 6 to 6-1/2 cm, somewhere in that ballpark. If they have a dissection, you probably want to get to them a little sooner. Also, if they are younger, you want to get to them sooner. The patients who we operated on had obvious indications for surgery. One quarter of them had rupture. All of them had chest pain or abdominal or flank pain. They had ongoing pain that suggested their dissection was either progressing or there was imminent rupture. So these patients are going to do worse. DR TRIBBLE: Forgive me for asking you to be more specific. I just want you to tell us one number, a number. What size aorta mandates an operation in your practice when you are caring for a patient with an acute type B dissection? In your manuscript, you said that size is one of the indications even if none of the other indications were present, but you specify a size that mandates an operation. What size would that be? DR BOZINOVSKI: If the aorta is 5-1/2 cm and you have an acute dissection and you have chest pain, go ahead and start making arrangements for surgery. They don t necessarily have

7 Ann Thorac Surg BOZINOVSKI AND COSELLI 2008;85: ACUTE DISTAL AORTIC DISSECTION OUTCOMES to go for surgery right then and there if you can manage their pain, but at some point they should come for surgery. DR TRIBBLE: But would you operate on somebody with a 5-1/2 cm aorta with acute type III dissection with none of the other indications? That is the question. DR BOZINOVSKI: Sorry, say that again. DR TRIBBLE: If you have an aorta that is 5-1/2 cm and you have a type III dissection, is that alone an indication? In your paper, you suggest that it is. DR BOZINOVSKI: It is not an indication for emergent surgery. These patients should be followed and you should give them consideration. That is something you have to sit down with the patient and talk about. You have to discuss what is the chance of rupture, what they would like to have done. DR TRIBBLE: Okay, fair enough. In your manuscript, you mentioned all sorts of things about the operative adjuncts, cerebrospinal fluid (CSF) drains, clamp-and-sew techniques, et cetera, but I ll bet a lot of people in the audience would enjoy knowing some specifics about how you handle these friable acutely dissected aortas. After all, the whole issue with this entity is that it is not the same as operating on a nice, leathery, mature aneurysm or a healed dissection. What size suture do you use? Do you use felt? Do you use glue? Maybe just a few comments about the technical side of this that would be helpful to us. DR BOZINOVSKI: On the proximal suture line, we generally use 4-0 polypropylene as a running stitch. We will reinforce the entire anastomosis with interrupted 4-0 polypropylene with pledgeted sutures. And then we completely remove the dissection flap. You need to do that because often your dissection flap is going to stop at your intercostals, and if you don t remove it, you might miss some of these intercostals vessels and then you are going to end up with a problem with reoperation for bleeding. So you want to completely remove that dissection flap, take a look at what is underneath, and make sure that you oversew or reattach whatever is necessary. And then you just 971 progress down. If you have aggressively back bleeding smaller vessels, then we don t reattach those, but if there are large vessels that aren t aggressively back bleeding, then we will reattach those, and just keep going down until you get to the viscerals, or if you don t need to go that far distally, we just fashion our distal anastomosis. At the distal end, if there is a dissection and the aorta is not aneurysmal, then we will tack the layers together and fashion our anastomosis there. That differs from our chronic dissections where you can sometimes take a wedge out of the tissue and thereby perfuse both lumens. You have a nice thick septum in that scenario. In acute dissections, it is very thin, friable, so we try to tack it together. We don t generally use glue on our type III dissections. We are not against using glue in that situation. And then again, it is 4-0 polypropylene on the distals. DR MARC MOON: (St. Louis, MO): I know from reviewing your elective series that you use CSF drains in almost every single case. So how come in this series of probably patients who were at higher risk you didn t use CSF drains except in a very minority of patients? DR BOZINOVSKI: If you have a patient with an acute type III dissection, with an indication for surgery as described earlier, you may not want to delay timely surgery to obtain a CSF drain. In emergent or urgent cases, we prefer to expedite surgical intervention. Additionally, anesthesiologists are going to have different experiences with putting in a CSF drain. DR MOON: I understand that with rupture, but the other ones presumably weren t ruptured and emergently rushed to the operating room. DR BOZINOVSKI: No, but all these patients were operated on either urgently or emergently, and we simply don t like to delay surgical intervention: in the case of our rupture patients, in none of them, and in the case of our nonrupture patients, I think 5 of the 59 had a CSF drain placed. Also, it depends on what portion of aorta needs to be replaced. Generally, if the proximal descending thoracic aorta is to be clamped, then we may institute a CSF drain. However, we are not going to take the time away from getting the repair under way to place a CSF drain.