- Email: [email protected]
Repair of Thoracoabdominal Aortic Aneurysms in Octogenarians
Repair of Thoracoabdominal Aortic Aneurysms in Octogenarians Leonard N. Girardi, MD, and Joseph S. Coselli, MD Department of Surgery, Baylor College of Medicine, Houston, Texas
Background. Reports on octogenarians undergoing coronary revascularization, valve replacement, and abdominal aneurysmorrhaphy demonstrate little increase in operative risk during elective procedures. However, the mortality in this group of patients increases rapidly when urgent or emergent procedures are performed. We analyzed the outcome of patients in their ninth decade of life undergoing repair of thoracoabdominal aortic aneurysms. Methods. A retrospective review of 39 consecutive octogenarians undergoing repair of thoracoabdominal aortic aneurysms. Results. Thirty-nine of the past 900 patients with thoracoabdominal aortic aneurysms (5.2%) repaired by us were between the ages of 80 and 89 years. The median age was 84 years with a male-to-female ratio of 1:3. Two of 39 patients (5%) had acute type III dissections, and the remainder had chronic aneurysms. Twelve patients had Crawford extent I aneurysms, whereas 7, 10, and 10 patients were extent II, III, and IV, respectively. The
overall in-hospital mortality was 10.3% (4 of 39 patients). Major postoperative complications included paraperesis/ paraplegia, 5% (n 5 2); renal failure, 18% (n 5 7) including hemodialysis in 3 patients; stroke, 5% (n 5 2); myocardial infarction or arrhythmia, 18% (n 5 7); and respiratory insufficiency, 36% (n 5 14) including 4 patients requiring tracheostomy. A univariate analysis of perioperative risk factors was performed using the Fisher’s exact test. The need for hemodialysis (p 5 0.035), a tracheostomy (p 5 0.0001), or a perioperative myocardial infarction (p < 0.001) significantly increased the risk of death. Conclusions. Repair of thoracoabdominal aortic aneurysms in octogenarians can be performed with acceptable morbidity and mortality. However, survival decreases dramatically with even single system organ failure. An extended period of recovery is usually required in these elderly, high-risk patients. (Ann Thorac Surg 1998;65:491–5) © 1998 by The Society of Thoracic Surgeons
R
elective and emergent repair are viable options in these high-risk patients.
epair of thoracoabdominal aortic aneurysms (TAAA) remains a formidable operation with substantial morbidity and mortality. Advances in surgical technique and perioperative care have reduced the mortality in most series to approximately 10%, and major perioperative morbidity such as paraplegia and renal failure continues to decrease in incidence [1–3]. The mean age at operation of patients undergoing repair of TAAA is 65 to 70 years in most series. However, the average life expectancy for a United States citizen is more than 76 years, with more than 10 million people older than 80 years. In addition, the average octogenarian can expect to live 6 to 10 years beyond his or her 80th birthday [4]. Major elective cardiovascular reconstruction in this population has not proven to be significantly more risky than in patients less than 80 years of age. Coronary artery bypass, valve replacement, and abdominal aortic aneurysm repair have been performed with acceptable perioperative risk. It is only when these elderly patients are subjected to emergent intervention that the mortality increases substantially [5–9]. We analyzed our experience with TAAA repair in octogenarians to determine whether Accepted for publication Aug 22, 1997. Address reprint requests to Dr Coselli, 6560 Fannin, Suite 1144, Houston, TX 77030 (e-mail: [email protected]).
© 1998 by The Society of Thoracic Surgeons Published by Elsevier Science Inc
Material and Methods Nine hundred consecutive patients underwent operation on the thoracoabdominal aorta from March 1989 through March 1997. Thirty-nine (4.3%) were 80 years of age or older. A retrospective analysis of these 39 patients was performed. Complete follow-up was available in all but 2 patients. Perioperative risk factors were subjected to univariate analysis using a Fisher’s exact test. Survival was determined using the method of Kaplan and Meier [10]. All statistics were performed using SAS for Windows statistical software program (SAS Institute Inc, Cary, NC). Mortality was defined as death within 30 days of the procedure or before discharge from the hospital. Paraplegia and paraperesis were defined as any lower extremity neuromuscular deficit, either immediate or delayed, that occurred in the hospital. Renal failure was defined as the need for hemodialysis or creatinine twice baseline level. Respiratory failure was defined as prolonged intubation (.48 hours), pneumonia, adult respiratory distress syndrome, or the need for a tracheostomy. Cardiac complications included arrhythmias, myocardial infarction, or congestive heart failure. 0003-4975/98/$19.00 PII S0003-4975(97)01300-3
492
GIRARDI AND COSELLI THORACOABDOMINAL AORTIC ANEURYSMS
Ann Thorac Surg 1998;65:491–5
Table 1. Complications and Mortality in Octogenarians Undergoing Thoracoabdominal Aortic Aneurysm Repair Complication Pulmonary Cardiac Renal Left recurrent nerve palsy Paraperesis/paraplegia Cerebrovascular accident Reexploration for bleeding Mortality
Fig 1. Number of patients in each category of aneurysm extent according to the Crawford classification system for thoracoabdominal aortic aneurysms [11].
Results Thirty-nine octogenarians underwent repair of TAAA. Ages ranged from 80 to 89 years with a median age of 84 years. There were 26 women and 13 men. Preoperative risk factors are as follows: hypertension, 24 (61%); chronic obstructive pulmonary disease, 14 (36%); coronary artery disease, 13 (33%); renal occlusive disease, 12 (31%); aneurysm rupture, 7 (18%); renal insufficiency (creatinine level .2.5g/dL), 6 (15%); diabetes mellitus, 3 (8%); cerebrovascular accident, 2 (5%); and acute type III dissection, 2 (5%). Of note, 7 patients (18%) presented with either a contained or free rupture of the aneurysm. The distribution of aneurysms according to the classification system of Crawford and colleagues [11] is seen in Figure 1. The technique for operative repair of TAAAs has been described in detail elsewhere [2]. Briefly, heparin (1 mg/kg) was administered to all patients before aortic cross-clamping. All large, patent intercostal arteries in the region between T8 and L1 were reimplanted into a Hemashield Dacron prosthetic (Meadox Medical, Oakland, NJ). A total of 19 patients had 1 to 5 pairs of intercostal arteries included in their repair. Seven of 19 patients believed to be at high risk for paraplegia (extents I and II) were chosen for atriofemoral or atriodistal aortic bypass using a Biomedicus pump (Minneapolis, MN). The 12 patients excluded from the use of partial bypass included patients with large friable intraluminal atheroma or thrombus, and those with extremely large aneurysms (9 to 13 cm) that would make distal cross-clamping unsafe. Complications and mortality are summarized in Table 1. Perioperative in-hospital mortality occurred in 4 patients (10.3%). Two patients had postoperative myocardial infarctions that progressed to multiple organ failure
No. of Patients (%) 14 (36%) 7 (18%) 7 (18%) 8 (21%) 2 (5.1%) 2 (5.1%) 1 (2.6%) 4 (10.3%)
and death. Another patient died of adult respiratory distress syndrome of unknown cause. The final death occurred on postoperative day 29 after an intraoperative cerebrovascular accident. All patients with ruptured aneurysms survived to discharge. There were no intraoperative deaths. There was a total of four major neurologic complications. Two patients (5.1%) had neurologic deficits secondary to spinal cord ischemia. One patient developed paraperesis and one developed delayed paraplegia on postoperative day 9 after TAAA repair in the setting of an acute type III dissection. The paraplegia in the latter patient was noted after hemodialysis during which the patient had multiple hypotensive episodes. Despite the institution of cerebrospinal fluid drainage, pressors, and steroids, the neurologic injury was permanent. Both of these patients had extent III aneurysms that were repaired without the support of atriofemoral bypass. The presence of an acute dissection was predictive of a postoperative neurologic event (p , 0.05) but not death. Extent III aneurysms were similarly statistically predictive of a postoperative neurologic deficit (p , 0.038). Two patients (5.1%) had perioperative cerebrovascular accidents. One patient never awakened postoperatively and subsequently died. The second developed a hemiparesis on postoperative day 7. No source for her stroke was found. Fortunately, she recovered and was ambulating with assistance before discharge. A total of 7 patients experienced postoperative renal failure (18%). Four patients had their renal function return to baseline level and never required hemodialysis. Three patients (7.7%) with aneurysms of different extent (II, III, IV) required hemodialysis; subsequently 2 died in the hospital. The development of postoperative renal failure could not be predicted based on a patient’s preoperative renal function (p . 0.49) or the presence of renal arterial occlusive disease (p . 0.15). Neither atriodistal bypass nor cold renal perfusion were protective against the development of postoperative renal failure (p . 0.05). In addition, cross-clamp time did not influence postoperative renal function (p . 0.05). However, the need for hemodialysis was predictive of postoperative mortality (p , 0.03). Postoperative pulmonary complications were present in 14 patients (36%). A majority of these complications were prolonged intubations (10%, 4 patients) or reintu-
Ann Thorac Surg 1998;65:491–5
Fig 2. Survival of 39 octogenarians undergoing repair of thoracoabdominal aortic aneurysms. The median survival is 31.5 months with an actuarial 5-year survival of 50%.
bations (15%, 6 patients). Four patients eventually required tracheostomies. All four died in the perioperative period. There were seven cardiac complications in the perioperative period. Five patients required medical management of supraventricular tachyarrhythmias and 2 (5.1%) patients experienced fatal transmural myocardial infarctions. Although the development of supraventricular tachyarrhythmias was not predictive of a poor outcome, a postoperative myocardial infarction was strongly predictive of mortality (p , 0.0013). One patient required reexploration for postoperative hemorrhage (2.6%). This patient had a splenic laceration and required splenectomy. Wound infections requiring local wound care or debridement developed in 2 patients and a left true vocal cord palsy, in 8 patients (21%). All of these patients had extent I or II aneurysms. All underwent thyroplasty before discharge from the hospital with good result. The univariate risk factors predictive for death are tracheostomy (p , 0.0001), myocardial infarction (p , 0.0013) and hemodialysis (p , 0.035) and for neurologic injury are acute type III dissection (p , 0.05) and extent III aneurysm (p , 0.038) in the perioperative period. The median length of stay for these 39 patients was 15 days with a range of 10 to 86 days. The median survival for all 39 octogenarians is 31.5 months with a 5-year actuarial survival of 50% (Fig 2).
Comment As the population continues to age, it is becoming increasingly common to perform major cardiovascular operations on those older than 80 years. Early series of valve replacement [12], coronary artery bypass [13], and abdominal aneurysmectomy [7] reported mortality rates 5 to 10 times higher than patients of lesser age undergoing similar procedures. Recently, however, the mortality for these elective procedures has been reduced to 5% to 10%. Specifically, elective abdominal aortic aneurysm repair has had a reproducible mortality of approximately 6% [8, 9, 14]. We sought to examine our results with repair of
GIRARDI AND COSELLI THORACOABDOMINAL AORTIC ANEURYSMS
493
TAAAs in octogenarians to determine whether repair is warranted in this high-risk population. Octogenarians with TAAAs should not be denied intervention because of excessive perioperative mortality. Our mortality of 10.3% is not significantly different from our previous report on 372 consecutive TAAA repairs in which the mean age was 66 years. Although there are no other series of TAAA repairs in this elderly patient population, comparison to similar patients undergoing repair of infrarenal abdominal aortic aneurysms (IAAA) is favorable. Dean and colleagues [8] reported an inhospital mortality of 38% in 34 consecutive octogenarians requiring IAAA repair. The mortality for elective cases was 6%, whereas the mortality rose to 40% for patients with symptomatic aneurysms and 91% in cases of ruptured IAAA. Results at the Cleveland Clinic were comparable with a slightly higher mortality for elective cases [9]. Examination of the causes of death in our series as well as these two reports is worthy of discussion. Two patients in our cohort died of a transmural myocardial infarction after TAAA repair, accounting for 50% of our mortality. Both of these patients were elective cases with no evidence of rupture or leak. Dean and associates [8] found that 23% of the deaths in octogenarians undergoing IAAA repair were secondary to acute myocardial infarctions. O’Hara and colleagues [9] encountered a similar problem with 31% of their patients’ deaths attributable to fatal myocardial ischemia. In addition, they reported that only 4% of their patients having previous myocardial revascularization had a fatal myocardial infarction, whereas 12% of those without revascularization succumbed to a fatal myocardial infarction. Certainly these are patients with a high incidence of coronary artery disease. Young and co-workers [15] reviewed the coronary angiograms in 302 patients undergoing IAAA repair. They noted that 46% of asymptomatic patients had at least single-vessel disease, with 20% having significant triple vessel disease. Patients with an anginal history had even higher rates of coronary artery disease requiring intervention. Orecchia and colleagues [16] confirmed these results in an additional 59 patients requiring aortic operation. Although no guidelines exist with respect to the proper cardiac evaluation in patients with TAAAs the high incidence of insidious coronary artery disease and myocardial infarction suggests a more exhaustive search for such disease is indicated in the elective surgical candidate. Obviously, patients with symptomatic or frankly ruptured aneurysms cannot undergo a thorough preoperative cardiac evaluation. One must also be cognizant that 7% to 10% of octogenarians requiring preoperative myocardial revascularization will not survive this intervention and an additional unknown percentage will rupture their aneurysm while undergoing these procedures. However, the risk of correcting significant coronary artery disease is clearly less than the uniformly fatal outcome when incurring a perioperative myocardial infarction. In addition, a survival benefit has been demonstrated for patients with known IAAAs undergoing preoperative coronary revascularization with angioplasty or bypass grafting when compared with a
494
GIRARDI AND COSELLI THORACOABDOMINAL AORTIC ANEURYSMS
nonrevascularized population (75% versus 30%, respectively) [9]. It is counterintuitive that we had no mortality for the 7 octogenarians presenting with ruptured TAAA. The mortality associated with this mode of presentation in IAAA is 33% to 91% [8, 9, 17, 18]. However, our patient population differs from that seen in these reports. All of our patients were hemodynamically stable at the time of presentation to our center. Dean and colleagues [8] reported that more than 50% of their patients with ruptured IAAA had unrecordable blood pressure before the application of a cross-clamp. The Mayo Clinic series reported 67% of their patients were hypotensive, whereas 24% had a cardiac arrest before operation [17]. Because a large percentage of our patients were transferred to our institution from distant geographic areas, those with hemodynamic instability or worse were selected out before our intervention. Some groups have advocated withholding operative intervention in octogenarians with ruptured aneurysms and hemodynamic compromise [8, 18]. Mortality rates are consistently greater than 90% with most patients dying of multiple organ failure after a prolonged intensive care unit stay. We have no data to refute this position and agree that resources may be better used in these desperate situations. The incidence of paraplegia and paraperesis in this group of patients is not significantly higher than in our previous report. A number of factors play a role in keeping this often fatal complication to a minimum. We continue to support an aggressive posture toward intercostal artery reimplantation, especially in patients at the highest risk for neurologic injury (extents I and II) [19]. All 19 patients in this group had at least one pair of intercostals reimplanted, either as a separate island of aorta or as part of the distal anastomosis. Extent III aneurysms were statistically more likely to be associated with postoperative neurologic injury in octogenarians. However, we believe the case of delayed paraplegia is more attributable to multiple hemodynamic insults associated with hemodialysis rather than aneurysm extent. If this case is eliminated from consideration, aneurysm extent no longer is a risk factor for postoperative neurologic deficits. Although often limited by large aneurysm size or mural atheromatous debris we prefer to use atriodistal bypass in this same high-risk group. The small sample size in this report precludes statistical support for its use, yet we believe that distal perfusion has a positive influence on neurologic outcome. It allows one to perform a technically sound proximal anastomosis without time constraint while maintaining intercostal and collateral flow to the cord. It also reduces total cord ischemic time in complex reconstructions in which multiple intercostal arteries are to be reimplanted along with visceral and renal artery reimplantation. This is especially true with aneurysms in the setting of an aortic dissection, a risk factor for paraplegia in this report and in others [20]. Kouchoukos and colleagues [21] support the use of profound hypothermia and circulatory arrest to limit spinal cord ischemia. Although we did not use this modality in
Ann Thorac Surg 1998;65:491–5
any of these 39 patients, we have had a 70% mortality with circulatory arrest for TAAA repair in patients older than 70 years. Although their neurologic function may be preserved, they are often too frail to withstand the systemic insult of this technique. We continue to reserve circulatory arrest for patients in whom the risk of stroke with proximal cross-clamping is extraordinarily high. Perhaps we could have avoided the single intraoperative cerebrovascular accident had we used profound hypothermia and circulatory arrest in that particular case. Acute renal failure after TAAA repair carries poor prognostic implications. A total of 18% of our patients experienced a reduction in renal function, with 7.7% requiring hemodialysis. In this small set of patients we were unable predict the development of renal failure based on preoperative renal function or renal occlusive disease. In addition, our small numbers do not statistically support the routine use of either atriodistal bypass or cold renal perfusion as adjuncts to preserving renal function. Nonetheless, we agree with Safi and colleagues [22] that such measures are protective against postoperative renal failure. In their series 18% of patients developed renal failure with 15% requiring hemodialysis. This complication carried a 50% mortality. However, the use of distal perfusion was protective against the development of acute renal failure. We also found hemodialysis to be highly predictive of postoperative mortality and support the use of distal perfusion or cold renal perfusion when technically feasible. Pulmonary complications were the most prevalent problem seen in octogenarians undergoing TAAA repair. The high incidence of chronic obstructive pulmonary disease and the fragility of most octogenarians contribute heavily to the development of this complication. Although the need for a tracheostomy was strongly predictive of death, a majority of tracheostomies were performed in cases of multiple organ failure or neurologic injury. It is interesting that the median length of stay for the entire series was 15 days, whereas patients in whom pulmonary complications developed had a median length of stay of 33 days. Many of these days were spent in a costly intensive care unit on a ventilator. A lower threshold for tracheostomy may be warranted in these patients. This may reduce their total hospital stay and cost and would most likely eliminate tracheostomy as a strong predictor of postoperative mortality. Left recurrent nerve palsy occurred in 42% of our patients with extent I and II aneurysms. Although we make every effort to preserve the nerve during dissection in the aortopulmonary window, we believe that two points must be kept in mind with respect to this complication. Proximal control is of paramount importance and supercedes nerve preservation. In addition, a high index of suspicion for nerve injury in the postoperative period allows us to identify such patients early and institute aspiration precautions and pulmonary physiotherapy once they are extubated. We aggressively involve our otolaryngology colleagues and all patients have undergone a type I thyroplasty with good results. Adherence to
Ann Thorac Surg 1998;65:491–5
these principles has made recurrent nerve palsy a much less threatening complication. In conclusion, we believe that age alone should not exclude a patient from undergoing repair of a TAAA. Octogenarians with TAAAs more than 6.0 cm in diameter should be considered for operative intervention if there are no systemic contraindications to operation. To avoid major postoperative morbidity and mortality a thorough investigation of their cardiac status is appropriate in elective cases. Although their rate of major postoperative morbidity is not greater than that of younger patient populations, the mortality associated with such complications is significant. Intraoperative adjuncts to preserve end organ function should be liberally applied. A more aggressive posture toward postoperative pulmonary complications may reduce the length of stay and cost.
GIRARDI AND COSELLI THORACOABDOMINAL AORTIC ANEURYSMS
9.
10. 11.
12. 13. 14. 15.
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. Coselli JS. Thoracoabdominal aortic aneurysms: experience with 372 patients. J Cardiol Surg 1994;9:638– 47. 3. 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. 4. Havlik RJ, Liu BM, Kovar MG, et al. National Center for Health Statistics: Health, United States, 1992. 5. Merrill WH, Stewart JR, Frist WH, et al. Cardiac surgery in patients age 80 years or older. Ann Thorac Surg 1990;211: 772–5. 6. Morris RJ, Strong MD, Grunewald KE, et al. Internal thoracic artery for coronary artery grafting in octogenarians. Ann Thorac Surg 1996;62:16–22. 7. Harris KA, Ameli FM, Lally M, et al. Abdominal aortic aneurysm resection in patients more than 80 years old. Surg Gynecol Obstet 1986;162:536– 8. 8. Dean RH, Woody JD, Enarson CE, Hansen KJ, Plonk GW.
16. 17. 18. 19.
20. 21.
22.
495
Operative treatment of abdominal aortic aneurysms in octogenarians. Ann Thorac Surg 1993;6:721– 8. O’Hara PJ, Hertzer NR, Krajewski LP, et al. Ten-year experience with abdominal aortic aneurysm repair in octogenarians: early results and late outcome. J Vasc Surg 1995;21: 830– 8. Kaplan EL, Meier P. Non-parametric estimation from incomplete observations. J Am Stat Assoc 1958;53:457– 81. 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. Fiore AC, Naunheim KS, Barner HB, et al. Valve replacement in the octogenarian. Ann Thorac Surg 1989;48:104– 8. Naunheim KS, Kern MJ, McBride LR, et al. Coronary artery bypass surgery in patients aged 80 years or older. Am J Cardiol 1987;59:804–7. Glock Y, Smile E, Dalous P, et al. Abdominal aortic aneurysmectomy in octogenarian patients. J Cardiovasc Surg 1990; 31:71– 6. Young JR, Hertzer NR, Beven EG, et al. Coronary artery disease in patients with aortic aneurysms: a classification of 302 coronary angiograms and results of surgical management. Ann Vasc Surg 1986;1:36– 42. Orecchia PM, Berger PW, White CJ, et al. Coronary artery disease in aortic surgery. Ann Vasc Surg 1988;2:28–36. Gloviczki P, Pairolero PC, Mucha P, et al. Ruptured abdominal aortic aneurysms: repair should not be denied. J Vasc Surg 1992;15:851–9. Johansen K, Kohler RT, Nicholls SC, et al. Ruptured abdominal aortic aneurysms: the Harborview experience. J Vasc Surg 1991;13:240–5. Svensson LG, Hess KR, Coselli JS, Safi HJ. Influence of segmental arteries, extent, and atriofemoral bypass on postoperative paraplegia after thoracoabdominal aortic operations. J Vasc Surg 1994;20:255– 62. Panneton JM, Hollier LH. Nondissecting thoracoabdominal aortic aneurysms: part I. Ann Vasc Surg 1995;9:503–14. Kouchoukos NT, Daily BB, Rokkas CK, Murphy SF, Bauer S, Abboud N. Hypothermic bypass and circulatory arrest for operations on the descending thoracic and thoracoabdominal aorta. Ann Thorac Surg 1995;60:67–77. Safi HJ, Harlin SA, Miller CC, et al. Predictive factors for acute renal failure in thoracic and thoracoabdominal aortic aneurysm surgery. J Vasc Surg 1996;24:338– 45.
Background. Reports on octogenarians undergoing coronary revascularization, valve replacement, and abdominal aneurysmorrhaphy demonstrate little increase in operative risk during elective procedures. However, the mortality in this group of patients increases rapidly when urgent or emergent procedures are performed. We analyzed the outcome of patients in their ninth decade of life undergoing repair of thoracoabdominal aortic aneurysms. Methods. A retrospective review of 39 consecutive octogenarians undergoing repair of thoracoabdominal aortic aneurysms. Results. Thirty-nine of the past 900 patients with thoracoabdominal aortic aneurysms (5.2%) repaired by us were between the ages of 80 and 89 years. The median age was 84 years with a male-to-female ratio of 1:3. Two of 39 patients (5%) had acute type III dissections, and the remainder had chronic aneurysms. Twelve patients had Crawford extent I aneurysms, whereas 7, 10, and 10 patients were extent II, III, and IV, respectively. The
overall in-hospital mortality was 10.3% (4 of 39 patients). Major postoperative complications included paraperesis/ paraplegia, 5% (n 5 2); renal failure, 18% (n 5 7) including hemodialysis in 3 patients; stroke, 5% (n 5 2); myocardial infarction or arrhythmia, 18% (n 5 7); and respiratory insufficiency, 36% (n 5 14) including 4 patients requiring tracheostomy. A univariate analysis of perioperative risk factors was performed using the Fisher’s exact test. The need for hemodialysis (p 5 0.035), a tracheostomy (p 5 0.0001), or a perioperative myocardial infarction (p < 0.001) significantly increased the risk of death. Conclusions. Repair of thoracoabdominal aortic aneurysms in octogenarians can be performed with acceptable morbidity and mortality. However, survival decreases dramatically with even single system organ failure. An extended period of recovery is usually required in these elderly, high-risk patients. (Ann Thorac Surg 1998;65:491–5) © 1998 by The Society of Thoracic Surgeons
R
elective and emergent repair are viable options in these high-risk patients.
epair of thoracoabdominal aortic aneurysms (TAAA) remains a formidable operation with substantial morbidity and mortality. Advances in surgical technique and perioperative care have reduced the mortality in most series to approximately 10%, and major perioperative morbidity such as paraplegia and renal failure continues to decrease in incidence [1–3]. The mean age at operation of patients undergoing repair of TAAA is 65 to 70 years in most series. However, the average life expectancy for a United States citizen is more than 76 years, with more than 10 million people older than 80 years. In addition, the average octogenarian can expect to live 6 to 10 years beyond his or her 80th birthday [4]. Major elective cardiovascular reconstruction in this population has not proven to be significantly more risky than in patients less than 80 years of age. Coronary artery bypass, valve replacement, and abdominal aortic aneurysm repair have been performed with acceptable perioperative risk. It is only when these elderly patients are subjected to emergent intervention that the mortality increases substantially [5–9]. We analyzed our experience with TAAA repair in octogenarians to determine whether Accepted for publication Aug 22, 1997. Address reprint requests to Dr Coselli, 6560 Fannin, Suite 1144, Houston, TX 77030 (e-mail: [email protected]).
© 1998 by The Society of Thoracic Surgeons Published by Elsevier Science Inc
Material and Methods Nine hundred consecutive patients underwent operation on the thoracoabdominal aorta from March 1989 through March 1997. Thirty-nine (4.3%) were 80 years of age or older. A retrospective analysis of these 39 patients was performed. Complete follow-up was available in all but 2 patients. Perioperative risk factors were subjected to univariate analysis using a Fisher’s exact test. Survival was determined using the method of Kaplan and Meier [10]. All statistics were performed using SAS for Windows statistical software program (SAS Institute Inc, Cary, NC). Mortality was defined as death within 30 days of the procedure or before discharge from the hospital. Paraplegia and paraperesis were defined as any lower extremity neuromuscular deficit, either immediate or delayed, that occurred in the hospital. Renal failure was defined as the need for hemodialysis or creatinine twice baseline level. Respiratory failure was defined as prolonged intubation (.48 hours), pneumonia, adult respiratory distress syndrome, or the need for a tracheostomy. Cardiac complications included arrhythmias, myocardial infarction, or congestive heart failure. 0003-4975/98/$19.00 PII S0003-4975(97)01300-3
492
GIRARDI AND COSELLI THORACOABDOMINAL AORTIC ANEURYSMS
Ann Thorac Surg 1998;65:491–5
Table 1. Complications and Mortality in Octogenarians Undergoing Thoracoabdominal Aortic Aneurysm Repair Complication Pulmonary Cardiac Renal Left recurrent nerve palsy Paraperesis/paraplegia Cerebrovascular accident Reexploration for bleeding Mortality
Fig 1. Number of patients in each category of aneurysm extent according to the Crawford classification system for thoracoabdominal aortic aneurysms [11].
Results Thirty-nine octogenarians underwent repair of TAAA. Ages ranged from 80 to 89 years with a median age of 84 years. There were 26 women and 13 men. Preoperative risk factors are as follows: hypertension, 24 (61%); chronic obstructive pulmonary disease, 14 (36%); coronary artery disease, 13 (33%); renal occlusive disease, 12 (31%); aneurysm rupture, 7 (18%); renal insufficiency (creatinine level .2.5g/dL), 6 (15%); diabetes mellitus, 3 (8%); cerebrovascular accident, 2 (5%); and acute type III dissection, 2 (5%). Of note, 7 patients (18%) presented with either a contained or free rupture of the aneurysm. The distribution of aneurysms according to the classification system of Crawford and colleagues [11] is seen in Figure 1. The technique for operative repair of TAAAs has been described in detail elsewhere [2]. Briefly, heparin (1 mg/kg) was administered to all patients before aortic cross-clamping. All large, patent intercostal arteries in the region between T8 and L1 were reimplanted into a Hemashield Dacron prosthetic (Meadox Medical, Oakland, NJ). A total of 19 patients had 1 to 5 pairs of intercostal arteries included in their repair. Seven of 19 patients believed to be at high risk for paraplegia (extents I and II) were chosen for atriofemoral or atriodistal aortic bypass using a Biomedicus pump (Minneapolis, MN). The 12 patients excluded from the use of partial bypass included patients with large friable intraluminal atheroma or thrombus, and those with extremely large aneurysms (9 to 13 cm) that would make distal cross-clamping unsafe. Complications and mortality are summarized in Table 1. Perioperative in-hospital mortality occurred in 4 patients (10.3%). Two patients had postoperative myocardial infarctions that progressed to multiple organ failure
No. of Patients (%) 14 (36%) 7 (18%) 7 (18%) 8 (21%) 2 (5.1%) 2 (5.1%) 1 (2.6%) 4 (10.3%)
and death. Another patient died of adult respiratory distress syndrome of unknown cause. The final death occurred on postoperative day 29 after an intraoperative cerebrovascular accident. All patients with ruptured aneurysms survived to discharge. There were no intraoperative deaths. There was a total of four major neurologic complications. Two patients (5.1%) had neurologic deficits secondary to spinal cord ischemia. One patient developed paraperesis and one developed delayed paraplegia on postoperative day 9 after TAAA repair in the setting of an acute type III dissection. The paraplegia in the latter patient was noted after hemodialysis during which the patient had multiple hypotensive episodes. Despite the institution of cerebrospinal fluid drainage, pressors, and steroids, the neurologic injury was permanent. Both of these patients had extent III aneurysms that were repaired without the support of atriofemoral bypass. The presence of an acute dissection was predictive of a postoperative neurologic event (p , 0.05) but not death. Extent III aneurysms were similarly statistically predictive of a postoperative neurologic deficit (p , 0.038). Two patients (5.1%) had perioperative cerebrovascular accidents. One patient never awakened postoperatively and subsequently died. The second developed a hemiparesis on postoperative day 7. No source for her stroke was found. Fortunately, she recovered and was ambulating with assistance before discharge. A total of 7 patients experienced postoperative renal failure (18%). Four patients had their renal function return to baseline level and never required hemodialysis. Three patients (7.7%) with aneurysms of different extent (II, III, IV) required hemodialysis; subsequently 2 died in the hospital. The development of postoperative renal failure could not be predicted based on a patient’s preoperative renal function (p . 0.49) or the presence of renal arterial occlusive disease (p . 0.15). Neither atriodistal bypass nor cold renal perfusion were protective against the development of postoperative renal failure (p . 0.05). In addition, cross-clamp time did not influence postoperative renal function (p . 0.05). However, the need for hemodialysis was predictive of postoperative mortality (p , 0.03). Postoperative pulmonary complications were present in 14 patients (36%). A majority of these complications were prolonged intubations (10%, 4 patients) or reintu-
Ann Thorac Surg 1998;65:491–5
Fig 2. Survival of 39 octogenarians undergoing repair of thoracoabdominal aortic aneurysms. The median survival is 31.5 months with an actuarial 5-year survival of 50%.
bations (15%, 6 patients). Four patients eventually required tracheostomies. All four died in the perioperative period. There were seven cardiac complications in the perioperative period. Five patients required medical management of supraventricular tachyarrhythmias and 2 (5.1%) patients experienced fatal transmural myocardial infarctions. Although the development of supraventricular tachyarrhythmias was not predictive of a poor outcome, a postoperative myocardial infarction was strongly predictive of mortality (p , 0.0013). One patient required reexploration for postoperative hemorrhage (2.6%). This patient had a splenic laceration and required splenectomy. Wound infections requiring local wound care or debridement developed in 2 patients and a left true vocal cord palsy, in 8 patients (21%). All of these patients had extent I or II aneurysms. All underwent thyroplasty before discharge from the hospital with good result. The univariate risk factors predictive for death are tracheostomy (p , 0.0001), myocardial infarction (p , 0.0013) and hemodialysis (p , 0.035) and for neurologic injury are acute type III dissection (p , 0.05) and extent III aneurysm (p , 0.038) in the perioperative period. The median length of stay for these 39 patients was 15 days with a range of 10 to 86 days. The median survival for all 39 octogenarians is 31.5 months with a 5-year actuarial survival of 50% (Fig 2).
Comment As the population continues to age, it is becoming increasingly common to perform major cardiovascular operations on those older than 80 years. Early series of valve replacement [12], coronary artery bypass [13], and abdominal aneurysmectomy [7] reported mortality rates 5 to 10 times higher than patients of lesser age undergoing similar procedures. Recently, however, the mortality for these elective procedures has been reduced to 5% to 10%. Specifically, elective abdominal aortic aneurysm repair has had a reproducible mortality of approximately 6% [8, 9, 14]. We sought to examine our results with repair of
GIRARDI AND COSELLI THORACOABDOMINAL AORTIC ANEURYSMS
493
TAAAs in octogenarians to determine whether repair is warranted in this high-risk population. Octogenarians with TAAAs should not be denied intervention because of excessive perioperative mortality. Our mortality of 10.3% is not significantly different from our previous report on 372 consecutive TAAA repairs in which the mean age was 66 years. Although there are no other series of TAAA repairs in this elderly patient population, comparison to similar patients undergoing repair of infrarenal abdominal aortic aneurysms (IAAA) is favorable. Dean and colleagues [8] reported an inhospital mortality of 38% in 34 consecutive octogenarians requiring IAAA repair. The mortality for elective cases was 6%, whereas the mortality rose to 40% for patients with symptomatic aneurysms and 91% in cases of ruptured IAAA. Results at the Cleveland Clinic were comparable with a slightly higher mortality for elective cases [9]. Examination of the causes of death in our series as well as these two reports is worthy of discussion. Two patients in our cohort died of a transmural myocardial infarction after TAAA repair, accounting for 50% of our mortality. Both of these patients were elective cases with no evidence of rupture or leak. Dean and associates [8] found that 23% of the deaths in octogenarians undergoing IAAA repair were secondary to acute myocardial infarctions. O’Hara and colleagues [9] encountered a similar problem with 31% of their patients’ deaths attributable to fatal myocardial ischemia. In addition, they reported that only 4% of their patients having previous myocardial revascularization had a fatal myocardial infarction, whereas 12% of those without revascularization succumbed to a fatal myocardial infarction. Certainly these are patients with a high incidence of coronary artery disease. Young and co-workers [15] reviewed the coronary angiograms in 302 patients undergoing IAAA repair. They noted that 46% of asymptomatic patients had at least single-vessel disease, with 20% having significant triple vessel disease. Patients with an anginal history had even higher rates of coronary artery disease requiring intervention. Orecchia and colleagues [16] confirmed these results in an additional 59 patients requiring aortic operation. Although no guidelines exist with respect to the proper cardiac evaluation in patients with TAAAs the high incidence of insidious coronary artery disease and myocardial infarction suggests a more exhaustive search for such disease is indicated in the elective surgical candidate. Obviously, patients with symptomatic or frankly ruptured aneurysms cannot undergo a thorough preoperative cardiac evaluation. One must also be cognizant that 7% to 10% of octogenarians requiring preoperative myocardial revascularization will not survive this intervention and an additional unknown percentage will rupture their aneurysm while undergoing these procedures. However, the risk of correcting significant coronary artery disease is clearly less than the uniformly fatal outcome when incurring a perioperative myocardial infarction. In addition, a survival benefit has been demonstrated for patients with known IAAAs undergoing preoperative coronary revascularization with angioplasty or bypass grafting when compared with a
494
GIRARDI AND COSELLI THORACOABDOMINAL AORTIC ANEURYSMS
nonrevascularized population (75% versus 30%, respectively) [9]. It is counterintuitive that we had no mortality for the 7 octogenarians presenting with ruptured TAAA. The mortality associated with this mode of presentation in IAAA is 33% to 91% [8, 9, 17, 18]. However, our patient population differs from that seen in these reports. All of our patients were hemodynamically stable at the time of presentation to our center. Dean and colleagues [8] reported that more than 50% of their patients with ruptured IAAA had unrecordable blood pressure before the application of a cross-clamp. The Mayo Clinic series reported 67% of their patients were hypotensive, whereas 24% had a cardiac arrest before operation [17]. Because a large percentage of our patients were transferred to our institution from distant geographic areas, those with hemodynamic instability or worse were selected out before our intervention. Some groups have advocated withholding operative intervention in octogenarians with ruptured aneurysms and hemodynamic compromise [8, 18]. Mortality rates are consistently greater than 90% with most patients dying of multiple organ failure after a prolonged intensive care unit stay. We have no data to refute this position and agree that resources may be better used in these desperate situations. The incidence of paraplegia and paraperesis in this group of patients is not significantly higher than in our previous report. A number of factors play a role in keeping this often fatal complication to a minimum. We continue to support an aggressive posture toward intercostal artery reimplantation, especially in patients at the highest risk for neurologic injury (extents I and II) [19]. All 19 patients in this group had at least one pair of intercostals reimplanted, either as a separate island of aorta or as part of the distal anastomosis. Extent III aneurysms were statistically more likely to be associated with postoperative neurologic injury in octogenarians. However, we believe the case of delayed paraplegia is more attributable to multiple hemodynamic insults associated with hemodialysis rather than aneurysm extent. If this case is eliminated from consideration, aneurysm extent no longer is a risk factor for postoperative neurologic deficits. Although often limited by large aneurysm size or mural atheromatous debris we prefer to use atriodistal bypass in this same high-risk group. The small sample size in this report precludes statistical support for its use, yet we believe that distal perfusion has a positive influence on neurologic outcome. It allows one to perform a technically sound proximal anastomosis without time constraint while maintaining intercostal and collateral flow to the cord. It also reduces total cord ischemic time in complex reconstructions in which multiple intercostal arteries are to be reimplanted along with visceral and renal artery reimplantation. This is especially true with aneurysms in the setting of an aortic dissection, a risk factor for paraplegia in this report and in others [20]. Kouchoukos and colleagues [21] support the use of profound hypothermia and circulatory arrest to limit spinal cord ischemia. Although we did not use this modality in
Ann Thorac Surg 1998;65:491–5
any of these 39 patients, we have had a 70% mortality with circulatory arrest for TAAA repair in patients older than 70 years. Although their neurologic function may be preserved, they are often too frail to withstand the systemic insult of this technique. We continue to reserve circulatory arrest for patients in whom the risk of stroke with proximal cross-clamping is extraordinarily high. Perhaps we could have avoided the single intraoperative cerebrovascular accident had we used profound hypothermia and circulatory arrest in that particular case. Acute renal failure after TAAA repair carries poor prognostic implications. A total of 18% of our patients experienced a reduction in renal function, with 7.7% requiring hemodialysis. In this small set of patients we were unable predict the development of renal failure based on preoperative renal function or renal occlusive disease. In addition, our small numbers do not statistically support the routine use of either atriodistal bypass or cold renal perfusion as adjuncts to preserving renal function. Nonetheless, we agree with Safi and colleagues [22] that such measures are protective against postoperative renal failure. In their series 18% of patients developed renal failure with 15% requiring hemodialysis. This complication carried a 50% mortality. However, the use of distal perfusion was protective against the development of acute renal failure. We also found hemodialysis to be highly predictive of postoperative mortality and support the use of distal perfusion or cold renal perfusion when technically feasible. Pulmonary complications were the most prevalent problem seen in octogenarians undergoing TAAA repair. The high incidence of chronic obstructive pulmonary disease and the fragility of most octogenarians contribute heavily to the development of this complication. Although the need for a tracheostomy was strongly predictive of death, a majority of tracheostomies were performed in cases of multiple organ failure or neurologic injury. It is interesting that the median length of stay for the entire series was 15 days, whereas patients in whom pulmonary complications developed had a median length of stay of 33 days. Many of these days were spent in a costly intensive care unit on a ventilator. A lower threshold for tracheostomy may be warranted in these patients. This may reduce their total hospital stay and cost and would most likely eliminate tracheostomy as a strong predictor of postoperative mortality. Left recurrent nerve palsy occurred in 42% of our patients with extent I and II aneurysms. Although we make every effort to preserve the nerve during dissection in the aortopulmonary window, we believe that two points must be kept in mind with respect to this complication. Proximal control is of paramount importance and supercedes nerve preservation. In addition, a high index of suspicion for nerve injury in the postoperative period allows us to identify such patients early and institute aspiration precautions and pulmonary physiotherapy once they are extubated. We aggressively involve our otolaryngology colleagues and all patients have undergone a type I thyroplasty with good results. Adherence to
Ann Thorac Surg 1998;65:491–5
these principles has made recurrent nerve palsy a much less threatening complication. In conclusion, we believe that age alone should not exclude a patient from undergoing repair of a TAAA. Octogenarians with TAAAs more than 6.0 cm in diameter should be considered for operative intervention if there are no systemic contraindications to operation. To avoid major postoperative morbidity and mortality a thorough investigation of their cardiac status is appropriate in elective cases. Although their rate of major postoperative morbidity is not greater than that of younger patient populations, the mortality associated with such complications is significant. Intraoperative adjuncts to preserve end organ function should be liberally applied. A more aggressive posture toward postoperative pulmonary complications may reduce the length of stay and cost.
GIRARDI AND COSELLI THORACOABDOMINAL AORTIC ANEURYSMS
9.
10. 11.
12. 13. 14. 15.
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. Coselli JS. Thoracoabdominal aortic aneurysms: experience with 372 patients. J Cardiol Surg 1994;9:638– 47. 3. 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. 4. Havlik RJ, Liu BM, Kovar MG, et al. National Center for Health Statistics: Health, United States, 1992. 5. Merrill WH, Stewart JR, Frist WH, et al. Cardiac surgery in patients age 80 years or older. Ann Thorac Surg 1990;211: 772–5. 6. Morris RJ, Strong MD, Grunewald KE, et al. Internal thoracic artery for coronary artery grafting in octogenarians. Ann Thorac Surg 1996;62:16–22. 7. Harris KA, Ameli FM, Lally M, et al. Abdominal aortic aneurysm resection in patients more than 80 years old. Surg Gynecol Obstet 1986;162:536– 8. 8. Dean RH, Woody JD, Enarson CE, Hansen KJ, Plonk GW.
16. 17. 18. 19.
20. 21.
22.
495
Operative treatment of abdominal aortic aneurysms in octogenarians. Ann Thorac Surg 1993;6:721– 8. O’Hara PJ, Hertzer NR, Krajewski LP, et al. Ten-year experience with abdominal aortic aneurysm repair in octogenarians: early results and late outcome. J Vasc Surg 1995;21: 830– 8. Kaplan EL, Meier P. Non-parametric estimation from incomplete observations. J Am Stat Assoc 1958;53:457– 81. 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. Fiore AC, Naunheim KS, Barner HB, et al. Valve replacement in the octogenarian. Ann Thorac Surg 1989;48:104– 8. Naunheim KS, Kern MJ, McBride LR, et al. Coronary artery bypass surgery in patients aged 80 years or older. Am J Cardiol 1987;59:804–7. Glock Y, Smile E, Dalous P, et al. Abdominal aortic aneurysmectomy in octogenarian patients. J Cardiovasc Surg 1990; 31:71– 6. Young JR, Hertzer NR, Beven EG, et al. Coronary artery disease in patients with aortic aneurysms: a classification of 302 coronary angiograms and results of surgical management. Ann Vasc Surg 1986;1:36– 42. Orecchia PM, Berger PW, White CJ, et al. Coronary artery disease in aortic surgery. Ann Vasc Surg 1988;2:28–36. Gloviczki P, Pairolero PC, Mucha P, et al. Ruptured abdominal aortic aneurysms: repair should not be denied. J Vasc Surg 1992;15:851–9. Johansen K, Kohler RT, Nicholls SC, et al. Ruptured abdominal aortic aneurysms: the Harborview experience. J Vasc Surg 1991;13:240–5. Svensson LG, Hess KR, Coselli JS, Safi HJ. Influence of segmental arteries, extent, and atriofemoral bypass on postoperative paraplegia after thoracoabdominal aortic operations. J Vasc Surg 1994;20:255– 62. Panneton JM, Hollier LH. Nondissecting thoracoabdominal aortic aneurysms: part I. Ann Vasc Surg 1995;9:503–14. Kouchoukos NT, Daily BB, Rokkas CK, Murphy SF, Bauer S, Abboud N. Hypothermic bypass and circulatory arrest for operations on the descending thoracic and thoracoabdominal aorta. Ann Thorac Surg 1995;60:67–77. Safi HJ, Harlin SA, Miller CC, et al. Predictive factors for acute renal failure in thoracic and thoracoabdominal aortic aneurysm surgery. J Vasc Surg 1996;24:338– 45.