---

Surgical treatment of abdominal aortic aneurysms in octogenarians: long-term results
(Portuguese PDF version)

Márcio Luís Lucas,1 Elton Luiz Schmidt Weber,2 Telmo Pedro Bonamigo3

1.Resident physician, Service of Angiology and Vascular Surgery, Santa Casa de Misericórdia de Porto Alegre, Porto Alegre, RS, Brazil.
2. Vascular Surgeon. Former-resident, Service of Angiology and Vascular surgery, Santa Casa de Misericórdia de Porto Alegre, Porto Alegre, RS, Brazil.
3. Chief of the Service of Angiology and Vascular Surgery, Santa Casa de Misericórdia de Porto Alegre, Porto Alegre, RS, Brazil.

Correspondence:
Telmo Pedro Bonamigo
Rua Coronel Bordini, 675/303
CEP 90440-001 - Porto Alegre, RS, Brazil
Tel.: +55 (51) 3333.1642
E-mail: telmobonamigo@terra.com.br

---

ABSTRACT

Objective: To evaluate the perioperative results of conventional open repair of non-ruptured abdominal aortic aneurysm in octogenarians by analyzing the early and late outcomes in patients with or without symptoms associated with abdominal aortic aneurysm.

Patients and method: Between March 1982 and March 2002, 41 patients (mean age = 82.7) were divided into two groups. Group A (n = 21): asymptomatic; and group S (n = 20): symptomatic. The results were evaluated by the 2 test, exact Fisher's test, and t test when necessary. The survival rate was studied by Kaplan-Meier study and long rank test. A value of P < 0.05 was considered significant.

Results: The presence of hypertension, coronary artery disease and renal insufficiency was similar between the groups. Males and smoking history were more prevalent in the asymptomatic group (P < 0.05). The overall mortality rate was 9.75%, been higher in the group S (15 vs 4.76%; P = 0.343). The survival rates in 1, 3, 5, and 7 years were 94.2, 67.8, 47.2, and 33%, respectively. There was not any difference between the studied groups.

Conclusions: The perioperative mortality obtained from octogenarians submitted to open repair of not-ruptured abdominal aortic aneurysm seems to be safe and it is in accordance with the current data. The presence of symptoms increased this mortality rate without statistical difference, but the long term results were similar between this two groups.

Key-words: abdominal aortic aneurysm, 80 and over aged, treatment outcome.

J Vasc Br 2004;3(4):339-47

---

Old age can be considered an important risk factor for the development of perioperative complications.¹ The mortality rate in the elective repair of abdominal aortic aneurysm (AAA) in patients older than 80 years can reach 8.6%.² This value can be even higher (11 to 33%) in urgent surgeries (management of enlarged aneurysm),^3,4 varying from 36 to 91% in cases of aneurysm rupture.^4,5

Normally, the aneurysmal disease in octogenarians is diagnosed when the aneurysm is already dilated, and there is a considerable risk of rupture in short or medium term.⁶ The indication for surgery should be carefully considered, as these patients have an increased risk of undergoing perioperative complications.7 On the other hand, it should be noticed that their life expectation can be of 6 or 7 years, and so the treatment of potentially fatal diseases, like the AAA, should be considered as well.^6,8

Our aim was to assess the surgical morbidity and mortality, as well as the long-term results of abdominal aortic aneurysmectomy in octogenarians. Besides, we analyzed the presence or absence of symptoms during the patients' recovery.

PATIENTS AND METHODS

The preoperative diagnosis was based on the patient's history and clinical examination, as well as abdominal ultrasonography, computed tomography, and aortography in specific cases. The surgical procedure was performed with a conventional technique of transperitoneal approach to the abdominal aorta and xiphopubic incision. The patients remained in the intensive care unit for the first 24 hours, and additional time of permanence was indicated in cases of occasional complications.

Morbidity and mortality rates were assessed within 30 days after surgery. The patients' follow-up - through medical consultations, interview by phone or questionnaire sent by mail - allowed us to know the patients' survival rates, as well as the causes of death during the late postoperative period.

The patients were divided into two groups according to the absence (group A) or presence (group S) of symptoms (pain) related to AAA. The demographic characteristics and patients' comorbidities, and the features inherent to the aneurysmal disease, were discriminated for each group, and possible statistical differences among them were checked.

The obtained data were expressed in absolute or relative values (percentages), the average and standard deviation of values being calculated when necessary. The statistical analysis was performed using the chi-square (χ²) test, Fisher exact test or t-test when indicated. The analysis of survival was done through the Kaplan-Meyer curve, and the log-rank test was used. A value of P < 0.05 was considered significant.

RESULTS

Mean age of the octogenarians was significantly higher than the age of patients under 80 (82.7 ± 2.3 versus 68.2 ± 6.5 years; P = 0.000). There was no age difference between the two groups of octogenarians; mean age in group A was 82.5 (± 1.74) years and in group S, 83 (±2.86) years (P = 0.5). Twenty-nine out of 41 patients (70.7%) were male, with a higher prevalence of males in patients of group A (95.2%) compared to group S (45%) (P = 0.002). Regarding comorbidities, high blood pressure, smoking, ischemic cardiopathy and renal insufficiency were present in 65.8, 51.2, 22 and 7.3% of patients, respectively. Diabetes mellitus was not diagnosed in any patient. Among the groups, the presence of these comorbidities did not have statistically significant difference (Table 1).

Table 1 - Demographic and clinical features of octogenarians who underwent aneurysmectomy
of abdominal aorta

	Group A		Group S		P
Comorbidity	n	%	n	%
Sex
Male	20	95.4	9	45	0.002
Female	1	4.6	11	55
Smoking	18	85.7	9	45	0.009
Hypertension	11	52.4	10	50	NS
Ischemic cardiopathy	6	28.6	3	15	NS
Renal insufficiency	3	14.3	-	0	NS

Fisher exact test; statistically significant difference when P < 0.05; NS = Not significant.

The main complaint was isolated abdominal pain in 10 patients of group S (50%), while lower back pain was present in eight patients (40%). The coexistence of lower back and abdominal pain was observed in only two cases (10%).

The aneurysm diameter in octogenarians ranged from 4 to 18 cm, and the average of these values was not statistically different from those obtained among patients with less than 80 years (6.95 ± 2.5 vs 6.47 ± + 2.2 cm, respectively; P = 0,179). For groups A and S, the averages of these dimensions were 6.47 (±1.94) cm and 7.46 (±3.08) cm, respectively (P = 0.223). The presence of inflammatory AAA was observed in two patients (4.9%), one in each group.

The mean operative time was 201.7 (±59,14) minutes for group A and 195 minutes (53.75) for group S (P = 0.707), probably depending on the type of vascular reconstruction used (Table 2). The blood loss was similar in the two groups, 686.9±177.95 ml in group A and 699.4± 182.01 ml in group S (P = 0.825). In group A, two simultaneous surgical procedures were undertaken besides the surgery for AAA in three different patients: a left aorto-renal bypass in a patient with previous renal insufficiency and renovascular hypertension; and a nefrectomy in a patient who had renal neoplasm on the left side. A third complementary procedure was necessary, a transoperative splenectomy due to accidental injury of the spleen in another patient.

Table 2 - Surgical parameters of correction of abdominal aortic aneurysm in
octogenarians with or without symptoms

Parameters	Group A	Group S	P
Number of patients	21	20	-
Operative time * (min)	201.7±59.14	195±53.75	NS
Blood loss * (ml)	686.9±177.95	694.4±182.01	NS
Combined surgery	3	-	NS
Type of reconstruction
Aorto-aortic	9	7
Aorto-bi-iliac	11	10
Aorto-iliacL+Aorto-femoralR	1	-
Aorto-bi-femoral	-	3

* t-test. Statistically significant difference when P < 0.05. NS = Not-significant.

The global mortality rate in 30 postoperative days was 3.13, 9.75% in octogenarians and 2.72% in patients under 80 (P = 0.041). The mortality rate in group S was 15%, while in group A it was 4.76% (P = 0.343). The causes of postoperative death among octogenarians are shown in Table 3, and the non-fatal complications are listed in Table 4. The morbidity rates for the asymptomatic and symptomatic groups were 14.3% and 20%, respectively, and the respiratory complications were more frequent in the group of patients with symptoms related to AAA (10 vs 4.76%; P >0.05).

Table 3 - Postoperative mortality in octogenarians who underwent aneurysmectomy

Cause of Death	Group A (n = 21)	Group S (n = 20)	P
Acute myocardial infarction	1	1
Mesenteric ischemia	-	1
Pulmonary embolism	-	1
	1 (4.76%)	3 (15%)	0.343
Total	4 (9.75%)

Fisher exact test; statistically significant difference when P < 0.05;.

Table 4 - Non-fatal complications after abdominal aortic aneurysmectomy in octogenarians

Complication	Group A	Group S	P
Atelectasis/respiratory infection	1	2
Urinary infection	1	-
Hemorrhage	1 *	1 †
Renal insufficiency ‡	-	1
Total	3/21 (14.3%)	4/20 (20%)	0.697

* Retroperitoneal hematoma.
† Hemorrhage requesting transoperative splenectomy.
‡ Transient worsening of postoperative renal function.

The patients discharged from hospital (n = 37) formed a retrospective cohort of this study. During the clinical follow-up, two patients were missed (5.4%). The average follow-up of all patients was 47.9 ± 41.5 months for the asymptomatic group (variation of 3 to 167 months). The average survival was 67 ± 10 months, being 61 ± 14 months for the asymptomatic group and 71 ± 16 months for the symptomatic group. The global rate of survival in 1, 3, 5 and 7 years was 94.2, 67.8, 47.2 and 33%, respectively (Figure 1). The analysis of the survival curve does not show any statistically significant difference among the rates obtained in groups A and S (Figure 2). In addition, these patients' global survival was around 20% in 10 years.

Figure 1 - Global survival of octogenarians after the surgical treatment of abdominal aortic aneurysm.

Figure 2 - Global survival of asymptomatic and symptomatic octogenarians after the surgery of abdominal aortic aneurysm

The causes of late death, during clinical follow-up, are summarized in Table 5. There was a predominance of cardiac (26.1%), neurological (21.7%) and respiratory (8.7%) causes.

Table 5 - Causes of death during the clinical follow-up of octogenarians who
underwent abdominal aortic aneurysmectomy

Causes of death	Group A	Group B	% (n = 35)
Cardiac	3	3	26.1
Neurological	4	1	21.7
Respiratory	-	2	8.7
Other 2 -			8.7
Unknown	4	2	26.1

DISCUSSION

The prevalence of AAA grows progressively as the age increases, with percentages of 1% in individuals of 55-60 years to 10% in octogenarians.⁹ In addition, aneurysms in octogenarians can present with wider diameters at the time of the diagnosis. This fact may be reflecting the tendency of most physicians to delay the referral of these patients to a vascular surgeon. Therefore, there is an increased risk of rupture, since there is a direct relation between the dimension of the AAA and the chance of rupture. Rates of rupture around 50% have been demonstrated in patients with AAA between 6 and 7 cm of diameter.¹⁰ Conway et al.¹⁰ also observed that only 17% of the octogenarians with AAA, who did not undergo surgical treatment, had a survival of three years. In our study, the patients treated with conventional surgery showed survival rates of 70% at the end of a three-year-follow-up.

For some authors, the percentage mortality with surgical treatment of AAA from 80 years old on is higher than in the younger groups.^1,11 However, for other authors, the morbidity and mortality rates were similar in the two groups.^12,13 In our study, the surgery of AAA in octogenarians showed to be feasible, with a global mortality rate of 9.75%, being comparable to the indexes obtained by other authors like Kazmers et al.⁷ (8.25%), Kruger et al.¹³ (6.25%) e Treiman et al.² (8.6%). In a survey by the NIS (Nationwide Inpatient Sample) made with a large sample of patients treated in the USA, Vemuri et al.¹ observed a perioperative mortality of 7.9% in octogenarians who were on treatment for AAA. On the other hand, some groups demonstrated indexes lower than 5%, like O'Donnell et al.³ (4.7%), Falk et al.⁶ (0%) and Robson et al.⁴ (0%). There are other very important cooperative works, like the one published by Kazmers et al.⁷ of the Veterans Medical Affairs Administration, which reported on 5,627 surgeries of AAA with a mortality rate of 4.1% in patients under 80, and 8.25% in 206 patients over 80. Dardik et al.¹⁴, reported an experiment carried out in the state of Michigan: the mortality rate in AAA surgery of 2335 patients was 2.2% for patients under 65; 2.5% for those between 65 and 69 years; 3.5% for those between 70 and 79 years, and 7.3% for the octogenarians. Another reference study published in 1995, carried out by the Cleveland Clinic, reported a mortality rate of 1.25% in 573 patients under 80 , and 3.8% in 53 patients over 80.11 Our study was based on an experiment of 701 cases, 660 patients (94.2%) under 80 years with a mortality rate of 2.72%, and 41 patients (5.8% of the total) over 80 years, with a mortality rate of 9.75% (P < 0.05).

In our patients, the frequency of octogenarians as compared to the patients treated for AAA was 5.8%. In other series, like the one studied by Kruger et al.,¹³ this ratio can reach 24.5%. In the American NIS sample, this proportion was of 12.2%.1 In Chile, Valdés et al.¹⁵ found a proportion of 12% of patients over 80. In the experiment by the Veterans Medical Affairs Administration, in a total of 6260 patients operated between 1991 and 1995, only 3.7% were octogenarians.⁷

The main comorbidities found in our patients were high blood pressure (65.8%) and smoking history (51.2%), as it is demonstrated in other studies.^6,13 In the experiment by Valdés et al.,¹⁵ the presence of hypertension and smoking was observed in 18.7 and 17.5%, respectively. The frequency of hypertension did not differ between the asymptomatic and symptomatic groups. Smoking history was almost twice more frequent in the first group.

In the analysis of the patients in question as for the presence or not of symptoms related to AAA, there was a threefold increased risk of death in symptomatic patients, although based on the statistical study this difference has not showed to be significant. This fact can occur because our sample is small and does not allow for a definite conclusions about the subject.

The clinical conditions of patients submitted to conventional surgery must be taken into account when mortality is being assessed.

In principle, it is obvious that the octogenarians have a higher degree of well-known respiratory, cardiac and renal dysfunctions. However, there is another very important factor, the patient's clinical condition when the indication for surgery is made. In general, the patient with symptomatic AAA is the one for whom the surgical treatment was not previously proposed by the clinician, due to the comorbidities. However, when AAA becomes symptomatic, painful due to expansion or with an enlarged diameter, the surgical indication is more clear and well-accepted by the family. These patients can also present a higher degree of more diffuse atherosclerotic disease, with stenosis or dilation of the iliac arteries, which can make the surgery more difficult and raise the volume of blood replacement during the procedure. In this context, the asymptomatic patients treated electively can have a mortality rate of 0 to 8.6%;^2,6 those who are ill with symptoms related to AAA (expansion) can have a mortality rate of 11 to 33%;^3,4 and the patients treated during the event of rupture can vary from 36 to 91%.^4,5 On the contrary, the symptomatic group of our study demonstrated a significantly higher proportion of female patients compared to group S (55 vs 4.6%; P = 0.002). This could be another possible explanation for the higher mortality rate in this group compared to the asymptomatic group, since a higher mortality among women who underwent aneurysmectomy of the abdominal aorta compared to men is well demonstrated.¹⁶ However, the review of mortality among symptomatic patients evidenced that only one out of the three deaths in the group was of a female patient.

When the mortality of our asymptomatic octogenarian patients is analyzed, there is not any statistically significant difference compared to our patients under 80 years (4.76 vs 2.72%, P > 0.05). In addition, these values are very near to those obtained by great series of multicentric works, like the French one, published by Kieffer et al.,¹⁷ with a mortality rate of 4.8%; the Canadian, published by Johnston & Scobie,¹⁸ with a mortality rate of 4.8%; and the one published by Dardik et al.,¹⁴ reporting the experiment of Michigan, of 2,335 cases with a mortality rate of 3.5%. It is important to note that, in the three reports mentioned, the mortality is of the total population of patients, involving all ages and both sexes.

When the causes of death were analyzed in our work, we observed that there was one death, due to myocardial infarction, among the 21 asymptomatic patients. In the group of symptomatic patients, there were three deaths, caused, respectively, by myocardial infarction, mesenteric thrombosis and pulmonary embolism, what corresponded to 15%. Therefore, the main cause of death (50% of patients) was myocardial infarction, and the same proportion was observed by Krueger et al.¹³ The main complications observed in all patients were respiratory (7.3%) and hemorrhagic (4.8%).

Another very interesting aspect to be described is the analysis of the clinical follow-up of patients who survived the surgical procedure. Due to a great effort, a clinical follow-up of 94.6% of patients was possible because we maintained contact with the assistant physician, by mail or telephone, and the follow-up of only two patients was missed. The average time of our patients' follow-up remained around 48 months, the global average survival being 67 months.

Our patients' follow-up showed that the most frequent causes of death were the cardiovascular diseases (acute myocardial infarction, cardiac failure and stroke), accounting for 47.6% of the known causes of death, while the respiratory diseases accounted for 9.5% of the total. It should be pointed out that, in six patients, the cause of death was not identified and that the clinical follow-up of two patients was missed.

The global survival in five years was 47.2%. This finding is similar to those by Valdés et al.¹⁵ and O'Hara et al.,¹¹ which were 41.3 and 48% in the same time interval. The best results were reported by Sugawara et al.¹⁹ and Falk et al.,⁶ with survival in five years of 85.7 and 67%, respectively. This difference in results can be accounted for the fact that these works are from countries which offer better assistance conditions during the pre- and postoperative follow-up of elderly patients.

In Latin America, the life expectation of 80-year-olds can come to seven years.8 In Brazil, the life expectation of sixty-year-old people is around 17 years,²⁰ this data reinforce the trend toward the surgical treatment of AAA in elderly people. The average survival among our patients was 67 months, result that is similar to that obtained by Aune et al.,²¹ who obtained an average survival around five years in octogenarians who underwent conventional surgical repair of ruptured AAA.

Regarding the endovascular treatment, the stability of results obtained by this type of procedure has not yet been well-established. The follow-up of these patients shows a rate around 40% of specific complications of this type of therapy (endoleaks), which are correlated with an increase in the rupture rate after endovascular procedure.²² The perioperative mortality in octogenarians who undergo endovascular repair of AAA has been similar to that obtained by the conventional surgery and there is no statistically significant difference in these patients' survival during the clinical follow-up.²³ However, the rates of perioperative complications seem to be higher in patients who undergo the conventional treatment. In the study by Sicard et al.,²³ the morbidity rate in the conventional group (37%) was significantly higher than the one in the endovascular group (11.5%). However, approximately 30% of the complications in the conventional group were related to gastrointestinal complications, the most of them being diagnosed as prolonged ileum. In addition, except for these complications, there was no difference in the complication rates between the two groups. Another aspect that limits the results of endovascular therapy for AAA in octogenarians is the short-term clinical follow-up to date. In the study published by Biebl et al.,²⁴ the average time of follow-up was 16 months, similar to that described in the work by Sicard et al.²³ During the follow-up, the occurrence of endoleaks in octogenarians, in 1, 2 and 3 years, was of 26.2, 30 and 61.7%, respectively. Another interesting fact is that these authors observed a rate of significant renal dysfunction (14%) in these patients after the procedure, probably related to the use of contrast.²⁴

Briefly, the surgery for AAA in octogenarians is feasible and should be considered, since the indexes of morbidity and mortality are acceptable and similar to the observations of other authors, and there are also good long-term outcomes. In addition, the octogenarians, in a near future, will form a subgroup of acceptable patients who are candidates for surgical repair of AAA. In this study, the high rate of mortality among the symptomatic patients compared to asymptomatic patients can reflect a delay in the referral of those patients to surgery.

REFERENCES

1. Vemuri C, Wainess RM, Dimick JB, et al. Effect of increasing patient age on complication rates following intact abdominal aortic aneurysm repair in the United States. J Surg Res 2004;118:26-31.

2. Treiman RL, Levine KA, Cohe JL, Cossman DV, Foran RF, Levin PM. Aneurysmectomy in octogenarian. A study of morbidity and quality of survival. Am J Surg 1982;144:194-7.

3. O'Donnell TFO, Darling RC, Linton RR. Is 80 years too old for aneurysmectomy? Arch Surg 1976;111:1250-7.

4.Robson AK, Currie IC, Poskitt KR, Scott DJ, Baird RN, Horrocks M. Abdominal aortic aneurysm repair in the over eighties. Br J Surg 1989;76:1018-20.

5. Dean RH, Woody JD, Enarson CE, Hansen KJ, Plonk GW. Operative treatment of abdominal aortic aneurysms in octogenarians. When is too much too late? Ann Surg 1993;217:721-8.

6. Falk V, Vettelschob M, Walther T, et al. Surgical treatment of abdominal aortic aneurysms of octogenarians. Cardiovasc Surg 1996;4:727-31.

7. Kazmers A, Perkins AJ, Jacobs LA. Outcomes after abdominal aortic aneurism repair in those > or = 80 years of age: recent Veteran Affairs experience. Ann Vasc Surg 1998;12:106-12.

8. Centro Latinoamericano y Caribeño de Demografía - CELADE. Boletín demográfico no. 67. América Latina: Tablas de Mortalidad. 1950-2025. CELADE; enero 2001.

9. Pleumeekers HJ, Hoes AW, van der Does E, et al. Aneurysms of the abdominal aorta in older adults. The Rotterdam Study. Am J Epidemiol. 1995;142:1291-9.

10. Conway KP, Byrne J, Townsend M, Lane IF. Prognosis of patients turned down for conventional abdominal aortic aneurysm repair in the endovascular and sonographic era: Szilagyi revisited? J Vasc Surg. 2001;33:752-7.

11. O'Hara PJ, Hertzer NR, Krajewski LP, et al. Ten-year experience with abdominal aortic repair in octagenarians: early results and late outcome. J Vasc Surg 1995;21:830-8.

12. Paty PS, Lloyd WE, Chang BB, Darling RC 3rd, Leather RP, Shah DM. Aortic replacement for abdominal aortic aneurysm in elderly patients. Am J Surg. 1993;166:191-3.

13. Kruger A, Foster W, Love A, Woodruff P, Blackford J. Abdominal aortic aneurysm repair in the veteran population. ANZ J Surg 2002;72:628-31.

14. Dardik A, Lin JW, Gordon TA, Williams GM, Perler BA. Results of elective abdominal aortic aneurysm repair in the 1990's: a population based analysis of 2335 cases. J Vasc Surg 1999;30:985-95.

15. Valdés F, Bergoeing M, Krämer A, et al. Aneurysma aórtico abdominal en pacientes mayores de 80 anos: tratamiento quirúrgico convencional en 80 casos consecutivos. Rev Med Chile 2003;131:981-6.

16. Katz DJ, Stanley JC, Zelenock GB. Gender differences in abdominal aortic aneurysm prevalence, treatment, and outcome. J Vasc Surg. 1997;25:561-8.

17. Kieffer E, Koskas F, Deailly J, Gouny P. Mortalité péri-opératoire de la chirurgie des anévrysmes de l'aorte abdominale: Étude multicentrique de l'AURC. In. Kieffer E, editor. Les Anevrysmes de l'Aorte Abdominale Sous-rénale. Paris: AERCV; 1990. p. 235-244.

18. Johnston KW, Scobie TK. Multicentric prospective study of nonruptured abdominal aortic aneurysm. Population and operative management . J Vasc Surg 1988;7:69-81.

19. Sugawara Y, Takagi A, Sato O, et al. Surgical treatment of abdominal aortic aneurysms in octogenarians. Jpn Circ J 1996;60:328-33.

20. Instituto Brasileiro de Geografia e Estatística (IBGE) [site na Internet]. Brasília: Ministério do Planejamento, Orçamento e Gestão; [citado em 19 de novembro de 2004]. Brasil: tábua completa de mortalidade - ambos os sexos - 1999; [aproximadamente 5 telas]. Disponível em: http://www.ibge.gov.br/home/estatistica/populacao/tabuadevida/tabuacompleta.shtm#

21. Aune S, Laxdal E, Pedersen G, Dregelid E. Lifetime gain related to cost repair of ruptured abdominal aortic aneurysm in octogenarian. Eur J Vasc Endovas Surg 2004;27:299-304.

22. Zarins CK, White RA, Moll FL, et al. The AneuRx stent graft: four-year results and worldwide experience 2000. J Vasc Surg. 2001;33(2 Suppl):S135-45. Review. Erratum in: J Vasc Surg 2001;33:1318.

23. Sicard GA, Rubin BG, Sanchez LA, et al. Endoluminal graft repair for abdominal aortic aneurysm in high-risk patients and octogenarians: is it better than open repair? Ann Vasc Surg 2001;234:427-37.

24. Biebl M, Lau LL, Hakaim AG, et al. Midterm outcome of endovascular abdominal aortic aneurysm repair in octogenarians: a single institution's experience. J Vasc Surg 2004;40:435-42.