Multiple aortic aneurysms: The surgical management

results of

Peter Gloviczki, ME), Peter Pairolero, MD, T imothy Welch, MD, Kenne th Cherry, MD, John HaUett, MD, Barbara Toomey, R N James Naessens, M P H , Thomas Orszulak, MD, and Hartzell Schaff, MD, Rochester, Minn.

During the past 2 decades 102 consecutive patients (77 men and 25 women) with multiple aortic aneurysms underwent 201 aortic reconstructions. These procedures (174 elective and 27 emergent) represented 3.4% of the 5837 aortic aneurysm operations performed. Seventy-five (30.9%) of the 243 aneurysms occurred in the infrarenal aorta, 65 occurred in the descending aorta (26.7%), 56 occurred in the thoracoabdominal aorta (23.0%), and 47 occurred in the ascending aorta or arch (19.3%). Ages ranged from 20 to 81 years (mean 63.3 years). Smoking history and abnormal electrocardiographic tracings were present in 84.3% of the patients, hypertension was present in 77.5%, and obstructive lung disease was present in 60.8%. Multiple aortic aneurysms were present at the time of the first repair in 55 patients (53.9%). Twelve patients had one procedure, 81 had two, and nine had three. Sixteen (17.8%) of the 90 patients who had multiple operations had a subsequent operation for complications of the unrepaired aneurysm (rupture 12, symptoms 4). Fourteen perioperative deaths occurred among the 174 elective repairs (8.0%), and 11 occurred among the 27 emergent procedures (40.7%). Procedure mortality increased with the ordinal number of elective operations and was 4.4% for the first, 10.4% for the second, and 33.3% for the third. Seven of 21 patients (33.3%) who had simul- taneous repair of at least two aortic aneurysms died in the perioperative period. Overall, 77 of the 102 patients (75.5%) survived all surgical procedures to repair their multiple aortic aneurysms; of these, 63 had complete resection of all known aneurysms. Follow- up was complete in all patients and averaged 6.3 years (ranges: 1 month to 19 years). There were 30 late deaths; the most frequent cause was myocardial infarction. Kaplan- Meier 5-year survival including perioperative deaths for all patients after the first operation was 76% and after the last operation 40%. We conclude that multiple aortic aneurysms can be safely managed, usually with staged repairs, and that long-term survival is probable. After the first aortic operation the presence of multiple aneurysms mandates close ob- servation with timely surgical intervention. (J VASC SURG 1990;11:19-28.)

Aneurysmal degeneratio n of the arterial system is frequently multifocal. Multiple aortic aneurysms were present in 12.6% of 1510 patients who under- went repair of aortic aneurysms by Crawford and Cohen.~ Among 1149 patients treated for abdominal aortic aneurysms, De Bakey et al. 2 found 62 cases (4%) with aneurysm of the thoracic aorta. Epide- miologic studies suggest that aneurysm in the ab- dominal aorta will eventually develop in 25% to 30% of the patients with thoracic aortic aneurysm. 3,~

From the Section of Vascular Surgery, Division of Cardiovascular Surgery, Department of Diagnostic Radiology and Section of Biostatistics, Mayo Clinic.

Presented at the Thirty-seventh Scientific Meeting of the North American Chapter, International Society for Cardiovascular Surgery, New York, N.Y., June 19-20, 1989.

Reprint requests: Peter Gloviczki, MD, Section of Vascular Sur- gery, Mayo Clinic, Rochester, MN 55905.

~_ !~/16620

The survival of patients with aortic aneurysms has significantly improved in the last decade because of earlier diagnosis, aggressive surgical management, and improved perioperative care. However, the pres- ence of a second aortic aneurysm continues to influ- ence both early and late survival. Next to coronary artery disease the most important cause of early death after abdominal aortic aneurysm repair is rupture of a second aortic aneurysm, s Plate et al. 6 from our in- stitution reported on late follow-up of 1112 patients who underwent repair of an abdominal aortic an- eurysm. There were 25 late deaths caused by rupture of aneurysms that were either not recognized or were not present at the time of the first operation. Thus early recognition and treatment of multiple aortic aneurysms is necessary to prolong the life of these patients.

The purpose of this article is to evaluate factors

19

20 Gloviczki et al.

Journal of VASCULAR

SURGERY

3O (12.3%)

(26.7%)

56 (23%)

Table I. Frequency of associated diseases in 102 patients with multiple aortic aneurysms

Disease No. %

Smoking 86 84.3 Abnormal electrocardiographic tracings 86 84.3 Hypertension 79 77.5 Obstructive pulmonary disease 62 60.8 Peripheral vascular disease 34 33.3 Obesity 30 29.4 Angina 26 25.5 Previous myocardial infarction 20 19.6 Aortic valve insufficiency 18 17.6 Renal insufficiency 16 15.7 Stroke/transient ischemic attacks 15 14.7 Elevated cholesterol, triglycerides 13 12.7 Congestive heart failure 11 10.8 Diabetes 7 6.9

7 5 (30.9%)

Fig. 1. Incidence and location of 243 aortic aneurysms in 102 patients.

affecting early and late survival in patients who un- derwent repair of at least two aortic aneurysms.

MATERIAL AND METHODS

Between March 11, 1968, and Nov. 4, 1988, 102 consecutive patients underwent 201 aortic recon- structions for multiple aortic aneurysms. There were 77 men and 25 women whose ages ranged from 20 to 81 years (mean: 63.3 years). The 201 operations (174 elective and 27 emergent) represented 3.4% of the 5837 aortic aneurysm repairs performed during the same period. Two percent of the patients with abdominal aortic aneurysm and 18% of patients with thoracic aneurysm underwent repair for multiple aor- tic aneurysms.

The cause was degenerative atherosclerotic dis- ease in 76 patients, chronic dissection in at least one aneurysm in 23 patients, Marfan's syndrome in two, and nonspecific aortitis in one. Follow-up informa- tion was obtained in all patients and was based on clinical examinations, questionnaires, or death cer- tificates. Total follow-up encompassed 625.5 years, and averaged 6.3 years (range: 1 month to 19 years).

Location of aneurysms. Two hundred forty'- three aneurysms were diagnosed in these 102 pa- tients. The number of aneurysms per patient ranged from two to four, and the most common location was the infrarenal aorta (70 patients). Of the 243 aneurysms, 75 (30.9%) occurred in the abdominal aorta, 65 in the descending aorta (26.7%), and 56 in the thoracoabdominal aorta (23.0%) (Fig. i)~ Twenty-five different patterns of aortic involvement were identified. The most frequent combinations in- volved the descending thoracic and infrarenal aortas (25 patients) and the thoracoabdominal and infra- renal aortas (23 patients).

Risk factors. A smoking history and abnor- mal electrocardiographic tracings were each present in 84.3% of patients, hypertension was present in 77.5%, and obstructive lung disease was present in 60.8% (Table I). Only eight patients had neither hypertension nor evidence of heart disease. Ten pa- tients underwent coronary artery bypass grafting, e- fore aneurysm repair, and nine patients had coronary revascularization at the time of repair of an ascending aortic aneurysm. Fifteen patients had transient iscl-f: emic attacks or previous stroke, and five of these underwent carotid endarterectomy before repair of the aneurysms. Sixteen patients had renal insuffi;. ciency with a serum creatinine greater than 2.0 mg/dl. Seven patients underwent prior repair of pe- ripheral aneurysms (lilac one, femoral four, poplite~ twO).

Method of detection. The most frequent method of confirming the diagnosis of an abdominal aortic aneurysm was ultrasonography (76 %). Aortography was performed in 64% of the patients, and com- puted tomography (CT) scanning was performed ill 48% of patients with abdominal aortic aneurysms. The most common examination to detect tho~<~ or thoracoabdominal aneurysms was routine chest

Volume 11 Number 1 January 1990 Multiple aortic aneurysms 21

Table II. Summary of 201 procedures for multiple aortic aneurysm in 102 patients

Site of aortic aneurysm First operation Second operation Third operation Total

Staged repair Ascending 18 6 2 26 Arch 4 3 - - 7 Descending 21 20 2 43 Thoracoabdominal 6 31 3 40 Abdominal 37 25 2 64

Subtotal 86 85 9 180 Simultaneous repair

Ascending & arch 2 1 - - 3 Descending (2) 5 - - - - 5 Arch & descending 1 1 - - 2 Ascending, arch & descending 1 - - - - 1 Thoracoabdominai & abdominal 4 2 - - 6 Arch & thoracoabdominal 1 1 - - 2 Descending and abdominal 2 - - - - 2

Subtotal 16 5 0 21 Total 10--2 9---6 9 201

roe~tgenography (65%), followed by aortography (58%), CT scanning (49%), and ultrasonography (17%). Mean diameter of all aortic aneurysms mea- sured preoperatively by any of these methods was 6.2 cm (range: 3 to 15 cm). The presence of multiple aortic aneurysms was recognized before the first op- eration in 55 patients. In 47 patients a second or third aneurysm was subsequently discovered, either during routine clinical evaluation or when the patient presented with signs and symptoms of expansion or ~'upture. The mean time from repair of the first an- eurysm to the presentation of a second aneurysm in these 47 patients was 5.19 years (standard deviation, 4.32 years), with 25% presented by 1.59 years, 50% by 3.83 years, and 75% by 7.6 years, with a range from 68 days to 17.8 years. The mean time interval between repair of an abdominal aortic aneurysm and aF. ~arance of a new thoracic aneurysm was 6.34 years (standard deviation, 4.7), whereas it was only 2.7 years (standard deviation, 1.7) between the repair of a thoracic aortic aneurysm and the appearance of an abdominal aortic aneurysm (p = NS).

Surgical treatment. All patients underwent re- pair of at least two aortic aneurysms, and 85 had all known aortic aneurysms repaired. Two hundred twenty-three of the 243 known aortic aneurysms (91%) were repaired during 201 operations. Twelve patients had one procedure, 81 had two, and nine ~-~ad three. There were 21 patients who underwent simultaneous repair of at least two aortic aneurysms (Table II). Of the 90 patients who had two or three operations, 16 (17.8%) had a later operation for complication of an unrepaired aneurysm (rupture 12, symptoms 4).

• There were 27 emergency procedures, 15 for rup- tured aneurysms, 11 for pain, and one for distal em-

bolization with leg ischemia. The diameter of the smallest infrarenal abdominal aneurysm which rup- tured was 5.6 cm. There were three ruptured de- scending thoracic aneurysms (4 cm, 4 cm, and 3.8 cm). One of the 4 cm aneurysms ruptured 2 days after repair of an abdominal aortic aneurysm. Among the six ruptured thoracoabdominal aneurysms there was one small aneurysm (4 cm).

Surgical technique has continuously evolved in the time span of this study. Aneurysms of the as- cending aorta and aortic arch have been repaired with extracorporeal circulation and hypothermia. In the past several years circulatory arrest has been used routinely for arch repair. Graft inclusion according to Crawford et al.7 was used in 96% of the patients with descending thoracic and thoracoabdominal an- eurysm repairs. Gott shunt, left atriofemoral bypass grafting, or left pulmonary vein-femoral artery bypass grafting with a magnetically coupled mechanical pump, Model 520D (Biomedicus, Eden Prairie, Minn.) was used in 15 patients with descending tho- racic or thoracoabdominal aneurysms. Thirteen pa- tients had continuous intraoperative spinal fluid pres- sure monitoring; in these patients spinal fluid was drained to maintain pressure less than 15 mm Hg.

STATISTICAL METHODS

Standard descriptive statistical methods were used. Association between early mortality and risk factors was assessed univariately by the two-sample t test, Wilcoxon rank sum test, or Chi-square test for continuous, ordinal, and dichotomous risk factors, respectively. Multivariate assessment of factors relat- ing to early mortality was based on a forward step- wise logistic regression model. These assessments were performed for the second aneurysm repair to

22 Gloviczki et al.

Journal of VASCULAR

SURGERY

Table III. Perioperative mortality of 102 patients undergoing 201 procedures for repair of multiple aortic aneurysms

Elective cases Emergency cases Total

Perioporative Perioperative Perioperative deaths deaths deaths

Site of aortic aneurysm repaired No. * No. % No. * No. % No* No. %

Staged repair Ascending 26 3 11.5 0 - - - - 26 3 11.5 Arch 7 1 14.3 0 - - - - 7 1 14.3 Descending 37 2 5.4 6 2 33.3 43 4 9.3 Thoracoabdominal 32 2 6.3 8 6 75.0 40 8 20.0 Abdominal 53 0 0 11 2 18.2 64 2 3.1

Subtotal 155 8 5.2 25 10 40.0 180 18 10.0 Simultaneous repair

Ascending and arch 3 0 0 0 - - - - 3 0 0 Decending (2) 5 1 20.0 0 - - - - 5 1 20.0 Arch and descending 2 2 100.0 0 - - - - 2 2 100.0 Ascending, arch & descending 1 1 100.0 0 - - - - 1 1 100.0 Thoracoabdominal & abdominal 4 0 0 2 1 50.0 6 1 16.7 Arch & thoracoabdominal 2 1 50.0 0 - - - - 2 1 50.0 Descending & abdominal 2 1 50.0 0 - - - - 2 i 50~

Subtotal 19 6 31.6 2 1 50.0 21 7 33~.~ Totals 174 14 8.0 27 11 40.7 201 25 12.4

*Number of procedures.

Table IV. Association between potential risk factors and perioperative death in 102 patients with multiple aortic aneurysms

Elective Emergency Total

Perioperative Perioperative Perioperative deaths deaths deaths

Risk factor No.~ No. % No.~ No. % No.~ No. %

Age: 0-64 23 3 13.0 5 2 40.0 28 65-74 43 7 16.3 13 3 23.1 56 75 + 14 3 21.4 4 2 50.0 18

Sex: Men 60 10 16.7 17 6 35.3 77 Women 20 3 15.0 5 1 20.0 25

Ever smoked 70 12 17.1 16 5 31.3 86 Hypertension 6 ! 11 18.0 18 7 38.9 79 History of heart disease 36 5 13.9 9 4 44.4 45 Abnormal ECG 70 11 15.7 16 5 31.2 86 CVA/TIA 11 4 36.4 4 1 25.0 15 Claudication / PVD 28 4 14.3 6 4 66.7* 34 COPD 48 8 16.7 14 6 42.9 62 Renal insufficiency 12 2 16.7 4 2 50.0 16 Obesity 22 4 18.2 8 4 50.0 30 Aortic insufficiency 17 4 23.5 1 0 0.0 18 Rupture at either repair 3 1 33.3 15 7 46.7* 18 Asymptomatic at both repairs 25 2 8.0 - - - - - - 25 Days between repair:

0-60 23 5 21.7 3 1 33.3 26 61-365 18 1 5.6 4 3 75.0 22 365 + 39 7 18.0 15 3 20.0 54

Cross-clamp time of thoracic or thoracoabdominal repair not noted

7 2 28.6 4 3 75.0 11 0-40 minutes 15 1 6.7* 6 0 0 21 41 + minutes 38 10 26.3* 9 3 33.3 47

5 17.9 10 17.9 5 27.8

16 20.8 4 16.0

17 19.8 18 22.8

9 20.0 16 18.6 5 33.3 8 23.5

14 22.6 4 25.0 8 26.7 4 22.2 8 44.4* 2 8.0

6 4

10

5 1

13

23.1 18.2 18.5

45.5 4.8

27.7**

CVA, Cerebrovascular accident; PVD, peripheral vascular disease; COPD, chronic *p < 0.05. **p < 0.005. tNo. of patients.

obstructive pulmonary disease.

Volume 11 Number 1 January 1990 Multiple aortic aneurysms 23

Table V. Major perioperative complications of 201 procedures in 102 patients

Emergent Elective Total (n* = 27) (n* = 174) (n* = 201)

No. % No. % No. %

Death 11 Respiratory failure tracheostomy 7

1 Myocardial infarction 4 Reoperation for bleeding 5 Renal failure requiring dialysis 1 Paraplegia 0 Paraparesis 1 Cerebrovascular accident 0 Cerebral anoxia/coma 2 Sepsis 2 Intestinal obstruction 0 DVT 1

(40.7) 14 (8.0) 25 (12.4) (25.9) 15 (8.6) 22 (10.9) (3.7) 4 (2.3) 5 (2.5)

(14.8) 10 (5.7) 14 (7.0) (18.5) 5 (2.9) 10 (5.0) (3.7) 5 (2.9) 6 (3.0)

3 (1.7) 3 (1.5) (3.7) 6 (3.4) 7 (3.5)

2 (1.1) 2 (1.0) (7.4) 2 (1.1) 4 (2.0) (7.4) 5 (2.9) 7 (3.5)

2 (1.1) z (1.0) (3.7) 0 1 (0.5)

DVT, Deep venous thrombosis. *n, number of procedures

avoid bias (e.g., only patients surviving the first op- eration were eligible to have a second repair). Long- term survival was described with Kaplan-Meier methods, s All tests were two-tailed. Tests were con- sidered significant a tp < 0.05.

RESULTS Early mortality and morbidity

Fourteen perioperative deaths occurred among the 174 elective repairs (procedure mortality 8 %) and 11 among the 27 emergency repairs (procedure mor- tality 40.7%) (Table III). Procedure mortality in- creased with the number of elective operations and was 4.4% for the first, 10.4% for the second, and 33.3% for the third operation. :~'Of the 21 patients who had concurrent repair of

at least two aortic aneurysms, the procedure mortality was 33%. Early mortality was significantly higher (p < 0.05) when descending thoracic or thoracoab- dominal aneurysms were simultaneously repaired with aneurysms of the ascending aorta or arch. Four out of five patients who had simultaneous repair of either ascending or arch aneurysm and either de- scending or thoracoabdominal aneurysm died within 30 days. However, no early deaths occurred after concurrent repair of ascending and arch aneurysms (three patients) or after simultaneous elective repairs of thoracoabdominal and infrarenal aneurysms (four patients). Overall, 77 of the 102 patients (75.5%) survived all surgical procedures to repair their mul- tiple aortic aneurysms. Of these 77 patients, 63 had complete resection of all known aneurysms.

The most frequent cause of early death was peri- operative bleeding (11 patients) followed by myo-

cardial infarction (10 patients). Four of the nine pa- tients who underwent coronary revascularization si- multaneously with ascending aortic aneurysm repair died in the perioperative period. Association of potential risk factors and patient mortality within 30 days of repair of the second aortic aneurysm is listed in Table IV. Ruptured aneurysm at any level (p < 0.025) and cross-clamp time over 40 minutes during repair of thoracic or thoracoabdominal an- eurysm (p < 0.005) significantly decreased early sur- vival. Peripheral vascular occlusive disease was more frequent (p < 0.035) among those who died early after emergency repair.

The most frequent major perioperative compli- cations were respiratory failure and myocardial in- farction (Table V). Ten patients needed reoperation for bleeding; six required postoperative dialysis. Ten of the 201 procedures resulted in permanent neu- rologic deficit as a result of spinal cord ischemia. The risk of paraplegia and paraparesis after emergency operation was 3.7% (1/27) and after elective repair 5.2% (9/174). However, early death often precluded complete neurologic examination. One of the 13 pa- tients who underwent spinal fluid drainage had para- plegia. This patient underwent replacement of the entire thoracic and upper abdominal aorta for a chronic dissecting aneurysm. Paraplegia developed despite reimplantation of two separate sets of inter- costal arteries in addition to spinal fluid drainage and left atriofemoral bypass grafting.

Late survival

There were 30 late deaths; the most frequent causes were myocardial infarction and cancer. The

24 Gloviczki et al.

Journal of VASCULAR

SURGERY

100

(23)

s° f I . . . - - 60 (11)

4 0 ~

f" 20

0 i i i i 0 1 2 3 4

Years C,G 120442B-4C

Staged repair Simultaneous repair

(17)

(6)

(2)

I I I I 5 6 7 8

Fig. 2. Probability of overall survival of 39 patients undergoing staged aortic repair for multiple aortic aneurysms as compared to probability of survival of 16 patients with simultaneous repair. Numbers in parentheses represent patients entering the various observation periods.

cause of death in two patients, however, was rupture of an unrepaired third aortic aneurysm. Fig. 2 dem- onstrates overall survival of 55 patients who had two or more aortic aneurysms at the first operation. Si- multaneous repair of the aneurysms was performed in 16 patients, whereas 39 had staged repairs. Three- and 5-year survival was 75% and 53%, respectively, after simultaneous repair, and 80% and 72%, re- spectively, after staged repair (p -- NS).

Five-year survival including perioperative deaths for all patients after the first operation was 76% (Fig. 3). Five-year survival after all operations for multiple aortic aneurysms was 40% (Fig. 4). Overall survival after operations for multiple aortic aneurysms was decreased (p < 0.001) when compared to a group of 1112 patients who underwent abdominal aortic aneurysm repair (Fig. 5). 9

DISCUSSION In a classic study by Szilagyi et al.l° 5-year survival

of 223 patients with untreated abdominal aortic an- eurysm was only 17.2%. Late survival with thoracic aneurysm without surgical treatment appears to be even worse (13%). 3 Today elective repair of abdom- inal and thoracic aortic aneurysms can be performed with a mortality of less than 3% 9 and 10%, 4'7'11 re- spectively. Late survival of these patients is between 50% and 65% at 5 years, and is limited primarily by cardiac disease. Survival of patients with multiple aor-

tic aneurysms, even those with complete repair, ap- pears to be less. In the study by Crawford and Cohen, 1 5-year survival after complete repair of mul- tiple aortic aneurysms was 48%, similar to the 40% observed in the present review.

After coronary artery disease, the main reason for increased mortality after repair of an aortic aneurysm is the risk of rupture of a second aortic aneurysm. Emergency procedures to treat the symptomatic or ruptured aneurysms have even greater morbidity and mortality. On the other hand, even if all aneurysms are repaired during elective operations, additional risks still are presentl If all aneurysms are repai'~d simultaneously, the operation is more extensive, and certain combinations of aneurysm repair (ascending aorta or arch with descending or thoracoabdominal aneurysms) have particularly high mortality. If repair is staged, mortality of the first procedure may be less, but there is the cumulative risk of a second major operation plus the risk of rupture while awaiting the second operation. Procedure mortality also increases with the number of elective operations.

If staged repair is planned, there is a definite risk that the second aneurysm will rupture during the interval between repairs. Among 76 patients who underwent partial resection for multiple aortic an- eurysms, Crawford and Cohen 1 reported 18 deaths (24%) as a result of rupture of the unrcpaired aortic aneurysm. Six of these ruptures occurred early after

Voltmae 11 Number 1 January 1990 Multiple aortic aneurysms 25

100

8O

. - - 6 0 ¢0

P4o CO

20

(35) Expected

MAA

0 I I I I I I I I I I I I I

o 1 2 a 4 5 6 7 8 9 lo 11 12 13

Years CG 120442B- 1C

MAA-- Multiple Aortic Aneurysms

Fig. 3. Probability of overall survival of 102 patients with multiple aortic aneurysms after the first aortic repair as compared to expected survival adjusted for age and sex. Numbers in parentheses represent patients entering the various observation periods.

the first operation. In our review among 55 patients who had at least two aneurysms at the first presen- tation, four patients had aneurysm rupture after re- pair of the first aneurysm. One of these, a 4 cm de- scending thoracic aneurysm, ruptured in the imme- diate postoperative period 2 days after infrarenal aortic aneurysm repair.

The mulfifocal nature of aneurysm involvement of the aorta mandates thorough preoperative evalu- ation of any patient with aortic aneurysm at any level. Similarly, close postoperative surveillance with search for another aortic aneurysm is indicated in patients who undergo repair of an aortic aneurysm. At present ~it is difficult to predict the time interval to the de- vc" )pment of a second aortic aneurysm. The mean time interval from repair of an aortic aneurysm until the discovery of a second aortic aneurysm in our study was 5.19 years. It appears that an abdominal aortic aneurysm may develop earlier (2.76 years) after ,repair of a thoracic aneurysm, than a thoracic an- eurysm developing after abdominal aortic aneurysm repair (6.34 years). However, to accurately predict the probability of the development of a second aortic aneurysm, information on two groups of patients not included in this study will also be needed. These are patients in whom a second aneurysm developed but they did not have it repaired and those who never developed a second aortic aneurysm. However, there are clues to suggest the presence of multiple aortic aneurysms. Both dissection and Marfan's syndrome often involve several sections of the thoracic or thor-

acoabdominal aorta. The most frequent cause of mul- tiple thoracic aneurysms in young patients is Marfan's syndrome. 1'12 Two young patients in this study had Marfan's syndrome with involvement of the ascend- ing and descending thoracic aorta in one and aortic arch and thoracoabdominal aorta in the other.

The probability of multiple aortic aneurysms de- veloping appears to be higher in those patients who first present with thoracic aneurysm. Bickerstaff et alY found an associated abdominal aortic aneurysm in 25% of the cases with thoracic aneurysm. In Craw- ford and Cohen's study, 68% of patients with de- scending thoracic aortic aneurysms had multiple aor- tic aneurysms as compared to only 12% in patients with abdominal aortic aneurysms. 1 Of 5837 aortic aneurysm repairs that we have performed during the 2 decades, 2% of the patients with abdominal aortic aneurysm and 18% of the patients with thoracic aneurysm underwent repair for multiple aortic an- eurysms.

Our current initial evaluation of patients with thoracic or thoracoabdominal aneurysms includes CT scanning of the chest and abdomen. If no in- frarenal aneurysm is found initially, annual follow- up with less expensive ultrasound examination is mandatory, considering the high chances of subse- quent development of abdominal aortic aneurysm. In addition, periodic CT scanning is used to evaluate subsequent aneurysm formation in the thorax in those patients who have had repair of thoracic or thoracoabdominal aneurysms.

26 Gloviczki et al.

Journal of VASCULAR

SURGERY

IO0

8O

- . - '60

CO 40

20

0

I !

| '--"-I

(43)

83% |

LI• %

08)

Expected

MAA (7)

I I I I I I I

0 1 2 3 4 5 6 7

Years MAA= Multiple Aortic Aneurysms

@ CG120442B-2C

Fig. 4. Probability of overall survival of 102 patients after all aortic procedures as compared to expected survival adjusted for age and sex. Numbers in parentheses represent patients entering the various observation periods.

100

80 I-

~ 6 0

40

(4a)

(705)

(517)

I (18) "1-1.- M A A '--t--I-- AAA* 20 (7)

0 0 1 2 3 4 5 6 7 8 9 10 11 12 13

Years MAA = Multiple Aortic Aneurysms AAA = Abdominal Aortic Aneurysm

@ CG 120442B-8C

Fig. 5. Probability of overall survival of 1112 patients after abdominal aortic aneurysm repair as compared to survival of 102 patients after repair of multiple aortic aneurysms. Numbers in parentheses represent patients entering the various observation periods. (*From HoUier et al. J VASC SuRe 1984;1:290-9.)

Patients with abdominal aortic aneurysms should undergo CT scanning to evaluate the entire aorta if there is any suspicion of thoracic aneurysm on the chest roentgenogram, if the patient is young, or if there is evidence of dissection or suprarenal aneurysm

on the abdominal scan, ultrasound, or aortogram. At the present, annual chest roentgenograms are sug- gested in those patients who undergo repair of an abdominal aortic aneurysm. Any abnormality on the chest roentgenogram or any symptom suggesti~d

Volume I i Number 1 January 1990 Multiple aortic aneurysms 27

aneurysmal disease is fur ther evaluated wi th C T scanning .

Planning surgical t rea tment for mult iple aortic aneurysms is influenced by location, diameter , and signs and symptoms o f the aortic aneurysms. For patients w h o have no symptoms and elective repair Is planned, ou r current approach, in general, is to stage the operat ions if aneurysm o f the ascending aor ta or arch is associated wi th descending thoracic or thoracoabdomina l aneurysm. Howeve r , simulta- neous repair is pe r fo rmed for ascending and arch aneurysms, for thoracoabdomina l and infrarenal an- eurysms, and for suprarenal and infrarenal abdominal aortic aneurysms. The m a n a g e m e n t o f descending thorac ic and infrarenal aortic aneurysms is individ- ualized, and the d iameter and symptoms o f the an- eurysms may dictate the sequence o f repair. M o r e proximal aneurysms are usually t reated first. O u r data suv~est that mult iple aortic aneurysms can be safely managed , usually wi th staged aortic operat ions, and t h a t long- te rm survival is possible. After repair o f the first aneurysm, the presence o f mult iple aortic an- eurysms mandates close observat ion wi th t imely sur- gical intervention.

REFERENCES 1. Crawford ES, Cohen ES. Aortic aneurysm: a multifocal dis-

ease. Arch Surg 1982; 117:1393-400. 2. De Bakey ME, Crawford ES, Cooley DA, Morris GC, Roys-

ter TS, Abbott WPI Aneurysm of abdominal aorta. Analysis

of results of graft replacement therapy one to eleven years after operation. Ann Surg 1964;160:622-39.

3. Bickerstaff LK, Pairolero PC, Hollier LH, et al. Thoracic aortic aneurysms: a population-based study. Surgery 1982; 92:1103-8.

4. Pressler V, McNamara JJ. Aneurysm of the thoracic aorta. Review of 260 cases. J Thorac Cardiovasc Surg 1985;89: 50-4.

5. Crawford ES, Saleh SA, Babb JW, Glaeser DH, Vaccaro PS, Silvers A. Infrarenal abdominal aortic aneurysm, Factors in- fluencing survival after operation performed over a 25-year period. Ann Surg 1981;193:699-709.

6. Plate G, HoUier LH, O'Brien P, Pairolero PC, Cherry KJ, Kazmier FJ. Recurrent aneurysms and late vascular compli- cations following repair of abdominal aortic aneurysms. Arch Surg 1985;120:590-4.

7. Crawford ES, Crawford JL, Sail HJ, et al. Thoracoabdominal aortic aneurysms: preoperative and intraoperative factors de- termining immediate and long-term results of operations in 605 patients. J VASC SURG 1986;3:389-404.

8. Kaplan EL, Meier P. Nonparametric estimation from incom- plete observations. J Am Stat Assoc 1958;53:457-81.

9. Hollier LH, Plate G, O'Brien PC, et al. Late survival after abdominal aortic aneurysm repair: influence of coronary ar- tery disease. J VASC SURG 1984;1:290--9.

10. Szilagyi DE, Smith RF, DeRusso FJ, EUiott JP, Sherrin FW. Contribution of abdominal aortic aneurysmectomy to pro- longation of life. Ann Surg 1966;164:678-99.

11. Hollier LH, Symmonds j'B, Palrolero PC, Cherry KJ, Hallett JW, Gloviczki P. Thoracoabdominal aortic aneurysm repair. Analysis of postoperative morbidity. Arch Surg 1988;123: 871-5.

12. Mohr R, Adar R, Rubinstein Z. Multiple aortic aneurysms in Marfan's syndrome. Case report and review of the litera- ture. J Cardiovasc Surg [Torino] 1984;25:566-70.

E', 3 C U S S I O N

Dr. James C. Stanley (Ann Arbor, Mich.). Patients with extensive aortic aneurysmal disease represent a ther- apeutic challenge to most vascular surgeons. My comments and questions are directed to three aspects of this w o r k - - "the classification, etiology, and treatment o f these an- eurysms.

The first relates to your use of the word multiple. It is important to understand whether you are referring to multiple discrete aneurysms or multiple levels of aortic in- volvement by aneurysmal disease. Given your five desig- nated sites of disease--the ascending aorta, the transverse arch, the descending thoracic aorta, the paradiaphragmatic thoracic and proximal abdominal aorta, and last, the in- frarenal aorta-- there could be a total of 25 combi- nations of multiplicity regarding level o f involvement. One presumes that you would classify a diffusely aneu- rysmal aorta--extending from the aortic annulus to the , .~:urcation--as representative of all five levels o f involve-

ment, yet as such it would not represent five discrete aneurysms. Thus the word multiple must not be miscon- strued to mean different lesions occurring as separate an- eurysms, except when an intervening level of the aorta is not aneurysmal. This is not a minor issue in contemporary reporting when many authors apply different definitions to multiple aneurysms, aneurysmosis, arteriomegaly, and aneurysmal ectasia.

A second issue relates to etiology. The underlying pathogenesis of most o f these aneurysms is probably not arteriosclerosis. That statement may sound like heresy, but arteriosclerosis in these cases is more likely a secondary event occurring after an as of yet not well understood process affecting the aortic elastic tissue. Certain of these cases may represent a genetically transmitted disorder. In the Mayo Clinic series, the 6% of their cases with discrete multiple aneurysms affecting the aorta as well as the femoral and popliteal vessels are most apt to reflect a very strong

28 Gloviczki et al.

Journal of VASCULAR

SURGERY

male phenotypic expression of a genetic defect. In a series of 172 femoral aneurysms affecting 100 patients at our institution, all occurred in men. We too called these an- eurysms arteriosclerotic, but this is more reflective of their eventual histopathology than true etiology.

The role o f smoking-activated neutrophil elastase ac- tivity, and perhaps a more subtle defect in elastin, as might have been evident in the 61% of your patients with ob- structive lung disease, deserves note. Although recent stud- ies have not related such to aortic aneurysmal expansion, earlier reports suggested a potential etiologic role for these factors. In your experience did these two factors exist as independent or codependent predictors of aneurysmal fate? Such might have been important in the 46% of your patients who had aneurysmal disease appear at a second level over a relatively short follow-up period, or in the 18% of your cases who experienced symptomatic expansion or rupture of their aneurysms.

Last is the issue of treatment. What role did size assume in decisions regarding simultaneous repair of descending, thoracic, and infrarenal aortic aneurysms. Would you treat the descending thoracic component of extensive disease if the aorta at this level were 4 cm in diameter? The 15 pa- tients who underwent left atriofemoral bypass surgery or left pulmonary artery-femoral bypass surgery are notewor- thy. Was this approach undertaken sporadically throughout your experience, or do you advocate it selectively?

Dr. Gloviczki. Thank you, Dr. Stanley, for discussing our paper. In response to your question concerning clas- sification of this disease, I entirely agree that what we are talking about is a multilevel manifestation of the same an- eurysmal disease; so I do not think that these are different aneurysms. I think these are aneurysms of similar origin,

although there are intervening aortic segments, which may be of normal size.

As far as the cause of the disease is concerned, we do not have sufficient data to prove that there is any genetic cause present in our patients; we found smoking in a very high number of these patients and that really did not differ at all, considering other articles about the incidence of smoking. We were impressed by the high incidence of obstructive pulmonary disease, 62%. In one of our pre- vious articles on abdominal aortic aneurysms, the incidence of obstructive pulmonary disease was only 40%, and I know that in Dr. Crawford's last series the incidence of obstructive pulmonary disease with thoracoabdominal aor- tic aneurysms was around 30%.

As far as the question concerning the size, it does not appear to be a major factor in the thoracic aorta. The reason I say that is because I think an aneurysm of the thoracic aorta is deadly. An epidemiologic study from our institu- tion showed that 5-year survival of patients with thoracic aneurysms is only 13%. In our present study, four of nine ruptured aneurysms at the thoracic or thoracoabdor~"~al level measured 4 cm or less. Although this is a selected group of patients, our feeling is, that in a patient who is a good risk who has a 4 cm thoracic aneurysm, repair should be considered.

In recent years we have selectively used left pulmonary vein-femoral artery or left atrial-femoral artery bypass graft- ing with a centrifugal pump in several patients with tho- racic or thoracoabdominal aneurysms. In a small group of 35 patients, we have collected some data that the centrif- ugal pump may provide better protection than a passive shunt. However, we have no evidence that any of these shunts eliminates the risk of paraplegia in these patients.