- Email: [email protected]
Familial thoracic aortic dilatations and dissections: A case control study
Familial thoracic aortic dilatations and dissections: A case control study Alan Biddinger, M S E , M a r n i e Rocklin, M S , J o s e p h Coselli, M D , and D i a n n a M. Milewicz, M D , P h D , Houston, Tex.
Purpose: Evidence suggesting that genetic factors contribute to the development of common disorders can be obtained by demonstrating familial aggregation of the disease. This study investigated whether thoracic aortic dilations and dissections aggregate in families by comparing the prevalence o f thoracic aortic aneurysms, thoracic aortic dissections, and sudden death in first-degree relatives of patients referred for thoracic aortic surgery. Methods: Families were ascertained through 158 nonsyndromic patients referred for surgical correction of either thoracic aortic aneurysms or dissections (probands) and their 843 first-degree relatives. A control group of 547 first-degree relatives was derived from 114 proband spouses. Groups were examined for statistical differences in the prevalence of thoracic aneurysms, thoracic aortic dissections, abdominal aortic aneurysms, sudden death, and myocardial infarctions. Results: First-degree relatives of probands demonstrated a higher prevalence of thoracic aortic aneurysms and sudden death when compared with the control group. Relative risks of thoracic aortic aneurysm development in proband fathers, brothers, and sisters were 1.8, 10.9, and 1.8, respectively. A pattern of inheritance of the thoracic aortic aneurysms could not be determined. Conclusions: This study indicates proband first-degree relatives are at higher risk o f thoracic aortic aneurysms and sudden death compared with a control group. This study supports the role of genetic factors-in the cause of thoracic aortic aneur;ysms and provides important information for identifying individuals at risk. (J Vase Surg 1997;25:506-11.) ;
Aneurysms and dissections o f the aorta are assod a t e d with a high degree o f morbidity, mortality, and medical expenditure despite continued improvements in diagnostic and surgical techniques. 1,2 Prevention o f these disorders through the early identification of predisposed individuals and the modification o f contributing environmental and genetic factors is a potential cost-effective m e t h o d for addressing these diseases. Establishing familial aggregation and estimating relative risks o f these disorders for family From the Department of Internal Medicine, University of Texas Houston Medical School, and Baylor College of Medicine. Supported in part by a Pfizer Scholars Award, an American Heart Association grant-in-aid (93014300), a Schissler Foundation Fellowship, and an NIH Clinical Research Center Grant (MOIRR-02558). Presented as a Poster (45I) at the Forty-sixthAnnual Meeting of the American Society for Human Generics in Minneapolis, Minn., Oct. 24-28, 1995. Reprint requests: Dianna M. Milewicz,MD, PhD, Department of Internal Medicine, University of Texas-Houston Medical SchooI, MSB 1.614, 6431 Fannin, Houston, TX 77030. Copyright © t997 by The Societyfor Vascular Surgeryand International Society for Cardiovascular Surgery, North American Chapter. 0741-5214/97/$5.00 + 0 24/1/76003 506
members is useful not only for identifying susceptible individuals but also for implicating a genetic contribution to these disorders. Although abdominal aortic aneurysms have been well characterized in regard to familial aggregation, 3 risk factors, 4,s possible canses, 6,7 and potential modes o f inheritance, 8,9 less is known regarding thoracic aortic aneurysms (TAA) and thoracic aortic dissections. 2,1°-12 For this reason this study was primarily designed to c o m p a r e firstdegree relatives o f patients referred for surgical repair o f dilated or dissected thoracic aortas with a control g r o u p for differences in occurrence o f these disorders. Because rupture or dissection o f asymptomatic aneurysms has a high mortality rate, ~ this study also examined the prevalence o f sudden deaths in the two groups. O n e h u n d r e d fifty-eight patients were inter¢iewed, and family histories were obtained. TAA prevalence, age and sex distributions, disease, familial aggregation, and potential m o d e s o f inheritance are reported. This study d e m o n s t r a t e s familial aggregation o f TAA and sudden death as a first step toward identifying individuals with potential genetic predispositions for these disorders.
IOURNAL OF VASCULAR SURGERY Volume 25, Number 3
METHODS P o p u l a t i o n selection. The study population consisted o f first-degree relatives o f 158 patients (100 men and 58 women) who were referred to Methodist Hospital, Houston, Texas, for surgical repair o f TAA, thoracic aortic dissections, or thoracoabdominal aortic aneurysms. These individuals underwent corrective surgery within the 21/2-year period of September 1990 to March 1993. Patients excluded from the study had either known syndromes (e.g., Marfan syndrome) or vasculitis associated with thoracic aorta anomalies or were unaware o f their first-degree relatives' medical history. There were 153 white, 3 black, 1 Hispanic, and 1 Asian proband patients. A control population of first-degree relatives was obtained by interviewing 114 of the probands' spouses about their family history. This study recorded occurrence of symptomatic or previously diagnosed thoracic and abdominal aneurysms, dissection o f the thoracic aorta, myocardial infarction, and sudden death. Only relatives whose age placed them at risk for symptomatic aneurysm development were included in the study. A minimum age o f 30 years was selected based on the age of the youngest proband patient. Excluding the 308 offspring common to both groups, there were 843 proband and 547 control first-deg~'ee family members. Data collection. After appropriate consent as required by the Institutional Review Board was obtained, one o f the authors (M. R.) interviewed the study subjects with a standardized questionnaire designed to maintain consistency and minimize recall bias in the data collection process. Medical records were reviewed for aortogram and surgical pathology results. Aneurysms in the ascending thoracic aorta, aortic arch, or descending aorta were classified as thoracic aortic aneurysms (TAA). Aneurysms below the diaphragm but superior to the renal arteries were classified as thoracoabdominal aortic aneurysms (TAA). Infrarenal aneurysms were classified as abdominal aortic aneurysms (AAA). Aortic dissections were classified according to the DeBakey system) 4 The predominant disease was reported as either cystic medial necrosis or atherosclerosis. Sudden death was defined as an unexpected death o f unlmown cause in the absence o f a history o f heart disease and subsequent autopsy. In reported cases o f thoracic, abdominal, or cerebral aneurysms in first-degree relatives, medical, surgical, and radiologic records were obtained whenever possible to corroborate elicited medical histories. Statistical analysis. Statistical analysis was performed with an SPSS statistical software package
Biddinger et al.
507
(SPSS, Inc., Chicago, Ill.). Tripartate Logistic Regression was used to verify that dilated and dissected thoracic aorta data could be combined for subsequent statistical analysis. 1S Mann-Whimey and 2 × 2 ×2 analyses were used to test for age and sex differences between the study and control groups, respectively. One-way analyses o f variance were used to compare ages among multiple groups. Log linear analysis was used to test for associations among the multiple categoric variables. Fisher's exact test was used to analyze 2 × 2 tables with multiple data entries smaller than five. Overall relative risk estimates for TAA and sudden death were computed with the Mantel-Haenszel estimator.16 Separate TAA relative risk estimates for proband parents and siblings were calculated as described by Webster et al)7 with TAA cumulative incidence functions ~8 derived from data published by Bickerstaff et al.19 and tables given in Bailer and Ederer. 2° RESULTS P r o b a n d characteristics. Proband characteristics, summarized in Table I, were' similar to previous findings in regard to proband sex, age, aneury,;m history, aortic anomaly at the time of surgery, and disease. 1°,12,17,21 Men had an average age o f 63.8 years (SD = 11.31), and women had an average age o f 67.0 years (SD = 11.37). The ratio o f affected men to women was 1.7:1, with men demonstrating a higher prevalence o f aortic dissections (p = 0.004). There were no statistically significant sex-related differences with respect to the occurrence o f previous TAA, previous AAA, dilated thoracic or thoracoabdominal aneurysm, aortic dissection, abdominal aortic aneurysm, hypertension, or smoldng history. Cystic medial necrosis was significantly associated with a history o f TAA (p = 0.006) and aortic dissection (p = 0.012). Atherosclerosis was associated with thoracoabdominal aortic aneurysm (p < 0.0001) and AAA (p < 0.0001). No age differences were noted when sex, aneurysm type (dilated vs dissecting), and aneurysm location were examined at the p = 0.05 significance level. Finally, there was a statistically significant association between hypertension and a family history o f dilated or dissecting thoracic aneurysms (p = 0.031). Disease prevalence in first-degree relatives. From a statistical point o f view Tripartate logistic regression o f the data revealed that proband aneurysm type (dilated vs dissecting) did not make a statistically significant contribution toward predicting first-degree relative aortic disease (data not shown), indicating that TAA and dissected thoracic
JOURNAL OF VASCULARSURGERY
508
Biddinger et al.
1997
March
Table I. Summary ofproband characteristics Previous aneurysm
Men
Total Women
Age (yr)
TAA
30-39 40-49 50-59 60-69 70-79 ->80
2 2 7 12 4
30-39 40-49 50-59 60-69 70-79 ->80
Total
27 1 2 10 4 1 18
Dilated thoracic aneurysm
AAA
TAA
TAAA
Both
I
1 1 2
1 1 3 6
1 3 6 10 5
4
11
25
1 4 3 17 5 2 32 2 1 1 12 8 3 27
1 7 12 1 21
3 3 1 7
Dissecting thoracic aorta classification H
Disease
III
CMN
1
3 9 10 22 12
2 5 2 1 10
3 5 8 1 18
56
ATH 1 3 13 13 2 32 2
1 1
7 5 1 14
1
2 3 3
2 1 3
8
2 1 3
10 6 2 20
6
5 10 10 3 30
TAA, Thoracic aortic aneurysm; TAA, thoracic aortic aneurysm; TAAA, thoracoabdominal aneurysm; Both, TAA with TAAA; CMN, cystic medical necrosis; ATH, atherosclerosis.
Table II. Disease prevalence in first-degree relatives Probands (N) Event TAA or dissecting thoracic aorta AAA MI Sudden death
Control group (N)
Statisics
Men (429)
Women (414)
Men (274)
Women (273)
P
12 11 48 38
5 1 17 10
2 7 31 9
0 0 9 5
0.0096 0.82 0.78 0.0057
TAA, Thoracic aortic aneurysm; AAA, abdominal aortic aneurysm; M/, myocardial infarction.
aorta data could be combined for subsequent analysis. However, dilated thoracic anemTsms and thoracic aortic dissections are thought to have different causes, suggesting that they should be analyzed separately.2226 Because of this discrepancy we provide the first-degree relative data on all the proband patients and then an analysis separating the proband patients into two groups based on whether the proband had aortic dilation or dissection. Overview o f the combined data. Comparison ofproband first-degree relatives with the spouse control groups revealed no statistically significant differences in age or sex. The prevalence of TAA and sudden death was significantly higher in the proband patients first-degree relatives compared with the control group relatives. As shown in Table II 17 of the proband patients' relatives had TAA, whereas only 2 relatives of the control group were affected (Mantel and Haenszel X2 p = 0.0096). Proband relatives reported 48 occurrences of sudden death compared with 14 sudden deaths in the control group (Mantel
and Haenszel X2, p = 0.0057). In contrast, no significant difference was seen in the occurrence of AAA (p = 0.82) or myocardial infarctions (p = 0.78) between the proband and control groups. With regard to sex, men were more likely than women to have thoracic aortic aneurysms or dissections (p = 0.042), sudden death (p < 0.0001), myocardial infarction (p < 0.00001), and abdominal aortic aneurysms (p < 0.001). Segregated data analysis. First-degree relatives of patients in the proband and control groups were segregated into two groups based on the type of aortic disease present in the proband patient (dilated aneurysm or aortic dissection) as shown in Table III. Relatives of proband patients with dilated TAA also had a significant prevalence of dilated TAA (p = 0.026) compared with those of patients in the control group. When proband fathers, mothers, brothers, and sisters were examined separately, there was insufficient power in the Mantel-Haenszel test to obtain relative risk estimates. As an alternate ap-
JOURNAL OF VASCULARSURGERY Volume 25, Number 3
Biddinger et al.
509
Table III. Prevalence in first-degree relatives segregated by proband thoracic aorta dilation or dissection TAA probands group
TAA controlgroup
Dissecting aorta probands group
Dissecting aorta controlgroup
(N)
(N)
(N)
(N)
Statistics
Event
Men (243)
Women (254)
Men (159)
Women (153)
Men (186)
Women (160)
Men (115)
Women (120)
TAA p value
Dissecting p value
TAA Dissection AAA MI Sudden death
7 0 5 31 24
1 1 0 10 5
0 0 0 24 4
0 0 0 6 6
4 1 6 17 14
2 1 1 7 5
2 0 7 7 3
0 0 0 3 1
.026 1.0 .163 .504 .089
.484 .517 .461
.177 .022
TAA, Thoracic aortic aneurysm; Dissection, thoracic aorta dissection; A_AA, abdominal aortic aneurysm; MI, myocardial infarction; TA.A p value, values obtained when comparing TAA probands and TAA controls; Dissectingp value, values obtained when comparing dissecting aorta probands with dissecting aorta Controls.
proach the technique described by Webster et al? 7 was used to obtain conservative relative risk estimates for each of these individuals. Relative risks of dilated thoracic aneurysm development in TAA proband fathers, brothers, and sisters were 1.8 (95% CI = 1.5, 2.2), 10.9 (95% CI = 9.4, 12.7), and 1.8 (95% CI = 1.6, 2.1), respectively. No estimate of risk in proband mothers could be calculated, because none of them had TAA. No difference was found in sibling risk with regard to proband sex (p = 0.267). Furthermore no significant difference was seen in prevalence of dissecting aortic aneurysms (p = 1.0), AAA (p = 0.163), myocardial infarctions (p = 0.504), or sudden death (p = 0.089). In contrast, first-degree relatives of probands with dissecting aortic aneurysms demonstrated higher prevalence of sudden death (relative risk = 3.35; CI% = 1.13, 9.99), but no significant differences were detected in the prevalence of dilated TAA (p = 0.484), aortic dissections (p = 0.517), AAA (p = 0.461)or myocardial infarction (p = 0.177). Analysis o f p r o b a n d pedigrees. O f the 158 proband pedigree patients, 16 multiplex pedigrees contained first-degree family members with either dilated or dissecting thoracic aortic aneurysms. With respect to TAA proband patients, dilated aneurysms were seen in one father, six brothers, and one sister including the single instance of a father-son affected pair. One sister of a proband patient had aortic dissection. In contrast, none of the spouse control pedigrees contained members affected with either TAA or thoracic aortic dissection. In the eight pedigrees ascertained from proband patients with aortic dissection, one father and one sister had aortic dissection. Dilated thoracic aneurysms were seen in two fathers, one mother, two
brothers, and one sister. Two control simplex pedigrees contained fathers with dilated thoracic aneurysms. Visual inspection of these pedigrees did not suggest a probable mode of inheritance for either dilated or dissecting thoracic aortic aneurysms (data not shown). DISCUSSION
This study demonstrates familial aggregation of TAA and represents a step towards identifying individuals predisposed to have the disorder. Although it is currently believed that the cause of dilated and dissecting aneurysms is different,23-26 there are families with autosomal dominant inheritance of aortic disease in which dilated aneurysms and aortic dissections are seen together. 12,27 In the families with aortic disease identified in this study, seven families had affected members with dilated aneurysms along with other family members with aortic dissections. These facts suggested that dividing the probands into two groups based on whether they had a dilated aneurysm or aortic dissection might not be appropriate when we were attempting to determine the familial aggregation of thoracic aortic aneurysms and dissections. Therefore we have presented our data both combining and separating the probands based on whether they had a dilated aneurysm or aortic dissection. Bickerstaff et al.19 reported a TAA incidence of 5.9 per 100,000 person-years in white patients. Applying this estimate to our study population, an average of 1.9 dilated aneurysms were expected in firstdegree relatives of dilated aneurysm proband patients (32,206 person-years). A statistically significant fourfold increase in the number of dilated aneurysms was observed. Subsequent examination revealed that
510
Biddinger et al.
proband fathers and sisters had a 1.8-fold increased risk of TAA, whereas proband brothers had an ll-fold increased risk of TAA. These findings demonstrate that TAAs aggregate in families and support a sex-influenced disease process that affects proband siblings. Dissecting aneurysms were rarely reported in first-degree relatives o f probands and were never reported in control family members. As a consequence no meaningful statistical comparisons could be drawn. However, asymptomatic dissecting aneurysms have a high mortality rate 2 and in the absence of radiographic studies, surgery, or autopsy can either go undetected or can be misclassified as sudden deaths. Examination o f sudden deaths in first-degree relatives o f dissecting aneurysm proband patients revealed a statistically higher prevalence and a threefold increased risk for this disorder. In contrast, relatives o f patients with dilated aneurysm showed no increased prevalence o f sudden death. The lack of a statistically significant difference in deaths attributed to myocardial infarction also lent s u p p o r t to the hypothesis that aneurysm rupture as opposed to another cause contributed to the increased number of unexplained sudden deaths. Hypertension has long been recognized as an important risk factor for the development of aortic ancurysms and aortic dissections. 4,19 In this study 60% o f p r o b a n d patients with dilated aneurysms and 70% o f p r o b a n d patients with aortic dissections were hypertensive (data not shown). No statistically significant difference was seen in the prevalence o f elevated blood pressure in the two groups o f p r o b a n d patients (p = 0,20). Although hypertension is associated with the two disorders, the formation o f dilating or dissecting aortas cannot be ascribed to elevated blood pressure alone. Thoracic aortic aneurysms and abdominal aortic aneurysms may have heterogeneous causes based on differences in familial aggregation and disease. Although abdominal aortic aneurysms have been shown to occur as multiple events in families, 3,17,28,29 no statistically significant difference was seen in AAA occurrence in proband and control relatives in th'is study. The observation that men are more often affected than women in families with TAA aggregation contrasts with familial abdominal aortic aneurysm studies, in which women were more likely to be affected than men. 3°,a~ A final difference between TAA and AAA is related to disease. There were more cases o f cystic medial necrosis (55%) than atherosclerosis (45%) reported in pathologic sections available for review. In contrast, the vast majority o f the aortic
J O U R N A L OF VASCULAR SURGERY March 1997
disease found in previous AAA studies were atherosclerotic lesionsY a4 Inspection o f the proband pedigrees revealed the occurrence o f both dilated thoracic aneurysms and aortic dissections regardless o f the proband's aortic anomaly. The predominance o f affected male patients at first suggested X-linked inheritance for eitiler dilated or dissected aortas. However, occurrences o f father-to-son transmission in both cases excluded this mode of inheritance but supported a sex-influenced process. There were seven TAA pedigrees and three aortic dissection pedigrees where only siblings were affected with dilated TAA, supporting either autosomal recessive inheritance or autosomal dominant inheritance with incomplete penetrance. Genetic heterogeneity could account for similar TAA phenotypes arising from different genotypes and alternative modes o f transmission. O f the 158 proband patients in this study, 153 (96.8%) were white. Other studies relating to the epidemiologic characteristics o f aortic aneurysms have also reported a predominance o f affected white patients. 19,3S Although the possibility exists that white people are more likely to have TAA than people o f other ethnicity, the high prevalence o f white patients in this study may also represent a selection bias. Because TAAs are usually asymptomatic, the high rate o f white patients referred for surgery could reflect a disproportionate access to medical care with a concomitant increase in asymptomatic aneurysm detection. In either case the results o f this study should be applied only to white populations. In conclusion, first-degree relatives of proband patients referred for dilated thoracic aneurysm repair are at an increased risk for having TAA themselves. This is particularly true for male siblings o f affected probands. First-degree relatives o f patients with dissecting aneurysm have a threefold increased risk o f sudden death possibly because o f aortic dissection. Formal segregation analysis may elucidate the mode(s) o f inheritance o f TAA and thoracic aortic dissection, providing a step toward identifying underlying genetic alterations that predispose individuals to these disorders. Studies are in progress to identify specific genetic and environmental influences that contribute to TAA pathogenesis before cost-effective screening methods are recomm e n d e d for identifying family members at increased risk. We thank the patients and their spouses for participating in this study. We also thank Dr. Max Buja for his
JOURNAL OF VASCULARSURGERY Volume 25, Number 3
expertise in cardiovascular p a t h o l o g y a n d Dr. Dennis J o h n s t o n for his assistance w i t h statistics. REFERENCES
1. Lilienfeld D, Baxter I, Sprafka J. Prevalence of aortic aneurysms in the twin cities metropolitan area, 1979-84. Public Health Rep 1995;108:506-9. 2. Lilienfeld DE, Gunderson PD, Sprat'ka JM, Vargas C. Epidemilogy of aortic aneurysms: I. Mortality trends in the United States, 1951 to 1981. Arteriosclerosis 1987;7:63743. 3. Tilson MD, Seashore MR. Fifty families with abdominal aortic aneurysms in two or more first-order relatives. Am l Surg 1984;147:551-2. 4. Spittell JA. Hypertension and arterial aneurysm. J Am Coll Cardiol 1983;1:533-40. 5. Auerbach O, Garfinkel L. Atherosclerosis and aneurysm of aorta in relation to smoking habits and age. Chest I980;78: 805-9. 6. Powell J, Greenhalgh RM. Cellular, enzymatic and genetic factors in the pathogenesis of abdominal aortic aneurysms. J Vasc Surg 1989;9:297-302. 7. Tilson MD, Reilly JM, Brophy CM, Webster EL, Barnett TR. Expression and sequence of the gene for tissue inhibitor of metalloproteinases in patients with abdominal aortic aneurysms. J Vasc Surg 1993;18:266-70. 8. Reilly JM, Tilson MD. Incidence and etiology of abdominal aortic aneurysms. Surg Clin North Am 1989;69:705-11. 9. Majumder PP, St. lean PL, Ferrell RE, Webster MW, Steed DL. On the inheritance of abdominal aortic aneurysm. Am l Hum Genet 1991;48:164-70. 10. DeBakey ME, Noon GP. Aneurysms of the thoracic aorta. Mod Conc Cardiovasc Dis 1975;44:53-8. I 1. Wilson S, Hutchins G. Aortic dissecting aneurysms: causative factors in 204 subjects. Arch Pathol Lab Med 1982;106:17580. 12. Nicod P, Bloor C, Godfrey M, et al. Familial aortic dissecting aneurysm. J Am Coil Cardiol 1989;13:811-9. 13. Skupin M, Blum U, Krause E, Ruhr-Leukart M. Results of surgical repair for 110 thoracic aortic aneurysms. Thorac Cardiovasc Surg 1990;38:175-80. 14. DeBakey M, Henly W, Cooley D. Surgical management of dissecting aneurysms of the aorta. J Thorac Cardiovas Surg 1965;49:130-48. 15. Greene WH. Econometric analysis. 2nd ed. New York: MacMillan, 1992. 16. Mantel N, Haenszel W. Statistical aspects of the analysis of data from retrospective studies of disease, l Natl Cancer Inst 1959;22:719-48. 17. Webster MW, St Jean PL, Steed DL, Ferrell RE, Majumder PP. Abdominal aortic aneurysm: results of a family study. J Vasc Surg 1991;13:366-72.
Biddinger et al.
511
18. Breslow NE, Day NE. Statistical methods in cancer research: Vol I. The analysis of case-control studies. Lyon: IARC Scientific Publications No. 32, 1980:49-51. 19. Bickerstaff LK, Pairolero PC, Hollier LH, et al. Thoracic aortic aneurysms: a population-based study. Surgery 1982; 92:1103-8. 20. Bailar III lC, Ederer F. Significance factors for the ratio of a Poisson variable to its expectation. Biometrics 1964;20:63942. 21. Pressler V, McNamara I. Thoracic aortic aneurysm: natural history and treatment. J Thorac Cardiovasc Surg 1980;79: 489-98. 22. Shores J, Berger KR, Murphy EA, Pyeritz RE. Progression of aortic dilatation and the benefit of long-term [3-adrenergic blockade in Marfan's syndrome. N Engl }- Med 1994;330: 1335-41. 23. Nakashima Y, Shiokawa Y, Sueishi K. Alterations of elastic architecture in human aortic dissecting aneurysm. Lab Invest 1990;62:751-60. 24. Hirst A, Gore I. Is cystic medionecrosis the cause of dissecting aortic anenrysms? Circulation 1976;53:915-6. 25. Larson E, Edwards W. Risk factors for aortic dissection: a necropsy study of 161 cases. Am J Cardiol 1984;53:849-55. 26. Wilson S, Hutchins G. Aortic dissecting aneurysm: causative factors in 204 subjects. Arch Pathol Lab Med 1982 ;106:17580. 27. Francke U, Berg MA, Tynan K, et al. A gly 1127 ser mutation in an EGFqike domain of the fibrillin-1 gene is a risk factor for ascending aortic aneurysm and dissection. Am J Hum Genet 1995;56:1287-96. 28. Tilson MD, Seashore MR. Human genetics of the abdon~inal aortic aneurysm. Surg Gynecol Obstet 1984;158:129-32. 29. Norrg~rd O, Rais O, 2mgquist I(A. Familial occurrence., of abdominal aortic aneurysms. Surgery 1984;95:650-6. 30. Powell JT, Greenhalgh RM. Multifactorial inheritance of abdominal aortic aneurysms. Eur J Vasc Surg 1987;1:29-31. 31. Johnston KW, ScoNe TK. Multicenter prospective study of nonruptured abdominal aortic aneurysms. I. Population and operative management. J Vasc Surg 1988;7:69-79. 32. Martin P. On abdominal aortic aneBrysms. J Cardiovasc Surg
1978;19:597-8. 33. Gore I, Hirst Jr AE. Arteriosclerotic aneurysms ofthe abdominal aorta: a review. Prog Cardiovasc Dis 1973;16:113-50. 34. McNamara lJ, Pressler VM. Natural history of arteriosclerotic thoracic aortic aneurysms. Ann Thorac Surg 1978;26:46872. 35. )~oyceJW, Fairbairn II ~F. Aneurysms of the thoracic aorta: a clinical study with special reference to prognosis. Circulation 1964;29:176-81.
Submitted Apr. 22, 1996; accepted June 21, 1996.
Purpose: Evidence suggesting that genetic factors contribute to the development of common disorders can be obtained by demonstrating familial aggregation of the disease. This study investigated whether thoracic aortic dilations and dissections aggregate in families by comparing the prevalence o f thoracic aortic aneurysms, thoracic aortic dissections, and sudden death in first-degree relatives of patients referred for thoracic aortic surgery. Methods: Families were ascertained through 158 nonsyndromic patients referred for surgical correction of either thoracic aortic aneurysms or dissections (probands) and their 843 first-degree relatives. A control group of 547 first-degree relatives was derived from 114 proband spouses. Groups were examined for statistical differences in the prevalence of thoracic aneurysms, thoracic aortic dissections, abdominal aortic aneurysms, sudden death, and myocardial infarctions. Results: First-degree relatives of probands demonstrated a higher prevalence of thoracic aortic aneurysms and sudden death when compared with the control group. Relative risks of thoracic aortic aneurysm development in proband fathers, brothers, and sisters were 1.8, 10.9, and 1.8, respectively. A pattern of inheritance of the thoracic aortic aneurysms could not be determined. Conclusions: This study indicates proband first-degree relatives are at higher risk o f thoracic aortic aneurysms and sudden death compared with a control group. This study supports the role of genetic factors-in the cause of thoracic aortic aneur;ysms and provides important information for identifying individuals at risk. (J Vase Surg 1997;25:506-11.) ;
Aneurysms and dissections o f the aorta are assod a t e d with a high degree o f morbidity, mortality, and medical expenditure despite continued improvements in diagnostic and surgical techniques. 1,2 Prevention o f these disorders through the early identification of predisposed individuals and the modification o f contributing environmental and genetic factors is a potential cost-effective m e t h o d for addressing these diseases. Establishing familial aggregation and estimating relative risks o f these disorders for family From the Department of Internal Medicine, University of Texas Houston Medical School, and Baylor College of Medicine. Supported in part by a Pfizer Scholars Award, an American Heart Association grant-in-aid (93014300), a Schissler Foundation Fellowship, and an NIH Clinical Research Center Grant (MOIRR-02558). Presented as a Poster (45I) at the Forty-sixthAnnual Meeting of the American Society for Human Generics in Minneapolis, Minn., Oct. 24-28, 1995. Reprint requests: Dianna M. Milewicz,MD, PhD, Department of Internal Medicine, University of Texas-Houston Medical SchooI, MSB 1.614, 6431 Fannin, Houston, TX 77030. Copyright © t997 by The Societyfor Vascular Surgeryand International Society for Cardiovascular Surgery, North American Chapter. 0741-5214/97/$5.00 + 0 24/1/76003 506
members is useful not only for identifying susceptible individuals but also for implicating a genetic contribution to these disorders. Although abdominal aortic aneurysms have been well characterized in regard to familial aggregation, 3 risk factors, 4,s possible canses, 6,7 and potential modes o f inheritance, 8,9 less is known regarding thoracic aortic aneurysms (TAA) and thoracic aortic dissections. 2,1°-12 For this reason this study was primarily designed to c o m p a r e firstdegree relatives o f patients referred for surgical repair o f dilated or dissected thoracic aortas with a control g r o u p for differences in occurrence o f these disorders. Because rupture or dissection o f asymptomatic aneurysms has a high mortality rate, ~ this study also examined the prevalence o f sudden deaths in the two groups. O n e h u n d r e d fifty-eight patients were inter¢iewed, and family histories were obtained. TAA prevalence, age and sex distributions, disease, familial aggregation, and potential m o d e s o f inheritance are reported. This study d e m o n s t r a t e s familial aggregation o f TAA and sudden death as a first step toward identifying individuals with potential genetic predispositions for these disorders.
IOURNAL OF VASCULAR SURGERY Volume 25, Number 3
METHODS P o p u l a t i o n selection. The study population consisted o f first-degree relatives o f 158 patients (100 men and 58 women) who were referred to Methodist Hospital, Houston, Texas, for surgical repair o f TAA, thoracic aortic dissections, or thoracoabdominal aortic aneurysms. These individuals underwent corrective surgery within the 21/2-year period of September 1990 to March 1993. Patients excluded from the study had either known syndromes (e.g., Marfan syndrome) or vasculitis associated with thoracic aorta anomalies or were unaware o f their first-degree relatives' medical history. There were 153 white, 3 black, 1 Hispanic, and 1 Asian proband patients. A control population of first-degree relatives was obtained by interviewing 114 of the probands' spouses about their family history. This study recorded occurrence of symptomatic or previously diagnosed thoracic and abdominal aneurysms, dissection o f the thoracic aorta, myocardial infarction, and sudden death. Only relatives whose age placed them at risk for symptomatic aneurysm development were included in the study. A minimum age o f 30 years was selected based on the age of the youngest proband patient. Excluding the 308 offspring common to both groups, there were 843 proband and 547 control first-deg~'ee family members. Data collection. After appropriate consent as required by the Institutional Review Board was obtained, one o f the authors (M. R.) interviewed the study subjects with a standardized questionnaire designed to maintain consistency and minimize recall bias in the data collection process. Medical records were reviewed for aortogram and surgical pathology results. Aneurysms in the ascending thoracic aorta, aortic arch, or descending aorta were classified as thoracic aortic aneurysms (TAA). Aneurysms below the diaphragm but superior to the renal arteries were classified as thoracoabdominal aortic aneurysms (TAA). Infrarenal aneurysms were classified as abdominal aortic aneurysms (AAA). Aortic dissections were classified according to the DeBakey system) 4 The predominant disease was reported as either cystic medial necrosis or atherosclerosis. Sudden death was defined as an unexpected death o f unlmown cause in the absence o f a history o f heart disease and subsequent autopsy. In reported cases o f thoracic, abdominal, or cerebral aneurysms in first-degree relatives, medical, surgical, and radiologic records were obtained whenever possible to corroborate elicited medical histories. Statistical analysis. Statistical analysis was performed with an SPSS statistical software package
Biddinger et al.
507
(SPSS, Inc., Chicago, Ill.). Tripartate Logistic Regression was used to verify that dilated and dissected thoracic aorta data could be combined for subsequent statistical analysis. 1S Mann-Whimey and 2 × 2 ×2 analyses were used to test for age and sex differences between the study and control groups, respectively. One-way analyses o f variance were used to compare ages among multiple groups. Log linear analysis was used to test for associations among the multiple categoric variables. Fisher's exact test was used to analyze 2 × 2 tables with multiple data entries smaller than five. Overall relative risk estimates for TAA and sudden death were computed with the Mantel-Haenszel estimator.16 Separate TAA relative risk estimates for proband parents and siblings were calculated as described by Webster et al)7 with TAA cumulative incidence functions ~8 derived from data published by Bickerstaff et al.19 and tables given in Bailer and Ederer. 2° RESULTS P r o b a n d characteristics. Proband characteristics, summarized in Table I, were' similar to previous findings in regard to proband sex, age, aneury,;m history, aortic anomaly at the time of surgery, and disease. 1°,12,17,21 Men had an average age o f 63.8 years (SD = 11.31), and women had an average age o f 67.0 years (SD = 11.37). The ratio o f affected men to women was 1.7:1, with men demonstrating a higher prevalence o f aortic dissections (p = 0.004). There were no statistically significant sex-related differences with respect to the occurrence o f previous TAA, previous AAA, dilated thoracic or thoracoabdominal aneurysm, aortic dissection, abdominal aortic aneurysm, hypertension, or smoldng history. Cystic medial necrosis was significantly associated with a history o f TAA (p = 0.006) and aortic dissection (p = 0.012). Atherosclerosis was associated with thoracoabdominal aortic aneurysm (p < 0.0001) and AAA (p < 0.0001). No age differences were noted when sex, aneurysm type (dilated vs dissecting), and aneurysm location were examined at the p = 0.05 significance level. Finally, there was a statistically significant association between hypertension and a family history o f dilated or dissecting thoracic aneurysms (p = 0.031). Disease prevalence in first-degree relatives. From a statistical point o f view Tripartate logistic regression o f the data revealed that proband aneurysm type (dilated vs dissecting) did not make a statistically significant contribution toward predicting first-degree relative aortic disease (data not shown), indicating that TAA and dissected thoracic
JOURNAL OF VASCULARSURGERY
508
Biddinger et al.
1997
March
Table I. Summary ofproband characteristics Previous aneurysm
Men
Total Women
Age (yr)
TAA
30-39 40-49 50-59 60-69 70-79 ->80
2 2 7 12 4
30-39 40-49 50-59 60-69 70-79 ->80
Total
27 1 2 10 4 1 18
Dilated thoracic aneurysm
AAA
TAA
TAAA
Both
I
1 1 2
1 1 3 6
1 3 6 10 5
4
11
25
1 4 3 17 5 2 32 2 1 1 12 8 3 27
1 7 12 1 21
3 3 1 7
Dissecting thoracic aorta classification H
Disease
III
CMN
1
3 9 10 22 12
2 5 2 1 10
3 5 8 1 18
56
ATH 1 3 13 13 2 32 2
1 1
7 5 1 14
1
2 3 3
2 1 3
8
2 1 3
10 6 2 20
6
5 10 10 3 30
TAA, Thoracic aortic aneurysm; TAA, thoracic aortic aneurysm; TAAA, thoracoabdominal aneurysm; Both, TAA with TAAA; CMN, cystic medical necrosis; ATH, atherosclerosis.
Table II. Disease prevalence in first-degree relatives Probands (N) Event TAA or dissecting thoracic aorta AAA MI Sudden death
Control group (N)
Statisics
Men (429)
Women (414)
Men (274)
Women (273)
P
12 11 48 38
5 1 17 10
2 7 31 9
0 0 9 5
0.0096 0.82 0.78 0.0057
TAA, Thoracic aortic aneurysm; AAA, abdominal aortic aneurysm; M/, myocardial infarction.
aorta data could be combined for subsequent analysis. However, dilated thoracic anemTsms and thoracic aortic dissections are thought to have different causes, suggesting that they should be analyzed separately.2226 Because of this discrepancy we provide the first-degree relative data on all the proband patients and then an analysis separating the proband patients into two groups based on whether the proband had aortic dilation or dissection. Overview o f the combined data. Comparison ofproband first-degree relatives with the spouse control groups revealed no statistically significant differences in age or sex. The prevalence of TAA and sudden death was significantly higher in the proband patients first-degree relatives compared with the control group relatives. As shown in Table II 17 of the proband patients' relatives had TAA, whereas only 2 relatives of the control group were affected (Mantel and Haenszel X2 p = 0.0096). Proband relatives reported 48 occurrences of sudden death compared with 14 sudden deaths in the control group (Mantel
and Haenszel X2, p = 0.0057). In contrast, no significant difference was seen in the occurrence of AAA (p = 0.82) or myocardial infarctions (p = 0.78) between the proband and control groups. With regard to sex, men were more likely than women to have thoracic aortic aneurysms or dissections (p = 0.042), sudden death (p < 0.0001), myocardial infarction (p < 0.00001), and abdominal aortic aneurysms (p < 0.001). Segregated data analysis. First-degree relatives of patients in the proband and control groups were segregated into two groups based on the type of aortic disease present in the proband patient (dilated aneurysm or aortic dissection) as shown in Table III. Relatives of proband patients with dilated TAA also had a significant prevalence of dilated TAA (p = 0.026) compared with those of patients in the control group. When proband fathers, mothers, brothers, and sisters were examined separately, there was insufficient power in the Mantel-Haenszel test to obtain relative risk estimates. As an alternate ap-
JOURNAL OF VASCULARSURGERY Volume 25, Number 3
Biddinger et al.
509
Table III. Prevalence in first-degree relatives segregated by proband thoracic aorta dilation or dissection TAA probands group
TAA controlgroup
Dissecting aorta probands group
Dissecting aorta controlgroup
(N)
(N)
(N)
(N)
Statistics
Event
Men (243)
Women (254)
Men (159)
Women (153)
Men (186)
Women (160)
Men (115)
Women (120)
TAA p value
Dissecting p value
TAA Dissection AAA MI Sudden death
7 0 5 31 24
1 1 0 10 5
0 0 0 24 4
0 0 0 6 6
4 1 6 17 14
2 1 1 7 5
2 0 7 7 3
0 0 0 3 1
.026 1.0 .163 .504 .089
.484 .517 .461
.177 .022
TAA, Thoracic aortic aneurysm; Dissection, thoracic aorta dissection; A_AA, abdominal aortic aneurysm; MI, myocardial infarction; TA.A p value, values obtained when comparing TAA probands and TAA controls; Dissectingp value, values obtained when comparing dissecting aorta probands with dissecting aorta Controls.
proach the technique described by Webster et al? 7 was used to obtain conservative relative risk estimates for each of these individuals. Relative risks of dilated thoracic aneurysm development in TAA proband fathers, brothers, and sisters were 1.8 (95% CI = 1.5, 2.2), 10.9 (95% CI = 9.4, 12.7), and 1.8 (95% CI = 1.6, 2.1), respectively. No estimate of risk in proband mothers could be calculated, because none of them had TAA. No difference was found in sibling risk with regard to proband sex (p = 0.267). Furthermore no significant difference was seen in prevalence of dissecting aortic aneurysms (p = 1.0), AAA (p = 0.163), myocardial infarctions (p = 0.504), or sudden death (p = 0.089). In contrast, first-degree relatives of probands with dissecting aortic aneurysms demonstrated higher prevalence of sudden death (relative risk = 3.35; CI% = 1.13, 9.99), but no significant differences were detected in the prevalence of dilated TAA (p = 0.484), aortic dissections (p = 0.517), AAA (p = 0.461)or myocardial infarction (p = 0.177). Analysis o f p r o b a n d pedigrees. O f the 158 proband pedigree patients, 16 multiplex pedigrees contained first-degree family members with either dilated or dissecting thoracic aortic aneurysms. With respect to TAA proband patients, dilated aneurysms were seen in one father, six brothers, and one sister including the single instance of a father-son affected pair. One sister of a proband patient had aortic dissection. In contrast, none of the spouse control pedigrees contained members affected with either TAA or thoracic aortic dissection. In the eight pedigrees ascertained from proband patients with aortic dissection, one father and one sister had aortic dissection. Dilated thoracic aneurysms were seen in two fathers, one mother, two
brothers, and one sister. Two control simplex pedigrees contained fathers with dilated thoracic aneurysms. Visual inspection of these pedigrees did not suggest a probable mode of inheritance for either dilated or dissecting thoracic aortic aneurysms (data not shown). DISCUSSION
This study demonstrates familial aggregation of TAA and represents a step towards identifying individuals predisposed to have the disorder. Although it is currently believed that the cause of dilated and dissecting aneurysms is different,23-26 there are families with autosomal dominant inheritance of aortic disease in which dilated aneurysms and aortic dissections are seen together. 12,27 In the families with aortic disease identified in this study, seven families had affected members with dilated aneurysms along with other family members with aortic dissections. These facts suggested that dividing the probands into two groups based on whether they had a dilated aneurysm or aortic dissection might not be appropriate when we were attempting to determine the familial aggregation of thoracic aortic aneurysms and dissections. Therefore we have presented our data both combining and separating the probands based on whether they had a dilated aneurysm or aortic dissection. Bickerstaff et al.19 reported a TAA incidence of 5.9 per 100,000 person-years in white patients. Applying this estimate to our study population, an average of 1.9 dilated aneurysms were expected in firstdegree relatives of dilated aneurysm proband patients (32,206 person-years). A statistically significant fourfold increase in the number of dilated aneurysms was observed. Subsequent examination revealed that
510
Biddinger et al.
proband fathers and sisters had a 1.8-fold increased risk of TAA, whereas proband brothers had an ll-fold increased risk of TAA. These findings demonstrate that TAAs aggregate in families and support a sex-influenced disease process that affects proband siblings. Dissecting aneurysms were rarely reported in first-degree relatives o f probands and were never reported in control family members. As a consequence no meaningful statistical comparisons could be drawn. However, asymptomatic dissecting aneurysms have a high mortality rate 2 and in the absence of radiographic studies, surgery, or autopsy can either go undetected or can be misclassified as sudden deaths. Examination o f sudden deaths in first-degree relatives o f dissecting aneurysm proband patients revealed a statistically higher prevalence and a threefold increased risk for this disorder. In contrast, relatives o f patients with dilated aneurysm showed no increased prevalence o f sudden death. The lack of a statistically significant difference in deaths attributed to myocardial infarction also lent s u p p o r t to the hypothesis that aneurysm rupture as opposed to another cause contributed to the increased number of unexplained sudden deaths. Hypertension has long been recognized as an important risk factor for the development of aortic ancurysms and aortic dissections. 4,19 In this study 60% o f p r o b a n d patients with dilated aneurysms and 70% o f p r o b a n d patients with aortic dissections were hypertensive (data not shown). No statistically significant difference was seen in the prevalence o f elevated blood pressure in the two groups o f p r o b a n d patients (p = 0,20). Although hypertension is associated with the two disorders, the formation o f dilating or dissecting aortas cannot be ascribed to elevated blood pressure alone. Thoracic aortic aneurysms and abdominal aortic aneurysms may have heterogeneous causes based on differences in familial aggregation and disease. Although abdominal aortic aneurysms have been shown to occur as multiple events in families, 3,17,28,29 no statistically significant difference was seen in AAA occurrence in proband and control relatives in th'is study. The observation that men are more often affected than women in families with TAA aggregation contrasts with familial abdominal aortic aneurysm studies, in which women were more likely to be affected than men. 3°,a~ A final difference between TAA and AAA is related to disease. There were more cases o f cystic medial necrosis (55%) than atherosclerosis (45%) reported in pathologic sections available for review. In contrast, the vast majority o f the aortic
J O U R N A L OF VASCULAR SURGERY March 1997
disease found in previous AAA studies were atherosclerotic lesionsY a4 Inspection o f the proband pedigrees revealed the occurrence o f both dilated thoracic aneurysms and aortic dissections regardless o f the proband's aortic anomaly. The predominance o f affected male patients at first suggested X-linked inheritance for eitiler dilated or dissected aortas. However, occurrences o f father-to-son transmission in both cases excluded this mode of inheritance but supported a sex-influenced process. There were seven TAA pedigrees and three aortic dissection pedigrees where only siblings were affected with dilated TAA, supporting either autosomal recessive inheritance or autosomal dominant inheritance with incomplete penetrance. Genetic heterogeneity could account for similar TAA phenotypes arising from different genotypes and alternative modes o f transmission. O f the 158 proband patients in this study, 153 (96.8%) were white. Other studies relating to the epidemiologic characteristics o f aortic aneurysms have also reported a predominance o f affected white patients. 19,3S Although the possibility exists that white people are more likely to have TAA than people o f other ethnicity, the high prevalence o f white patients in this study may also represent a selection bias. Because TAAs are usually asymptomatic, the high rate o f white patients referred for surgery could reflect a disproportionate access to medical care with a concomitant increase in asymptomatic aneurysm detection. In either case the results o f this study should be applied only to white populations. In conclusion, first-degree relatives of proband patients referred for dilated thoracic aneurysm repair are at an increased risk for having TAA themselves. This is particularly true for male siblings o f affected probands. First-degree relatives o f patients with dissecting aneurysm have a threefold increased risk o f sudden death possibly because o f aortic dissection. Formal segregation analysis may elucidate the mode(s) o f inheritance o f TAA and thoracic aortic dissection, providing a step toward identifying underlying genetic alterations that predispose individuals to these disorders. Studies are in progress to identify specific genetic and environmental influences that contribute to TAA pathogenesis before cost-effective screening methods are recomm e n d e d for identifying family members at increased risk. We thank the patients and their spouses for participating in this study. We also thank Dr. Max Buja for his
JOURNAL OF VASCULARSURGERY Volume 25, Number 3
expertise in cardiovascular p a t h o l o g y a n d Dr. Dennis J o h n s t o n for his assistance w i t h statistics. REFERENCES
1. Lilienfeld D, Baxter I, Sprafka J. Prevalence of aortic aneurysms in the twin cities metropolitan area, 1979-84. Public Health Rep 1995;108:506-9. 2. Lilienfeld DE, Gunderson PD, Sprat'ka JM, Vargas C. Epidemilogy of aortic aneurysms: I. Mortality trends in the United States, 1951 to 1981. Arteriosclerosis 1987;7:63743. 3. Tilson MD, Seashore MR. Fifty families with abdominal aortic aneurysms in two or more first-order relatives. Am l Surg 1984;147:551-2. 4. Spittell JA. Hypertension and arterial aneurysm. J Am Coll Cardiol 1983;1:533-40. 5. Auerbach O, Garfinkel L. Atherosclerosis and aneurysm of aorta in relation to smoking habits and age. Chest I980;78: 805-9. 6. Powell J, Greenhalgh RM. Cellular, enzymatic and genetic factors in the pathogenesis of abdominal aortic aneurysms. J Vasc Surg 1989;9:297-302. 7. Tilson MD, Reilly JM, Brophy CM, Webster EL, Barnett TR. Expression and sequence of the gene for tissue inhibitor of metalloproteinases in patients with abdominal aortic aneurysms. J Vasc Surg 1993;18:266-70. 8. Reilly JM, Tilson MD. Incidence and etiology of abdominal aortic aneurysms. Surg Clin North Am 1989;69:705-11. 9. Majumder PP, St. lean PL, Ferrell RE, Webster MW, Steed DL. On the inheritance of abdominal aortic aneurysm. Am l Hum Genet 1991;48:164-70. 10. DeBakey ME, Noon GP. Aneurysms of the thoracic aorta. Mod Conc Cardiovasc Dis 1975;44:53-8. I 1. Wilson S, Hutchins G. Aortic dissecting aneurysms: causative factors in 204 subjects. Arch Pathol Lab Med 1982;106:17580. 12. Nicod P, Bloor C, Godfrey M, et al. Familial aortic dissecting aneurysm. J Am Coil Cardiol 1989;13:811-9. 13. Skupin M, Blum U, Krause E, Ruhr-Leukart M. Results of surgical repair for 110 thoracic aortic aneurysms. Thorac Cardiovasc Surg 1990;38:175-80. 14. DeBakey M, Henly W, Cooley D. Surgical management of dissecting aneurysms of the aorta. J Thorac Cardiovas Surg 1965;49:130-48. 15. Greene WH. Econometric analysis. 2nd ed. New York: MacMillan, 1992. 16. Mantel N, Haenszel W. Statistical aspects of the analysis of data from retrospective studies of disease, l Natl Cancer Inst 1959;22:719-48. 17. Webster MW, St Jean PL, Steed DL, Ferrell RE, Majumder PP. Abdominal aortic aneurysm: results of a family study. J Vasc Surg 1991;13:366-72.
Biddinger et al.
511
18. Breslow NE, Day NE. Statistical methods in cancer research: Vol I. The analysis of case-control studies. Lyon: IARC Scientific Publications No. 32, 1980:49-51. 19. Bickerstaff LK, Pairolero PC, Hollier LH, et al. Thoracic aortic aneurysms: a population-based study. Surgery 1982; 92:1103-8. 20. Bailar III lC, Ederer F. Significance factors for the ratio of a Poisson variable to its expectation. Biometrics 1964;20:63942. 21. Pressler V, McNamara I. Thoracic aortic aneurysm: natural history and treatment. J Thorac Cardiovasc Surg 1980;79: 489-98. 22. Shores J, Berger KR, Murphy EA, Pyeritz RE. Progression of aortic dilatation and the benefit of long-term [3-adrenergic blockade in Marfan's syndrome. N Engl }- Med 1994;330: 1335-41. 23. Nakashima Y, Shiokawa Y, Sueishi K. Alterations of elastic architecture in human aortic dissecting aneurysm. Lab Invest 1990;62:751-60. 24. Hirst A, Gore I. Is cystic medionecrosis the cause of dissecting aortic anenrysms? Circulation 1976;53:915-6. 25. Larson E, Edwards W. Risk factors for aortic dissection: a necropsy study of 161 cases. Am J Cardiol 1984;53:849-55. 26. Wilson S, Hutchins G. Aortic dissecting aneurysm: causative factors in 204 subjects. Arch Pathol Lab Med 1982 ;106:17580. 27. Francke U, Berg MA, Tynan K, et al. A gly 1127 ser mutation in an EGFqike domain of the fibrillin-1 gene is a risk factor for ascending aortic aneurysm and dissection. Am J Hum Genet 1995;56:1287-96. 28. Tilson MD, Seashore MR. Human genetics of the abdon~inal aortic aneurysm. Surg Gynecol Obstet 1984;158:129-32. 29. Norrg~rd O, Rais O, 2mgquist I(A. Familial occurrence., of abdominal aortic aneurysms. Surgery 1984;95:650-6. 30. Powell JT, Greenhalgh RM. Multifactorial inheritance of abdominal aortic aneurysms. Eur J Vasc Surg 1987;1:29-31. 31. Johnston KW, ScoNe TK. Multicenter prospective study of nonruptured abdominal aortic aneurysms. I. Population and operative management. J Vasc Surg 1988;7:69-79. 32. Martin P. On abdominal aortic aneBrysms. J Cardiovasc Surg
1978;19:597-8. 33. Gore I, Hirst Jr AE. Arteriosclerotic aneurysms ofthe abdominal aorta: a review. Prog Cardiovasc Dis 1973;16:113-50. 34. McNamara lJ, Pressler VM. Natural history of arteriosclerotic thoracic aortic aneurysms. Ann Thorac Surg 1978;26:46872. 35. )~oyceJW, Fairbairn II ~F. Aneurysms of the thoracic aorta: a clinical study with special reference to prognosis. Circulation 1964;29:176-81.
Submitted Apr. 22, 1996; accepted June 21, 1996.