Natural history of descending thoracic and thoracoabdominal aneurysms

Natural History of Descending Thoracic and Thoracoabdominal Aneurysms Randall B. Griepp, MD, M. Arisan Ergin, MD, PhD, Jan D. Galla, MD, PhD, Steven L. Lansman, MD, PhD, Jock N. McCullough, MD, Khanh H. Nguyen, MD, James J. Klein, MD, and David Spielvogel, MD Department of Cardiothoracic Surgery, Mount Sinai School of Medicine, New York, New York

Background. A review of 165 patients with chronic dissecting and degenerative aneurysms of the descending thoracic and thoracoabdominal aorta initially managed nonoperatively was carried out to ascertain factors associated with a high risk of rupture. Methods. Changes in the aneurysms were followed with three-dimensional reconstructions of computed tomograph scans. Risk factors were compared in patients with dissecting and nondissecting aneurysms who experienced rupture, in whom operation was recommended during the course of follow-up, and in those without rupture or operation. Results. Nondimensional variables associated with an enhanced risk of rupture include age, the presence of chronic obstructive pulmonary disease, and even uncharacteristic continued pain. Patients with rupture of dissections had significantly higher blood pressures than survivors, and significantly smaller maximal descending

thoracic aortic diameters (median 5.4 cm) than patients with rupture of degenerative aneurysms (median 5.8 cm). The extent of the aneurysm, as reflected by the maximal abdominal aortic diameter, was a significant risk factor for rupture only in nondissecting aneurysms. Mortality from rupture was significantly higher in patients with chronic dissections than in patients with nondissecting aneurysms: 9/10 vs 26/34 (p 5 0.004). Conclusions. Almost 20% of patients followed nonoperatively succumbed to rupture, suggesting that a more aggressive surgical approach toward patients with chronic aneurysms of the descending thoracic and thoracoabdominal aorta is warranted. An individualized risk of rupture within 1 year can now be calculated, and patients whose operative risk is lower than their calculated risk should be offered elective surgery. (Ann Thorac Surg 1999;67:1927–30) © 1999 by The Society of Thoracic Surgeons

A

follow-up is also variable, so that is often difficult to be sure whether deaths occurred from rupture or from other causes. Recent improvements in the results of surgery and in the ability to image aneurysms accurately have resulted in the withdrawal of increasing numbers of patients for elective surgery, even in the most careful, rigorous studies. The removal of these patients weakens our ability to assess the possible contribution to rupture risk of factors that are frequent indications for surgery, such as aneurysm size, extent, growth rate, and the presence of pain. Thus, the conclusions of recent natural history studies must be viewed in context: they are most applicable to relatively asymptomatic patients with moderate-size or small aneurysms. Fortunately, these are the patients for whom the decision whether or not to operate is most often a dilemma.

lthough results after elective surgery for descending thoracic and thoracoabdominal surgery have improved markedly in recent years, unexpected rupture of thoracic aneurysms is still almost invariably fatal [1]. The ability to make an accurate prediction of the probability that an aneurysm will rupture would greatly improve selection of patients for elective surgery. From a number of recent investigations into the natural history of thoracic and thoracoabdominal aneurysms, a consensus is emerging about some specific risk factors that may increase the likelihood of rupture, but some uncertainties remain. It has been difficult to extract information relevant for future patient care from studies of the natural history of aneurysms for several reasons. Most studies have included a mixture of different proportions of patients with aneurysms with different etiologies, different locations within the thorax, at varying intervals from acute onset, classified by several conflicting and overlapping systems of nomenclature. The completeness and accuracy of

Presented at the Aortic Surgery Symposium VI, April 30 –May 1, 1998, New York, NY. Address reprint requests to Dr Griepp, Department of Cardiothoracic Surgery, Mount Sinai School of Medicine, One Gustave Levy Place, New York, NY 10029.

© 1999 by The Society of Thoracic Surgeons Published by Elsevier Science Inc

Risk Factors for Rupture: Dimensional Variables Size Recent studies have confirmed the importance of aneurysm size as a major factor predicting rupture. Rupture is much more likely to occur when an aneurysm exceeds 5 cm in diameter, and the risk increases as the aneurysm increases further in size [2– 6]. In a recent multivariate 0003-4975/99/$20.00 PII S0003-4975(99)00427-0

1928

GRIEPP ET AL RISK FACTORS FOR ANEURYSM RUPTURE

Ann Thorac Surg 1999;67:1927–30

analysis of 114 patients with nondissecting thoracic and thoracoabdominal aneurysms, each 1-cm increment in the maximal descending aortic diameter increased the risk of rupture by a factor of 1.9 ( p 5 0.003) [6]. The extent of the aneurysm is also important, as reflected by the fact that each increase of 1 cm in the maximal abdominal diameter further increases rupture risk by a factor of 1.5 ( p 5 0.05) [6].

Growth Rate Most studies have demonstrated that the rate of expansion of aneurysms is exponential, as might be predicted from physiological principles: enlargement accelerates as aneurysms get bigger [7]. Recently, the widely held belief that a faster growth rate is an independent risk factor for thoracic aneurysm rupture has been confirmed in a study by Lobato and Puech-Leao [4].

Risk Factors for Rupture: Nondimensional Variables Perhaps because selection of patients with chronic thoracic aneurysms for elective surgery is usually based on the size, extent, and rate of expansion of aneurysms, historical and constitutional factors that may contribute to rupture risk have received far less attention than characteristics describing the aneurysm itself.

Smoking/Chronic Obstructive Pulmonary Disease The presence of chronic obstructive pulmonary disease (COPD) is surprisingly powerful predictor of aneurysm rupture, as first pointed out by Cronenwett and associates in relation to abdominal aneurysms [8]. In our multivariate analysis of nondissecting thoracic and thoracoabdominal aneurysms, a history of COPD increased the odds of rupture by a factor of 3.6 ( p 5 0.04) [6]. A history of smoking is found in the majority of patients with aortic aneurysms, and is also a risk factor for rupture, but COPD has eclipsed smoking in several studies in which the two are looked at in the same multivariate analysis [6, 7]. The greater predictive power of COPD may be related to individual differences in response to the toxic effects of smoking: the presence of COPD may be a sensitive indicator of intolerance of connective tissue to smoking-related toxicity both in the lung and in the aorta. In a survey of abdominal aneurysms by Strachan, the risk of rupture was unequivocally increased by smoking, with an odds ratio of 6.5 [9]. The effect on abdominal aneurysm expansion of continued smoking was monitored in another study by measuring nicotine metabolites in blood. Significantly more rapid growth was seen in smokers, justifying continuing to recommend cessation of smoking in all patients with aneurysmal disease [10]. In one of our early studies, we documented more rapid expansion of thoracic aneurysms in patients with a history of smoking: 0.70 cm/year in smokers vs 0.35 cm/year in nonsmokers [7].

Fig 1. Ruptured thoracic aneurysms in a Swedish population study, showing the increase found with increasing patient age. Data are from Johansson and colleagues [1].

Age The risk that a thoracic aneurysm will rupture increases quite dramatically with age. This is illustrated in Figure 1, constructed from a Swedish autopsy study of deaths from thoracic aneurysms [1]. The proportion of women with thoracic aneurysms also increases with age, almost equaling male incidence in the elderly. In our multivariate analysis of descending thoracic and thoracoabdominal aneurysms, the relative risk of rupture increased by a factor of 2.6 for every decade of age ( p 5 0.02) [6]. The recognition that there is a greater risk of aneurysm rupture in elderly patients should lead to a reexamination of the predilection of most surgeons to propose elective surgery more readily to younger patients.

Pain The evaluation of pain as a risk factor for rupture is complicated by the fact that most surgeons consider the presence of aneurysm-related pain an indication for surgery. This results in the removal of most patients with characteristic chest or back pain from those at risk for rupture. Therefore, we were surprised to discover that even vague and uncharacteristic pain at the time of nonoperative follow-up in patients with thoracic or thoracoabdominal aneurysms is significantly associated with subsequent rupture, with an odds ratio of 2.3 in our multivariate analysis ( p 5 0.04) [6].

Hypertension A history of hypertension is present in most patients with aneurysms, and it is widely recognized that hypertension, especially diastolic hypertension, is very highly correlated with the initial development of aneurysms [7]. Consequently, treatment of patients with aneurysms with beta-adrenergic blocking agents and other antihypertensive agents is almost universally recommended to minimize progression of their disease. Perhaps because such treatment is often effective, hypertension has only intermittently emerged as a risk factor for rupture in recent studies. In our own just-completed study of pa-

Ann Thorac Surg 1999;67:1927–30

tients with chronic type B dissection, however, patients who experienced rupture had significantly higher diastolic and mean blood pressures during follow-up visits than patients who survived without rupture, even though a history of hypertension was equally prevalent in both groups [11]. This observation reinforces the need for vigorous treatment of hypertension, as well as consideration of elective surgery for those patients in whom hypertension cannot adequately be controlled.

Renal Failure The presence of renal failure has been identified as a risk factor for rupture of thoracic, abdominal, and thoracoabdominal aneurysms by Cambria and colleagues and Perko and colleagues, and Masuda and colleagues cite renal failure as a risk factor for rapid expansion of thoracic aneurysms [2, 5, 12]. In our own studies, however, renal failure has not emerged as a significant risk factor for aneurysm rupture.

Etiology: Degenerative Aneurysms vs Chronic Dissections There are conflicting reports regarding whether or not the etiology of an aneurysm plays a decisive role in the likelihood of rupture. Certainly, there is a consensus that patients with known disorders of connective tissue, particularly those with Marfan’s syndrome, are likely to experience rupture more readily than other patients with aneurysms, and that patients with a family history of early dissection and rupture may be particularly at risk. A number of studies over the years have also suggested that the presence of dissection is a risk factor predisposing to rupture even in patients without Marfan’s syndrome [13, 14]. Our recent natural history study of descending aortic and thoracoabdominal aortic aneurysms excluded chronic dissections from the initial analysis with the idea that they might behave differently; they were subsequently studied separately. Comparisons of patients who experienced rupture revealed that those with chronic type B dissection had smaller maximal diameters in the descending aorta before rupture than patients with degenerative aneurysms: a median of 5.4 cm in dissections vs 5.8 cm in patients with degenerative aneurysms ( p 5 0.05) [6, 11]. Furthermore, the extent of the aneurysm, as reflected by the presence of a large abdominal diameter, was not a significant risk factor for rupture in dissections, in contrast to its significant role in predicting rupture in degenerative aneurysms. The contribution of nondimensional risk factors in predicting rupture is no different in patients with dissection than in patients with degenerative aneurysms, with a higher risk of rupture in older patients, and in those with COPD and/or continued pain [6, 11]. Continued hypertension was significantly more prevalent among patients with chronic dissection who experienced rupture than in those who remained alive; unfortunately, follow-up blood pressures were not available for comparison in patients with degenerative aneurysms [11]. The characteristics of patients with degenerative and dissect-

GRIEPP ET AL RISK FACTORS FOR ANEURYSM RUPTURE

1929

Table 1. Comparison of Patients With Rupture of Degenerative Aneurysms and Chronic Dissections Rupture Data Degenerative Aneurysms

Chronic Dissections

p Value

26/106 26/34 74.9 62% 35% 69% 5.8 4.7

9/50 9/10 73.4 56% 67% 78% 5.4 3.8

NS 0.004 0.15 NS NS NS 0.05 NS

Rupture rate Rupture deaths Age (mean, years) Pain COPD Hypertension Thoracic diameter (cm) Abdominal diameter (cm)

COPD 5 history of chronic obstructive pulmonary disease.

ing aneurysms who underwent rupture during nonoperative follow-up are shown in Table 1.

Estimating Risk of Rupture Regardless of the etiology of aortic enlargement, nearly 20% of patients in our studies underwent rupture despite periodic careful surveillance. An overwhelming majority of those who died during follow-up experienced aortic rupture; 90% of patients with chronic type B dissection and 75% of patients with degenerative aneurysms succumbed to rupture [6, 11]. The high proportion of patients dying as a consequence of rupture, compared with earlier reports, probably reflects recent improvements in treatment of potentially lethal ailments which often coexist with aneurysms, particularly hypertension and coronary artery disease. Our recent study of the natural history of descending aortic and thoracoabdominal aortic aneurysms has allowed us to generate an equation for the probability of rupture within 1 year for each patient [6]. The equation incorporates the patient’s age, history of COPD, presence of pain, and the maximal thoracic and abdominal diameters of the aneurysm (Fig 2). An individualized risk of death, paraplegia, or other serious complication of oper-

Fig 2. The formula for calculating the probability of rupture of a degenerative thoracic aortic aneurysm: age is entered in years; maximal descending (desc dia) and abdominal diameter (abd dia) are entered in cm, and pain and COPD are scored as 0 if absent and 1 if present. To calculate the probability of rupture in a patient with chronic dissection, 0.4 cm should be added to the thoracic diameter in order to compensate for the smaller size of chronic dissections at the time of rupture. Data derived from Juvonen et al [6].

1930

GRIEPP ET AL RISK FACTORS FOR ANEURYSM RUPTURE

ation can also be calculated, based on some of the same factors, including the patient’s age and the extent of the aneurysm to be resected [15]. For patients with degenerative aneurysms, our current practice is to recommend operation if the calculated risk of rupture within 1 year exceeds the estimate of operative risk. In patients with chronic type B dissection, in whom death is overwhelmingly likely to be related to rupture, and in whom rupture tends to occur more readily in less dilated and perhaps less extensive aneurysms, we recommend adding 0.4 cm to the thoracic diameter before carrying out the calculation; this will result in a conservative but still individualized estimate of rupture risk. Recent studies of the natural history of aneurysms have thus equipped us to refine the way in which we decide whether to recommend operation for a patient with a chronic degenerative or dissecting thoracic aneurysm: by balancing the risk of rupture against the operative risk in each patient rather than relying on general guidelines. Since our data come from a prospective study of all patients being followed nonoperatively, unlike some studies in which growth rates and other risk factors have been analyzed retrospectively after rupture has occurred, they accurately reflect the population of patients in whom the decision whether or not to operate electively is most problematic. The mortality figures argue that a more aggressive surgical approach is warranted, and the surgical results suggest that most patients with a high risk of rupture, while not ideal surgical candidates, do nevertheless survive elective aneurysm resection without sequelae. Since death after unexpected aortic rupture is the most likely outcome if elective aneurysm resection is not carried out, more patients with a high risk of rupture should be persuaded to undergo surgery, and a greater effort to control smoking and blood pressure should be made in those who remain in nonoperative follow-up.

Ann Thorac Surg 1999;67:1927–30

References 1. Johansson G, Markstrom U, Swedenborg J. Ruptured thoracic aortic aneurysms: a study of incidence and mortality rates. J Vasc Surg 1995;21:985– 8. 2. Cambria RA, Gloviczki P, Stanson AW, et al. Outcome and expansion rate of 57 thoracoabdominal aortic aneurysms managed nonoperatively. Am J Surg 1995;170:213–7. 3. Elefteriades JA, Hartleroad J, Gusberg RJ, et al. Long-term experience with descending aortic dissection: the complication-specific approach. Ann Thorac Surg 1992;53:11–21. 4. Lobato AC, Puech-Leao P. Predictive factors for rupture of thoracoabdominal aortic aneurysm. J Vasc Surg 1998;27: 446–53. 5. Perko MJ, Norgaard M, Herzog TM, et al. Unoperated aortic aneurysm: a survey of 170 patients. Ann Thorac Surg 1995; 59:1204–9. 6. Juvonen T, Ergin MA, Galla JD, et al. Prospective study of the natural history of thoracic aortic aneurysms. Ann Thorac Surg 1997;63:1553– 45. 7. Dapunt OE, Galla JD, Sadeghi AM, et al. The natural history of thoracic aortic aneurysms. J Thorac Cardiovasc Surg 1994; 107:1323–33. 8. Cronenwett JL, Murphy TF, Zelenock GB, et al. Actuarial analysis of variables associated with rupture of small abdominal aortic aneurysms. Surgery 1985;98:472– 83. 9. Strachan DP. Predictors of death from aortic aneurysms among middle-aged men: the Whitehall study. Br J Surg 1991;78:401– 4. 10. MacSweeney STR, Ellis M, Worrell PC, et al. Smoking and growth of small abdominal aortic aneurysms. Lancet 1994; 344:651–2. 11. Juvonen T, Ergin MA, Galla JD, et al. Risk factors for rupture of chronic type B dissections. J Thorac Cardiovasc Surg 1999; 117:776 – 86. 12. Masuda Y, Takanishi K, Takasu J, et al. Expansion rate of thoracic aneurysms and influencing factors. Chest 1992;102: 461– 6. 13. Bickerstaff LK, Pairolero PC, Hollier LH, et al. Thoracic aortic aneurysms: a population-based study. Surgery 1982;92: 1103– 8. 14. Pitt MPI, Bonser RS. The natural history of thoracic aortic aneurysm disease: an overview. J Cardiovasc Surg 1997;12: 270– 8. 15. Griepp RB, Ergin MA, Galla JD, et al. Looking for the artery of Adamkiewicz: a quest to minimize paraplegia following surgery for aneurysms of the descending thoracic and thoracoabdominal aorta. J Thorac Cardiovasc Surg 1996;112: 1202–15.