Predictors of morbidity and mortality with endovascular and open thoracic aneurysm repair

Article Outline

• Abstract
• Methods
• Results
  • Mortality
  • Morbidity
  • Graft-related events
• Discussion
• Conclusion
• Author contributions
• References
• Copyright

Background

Open and endovascular thoracic aneurysm repairs are associated with significant complications including paraplegia, stroke, vascular insufficiency, and death. Predictors of adverse outcomes are not well-defined in this patient population.

Methods

The database of the GORE TAG (W.L. Gore, Flagstaff, Ariz) Pivotal Trial comparing the TAG endograft to open repair was interrogated. Univariate (UVA) and multivariate analyses (MVA) of demographic, clinical, anatomic, and procedural variables were conducted to discover possible predictors of serious adverse events for the whole group and for the TAG and open cohort groups separately. Early adverse outcomes occurred within 30 days or the initial hospitalization. P value of ≤ .05 was significant.

Results

A total of 140 TAG and 94 open descending thoracic aneurysm (DTA) patients were analyzed, consisting of 128 men and 106 women. Perioperative deaths were 9/94 for open surgery and 3/140 for TAG patients, with 10/12 (7 open, 3 TAG) deaths occurring in men. Two female deaths were both after open surgery. Multivariate analysis showed predictors of death for all patients were symptomatic aneurysms and male gender. Analysis of a combined morbidity/mortality endpoint (stroke/paralysis/MI/death) showed elevated creatinine predicted these events for the whole group. Open surgery (P < .001) and increasing aneurysm diameter (P < .001) predicted an increased likelihood of any major adverse event. Open surgery was significantly associated with an increased risk of paraplegia (P = .002). Vascular complications were more frequent in the TAG (19%) than in open DTA patients (9%) (P = .038). Female gender (P = .01) predicted vascular complications within the endovascular group. For all analyses, long procedure times were correlated with adverse events. Women were noted to have longer procedure times for both TAG and open repairs.

Conclusion

Elevated creatinine levels and symptomatic aneurysms predict morbidity and mortality, respectively, regardless of repair type. Male gender predicted death after open surgery, and since most deaths (9 of 12) were in this group, male gender predicted death overall, despite women's more difficult endovascular TAA repairs as evidenced by longer procedure times and higher vascular complication rates. All major adverse events and paraplegia were more common for open surgery patients.

With an aging population and new treatment options available, repairs of descending thoracic aneurysms (DTA) are likely to become more common.¹, ², ³, ⁴ Both open and endovascular repair of descending thoracic aneurysm have been shown to have multiple and serious complications.⁵, ⁶, ⁷ There are few predictors of these adverse events, for open or thoracic endovascular aneurysm repair (TEVAR), making patient education and selection difficult.

Until recently, the Gore TAG stent graft (W.L. Gore, Flagstaff, Ariz) was the only Food & Drug Administration (FDA)-approved thoracic stent graft for treatment of DTA. We examined patients treated with the TAG device as well as the open surgery cohort for the Phase II TAG clinical trial to determine if there were demographic or intraoperative factors that could predict the primary endpoint of death within 30 days. Other endpoints included the combined outcomes of death, paraplegia, stroke, or myocardial infarction. We also sought to determine predictors of vascular complications and of long-term graft-related events such as migration, endoleak, or conversion. All analyses of immediate and long-term complications and possible predictors of major adverse events (MAEs) are new to this work. Similarly, this is the first work to examine graft-related complications in a large population of open and endovascular patients to try to determine predictors of late interventions so that this information could be used in patient selection.

Back to Article Outline

Methods 

The Gore TAG pivotal trial was a multicenter, prospective, non-randomized Phase II study that recruited surgical candidates with DTA from September 1999 through May 2001. Treatment modality (open or endovascular) was determined by the surgeon based on patient characteristics and availability of endovascular grafts. One hundred forty endovascular patients were enrolled. Ninety-four patients with DTA treated by open surgery were used as a control group. Of the 94 open surgical control patients, 44 were concurrent subjects and 50 were historic controls from the enrolling institutions. Details on the two populations have been previously published.⁸, ⁹ Briefly, patients with DTA of at least twice the diameter of the normal thoracic aorta and with 2 cm of non-aneurysmal neck for sealing distal to the left carotid artery and proximal to the celiac artery were eligible for endovascular treatment. TAG devices ranged from 26 to 40 mm in diameter. All devices in this analysis were the original TAG endograft available prior to the revision which eliminated the longitudinal spine and replaced the fabric with a low-porosity material. Open repair was performed according to local protocols at the participating institutions. The extent of the open repair could not extend more proximally than the left carotid artery or more distally than the celiac axis. There were no mandates regarding use of spinal cord protection or use of left heart bypass. High-risk patients, including those with dissection, rupture, mycotic aneurysm, or trauma were excluded, as were medically high risk patients.⁸, ⁹ Follow-up exams, 4-view chest x-rays and spiral CT scans were performed at 1, 6, and 12 months and yearly thereafter. These exams were performed at 3 months if an endoleak was present. Five-year follow-ups were concluded for all available patients in August of 2006.

The 5-year results of the Gore TAG pivotal trial were analyzed for predictors of morbidity and mortality. Preoperative demographic and anatomic characteristics were used as independent variables to predict intraoperative results, postoperative complications and long-term outcomes. Univariate (UVA) and multivariate analyses (MVA) were conducted to discover possible predictors of major adverse events. Major adverse events (MAEs) as defined per Sacks criteria,¹⁰ were reported by the study sites and verified by clinical events coordinators. MAEs were those that resulted in a prolongation of treatment, new hospitalization, major disability, or death. Major and minor adverse events were adjudicated by a clinical events committee. Perioperative deaths were those deaths which occurred in the hospital or within 30 days of the initial procedure. Pre and postoperative DTA measurements and endoleak assessments were performed at the study sites. Significant sac size change was defined as ≥5 mm change of the largest diameter of the aneurysm from the baseline 1-month measurement. Migration was defined as a graft shift of ≥10 mm either cranial or caudal compared to the 1-month baseline. Spine fractures, seen in 20 patients in the original TAG study group prior to the revision of the product, were not counted as material or device failures in this analysis due to the acknowledged problem with the product causing the fractures rather than with patient characteristics as a causative agent. Endoleaks were infrequently seen with spinal fractures⁹, ¹¹ and, therefore, were counted as late graft-related events.

Due to a low frequency of events, the most serious complications of TEVAR and open DTA repair were combined for an endpoint of death, myocardial infarction (MI), stroke, or paraplegia. These events were also examined separately. Patients were analyzed by treatment group and as an entire cohort for predictors of major adverse events.

All variables with a P value of .10 or lower in univariate analysis were included in multivariate analyses. Multivariate analysis modeling was assessed using the Hosmer-Lemeshow test for goodness of fit. Early adverse outcomes occurred within 30 days or during the initial hospitalization. Predictors of late device-related events at 5 years were analyzed for the TAG group only. A P value of ≤ .05 was considered significant. Table I lists variables used as predictors of adverse events.

Table I. Variables used as predictors of adverse events including death

Back to Article Outline

Results 

A total of 140 TAG patients (60 women, 80 men) and 94 open patients (46 women, 48 men) were analyzed. Details of demographic comparisons between the populations have been previously published.⁸ The only significant differences in the preoperative characteristics or presentations of the patients was that there were significantly more symptomatic aneurysms in the open surgical group (36/94 [38%] vs 30/140 [21%], P = .007). The designation of “symptomatic” was made by the treating surgeon without standardized criteria for this status. The percentage of patients with a history of coronary artery disease in the TAG group was slightly higher (49% TAG vs 36% open) and this trended toward significant (P = .06). With these exceptions, there were no other notable preoperative differences.

Mortality 

Perioperative deaths were 3/140 (2.1%) for TAG patients and 9/94 (9.6%) for open procedures (P = .01 in univariate analysis). Details on the perioperative deaths have been previously reported.⁸, ⁹, ¹¹ The three TAG deaths were from stroke, cardiac arrest, and sepsis. The 9 open repair deaths were due to respiratory failure (n = 4), stroke (n = 3), cardiac causes (n = 1), and aorto-enteric fistula (n = 1). Ten of 12 deaths were in men (7 open, 3 TAG) and two were in women (both open surgery DTA repairs). Patients who were categorized as presenting with a symptomatic aneurysm had a 12% death rate (8/66) vs a 2% death rate (4/168) in non-symptomatic patients (P = .004 in univariate analysis). The results of the univariate analysis for predictors of death in the full TAG/open surgery cohort are presented in Table II. Even though procedure time was initially found to be significantly associated with death, this was not included in the multivariate analysis as it was thought to be a marker for difficult surgery rather than a causative agent for morbidity or mortality. Additionally, it was also strongly associated with the type of repair, with open surgery having predictably longer operating times. The multivariate analysis for predictors of death in the entire cohort showed only symptomatic aneurysms (P = .004) and male gender (P = .048) were significant.

Table II. Univariate analysis of predictors of death for open and endovascular DTA repair (n = 234)

Max, maximum; Hx, history; CI, confidence interval; PAD, peripheral arterial disease; CHF, congestive heart failure; CAD, coronary artery disease; MI, myocardial infarction; ASA, American Society of Anesthesiology.

Shaded variables did not meet the P ≤ .1 threshold for inclusion in the multivariable model.

Morbidity 

A composite endpoint of death/stroke/MI/paraplegia was constructed to try to predict the factors causing the most serious adverse outcomes. Of 17 patients who suffered spinal cord ischemia, 6 died, all in the open surgery cohort. Table III shows the incidence of these events for the entire cohort and by treatment group. Variables were screened in univariate analysis for both the group as a whole and for the treatment groups separately. The results of the univariate analysis are shown in Table IV. The multivariate model was run using the listed variables as well as including a variable for treatment group (P = .006 in univariate analysis for entire cohort). On MVA, only elevated creatinine (P = .002) and treatment group (P = .006) predicted combined morbidity endpoint in the entire cohort of DTA repair patients. The predictive value of elevated creatinine is of questionable validity in the open surgery group as many of the historic controls did not have a preoperative creatinine value available. Mean creatinine in TAG patients with the composite endpoint was 1.54 mg/dL vs 1.06 mg/dL in patients without the composite endpoint. For open surgery patients this difference was less pronounced at 1.18 mg/dL with the composite endpoint and 1.09 mg/dL without. Treatment group appears to be the best predictor of the combined adverse events. Analysis of myocardial infarction and stroke as separate events did not yield significant predictors, partially due to the low frequency of these events.

Table III. Incidence of death/stroke/MI or paraplegia

	TAG (n = 140)	Open control (n = 94)	Total (n = 234)
Composite	11(7.9%)	17(18.1%)	28(12%)
Death	3(2%)	9(10%)	12(5%)
Stroke	5(4%)	4(4%)	9(4%)
MI	1(0.7%)	1(1.1%)	2(0.9%)
Paraplegia	4(3%)	13(14%)	17(7%)

MI, Myocardial infarction.

Note: Patients may have had more than one complication.

Table IV. Significant predictors of the composite endpoint of death/stroke/MI/paraplegia by treatment group in univariate analysis, separated by treatment group

CI, Confidence interval.

Analysis of paraplegia alone showed that in univariate analysis, treatment group (P = .002), preoperative hematocrit value (P = .060) history of arrhythmia (P = .059), and presence of patent intercostal arteries (P = .091) were significant. In multivariate analysis, only treatment group was significant (P = .002). Univariate analysis of the individual groups failed to show any significant predictors of paraplegia to the 0.05 level in the surgical control group. In the TAG group, decreasing weight was predictive of paraplegia/paresis (P = .013), but with so few patient affected (n = 4), this finding should be interpreted with caution.

An analysis of predictors of any MAE was conducted for both TAG and open surgery groups. Perioperatively, 28% of TAG and 70% of open DTA patients had at least one MAE (P < .001).¹¹ At 5 years, there was still a significant difference in the occurrence of MAEs, with 57.9% of TAG and 78.7% of open patients having an MAE at any point (log rank < .001).¹¹ In the open group UVA, increasing aneurysm diameter was found to predict any adverse event (P = .02). For TAG patients, a low preoperative hematocrit value predicted MAEs (P = .03), with increasing creatinine value (P = .057) and increasing aneurysm diameter (P = .082) trending toward significant. In MVA, using the entire cohort, open surgery treatment group (P < .001) and increasing aneurysm diameter (P < .001) predicted a significantly increased incidence of any major adverse event.

When asymptomatic aneurysm patients were analyzed as a separate cohort, univariate analysis in the open surgery group showed a history of smoking predicted any MAE. Multivariate analysis demonstrated a high creatinine value (P = .005) and short proximal neck length (P = .009) predicted events in the endovascular group. Multivariate analysis for all non-symptomatic DTA patients was significant for treatment group (open surgery predicting any MAE) (P =.01). Meaningful analysis of the asymptomatic group for risk factors for death was not performed due to the low incidence of this event.

Vascular complications of intraoperative hemorrhage, thrombosis, or acute ischemia were seen in 19% of TAG patients and 9% of open surgical patients (P = .038). Male patients had a 10% risk of perioperative vascular events (13/128) vs a 20% rate in women (21/106) (P = .01). Results of the univariate analysis demonstrated that decreasing weight (P = .007), endovascular treatment (P = .027), decreasing iliac diameter (P = .034), female gender (P = .037), increasing age (P =.038), and symptomatic aneurysms (P = .046) were all predictors of vascular complications. However, when put into a multivariate analysis, missing data from the open control group (iliac diameters) prevented a reliable model from being constructed for the entire cohort. Within the TAG treatment group only, gender (P = .01) was found to predict vascular complications. Age, body-mass index (BMI), iliac diameter, and symptomatic aneurysm were not found to have a predictive value. Longer procedure times were again found to be correlated with adverse events. TAG procedure times were significantly longer for women at a mean of 180 minutes (range, 65-580) vs 129 minutes (range, 61-290) for men (P = .001).

Graft-related events 

Graft-related events at 5 years were examined and analyzed for predictive factors. Table V lists these events from the operative period through long-term follow-up. The operative events have been previously described.⁹ The number of patients with an endoleak at any time was 10%, and at 5 years of follow-up 4.3% of patients had an active endoleak. A detailed description of the endoleaks, treatment, and all graft-related events in follow-up has been released.¹¹ In brief, 3 patients have undergone additional extension procedures for endoleak. Nineteen percent of patients have been noted to have sac enlargement at 5 years of follow-up, and no interventions have been performed for this. One instance of proximal migration was noted and was most likely due to a poor seal at the original procedure. This patient underwent successful conversion. Two additional complications during follow-up were infections due to aorto-esophageal fistula. Both of these patients died as a result of this condition. Analysis of the TAG cohort failed to demonstrate any predictive variable for these events, or for graft events in general.

Table V. Graft-related events at 5 years

Event	30(21%)⁎⁎
Branch vessel occlusion⁎	4(3%)
Deployment failure⁎	1(<1%)
Lumen obstruction⁎	1(<1%)
Device complication at treatment⁎	4(3%)
Endoleak (total number of patients with endoleak at any time)	15(10.6%)
Increase in diameter ≥5 mm	8(19%)
Prosthesis migration	1(<1%)
Prosthesis material failure	1(<1%)
Extrusion/erosion	0(0%)
Aneurysm rupture	0(0%)
Other complication at follow-up	2(1.4%)

Back to Article Outline

Discussion 

This is the first attempt to define predictors of morbidity and mortality in a combined open and endovascularly treated population of patients treated for descending thoracic aneurysms. We found that symptomatic aneurysms and male gender predicted death in the entire cohort of DTA repair patients. Although the multivariate analysis did not demonstrate a treatment group effect for mortality, this was thought to be related to a confounding effect of symptomatic aneurysms. The negative predictive effect of symptomatic aneurysms is not difficult to understand, even though none of these patients were thought to have an acute rupture. The lack of patient optimization in urgent surgery coupled with the potential hemodynamic instability and inflammatory cascade involved in an acute presentation is known to translate into worse outcomes.¹², ¹³ The fact that significantly more symptomatic aneurysms were found in the open treatment group does have a confounding effect with the treatment group. In univariate analysis, symptomatic aneurysms (P = .004) and open repair (P = .01) both predicted death. When constructing the multivariate model, symptomatic aneurysms had more weight due to the lower P value, and this, in part, negated the effect of treatment group. The strong predictive effect of male gender, however, is more surprising. There is ample literature describing worse outcomes after infrarenal abdominal aortic aneurysm (AAA) repairs in women vs men,¹⁴, ¹⁵, ¹⁶, ¹⁷, ¹⁸, ¹⁹ and women were noted in our series to have longer procedure times and a higher incidence of vascular complications. We attribute both the longer procedure times and the higher number of vascular complications to smaller iliac arteries found in women. Although there was an association with longer procedure times and all complications, this did not translate into increased mortality in women.

The predictor of serious morbidity in the entire cohort was found to be an elevated creatinine level, although no patients were in renal failure. There is strong evidence that even subtle elevations in creatinine can translate into adverse perioperative events.²⁰, ²¹ However, this finding is more meaningful in the TAG group as data for the open surgical patients did not always include creatinine values. We interpret these results to indicate that in patients with even subtle elevations in creatinine renal dysfunction may be a marker for systemic vasculopathy and that these patients should be warned that they may have a higher incidence of complications. Although procedure time was found to have a significant predictive effect for many adverse events, the authors determined that this was a marker for difficult surgery rather than a risk factor in and of itself. Additionally, long procedure times were strongly correlated with open surgery and would have altered comparisons within the entire DTA cohort, falsely biasing analyses in favor of the TAG group.

Open surgery and larger aneurysms were found to predict the occurrence of any adverse event in follow-up, and open surgery was the only reliable predictor of paraplegia in the entire DTA cohort. These associations are not surprising, but do lend weight to the argument in favor of TEVAR. This is especially compelling in that adverse events, including endoleaks, and sac enlargements were tabulated out to 5 years. The finding that even with late graft-related events there are significantly fewer MAEs with endovascular DTA repair illustrates that this is a durable procedure and less morbid than open surgery in appropriate patients. Unfortunately, no reliable predictors could be determined for late graft-related events.

The strengths of this study are that data was prospectively collected, and all complications have been verified by independent practitioners and a clinical events committee. There are a relatively large number of patients involved, although even with 234 patients we were unable to perform many important analyses due to the low number of index events or missing data points. Missing data points were primarily seen in the open surgical patients, especially historic controls. These restrictions lead to the formation of the composite morbidity point in order to be able to predict the most serious perioperative events, as each separate event happened too infrequently.

Unfortunately, this population was not randomized to open vs endovascular repair – a study that will most likely never be completed. We know that there were more symptomatic aneurysm patients in the open repair group, and this fact had a profound effect on the overall conclusions of our analysis due to the high death rate within this population. Although there was marginally more cardiac disease within the TAG group, we did not see significant differences in postoperative cardiac events between the 2 groups.

Back to Article Outline

Conclusion 

Morbidity and mortality within the TAG and open surgery cohorts were low for DTA repair. Symptomatic aneurysms and male gender were found to predict death overall, and both of these variables were found more often in the open surgery group. There was increased morbidity associated with an elevated creatinine for the entire cohort. Vascular complications were seen more often in women and with endovascular repair. Open surgery predicted any major adverse event and paraplegia.

Back to Article Outline

Author contributions 

Back to Article Outline

References 

1. Ehrlich M, Grabenwoeger M, Cartes-Zumelzu F, Grimm M, Petzl D, Lammer J, et al. Endovascular stent graft repair for aneurysms on the descending thoracic aorta. Ann Thorac Surg. 1998;66:19–24
   • View In Article
   • MEDLINE
   • CrossRef
2. Clouse WD, Hallett JW, Schaff HV, Gayari MM, Ilstrup DM, Melton LJ. Improved prognosis of thoracic aortic aneurysms: a population-based study. JAMA. 1998;280:1926–1929
   • View In Article
   • MEDLINE
   • CrossRef
3. Cambria RA, Gloviczki P, Stanson AW, Cherry KJ, Bower TC, Hallett JW, et al. Outcome and expansion rate of 57 thoracoabdominal aortic aneurysms managed nonoperatively. Am J Surg. 1995;170:213–217
   • View In Article
   • Abstract
   • Full-Text PDF (665 KB)
   • CrossRef
4. Cambria RP, Clouse WD, Davison JK, Dunn PF, Corey M, Dorer D. Thoracoabdominal aneurysm repair: results with 337 operations performed over a 15-year interval. Ann Surg. 2002;236:471–479
   • View In Article
   • MEDLINE
   • CrossRef
5. Hamerlijnck RP, Rutsaert RR, De Geest R, Brutel de la Riviere A, Defauw JJ, Vermeulen FE. Surgical correction of descending thoracic aortic aneurysms under simple aortic cross-clamping. J Vasc Surg. 1989;9:568–573
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (511 KB)
   • CrossRef
6. Huynh TTT, Miller CC, Estrera AE, Porat EE, Safi HJ. Thoracoabdominal and descending thoracic aortic aneurysm surgery in patients aged 79 years or older. J Vasc Surg. 2002;36:469–475
   • View In Article
   • Abstract
   • Full-Text PDF (111 KB)
   • CrossRef
7. Demers P, Miller DC, Mitchell RS, Kee ST, Sze D, Razavi MK, et al. Midterm results of endovascular repair of descending thoracic aortic aneurysms with first-generation stent grafts. J Thorac Cardiovasc Surg. 2004;127:664–673
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (283 KB)
   • CrossRef
8. Bavaria JE, Appoo JJ, Makaroun MS, Verter J, Yu ZF, Mitchell RS Gore TAG Investigators. Endovascular stent grafting versus open surgical repair of descending thoracic aortic aneurysms in low-risk patients: a multicenter comparative trial. J Thorac Cardiovasc Surg. 2007;133:369–377
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (142 KB)
   • CrossRef
9. Makaroun MS, Dillavou ED, Kee ST, Sicard G, Chaikof E, Bavaria JE, et al. Endovascular treatment of thoracic aortic aneurysms: results of the phase II multicenter trial of the GORE TAG thoracic endoprosthesis. J Vasc Surg. 2005;41:1–9
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (302 KB)
   • CrossRef
10. Sacks D, Marinelli DL, Martin LG, Spies JB. Reporting standards for clinical evaluation of new peripheral arterial revascularization devices. J Vasc Interv Radiol. 1997;8:137–149
    • View In Article
    • Full-Text PDF (1328 KB)
    • CrossRef
11. Makaroun MS, Dillavou ED, Wheatley GH, Cambria RP. Five-year results of endovascular treatment compared to open repair of thoracic aortic aneurysms. J Vasc Surg. 2008;47:912–918
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (489 KB)
    • CrossRef
12. Godet G, Fiqon M-H, Vicaut E, Zubicki A, Bertrand M, Riou B, et al. Risk factors for acute postoperative renal failure in thoracic or thoracoabdominal aortic surgery: a prospective study. Anesth Analg. 1997;85:1227–1232
    • View In Article
    • MEDLINE
    • CrossRef
13. Yadi T, Atay Y, Cikirikciolu M, Boa M, Posaciolu H, Ozbaran M, et al. Determinants of early mortality and neurological morbidity in aortic operations performed under circulatory arrest. J Card Surg. 2000;15:186–193
    • View In Article
    • MEDLINE
    • CrossRef
14. Dimick JB, Stanley JC, Axelrod DA, Kazmers A, Henke PK, Jacobs LA, et al. Variation in death rate after abdominal aortic aneurysmectomy in the United States impact of hospital volume, gender, and age. Ann Surg. 2002;235:579–585
    • View In Article
    • MEDLINE
    • CrossRef
15. Akkersdijk GJM, Prinssen M, Blankensteijn JD. The impact of endovascular treatment on in-hospital mortality following non-ruptured AAA repair over a decade: a population based study of 16,446 patients. Eur J Vasc Endovasc Surg. 2004;28:41–46
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (338 KB)
    • CrossRef
16. Dueck AD, Kucey DS, Johnston KW, Alter D, Laupacis A. Survival after ruptured abdominal aortic aneurysm: effect of patient, surgeon, and hospital factors. J Vasc Surg. 2004;39:1253–1260
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (118 KB)
    • CrossRef
17. Semmens JB, Norman PE, Lawrence-Brown M, Holman CD. Influence of gender on outcome from ruptured abdominal aortic aneurysm. Brit J Surg. 2000;87:191–194
    • View In Article
18. Dillavou ED, Muluk SC, Makaroun MS. A decade of change in AAA repair in the US: have we improved outcomes equally among men and women?. J Vasc Surg. 2006;43:230–238
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (244 KB)
    • CrossRef
19. The U.K. Small Aneurysm Trial ParticipantsBrown LC, Powell JT. Risk factors for aneurysm rupture in patients kept under ultrasound surveillance. Ann Surg. 1999;230:289–297
    • View In Article
    • MEDLINE
    • CrossRef
20. Sugawara Y, Sato O, Miyata T, Deguchi J, Kimura H, Namba T, et al. Surgical results of abdominal aortic aneurysm repair in patients with chronic renal dysfunction. Japan Circ J. 1997;61:762–766
    • View In Article
21. Yusuf K, Murat B, Unal A, Ulku K, Taylan K, Ozerdem O, et al. Inflammatory abdominal aortic aneurysm: predictors of long-term outcome in a case-control study. Surg. 2007;141:83–89
    • View In Article