|
|
Risk Factors Of Death After Thoracic Endovascular Aortic Repair: A Multivariate Analysis.
A. Khoynezhad1, J. Selzle1, C. E. Donayre2, R. A. White2. 1University Of Nebraska Medical Center, Omaha, NE, 2Harbor-UCLA Medical Center, Torrance, CA,
BACKGROUND:
Thoracic endovascular aortic repair is an emerging option to open repair in a selected population with aortic pathologies. The risk factors associated with early and late mortality following thoracic endovascular aortic repair are poorly understood. The aim of this study is to analyze the risk factors associated with death following Thoracic endovascular aortic repair.
METHODS:
153 patients underwent 184 thoracic endovascular aortic repairs between 1998 and 2005 through a Single-Institution-Investigator Investigational Device Exemption approved by the Food and Drug Administration. Computer tomography scans, angiograms and medical records were reviewed. Statistical analysis was performed using SPSS 13.0 software. Fisher’s Exact Test and Cox Regression were performed for univariate analysis of risk factors for early and late mortality after thoracic endovascular aortic repair. Variables with p-values less than 0.05 in the univariate analysis were included in the stepwise model selection procedure of multivariable regression analysis.
RESULTS:
The underlying pathologies included descending thoracic aortic aneurysm (n=91), acute Type B aortic dissection (n=25), chronic Type B aortic dissection (n=42), aortic transection (n=12), and penetrating aortic ulcer (n=14). Thoracic endovascular aortic repair was technically successful in all but three patients. Another three required an open repair within the first month. 4.3% (8/184) of patients developed stroke, and another eight patients developed paraplegia/paraperesis. Early and late mortality was 9.8% (n=18), and 19% (n=35) in a 16-month average period of follow-up, respectively. The results of univariate analysis of variables associated with early mortality are represented in the following tables. Procedural endoleak was the only significant predictor of early death in the multivariate model (P-value: 0.0036; odds ratio: 8.4; 95% confidence interval: 1.6-43.9).
Predictors of late death in the univariate analysis were coronary artery disease, congestive heart failure, chronic obstructive pulmonary disease, peripheral vascular disease, intraoperative vascular injury or thrombosis, intraoperative or postoperative bleeding, prolonged ICU stay, postoperative myocardial infarction and acute renal failure. Multivariate Cox regression revealed chronic obstructive pulmonary disease (P-value: 0.024; odds ratio: 3.8; 95% confidence interval: 1.2-12.1), postoperative myocardial infarction (P-value: 0.0053; odds ratio: 9.7; 95% confidence interval: 2.0-48.4), and acute renal failure (P-value: 0.0006; odds ratio: 22.8; 95% confidence interval: 3.8-137.6) to be independent risk factors for late mortality.
CONCLUSIONS:
Procedural type I endoleak is an independent risk factor of early mortality following Thoracic endovascular aortic repair. Chronic obstructive pulmonary disease, postoperative myocardial infarction and acute renal failure are predictors of late death in the multivariate analysis.
Univariate Analysis for Early Mortality (Fisher’s Exact Test)| Variable | Description | Alive at One Month | Early Mortality | P-value | | Peripheral vascular disease | No; Yes | 93 (93%); 13 (72%) | 7 (7%); 5 (28%) | 0.019 | | General anesthesia | No; Yes | 72 (94%); 39 (81%) | 5 (6%); 9 (19%) | 0.044 | | Intraoperative hypotension | No; Yes | 61 (94%); 5 (63%) | 4 (6%); 3 (37%) | 0.025 | | Procedural endoleak | None; Type 1 | 84 (94%); 5 (63%) | 6 (6%); 3 (37%) | 0.024 | | Blood transfusion | No; Yes | 60 (95%); 17 (74%) | 3 (5%); 6 (26%) | 0.0098 | | postoperative myocardial infarction | No; Yes | 78 (91%); 3 (37%) | 8 (9%); 5 (63%) | 0.001 | | Postoperative bleeding | No; Yes | 78 (90%); 2 (40%) | 9 (10%); 3 (60%) | 0.015 | | Vascular injury | No; Yes | 77 (92%); 4 (50%) | 7 (8%); 4 (50%) | 0.0063 | | ICU days | 0-1; >=2 | 37 (97%); 41 (82%) | 1 (3%); 9 (18%) | 0.039 |
Univariate Analysis for Late Mortality (Cox Regression)| Variable | Description | Hazard ratio | 95% Confidence Interval | P-value | | ICU days | 0-1; >=2 | 1; 3.2 | -; 1.1 - 8.8 | 0.026 | | Coronary artery disease | No; Yes | 1; 2.7 | -; 1.3 - 5.6 | 0.0067 | | Congestive heart failure | No; Yes | 1; 3.3 | -; 0.1 - 2.7 | 0.0068 | | Peripheral vascular disease | No; Yes | 1; 3.1 | -; 1.3 - 7.5 | 0.009 | | chronic obstructive pulmonary disease | No; Yes | 1; 3.2 | -; 1.4 - 7.1 | 0.0045 | | Blood loss (mL) | < 400; >=400 | 1; 2.6 | -; 1.1 - 6.5 | 0.039 | | postoperative myocardial infarction | No; Yes | 1; 8.8 | -; 3.6 - 21.6 | 0.0009 | | Postoperative bleeding | No; Yes | 1; 4.9 | -; 1.4 - 17.6 | 0.014 | | Vascular injury | No; Yes | 1; 4.2 | -; 1.5 - 11.8 | 0.0058 |
Back to 33rd Annual Meeting
Back to Program Outline
|
