Outcomes of endovascular abdominal aortic aneurysm repair compared with open surgical repair in high-risk patients: Results from the Swedish Vascular Registry
Article Outline
- Abstract
- Patients and methods
- Results
- Discussion
- Conclusion
- Author contributions
- Acknowledgment
- References
- Copyright
Background
The management of infrarenal aortic aneurysms in high-risk patients remains a challenge. Endovascular aneurysm repair (EVAR) is associated with superior short-term mortality rates but unclear long-term results and has not been shown to improve survival in patients unfit for open repair (OR). The aim of this population-based study was to evaluate the outcome after elective EVAR compared with OR in a high-risk patient cohort.
Methods
Prospectively collected data from January 2000 to December 2006 were retrieved from the Swedish Vascular Registry. The high-risk cohort was defined as age ≥60 years, American Anesthesiologists Association (ASA) class 3 or 4, and at least one cardiac, pulmonary, or renal comorbidity. These criteria were met by 217 of 1000 EVAR patients and 483 of 2831 OR patients. Primary end points were 30-day and 1-year all-cause mortality. Kaplan-Meier curves for survival and multivariate Cox regression analyses were performed.
Results
The crude 30-day and 1-year all-cause mortality rates for EVAR vs OR for the whole treatment group (n = 3831) were 1.8% vs 2.8% and 8.0% vs 7.2%, respectively. In the high-risk cohort (n = 700), EVAR patients were approximately 2 years older and renal insufficiency and diabetes mellitus were more common, and smoking was more prevalent in the OR group. About two-thirds of EVAR procedures were performed at university hospitals and one-half of OR procedures were performed at county hospitals. In the high-risk cohort, there was no difference in mortality at 30-days (EVAR, 4.6% vs OR, 3.3%), but OR had lower 1-year mortality (8.5% vs 15.9%; P = .003). More bleeding complications occurred in the EVAR group, but more pulmonary complications occurred in the OR group; however, there was no difference in cardiac, cerebrovascular, or renal complications. The mean follow-up was 3.4 years. EVAR was associated with increased mortality risk after adjusting for age, ASA class, and comorbidities (hazard ratio, 1.50; 95% confidence interval, 1.07-2.12; P = .02). Kaplan-Meier survival analysis showed a lower mortality rate for patients undergoing OR, which remained during follow-up (P = .001).
Conclusions
Elective OR of aortic aneurysms seems to have a better outcome compared with EVAR in this specific, population-based, high-risk patient cohort after adjusting for covariates. We cannot confirm the benefit of EVAR from previous registry studies with a similar high-risk definition. In clinical practice, OR may be at least as good as EVAR in high-risk patients fit for surgery.
Endovascular aneurysm repair (EVAR) has rapidly been integrated into routine clinical care worldwide for the treatment of abdominal aortic aneurysm (AAA). The minimally invasive endovascular technology that was originally intended to expand the treatment options for patients deemed not good candidates for open surgical repair has evolved substantially since its introduction.1 The European randomized trials have provided evidence to show that EVAR is associated with lower operative mortality than open repair (OR) for treatment of AAA, but no difference in midterm all-cause mortality.2, 3
Endovascular repair providing a less invasive alternative has been considered an attractive treatment for high-risk individuals with AAA. For patients who are unfit for OR, the randomized EVAR-2 trial did not find a survival benefit from EVAR compared with observation.4 However, high crossover and procedural mortality rates have led to controversy regarding the validity of this trial. Recent registry reviews have shown that a selected cohort of high-risk patients does well with EVAR compared with OR.5, 6 This study used a large, population-based prospective vascular registry to determine the operative mortality and long-term survival of elective EVAR compared with OR in high-risk patients.
Patients and methods
Prospectively collected data from January 2000 to December 2006 for all patients who underwent elective EVAR or OR of nonruptured infrarenal aortic aneurysms were retrieved from the Swedish Vascular Registry (Swedvasc). The registry has national coverage and includes all centers performing EVAR and OR in Sweden: 36 hospitals were performing OR and 17 hospitals performing EVAR. Prospective data on basic demographics and risk factors, together with postoperative 30-day and 1-year outcomes, were registered. All patients were cross-matched with the National Population Registry in March 2008 to update mortality data, including date of death. Ethical approval for the study was obtained from the Registry Steering Committee, which is the authority for research based on the registry data. Each patient gave informed consent before registration.
Sweden is divided in health service regions; each includes several counties and has a central hospital (university hospital). Each county within a region has a county hospital with up to 700 beds and with specialized and outpatient facilities to serve a population of about 300,000. The counties, in turn, are divided into districts; each has a population of about 75,000 and is served by a district hospital, which usually has <300 beds.
High-risk was defined, in line with recent registry studies,5, 6 as age ≥60 years, American Society of Anesthesiology (ASA) physical status classification 3 or 4, and at least one cardiac, pulmonary, or renal comorbidity. ASA class 3 is a patient with severe systemic disease, and ASA class 4 is a patient with severe systemic disease that is a constant threat to life.7 Risk factors had the same definitions throughout the study period. Cardiac disease is defined as previous myocardial infarction, atrial fibrillation, heart failure, angina pectoris, coronary artery bypass graft surgery, surgery for valvular heart disease or signs of myocardial ischemia on electrocardiogram; pulmonary disease, diagnosed pulmonary disease such as chronic obstructive pulmonary disease (COPD) or emphysema; renal insufficiency, a serum creatinine level >150 μmol/L (>1.7 mg/dL); hypertension, diastolic blood pressure >110 mm Hg or antihypertensive medication; diabetes mellitus, antidiabetic treatment with diet, oral hypoglycemics, or insulin; cerebrovascular disease, previous stroke or transient ischemic attack; previous vascular surgery, previous open or endovascular procedure or amputation for peripheral vascular disease; hyperlipidemia, cholesterol value >2 standard deviations from normal value; statins, statin treatment >1 month preoperatively; and smoking, regular smoking for the past 5 years.
The outcome measures were (1) 30-day operative mortality, defined as death during the initial hospitalization or death from any cause ≤30 days of the index procedure; (2) postoperative complications within ≤30 days; and (3) long-term survival and all-cause mortality.
Statistical methods
Statistical analyses were performed using SPSS software (SPSS Inc, Chicago, Ill). Continuous demographic variables were expressed as mean ± standard deviation and categoric variables as percentages. The Mann-Whitney U test was used for continuous variables and the Pearson χ2 test for categoric variables. Kaplan-Meier curves, using the log-rank test, were used to compare crude cumulative survival between EVAR and OR. We compared OR and EVAR by calculating adjusted hazard ratios (HR) using a Cox proportional hazards model and adjusted for potential confounders (age at operation, sex, and major comorbidities). Visual inspection of log–log plots indicated that the proportional hazard assumption was not violated. All HRs were reported with 95% confidence intervals (CIs). All tests were two-tailed, with a P = .05 judged statistically significant.
Results
Patient population
During the study period between 2000 and 2006, 3831 patients undergoing elective repair of nonruptured infrarenal aortic aneurysm were identified from the registry, of which 2831 had OR and 1000 had EVAR. The annual volume per hospital varied, with OR at 0 to 52 cases per year and EVAR at 0 to 80 cases per year. The yearly number of EVAR and OR is summarized in Table I. The high-risk criteria narrowed the cohort to 700 patients, consisting of 217 EVAR and 483 OR.
Table I. The number of endovascular and open aortic aneurysm repairs per year during the study period
| Year | EVAR | OR |
|---|---|---|
| 2000 | 103 | 416 |
| 2001 | 440 | 96 |
| 2002 | 395 | 101 |
| 2003 | 377 | 117 |
| 2004 | 449 | 145 |
| 2005 | 373 | 166 |
| 2006 | 381 | 272 |
| Total | 1000 | 2831 |
Table II reports the baseline characteristics for all patients undergoing AAA repair, and Table III summarizes data for the high-risk cohort. In the high-risk cohort, it is noteworthy that EVAR patients were approximately 2 years older and renal insufficiency and diabetes mellitus were more common, and smoking was more prevalent in the OR group. Half of the patients were taking statins and about one-third had at least two high-risk comorbidities. About two-thirds of EVAR procedures were performed at university hospitals, and one-half of OR procedures were performed at county hospitals. The mean follow-up time was 3.4 ± 2.0 years for the high-risk cohort, and all patients were followed up at least 1 year for survival.
Table II. Baseline characteristics for 3831 patients undergoing endovascular or open abdominal aortic aneurysm repair during the study period
| Characteristics | EVAR (n = 1000) | OR (n = 2831) | Pa |
|---|---|---|---|
| Demographic factors | |||
 Age, mean (SD), y | 74±7 | 71±8 | <.0001 |
| Age >60 y, % (No.) | 96(965/1000) | 92(217/2831) | <.0001 |
| Male sex, % (No.) | 86(861/1000) | 82(2308/2831) | .001 |
| Setting, % (No.) | |||
| District hospital | 6(58/1000) | 12(332/2831) | |
| County hospital | 23(233/1000) | 50(1412/2831) | <.0001 |
| University hospital | 71(709/1000) | 38(1087/2831) | |
| Risk factors, % (No.) | |||
| Cardiac disease | 59(548/928) | 53(1419/2682) | .001 |
| Pulmonary disease | 20(178/890) | 19(497/2640) | .44 |
| Renal impairment | 15(134/891) | 10(257/2616) | <.0001 |
| Any of cardiac, pulmonary, renal | 63(629/1000) | 58(1636/2831) | .005 |
| Cerebrovascular disease | 16(145/883) | 14(373/2627) | .11 |
| Diabetes mellitus | 13(117/888) | 8(202/2618) | <.0001 |
| Hypertension | 61(555/912) | 62(1661/2663) | .42 |
| Smoking | 35(308/877) | 50(1254/2534) | <.0001 |
| Hyperlipidemia | 38(161/424) | 43(562/1319) | .092 |
| Statin treatment | 17(155/898) | 14(380/2628) | .043 |
| Previous vascular surgery | 44(179/407) | 44(501/1150) | .43 |
| ASA class | |||
| I | 9(58/666) | 10(208/2033) | |
| II | 52(349/666) | 57(1162/2033) | .016 |
| III | 36(240/666) | 31(627/2033) | |
| IV | 3(19/666) | 2(36/2033) |
aCalculated using the Mann-Whitney U test for continuous variables and the χ2 test for categoric variables, unless otherwise specified. |
Table III. Baseline characteristics for 700 high-risk patients undergoing endovascular or open abdominal aortic aneurysm repair during the study period
| Variables | EVAR (n = 217) | OR (n = 483) | Pa |
|---|---|---|---|
| Demographics | |||
| Age, mean (SD), y | 75±6 | 73±6 | <.0001 |
| Male sex, % (No.) | 85(184/217) | 83(401/483) | .56 |
| Setting, % (No.) | |||
| District hospital | 4(8/217) | 14(65/483) | |
| County hospital | 34(74/217) | 51(248/483) | <.0001 |
| University hospital | 62(135/217) | 35(170/483) | |
| Risk factors, % (No.) | |||
| Cardiac disease | 87(188/215) | 88(414/473) | .98 |
| Pulmonary disease | 32(69/213) | 34(159/463) | .62 |
| Renal impairment | 26(56/213) | 19(86/455) | .030 |
| High-risk comorbidity, No.b | |||
| 1 | 62(134/217) | 67(326/483) | .24 |
| 2 | 32(70/217) | 29(138/483) | |
| 3 | 6(13/217) | 4(19/483) | |
| Cerebrovascular disease | 21(44/212) | 18(81/454) | .37 |
| Diabetes mellitus | 16(33/211) | 9(42/454) | .015 |
| Hypertension | 72(154/214) | 71(327/461) | .78 |
| Smoking | 34(66/192) | 52(229/440) | <.0001 |
| Hyperlipidemia | 47(40/85) | 59(136/231) | .061 |
| Statin treatment | 47(62/133) | 50(115/228) | .48 |
| Previous vascular surgery | 22(47/213) | 18(80/455) | .17 |
| ASA class | |||
| III | 92(199/217) | 95(459/483) | .087 |
| IV | 8(18/217) | 5(24/483) |
aCalculated using the Mann-Whitney U test for continuous variables and the χ2 test for categoric variables, unless otherwise specified. |
bNumber of cardiac, pulmonary and renal risk factors present. |
Outcome measures
The respective crude 30-day operative and 1-year all-cause mortality rates for the 3831 treated patients for EVAR vs OR were 1.8% vs 2.8% (P = .08) and 8.0% vs 7.2% (P = .15; Fig 1). There was no statistical difference in 30-day mortality looking at type of hospital (OR: university 3.0%, county 2.6%, district 3.0% [P = .81]; EVAR: university 2.1%, county 1.3%, district 0% [P = .40]).
Fig 1.
Kaplan-Meier curve of survival for all patients with abdominal aortic aneurysm undergoing open (OR, dashed line) or endovascular (EVAR, solid line) repair. Error bars denote 95% confidence intervals.
In the high-risk cohort (n = 700), operative mortality at 30-days was similar, with EVAR at 4.6% compared with OR at 3.3% (P = .40; Fig 2). There was no significant difference in reported 30-day major complications (Table IV). More bleeding complications occurred in the EVAR group and the OR group had more pulmonary complications; however, no difference was noted in cardiac, cerebrovascular, or renal complications. A prolonged intensive care unit stay (>5 days) was more common in patients undergoing OR.
Fig 2.
Kaplan-Meier curve of survival for high-risk patients with abdominal aortic aneurysm undergoing open (OR, dashed line) or endovascular (EVAR, solid line) repair. Error bars denote 95% confidence intervals.
Table IV. Reported 30-day complications and freedom from 30-day complications in the high-risk abdominal aortic aneurysm group
| Variable | OR (n = 483), % No. | EVAR (n = 217), % (No.) | Pa |
|---|---|---|---|
| 30-day complicationsb | |||
| Bleeding | 5.0(24/479) | 9.2(19/206) | .037 |
| Ileus | 1.0(5/479) | 0(0/206) | .14 |
| Bowel ischemia | 2.5(12/479) | 1.5(3/206) | .39 |
| Bowel resection | 0.4(2/479) | 1.0(2/206) | .38 |
| Superficial infection | 2.9(14/479) | 2.4(5/206) | .72 |
| Deep infection | 0.4(2/479) | 0.5(1/206) | .90 |
| Sepsis | 0.4(2/479) | 1.0(2/206) | .38 |
| Wound dehiscence | 4.0(19/479) | 0.5(1/206) | .013 |
| Resuture of wound | 3.1(15/479) | 0(0/206) | .010 |
| Drainage | 0.6(3/479) | 1.5(3/206) | .28 |
| Graft occlusion | 3.5(17/479) | 4.9(10/206) | .42 |
| Distal embolization | 4.2(20/479) | 2.9(6/206) | .43 |
| Fasciotomy | 0.8(4/479) | 0(0/206) | .19 |
| Major amputation | 0.8(4/479) | 1(2/206) | .86 |
| Minor amputation | 0.4(2/479) | 0(0/206) | .35 |
| Relaparotomy | 2.7(13/479) | 0.5(1/206) | .059 |
| Cerebrovascular | 1.9(9/479) | 0.5(1/206) | .16 |
| Cardiac | 6.5(31/479) | 6.3(13/206) | .94 |
| Pulmonary | 6.9(33/479) | 1.0(2/206) | .001 |
| Renal | 4.2(20/479) | 5.8(12/206) | .35 |
| Venous thromboembolism | 0.2(1/479) | 0(0/206) | .51 |
| Multiple organ failure | 1.7(8/479) | 1.5(3/206) | .84 |
| ICU stay >5 days | 6.3(30/479) | 1.5(3/206) | .007 |
| Freedom from 30-day complication | |||
| No surgical complicationc | 78(374/479) | 80(165/206) | .44 |
| No general complicationd | 76(365/479) | 82(168/206) | .12 |
| No reoperatione | 90(432/479) | 96(197/206) | .017 |
| No complication or reoperation | 62(298/479) | 64(132/206) | .64 |
| No major complicationf | 81(386/479) | 85(175/206) | .17 |
aP was calculated using χ2 test. |
bRegistration of complications was missing in 4 OR patients and 11 EVAR patients. |
cSurgical complication includes bleeding, ileus, superficial and deep infection, graft occlusion, wound dehiscence, distal embolization, and bowel ischemia. |
dGeneral complication includes cerebrovascular, cardiac, pulmonary, renal, sepsis, multiorgan failure, venous thromboembolism, and ICU stay >5 days. |
eReoperation includes minor and major amputations, drainage, fasciotomy, resuture of wound, bowel resection, and relaparotomy. |
fMajor complication includes bowel ischemia or resection, cerebrovascular, cardiac, multiorgan failure, pulmonary, renal, sepsis, or major amputation. |
The 1-year mortality rate for OR was 8.5%, which was significantly lower compared with 15.9% for EVAR (P = .003; Fig 2). Kaplan-Meier survival analysis showed a lower mortality rate for patients undergoing OR, which remained during follow-up (P = .001). The estimated overall mortality at 4 years was 26% for OR and 41% for EVAR. EVAR was associated with increased mortality risk after adjusting for age, ASA class, and comorbidities (HR, 1.50; 95% CI, 1.07-2.12; P = .02).
Discussion
The significant mortality rate associated with aneurysm repair and a trend toward minimally invasive techniques has led to a rapid development of the endovascular technique for AAA treatment. A natural indication for the endovascular approach was assumed to be patients who are at higher risk from OR. However, this population-based study found a better long-term survival for high-risk patients undergoing elective OR of nonruptured infrarenal aortic aneurysms compared with EVAR. There was no difference between treatments in terms of operative mortality and major 30-day complications in the high-risk cohort.
When we compared our 30-day operative mortality rate for the whole patient cohort with the randomized EVAR-12 and Dutch Randomised Endovascular Aneurysm Management (DREAM)3 trials, there seemed to be a similar mortality rate for EVAR (1.8% vs 1.2% to 1.7%) but a lower mortality rate for OR (2.8% vs 4.6% to 4.7%) in our cohort. We still found a short-term survival advantage for the EVAR group, but this did not reach significance because of the low mortality rate for OR. When we looked at long-term outcome, there was no difference between treatment groups in the whole patient cohort, and the results were similar compared with the large randomized studies. The trend towards an early survival advantage for the EVAR group was lost after 12 months of follow-up.
Two recent large registry studies examined the results of EVAR and OR in high-risk patients with AAA.5, 6 Using the Department of Veteran Affairs National Surgical Quality Improvement Program database, Bush et al5 identified a high-risk cohort of 2368, including veterans (99% men) aged ≥60 years, ASA classification 3 or 4, and comorbidity variables of history of cardiac, respiratory, or hepatic disease; cardiac revascularization, renal insufficiency, or low serum albumin level. They found the EVAR mortality rate to be significantly lower than OR at 30 days (3.4% vs 5.2%) and at 1 year (9.5% vs 12.4%).
The Outcomes Committee of the Society for Vascular Surgery selected a cohort of 626 high-risk patients from five multicenter investigational device exemption clinical trials leading to Food and Drug Administration (FDA) approval.6 High-risk in this study was defined as age ≥60 years with aneurysm size ≥5.5 cm plus at least one cardiac, pulmonary, or renal comorbidity. When they compared the EVAR and OR groups, they found no significant difference in operative mortality (EVAR, 2.9%; OR, 5.1%), AAA-related death, or overall survival at 1 year (EVAR, 87%; OR, 86%) or 4 years (EVAR, 56%; OR, 66%). Limitations in this study were a low number of 61 patients in the OR group, and 75% of patients classified as high-risk had only one comorbidity, whereas <1% had all three comorbidities.
Smaller retrospective registry studies in high-surgical-risk patients undergoing EVAR have shown operative mortality rates of 4.3% to 5% and a 3-year survival of 70% to 85%.8, 9 A recent analysis of patients from the United Kingdom (UK) EVAR trials, using a modified version of the Customized Probability Index to allocate fitness scores for all patients, found EVAR convincingly to favor only the good-fitness group in terms of 30-day mortality.10 For midterm survival, no benefit was found for either EVAR or OR across all fitness scores.
The EVAR 2 trial included 338 patients deemed unfit for OR, with a minimum age of 60 years and aneurysm diameter ≥5.5 cm, and is the only randomized trial that compared EVAR with observation.4 The study reported no difference in all-cause or AAA-related mortality with a trend favoring observation. The trial was complicated by long delays in EVAR after randomization and a 27% patient crossover rate from the no intervention group. The 30-day operative mortality rate in the EVAR group for elective cases was 7%, and the overall mortality rate after 4 years was 64%.
When compared with the two large registry studies from Bush et al5 and Sicard et al,6 the 30-day operative mortality rate in the present study seems to be higher for EVAR and lower for OR. The lower operative mortality for OR in general might reflect a centralization of AAA treatment to larger hospitals and procedures performed predominantly by vascular surgeons. Other possible explanations are selection of patients with a better risk profile, and improved anesthesia and surgical techniques as well as postoperative care. Most of the EVAR procedures were performed at university hospitals. The early survival advantage usually seen for EVAR was not present in this high-risk cohort. We do not believe that the loss of early survival advantage for EVAR is explained by learning curve because EVAR was introduced in the mid-1990s in Sweden. Also, low-volume centers did not in general have a worse 30-day mortality rate compared with high-volume centers (data not shown).
The overall 1-year mortality in the present study is higher for EVAR compared with the other registry studies. This was clearly surprising given that we used the same high-risk definitions. The advantage of this population-based registry is its external validity (ie, the applicability of its results to the defined population). Looking at the EVAR 2 trial, the 30-day and 1-year mortality rates for EVAR in our study were lower. However, we do not intend to compare our results with the EVAR 2 trial, which included a different subpopulation of patients, those unfit for OR. Patients in the registry studies undergoing EVAR could be suitable for OR and thus not comparable with patients unfit for OR. The registry studies are also lacking a surveillance arm for comparison.
Endovascular repair was associated with increased mortality risk after adjusting for age, ASA class, and risk factors. Cardiac complications are the most common serious perioperative complication of EVAR and the most common cause of late death.4 There was no difference in cardiac risk factors or ASA classification between the EVAR and the OR group in this registry, but we did not have access to the cause of death. All-cause mortality was reported, which is very accurate due to every citizen's personal identity code registered in the National Population Registry. Regarding aneurysm-related death, we believe it is an unreliable end point considering today's low autopsy rate and inaccurate recordings outside the hospital. Aneurysm-related deaths are not common beyond the perioperative period, and patients generally die from causes not related to their aneurysm.5 There was also no significant difference in 30-day major complications between treatment groups.
We can only speculate about causes responsible for the increased 1-year mortality rate after EVAR in this cohort. The effect of contrast-induced nephropathy (CIN) in high-risk patients undergoing EVAR is not known. This is an important complication that accounts for a significant number of cases of hospital-acquired renal failure, with adverse effects on prognosis and health care costs.11 The EVAR procedure itself, with intraluminal manipulation of the renal arteries, device placement across the renal ostia, and contrast administration during the procedure, as well as repeated contrast-infused computed tomography surveillance could possibly all together worsen renal function in high-risk patients. A higher incidence of in-hospital and late cardiovascular events as well as death has been reported when CIN develops.12 At 1 year, the cumulative rate of major adverse cardiac events was significantly higher in patients with CIN according to a study by Dangas et al.13 The risk of CIN is elevated and of clinical importance in patients with chronic kidney disease, particularly when diabetes mellitus is also present. In this study, it is noteworthy that high-risk EVAR patients had significantly more diabetes mellitus and renal insufficiency than OR patients.
Mills et al14 recently showed significantly greater renal function decline for EVAR compared with OR during long-term follow-up. Age >70 years strongly correlated with serial renal function decline in all patients. A decreased glomerular filtration rate in elderly patients is an independent predictor of adverse outcomes, including not only complications related to chronic kidney disease but also cardiovascular events and death.15
We did not have information on size of the aneurysms or complexity of aneurysm anatomy. Zarins et al16 reported that patients with large AAAs (≥6.0 cm) have shorter life expectancy and have a higher risk of rupture, surgical conversion, and aneurysm-related death after EVAR compared with patients with smaller aneurysms. The general consensus in Sweden is to perform AAA repair when the aneurysm has reached a diameter of 5.0 to 5.5 cm, and we have no reason to believe that the size would be different between the treatment groups.
Several validated scoring systems have been developed for risk stratification of aneurysm repair.17 However, there is no validated scoring system for high-risk patients with AAA. Risk factors associated with excessive surgical risk in patients with AAA and frequent variables in scoring systems are age and cardiac, pulmonary, and renal comorbidities.18 The definition of high-risk patients has, unfortunately, varied in different studies. The differences in outcome among studies might reflect the lack of consistent definition of high-risk patients. For comparison reasons, we tried to use definitions similar to those used in previous registry studies, including age, ASA class, and cardiac, pulmonary, and renal risk factors. The EVAR 2 trialists took a pragmatic approach to fitness for OR but noted details of respiratory, renal, and cardiac risk. A problem with the ASA classification is, despite its simplicity, a possibility for wide interpretation by the surgeon entering the variable. A consistent, internationally accepted definition of risk factors would facilitate comparisons between registries and trials in the future.
These data represent all hospitals performing AAA repair covering a population of 9 million. Registry data have several inherent limitations, but Swedvasc has been validated on several occasions.19, 20, 21 Validity control has been performed comparing data with computerized anesthesia registries and by refilling random samples of protocols. More than 90% of open and endovascular arterial procedures performed in the country are reported to the registry compared with the National Inpatient Registry that is used for health care development, research, and planning. Data monitoring by independent outside assessors have not been undertaken. Accurate updated survival data are obtained every week by cross checking with the National Population Registry.
The registry has no data on patients not undergoing EVAR or OR. Different selection criteria for surgery could possibly indicate that less healthy patients are not chosen for OR. However, the annual operative frequency for AAA repair is not lower in Sweden compared with several other countries. Baseline risk factors are prospectively registered, but the reliability depends on the accuracy of the responsible surgeon while completing the protocol. Most risk factors have a >95% registration rate. Analyses of cost-effectiveness and quality of life issues were not performed in this study.
Conclusion
Patients deemed fit for OR have a better long-term outcome compared with patients deemed fit and suitable for EVAR in this high-risk cohort. We cannot confirm the benefit of EVAR from previous registry studies with a similar high-risk definition. In clinical practice, OR may be at least as good as EVAR in high-risk patients fit for surgery. If the patient is at lower risk, EVAR tends to reduce operative death, but this early survival advantage disappears after 1 year. We believe that EVAR for infrarenal AAAs is the first choice in line with the large randomized studies, but there could be subgroups, such as high-risk patients, that benefit more from OR than EVAR. From this study we cannot draw any conclusions about patients unfit for surgery. Improved criteria for patient selection, including risk stratification of consistent defined comorbidities, will better help us to find those patients that will benefit from open or endovascular AAA treatment.
Author contributions
We thank Professor Jesper Swedenborg for advice and for providing helpful comments. All vascular surgeons in Sweden who meticulously registered their procedures are gratefully acknowledged for their effort. The Swedvasc Steering Committee: Anders Lundell (chairman), Ken Eliasson, Claes Forsell, Ingvar Jansson, Lars Karlström, Björn Kragsterman, Becke Lundkvist, Jonas Malmstedt (secretary), and Joakim Nordanstig. Senior advisors: David Bergqvist, Lars Norgren, and Thomas Troëng.
References
- . Systematic review: repair of unruptured abdominal aortic aneurysm. Ann Intern Med. 2007;146:735–741
- . Endovascular aneurysm repair versus open repair in patients with abdominal aortic aneurysm (EVAR trial 1): randomised controlled trial. Lancet. 2005;365:2179–2186
- Two-year outcomes after conventional or endovascular repair of abdominal aortic aneurysms. N Engl J Med. 2005;352:2398–2405
- . Endovascular aneurysm repair and outcome in patients unfit for open repair of abdominal aortic aneurysm (EVAR trial 2): randomised controlled trial. Lancet. 2005;365:2187–2192
- . Performance of endovascular aortic aneurysm repair in high-risk patients: results from the Veterans Affairs National Surgical Quality Improvement Program. J Vasc Surg. 2007;45:227–233
- . Endovascular abdominal aortic aneurysm repair: long-term outcome measures in patients at high-risk for open surgery. J Vasc Surg. 2006;44:229–236
- . Grading of patients for surgical procedures. Anesthesiology. 1941;2:281–284
- . Endovascular repair of infrarenal abdominal aortic aneurysms in high-risk-surgical patients. Eur J Vasc Endovasc Surg. 2007;34:145–151
- Midterm results of endovascular infrarenal abdominal aortic aneurysm repair in high-risk patients. Vasc Endovasc Surg. 2007;41:301–309
- . Patient fitness and survival after abdominal aortic aneurysm repair in patients from the UK EVAR trials. Br J Surg. 2007;94:709–716
- . Contrast-induced acute kidney injury. J Am Coll Cardiol. 2008;51:1419–1428
- Epidemiology and prognostic implications of contrast-induced nephropathy. Am J Cardiol. 2006;98:5K–13K
- Contrast-induced nephropathy after percutaneous coronary interventions in relation to chronic kidney disease and hemodynamic variables. Am J Cardiol. 2005;95:13–19
- Comparison of the effects of open and endovascular aortic aneurysm repair on long-term renal function using chronic kidney disease staging based on glomerular filtration rate. J Vasc Surg. 2008;47:1141–1149
- . Chronic kidney disease and the risks of death, cardiovascular events, and hospitalization. N Engl J Med. 2004;351:1296–1305
- . Endovascular aneurysm repair at 5 years: Does aneurysm diameter predict outcome?. J Vasc Surg. 2006;44:920–929
- . Objective scoring systems of medical risk: a clinical tool for selecting patients for open or endovascular abdominal aortic aneurysm repair. J Vasc Surg. 2007;45:1102–1108
- Identifying and grading factors that modify the outcome of endovascular aortic aneurysm repair. J Vasc Surg. 2002;35:1061–1066
- Auditing surgical outcome: ten years with the Swedish Vascular Registry-Swedvasc (The Steering Committee of Swedvasc). Eur J Surg Suppl. 1998;581:3–8
- Twenty years with the Swedvasc Registry. Eur J Vasc Endovasc Surg. 2008;35:129–130
- . Outcome after abdominal aortic aneurysm repair in Sweden 1994-2005. Br J Surg. 2008;95:564–570
Competition of interest: none.
PII: S0741-5214(08)01144-0
doi:10.1016/j.jvs.2008.07.009
© 2008 Published by Elsevier Inc.
