Early and mid-term results of ruptured abdominal aortic aneurysms in the endovascular era in a community hospital
Article Outline
Objective
Endovascular repair (EVAR) has been increasingly used for ruptured abdominal aortic aneurysms (rAAAs), especially in major academic centers. The goal of this article is to report our results with an EVAR-first approach for rAAA which we adopted in 2001 in our community hospital.
Methods
All consecutive patients who underwent attempted repair for rAAA between February 2001 and July 2006 were analyzed. Only patients with computed tomographic or visual verification of extraluminal blood were included.
Results
A total of 40 patients (30 men; mean age, 76.4 ± 7.2 years; range, 57-89 years) presented with rAAA. Thirty patients underwent attempted EVAR for rAAA, constituting 4.1% of all EVAR cases (n = 738), and 10 patients had attempted open repair. Twenty-one (53%) were transferred from another institution. Computed tomography was performed in 97.5%. On arrival to the emergency department, 43%% were hypotensive (systolic blood pressure <80 mm Hg). Transfemoral balloon occlusion was used in 12 cases (30%; 10 in the EVAR group and 2 in the open group). The length of operation was 128 ± 35 minutes (range, 77-210 minutes) in EVAR cases. EVAR was completed in 93.3% (iliac anatomy and proximal endoleak caused open conversion in two cases). Out of the 10 open treated cases, 1 was converted to EVAR and survived. The grafts used for EVAR were AneuRx (n = 21), Zenith (n = 5), and Ancure (n = 4), and 97% were bifurcated. Five patients (16.6%) in the EVAR group died within 30 days (four required balloon occlusion). The mean length of stay was 9.1 ± 6.2 days (range, 4-30 days) in survivors of EVAR. In the EVAR-treated group, two patients died (7 and 9 months; unrelated), and six of the surviving patients (23%) required secondary procedures (five femorofemoral bypasses for limb occlusions and one proximal cuff for a type I endoleak that caused repeat rupture) during a mean follow-up of 13.8 ± 10.4 months (range, 3-39 months). The mortality rate was 40% (4/10) in patients who underwent open procedures during this period, with an overall mortality rate of 22.5% for all ruptures treated. The difference in 30-day mortality in the EVAR and open groups did not reach statistical significance (17% vs 40%; P = .19). In the entire cohort, hypotension (systolic blood pressure <80 mm Hg) on arrival and loss of consciousness were associated with 30-day mortality. Balloon occlusion was correlated with mortality in the EVAR-treated group (44% vs 4%; P = .019). The multivariate analysis using logistic regression showed that hypotension (odds ratio [OR], 7.4; 95% confidence interval [CI], 1.3-42.0; P = .025), loss of consciousness (OR, 37.5; 95% CI, 3.4-40.8; P = .003), and the need for balloon occlusion (OR, 5.2; 95% CI, 1.8-25.5; P = .042) were correlated with higher perioperative mortality, whereas age greater than 76 years, coronary artery disease, chronic obstructive pulmonary disease, hypertension, diabetes, renal insufficiency, and type of procedure did not.
Conclusions
Our results show that EVAR is feasible with favorable outcomes in patients presenting with rAAA in a busy community hospital. There is a high secondary intervention rate, which can potentially be decreased by ensuring good iliac limb anatomy at the end of the procedure and by a closer follow-up.
Despite advances made in critical care, prehospital care, anesthesia care, and postoperative techniques in recent decades, the mortality after open repair of ruptured abdominal aortic aneurysms (rAAAs) remains 35% to 80%.1, 2 This has resulted in the use of the less invasive endovascular aneurysm repair (EVAR)3, 4 in an effort to improve survival in these patients. Although some authors have reported significantly decreased mortality (9%-45%) when compared with historical controls,5, 6, 7, 8, 9 others have not shown any significant differences between these two modalities.10, 11 However, there is a wide variation between studies in the distribution of unstable patients, inclusion of symptomatic patients without rupture, exclusion criteria, and percentage of patients treated by EVAR, thus making it extremely hard to determine the true effect of EVAR on mortality after rAAA.
In addition, most of the studies were reported from either large academic centers4, 5, 12 or large referral centers where regionalized medicine is practiced,9, 10, 11 and the effect of EVAR on rAAA mortality in the community at large is still largely unknown. We started using EVAR in elective repair of AAAs in 1999 in our community hospital after approval of the devices by the US Food and Drug Administration and developed a very high volume EVAR practice. We adopted an EVAR-first approach for all comers with rAAAs since our first EVAR of an rAAA in 2001. The goal of this article is to report our management protocol, techniques, and results with the EVAR-first approach in patients presenting with rAAA to our community hospital.
Materials and methods
All consecutive patients who presented with rAAA to the Sisters of Charity Hospital between February 2001 and October 2006 were included. Rupture was defined as visualization of extraluminal blood by computed tomography (CT), operative findings, or both. Patients with symptomatic AAA without CT evidence of extraluminal blood were not included. Demographic characteristics, comorbidities, clinical presentation, hemodynamic stability on arrival, preoperative serum creatinine, hemoglobin level, electrocardiographic findings of myocardial ischemia, transfer from another institution, maximum aneurysm diameter (millimeters), operative time, type of stent graft used, intraoperative events (including adjunctive use of intra-aortic balloon placement), and 30-day postoperative outcomes and follow-up (CT and clinical) were all obtained from patients’ hospital and office charts. Institutional review board approval was obtained for this retrospective study involving chart review.
The diagnosis of rAAA in a stable patient was usually made by the CT scan ordered by the emergency department (ED) physician and read by an on-call radiologist. The operating room team was activated by the ED physician, who would notify the vascular surgeon on call. We have 24-hour coverage by a radiologist in our hospital. The same would occur when a patient with a diagnosis of rAAA was to be transferred to our ED. When an unstable patient arrived in the ED with a suspected rAAA, the patient was immediately transferred to the CT scanner as the operating room was prepared for both EVAR and open repair. When the vascular surgeons were in the hospital, they accompanied the patient to the CT scanner and/or the operating room. When they were not in the hospital, they would be notified of the patient’s transfer to CT scanner or the operating room, and the surgeon would see the patient either in the CT scanner, en route to the operating room, or in the operating room. The vascular surgeons would decide the type of approach to use (open vs EVAR), as well as the graft type to be used, with no further contact with the radiologist. Two vascular surgeons were involved whenever feasible; however, both vascular surgeons are equally capable and comfortable with both types of repair and operated alone in 12 of the 40 cases in this cohort. One very unstable patient with a known AAA who developed hypotension while in the hospital was taken straight to the operating room from the floor. This was the only patient who did not have a preoperative CT scan. His rupture was verified with the postprocedure CT scan, as was routinely obtained in all patients after EVAR. The time of arrival to our ED and the time of start of the operation were recorded, and the intervals were calculated.
Hemodynamic instability was defined as systolic blood pressure less than 80 mm Hg and/or loss of consciousness. Blood pressure less than 80 mm Hg on arrival to the ED was noted. If the patient became unstable at any time before the operation, this was also noted. Permissive hypotension was practiced and fluid resuscitation kept to a minimum as long as the patient did not lose consciousness. CT with intravenous contrast using 3-mm slices was performed. Increased creatinine did not change the protocol used, and all patients received intravenous contrast. Rupture was defined when extravasated blood was seen outside of the aneurysm sac either in the CT scan or in the operating room. The eligibility of EVAR vs open repair and the type of graft to be used was determined by the surgeons, both of whom have been reading CTs and performing EVAR since 1999, using very similar criteria. The anatomic inclusion criteria for EVAR included a neck length of at least 5 mm, a neck diameter less than 28 mm (<26 mm before the 2003 approval of the Zenith [Cook Inc, Bloomington, Ind] graft), and accessibility of the iliac arteries (heavy calcification, tortuosity, and size <7 mm). Patients with circumferential heavy calcification in the neck were excluded, but tortuous necks were evaluated by intraoperative angiogram for the feasibility of EVAR. Because of our high volume of EVAR practice, we have always maintained the most commonly used sizes of the commercially available endografts (Ancure [Guidant, St Paul, Minn], AneuRx [Medtronic, Minneapolis, Minn], Excluder [Gore Inc, Flagstaff, Ariz], and Zenith [Cook Inc, Bloomington, Ind]) in our stock for treatment of rAAA with EVAR. AneuRx grafts were used preferentially unless the neck diameter was greater than 26 mm, at which time the Zenith graft was used. Bifurcated grafts were preferentially used. Conversion to a uni-iliac graft was not necessary in any patient. Before July 2003, one patient was treated with an aortouni-iliac graft (Ancure) with contralateral common iliac artery ligation and a femorofemoral bypass using a ringed 8-mm polytetrafluoroethylene graft. If the patient’s anatomy was not clear by the CT scan (patients who had CT scans in an outside hospital with 5-mm cuts), if this could not be performed, or if the patient was unstable, the patient was taken straight to the operating room, and a single femoral cutdown was performed with the patient under local anesthesia. An aortogram was performed after access to the aorta to evaluate the proximal neck for adequacy of EVAR. In seven patients who did not seem to be candidates for EVAR, EVAR was attempted. One of these patients had persistent endoleak that led to immediate conversion to open repair, and the second patient had a late migration and rupture at 11 months and underwent successful repair by using a cuff extension. If patient was not deemed a candidate for EVAR, the femoral access was maintained for possible use of balloon occlusion.
Dedicated vascular operating room personnel consisting of a circulating nurse, scrub technician, and radiology technician, in addition to the vascular surgeon (one or two) experienced in both open repair and EVAR, were always available. The room was prepared for open repair if needed. Unlike the elective EVAR patients in our practice, who undergo regional anesthesia, all cases with rAAA were performed under general anesthesia; however, local anesthesia was used in unstable patients for initial access, later to be converted to general anesthesia when proximal control was obtained. Similar to Mehta et al,12 we prefer general anesthesia in these cases to expedite the process and to leave open the possibility for open conversion. All procedures were performed by vascular surgeons in the operating room using the OEC 9800 system (GE Medical Systems, Salt Lake City, Utah).
After the initial cutdown, usually performed simultaneously by two surgeons, a guidewire was advanced to the descending thoracic aorta and was then changed to a superstiff guidewire, over which a 16F sheath was placed into the aorta. The balloon occluder (Reliant [Medtronic] or Coda [Cook]) was made available and was placed through this sheath at any time during the procedure if the patient became more unstable with a worsening of hypotension or lost consciousness. The contralateral femoral artery was used to access the suprarenal aorta, and a marker flush pigtail catheter was used to perform an angiogram. Heparin was not used during these procedures, as is our routine in rAAA cases, to decrease coagulopathy in a bleeding patient. The stent graft deployment was completed as dictated by each device in a routine fashion. We favored femoral access in all patients who required balloon occlusion. We did not find it necessary to use brachial access for balloon insertion in any case in this series and prefer to support the balloon via the sheath inserted via the femoral artery. After the inflation of the balloon above the celiac artery, we delivered the main body delivery device through the contralateral sheath, and after we marked the renal arteries by using retrograde injections through the sheath where the balloon was inserted, the balloon was deflated, and the graft was deployed. The balloon had to be reinserted and inflated in the main body of the graft, below the renal arteries, in 8 of the 10 cases in whom balloon occlusion was used. The contralateral limb was cannulated, and iliac limb extensions were deployed to complete the procedure. After completion of the graft deployment, the completion angiograms were performed for type I or type III endoleaks, and no patient left the operating room with these types of endoleaks. A CT scan with contrast was repeated before discharge on all patients to exclude any endoleaks. We would plan to treat type II endoleaks only if the patient remained unstable. One stable patient with a type II endoleak was observed in our series.
Open repair was performed with the patient under general anesthesia via the transabdominal approach in a standard fashion with a Dacron graft (DuPont, Wilmington, Del) in all cases, and supraceliac or infrarenal clamping was used as deemed necessary. Cell-saving devices were used only in patients who underwent open repair.
The measured outcomes included technical success with complete sealing of the aneurysm, conversion, intensive care unit (ICU) length of stay (LOS), total hospital LOS, 30-day in-hospital mortality, complications, early and late reinterventions, aneurysm-related events, and condition on last follow-up. The patients who underwent EVAR were followed up with our follow-up protocol for elective EVAR, which consisted of clinic visits and CT scans at 3, 6, and 12 months and then yearly thereafter. The patients who survived open repair were followed up yearly after being seen within 1 month after discharge. These patients did not have the vigorous post-EVAR surveillance with repeated CT scans.
Data analysis was performed with SPSS 14.0 software (SPSS Inc, Chicago, Ill). The descriptive statistics are given as mean ± SD, with median and range numbers indicated in parentheses when data were skewed. Cross-tabulation of predictive covariates was performed between two groups (survivors vs nonsurvivors) by using the Fisher exact test for nonparametric variables. Multivariate analysis was performed with stepwise (forward: likelihood ratio) binary logistic regression analysis for predicting perioperative mortality (death within 30 days). Kaplan-Meier life tables were used for statistical analysis of overall, reintervention-free, and endoleak-free survival rates. All P values were considered significant if <.05.
Results
A total of 40 patients (30 men; mean age, 76.4 ± 7.2 years; range, 57-89 years) presented with rAAAs. Thirty patients underwent attempted EVAR for rAAA, constituting 4.2% (n = 738) of all EVAR cases and 75% (n = 40) of all patients who presented with rAAA during this period. The patients’ demographic characteristics are shown in Table I. The periprocedural characteristics are shown in Table II. Twenty-one (53%) were transferred from another institution. CT was performed in 39 (97.5%). The mean time to the start of the operation from arrival to the ED was 57 ± 18 minutes (range, 19-98 minutes), and there was a significant difference between stable and unstable patients (65 ± 17 minutes vs 49 ± 15 minutes; P = .02). The mean size of the AAA was 6.7 ± 1.3 cm (range, 4.2-12.0 cm). On arrival to the ED, 17 (43%) patients were hypotensive. Loss of consciousness was seen in 6 (15%), and 21 (53%) became unstable at some point before the operation. Transfemoral balloon occlusion was used in 10 patients in the EVAR group and 2 patients in the open group. Bifurcated devices were used in all patients but one. One patient who presented with a stable rupture and a marginal neck length (8 mm) was found to have a ruptured inflammatory aortic aneurysm on exploration, and as a result of excessive difficulty with exposure, he was successfully converted to EVAR and survived. He remains problem free 24 months after the procedure. AneuRx (n = 22; 71%), Zenith (n = 5; 16%), and Ancure (n = 4; 13%) grafts were used (97%; bifurcated).
Table I. Demographic characteristics of all patients presenting with ruptured abdominal aortic aneurysms (age, 76.4 ± 7.2 years; range, 57-89 years)
| Variable | % |
|---|---|
| Male | 75% |
| Coronary artery disease | 63% |
| Hypertension | 80% |
| Diabetes mellitus | 20% |
| Cerebrovascular disease | 8% |
| Hyperlipidemia | 55% |
| Chronic pulmonary occlusive disease | 43% |
| Renal insufficiency (creatinine >1.5 mg/dL) | 25% |
| Cirrhosis | 3% |
| Smokers (active) | 35% |
Table II. Periprocedural characteristics of patients who had endovascular aneurysm repair (EVAR) for ruptured abdominal aortic aneurysms
| Variable | EVAR | Open | P value | Overall |
|---|---|---|---|---|
| CT performed | 97% | 100% | .99 | 97.5% |
| Transfer | 57% | 40% | .473 | 53% |
| Unstable | 50% | 60% | .721 | 53% |
| Initial systolic blood pressure <80 mm Hg | 43% | 40% | .99 | 43% |
| Loss of consciousness | 10% | 30% | .153 | 15% |
| Ischemic changes on ECG | 13% | 40% | .089 | 20% |
| Need for aortic occlusion balloon | 30% | 20% | .696 | 28% |
| Bifurcated device/graft | 97% | 100% | NR | NR |
| Operation time (min) | <.001 | NR | ||
 Mean ± SD | 126 ± 32 | 192 ± 27 | ||
| Range | 77-210 | 150-235 | ||
| Mean PRBCs transfused (units) | 2.4 ± 2.7 | 5.0 ± 1.9 | .008 | NR |
The mean operation time was 126 ± 32 minutes (range, 77-210 minutes) in the EVAR group and 192 ± 27 minutes (range, 150-235 minutes) in open cases, all of whom had bifurcated grafts placed. EVAR was completed in 29 cases: 28 (93%) of EVAR-attempted cases and 1 conversion from the open group. Inability to deliver the device because of iliac artery tortuosity and calcification (one patient) and a proximal endoleak in another patient with an 8-mm, 90° angled tortuous neck caused conversion to open procedures in two patients (AneuRx was attempted in both). Mean blood transfusion requirements were 2.6 ± 2.7 units (range, 0-13 units; median, 2 units) in the EVAR group and 5.0 ± 1.9 units (range, 3-8 units; median, 4 units) in open cases (P = .008). The reasons for choosing open repair were unfavorable neck anatomy in seven cases and iliac anatomy in the remaining three cases. The clamp could be placed infrarenally in three and suprarenally in six cases, and the remaining patient with marginal neck anatomy was successfully converted to EVAR.
A total of 12 patients had common iliac aneurysms larger than 2 cm in diameter. Two simply had extension of the iliac limbs to external iliac arteries, and three had additional coil embolizations of their internal iliac arteries before placement of external iliac artery extensions. The remaining common iliac aneurysms were sealed by using an aortic extension cuff (24-28 mm). At least one internal iliac artery was preserved in all patients.
Five patients (17%) in the EVAR group and four patients (40%) in the open group died within 30 days (P = .19). The 30-day mortality rate for the entire cohort was 22.5%. In the EVAR group, one patient with oxygen-dependent chronic obstructive pulmonary disease and cirrhosis with portal hypertension died 7 hours after the procedure. Another patient who had a cardiopulmonary arrest before the procedure had a persistent type I endoleak from his severely angulated neck, which was recognized in the operating room; additional placement of an aortic cuff was not successful, and the procedure was immediately converted to an open approach. Balloon occlusion was kept until a suprarenal clamp could be placed and the repair completed. Unfortunately, the patient died 8 hours later in the ICU. The remaining three patients died of multiple system organ failure within 48 hours after successful exclusion of their aneurysms. They were all anuric and acidotic, had increased liver function tests, and had acute respiratory distress syndrome. In the open-treated group, four patients died on postoperative days 1 (two patients), 5, and 8 from multiple organ system failure. Three of these four patients had suprarenal clamps placed, and all were unstable; two lost consciousness before surgery.
The mortality rates in patients with various preoperative and preoperative characteristics are shown in Table III for all patients and for those treated with EVAR. Mortality in patients who required balloon occlusion was 44% vs 4% in those who did not (P = .019) in the EVAR group and was 45% and 14%, respectively, in the entire cohort (P = .083). The mortality rate was 5% in those who presented and remained stable, whereas it was 38% in those who were not stable (P = .021). There was no difference in mortality in those transferred from another institution (14%) as compared with those presenting primarily to our ED (31%; P = .265). The stability rate was the same between transferred patients (43%) and nontransferred patients (53%; P = .752). Mortality was higher in those who experienced loss of consciousness (83% vs 12%; P = .001) and ischemic changes on the electrocardiogram (50% vs 16%; P = .059). The multivariate analysis using logistic regression showed that hypotension on admission (systolic blood pressure <80 mm Hg; odds ratio [OR], 7.4; 95% confidence interval [CI], 1.3-42.0; P = .025), loss of consciousness (OR, 37.5; 95% CI, 3.4-40.8; P = .003), and the need for balloon occlusion (OR, 5.2; 95% CI, 1.8-25.5; P = .042) were correlated with higher perioperative mortality, whereas age older than 76 years, coronary artery disease, chronic obstructive pulmonary disease, hypertension, diabetes, renal insufficiency, and type of procedure did not. The perioperative mortality of unstable patients was 67% (4/6) in the open group and 27% (4/15) in the EVAR-treated group (P = .146).
Table III. The effect of different factors on mortality
| Variable | Perioperative mortality | P value | |
|---|---|---|---|
| Yes | No | ||
| Male | 20% | 30% | .665 |
| Age >76 y | 29% | 16% | .457 |
| Coronary artery disease | 32% | 7% | .117 |
| Hypertension | 28% | 0% | .162 |
| Diabetes mellitus | 38% | 19% | .348 |
| Cerebrovascular disease | 0% | 24% | .99 |
| Hyperlipidemia | 23% | 22% | .99 |
| Chronic pulmonary occlusive disease | 29% | 17% | .456 |
| Renal insufficiency (creatinine >1.5 mg/dL) | 30% | 20% | .665 |
| Cirrhosis | 100% | 21% | .225 |
| Smokers (active) | 14% | 27% | .453 |
| Transfer | 14% | 20% | .265 |
| Hypotension on arrival (<80 mm Hg) | 41% | 9% | .023* |
| Unstable before surgery | 38% | 5% | .021* |
| Loss of consciousness | 83% | 12% | .001 |
| Ischemia on ECG | 50% | 16% | .059* |
| Hardman score >1 | 35% | 10% | .127 |
| Balloon occlusion | 45% | 14% | .083* |
| PRBCs >2 units | 47% | 0% | .001* |
| Type of repair (EVAR) | 17% | 40% | .190 |
The number of blood transfusions inversely correlated with survival. All patients who required two units or less of packed red blood cells in the EVAR group (n = 21) survived, whereas five (56%) died if more than two units were used (P = .001). In addition, patients who required three or four units of packed red blood cells but did not require balloon occlusion survived, whereas those who required balloon occlusion died. All three patients who required more than four units of packed red blood cells died.
The mean length of stay in the ICU was 3.2 ± 1.4 days, and the total hospital LOS was 9.4 ± 6.4 days (range, 4-30 days) in surviving EVAR-treated patients. The major 30-day morbidities in survivors included one (4%) myocardial infarction, one (4%) deep vein thrombosis with pulmonary embolism, and one (4%) limb ischemia secondary to graft limb occlusion necessitating femorofemoral bypass. We did not diagnose any abdominal compartment syndrome or bowel ischemia necessitating re-exploration in this group of patients. In the open group, two surviving patients had postoperative pneumonia and renal insufficiency, and one had limb ischemia necessitating thrombectomy. The mean ICU and total hospital LOS were 4.7 ± 2.0 days and 10 ± 3.7 days, respectively, in the open-treated patients who survived (P = .046 for ICU stay; P = .149 for total LOS when compared with EVAR cases).
The mean follow-up for the EVAR group was 13.8 ± 10.4 months (range, 3-48 months; median, 12 months [16.0 ± 11.2 months in survivors]). An additional four patients presented with limb ischemia and underwent femorofemoral bypass 4, 5, 7, and 7 months after the EVAR, giving a total of five patients who underwent this procedure (16%; 19% in those surviving >30 days). Only one of these had an Ancure graft, and the rest had AneuRx grafts placed. The overall late endoleak rate was 8%. The 6- and 12-month endoleak-free survival rates were 96% ± 4% (89% ± 8%) in survivors. One patient presented with a repeat rupture 11 months after the first one, and the type I endoleak caused by graft migration was successfully repaired with a proximal cuff extension; this patient remains alive 5 months after the second repair. There was one type II endoleak, which was managed conservatively and disappeared at the 6-month follow-up. Sac shrinkage of 5 mm or more was observed in 62%, as measured by the surgeon in the last CT, with only one sac increase noted in the patient on CT when he presented with repeat rupture due to proximal migration. This patient’s sac remained stable at the last follow-up. The mean sac size on the last follow-up CT was 5.6 ± 1.0 cm, with a mean decrease of 1.0 ± 0.9 cm. The overall aneurysm-related secondary reintervention rate was 23% (6/26) in survivors. The 6- and 12-month reintervention-free survival rates were 83% ± 8% and 67% ± 11%, respectively, in survivors.
The 12-month survival of the entire EVAR-treated group was 77% ± 8% (91% ± 6% in those surviving the initial operation). Late mortalities included two patients: One patient with chronic steroid use and multiple medical comorbidities died as a result of sepsis 7 months after a perforated appendix, which fistulized into the aneurysm sac. The second patient died 9 months later from pancreatic cancer.
Discussion
There has been a significant increase in recent years of reports involving EVAR in the treatment of rAAA.5, 6, 7, 8, 9, 10, 11, 12, 13, 14 This is likely the result of dissatisfaction on the surgeons’ part on the lack of improvement on overall mortality with the traditional open surgical repair. This increase in EVAR use in rAAA also paralleled the dramatic increase in expertise and availability of newer grafts in recent years. Although earlier reports included only high-risk patients,3, 4 more and more centers are adopting an EVAR-first approach whenever feasible and are reporting improved outcomes with decreased morbidity, mortality, blood transfusions, and LOS.7, 9, 12, 14 However, there is a wide variability between studies, making comparisons and conclusions nearly impossible regarding EVAR’s value in rAAAs. These variabilities include the percentage of patients with hemodynamic instability, the minimum neck length criteria used for inclusion for EVAR (5-15 mm), the willingness to proceed with EVAR without preoperative CT scan in unstable patients, the willingness to accept lower blood pressures before rushing the patient to the OR, the availability of experienced vascular surgeons, and the availability of grafts accommodating more types of anatomy.3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14
We adopted the EVAR-first approach for rAAA after acquiring a significant experience with elective EVAR after its Food and Drug Administration approval in the United States in 1999. After our first EVAR of rAAA, we adopted this for all comers relatively easily because of the level of familiarity our ED, OR, and radiology staff developed in managing these patients as a result of our large volume of EVARs in elective and urgent nonruptured aneurysms. We emphasized (1) fluid restriction unless the patient lost consciousness and (2) instant and simultaneous notification of the vascular surgeon, OR, and CT scanner by the ED staff in the event of an rAAA diagnosis or transfer. Although we did not develop a protocol and performed test runs before we started our EVAR experience in rAAA patients, we do recommend this approach to centers that contemplate starting their own rAAA/EVAR program, especially if their AAA volume is not high.
Because of its minimally invasive nature, EVAR should intuitively have a bigger effect on patients with hemodynamic instability; however, this is used as an exclusion criterion in some studies11 or used as a criterion for skipping the CT scanner, thus resulting in many fewer patients receiving EVAR in this subgroup. Mehta et al12 reported that only 3 of the 45 surgically treated patients who were treated after they launched their EVAR for rAAA protocol had absolutely prohibitive anatomy for EVAR. Reporting on a mostly (65%) unstable group of patients, Moore et al9 reported a mortality rate of 5% after EVAR vs 28% in surgically treated patients, with corresponding figures of 14% and 56% in hemodynamically unstable patients; this is not too dissimilar to our 27% and 66% rates in unstable patients. In contrast, Coppi et al6 reported 53% mortality after EVAR, compared with 61% in the open group in unstable rAAA patients. It is of note, however, that although 52% of all patients were eligible, only 27% were treated by EVAR in this series. Overall, 53% of our patients were unstable before surgery (50% in the EVAR group and 60% in the open group). Hypotension on arrival to the ED, loss of consciousness, and need for balloon occlusion, all of which were related to instability, were correlated with poorer outcomes, and the type of treatment did not seem to affect perioperative survival, likely because of the small number of patients treated. Although our mortality was significantly higher in unstable than in stable patients (38% vs 5%; P = .021), it is still reasonably low, especially when one considers that the study included all comers and that hemodynamic instability was not an exclusion criterion. Successful exclusion of rAAA with EVAR was accomplished in 72.5% of all patients with rAAA, including the two conversions: one for iliac anatomy and the other for persistent proximal endoleak. Incidentally, this seems to be very high when compared with other studies that include all comers (18%-49%).6, 7, 10, 11, 12 Our number is also almost double the previously quoted 40% of EVAR suitability for all comers with rAAA.16 However, the criteria used in those studies for anatomic suitability, including neck length and quality, were very stringent, and our willingness to complete the procedure even in those with unclear neck anatomy by CT scan may have contributed to this high percentage. We think that sealing the rAAA even when the anatomic features are not ideal and converting a life-threatening situation into a more manageable and stable condition is an acceptable short-term goal, especially in this high-risk patient population.
The CT scan is usually skipped in those with blood pressures less than 80 mm Hg in most protocols, with the fear of losing the patient en route; however, this is rarely the cause of patient death as long as fluid restriction is followed and unnecessary delays are avoided between the ED, CT scanner, and OR, especially when the speed of the modern CT scanners and mandatory times for OR setup are considered. Lloyd et al17 reported the time to death from admission to be more than 2 hours in nonoperated rAAA patients, with the median time being 10.5 hours from admission to the ED, which should give ample time for both CT scan and OR preparation for most patients. Our CT scan acquisition rate of 97.5% is one of the highest in the literature reporting on all comers with rAAA, and the mean time from arrival to the ED and start of the operation (49 minutes) in unstable patients is acceptable.
There is no 24-7 in-house coverage by vascular surgeons in our hospital, and there are no surgery residents or vascular surgery fellows taking the first call from the ED, as is reported from the large academic centers.9, 10, 12 We have not found this to be a problem because of the simultaneous activation of all vascular surgeons, CT scanner, and OR staff by the ED physicians, mostly even before the patient with a high likelihood of rAAA reaches the hospital. Those transferring from other institutions would be accompanied by their CT results, thus making it even easier to make plans ahead of time, but there was never a time that surgery was delayed because of surgeon’s late arrival. It is important to note that the surgeons live within 20 to 25 minutes of driving distance from the hospital, and the entire surgical team would have been mobilized before the surgeons’ arrival.
Although balloon occlusion was initially described via the transbrachial approach with a cutdown, the transfemoral approach has been our preferred approach, as is the case for most surgeons.15 This maneuver was necessary in a third of our patients but was necessary in up to 73% of patients in some series.14 Because of its inherent potential complications, such as mesenteric or renal artery ischemia, with additional risks of embolization or thrombosis, it is very important not to use balloon occlusion unless it is absolutely necessary. We did not encounter any complications related directly to balloon occlusion in this series involving 12 patients, and the 45% mortality in this group was a reflection of the patient’s condition rather than being a result of it.
Uni-iliac grafts have been suggested as a means of expediting exclusion of AAA by some authors6, 10; however, we have not found placement of bifurcated grafts to be an issue during surgery. Still, we would not hesitate to convert to an aortouni-iliac configuration, followed by contralateral occlusion and femorofemoral bypass, if continued excessive bleeding coupled with difficult contralateral cannulation were encountered. We have encountered additional limb occlusions (19% in survivors) in our patients, all of whom were treated with femorofemoral bypass grafts. This constituted the most common reason for aneurysm-related secondary intervention in our series. We think that this was the result of not performing the oblique views of the iliac arteries at the end of the procedure in these emergency cases, as is our routine in elective cases. Avoidance of contrast load and the tendency to end the emergency operation as soon as possible may have been the reasons for this practice, but we have since been paying more attention to the quality of the iliac arteries and graft limbs at the completion of the EVAR in these patients. Our limb occlusion rate in the elective EVAR population is 2.7%, and we think that the higher rate in rAAA patients resulted from failing to recognize the anatomic defects in the final repair.
The secondary intervention rate following emergency EVAR has been reported to be very similar to that following elective EVAR by Oranen et al,18 with a 2-year intervention rate of 16%, whereas Hechelhammer et al13 reported this to be much higher, at 35%. Endoleaks with or without migration were the most common causes for reintervention in these series, whereas we had only one late type I endoleak due to migration with rupture in our series. One patient with type II endoleak was treated conservatively; type II endoleaks have not been reported to cause problems after rAAA treatment with EVAR.19
We have not diagnosed any abdominal compartment syndrome, which has been reported in up to 18% of patients after EVAR for rAAA,12 in our series. However, we have not routinely performed bladder pressure measurements. We rely on other criteria (tense distended abdomen, increased airway pressures, hypotension, and oligoanuria). We have recently adopted routine bladder pressure measurements. Mehta et al12 suggested that they started seeing less abdominal compartment syndrome after they stopped using heparin in their patients. We have not used heparin in rAAA patients from the beginning of our experience, and it is possible that this may have affected our favorable experience with this particular complication.
There are some weaknesses of our study. It is retrospective, and data were not prospectively collected. We have not analyzed the outcome of patients treated before the EVAR-first approach was started, thus making it hard to prove that we improved the mortality in patients with rAAA in our hospital, although comparing results with historical controls has its own drawbacks. Because the duration encompasses nearly 4 years, there has been a significant variation in the availability of the types of stent grafts; therefore, our current experience is likely not represented uniformly by the study group. Finally, the lack of a written protocol may have caused some variations in the way some patients were treated, especially before reaching the OR. However, we do not believe that such small variations were significant enough to have any effect on the final outcome of any patient in this series.
Conclusions
Our results, along with the results of other studies, show that EVAR is feasible, with favorable outcomes in patients presenting with rAAA in a busy community hospital, as long as there are trained personnel who are capable of performing open or EVAR procedures in rAAA. This should be performed in the OR so the patient can undergo emergency conversion if necessary. The perioperative mortality is still largely determined by the patient’s hemodynamic status at presentation. The relatively high reintervention rate can be potentially decreased by ensuring good iliac anatomy at the end of the procedure, and close follow-up is necessary to ensure early diagnosis of late-onset endoleaks to prevent secondary ruptures.
Author contributions
References
- Ruptured abdominal aortic aneurysms: the excessive mortality rate of conventional repair. J Vasc Surg. 2001;34:41–46
- Factors increasing the mortality rate for patients with ruptured abdominal aortic aneurysms. Ann Surg. 2001;15:6001–6007
- . Emergency endovascular repair of leaking aortic aneurysm. [letter] Lancet. 1994;344:1645
- Endovascular repair of ruptured aortoiliac aneurysms. J Am Coll Surg. 1999;189:102–112
- . Impact of endovascular repair on early outcomes of ruptured abdominal aortic aneurysms. J Vasc Surg. 2004;40:211–215
- . A single-center experience in open and endovascular treatment of hemodynamically unstable and stable patients with ruptured abdominal aortic aneurysms. J Vasc Surg. 2006;44:1140–1147
- An intention-to-treat by endovascular repair policy may reduce overall mortality in ruptured abdominal aortic aneurysm. J Vasc Surg. 2006;44:467–471
- . Endovascular repair of ruptured AAA: a literature review. Acta Chir Belg. 2005;105:134–139
- . Improved survival after introduction of an emergency endovascular therapy protocol for ruptured abdominal aortic aneurysms. J Vasc Surg. 2007;45:443–450
- Endograft treatment of ruptured abdominal aortic aneurysms using the Talent aortouniiliac system: an international multicenter study. J Vasc Surg. 2006;43:1111–1122
- Endovascular repair versus open surgery in patients with ruptured abdominal aortic aneurysms: clinical outcomes with 1-year follow-up. J Vasc Surg. 2006;44:1148–1155
- Establishing a protocol for endovascular treatment of ruptured abdominal aortic aneurysms: outcomes of a prospective analysis. J Vasc Surg. 2006;44:1–8
- . Midterm outcome of endovascular repair of ruptured abdominal aortic aneurysms. J Vasc Surg. 2005;41:752–757
- . Endovascular treatment of ruptured abdominal aortic aneurysms: a shift of the paradigm?. J Endovasc Ther. 2005;12:548–555
- . Balloon occlusion of the aorta during endovascular repair of ruptured abdominal aortic aneurysm. J Endovasc Ther. 2005;12:556–559
- . Suitability of ruptured AAA for endovascular repair. J Endovasc Ther. 2004;11:635–640
- Feasibility of preoperative computed tomography in patients with ruptured AAA: time-to-death study in patients without operation. J Vasc Surg. 2004;39:788–791
- Is emergency endovascular aneurysm repair associated with higher secondary intervention risk at mid-term follow-up?. J Vasc Surg. 2006;44:1156–1161
- . Endoleak after endovascular repair of ruptured abdominal aortic aneurysms: is it a problem?. J Endovasc Ther. 2003;10:766–771
Competition of interest: none.
PII: S0741-5214(07)01056-7
doi:10.1016/j.jvs.2007.06.037
© 2007 The Society for Vascular Surgery. Published by Elsevier Inc. All rights reserved.
