Is emergency endovascular aneurysm repair associated with higher secondary intervention risk at mid-term follow-up?

Patients 

Between May 1998 and August 2005, 56 patients underwent eEVAR for an acute infrarenal AAA, of whom 34 had a rAAA (61%) and 22 an acute nrAAA (39%). From the beginning of the study period on, patients presenting with an acute AAA were evaluated for eEVAR. Prerequisites for evaluation were the availability of an endovascular team and a sufficient stock of devices. From 2003 on, eEVAR was performed on an intention-to-treat basis. Indeed, with increasing experience, an on-call team could be provided at all times, and from then on devices were always available.

Interpretation of hemodynamic stability in rAAA to allow evaluation for eEVAR was left to the discretion of the attending surgeon. Systolic blood pressures as low as 50 to 70 mm Hg were accepted in applying the “hypotensive hemostasis” principle.19 This means that fluid administration was restricted to avoid rise in blood pressure with the risk of subsequent bleeding.

Anatomic suitability for eEVAR was determined by computed tomographic angiography (CTA) according to guidelines for elective EVAR. These included a proximal neck length >15 mm with <60° angulation and access vessels large enough to accommodate the introducer sheaths.20 With time patients with more peculiar anatomy were also accepted, including those with severe angulations of neck and iliac arteries. This decision was reached between the attending vascular surgeon and interventional radiologist.

As a frame of reference for outcome, we defined a case-matched cohort of the 322 patients who underwent elective endovascular AAA repairs during the study period. During the same period, 211 patients underwent acute open repair, of whom 161 (76%) had a rAAA and 50 (24%) an acute nrAAA. Only 6% of patients presenting to our hospital with an acute AAA were not treated, mainly because of old age and comorbidity. Patient and aneurysm characteristics and type of anesthesia used are listed in Table I.21

Table I. Patient and aneurysm related characteristics and type of anesthesia used

		rAAA	Acute nrAAA	Elective AAA
	Age (year)	73±9	73±9	71±8
	Male/female (%)	91/9	91/9	94/6
	Aneurysm diameter (mm)	74±19	64±11	59±10
	Prox. neck length (mm)	22±6	30±16	28±11
	Anesthesia (%)
	Local	80	86	74
	General	20	14	14
	Spinal	0	0	12
	ASA classification 21(%)
	2	0	9	24
	3	32	68	72
	4	68	23	4
	Comorbidity (%)
	Coronary artery disease	44.1	45.5	46.6
	Congestive heart failure	11.8	13.6	7.8
	Arrhythmia	17.6	13.6	13.4
	COPD	26.5	18.2	25.5
	Diabetes mellitus	11.8	9.1	9.3
	Hypertension	29.4	40.9	51.9
	Chronic renal failure⁎	11.8	4.5	6.2

rAAA, Ruptured abdominal aortic aneurysm; nrAAA, nonruptured abdominal aortic aneurysm; ASA, American Society of Anesthesiology; COPD, chronic obstructive pulmonary disease.

⁎ Serum creatinine, >150 umol/L.

Procedure 

The first choice for anesthesia was local.7A bifurcated device was used whenever possible, which represents the most physiologic solution and is our daily practice in elective EVAR. Several types of stent grafts were inserted. In the beginning the choice was determined by market availability (Table II); now we have a full stock of both Zenith (Cook, Bloomington, Ind) and Excluder (W.L. Gore & Associates, Flagstaff, Ariz) stent grafts, which explains their use in acute cases in our hospital.

Table II. Stent grafts inserted

	Type⁎	rAAA n(%)	Acute nrAAA n(%)	Elective AAA n(%)
	Vanguarda	0(0)	4(18)	27(8)
	Talentb	2(6)	1(5)	51(16)
	Excluderc	3(9)	0(0)	61(19)
	Zenithd	9(85)	17(77)	170(53)
	Quantume	0(0)	0(0)	13(4)

rAAA, Ruptured abdominal aortic aneurysm; nrAAA, nonruptured abdominal aortic aneurysm.

⁎ The stent grafts inserted are listed in chronological order. a Boston Scientific Corp., Waterston, Mass. b World Medical/Medtronic Corp., Sunrise, Fla. c W.L. Gore & Associates, Flagstaff, Ariz. d Cook, Bloomington, Ind. e Cordis, Miami Lakes, Fla.

Follow-up 

Duplex ultrasound scanning (DUS) and plain radiographs of the abdomen in four directions were performed at discharge, 6 months, 1 year, and then yearly. CTA was performed ≤4 weeks after the procedure to confirm exclusion of the aneurysm. Thereafter CTA was only done if routine follow-up DUS or radiographs suggested any problem, including a type I or III endoleak, a type II endoleak with increase of the aneurysmal sac diameter, or in the event of migration, severe kinking, or structural damage of the stent graft. Indications for secondary intervention included type I and III endoleak or migration with imminent type I or III endoleak. Aneurysmal sac growth, with or without type II endoleak (endotension), was an indication for secondary intervention, but the indication was balanced with the patient’s age and comorbidity. Limb occlusion with disabling claudication or critical ischemia or severe kinking with imminent limb occlusion was also an indication for secondary intervention.

Definitions 

An acute AAA was defined as any AAA requiring treatment ≤24 hours. A differentiation was made between the rAAA and the acute nrAAA. The rAAA classification was only awarded in the presence of a retroperitoneal hematoma on CTA. All other acute AAAs were classified as acute nrAAAs as determined by acute onset of abdominal or back pain combined with pain at aneurysm palpation.

Aneurysm-related deaths were defined as all deaths due to aneurysm rupture after a primary or secondary intervention or open conversion.22 Early deaths (<30 days after the primary intervention or within the same hospital admission) were all classified as aneurysm related. Late deaths were only classified as unrelated if a nonrelated cause of death could be attributed.

In the context of this study, complications other than death of the patient refer only to those related to the aneurysm or stent graft.

A secondary intervention was defined as any subsequent endovascular or open surgical treatment related to aneurysm repair or complications thereof. An open conversion was defined as a laparotomy with removal of the stent graft and insertion of a surgical prosthesis. The term laparotomy implied nonconversion laparotomy only.

Statistics 

Data were prospectively collected in an Access database (Microsoft Corp, Redmond Wash) and analyzed using SPSS 12.0.1 (SPSS, Chicago, Ill) and GraphPad Prism 4 (GraphPad Software, San Diego, Calif). The outcome measures were all-cause and aneurysm-related mortality, complications, and secondary interventions. Variables were expressed as mean ± standard deviation. Time-to-event variables were studied with Kaplan-Meier survival analysis. If the standard error >10%, data were not presented in the figures. Comparison of time-to-event curves was conducted with Peto log-rank test. Values of P < .05 were considered statistically significant.

Early complications 

In the rAAA group, six (18%) of 34 patients died during the in-hospital period or ≤30 days. One patient died during the procedure. This 85-year-old patient presented with an AAA extending to both iliac bifurcations, but he died before complete exclusion. Five patients died in the postoperative period. One patient died the same day on the intensive care unit, probably due to a hypovolemic shock, although the completion angiogram showed exclusion of the AAA. Another patient died on day 4 due to cardiac failure. One patient, who underwent open conversion because of insufficient access, died of multiple organ failure on day 12. Another patient underwent a secondary intervention (laparotomy) on the day of the initial procedure for evacuation of a large retroperitoneal hematoma and died of respiratory insufficiency on day 14. The last patient who died had a complicated postoperative outcome with three reinterventions (laparotomies) for compartment syndrome, ischemic colitis, and sepsis. A severe pneumonia developed, and he died on day 20.

Three (other) intraoperative complications occurred. In one patient, a renal artery was inadvertently covered, which was accepted. In another, a contralateral limb was malpositioned. This patient also had a large retroperitoneal hematoma; therefore a laparotomy was performed to reposition the limb manually and to evacuate the hematoma. In the third, a local arterial access problem was solved with a patch plasty.

One (5%) of 22 patients in the acute nrAAA group died on day 6 because of respiratory insufficiency. Another patient underwent a secondary intervention because of a progressive groin hematoma after he had been treated with an aortouniiliac system with a femorofemoral crossover bypass.

In the elective AAA (control) group, the in-hospital or 30 day mortality was 3 (1%) of 322. One patient died due to bleeding from an intracerebral metastasis and two others due to myocardial infarction with ventricle fibrillation. Complications requiring secondary intervention are listed in Table III.

Table III. Early and late secondary interventions

	Secondary interventions	N	Patients	%
	In ruptured AAA
	Early secondary interventions	4	2	6
	Laparotomy (4)
	Late secondary interventions	3	3	9
	Fenestrated aortic cuff (1)
	Laparotomy + ligation of side branches (1)
	Coil-embolization (1)
	Patients without secondary interventions		29	85
	Total	7	34	100
	In acute nonruptured AAA
	Early secondary interventions	1	1	4
	Groin exploration (1)
	Late secondary interventions	3	3	14
	Wallstents + extensions (2)
	Extension, custom made (1)
	Patients without secondary interventions		18	82
	Total	4	22	100
	In elective AAA
	Early secondary interventions	4	4	1
	Thrombolysis + Wallstent iliac limb (1)
	Aortic cuff (1)
	Bypass from SMA to renal artery (1)
	Embolectomy + Wallstent iliac limb (1)
	Late secondary interventions	47	34	11
	Embolectomy + Wallstent iliac limb (2)
	Embolectomy + extensions (1)
	Thrombolysis + Wallstent iliac limb (1)
	Iliofemoral cross-over bypass (5)
	PTA iliac limb (1)
	PTA + Wallstent iliac limb (2)
	Wallstent (1)
	Coil-embolization (7)
	Aortic cuff (4)
	Fenestrated aortic cuff (2)
	Wallstents + extensions (3)
	Extension, custom made (6)
	Laparotomy + ligation of side branches (3)
	Laparotomy + suture aneurysm sac (1)
	Bridging stent graft (1)
	Conversion to open repair (7)
	Patients without secondary interventions		284	88
	Total	51	322	100

Late complications 

Mean follow-up in the rAAA group was 20 ± 21 months. During that period, five patients died after a mean of 11 months (range, 4 to 24 months). Two late deaths were due to cardiac events. One patient died as a result of acute leukemia and another related to an excess amount of different medications combined with severe anemia, for which treatment failed. The last late death occurred in an 86-year-old man with hypovolemic shock, most probably due to rupture of a concomitant 67-mm-diameter thoracic aortic aneurysm. Cumulative survival was 67.8% ± 8.6% at 1 year and 62.1% ± 9.5% at 2 and 3 years (Fig 1, A).

View Large Image Download to PowerPoint

Fig 1. A, Kaplan-Meier curves of patient survival expressed as all cause mortality. B, Kaplan-Meier curves of patient survival expressed as aneurysm-related mortality. rAAA, Ruptured abdominal aortic aneurysm; nrAAA, nonruptured abdominal aortic aneurysm.

Three late complications required secondary intervention. (Table III). One patient developed a proximal type I endoleak after 4 years due to caudal migration of the graft (Talent) that was probably caused by extension of disease. This resulted in an overly short neck for standard EVAR, and he was therefore treated with a fenestrated cuff (Cook). Finally, two patients with secondary type II endoleaks resulting in aneurysm growth were treated. In the first, this was done by coil-embolization after 19 months, and in the second, by laparotomy with ligation of the inferior mesenteric artery and lumbar arteries after 3 months. Cumulative secondary intervention risk (early included) was 9.2% ± 5.1% at 1 year and 16.2% ± 8.2% at 2 and 3 years (Fig 2).

View Large Image Download to PowerPoint

Fig 2. Kaplan-Meier curves of secondary intervention risk. rAAA, Ruptured abdominal aortic aneurysm; nrAAA, nonruptured abdominal aortic aneurysm.

In addition to the complications requiring treatment, three secondary type II endoleaks (1 with growth and 2 without growth) occurred. In all cases a watchful waiting policy was adopted with a 6-month follow-up interval. Finally, in one Excluder case, the aneurysm did grow without appearance of an endoleak on DUS and CTA. This was therefore classified as endotension, and that patient also is on a closer follow-up regimen at 6-month intervals.

Mean follow-up was 32 ± 25 months in the acute nrAAA group. During that period three patients died after a mean of 17 months (range, 6 to 31 months). Two late deaths were due to cardiac events, and the last death was due to cholangiocarcinoma. Cumulative survival was 90.9% ± 6.1% at 1 year, 84.8% ± 8.2% at 2 years, and 76.4% ± 10.9% at 3 years. (Fig. 1A).

Three late complications required secondary intervention (Table III). In these cases, a limb extension was applied to correct kinking and upward migration of the iliac limbs (2 Vanguards) or an overly short initial positioning in the common iliac artery (Talent). In both Vanguard cases, treatment consisted of insertion of Wallstents combined with Passager iliac extensions (Boston Scientific, Watertown, Mass), after 35 and 38 months, respectively. In the Talent case, a custom-made tapered iliac extension (Cook) was inserted after 6 months. Cumulative secondary intervention risk (early included) was 9.6% ± 6.5% at 1 and 2 years and 20.9% ± 12.0% at 3 years (Fig 2). There were two secondary type II endoleaks (1 each with and without growth), for which a watchful waiting policy was adopted, as described earlier.

In the elective AAA (control) group, mean follow-up was 40 ± 25 months. During that period, 53 patients died after a mean of 32 months (range, 2 to 84 months). Two deaths were classified as aneurysm-related (Fig 1, B) because the cause of death was unknown, whilst there was an endotension in one patient and a type II endoleak in the other. Cumulative survival was 95.2% ± 1.2% at 1 year, 89.9% ± 1.8% at 2 years, and 86.2% ± 2.1% at 3 years (Fig 1, A). Complications requiring secondary intervention are listed in Table III. Cumulative secondary intervention risk (early included) was 4.2% ± 1.1% at 1 year, 7.6% ± 1.6% at 2 years, and 12.9% ± 2.2% at 3 years (Fig 2).

The 30-day mortality for patients who underwent open repair for an acute infrarenal AAA was 32% in the rAAA group and 8% in the acute nrAAA group.