Endovascular repair of proximal para-anastomotic aneurysms after previous open abdominal aortic aneurysm reconstruction

Introduction

Conventional prosthetic reconstruction for repairs of abdominal aortic aneurysm (AAA) continues to be widely performed. Proximal para-anastomotic aneurysm (PPAA) is a relatively uncommon complication after conventional prosthetic reconstruction and the reported incidence varies widely, depending on the length of follow-up.^1–4 PPAAs after previous open reconstruction may present as a continuing dilation of the aortic arteries adjacent to the anastomosis (true PPAAs) or as a disruption of the anastomosis leading to pseudo-PPAA formation (false PPAAs). PPAAs are associated with high rupture rate, which range from 15 to 55% without intervention. ⁵

Open reinterventions are technically challenging, with higher mortality rates and morbidity rates compared with primary prosthetic reconstruction.^1,6,7 Some clinical reports have suggested that endovascular repair is an alternative management of PPAAs with significantly lower peri-operative mortality rates.^8–11 However, reports of this procedure with a large series and longer-term follow-up are still quite limited. In this single center study, we presented our results with 24 cases of PPAAs and described our experience.

Methods

Procedures

General anesthesia and groin incisions with open femoral arteriotomy were made for all the patients. The device types were chosen according to anatomic suitability, the configuration of previous prosthetic graft, PPAAs neck condition and the presence of true iliac aneurysm (Tables 1 and 2). If unibody bifurcated stent graft was used, sitting on the aortic bifurcation with “anatomical fixation”, an additional proximal cuff would be implanted for PPAAs. Whether to deploy the Palmaz stent depended on the DSA outcome, and we normally implanted the Palmaz stents if the aneurysm necks were still angulated after the main body and the proximal cuff had been deployed.

OPEN IN VIEWER

Table 1.

Characteristics of the patients and PPAAs.

	Tube (n = 19)	Bifurcated (n = 5)	All (n = 24)
Age	70.5	71.2	70.6
Male	16	4	20
ASA class
2	6	2	8
3	12	3	15
4	1	0	1
Time, years	10.1	11.3	10.3
Clinical details
Emergent	4	2	6
PAA. No	28	6	34
Proximal	19	5	24
Hostile neck	9	0	9
Pseudo	12	4	16
True	7	1	8
Distal	9	0	9
Pseudo	4	0	4
True	5	0	5
Iliac	0	1	1
Pseudo	0	1	1
True	0	0	0
Iliac true aneurysm	2	0	2

ASA: American Society of Anesthesiologists; PPAA: proximal para-anastomosis aneurysm; No: number; Time: time interval between the previous open aortic prosthetic reconstruction and endovascular repair; PAA: para-anastomotic aneurysm.

OPEN IN VIEWER

Table 2.

PPAAs’ anatomical features and endovascular repair methods.

GC	PPAA anatomical features	Endovascular repair methods (number of patients)
		A	B
Tube	Proximal (9)	Unibody bifurcated + Cuff (5)	Tube (4)
	Proximal + Distal (8)	Unibody bifurcated + Cuff (8)
	Proximal + true CIA (1)	Unibody bifurcated + Cuff + Extension (1)
	Proximal + Distal + true CIA (1)	Unibody bifurcated + Cuff + Extension (1)
Bifurcated	Proximal (4)	Tube (4)
	Proximal + Unilateral Iliac (1)	Unibody bifurcated + Cuff + Extension (1)

PPAA: proximal para-anastomotic aneurysm; GC: graft configuration of the initial open reconstruction; CIA: common iliac aneurysm.

For patients with iliac para-anastomotic aneurysms (PPAs) or true iliac aneurysms, the internal iliac arteries would be covered by limb extension, and the internal iliac arteries embolization (IIAE) had been done 3–4 weeks prior.

Results

From May 2003 to January 2013, totally 28 patients were identified with proximal pseudo-PAAs or true PAAs in our center. Among them 24 patients (20 men; mean age, 71 ± 7) fit the inclusion criteria for endovascular therapy and were treated with endovascular repair. Baseline characteristics of these patients and PPAAs are shown in Table 1. There were nine patients with hostile PPAAs neck. All were treated by using unibody bifurcated stent graft with cuff. Through preoperative risk assessment, eight patients were classified as American Society of Anesthesiologists (ASA) 2, 15 were ASA 3 and 1 were ASA 4. Nineteen patients were previously treated with a tube graft and five with bifurcated prosthesis grafts. The various PPAAs anatomical features and corresponding patient number are shown in Table 2.

The PPAAs in 18 patients were identified by routine surveillance protocol that included ultrasound imaging every one year after open aortic surgery. Six patients presented with a symptomatic PPAA, three of whom with a ruptured PPAA, two with abdominal pain and one with back pain. None of the patients in this series had symptoms or signs at CTA suggesting graft infection. The median interval between the initial open reconstruction and endovascular repair was 10.3 (5–17) years. Types and numbers of devices used are shown in Table 3.

OPEN IN VIEWER

Table 3.

Configurations and types of stent grafts that were used.

Configurations of stent grafts	Types	Number	Total
Bifurcated endograft	Powerlink	16	16
Tube endograft	Talent	6	8
	Vanguard/Mintec	2
Cuff	Powerlink	16	16
Palmaz stent	Palmaz	11	11
Limb extension	Powerlink	3	3

Three patients had unilateral IIAE 3–4 weeks prior to endovascular repair. No patients died during the procedure nor developed colon ischemia after endovascular repair. Stent grafts were deployed successfully in all the patients with a technique success rate of 100% (24/24). The 30-day clinical success rate was 95.8% (23/24). One patient (with previous tube prosthesis graft) was admitted for ruptured PPAA (with hostile neck) and emergently treated using a unibody bifurcated stent graft with a cuff. He had a minor type Ia endoleak, which was detected by the final angiography and under close observation. There were no peri-operative open conversions or deaths. The mean hospital stays were 6.7 days (range, 4–9 days).

Median follow-up was 43 months (range, 7–67 months) with no patients lost to follow-up. Follow-up CTA revealed four patients with type Ia endoleaks. Three patients were detected with a secondary type Ia endoleak after endovascular repair (with tube stent graft) and were treated by adding a proximal cuff. One patient with the primary minor type Ia endoleak became worse with abdominal pain six months after endovascular repair and the worsened endoleak could not be treated by standard endovascular approach according to our judgement. So he was converted to open repair immediately with prosthetic graft. The overall reintervention and open conversion rate during follow-up were 16.7% (4/24) and 4.2% (1/24), respectively. Estimates of freedom from reintervention (Figure 1) after endovascular PPAAs repair were 91.7% at 1 year, 87.1% at 3 years, and 80.9% at 5 years. There was significant difference in freedom from reintervention between PPAA patients treated with tube and unibody bifurcated stent grafts (Figure 2; p = 0.034). OPEN IN VIEWER

OPEN IN VIEWER

Figure 2.

Kaplan–Meier curve showing freedom from reintervention after endovascular repair of PPAAs with tube or unibody bifurcated stent graft.

The cumulative mortality rate was 12.5% (3/24). The patient treated by open conversion died one month post the open surgery because of renal and liver failure. Another two patients died 16 months and 33 months later, respectively, after the endovascular repair for myocardial infarction. Actuarial analysis for all patients estimated survival rates of 95.8% at 1 year and 87.3% at 5 years (Figure 3). OPEN IN VIEWER

Discussion

PPAAs are associated with high rupture rates in patients who did not undergo revision surgery.^6,12,13 In this group, three patients (12.5%) were admitted because of PPAA rupture and treated with emergent endovascular repair. So PPAA is one serious complication after open AAA surgery and close follow-up is mandatory. Early treatment is recommended once PPAA is confirmed by CTA or DSA.

Reported mortality rates of open PPAAs reconstruction vary considerably, from 8% to 70%, with morbidity rates of 70% to 83%. ⁸ Gawenda et al. reported that lower morbidity and mortality rates post endovascular repair than post open repair. ¹⁴ Endovascular repair for PAAs was safer without dissection through the scars from previous operations nor the suprarenal aortic clamping. Bosch et al.’s study, with a mean follow-up duration of 41 months, has shown endovascular repair is a feasible and durable alternative to open reconstruction for PAAs. ⁸ In our study, most of the patients whose ASA class were ≥2, so open surgical reconstruction was obviously unsuitable for these cases and endovascular repair was an alternative.

However, endovascular repair of PPAAs has some drawbacks as well. Firstly, proximal fixation zone for PPAAs may not be adequate. Sachdev et al. stated that short and angulated aortic necks are typically characteristic of PPAAs of the infrarenal aorta. ¹¹ In this cohort, there were nine (37.5%) patients with hostile necks. For PPAAs with hostile necks, common modular bifurcated stent graft will also not be recommended because of inadequate proximal landing zone. Secondly, there are reports indicating the insecure anchoring of the distal part of stent graft inside the previous polyester graft. As a result, tube stent graft is deemed unsuitable for PPAAs because of higher distal migration and type Ia endoleak rates. ⁸ In our results, type Ia endoleaks rates were significantly higher in PPAA patients treated using tube stent graft. Thirdly, space inside the previous polyester graft is very limited, so normally full expansion of the bifurcated main body is impossible, and cannulation of the contralateral limb is difficult too. Thus, using modular bifurcated stent graft in such patients seems inappropriate.

Suitable types of stent grafts should be chosen according to PPAAs anatomic features. We prefer to use unibody bifurcated stent grafts with transrenal fixation for PPAAs. Totally different from common modular bifurcated devices, unibody bifurcated stent graft can be deployed sitting on the native abdominal aortic bifurcation. This technique is named as “anatomical fixation” by us, and it has been shown to be effective in preventing distal migration in hostile neck AAAs.^15–19 Similarly, this technique can be utilized in endovascular repair for patients who have PPAAs with hostile necks. Moreover, unibody stent graft is directly deployed sitting on the aortic bifurcation and does not need to cannulate the contralateral limb so the procedure is quite simple and quick, very suitable for emergent case. Although fenestrated or branched grafts are particularly useful for PPAA cases with short necks, this technique requires great experience and the time needed to manufacture custom-made devices may make it unfeasible for symptomatic patients. Ziegler et al. reported their data of nine patients who received custom-designed fenestrated endoprostheses for para-aastomotic aortic aneurysm repair. During mean 12 months’ follow-up, there was one reintervention due to a break in a side branch stent graft. ²⁰ Compared with Ziegler’results, our mid-term results were considered acceptable. In this group, 16 PPAA patients (including nine PPAAs patients with hostile necks) were treated with Powerlink® unibody stent grafts, and the overall reintervention rate during mean 43 months’ follow-up was 6.3% (1/16). Moreover, the Palmaz stent can be used to strengthen the radial fixation force and oppose the graft to the aortic wall for better seal.

Conclusions

PPAAs are severe complications after AAA open reconstructions. Closer follow-up and prompt treatment are necessary. Endovascular treatment for PPAAs is effective, safe and durable. Unibody bifurcated stent graft proved to be suitable for most PPAAs with various anatomical features.