Endovascular repair of para-anastomotic aortic aneurysms

Results 

From October 1997 to January 2005, 53 patients (41 men) with 65 para-anastomotic aneurysms involving prosthetic grafts placed in the aortic arch, thoracic aorta, abdominal aorta, or the iliac arteries, were repaired using endovascular stent grafts at the Mount Sinai Medical Center in New York. Patient age averaged 69.7 years (range, 26 to 90 years). The mean number of aneurysms was 1.4 per patient. The mean time from the initial aortic reconstruction to the repair of the para-anastomotic aneurysm was 9.9 years (range, 0.2 to 25.5 years). Para-anastomotic aneurysms were detected at an average of 15.8 years after surgery for occlusive disease, and 8.9 years after repair for aneurysmal disease (P < .01). Major comorbidities and previous surgical history are summarized in Table I.

Table I. Medical comorbidities and previous aortic repair in patients treated with endovascular stent grafts for para-anastomotic aortic aneurysms

	Characteristics	Patients, n
	Comorbidities
	Coronary artery disease	27
	Hypertension	40
	Diabetes mellitus	6
	Hypercholesterolemia	25
	Chronic renal insufficiency	2
	Cerebrovascular disease	9
	COPD	25
	Malignancy	8
	Previous aortic repair
	Aortobiiliac graft	17
	Aortobifemoral graft	8
	Tube graft
	Infrarenal aorta	13
	Thoracoabdominal aorta	1
	Thoracic aorta	9
	Aortic arch	5

COPD, Chronic obstructive pulmonary disease.

In 43 patients, the para-anastomotic aneurysms were discovered as incidental findings during work-up of other medical conditions and were repaired electively. In 10 patients, aneurysms were symptomatic. Of these, nine presented with pain, one presented with an aortobronchial fistula, and all were treated on an emergency basis. All para-anastomotic aneurysms were repaired in the operating room under spinal anesthesia 23 patients, epidural in 23, local in 1, and general in 11.

The original prosthetic graft was placed for elective aneurysm repair in 39 patients, emergency aneurysm repair in 4, aortoiliac occlusive disease in 8, dissection in 1, and angiosarcoma in 1. In those patients originally treated for aneurysm, aneurysmal disease was originally located in the abdominal aorta in 30 patients, the aortic arch in 5 patients, the descending thoracic aorta in 7 patients, and the thoracoabdominal aorta in 1.

The mean size of proximal and, in tube grafts, distal para-anastomotic aortic aneurysms was 60.2 ± 11.1 mm and 65.9 ± 14.8 mm, respectively. Aneurysms involving the iliac anastomosis of bifurcated grafts averaged 47.1 ± 17.6 mm in diameter. Two patients had para-anastomotic aneurysms involving femoral anastomoses, with a mean diameter of 47 ± 22.5 mm. Both patients also had para-anastomotic aneurysms of the proximal graft. The median length of stay was 2.9 days (range, 1 to 82 days).

Technical success with endovascular repair of para-anastomotic aneurysms was determined by the absence of an endoleak on completion angiography in the operating room and was achieved in 52 patients (98%). One patient with a short angulated neck was identified intraoperatively to have a persistent proximal type I endoleak despite the presence of transrenal fixation and placement of a proximal cuff. Endoleaks were identified in eight patients (15%) on follow-up CT imaging. The mean time between operation and identification of an endoleak was 1.7 months (range, 0 to 7.5 months). Four type I endoleaks were identified, three ≤1 month postoperatively and one at 7.5 months. Three type II endoleaks were detected ≤1 month of surgery without sac expansion. One type III endoleak was noted at 5.3 months. The late type I endoleak and the type III endoleak were associated with contained rupture.

Patients who developed endoleaks were significantly more likely to have had emergency endovascular repair of the para-anastomotic aneurysm compared with those in whom endoleaks were not identified (P = .01). Table II correlates the type of endoleak identified with aneurysm morphology and describes their treatment and outcome of intervention.

Table II. Characterization of and treatment of endoleaks after endovascular repair of para-anastomotic aortic aneurysms

	Type	Time to diagnosis (m)	Previous aortic graft	Location of PAA	EVSG device⁎	EVSG configuration	Treatment
	I	0	Abdomen	Proximal	Talent	AUI	Observation
	I	0.1	Thoracic	Proximal	Talent	Thoracic	EVSG
	I	0.3	Arch	Distal	Talent	Thoracic	Open conversion
	I	7.5	Abdomen	Proximal	Talent	Bifurcated	Open conversion
	II	0	Thoracic	Distal	Excluder	Thoracic	Observation
	II	0.1	Abdomen	Proximal	Talent	Tube	Open conversion†
	II	0.4	Abdomen	Proximal	Talent	Bifurcated	Observation
	III	5.3	Thoracic	Proximal	Talent	Thoracic	EVSG

PAA, Para-anastomotic aneurysm; EVSG, endovascular stent graft; AUI, aortouniiliac.

⁎ Talent, Medtronic, Sunrise, Fla; Excluder, W. L. Gore and Associates, Flagstaff, Ariz. † This patient underwent open conversion because of persistent back pain in the setting of a thoracoabdominal aneurysm. The endograft was thus explanted to facilitate the repair.

Clinical or radiologic evidence of graft or perigraft infection was present in five patients. The mean time to presentation in patients with infection was 5.2 months, compared with 10.4 months in patients without infection. This difference was not statistically significant. Two patients with graft infections died after endovascular repair: one of myocardial infarction within the first postoperative week and the other of unknown causes 18 months after surgery. The remaining three patients with graft infections had event-free follow-up of 3, 12, and 18 months, respectively, and were maintained on lifelong oral antibiotics.

Two patients were lost to follow-up. In the remaining patients, the mean time to follow-up was 18.1 months (range, 0.6 to 76.6 months).

Six patients had complications that were not associated with endoleak within the first 30 days. These included groin hematoma in three and one each with nonfatal myocardial infarction, pneumonia, and ileus.

Seven patients experienced late complications during the follow-up period that did not involve endoleaks. The most common late complication other than endoleak was lower extremity ischemia in four patients due to stent graft limb occlusion (n = 2) or stenosis (n = 2), which occurred at 2, 3, 7, and 11 months after surgery. Extra-anatomic bypass was performed in three patients. The fourth patient underwent lower extremity amputation after presentation to another hospital. Stent graft limb occlusions occurred in two patients with common iliac para-anastomotic aneurysms, during which time the endograft was extended into the external iliac artery. All patients with lower extremity ischemia had been treated with either an aortouniiliac device or an endograft limb at the time of endovascular aneurysm repair. Other late complications included femoral artery pseudoaneurysm, aortotracheal fistula, and anastomotic disruption of a visceral segment reimplantation site that did not involve the stent graft site.

Five patients underwent open conversion. Three patients with proximal type I endoleaks underwent conversion: one occurred at the time of attempted endovascular repair and one at 7.4 months after endovascular repair. Both patients died in the perioperative period. The third patient had a type II endoleak and underwent conversion to open repair because of persistent back pain in the setting of a concomitant thoracoabdominal aneurysm involving the visceral segment. Spinal cord ischemia and paralysis developed 2 weeks after surgery. In two patients, conversion was unrelated to endoleak: one had disruption of the visceral reimplantation anastomosis of a thoracoabdominal aneurysm repair, as mentioned previously, and the other developed an aortotracheal fistula requiring open aortoaortic bypass.

Two patients died in the early postoperative period, for a 30-day mortality rate of 3.8%. One patient died of myocardial infarction, and one died suddenly of unknown causes. No postmortem examination was performed in this patient. A total of 26 patients died within the follow-up period, for an overall mortality rate of 49%. Kaplan-Meier estimates of survival predicted 79% survival at 12 months, 51% at 24 months, and 28% at 60 months. The mean time to death was 19.6 months (range, 0.3 to 76.3 months). Death during the follow-up period was significantly associated with advanced age at the time of para-anastomotic aneurysm repair (P = .03). The mean age at surgery was 75.9 ± 8.1 years in patients who died during the follow-up period compared with 65.4 ± 13.6 years in those who survived. Mortality was not significantly associated with gender, mean number of comorbidities per patient, aneurysm location or presentation, initial aortic pathology, presence of endoleak, need for reintervention, or conversion rate.

Discussion 

The number of series describing endovascular management of para-anastomotic aortic and iliac aneurysms has increased in the last 5 years, and each has determined that endograft treatment of these types of aneurysms is both feasible and safe. However, the available series, including our own, are small, encompassing no more than 30 patients in one report.4, 5, 6, 7, 8, 9, 10 This study characterizes our on-going experience with treating patients who underwent endovascular repair of these difficult aneurysms.

The true incidence of para-anastomotic aneurysms after open aortic reconstruction for either aneurysmal or occlusive disease is unknown. The literature reports a frequency of <1% to 12%,2, 10, 11, 12 although the true risk may be underestimated because of a general lack of routine, postoperative imaging of patients who have undergone open aortic repair. The diagnosis is usually made 8 to 10 years after the original surgery.1, 7, 13

Para-anastomotic aneurysms that develop in the setting of previous aneurysmal disease may represent true progression of disease or may be the result of anastomotic disruption leading to pseudoaneurysm formation.2, 13 True aneurysmal change is rarely seen in the setting of previous aortic bypass for aortoiliac occlusive disease.13 Although the distinction between true and false aneurysmal change may not always be clear, differences may exist in the indications for treatment. True para-anastomotic aneurysms may be subjected to the same size criteria as abdominal or iliac aneurysms in regards to repair. Pseudoaneurysms, on the other hand, may be more unpredictable in terms of rupture risk and should be expeditiously fixed in the appropriate surgical candidate.13

Open repair for para-anastomotic aneurysms is associated with a perioperative mortality of 17% in the elective setting2, 3 and 24% in emergency cases.1 Overall morbidity is also high, and includes cardiac, pulmonary, neurologic, and infectious complications.3 One report comparing open with endovascular repair of para-anastomotic aneurysms found a decreased operative time, blood loss, and length of stay in the endovascular group. Perioperative mortality could not be compared in that experience owing to the small number of patients.7 In our series, perioperative mortality ≤30 days of endovascular repair was 3.7%, and the 30-day complication rate, including endoleaks, was 23%. However, late mortality within a mean follow-up period of 18 months was 49% and was significantly associated with older age at the time of endovascular repair. Although endovascular repair of aneurysmal disease in older age groups, particularly octogenarians, is known to be feasible and associated with acceptable outcomes,14, 15 long-term survival after para-anastomotic aneurysm repair in this age group may be limited.

Immediate technical success was achieved in 98% of patients, as documented on completion angiography. Six (75%) of eight endoleaks were diagnosed ≤30 days, although three (38%) were classified as type II. Two of the patients with type II endoleaks were observed without consequence, but one required explantation of the graft for other reasons. Thus, the 30-day clinical success rate was 84% (49/53). Type I endoleaks were significantly associated with nonelective repair in patients who were symptomatic and may have been related to the emergency use of imperfectly sized endografts in patients with difficult anatomy.

Of interest was that the location of the para-anastomotic aneurysm did not significantly correlate with the development of endoleaks. Predictors of type I endoleaks include short, angulated aortic necks, which are characteristic of proximal para-anastomotic aneurysms of the infrarenal aorta after aortoiliac bypass for occlusive disease or aortic replacement for aneurysms.16 In our series, para-anastomotic aneurysms in this location were treated uniformly with transrenal device fixation, which may help prevent some proximal type I endoleaks and promote aneurysm exclusion.17

Most patients who presented with para-anastomotic aneurysms had previously been treated for true abdominal, thoracic, or aortic arch aneurysms. Eight patients (15%) had a history of previous aortoiliac occlusive disease, and aortobifemoral bypass surgery. Patients with aortoiliac occlusive disease had a significantly longer time to presentation than those who were reconstructed for aneurysmal disease. This may reflect a higher incidence of true aneurysmal degeneration in the patients originally treated for aneurysms than those treated for ischemia. Data from Edwards et al12 supports the finding that para-anastomotic pseudoaneurysms develop later than true aneurysmal degeneration at or near a suture line, although early infection and postoperative complications correlate with the early formation of pseudoaneurysms.

Lower extremity ischemia represented the most common late complication, occurring in four patients (7.5%). Ischemia was secondary to limb occlusion in two patients (3.7%) who were treated for common iliac para-anastomotic aneurysms. All patients with limb ischemia presented ≤1 year of endovascular stent graft repair, and three were successfully treated with an extra-anatomic bypass. Other series have reported a slightly lower incidence of limb occlusion of 2% to 3% and have found extension of the endograft into the external iliac artery to be a significant risk factor.18 In our series, two patients with limb occlusion had endografts extended to the external iliac artery, although these numbers are too small to assess the significance.

Conclusion 

Para-anastomotic aneurysms after open aortic repair represent a high-risk lesion that warrants repair in the appropriate candidates to prevent rupture. Endovascular management is feasible and is associated with lower perioperative morbidity and mortality rates compared with those that are historically reported for open repair. Devices that offer transrenal fixation may be advantageous for managing para-anastomotic aneurysms that involve the proximal anastomosis of conventional infrarenal aortic prostheses. Similar to endovascular treatment of primary abdominal and thoracic aneurysms, follow-up imaging at regular intervals is mandatory to assess for endoleaks that warrant reintervention. Perioperative mortality after these procedures is low, but long-term survival is limited, which is partly related to the age at which patients present with these types of aneurysms. Fig 1.

View Large Image Download to PowerPoint

Fig. A, The computed tomography image demonstrates a proximal para-anastomotic aneurysm that was found incidentally 10 years after open repair for an abdominal aortic aneurysm. B, The intraoperative image demonstrates the aneurysm before deployment of a Talent device. C, Intraoperative image after device deployment.

Author contributions