Endoleak following endovascular aortic repair

INTRODUCTION — Following endovascular repair of abdominal aortic aneurysm (AAA), endoleak (all types) has been reported in 20 to 50 percent of patients. The reported rates vary depending on the follow-up length and the imaging modality used for their detection [1]. Manifestations of endoleak range from asymptomatic to uncontrolled aortic rupture. Endoleaks result from a poor seal at the proximal or distal fixation sites or between the graft components, from patent lumbar or visceral vessels, or from the result of graft material failure.

Endoleaks may be identified on completion arteriography at the time of endovascular aneurysm repair (EVAR)or during later follow-up at the time of endograft surveillance imaging. The type of endoleak, classified as type I through type IV, determines its clinical significance and treatment. Some types of endoleak clearly affect short-term and long-term outcomes. There is little controversy that type I and III endoleaks should be treated, but there is debate about the clinical importance of type II endoleak. Reports conflict regarding the optimal timing and type of treatment. Most endoleaks are managed successfully with the placement of additional stents or by using embolization techniques, but sometimes open surgical repair is needed.

The classification and management of endoleaks that result from EVAR will be reviewed here. Types of endovascular devices and endovascular repair of abdominal aorta and thoracic aorta are reviewed separately. (See "Endovascular devices for abdominal aortic repair" and "Endovascular devices for thoracic aortic repair" and "Endovascular repair of abdominal aortic aneurysm" and "Endovascular repair of the thoracic aorta".)

ETIOLOGY AND CLASSIFICATION — For patients with suitable anatomy, an endovascular approach is widely accepted as the preferred initial approach for repair of an infrarenal abdominal aortic aneurysm (AAA) or descending thoracic aortic aneurysm [2-7]. One of the limitations of EVAR is the need for ongoing surveillance, primarily to identify endoleaks, which may require reintervention [8]. (See "Management of asymptomatic abdominal aortic aneurysm" and "Endovascular repair of abdominal aortic aneurysm".)

Successful endovascular repair must completely exclude the aneurysm from the arterial circulation to eliminate the effects of systemic blood pressure on the aneurysm sac, which prevents aneurysm expansion and rupture. Persistent arterial perfusion of the aneurysm sac after endovascular treatment indicates a failure to completely exclude the aneurysm and is defined as endoleak [9,10]. Endoleak is associated with an ongoing risk for aneurysm expansion or rupture.

The types of endoleak are classified by the source of the leak (table 1 and figure 1) [4,9]. The type of endoleak determines the magnitude of intrasac pressure.

●Type I endoleak is due to an incompetent seal at the proximal or distal attachment sites. Type I endoleaks can occur immediately after device placement or can develop later over time. (See 'Type I endoleak' below.)

•Type Ia – Proximal type I endoleak typically reflects incomplete apposition of the stent-graft to the aortic neck (image 1). Contributing factors include mural thrombus; a short, angulated, dilated, severely calcified, or reverse-tapered aortic neck; and incorrect device sizing. The use of an endograft with or without suprarenal fixation does not seem to affect the risk of type I endoleak [11,12].

•Type Ib – Distal type I endoleak is usually due to lack of an appropriate seal zone as recommended by the device manufacturer instructions for use, incorrect sizing of the iliac limbs or inadvertent deployment of the endograft limb, often because of excessive iliac tortuosity within the more proximal, larger iliac vessel (image 2).

Late type Ia or type Ib endoleak can develop as a result of conformational changes in the aneurysm sac, aneurysmal degeneration of the aortic neck or iliac arteries, severe angulation at the fixation sites, or graft migration (image 3 and image 4) [13-15]. Patients requiring large proximal diameter devices (34 to 36 mm) to seal dilated aortic necks or flared iliac limbs (>20 mm) to seal dilated iliac arteries are reported to have a higher risk of late type 1a or 1b endoleak, respectively [16,17].

●Type II endoleaks are due to the presence of a patent inferior mesenteric artery or patent lumbar branches or, in the case of thoracic aortic aneurysm repair, patent intercostal arteries. These allow retrograde flow into the aneurysm sac (image 5). The incidence of type II endoleak has been correlated with the number of patent aortic branches prior to endovascular repair of the aneurysm [18]. (See 'Type II endoleak' below.)

●Type III endoleaks (image 6) are due to a junctional leak, disconnection of the endograft components (type IIIa), suture breaks, metal ring fracture, or holes in the endograft fabric (type IIIb) [19,20]. As with type I endoleak, type III endoleaks can occur early or late. (See 'Type III endoleak' below.)

●Type IV endoleak is defined as increased graft wall "porosity" leading to exudation of plasma components across the wall of the endograft. (See 'Type IV endoleak' below.)

●Continued aneurysm sac expansion without a demonstrable endoleak on any imaging modality is referred to as endoleak of undefined origin but has also been termed type V endoleak or endotension [15]. (See 'Endoleak of undefined origin' below.)

While there is little controversy over type I and III endoleaks, reports conflict regarding the natural history, detection and follow-up, optimal timing, and type of treatment for type II endoleaks. The most common types of endoleak (I and II) are usually managed successfully with the placement of additional stents or embolization techniques, but sometimes surgery is needed.

CLINICAL MANIFESTATIONS — Most endoleaks manifest no symptoms. For patients who do exhibit clinical manifestations, these may range from findings consistent with an unrepaired aneurysm to contained rupture (back/flank pain) or overt rupture (hemodynamic instability) [8,21,22]. (See "Clinical features and diagnosis of abdominal aortic aneurysm", section on 'Ruptured AAA'.)

Endoleaks that pressurize the aneurysm sac sufficiently may be suspected on physical examination when the aneurysm remains pulsatile on palpation. (See "Clinical features and diagnosis of abdominal aortic aneurysm", section on 'Abdominal palpation'.)

DIAGNOSIS — The diagnosis of endoleak is frequently made at completion arteriography during endovascular aortic repair (EVAR) but may be also identified on follow-up imaging that demonstrates blood outside the bounds of the endograft. For some types of endoleak, the source can be difficult to determine.

Post-EVAR imaging protocols to detect endoleaks and/or endograft migration rely on computed tomographic (CT) angiography, given its availability and standardization. CT angiography can provide sac size and volume measurements, which can help guide surveillance and management, particularly when an endoleak cannot be clearly identified [23]. Duplex ultrasound with or without contrast as well as magnetic resonance (MR) angiography have been suggested as alternative imaging modalities [24-26]. In systematic reviews, MR imaging detected more endoleaks compared with CT angiography, particularly type II leaks [1,27]. Duplex ultrasound is less expensive and safer and may be more sensitive for the diagnosis of endoleaks [8,24,26,28-34].

Duplex ultrasound has the specific advantage of detecting flow direction of endoleaks, facilitating identification of the type of endoleak; Doppler waveforms may even be able to predict the natural history of type II endoleaks [35]. Contrast-enhanced ultrasound (CEUS) may have better accuracy, particularly for endoleak detection and classification [24]. In meta-analyses, the accuracy of CEUS was comparable to CT angiography, if not better, particularly for the detection of late type II endoleaks [26,31-33]. In the latest review, which included 26 studies, the pooled sensitivities and specificities of CEUS for all endoleaks were 0.94 (95% CI 0.89-0.97) and 0.93 (95% CI 0.89-0.96), respectively [34]. The pooled sensitivities and specificities for type I and type III endoleaks were higher at 0.97 (95% CI 0.8-1.00) and 1.00 (95% CI 0.99-1.00), respectively. Unfortunately, ultrasound accuracy is operator-dependent, and technical difficulties limit its universal application, particularly in patients with large body habitus. Wireless pressure sensors implanted during EVAR have shown some value in monitoring intrasac pressure, but sufficient data supporting their use are still lacking [36].

Most type I and type III endoleaks will be identified with completion arteriography. Some type II endoleaks may also be identified during the completion angiogram, provided that delayed images are taken. If type II endoleak is not apparent initially, it may be identified during surveillance. Since the aneurysm sac fills through collaterals, to properly detect type II endoleak on CT angiography, three runs are required: one without intravenous contrast, one with contrast, and a third delayed run during the venous phase [8]. Nonetheless, detection of type II endoleaks can be challenging, and they can be easily missed due to low flow, or they may not be seen if other types of endoleak cause intravenous contrast overlap.

Component separation associated with type III endoleak can be identified with plain abdominal films and confirmed with CT angiography. It may not be easy to distinguish component separation from a type I or type II endoleak, and it is even more difficult to differentiate type II endoleak when the source of the leak is a fabric tear. Type III endoleak is usually easily recognized with conventional arteriography but may require selective limb imaging with contralateral limb balloon occlusion.

ENDOLEAK MANAGEMENT AFTER EVAR — Endoleaks occurring at any time after endovascular abdominal aortic aneurysm repair (EVAR) are reported to occur in 20 to 50 percent of patients [37-39]. Type II endoleak is the most commonly reported endoleak, representing over 50 percent of all endoleaks. Studies have differed widely, though, depending on the type and generation of device used and duration of follow-up.

The Veterans Affairs Open Surgery Versus Endovascular Repair (OVER) trial reported an approximately 30 percent rate, with an average follow-up of six years [38]. Fifty-three percent of endoleaks resolved spontaneously, and 32 percent required secondary interventions. The distribution was as follows:

●Type I – 12 percent

●Type II – 76 percent

●Type III – 3 percent

●Type IV – 3 percent

●Indeterminate – 6 percent

The management of endoleak depends up on the type of endoleak, and each is discussed below (algorithm 1).

Type I endoleak — Type I endoleak is due to an incompetent seal at the proximal (Ia) or distal (Ib) attachment sites and represents a failure of EVAR (table 1 and figure 1). Type I endoleak can occur immediately after device placement or can develop later over time. The majority of type I endoleaks are identified during completion arteriography at the time of the endovascular procedure (image 1). If not captured or present at completion, they are usually identified during the one-month or annual surveillance imaging (image 7 and image 4). Type I endoleaks should be repaired as soon as discovered to prevent adverse events [15]. Several studies have shown a significant association between type I endoleak and late rupture, open conversion, and death [11,15,40,41].

Even with meticulous planning including judicious patient and stent-graft selection/sizing and implantation of the stent-graft with optimal positioning, type I endoleak is still reported. The incidence of type I endoleak ranges between 3 and 21 percent of endovascular aortic aneurysm repairs, with the higher incidences reflecting difficult anatomic situations (eg, short neck), failing first-generation endografts, endografts used outside the manufacturer's instructions for use (IFU), or very small endoleaks captured on higher-quality imaging studies [11,37,42].

In the largest published series coming from the EUROSTAR registry (3595 patients, 1996 to 2002), the overall reported incidence of type I endoleak was approximately 7 percent (3.3 percent type Ia endoleak, 3.7 percent type Ib endoleak), but the overall incidence increased to 10 percent during follow-up. Many of these eventually resolved [37,43,44]. In a review of 2402 patients from the Vascular Study Group of New England database from 2003 to 2012, type I endoleak was identified in 3.3 percent of EVARs at completion of surgery. Approximately 1 percent of patients with no type I endoleak during initial completion arteriography developed new type I endoleak at one-year follow-up [11].

Initial approach to type I endoleak — For type I endoleaks identified at the time of endograft placement, the initial approach consists of reballooning the fixation sites and possibly the reversal of intraoperative anticoagulation (algorithm 1). (See "Endovascular repair of abdominal aortic aneurysm", section on 'Anticoagulation'.)

●For proximal type Ia endoleaks that persist after reballooning, the placement of additional aortic cuffs or a balloon-expandable stent (eg, Palmaz) increases the radial force exerted by the proximal graft and maximizes apposition of the graft to the aortic wall [45,46]. In cases where the graft is deployed flush with the lowest renal artery and appears well apposed to the aortic wall, observation is reasonable since many of these will resolve at one-month follow-up computed tomography (CT) scan surveillance.

●For distal type Ib endoleaks that persist after balloon angioplasty of the distal attachment site, iliac limb extensions are used. If the iliac limb has been undersized, a flared iliac extension limb can be placed to exclude the endoleak. If the distal common iliac artery does not have an adequate length to provide a proper seal, coil embolization of the origin of the hypogastric artery and placement of a limb extension into the external iliac artery may be necessary. It is important to maintain pelvic perfusion through the contralateral hypogastric artery to minimize the risk for pelvic ischemia. (See "Complications of endovascular abdominal aortic repair", section on 'Pelvic ischemia'.)

These techniques are usually sufficient to repair type I endoleak and exclude the aneurysm. If unsuccessful, a conservative approach may be justified, particularly when remaining endovascular options are not feasible and conversion to open repair is considered high-risk. Watchful waiting of selected patients may lead to resolution of type I endoleak. Based on early experience, spontaneous closure of the type I endoleak was thought to be uncommon; however, later studies indicated that small, low-flow endoleaks may have high resolution rates [11,42,47]. The spontaneous disappearance of type Ia endoleak after appropriate stent-graft sizing and deployment in this subgroup is not surprising. Once perioperative coagulation abnormalities return to normal, the self-expanding nature of the stent-graft leads to gradual neck remodeling and improved aortic wall apposition. In a review of the Vascular Study Group of New England database, 90 percent of type I endoleaks resolved without the need for endovascular intervention or open conversion [11]. However, high-flow type I endoleaks will typically lead to significant pressure elevation in the aneurysm sac and aneurysm expansion, which has been linked to an ongoing risk for rupture, open conversion, and death [11,40].

Refractory and late type I endoleak — The percutaneous treatment of late or persistent type Ia endoleak can be challenging. In refractory cases, more advanced endovascular options may be needed. In situations where visceral branches preclude graft extension, glue or coil deposition in the tract between the stent-graft and the aortic wall has been used successfully [46]. Alternatively, placement of stents in the renal and potentially superior mesenteric arteries alongside a proximal endograft extension (Chimney technique) potentially allows the resolution of type I endoleak with reasonable results in the mid-term [48-50]. Fenestrated or branched devices may also provide a solution for patients with appropriate anatomy but require advanced endovascular skills and expertise [51,52]. Devices (eg, endostaples, EndoAnchors) are also available and may be used preemptively to enhance proximal fixation and seal [53-55]. These penetrate the stent-graft and secure it to the full thickness of the aortic wall. If used to treat an existing type I endoleak, they may not be as successful. Early reviews of the Aneurysm Treatment Using the Heli-FX Aortic Securement System Global Registry (ANCHOR) showed that 34 percent of the treated patients with type Ia endoleak remote from initial EVAR had evidence of type Ia endoleak during follow-up [56]. In a propensity-matched cohort of subjects undergoing EVAR, the rate of sac regression in subjects treated with EndoAnchors was significantly higher (81 versus 49 percent) [57]. Later studies support early use in hostile necks but long-term data are lacking [53-55].

Late distal type I endoleak can be treated most of the time with graft extensions to the iliac bifurcation or using iliac branch devices to preserve flow to the hypogastric artery [58-60]. Nevertheless, embolization of the ipsilateral internal iliac artery and distal graft extension is usually the most straightforward solution, particularly when the contralateral internal iliac is patent.

Open conversion is almost never required at the time of EVAR but is sometimes required for delayed type I endoleak. Conversion can be associated with increased morbidity and mortality, although case series report rates comparable with de novo open repair of abdominal aortic aneurysm (AAA). Open conversion can be performed via a transperitoneal or retroperitoneal approach, with initial proximal aortic control often above the renal arteries, depending upon the graft location and the presence of suprarenal fixation. Stepwise clamping more distally on the aorta is performed to reduce ischemic time, with complete or partial endograft removal. Partial explantation with in-situ graft replacement can be performed in the absence of graft infection, with the proximal suture line incorporating the suprarenal stents and the distal suture line of either a tube or bifurcated graft incorporating the iliac limbs [61]. In high-risk patients with multiple comorbidities, banding of the aortic neck can be considered with anecdotal reports of success. (See "Open surgical repair of abdominal aortic aneurysm".)

Type II endoleak — Following EVAR, type II endoleaks are due to a patent inferior mesenteric artery or lumbar branches that allow retrograde flow into the sac. Type II endoleak may also arise from other aortic branches, such as the median sacral artery or accessory renal arteries (table 1 and figure 1) [8,62,63]. Type II endoleaks can have a simple inflow and outflow vessel pattern or may be more complex with a nidus of involved vessels behaving like an arterial–venous malformation with recruitment of vessels over time.

The EUROSTAR registry (3595 patients, 1996 to 2002) reported a 9 percent rate of type II endoleak diagnosed at any time during follow-up [44,64], similar to the 8 to 10 percent rates reported in large systematic reviews and meta-analyses [3,11,37,43,44,65-68].

The incidence of Type II endoleak has been correlated, though not uniformly, with the number and/or size of patent branches prior to exclusion [18,63,65,69-71]. Early reports showed a possible relationship to the type of endograft, but studies with longer follow-up did not confirm this notion [71-73]. In a systematic review, a lower risk for type II endoleak was associated with smoking, peripheral artery disease, thrombus load (thickness at the maximum aneurysm diameter), presence of circumferential thrombus, and presence of thrombus at the level of the inferior mesenteric artery [63,74]. Risk factors associated with persistent or new type II endoleak appear to differ. In a review of 2367 endoleaks from the Vascular Study Group of New England database, multivariate analysis identified hypogastric artery embolization, distal graft extension, older age, and type of graft, but not aneurysm size, as significant risk factors [73].

Most type II endoleaks are identified early in the course following endovascular repair (image 5). The incidence is generally higher at one to six months, and, while some type II endoleaks may be identified at a later date, the rate gradually decreases to <10 percent at two years [66,71]. However, occult type II endoleaks may be revealed up to five years following endovascular repair [8,75-77].

While there is no clear evidence that sac expansion in this setting is a surrogate for future aneurysm rupture, aneurysm rupture does rarely occur [8,62,67]. Among patients with type II endoleak from the EUROSTAR registry, the cumulative two-year incidence of rupture was 1.8 percent (1 out of 55 patients); however, this rate, although twice as high, was not significantly different compared with patients without any detected endoleak (0.9 percent, 5 out of 548 patients) [65]. In a meta-analysis that involved 21,744 patients with 1515 type II endoleaks following EVAR, the rupture rate of all type II endoleaks reviewed was 0.9 percent; 43 percent of these ruptures had no evidence of sac expansion [67].

Criteria for treatment — Based upon their benign natural history and very rare association with rupture (approximately 1 percent), observation and surveillance are suggested, rather than any form of intervention, for most patients with type II endoleak (algorithm 1). This practice is supported by a multicenter study involving 2000 patients following endovascular repair with a five-year median follow-up, in which overall survival was similar for patients with and without type II endoleak. In addition, survival rates were similar for patients who underwent a secondary intervention compared with those who did not undergo a secondary intervention [78]. This is also supported by available United States and European guidelines [2,4,79].

Spontaneous resolution of type II endoleak is common. Approximately 60 percent of immediate postoperative type II endoleaks will resolve spontaneously within six months, with the remainder regressing over time [69,80]. Interestingly, the aneurysm sac may shrink in spite of the presence of a patent type II endoleak (image 8). Although sac expansion may or may not be a surrogate for an increased risk for rupture, expansion >5 mm during interval follow-up remains the best guide available [2,4]. Intervention is not warranted for endoleak persistence alone.

Factors that have been linked to spontaneous endoleak resolution include coronary artery disease, peripheral artery disease, chronic obstructive pulmonary disease, cancer, and smoking [65,73,80-82]. Internal iliac artery coil embolization, long-term anticoagulation, and dual antiplatelet therapies may cause them to persist. (See 'Alterations to antithrombotic therapies' below.)

The fate of persistent type II endoleaks is variable. Aneurysm sac size may decrease, remain stable, or increase. In 5 to 24 percent of patients, the aneurysm sac will remain stable or may expand >5 mm. For type II endoleaks that persist for more than six months, up to 55 percent of patients will have some aneurysm sac expansion [65,80,83,84]. The decision whether to treat persistent type II endoleak is based on sac size and expansion rate (Society for Vascular Surgery [SVS] guidelines: ≥5 mm; European Society for Vascular Surgery [ESVS] guidelines: ≥1 cm), the type and size of patent inflow and outflow vessels, and the presence of symptoms. This selective intervention approach is safe and recommended by current guidelines [2,4,79].

Treatment techniques — Obliteration of type II endoleaks can be difficult. The main goal of treatment is to eliminate the aortic branches at their junction with the aneurysm to abolish side branch perfusion. A variety of endovascular, laparoscopic, and open techniques has been proposed to accomplish this goal. While all alternatives may have a role, endovascular techniques are preferred given their minimally invasive nature.

●Transarterial embolization – Transarterial embolization is the most common intervention for type II endoleak. Transfemoral or transbrachial access is obtained, and a coaxial system with microcatheters is used to reach the inferior mesenteric or lumbar arteries to embolize the nidus and the feeder branches of the endoleak. The approach is generally feasible through the internal iliac (superior gluteal) or superior mesenteric (middle colic and marginal) arteries depending upon the target and associated collateral vessels (image 9). Microcoils are mainly used, but glue, thrombin, and Onyx (ethylene-vinyl alcohol copolymer) can also be delivered, with no proven superiority of one agent over the other (image 10) [8].

●Translumbar embolization – Translumbar embolization is a reasonable alternative when transfemoral access to the endoleak is not available (eg, occluded internal iliac or inferior mesenteric arteries, prior transarterial embolization). Not all vascular surgeons are familiar with this technique given the rare need for its use [8,82,85-87]. With the patient in prone position, access to the aneurysm sac is obtained under fluoroscopy, typically from a left paraspinal approach (approximately four fingers' breadth from midline). A sheath is introduced, and, following an aortic "sacogram," a coaxial system with microcatheters is created to reach the nidus and the target arteries, similar to the transarterial approach (image 11). Some authors have also suggested ventral or transcaval aneurysm sac puncture, but data are limited for these approaches [82,88,89].

An important aspect of success, regardless of the approach used, is gaining access and embolizing the sac (nidus) with as many feeders (inflow and outflow) as possible, a task that requires advanced endovascular skills [30,67,87]. Despite high technical success, recurrence of endoleak is common, and multiple interventions are often needed to stabilize aneurysm sac size in patients diagnosed with type II endoleak-associated sac expansion. In a systematic review, the success rate, defined as no recurrence during follow-up, averaged 62.5 percent (range 15 to 89 percent) for the transarterial approach and 81 percent (range 67 to 100 percent) for the translumbar approaches [67]. All studies that compared transarterial with translumbar approaches have shown better results for the translumbar method and less fluoroscopy and procedure time; however, in the majority of cases, translumbar embolization was the second-line treatment after a transarterial approach had failed, and thus such comparisons may not be valid [90].

Unsuccessful attempts treating type II endoleaks and/or a rapid aortic sac growth of 5 mm per year or greater should raise the suspicion of a delayed or occult type I or III endoleak, which should be investigated and treated [91]. (See 'Type I endoleak' above and 'Type III endoleak' below.)

Should type II endoleak persist and the sac continue to expand following "technically" successful embolizations, or when embolization is not possible, a surgical approach may become necessary [90].

Laparotomy or laparoscopy with ligation of the feeding side branches, suturing of the side branch ostia within the aneurysmal sac but leaving the stent-graft intact, or, ultimately, conversion to open repair are all alternatives that have increasing complexity, morbidity, and mortality [92-94].

Although pre-emptive embolization of aortic side branches has been reported an lowers the incidence of type II endoleak post-EVAR , its cost-effectiveness and any impact on the risk of rupture are not known, and this approach is not recommended by current guidelines [95].

Type III endoleak — Type III endoleaks are due to a junctional leak or disconnect of the endograft components (type IIIa) or holes in the endograft fabric (type IIIb) (table 1 and figure 1) [20].

As with type I endoleaks, type III endoleaks can occur early or late (image 6).

●Early type III endoleaks are leaks between endograft components noted on completion arteriography. These are most often due to insufficient overlap between graft components or inadequate balloon expansion at the junctions of the endograft components.

●Late type III endoleaks can develop months to years later, with partial or complete separation of endograft components. Late type III endoleaks have been related to conformational changes in the aneurysm sac, endograft migration, and/or dilation of aortic and iliac attachment sites [96]. The dominant forces on modular endovascular graft components are directed laterally and increase with aortic or endograft angulation [97,98]. The resulting endograft displacement is more prevalent with larger aneurysms and is associated with an increased incidence of type IIIa (and type I) endoleaks.

The reported incidence of type IIIa endoleaks across multiple endograft platforms is 2 to 3 percent [98-100]. The incidence of type III endoleak was relatively high with early generation endografts, which were susceptible to material fatigue, fabric rents, and limb disconnection. The improved technology and testing of later-generation endografts has nearly eliminated material failure, and longer overlap zones with larger inserts have decreased the occurrence of limb disconnections [96]. In 2017, the Food and Drug Administration reported an apparent increase in the occurrence of Type III endoleaks, potentially related to early-generation graft materials, but also to the presence of calcified plaque and inadequate overlap between graft components. The recommendation was to consider lifelong surveillance of patients who have been treated with EVAR and consider type III endoleaks in the differential diagnosis of patients who present with symptoms of potential aneurysm expansion or rupture. With device improvements, it is likely that type IIIa endoleaks are more common compared with type IIIb endoleaks. These are more prevalent with complex fenestrated EVAR procedures that have an increased number of endograft junctions. 

The most serious consequence of type III endoleak is pressurization of the aneurysm sac with subsequent expansion of the aneurysm and rupture. In an analysis of EUROSTAR registry data, patients with late type III endoleak had an almost nine-times-greater risk of aneurysm rupture compared with other registry patients, even when compared with those who had a type I endoleak [41]. As such, type III endoleaks are at least as serious type I endoleaks and should be treated as soon as they are identified to prevent the possibility of aortic rupture (algorithm 1). Endograft uncoupling can also obstruct aortic blood flow, leading to AAA thrombosis and acute lower extremity ischemia [98]. (See "Clinical features and diagnosis of acute lower extremity ischemia".)

Endovascular treatment is typically straightforward, involving the deployment of additional stent-graft components to seal the fabric defect or bridge the disconnected components. The main technical challenge is cannulation of the main body gate, which can be particularly difficult if the components are offset. Once cannulation has been accomplished, a new iliac limb graft can be deployed. Another option is to deploy a new bifurcated stent-graft, thereby rebuilding an entirely new device within the existing device. This technique is appropriate for situations of significant migration from the proximal landing zone or in cases of multiple component separations [96].

Type IV endoleak — Increased graft wall "porosity" was first defined as a type IV endoleak early in the endograft experience because of frequently noted small "blushes" within the aneurysm sac during completion arteriography (table 1 and figure 1) [20]. It was hypothesized that these small holes would thrombose in the immediate postoperative period; however, first-generation endografts with high permeability still allowed for aneurysm expansion [101,102].

This problem has typically been associated with some thin polyester grafts but has peculiarly not been universally noted in all procedures with endografts prone to exhibit type IV endoleak. With the current-generation endografts, type IV endoleak is self-limited, typically resolving in 24 hours (algorithm 1). It has not been associated with any long-term adverse events and does not require treatment. It is important to note that type IV endoleak can be quite disconcerting to see at completion arteriography, and it can obscure more serious type I or type III endoleaks.

Endoleak of undefined origin — Endoleak of undefined origin is continued aneurysm sac expansion without a demonstrable leak on any imaging modality and is also referred to as type V endoleak or endotension (table 1 and figure 1) [15]. This phenomenon is poorly understood but has been linked in a few cases to an incompletely sealed landing zone with laminated thrombus preventing the demonstration of an endoleak.

More commonly, endoleak of undefined origin is associated with a semiporous graft material reported with the original expanded polytetrafluoroethylene (e-PTFE) endografts (eg, Excluder). Sac enlargement was associated with a protein-rich exudate related to a transmural fluid leakage, also noted with open surgical grafts. A later reduction in material porosity resolved this issue, confirming the source of this type of endoleak. Treatment of such cases has consisted of relining the existing endograft with a new lower-porosity graft or explantation. (See "Endovascular devices for abdominal aortic repair", section on 'Excluder'.)

Alterations to surveillance — Guidelines for surveillance after endovascular repair of abdominal aortic aneurysm suggest obtaining CT angiography at postoperative day 30. Initially recommended surveillance protocols were consistent with those used by FDA-sponsored pivotal trials, including CT angiography or duplex ultrasound, combined with plain abdominal films or non-contrast CT at six months, and, provided there is no evidence for aneurysm expansion, annual duplex ultrasound surveillance thereafter [2,4,28-30]. (See "Endovascular repair of abdominal aortic aneurysm", section on 'Endograft surveillance'.)

The updated guidelines of the Society for Vascular Surgery recommend baseline imaging in the first month after EVAR with contrast-enhanced CT and color duplex ultrasound imaging [103]. In the absence of an endoleak or sac enlargement, imaging should be repeated in 12 months using contrast-enhanced CT or color duplex ultrasound imaging. Accumulating evidence suggests that omission of the six-month follow-up visit may be acceptable even after detection of a type II endoleak at the one-month postoperative study since it is more often treated conservatively [24,28-30].

Alterations to antithrombotic therapies — Whether to consider temporarily withholding antithrombotic therapies to resolve endoleak, in particular type II endoleak, is controversial and may depend upon the strength of the indication for the treatment (atrial fibrillation, intravascular device) as well as characteristics of the endoleak and aneurysm sac behavior.

The contribution of antithrombotic therapies (antiplatelet agents, warfarin, thrombolytics) to the incidence of endoleak following EVAR is not well studied, and available data are conflicting [81,104-107]. Among patients with a type II endoleak, antithrombotic therapies may be associated with lack of aneurysm sac shrinkage and endoleak persistence. In one systematic review, long-term warfarin therapy compared with no anticoagulation significantly increased the risk for any type of endoleak (odds ratio [OR] 1.77, 95% CI 1.26-2.48) and for persistent type II endoleak (OR 1.58, 95% CI 1.05-2.37) [81].

The effect of thrombolytics on the exacerbation or initiation of an endoleak has been reported in few case reports. A review identified several cases of acute endoleak following administration of thrombolytics [108]. In two of the cases, the acute leak stabilized spontaneously, and in one case, a type III leak led to rupture and eventual death in spite of repair attempts. On the whole, the indications leading to the use of thrombolytic therapy (eg, stroke, pulmonary embolism, acute limb ischemia) are likely to outweigh the seemingly rare occurrence of endoleak in patients with prior EVAR. At a minimum, these patients should be monitored for EVAR-related complications. Provided overt rupture has not occurred (or the indication for EVAR was not rupture itself), observation and follow-up imaging may be all that is required.

ENDOLEAK MANAGEMENT AFTER TEVAR — As with abdominal aortic endografting, patients undergoing thoracic aortic endografting may develop any of the types of endoleaks, and these patients also require ongoing surveillance to evaluate the size of the sac and graft morphology. Compared with endografting for abdominal aortic aneurysm, there are considerably less data on endoleaks available following thoracic endovascular aortic aneurysm repair (TEVAR).

In 2006, analysis of the European Talent Thoracic Retrospective Registry identified a type I and III endoleak rate of 10.4 percent [109]. Subsequent single-institution series and registries have confirmed that the overall rate of any persistent endoleak after TEVAR ranges from 5 to 11 percent, with type I endoleak as the most frequent type [110-113].

Type I and III endoleaks occurring after TEVAR require definitive repair, such as restenting or open surgical management [114-117]. While a type III endoleak can be treated simply with a bridging graft, a type I endoleak may be challenging to treat when the proximal or distal landing zone is short or aneurysmal. Advanced endovascular techniques using side branches are options that require special expertise, particularly when the aortic arch is involved. Endostapling in TEVAR is associated with lower technical success, higher perioperative mortality, and potential serious adverse events [55]. If endovascular treatment of a type I endoleak is not feasible, a hybrid procedure that uses debranching of the supraaortic vessels may be needed to allow extension of the seal zones for supplemental endografting. (See "Endovascular devices for thoracic aortic repair", section on 'Advanced devices'.)

Type II endoleaks require reintervention only if they are associated with aneurysm expansion. When related to the origin of the left subclavian artery, type II endoleaks can be treated with ligation, plugging, or coiling the subclavian with or without subclavian revascularization and/or regrafting more proximally [118]. The left upper extremity is almost always resilient to subclavian artery occlusion given the collateralization around the shoulder (figure 2). The decision to revascularize depends on the estimated risk of spinal ischemia, which is increased with prior abdominal aortic repair and occluded internal iliac arteries. Revascularization options include left subclavian to carotid transposition and carotid-subclavian bypass. (See "Surgical and endovascular techniques for aortic arch branch and upper extremity revascularization", section on 'Extra-anatomic bypass'.)

Other branch vessel endoleak sources can be embolized via translumbar or transthoracic techniques, with coils or biologic glues. Ligation via thoracoscopy or thoracotomy has also been described [114].

ENDOLEAK MANAGEMENT AFTER COMPLEX ABDOMINAL ENDOGRAFTING — Complex fenestrated, branched, and chimney endograft designs remove the barrier of the visceral aorta in the repair, thus affording the operator the option of customizing the proximal aortic landing zone to achieve an optimal seal. While such configurations are of proven durability, the increased complexity and modular nature of these devices introduce the potential for additional modes of failure in terms of component separation [119,120]. (See "Endovascular devices for abdominal aortic repair", section on 'Advanced devices'.)

Fenestrated and branched endografts are customized devices incorporating side holes or stents to accommodate the visceral or arch branches. The incidence of proximal type Ia endoleak was 2.8 percent in one of the largest institutional series of 969 fenestrated and branched aortic aneurysm repairs performed between 2001 and 2013 [120]. Increased component instability and aortic mortality in these patients underscore the need to obtain a durable seal with the index operation. Interestingly, completion type I and III endoleaks are common, but many resolve spontaneously by the initial postoperative imaging, suggesting that not all require intervention at the index procedure, a decision to be made by an experienced operator [121].

The chimney technique preserves vital aortic branches in aortic endovascular repair by deploying a covered stent into a vital aortic side branch parallel to the main aortic stent-graft. In this manner, the sealing zone of the aortic stent-graft can be extended beyond the origin of the vital side branch. However, one concern is the so-called gutter endoleak (leak between the main aortic graft and the parallel stent) causing a type I endoleak. A systematic review that involved 831 patients showed a 13 percent rate of early type I endoleak for the visceral chimneys and 11 percent rate for the arch chimneys. Most type I endoleaks were handled conservatively. Late type I endoleak was reported in 25 visceral and 4 percent of arch chimney cases [122].

Comparisons between fenestrated/branched and chimney techniques has not shown differences in endoleak rates [49].

PREVENTION STRATEGIES — Endoleaks are best prevented by proper patient and device selection. (See "Endovascular repair of abdominal aortic aneurysm", section on 'Anatomic suitability' and "Endovascular repair of the thoracic aorta", section on 'Anatomic considerations'.)

Types I and III endoleaks in large part can be avoided by careful stent-graft size selection, ensuring adequately long landing zones proximally and distally and generous graft overlapping in the setting of modular stent-grafts. Assuming appropriate patient selection and graft sizing, measures to prevent type I endoleak include correcting radiologic parallax to ensure correct positioning and appropriate balloon inflation of the attachment sites once the device is correctly deployed.

Techniques have been proposed to prevent type II endoleak, but their use remains controversial [62,123-129]. Some advocate preoperative or intraoperative embolization (or other techniques) to occlude patent aortic side branches with variable success rates. However, this is associated with longer procedure duration, longer fluoroscopy times, risk of coil dislocation, and higher costs [62]. These risks may outweigh any benefit gained so these prevention strategies for type II endoleak are not widespread. (See "Endovascular repair of abdominal aortic aneurysm", section on 'Measures to prevent endoleak' and "Endovascular repair of the thoracic aorta".)

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Aortic and other peripheral aneurysms".)

SUMMARY AND RECOMMENDATIONS

●Endoleak – Endoleak is a term that describes the presence of persistent flow of blood into the aneurysm sac after aortic endograft placement (eg, endovascular aneurysm repair [EVAR]; thoracic endovascular aneurysm repair [TEVAR]). Endoleaks are associated with a continued risk for aneurysm expansion or rupture. Endoleaks are classified based upon the source of the leak (table 1 and figure 1), which determines their clinical significance and treatment. Continued aneurysm sac expansion without a demonstrable endoleak on any imaging modality is referred to as endoleak of undefined origin. (See 'Etiology and classification' above.)

●Diagnosis – A diagnosis of endoleak is made when imaging demonstrates blood outside the bounds of the endograft. This often occurs during completion arteriography at the time of endograft placement. If the endoleak is not apparent initially, it may be identified during postoperative surveillance. Endoleaks have been reported as late as five years after endograft placement, underscoring the importance of ongoing surveillance. (See 'Clinical manifestations' above and 'Diagnosis' above.)

●Endoleak after EVAR

•Type I endoleak – Type I endoleak is due to an incompetent seal at the proximal (type Ia) or distal (type Ib) attachment sites. These can occur early at the time of endograft placement or late as a result of graft migration or conformational changes of the aorta or graft. Type I endoleaks are repaired as soon as they are discovered, which typically involves reballooning the attachment sites or the placement of additional stent components. Conversion to open repair for type I endoleak is rarely necessary at the time of the initial endograft placement but may be necessary for persistent type I endoleak. Watchful waiting of selected patients (persistent low-flow endoleak, high risk for open surgery) may lead to resolution of type I endoleak. (See 'Type I endoleak' above.)

•Type II endoleak – Type II endoleak is the most prevalent type and is due to a patent inferior mesenteric artery or other aortic branches (eg, lumbar, accessory renal). Detecting type II endoleak on computed tomographic (CT) angiography requires delayed image acquisition. For most patients with type II endoleak, we suggest observation and surveillance, rather than any form of intervention (Grade 2C). Spontaneous resolution of type II endoleak is common. Repeat CT imaging can be performed at 12 months, and provided there is no evidence for aneurysm expansion, subsequent duplex ultrasound can be performed annually. Treatment is indicated only for sac expansion >5 mm during interval follow-up. Options for obliteration of type II endoleaks include transarterial embolization (most common), translumbar embolization, or surgical intervention, which is generally reserved for those that do not resolve by other means. (See 'Type II endoleak' above and 'Alterations to surveillance' above.)

•Type III endoleak – Type III endoleak is due to junctional leak or disconnection of the endograft components (type IIIa) or holes in the endograft fabric (type IIIb). The incidence of type III endoleak is overall low due to improved endograft technology and longer overlap zones as well as improvements in fabric and material testing that have nearly eliminated this type of failure. Component separation associated with type III endoleak can be identified with plain abdominal films and confirmed with CT angiography. As with type I endoleak, type III endoleaks can occur early or late and require treatment when identified. The treatment typically involves placement of additional stent-graft components to seal the fabric defect or provide a bridge across the disconnected components and, in some cases, relining the endograft. (See 'Type III endoleak' above.)

•Type IV endoleak – Type IV endoleak is associated with graft porosity and is self-limited, typically resolving in 24 hours. It has been associated with some thin polyester grafts but has not been universally noted with endografts prone to it. Type IV endoleak has not been associated with any long-term adverse events and does not require specific treatment. It is important to note that type IV endoleak can be quite disconcerting to see at completion arteriography, and, more importantly, it can obscure the identification of more serious type I or type III endoleaks. (See 'Type IV endoleak' above.)

●Endoleak after TEVAR – Patients undergoing thoracic TEVAR similarly require ongoing surveillance to evaluated for endoleak. Following TEVAR, type I and type III endoleak requires definitive repair. A hybrid procedure that uses debranching of the supraaortic vessels may be needed to allow extension of the seal zones for supplemental endografting. As with type II endoleaks following EVAR, post-TEVAR type II endoleaks require reintervention only if they are associated with aneurysm expansion. Those related to the left subclavian artery can be treated by ligation, plugging, or coiling of the subclavian artery with or without left subclavian revascularization. (See 'Endoleak management after TEVAR' above.)

●Endoleak after complex aortic repair – Complex fenestrated, branched, and chimney endografts introduce the potential for additional modes of failure in terms of component separation and endoleaks. Treatment of endoleaks associated with these devices remains challenging. (See 'Endoleak management after complex abdominal endografting' above.)