An endoleak is defined as persistent blood flow outside the lumen of the endograft but within the aneurysm sac, verified by an imaging study [1]. Endoleaks can be classified according to Veith et al. [2].
Type I | Inadequate sealing of the landing zones
|
|---|---|
Type II | Retrograde blood flow in the aneurysm sac via collateral vessels (mainly the inferior mesenteric artery and lumbar arteries, occasionally an accessory renal artery)
|
Type III |
|
Type IV | Porosity of the graft fabric (usually self-limiting) |
Type V | Endotension (aneurysm growth without evidence of endoleak) |
Endoleaks can be subdivided into three groups based on when they manifest [3]:
- Early endoleak: within 24 hours
- Intermediate endoleak: postinterventional day 1 to 90
- Late endoleak: after 90 days
Type I endoleak
With type I endoleaks, there is a significant increase in pressure within the aneurysm sac with a corresponding risk of rupture. The incidence of type Ia endoleaks increases with:
- a short aneurysm neck (< 15 mm)
- a wide neck diameter (> 32 mm)
- an aneurysm neck with distal enlargement
- increased angulation (> 60 degrees)
- calcification and thrombus material in the region of the landing zones
Type I endoleaks should therefore be repaired where possible before the patient leaves the intervention room [5, 6], although this is not always achievable. In a cohort study by the Vascular Study Group of New England (VSGNE), 80 (3.3%) of 2402 patients who underwent elective EVAR had a persistent type I endoleak [6]. Patients with a type I endoleak had a higher in-hospital mortality compared with others, but did not differ in 1-year mortality, and in 94% of cases the endoleak had resolved within 1 year without intervention. Similar experiences have been reported by other studies [7]. If interventional repair of a type I endoleak is not achieved during the intervention, the above data suggest that watchful waiting is initially an option, but patients must be carefully followed up. Any increase in size of the aneurysm sac can be monitored by ultrasonography where appropriate expertise is available; follow-up intervals of three months appear to be adequate.
Early type I endoleaks can undergo balloon remodelling, and so-called extension cuffs and bare metal stents may also be used. There is also the endoluminal option of pinning stent graft components to the vessel wall by means of a stapler or endo-anchors [8].
In late phases with a persistent type I endoleak, an attempt can be made to fill the lumen between the endograft and the vessel wall using embolic agents. If the aneurysm diameter continues to progress, extension of the proximal or distal landing zone by stent grafts or open surgical repair may be considered.
Type II endoleak
There is a systematic review of the significance of type II endoleaks based on 32 non-randomised retrospective studies involving 21,744 patients [9]. Type II endoleaks were observed in 10.2% of patients after EVAR, and spontaneous regression occurred in 35.4% of cases. Aneurysm rupture was seen in 14 patients (0.9%) with an isolated type II endoleak, although in 6 of these patients the aneurysm had not expanded. Out of 393 interventions for type II endoleak, 28.5% were unsuccessful. The authors of the review concluded that conservative management of type II endoleak is safe, which is in line with data from other studies [10]. Another systematic review was unable to identify a threshold beyond which a type II endoleak requires intervention [11]. Given the rarity of aneurysm sac enlargement and rupture in type II endoleak, it is referred to as a "benign complication", the treatment of which should be decided on an individual basis. A retrospective study from 2014 reached the same conclusion [12].
Large-calibre lumbar arteries can lead to type II endoleaks [13], as can a large inferior mesenteric artery. Treatment of the endoleaks can include embolisation of the aforementioned vessels via a microcatheter. After EVAR, the inferior mesenteric artery is accessed via the arc of Riolan with probing of the superior mesenteric artery; the lumbar arteries are reached via the collateral circulation from the internal iliac artery. Embolisation often has to be repeated, because after occlusion of the vessels further collateral pathways open up. If no arterial access is available, treatment can also be carried out by direct puncture of the aneurysm sac and localisation of the endoleak (from posterior under local anaesthesia) [14]. Minimally invasive stapler transection of the inferior mesenteric artery is also a treatment option. The success rate of interventional treatment of type II endoleaks is 60% to 80%. If the feeder vessels cannot be excluded by interventional means and an increase in size is observed, surgical treatment is indicated.
Type III endoleak
The incidence of type III endoleaks is less than 4% at 1-year follow-up [15, 16]. The leaks are associated with an increased risk of rupture and should be treated promptly [17]. For treatment, the endograft is lined internally, either wholly or partially, with a second endograft. The same also applies to disconnection of individual stent graft components. Open surgical repair is rarely required.
Type IV endoleak
Type IV endoleaks result from endograft porosity or leakage. Typically, toward the end of the intervention, the leaks present as a perigraft contrast cloud persisting for several seconds during the final verification angiogram [17]. In most cases, the leaks resolve within 24 hours once the heparin effect wears off. Persistent type IV endoleaks generally do not require treatment [18]. In the case of unexpected progressive aneurysm growth, internal stenting ("relining") of the endograft becomes necessary.
Type V endoleak
In this endoleak type, also known as endotension, there is a continuous increase in the size of the aneurysm sac with no apparent contrast leakage. Usually, these endoleaks are self-limiting. An increase in size and imminent rupture require reintervention, either by implanting a second endograft within the first endoluminal graft or by open surgical repair.