Endoleak after stent-graft treatment of abdominal aortic aneurysm: a meta-analysis of clinical studies

Background: Endoleak is the major complication after endovascular treatment of abdominal aortic aneurysm (AAA) and its incidence seems to remain significant. Little is known about the association of device type and configuration with respect to the incidence, location, time of onset and fate of endoleakage. Methods: A meta‐analysis was performed via a Medline search of clinical studies after 1995 dealing with the endovascular treatment of AAA. Details of number of patients treated, configuration and type of endovascular device were collected. Data concerning site of origin, time of occurrence and fate of the endoleak were retrieved, along with information on change in diameter of the aneurysm with time. Results: The 23 publications included reported on 1189 patients. The 1118 patients with successfully inserted transfemoral endovascular grafts experienced 270 endoleaks (24 per cent). The majority arose from the distal stent attachment site (36 per cent), were present immediately after stent‐graft placement (66 per cent) and were persistent in time (37 per cent). Tube grafts were more frequently affected by endoleakage (35 per cent; P < 0·0001), especially at the distal stent attachment site (51 per cent), than bifurcated grafts (18 per cent; P = 0·004) and aortounilateral devices (20 per cent; P = 0·70). Self‐ expandable stent‐grafts were more frequently associated with endoleaks (25 per cent) than balloon‐expandable stent‐grafts (17 per cent) (P = 0·037). Conclusion: Endovascular treatment of AAA is an evolving field. Even after the initial learning curve and attention to device‐related problems, it is still accompanied by a significant number of endoleaks. Uniform presentation of results of treatment is necessary for analysing the effect of differences between patients, aneurysm morphology and device type. © 1999 British Journal of Surgery Society Ltd