The primary objective of treatment for aortic anrecognition of inappropriate blood flow within the aneurysm sac. The source may be one or more of eurysms is to prevent the death of the patient from rupture of the aneurysm. Endovascular grafting, with the anastomoses, the body of the graft itself through defects in the fabric, or aortic side branches backthe attributes of a minimally invasive procedure, has potential for treating patients with this condition more bleeding into the sac. Endoleaks may be present from the time of operation – early or primary endoleak – safely and at less cost than conventional open surgery. However, there are early indications that this new or may develop later – secondary endoleak. Methods that have been employed to detect them include conapproach may not be as effective as established methods in achieving the primary objective. Rupture trast enhanced computerised tomography (CT), duplex ultrasound scanning, angiography and magnetic resof the aneurysm within 1 year following endovascular repair was reported in six out of 895 patients entered onance imaging (MRI). The reported incidence of primary endoleaks ranges onto the Eurostar Registry and there have been other anecdotal reports of late ruptures. Three of the six from 5 to 44%. In some cases up to half of these primary endoleaks apparently sealed spontaneously. Eurostar patients died. A small risk of primary failure does not necessarily The occurrence of endoleaks may be dependent upon the type of device used and the experience of the outweigh the benefits of endovascular repair, paroperating team. But the frequency and completeness ticularly if failure can be anticipated and averted, in of follow-up and the method of examination used good time, by elective secondary intervention. But, could also influence the observed incidence. there is clearly a need to identify patients who are still The Eurostar protocol requires patients to be exat risk from rupture of their aneurysm and postamined by CT on discharge and at 6 weeks, 3 months, operative surveillance of all aortic endografts is im6 months and 1 year, and by angiography on discharge perative: but by which method? and at 1 year after operation. In a population of 895 Thus far, follow-up of patients after endovascular patients an endoleak was present on discharge in 14% aneurysm repair has been focussed, primarily, upon and another 18% developed a new endoleak during the identification of endoleak. This has been defined the first year. Over half (53%) of the patients in whom by White as: ‘‘a condition associated with endoluminal endoleaks were identified showed evidence of convascular grafts, defined by the persistence of blood tinued expansion of the aneurysm sac. By comparison, flow outside the lumen of the endoluminal graft but continued expansion of the sac occurred in only 11% within an aneurysm sac or adjacent vascular segment of those in whom no endoleak was detected. The being treated by the graft’’. Diagnosis of an endoleak association between the endoleaks identified by this defined in this way, is, therefore, dependent upon the programme and continued expansion of the aneurysm sac is clear confirmation that an endoleak signifies a ∗ Please address all correspondence to: P. L. Harris, Regional Vasrisk of treatment failure and consequently an incular Unit, 8c link, Royal Liverpool University Hospital, Prescot Street, Liverpool L7 8XP, U.K. dication for secondary intervention. However, two