OBJECTIVE: the outcomes for patients after endovascular treatment of abdominal aortic aneurysm (AAA) are determined primarily by the endpoints of death and endoleaks, the latter representing continued risk of rupture. The data of a multicentre registry were analysed with regard to the early outcome of stent-graft procedures for AAA and the complications associated with this treatment. In addition, the results during follow-up were analysed by determining mortality and endoleak development as separate endpoints and as a combined endpoint defined as endoleak-free survival. SETTING: 38 European institutions of Vascular Surgery collaborating in a multicentre registry project. PATIENTS AND METHODS: 899 patients with AAA underwent between May 1994 and March 1998 elective endovascular repair (818 men and 81 women; mean age 69 years). 80 (8.9%) of the patients had medical conditions that excluded them from open repair. 818 (91%) of patients had a bifurcated device, 63 (7%) had a straight tube graft, and only 18 (2%) had an aorto-uni-iliac device. Clinical examination and contrast-enhanced computed tomography was performed at fixed follow-up intervals to assess increase or decrease of the maximum transverse diameter (MTD). Endoleaks observed at follow-up were discriminated into persistent endoleak and temporary endoleak. The latter is defined as single time observed endoleaks or with two or more negative imaging studies between observed endoleaks. Life-table analyses were used to calculate the rates of freedom-from-endoleak (no endoleak at any time), freedom-from-persistent endoleak (no persistent endoleak), patient survival, and persistent-endoleak-free-survival. RESULTS: the median follow-up of this patient series was 6.2 months. The ratio between observed and expected follow-up data was 82% for the overall follow-up period. However, at 18 months of follow-up this rate was only 45%. The number of patients followed during this period was sufficient to allow statistically meaningful assessment. The MTD in patients with temporary endoleaks demonstrated a significant decrease at 6 to 12 months compared to preoperative values (mean 57 and 53 respectively, p =0.004). In patients with persistent endoleaks there was no change between the preoperative and 6-month MTD (mean 57 and 60 mm respectively). At 6 and 18 months freedom-from-endoleak was 83% and 74% and freedom-from-persistent endoleak was 93% and 90%, respectively. The 18-month cumulative patient survival was 88% and the main outcome measure, the persistent endoleak-free-survival was 79%. CONCLUSIONS: the MTD decreases in patients with temporary endoleak, but not in patients with persistent endoleak. Therefore, the use of the rate of freedom-from-persistent endoleak, reflecting absence of persisting endoleaks to estimate the prognosis with regard to the AAA, is justified. Determining persistent endoleak-free survival appears a rational approach to provide a realistic outlook for patients with stent-grafted AAA. The observed 18-month endoleak-free survival reflects a satisfactory mid-term result. Copyright 1999 W.B. Saunders Company Ltd.
OBJECTIVE: the outcomes for patients after endovascular treatment of abdominal aortic aneurysm (AAA) are determined primarily by the endpoints of death and endoleaks, the latter representing continued risk of rupture. The data of a multicentre registry were analysed with regard to the early outcome of stent-graft procedures for AAA and the complications associated with this treatment. In addition, the results during follow-up were analysed by determining mortality and endoleak development as separate endpoints and as a combined endpoint defined as endoleak-free survival. SETTING: 38 European institutions of Vascular Surgery collaborating in a multicentre registry project. PATIENTS AND METHODS: 899 patients with AAA underwent between May 1994 and March 1998 elective endovascular repair (818 men and 81 women; mean age 69 years). 80 (8.9%) of the patients had medical conditions that excluded them from open repair. 818 (91%) of patients had a bifurcated device, 63 (7%) had a straight tube graft, and only 18 (2%) had an aorto-uni-iliac device. Clinical examination and contrast-enhanced computed tomography was performed at fixed follow-up intervals to assess increase or decrease of the maximum transverse diameter (MTD). Endoleaks observed at follow-up were discriminated into persistent endoleak and temporary endoleak. The latter is defined as single time observed endoleaks or with two or more negative imaging studies between observed endoleaks. Life-table analyses were used to calculate the rates of freedom-from-endoleak (no endoleak at any time), freedom-from-persistent endoleak (no persistent endoleak), patient survival, and persistent-endoleak-free-survival. RESULTS: the median follow-up of this patient series was 6.2 months. The ratio between observed and expected follow-up data was 82% for the overall follow-up period. However, at 18 months of follow-up this rate was only 45%. The number of patients followed during this period was sufficient to allow statistically meaningful assessment. The MTD in patients with temporary endoleaks demonstrated a significant decrease at 6 to 12 months compared to preoperative values (mean 57 and 53 respectively, p =0.004). In patients with persistent endoleaks there was no change between the preoperative and 6-month MTD (mean 57 and 60 mm respectively). At 6 and 18 months freedom-from-endoleak was 83% and 74% and freedom-from-persistent endoleak was 93% and 90%, respectively. The 18-month cumulative patient survival was 88% and the main outcome measure, the persistent endoleak-free-survival was 79%. CONCLUSIONS: the MTD decreases in patients with temporary endoleak, but not in patients with persistent endoleak. Therefore, the use of the rate of freedom-from-persistent endoleak, reflecting absence of persisting endoleaks to estimate the prognosis with regard to the AAA, is justified. Determining persistent endoleak-free survival appears a rational approach to provide a realistic outlook for patients with stent-grafted AAA. The observed 18-month endoleak-free survival reflects a satisfactory mid-term result. Copyright 1999 W.B. Saunders Company Ltd.