The incidence and natural history of type I and II endoleak: a 5-year follow-up assessment with color duplex ultrasound scan.

PURPOSE: The purpose of this study was the demonstration of the value of color duplex ultrasound (CDU) scanning in the detection of type I endoleak (T1EL) and type II endoleak (T2EL), the correlation of Doppler scan waveform pattern to endoleak persistence or seal, and the description of the natural history of endoleak.
METHODS: The study was a retrospective review of 83 patients who underwent periodic CDU scan and computed tomographic (CT) scan surveillance of the endograft and aneurysm sac after insertion of an aortic endograft for abdominal aortic aneurysm (AAA). Forty-one patients (49%) with an endoleak at anytime in the follow-up period form the basis of this report.
RESULTS: T1EL was detected in all five patients with CDU and CT scans. T2EL was detected in 36 patients with CDU scan as compared with 18 patients with CT scan. With CT scan, endoleak was not detected when CDU scan showed no endoleak. Conversely, all CT scan--detected endoleaks were found with CDU scanning. The T2EL source artery was identified with CT scan in seven patients, whereas the source was identified in all 36 patients with CDU scan. Endoleak source did not correlate with outcome (seal or persistence). However, a to/fro Doppler scan waveform pattern was associated with spontaneous T2EL seal in seven of 12 patients, and a monophasic or biphasic waveform was associated with endoleak persistence in 14 of 17 patients (P =.023, with chi(2) test). Thirteen of 36 T2ELs underwent spontaneous seal by 6.2 +/- 2.8 months. T2ELs without increasing AAA diameter were observed. Eight patients with persistent T2EL present for more than 12 months did not undergo treatment. However, two patients underwent T2EL obliteration with coils because of AAA sac enlargement. T1EL of the distal attachment site was the initial endoleak identified in five patients, but seven patients harboring T2ELs had subsequent T1ELs develop. For the entire 83 patients, the combined T1EL and T2EL prevalence rate was 20% of patients at a 6-month follow-up period, but this rate increased to 50% after 24 months. The incidence rate of newly detected endoleaks and of spontaneous sealing was 24.4% at 12 months and 12.5% in longer-term follow-up period.
CONCLUSION: CDU scan is effective in the identification of the type of endoleak, the delineation of the vessel involved, and the hemodynamic information not available with any other testing method. Endoleaks have a dynamic natural history characterized by a variable onset with changing branch vessel involvement and spectral flow patterns. Periodic long-term endograft surveillance with CDU scanning is necessary for following existing endoleaks and for detecting new ones. Corroboration of these findings in larger multicenter prospective trials will be needed to determine whether CDU scan analysis of endoleaks would be predictive of long-term success in endovascular AAA repair.