Angio-TDM vs écho-Doppler dans la surveillance des endoprothèses aortiques : l'étude ESSEA

BACKGROUND: Color-duplex ultrasonography (DUS) could be an alternative to computed tomography-aortography (CTA) in the lifelong surveillance of patients after endovascular aneurysm repair (EVAR), but there is currently no level 1 evidence. The aim of this study was to assess the diagnostic accuracy of DUS as an alternative to CTA for the follow-up of post-EVAR patients. METHODS: Between December 16, 2010, and June 12, 2015, we conducted a prospective, blinded, diagnostic-accuracy study, in 15 French university hospitals where EVAR was commonly performed. Participants were followed up using both DUS and CTA in a mutually blinded setup until the end of the study or until any major aneurysm-related morphological abnormality requiring reintervention or an amendment to the follow-up policy was revealed by CTA. Database was locked on October 2, 2017. Our main outcome measures were sensitivity, specificity, positive predictive value, negative predictive value, positive and negative likelihood ratios of DUS against reference standard CTA. CIs are binomial 95% CI. RESULTS: This study recruited prospectively 659 post-EVAR patients of whom 539 (82%) were eligible for further analysis. Following the baseline inclusion visit, 940 additional follow-up visits were performed in the 539 patients. Major aneurysm-related morphological abnormalities were revealed by CTA in 103 patients (17.2/100 person-years [95% CI, 13.9–20.5]). DUS accurately identified 40 patients where a major aneurysm-related morphological abnormality was present (sensitivity, 39% [95% CI, 29–48]) and 403 of 436 patients with negative CTA (specificity, 92% [95% CI, 90–95]). The negative predictive value and positive predictive value of DUS were 92% (95% CI, 90–95) and 39% (95% CI, 27–50), respectively. The positive likelihood ratio was 4.87 (95% CI, 2.9–9.6). DUS sensitivity reached 73% (95% CI, 51–96) in patients requiring an effective reintervention. CONCLUSIONS: DUS had an overall low sensitivity in the follow-up of patients after EVAR, but its performance improved meaningfully when the subset of patients requiring effective reinterventions was considered.