Proper technical procedures improved outcomes in a retrospective analysis of EVAS FORWARD IDE trial 3-year results

Adverse outcomes observed late in the Nellix EndoVascular Aneurysm Sealing (EVAS) System (Endologix, Inc, Irvine, Calif) investigational device exemption trial prompted refinement of the anatomic instructions for use (IFU). This study aimed to investigate the association of procedural factors during Nellix endograft deployment and patient outcomes. We retrospectively reviewed 1-month imaging of 333 patients enrolled in the prospective, multicenter EVAS investigational device exemption trial between January 2014 and September 2016. Initial observations of those patients who met revised anatomic IFU yet still experienced late adverse events suggested that inadequate seal and low graft placement were common among these patients. Key procedural variables identified from a univariate analysis were applied to construct four cohorts stratified by procedural technical performance (technically adequate [P+] or technically inadequate [P−]) and the revised anatomic indications for use (anatomically within IFU [A+] or anatomically outside of IFU [A−]) and to compare them for aneurysm-related outcomes. A logistic regression analysis was performed to identify significant predictors of sac expansion or migration. Proximal and distal seal zones and low graft placement were identified by logistic regression analysis as significant predictors of sac expansion or migration. Accordingly, acquisition of ≥10-mm proximal and distal seal zones and the position of the lowest stent within 10 mm of the lowest renal artery were clinically justified as thresholds for a technically adequate procedure. Patients who did not achieve these parameters were deemed to have a technically inadequate procedure. By use of the proposed procedural adequacy criteria and established anatomic criteria, patients were stratified into four cohorts: A+/P+ (n = 77), A+/P− (n = 54), A−/P+ (n = 71), and A−/P− (n = 131). Three-year estimates of freedom from migration of 10 mm were 98.6% in A+/P+, 95.9% in A+/P−, 85.8% in A−/P+, and 80.1% in A−/P−; freedom from type IA endoleak estimates were 98.6% in A+/P+, 100% in A+/P−, 96.4% in A−/ P+, and 90.3% in A−/P− cohorts. Freedom from sac expansion and secondary intervention were 96.9% and 90.6% in A+/P+, 86.0% and 93.6% in A+/P−, 87.1% and 83.1% in A−/P+, and 80.5% and 79.8% in A−/P− groups, respectively. Two early deaths (aneurysm-related mortality) on days 4 and 12 postoperatively were reported within the A+/P− group. Kaplan-Meier estimates of freedom from all-c