Comparison of Unibody and Non-Unibody Endografts for Abdominal Aortic Aneurysm Repair in Medicare Beneficiaries: The SAFE-AAA Study

Abstract

BACKGROUND: Concerns have been raised about the long-term performance of aortic stent grafts for the treatment of abdominal aortic aneurysms, in particular, unibody stent grafts (eg, Endologix AFX AAA stent grafts). Only limited data sets are available to evaluate the long-term risks related to these devices. The SAFE-AAA Study was designed with the Food and Drug Administration to provide a longitudinal assessment of the safety of unibody aortic stent grafts among Medicare beneficiaries. METHODS: The SAFE-AAA Study was a prespecified, retrospective cohort study evaluating whether unibody aortic stent grafts are noninferior to non-unibody stent grafts with respect to the composite primary outcome of aortic reintervention, rupture, and mortality. Procedures were evaluated from August 1, 2011, through December 31, 2017. The primary end point was evaluated through December 31, 2019. Inverse probability weighting was used to account for imbalances in observed characteristics. Sensitivity analyses were used to evaluate the effect of unmeasured confounding, including the falsification end points of heart failure, stroke, and pneumonia. A prespecified subgroup includes patients treated from February 22, 2016, through December 31, 2017, corresponding to the market release of the most contemporary unibody endograft (Endologix AFX2 AAA stent graft). RESULTS: Of 87 163 patients who underwent aortic stent grafting at 2146 US hospitals, 11 903 (13.7%) received a unibody device. The average age was 77.0±6.7 years, 21.1% were female, 93.5% were White, 90.8% had hypertension, and 35.8% used tobacco. The primary end point occurred in 73.4% of unibody device-treated patients versus 65.0% of non-unibody device–treated patients (hazard ratio, 1.19 [95% CI, 1.15–1.22]; noninferior P value of 1.00; median follow-up, 3.4 years). Falsification end points were negligibly different between groups. In a contemporary subgroup of procedures, the cumulative incidence of the primary end point was 37.5% of unibody device–treated patients and 32.7% of non-unibody device–treated patients (hazard ratio, 1.06 [95% CI, 0.98–1.14]). CONCLUSIONS: In the SAFE-AAA Study, unibody endografts failed to meet noninferiority compared with non-unibody endografts with respect to aortic reintervention, rupture, and mortality. These data support the urgency of instituting a prospective longitudinal surveillance program for monitoring safety events related to aortic stent grafts.

The SAFE-AAA Study: Is One Limb at a Time Better Than Two?

“…The SAFE-AAA trial is a well-executed and well-designed observational study; its limitations derive from the study design itself. We agree with Secemsky and colleagues 5 when they suggest the establishment of a prospective surveillance system in order to monitor endograft performance and graft-related adverse events. Such a quality control measure should enable better overall results after EVAR, thereby lowering the expenses associated with subsequent interventions after EVAR and leading to the natural selection of those devices and design solutions that provide the highest level of patient safety over the long term.…”

“…In the current issue of Circulation , Secemsky and colleagues 5 describe findings from the SAFE-AAA study (Comparison of Unibody and Non-Unibody Endografts for Abdominal Aortic Aneurysm Repair in Medicare Beneficiaries Study) on the outcomes associated with unibody and nonunibody endografts in the treatment of intact infrarenal AAA. The main goal of the SAFE-AAA study was to test whether unibody endografts are noninferior compared with nonunibody devices in terms of a composite primary outcome of post–index procedure aneurysm rupture, aortic reintervention, or all-cause mortality.…”

“…Secemsky and colleagues 5 pooled all the unibody grafts in one cohort and all the nonunibody endografts into another one. The unibody group was used to create a subgroup from those patients who received unibody devices between May 9, 2016, and December 31, 2017, because this period coincided with the launch of the AFX2 device (February 2016).…”

“…Given that the AFX2 is the only currently available unibody graft on the market, the opportunity to compare this subgroup with nonunibody devices deployed during the same period is especially valuable. After weighting, Secemsky and colleagues 5 found that at 3.9 years, the cumulative incidence of the primary outcome was higher in the AFX2 subgroup compared with the nonunibody group (37.5% versus 32.7%). However, data over 3 years could not be interpreted with certainty, because the wide confidence intervals.…”

Unique Device Identifiers—Missing in Action

“…The investigation by Wang and colleagues 6 comes on the heels of 2 recent studies evaluating the safety of AFX stent grafts. 7,8 Despite using methodologies with different data sources, patient populations, and analytical approaches, all 3 groups found that the AFX stent grafts were associated with higher safety risks than were other EVAR devices. Nonetheless, each study illustrated key limitations of current mechanisms for postmarket surveillance using claims or registry data.…”