Comparison of self- and balloon-expandable bridging stent-grafts in branched endovascular aortic repair (COVIBRI study)

Background: Endovascular treatment of thoracoabdominal aortic aneurysms has become common, with satisfactory results. Nevertheless, long-term durability remains an issue mainly because of target visceral vessel (TVV) instability. Currently, no covered stent has been approved as a bridging stent graft (BSG), demanding continuous research on this topic. Methods: This was a multicenter observational retrospective cohort study comparing the midterm results of the Bard Covera Plus and Gore VBX as BSGs during branched endovascular aneurysm repair. The primary outcome was the comparison of the target vessel instability between the two groups. Primary patency, freedom from branch-related type I and III endoleaks and reintervention, and technical and clinical success were considered secondary outcomes. Logistic regression analysis was used to assess the association between selected baseline factors and TVV instability. TVV instability during follow-up was then evaluated using the Kaplan-Meier cumulative function. Results: Three hundred forty-five TVVs in 106 patients were considered suitable for the analysis. Two hundred twenty vessels were stented with the Covera stent graft (64%) and 125 with VBX (36%). Two hundred ninety-nine TVVs received a single BSG, 45 two BSGs, and only 1 three BSGs. Bare metal stent relining was required in 36% of TVVs, mostly in the Covera group (89 [41%] vs 36 [29%]) (P = .030). The primary technical success rate was 96% (331/345), and the assisted primary technical success rate was 99% (342/345). The TVV instability rate within 30 days was 2% (one Covera and five VBX; P = .015). Three BSG occlusions (one Covera and two VBX) and three type Ic endoleaks (three VBX) were detected. The median follow-up was 13.9 months (range, 5.8-25.5 months). Sixteen TVV instabilities were detected during the follow-up. Twelve BSG occlusions (six Covera and six VBX), three type Ic endoleaks (one Covera and two VBX), and one type IIIc endoleak (VBX). The overall target vessel instability rate was 5% (16/342). TVV instability was associated with the use of Gore VBX in the univariable logistic regression (odds ratio, 3.0; 95% confidence interval, 1.1-8.0; P = .027). Aneurysm rupture and aneurysm diameter were also associated with TVV instability in the univariable analysis (P = .002 and P = .008, respectively). The only factor predisposing to TVV instability in the multivariable logistic regression analysis was the use of Gore VBX as a BSG (odds ratio, 2.9; 95% confidence interval, 1.0-8.0; P = .043). Kaplan-Meier analysis showed a significantly higher risk of TVV instability in the VBX group (P < .001). Conclusions: Overall midterm outcomes in this cohort were satisfactory. Patency rates were similar between the two stents. Nevertheless, VBX seems to be associated with worse TVV instability. These results may be correlated with a higher incidence of type Ic endoleaks, which require an extensive learning curve for correct stent selection and deployment.