Learning Curve and Long-Term Outcomes of Thoracic Endovascular Repair With the Relay Stent-Graft

Purpose: To define the learning curve of a widely employed stent-graft for thoracic endovascular repair (TEVAR) by analyzing procedural variables and their impact on long-term outcomes. Materials and Methods: Standard TEVARs for each major aortic thoracic disease were included excluding procedures using thoracoabdominal, arch fenestrated/branched devices and/or chimneys. The primary outcome was the learning curve analysis using the cumulative sum chart method. The secondary outcomes, presented as early (Q1–Q2) versus latest (Q3–Q4) quartiles of experience, were 30-day major adverse events (MAEs); procedural details (additional maneuvers, operative and fluoroscopy time, and contrast volume); 30-day clinical success; endoleak; aorta-related reintervention; and overall and aorta-related survival. Results: Between November 2005 and September 2021, 220 consecutive TEVAR procedures involving the Relay endograft (Terumo Aortic, Sunrise, FL, USA) were performed and included in the present analysis. The mean follow-up was 4.5 ± 3.9 years. The learning curve was reached after 10 patients. Secondary outcomes improved significantly over experience. Thirty-day MAE occurrence was 14.5% versus 11.8% (p=0.550). Additional maneuvers (p=0.009), access-vessel additional maneuvers (p=0.010), operative time (p=0.004), Relay Plus (p=0.001), and implantation of multiple stent-grafts (p=0.034) were independent risk markers for MAEs. Operative time (125.9 ± 66.7 vs 86.7 ± 48.5 minutes, p<0.001), contrast volume (185.4 ± 112.8 vs 140.5 ± 88.2 mL, p=0.003), and fluoroscopy time (12.4 ± 12.7 vs 8.8 ± 7.5 minutes, p=0.017) decreased significantly. Late endoleak occurrence was 19.5% with a nonsignificant reduction (21.8% vs 17.3%, p=0.395). Fluoroscopy time (hazard ratio [HR]=1.0; 95% confidence interval [CI]=1.0–1.1; p=0.008), contrast volume (HR=1.0; 95% CI=1.0–1.1; p=0.018), and type III aortic arch (HR=3.3; 95% CI=1.7–6.4; p<0.001) were independent risk markers for endoleak. Fluoroscopy time (HR=1.0; 95% CI=1.0–1.1; p=0.032) and type III aortic arch (HR=3.6; 95% CI=1.7–7.4; p=0.001) confirmed their significant association in a multivariable analysis. Conclusion: In a high-volume center with a consistent previous endovascular experience, the Relay graft presented satisfactory long-term results with a short learning curve supporting its reliability. Clinical Impact: The manuscript addresses the understanding of how many TEVAR procedures with a currently implanted device have been required to reach the learning curve for an endovascular skilled center. Our TEVAR experience with the Relay stent-graft (Terumo Aortic) demonstrated that ten implantations were required to achieve the device-related learning curve. The fifteen-year analyzed period showed that intraoperative learning-related variables were associated with long-term clinical outcomes and both improved over time. The Relay stent-graft presented satisfactory long-term results along with a short learning curve in a high-volume endovascular center supporting its ongoing implantation.