Background
Bioresorbable scaffold (BRS) technology was developed to overcome the long-term limitations of metallic drug-eluting stents (DES) by providing temporary vessel support and subsequently restoring vasomotion after complete resorption. However, first-generation devices such as the everolimus-eluting BVS were associated with higher rates of scaffold thrombosis and target lesion revascularization (TLR) in both randomized trials and real-world use. Despite improved implantation techniques and more selective lesion targeting, outcomes remained inferior to contemporary DES. The German–Austrian ABSORB Registry (GABI-R; NCT02066623) is the largest multicenter observational study evaluating real-world performance of this first-generation BRS, allowing long-term assessment beyond the expected period of bioresorption. This analysis presents the final 5-year outcomes.
Methods
GABI-R is a prospective, multicenter, observational post-market registry conducted across approximately 100 centers in Germany and Austria. Consecutive patients with significant coronary artery disease treated with the BVS were enrolled and managed according to standard clinical practice without protocol-mandated interventions. Follow-up was scheduled at 30 days, 6 months, 2 years, and 5 years. Data were collected independently and analyzed on an intention-to-treat basis. Clinical outcomes were assessed using Kaplan–Meier estimates, and time-dependent predictors were evaluated with Cox regression analysis.
Results
A total of 3325 patients were enrolled, with successful scaffold implantation in 3266 (98.2%). At 5 years, vital status was known for 94.4% of patients, and complete clinical follow-up was achieved in 91.2%. The mean follow-up duration was 1872.9 ± 47.8 days. Most patients were asymptomatic or mildly symptomatic at 5 years (CCS I: 69.8%; NYHA I: 58.4%).
At 5 years, all-cause mortality was 7.0%, definite scaffold thrombosis occurred in 3.0%, and target lesion failure (TLF) reached 14.7%, primarily driven by target lesion revascularization. The TLF rate showed a polyphasic distribution with an early steep rise within the first month after PCI (2.3%), a gradual increase up to year 3, and another decrease of rates thereafter.
Discussion
In this large, real-world cohort of over 3300 patients treated with the first-generation BVS, procedural safety was favorable, and short-term outcomes were satisfactory. However, long-term results demonstrated a continued but diminishing risk of scaffold-related events up to three years, stabilizing thereafter. The five-year TLF (14.7%) and definite thrombosis (3.0%) rates were consistent with randomized data but remained higher than those typically observed with current-generation DES.
The temporal event pattern—with an early hazard phase followed by stabilization—suggests that most adverse events occur before complete polymer resorption, after which event rates approximate those of metallic stents. These findings emphasize the importance of meticulous implantation technique, appropriate lesion selection, and long-term secondary prevention to optimize outcomes.
Despite inherent limitations of an observational design and absence of a DES control arm, the GABI-R registry provides the most comprehensive real-world long-term data for first-generation bioresorbable scaffolds. It confirms that, while early device limitations led to initial setbacks, acceptable outcomes can be achieved when implantation protocols are optimized. These insights can promote the development and evaluation of next-generation, thinner-strut bioresorbable devices aimed at minimizing early procedural risks and achieving true long-term vascular restoration.
