Background: In patients with coronary in-stent restenosis (ISR) requiring reintervention, it is unclear if the choice of treatment should depend on whether the restenotic stent was a bare-metal stent (BMS) or a drug-eluting stent (DES).

Objectives: This study aimed to assess the comparative efficacy and safety of the 2 most frequently used treatments - angioplasty with drug-coated balloon (DCB) and repeat stenting DES - in patients with BMS-and DES-ISR.

Methods: The DAEDALUS (Difference in Antirestenotic Effectiveness of Drug-Eluting Stent and Drug-Coated Balloon Angioplasty for the Occurrence of Coronary In-Stent Restenosis) study was a pooled analysis of individual patient data from all 10 existing randomized clinical trials comparing DCB angioplasty with repeat DES implantation for the treatment of coronary ISR. In this pre-specified analysis, patients were stratified according to BMS- versus DES-ISR and treatment assigned. The primary efficacy endpoint was target lesion revascularization (TLR) at 3 years. The primary safety endpoint was a composite of all-cause death, myocardial infarction, or target lesion thrombosis at 3 years. Primary analysis was performed by mixed-effects Cox models accounting for the trial of origin. Secondary analyses included nonparsimonious multivariable adjustment accounting also for multiple lesions per patient and 2-stage analyses.

Results: A total of 710 patients with BMS-ISR (722 lesions) and 1,248 with DES-ISR (1,377 lesions) were included. In patients with BMS-ISR, no significant difference between treatments was observed in terms of primary efficacy (9.2% vs. 10.2%; hazard ratio [HR]: 0.83; 95% confidence interval [CI]: 0.51 to 1.37) and safety endpoints (8.7% vs. 7.5%; HR: 1.13; 95% CI: 0.65 to 1.96); results of secondary analyses were consistent. In patients with DES-ISR, the risk of the primary efficacy endpoint was higher with DCB angioplasty than with repeat DES implantation (20.3% vs. 13.4%; HR: 1.58; 95% CI: 1.16 to 2.13), whereas the risk of the primary safety endpoint was numerically lower (9.5% vs. 13.3%; HR: 0.69; 95% CI: 0.47 to 1.00); results of secondary analyses were consistent. Regardless of the treatment used, the risk of TLR was lower in BMS- versus DES-ISR (9.7% vs. 17.0%; HR: 0.56; 95% CI: 0.42 to 0.74), whereas safety was not significantly different between ISR types.

Conclusions: At 3-year follow-up, DCB angioplasty and repeat stenting with DES are similarly effective and safe in the treatment of BMS-ISR, whereas DCB angioplasty is significantly less effective than repeat DES implantation in the treatment DES-ISR, and associated with a nonsignificant reduction in the primary composite safety endpoint. Overall, DES-ISR is associated with higher rates of treatment failure and similar safety compared with BMS-ISR.