Coronary disease & ACSRandomised controlled trial

Sirolimus-Eluting Balloon vs Repeat Drug-Eluting Stent or Balloon Angioplasty for Coronary In-Stent Restenosis

J Am Coll CardiolPublished

  • Randomised controlled trial
  • non-inferiority
  • Bayesian
  • July 2020 to July 2024

Authors' conclusion

The sirolimus DEB was noninferior to a usual care control strategy including 80% repeat DES but not noninferior to DES for single-layer ISR for TLF at 12 months. There was significant interaction on the basis of operator selection of DES vs BA.

Study context

Population
Patients with coronary in-stent restenosis after successful lesion predilation390 per protocol
Participants
418 patients
Comparison
Sirolimus drug-eluting balloon (DEB) vs usual care
  • DEBn = 210
  • Controln = 208reference
Primary endpoint
TLF (cardiac death, target vessel MI, or clinically driven target lesion revascularization), per protocol
Follow-up
1 year

Primary result

TLF

Difference

1.7%

95% CrI −5.5% to 8.9%

Events by treatment group

DEB vs Control
difference 1.7% (95% CrI −5.5% to 8.9%)

Non-inferiority margin 10 as printed, not drawn

As reported in the abstract

Further findings

OutcomeDEBControldifference (95% CrI)
TLF, DEB vs DES in single-layer ISR (sequential secondary hypothesis test) · CrI; posterior probability for noninferiority 76.07%14.2%Not reported in the abstract7.7% (0.6%–14.6%)
TLF by operator-selected control (subgroup) — DEB vs DES · CrI; 15.3% vs 7.1%8.1% (1.4%–15.0%)
Outcomedifference (95% CrI)
TLF by operator-selected control (subgroup) — DEB vs BA · CrI; 23.6% vs 43.6%23.7% (−41.4% to −1.5%)
  • Target lesion failure (TLF), DEB vs DES in single-layer in-stent restenosis (ISR) (sequential secondary hypothesis test): DES 6.5%
  • TLF by operator-selected control (subgroup): interaction P = 0.0026

Original publication

Sirolimus-Eluting Balloon vs Repeat Drug-Eluting Stent or Balloon Angioplasty for Coronary In-Stent Restenosis. J Am Coll Cardiol. Published 15 April 2026. doi:10.1016/j.jacc.2026.03.021

Summarised from the published abstract; every number is checked against it