Individual patient data meta-analysis of paclitaxel-coated balloons vs. drug-eluting stents for small-vessel coronary artery disease: the ANDROMEDA study

Abstract Background and Aims In randomized clinical trials of patients undergoing percutaneous coronary intervention (PCI) for de novo small-vessel coronary artery disease (SV-CAD), paclitaxel-coated balloon (PCB) angioplasty showed mid-term angiographic or clinical non-inferiority to drug-eluting stent (DES) implantation. Nevertheless, these trials have sample size limitations, and the relative safety and efficacy beyond the first year remain uncertain. Methods The ANDROMEDA study was a collaborative, investigator-initiated, individual patient data meta-analysis comparing 3 year clinical outcomes between PCB angioplasty and DES implantation for the treatment of de novo SV-CAD. Multiple electronic databases (PubMed, Scopus, ScienceDirect, and Web of Science) were searched from May 2010 to June 2024 to identify eligible trials. All the following eligibility criteria were required: (i) random allocations of treatments; (ii) patients with SV-CAD; (iii) treatment with PCB or DES; and (iv) clinical follow-up of at least 36 months. The primary and co-primary endpoints were major adverse cardiac events (MACE) and target lesion failure (TLF), respectively. The protocol was registered with PROSPERO (CRD42023479035). Results Individual patient data from three randomized trials, including a total of 1154 patients and 1360 lesions, were combined. At 3 years, PCB was associated with a lower risk of MACE compared with DES [hazard ratio (HR) 0.67, 95% confidence interval (CI) 0.47–0.96], due to a lower risk of myocardial infarction and target vessel revascularization. This benefit persisted after multivariable adjustment (HR 0.75, 95% CI 0.58–0.96), but did not reach statistical significance in the two-stage analysis (HR 0.67, 95% CI 0.43–1.04). At the landmark analysis, the risk of MACE between groups was consistent over time. At 3 years, TLF was not significantly different between PCB and DES groups. Reconstructed time-to-event information from a fourth trial was included in a sensitivity analysis (1384 patients and 1590 lesions), showing consistent results in terms of TLF (HR 0.87, 95% CI 0.63–1.20). The comparison between PCB and second-generation DES did not reveal significant differences in 3 year TLF (HR 1.03, 95% CI 0.70–1.50). Conclusions In patients undergoing PCI for de novo SV-CAD, PCB angioplasty is associated with a reduction in MACE and a non-significant difference in TLF at 3 year follow-up compared with DES implantation. The restriction of the comparator group to second-generation DES does not alter the main conclusions. Larger trials comparing contemporary devices at a more prolonged follow-up are warranted to confirm these findings.