AngiogRaphic pattErns and management straTegies for coRonary restEnosis following drug-coATed balloon angioplasty in de novo lesions

Abstract Background With the expanding use of DCBs for de novo CAD, understanding the optimal assessment and management of restenosis in DCB-treated segments has become increasingly important. The RETREAT study aims to characterize the angiographic patterns of restenosis following DCB angioplasty in de novo lesions and to evaluate its management strategies. Methods RETREAT is an observational, retrospective, international, multicenter, investigator-initiated study including consecutive patients undergoing PCI for restenosis after prior DCB treatment of de novo coronary lesions across 15 European centers. Patients with in-stent restenosis or those managed conservatively were excluded. Restenosis was classified based on its angiographic pattern as focal (≤10 mm), diffuse (>10 mm), or occlusive. The primary endpoint was the 1-year cumulative incidence of target lesion revascularization (TLR) after reintervention. Key secondary endpoints included target vessel revascularization (TVR) and target-vessel myocardial infarction (TV-MI). Results A total of 262 patients with DCB restenosis treated between 2015 and 2023 were included. Median lesion length and reference vessel diameter were 20 (15-30) mm and 3.0 (2.5-3.5) mm respectively. Paclitaxel-coated DCBs were used in 70% of index procedures, while 30% of patients received sirolimus-coated DCBs. Restenosis occurred at a median time of 209 (IQR 106–476) days from the index PCI. Most patients were treated with DES PCI (N=134; 51%), followed by repeat DCB-based PCI (N=112; 43%) and plain old balloon angioplasty (N=16; 6%). Among patients with available site-reported angiographic assessment (N=174), most presented with a focal restenosis (54%), followed by diffuse (30.5%) and occlusive (15.5%) patterns. Clinical follow-up was recorded at a median time of 688 (IQR 334–1003) days after the restenosis. The 1-year cumulative incidences of TLR, TVR, and TV-MI were 5.4%, 6.8%, and 2.6%, respectively. These outcomes were numerically higher in patients treated with DES compared to repeat DCB, though not statistically significant (TLR: 6.8% vs. 3.7%, p=0.369; TVR: 7.5% vs. 5.9%, p=0.699; TV-MI: 3.7% vs. 1.2%, p=0.263). Compared to those with focal restenosis, patients with a diffuse or occlusive patterns displayed comparable risks of TLR (3.4% vs. 5.5%, p=0.440), TVR (4.7% vs. 5.5%, p=0.761) and TV-MI (1.7% vs. 1.4%, p=0.970). The final analysis, to be presented at the time of the meeting, is expected to include an independent assessment of angiographic patterns of restenosis by a core laboratory. Conclusions Restenosis following DCB angioplasty in de novo coronary lesions prevalently presents with focal patterns, similar to what has historically been observed after DES implantation. Both repeat DCB angioplasty and DES implantation are commonly used to treat post-DCB restenosis, with relatively low one-year rates of lesion-related adverse events, that are comparable between the two treatment strategies.