Optical coherence tomography predictors of clinical outcomes after stent implantation: the ILUMIEN IV trial

Background and Aims Observational registries have suggested that optical coherence tomography (OCT) imaging-derived parameters may predict adverse events after drug-eluting stent (DES) implantation. The present analysis sought to determine the OCT predictors of clinical outcomes from the large-scale ILUMIEN IV trial. Methods ILUMIEN IV was a prospective, single-blind trial of 2487 patients with diabetes or high-risk lesions randomized to OCT-guided versus angiography-guided DES implantation. All patients underwent final OCT imaging (blinded in the angiography-guided arm). From more than 20 candidates, the independent OCT predictors of 2-year target lesion failure (TLF; the primary endpoint), cardiac death or target-vessel myocardial infarction (TV-MI), ischaemia-driven target lesion revascularization (ID-TLR), and stent thrombosis were analysed by multivariable Cox proportional hazard regression in single treated lesions. Results A total of 2128 patients had a single treated lesion with core laboratory-analysed final OCT. The 2-year Kaplan-Meier rates of TLF, cardiac death or TV-MI, ID-TLR, and stent thrombosis were 6.3% (n = 130), 3.3% (n = 68), 4.3% (n = 87), and 0.9% (n = 18), respectively. The independent predictors of 2-year TLF were a smaller minimal stent area (per 1 mm2 increase: hazard ratio 0.76, 95% confidence interval 0.68-0.89, P < .0001) and proximal edge dissection (hazard ratio 1.77, 95% confidence interval 1.20-2.62, P = .004). The independent predictors of cardiac death or TV-MI were smaller minimal stent area and longer stent length; of ID-TLR were smaller intra-stent flow area and proximal edge dissection; and of stent thrombosis was smaller minimal stent expansion. Conclusions In the ILUMIEN IV trial, the most important OCT-derived post-DES predictors of both safety and effectiveness outcomes were parameters related to stent area, expansion and flow, proximal edge dissection, and stent length.