Development and validation of a prediction model for angiographic perforation during chronic total occlusion percutaneous coronary intervention: OPEN-CLEAN perforation score.

BACKGROUND Perforation is the most frequent complication of chronic total occlusion (CTO) percutaneous coronary intervention (PCI) and is associated with adverse events including mortality. METHODS Among 1,000 consecutive patients enrolled in 12 center prospective CTO PCI study (OPEN CTO), all perforations were reviewed by the angiographic core-lab. Eighty-nine patients (8.9%) with angiographic perforation were compared to 911 patients without perforation. We sought to describe the clinical and angiographic predictors of angiographic perforation during CTO PCI and develop a risk prediction model. RESULTS Among eight clinically important candidate variables, independent risk factors for perforation included prior CABG (OR 2.0 [95% CI, 1.2-3.3], p < .01), occlusion length (OR 1.2 per 10 mm increase [95% CI, 1.1-1.3], p < .01), ejection fraction (OR 1.2 per 10% decrease [95% CI, 1.1-1.5], p < .01), age (OR 1.3 per 5 year increase [95%CI, 1.1-1.5], p < .01), and heavy calcification (OR 1.7 [95% CI, 1.0-2.7], p = .04). Three other potential candidate variables, glomerular filtration rate, proximal cap ambiguity, and target vessel, were not independently associated with perforation. The model was internally validated using bootstrapping methods. From the full model, a simplified perforation prediction score (OPEN-CLEAN score: CABG, Length [occlusion], EF < 50%, Age, CalcificatioN) was developed, which discriminated the risk of angiographic perforation well (c-statistics = 0.75) and demonstrated good calibration. CONCLUSION This simple 5-variable prediction score may help CTO operators to risk-stratify patients for angiographic perforation using variables available prior to CTO PCI procedures.