Predictors of Adverse Outcome in High-Risk Percutaneous Coronary Interventions Patients

Background: Identifying high-risk percutaneous coronary intervention (PCI) patients is challenging. We aimed to evaluate which high-risk patients are prone to adverse events. Methods: We performed a retrospective study including consecutive high-risk PCIs from 2005 to 2018 in a large tertiary medical centre. Patients with unprotected left main (LM) disease, last patent coronary vessel, or 3-vessel coronary artery disease with left ventricular ejection fraction < 35% were included. A predictive 30-day major adverse cardiac events (MACE) score consisting of any myocardial infarction, all-cause death, or target-vessel revascularisation was constructed. Results: From 2005 to 2018, a total of 1890 patients who underwent PCI met the predefined high-risk PCI criteria. Mortality rate was 8.8% at 30 days and 20.7% at 1 year, and 30-day MACE rate was 14.2% and 33.5% at 1 year. Predictors of short-term MACE were New York Heart Association functional class (NYHA) 4 (hazard ratio [HR] 6.65; P < 0.001), systolic blood pressure (SBP) < 90 mm Hg (HR 4.93; P < 0.001), creatinine > 1.3 mg/dL (HR 3.57; P < 0.001), hemoglobin < 11.0 g/dL (HR 3.07; P < 0.001), pulmonary artery systolic pressure > 50 mm Hg (HR 2.06; P < 0.001), atrial fibrillation (HR 1.74; P < 0.001), and LM disease (HR 2.04; P < 0.001) or last patent vessel (HR 1.70; P = 0.002). A score constructed from these parameters reached a sensitivity of 90% and a specificity of 81% with areas under the receiver operating characteristic curve of 0.92 for MACE and 0.94 with 89% sensitivity and 87% specificity for all-cause mortality. Conclusions: Specific features such as LM lesion or last patent conduit, pulmonary hypertension, atrial fibrillation, anemia, and renal failure, along with low SBP and NYHA 4, aid risk stratification and consideration of further treatment measures.