Choroidal thickness and granulocyte colony-stimulating factor in tears improve the prediction model for coronary artery disease

Abstract

Abstract Background Coronary artery disease (CAD) detection in asymptomatic patients still remains controversial. The aim of our study was to evaluate the usefulness of ophthalmologic findings as predictors of the presence of CAD when added to cardiovascular classic risk factors (CRF) in patients with acute coronary cardiopathy suspicion. Methods After clinical stabilization, 96 patients with acute coronary cardiopathy suspicion were selected and divided in two groups: 69 patients with coronary lesions and 27 patients without coronary lesions. Their 192 eyes were subjected to a complete routine ophthalmologic examination. Samples of tear fluid were also collected to be used in the detection of cytokines and inflammatory mediators. Logistic regression models, receiver operating characteristic curves and their area under the curve (AUC) were analysed. Results Suggestive predictors were choroidal thickness (CT) (OR: 1.02, 95% CI 1.01–1.03) and tear granulocyte colony-stimulating factor (G-CSF) (OR: 0.97, 95% CI 0.95–0.99). We obtained an AUC of 0.9646 (95% CI 0.928–0.999) when CT and tear G-CSF were added as independent variables to the logistic regression model with cardiovascular CRF: sex, age, diabetes, high blood pressure, hypercholesterolemia, smoking habit and obesity. This AUC was significantly higher (p = 0.003) than the prediction derived from the same logistic regression model without CT and tear G-CSF (AUC = 0.828, 95% CI 0.729–0.927). Conclusions CT and tear G-CSF improved the predictive model for CAD when added to cardiovascular CRF in our sample of symptomatic patients. Subsequent studies are needed for validation of these findings in asymptomatic patients.

Keywords

• Coronary artery disease
• Choroidal thickness
• Granulocyte colony-stimulating factor
• Cardiovascular prevention
• ROC curves
• Predictive model