Prevalence and predictors of nonobstructive coronary artery disease identified with coronary angiography in contemporary clinical practice

Abstract

Guidelines recommend noninvasive tests (NITs) to risk stratify and identify patients with higher likelihood of coronary artery disease (CAD) prior to elective coronary angiography. However, a high percentage of patients are found to have nonobstructive CAD. We aimed to understand the relationship between patient characteristics, NIT findings, and the likelihood of nonobstructive CAD. Patients undergoing elective catheterization without history of CAD were identified from 1,128 hospitals in National Cardiovascular Data Registry's CathPCI Registry between July 2009 and December 2011. Noninvasive tests included stress electrocardiogram, stress echocardiogram, stress radionuclide, stress cardiac magnetic resonance, and computed tomographic angiography. Patient demographics, risk factors, symptoms, and NIT results were correlated with the presence of nonobstructive CAD, defined as all native coronary stenoses <50%. Of 661,063 patients undergoing elective angiography, 386,003 (58.4%) had nonobstructive CAD. Preprocedure NIT was performed in 64% of patients; 51.9% were reported to be abnormal, but only 9% had high-risk findings. Independent factors associated with nonobstructive CAD were younger age, female sex, atypical chest pain, and a low-risk NIT. Patients with high-risk findings on NIT were more likely to have obstructive CAD (adjusted odds ratio 3.03 [2.86-3.22]). Noninvasive test findings had minimal incremental value beyond clinical factors for predicting obstructive disease (C index = 0.75 for clinical factors vs 0.74 for NIT findings). In current practice, about two-thirds of patients undergo NIT prior to elective cardiac catheterization, yet most patients have nonobstructive CAD. The weak correlation between most NIT results and the likelihood of obstructive CAD provides further impetus for improving preangiography assessment of likelihood of disease.