False positive results of SPECT - we should minimize the unnecessary referrals to invasive coronary angiography

Purpose: False positive (FP) results are common in single-photon emission computed tomography (SPECT) performed for obstructive coronary artery disease (OCAD) investigation. It is important to discriminate whether an ischemic result of SPECT is a FP, to avoid unnecessary referrals for invasive coronary angiography (ICA). We determined the clinical factors that were independently associated with a FP SPECT result. Methods: We analyzed a cohort of patients referred for ICA for stable coronary artery disease diagnosis, based on clinical judgment and a SPECT positive for ischemia, in a single tertiary-care center (2006-2011). Traditional and nontraditional coronary artery disease risk factors, modified Framingham risk score (FRS), symptoms, pre-test (SPECT) probability of OCAD, left ventricular ejection fraction and ICA results were assessed. OCAD was defined as any luminal narrowing ≥70%, or ≥50% for the left main artery. A FP SPECT was defined as reversible stress-induced perfusion defect and no OCAD. The predictors of a FP result were determined by the chi-square, exact Fisher and t-student tests when appropriate, and multivariate analysis (logistic regression). The discriminatory power for a FP result, of a model based on those predictive factors, was assessed by the area under the ROC curve (AUC) analysis. Results: 960 patients were included: 66.6±9.9 years, 58.1% male, mean 10-year Framingham risk 19.0%; 45.9% typical angina and 17.5% atypical angina; 55% high pre-test probability of OCAD; 13.6% depressed left ventricular ejection fraction (<55%). Globally, 51.0% of SPECT were FP results. The factors independently associated with a FP SPECT were: absence of severe angina (OR 6.1, 95% CI 2.1-17.4), absence of typical angina (OR 4.2, 95% CI 2.9-5.9), presence of atypical angina (OR 2.8, 95% CI 1.6-5.0), normal left ventricular ejection fraction (≥55%) (OR 2.5, 95% CI 1.7-3.5), female gender (OR 2.4, 95% CI 1.8-3.4) and lower FRS (OR 1.2, 95% CI 1.1-1.2), (all p<0.05). A model considering these factors together had good discriminatory power for predicting a FP result: AUC 0.80, 95% CI 0.78-0.83. Conclusions: Half of SPECT were FP in a population of patients referred for ICA following clinical judgment and a SPECT positive for ischemia. When analyzed together, the absence of severe angina/typical angina, the presence of atypical angina, normal left ventricular ejection fraction, female gender and lower FRS have good power for discriminating a FP result. These parameters should be given more relevance in order to avoid some unnecessary referrals for ICA.