Abstract 4359518: Accuracy of Angiographic Identification of Culprit Lesions in Women with MINOCA

Abstract

Background: Myocardial infarction with nonobstructive coronary arteries (MINOCA) accounts for 6-15% of myocardial infarctions and disproportionately affects women. Coronary culprit lesions in MINOCA occur in ~50% of cases and may be identified by multivessel optical coherence tomography (OCT). Identification of culprit lesions may guide pharmacologic therapy, but multivessel intravascular imaging remains infrequently utilized in MINOCA due to procedural complexity, time, risk, and uncertainty regarding culprit vessel. Research Question: How accurate are interventional cardiologists at identifying OCT-defined culprit lesion location on invasive coronary angiography in MINOCA patients? Methods: We conducted a retrospective substudy of women with MINOCA and a culprit lesion as identified by the OCT core laboratory in the AHA-funded Heart Attack Research Program (HARP). The OCT core laboratory identified the location of each culprit lesion on angiography. Separately, two interventional cardiologists blinded to the results of OCT independently reviewed invasive coronary angiography to identify the culprit vessel and segment, before and after review of electrocardiography (ECG) and echocardiography (echo). We assessed physician accuracy in determining culprit lesion location and the impact of ECG and echo on accuracy using chi-square tests. Results: A total of 67 women with MINOCA and an OCT-defined culprit lesion were included in the analysis (median age 64, 97% NSTEMI). Culprit lesions were most often located in the LAD (49%) or RCA (33%). At least one interventional cardiologist correctly identified the culprit vessel in 52% (95% CI: 41–64%) of cases. Both were correct in only 27% (95% CI: 18–39%). The addition of ECG and echo did not improve accuracy of culprit lesion identification (P=0.98). Accuracy decreased further when interventional cardiologists were asked to identify the culprit coronary segment, with 39% (95% CI: 28–51%) correctly identified by ≥1 cardiologist. Accuracy of culprit lesion localization was highest for plaque rupture (88%). Accuracy was particularly poor for left circumflex culprits, which were missed by both cardiologists in all 11 cases. Conclusion: Angiographic identification of culprit lesions in women with MINOCA is challenging, even after consideration of ECG and echo data. These findings reinforce the utility of multivessel OCT to find culprit lesions in patients with MINOCA.