Abstract 1794: Evaluation of Usefulness and Efficacy of a “Segmental Perfusion Abnormality Index”: A Novel Diagnostic Approach with Conventional Enhanced Computed Tomography for the Differentiation of Acute Coronary Syndrome in Patients with Acute Onset of Chest or Back Pain
Background: Immediate and accurate differential diagnosis is critically important for the patients who present acute onset of chest and/or back pain. We have noticed that regions of decreased contrast enhancement existed within the myocardial wall of the patients with acute coronary syndrome (ACS) on conventional computed tomography (CT) images, performed to differentiate other acute thoracic diseases (e.g. aortic dissection, pulmonary embolism). This study was designed to utilize these features for the differential diagnosis of ACS.
Methods: The study population consisted of 95 consecutive patients who required enhanced CT for the differential diagnosis of acute chest discomfort at the emergency room. Using the representative apical four-chamber view, the left ventricular (LV) wall was divided into the following 6 segments: 1) basal septum, 2) mid septum, 3) apex, 4) mid lateral, 5) basal lateral and 6) basal posterior. A CT number was measured at each segment, and was then fitted into either of the following formula to calculate a segmental perfusion abnormality (SPA) index based on the existence of visually-apparent low density segments (VALDS): 1) In case of VALDS presence: average CT numbers of VALDS/average CT number of the rest normal segments × 100 (%), 2) in case of VALDS absence: the minimum CT number/average CT number of the rest 5 segments × 100 (%). The SPA index was compared between the two groups, one with ACS and another without ACS, both confirmed by coronary angiography.
Results: Forty-five patients (47.4%) were eventually diagnosed as ACS by coronary angiography. Importantly, the SPA index was significantly smaller in the ACS group compared to the non ACS group (48.7±17.2% vs 85.5 ± 7.1%, p<0.0001). The standard receiver operating characteristic curve indicated that an optimal threshold of the SPA index is estimated to be 65%, with an achievement of high diagnostic accuracy (sensitivity: 86.7%, specificity: 98.0%, positive predictive value: 97.5%, and negative predictive value: 89.0%).
Conclusions: The SPA index, a novel CT-based diagnostic approach for the non-invasive visualization of the ischemic LV wall, is confirmed to be easy and accurate method to differentiate ACS among the various patients who present acute chest or back pain.