934-28 Sensitivity and Specificity of Angiographic Markers for Thrombus: A Prospective Comparison with Angioscopy

The limitations of angiography for the detection of intracoronary thrombus are well recognized. Between November 1991 and July 1994, we performed 402 angioscopy procedures in 225 vessels in 202 patients, with the Image-Cath (Baxter). We performed a prospective study in 190 of these patients, who had an interpretable angioscopy performed just before PTCA to determine the sensitivity and specificity of predetermined angiographic criteria that are considered to be indicative of the presence of intracoronary thrombus. Angiographically verified thrombus was used as the gold standard for comparison. Lesions were classified on angiography (2 orthogonal views) by independent observers. The presence of an intraluminal filling defect, of overhanging edges, of haziness, or of ulceration were noted. The characteristic ulceration was not mutually exclusive of the other 3 characteristics. Of 15 filling defects on angiography 14 (93%) had thrombus on angiography; in the 23 lesions with overhanging edges 19 (83%) had thrombus on angioscopy; in the 27 ulcerated lesions 21 (78%) had angioscopic thrombus; in the 6 lesions that were hazy on angiography 5 had angioscopic thrombus.Empty CellEmpty CellAngioscopyEmpty CellEmpty CellEmpty CellThrombus+Thrombus-AngiographyThrombus+4512Thrombus-4093 In our model, using 5 prespecified angiographic characteristics, angiography had high specificity (89%) but relatively low sensitivity (53%) for the detection of thrombus compared to angioscopy.