Angioscopic evaluation of incompletely obstructing coronary intraluminal filling defects: comparison to angiography.

At 66 sites in 40 patients, we evaluated the sensitivity and specificity of coronary angiography in detecting intraluminal filling defects of varying sizes and in characterizing the contents (thrombus, intimal flap, both) of such defects using coronary angioscopy as "the gold standard." Overall angiographic sensitivity for thrombus was 37% and for intimal flap 45%. Specificity for thrombus was 100% and intimal flaps 96%. Angioscopically small flaps were less frequently seen angiographically (28%) than larger sizes (65%, p = 0.03). Angioscopically small thrombi were seen less often angiographically (30%) than larger ones (75%, p = 0.13). Filling defects (intimal flaps, thrombus, both) were characterized correctly in only 37% of sites. Angiography is relatively insensitivity in detecting intraluminal filling defects. Angioscopy may be preferred to or adjunctive with angiography in detecting these lesions.