Histopathologic validation of in-vivo angioscopic observation of coronary thrombus after angioplasty in a porcine model

Background Coronary angioscopy has been reported to be superior to angiogra phy and intravascular ultrasound for detecting intracoronary thrombus. Howe ver, in-vivo histopathologic validation of angioscopic detection of intraco ronary thrombus had not been performed. Objective To perform histopathologic validation of in-vivo angioscopic dete ction of coronary thrombus. Design An experimental, blinded comparison of angioscopy and histopathology . Methods Coronary angioscopy was performed from 0 to 14 days after angioplas ty in 39 porcine coronary arteries. When thrombus was detected by angioscop y, it was subclassified into white, mixed red-white, or red thrombus accord ing to color. By histopathology the presence of thrombus was determined and subclassified into platelet-rich, mixed platelet-erythrocyte, or erythrocy te-rich thrombus. Results Angioscopy correctly classified 19 of 21 coronary thrombi (sensitiv ity 90%) but incorrectly classified nine of 18 arteries without formation o f thrombus as having a thrombus (specificity 50%). Positiveand negative pre dictive values were 68 and 82%, respectively. The angioscopic subclassifica tion of thrombus into white, mixed red-white, or red thrombi was not correl ated to the corresponding histopathologic morphology (platelet-rich, mixed platelet-erythrocyte, or erythrocyte-rich) of the observed thrombi (chi (2) test: P = 0.5). Conclusions Angioscopic detection of thrombus in vivo had high sensitivity and negative predictive value but low-to-moderate specificity and positive predictive value. Visual assessment of color of angioscopically detected th rombi seemed not to reflect histopathologic morphology of thrombus accordin g to the definitions used in the present study. Coron Artery Dis 12:53-59 ( C) 2001 Lippincott Williams & Wilkins.