The role of coronary arteriography in demonstration of mural thrombosis after angioplasty. Insights from an experimental model.

Although intracoronary thrombosis often occurs after angioplasty and may affect its outcome, the accuracy of arteriography for identification of mural thrombi is unclear. This study analyzed the relationship between arteriographic abnormalities immediately before death and the histologic extent of thrombosis in 77 dogs submitted to balloon injury of intact left anterior descending coronary arteries. Survival time after angioplasty was 120 min. The incidence of mural thrombosis, defined on serial histologic sections, was 65.0 percent. A positive diagnosis of intracoronary thrombus at arteriography (AT+) was based on the presence of any of the following signs: filling defects, retention of contrast material, and slowed or interrupted flow. Seventeen dogs were AT+, and 60 were AT-. The overall sensitivity of arteriography was 34 percent, and the specificity was 100 percent. Even considering as significant only thrombi greater than 25.0 percent of the arterial lumen area, 11 of 27 dogs were AT- despite thrombus sizes between 27 percent and 75 percent of lumen area (sensitivity, 59 percent); arteriography consistently missed smaller thrombi (22 of 23 dogs were AT-). Arterial diameters and balloon-induced injury were similar between AT- and AT+ dogs. Scanning electron microscopy depicted a fibrin-poor thrombus in 14 of 19 AT+ dogs and a fibrin-rich thrombus in five, whereas all seven AT+ dogs had fibrin-rich thrombi. Logistic regression analysis showed a correlation between thrombus size and arteriographic positivity, whereas the presence of fibrin and slowed flow of contrast material did not independently predict positive arteriographic results. Thus, arteriography is inaccurate for identification of mural thrombosis after angioplasty, mostly because of its poor sensitivity.