Khaled R. Hassan M.D.            Washington University
Angiographic predictors of plaque progression in mildly to 
moderately diseased coronary arteries   
         
 

Progression of atherosclerotic plaques has not been predicted by angiography. We hypothesized that progression of plaques creating <50% stenosis is predicted by fresh mural thrombus, lesion location at a branch point, and plaque blush –delayed clearance of contract possibly due to angiogenesis or cap fissuring. 

Methods: Films of 200 patients undergoing repeat angiography for symptoms of ischemia, 5.6±4.8 (mean±sd) months apart, were viewed by two blinded observers. 123 patients were excluded due to prior PTCA or CABG, initial lesion severity >50% and a non-comparable paired angiograms. Presence of plaque blush, calcification, clot (mobile defect), eccentricity, and branch point location were compared in progressing (>20% stenosis increase) and non-progressing plaques. 

Results: 16 lesions in 15 patients progressed from 29±13% to 68±14% over 8.1±7.9 months. Patients with and without progression were similar in gender mix, age, CHD risks, medications, days between angiograms, clinical presentation and initial stenosis severity. Logistic regression identified plaque blush (p=.002), calcification (p=.024) and a branch point location (p=.001) as the predictor of plaque progression. Using these signs, the model predicted the odds ratio for plaque progression (Orp) as : Orp =e 2.5*BL+1.8*CA+2.6*BR. The model has a 81% sensitivity, 77% specificity and a overall accuracy of 78% when Orp of 1/3 was used to classify the groups. In other words, a moderate (<50%) stenosis with both "blush" and branch point signs had a 25% chance of progressing within 8 months if calcified as well. Such lesions had a 100% likelihood of progressing but only 40% progressing lesions had all 3 signs. 

Conclusion: In mild to moderate coronary stenoses, plaque blush (a novel sign) branch point location and calcification are predictive of plaque progression. If confirmed by prospective analysis, these criteria may be helpful in clinical decision making.