Purpose: To identify, using logistic regression models, factors impairing evaluation of patency of drug-eluting stents (DES) and bare metal stents (BMS) in coronary arteries by 64-slice MSCT versus conventional coronary angiography (CAG).
Materials and methods: We evaluated 75 stents (16 DES and 59 BMS; 57 in males) implanted in 49 consecutive subjects (35 males; aged 68+/-10 years) by enhanced ECG-gated MSCT (Light Speed VCT, General Electrics) and CAG. Stents were classified by implantation site: (1) right coronary artery, (2) left anterior descending branch (LAD), and (3) left circumflex branch. Logistic regression models were used to predict agreement of findings between CT and CAG and impossibility of evaluating stent patency by CT, using age, sex, body mass index (BMI), heart rate, stent type, stent size, diabetes mellitus, and hypertension.
Results: By CAG, patency was confirmed in 15 DES (94%) and 54 BMS (91%); significant in-stent restenosis of >/=50% was observed in 1 DES (6%) and 4 BMS (7%), and total occlusion in 1 BMS (2%). By CT, patency was confirmed in 69% DES and 78% BMS; 5 BMS (8%) revealed a low CT area inside the stent with significant in-stent restenosis, and 2 BMS (3%) revealed total occlusion. Patency of 31% DES and 22% BMS was unconfirmed by CT. Agreement of findings between CT and CAG was observed in 69% DES and 80% BMS. Significant negative predictors of agreement of findings between CT and CAG were, by category: (1) BMI (relative risks, 0.77; 95% confidence intervals [95% CI], 0.62-0.97); (2) site of implantation in LAD only (positive predictor) and BMI (relative risks, 4.59 and 0.77; 95% CI, 1.02-20.6 and 0.61-0.97, respectively); (3) BMI (relative risks, 0.77; 95% CI, 0.62-0.97). Significant predictors of impossibility of evaluating stent patency by CT were, by category: (1) BMI (relative risks, 1.54; 95% CI, 1.01-2.36); (2) none; (3) BMI (relative risk, 1.65; 95% CI, 1.03-2.63). Stent type and size were not significant predictors in any category. The percentage agreement of findings between CT and CAG was significantly higher in subjects with BMI<22 kg/m(2) than in those with 22 kg/m(2)</=BMI<25 kg/m(2) (p=0.035) and 30 kg/m(2)</=BMI (p<0.01). Analysis of evaluable stents by CT only revealed no significant negative predictor of agreement of findings between CT and CAG.
Conclusion: 64-slice MSCT is a non-invasive diagnostic modality with potential for evaluating patency of DES with a low incidence of restenosis. When BMI is high, however, it remains difficult to accurately evaluate stent patency irrespective of stent type and size.