This quantitative model demonstrates that the late coronary lumen diameter and the probability of restenosis after Palmaz-Schatz stenting or directional atherectomy are influenced strongly by the lumen diameter present immediately after the procedure rather than by the specific device used. Although the influence of a larger acute result on reduced restenosis appears to be well established in this treatment population, the interplay among the multiple other biological influences on restenosis limits the ability to predict the probability of restenosis for the individual patient based on a large acute result alone. Future studies of restenosis, however, can further refine this multivariable quantitative model by adjusting for the effects of other clinical variables, mechanical interventions, or drug therapies in addition to the clear effect of postprocedure luminal diameter.

After arterial injury produced by stent placement, diabetic patients have a significantly greater incidence of restenosis because of greater late loss at the treatment site. Because elastic recoil or vasospasm contributes little to stent restenosis, the increased late loss of minimum lumen diameter in diabetic patients suggests that they have a greater predisposition to intimal hyperplasia.

Palmaz-Schatz stents can be placed successfully with a low incidence of major complications. The angiographic restenosis rate was 25%, and 70% of patients remained free of cardiovascular events at 3 years. Diabetes, small postprocedure lumen diameter and stenting of the left anterior descending artery are associated with higher rates of restenosis.

Six-month angiographic follow-up of patients who underwent directional coronary atherectomy during the 1st 3 years of our experience shows an overall restenosis rate of 32%, with lower rates in patients with a postatherectomy lumen diameter > or = 3 mm, cholesterol level < or = 200 mg/dl or a recent myocardial infarction. Few if any events relating to the site of atherectomy developed after the 1st year of follow-up.