Abstract 4558: An Optimal Selective Strategy for Use of Drug-Eluting Stents Based on a Markov Model of Restenosis and Thrombosis
Background Drug-eluting stents (DES) reduce the risk of restenosis but increase the risk of thrombosis compared to bare metal stents (BMS).
Methods We developed a Markov model from published meta-analytic data over 4 yrs (Stone et al. NEJM 2007) on DES (n=2,633) vs BMS (n=2,628) to define the optimal strategy with respect to restenosis, thrombosis, and adverse outcome events of death or myocardial infarction (MI). The model mirrors the underlying dynamic processes involved in physiological responses to stenting, and yields transition rates for restenosis, stent thrombosis and final outcomes (myocardial infarction or death), as well as estimates of the fraction of patients at higher than average risk for adverse complications following stenting.
Results The model identified a high-risk group comprising 19% of BMS patients experiencing 94% of restenosis at 1 yr and 85% of thrombosis at 4 yrs of follow-up. The event rate associated with unconditional use of BMS or DES vs selective use of DES in high-risk BMS patients (BMS|DES) is summarized in Table⇓. The rate of restenosis was higher for BMS than for DES, but the rate for thrombosis was higher for DES. Selective use of DES in high-risk BMS patients resulted in reduction in restenosis at 1 yr (89% vs BMS alone, 58% vs DES alone) and in thrombosis at 4 yrs (38% vs BMS alone, 62% vs DES alone). Adverse outcome events were reduced 40% at 1 year and 20% at 4 yrs.
Conclusions Markov modeling identifies a high-risk group comprising nearly 20% of stent patients experiencing >80% of adverse outcomes. Selective use of DES in these high-risk patients reduces restenosis, thrombosis and adverse events compared to nonselective use of BMS or DES. This strategy therefore improves overall clinical outcome at lower overall cost.