Abstract P271: Mortality Benefit With Drug-Eluting Stents in Observational Studies: Fact or Artifact? Insights From the Prospective, Multicenter EVENT (Evaluation of Drug Eluting Stents and Ischemic Events) Registry

Abstract

Background: Both randomized trials and observational studies comparing drug-eluting stents (DES) with bare metal stents (BMS) have demonstrated substantial reductions in target lesion revascularization (TLR) with DES use. Data on mortality, however, are strikingly disparate with only observational studies supporting a benefit of DES. We hypothesized that these differences are related predominantly to residual selection biases in non-randomized studies. Methods: We examined the association between stent type (DES vs. BMS) and both TLR and mortality over one year among 9266 unselected pts who underwent PCI at 55 US centers and were enrolled in the prospective EVENT registry from 2004 to 2007. In addition to standard approaches to risk-adjustment (i.e., multivariable regression and propensity matching), we performed an instrumental variable (IV) analysis using the treatment era (2007 vs. 2004-6) as the instrument. Results: Overall 88% pts received DES, and there was a substantial difference by treatment year (93% in 2004-06, 73% in 2007, p<0.001). Unadjusted event rates over one year were significantly lower with DES compared to BMS (Death: 2.5% vs. 5.6%, TLR: 4.2% vs. 6.9%, p for both <0.001). Both multivariable and propensity-matched Cox regression analyses demonstrated a significantly lower risk of death and TLR with DES use (Table). IV analysis showed a strong trend towards lower TLR with DES but no difference in mortality. Conclusion: Among unselected PCI patients in contemporary practice, DES use is associated with a consistent reduction in TLR regardless of risk-adjustment methodology but no effect on mortality using IV analysis. These findings underscore the limitations of conventional multivariable and propensity score analyses to adequately address both known and unknown factors influencing treatment selection in non-randomized studies. Adjusted * risk of post discharge clinical outcomes over one year after PCI with drug-eluting stents (reference: bare metal stents) Death Any TLR Hazard ratio (95% CI) P value Hazard ratio (95% CI) P value Multivariable Cox regression 0.51 (0.37-0.69) <0.001 0.49 (0.37-0.64) <0.001 Propensity-matched Cox regression 0.54 (0.39-0.74) <0.001 0.52 (0.39-0.69) <0.001 Absolute risk difference(%) (95% CI) P value Absolute risk difference (%) (95% CI) P value Instrumental variable analysis (instrument: treatment era) 1.9 (-1.9, 5.7) 0.33 -4.5 (-9.0, 0.3) 0.07 TLR: Target Lesion revascularization; * Adjusted/ matched for: Age, gender, insurance status, hyperlipidemia, hypertension, diabetes, smoking status, medical history (myocardial infarction, peripheral artery disease, stroke, heart failure), history of PCI or bypass surgery, e Glomerular filtration rate, ejection fraction, multivessel disease, attempted lesion location (left main, LAD, SVG), # lesions treated, stent length, # of stents, maximum stent diameter