Abstract 316: Favorable Long Term Rates of Revascularization and Myocardial Infarction with Optimal Medical Therapy: No Effect of Second Opinion Status for Invasive Management

Abstract

Long-term outcomes of patients (pts) with stable coronary artery disease (CAD) are needed to properly assess the efficacy of Optimal Medical Therapy (OMT). Whether pts who seek a second opinion (SO) regarding OMT are at higher risk by virtue of the first opinion for invasive management is unknown. As part of a prospective cohort study of OMT in pts with stable CAD without prior coronary artery bypass graft (CABG), we ascertained opinion status at entry and the end-points of myocardial infarction (MI) and revascularization (revasc) as of 2005. Out of 693 pts, opinion status was available in 646 pts; of these, 118 were SO for invasive evaluation and management. Management decisions were based on history, physical examination, and non-invasive data; coronary anatomy was available in only 50% of the SO cohort. During follow up, the decision to intervene depended on stability of symptoms, LV function, and treadmill exercise duration. Revasc was defined as percutaneous coronary intervention (PCI) or CABG. MI was defined as meeting two of the three WHO criteria or a typical pattern of MI on a myocardial perfusion study or 2-D Echocardiography. Assessment of clinical predictors of each endpoint used stepwise multivariate Cox modeling with 0.15 as level of entry and 0.8 as level of removal. Variables significant at the 0.05 level in a univariate model as well as SO status were included. Results: The mean age of the cohort was 67.3 years, and the majority of pts (53.9%) had class 1 or 2 angina. During mean follow-up time of 6.9 years there were a total of 252 revascularization events and 114 MIs, of which 11 were fatal. SO patients were younger with a mean age of 63.9 years. Coronary anatomy was available in 39% of the entire cohort, and 50% of the SO group. The strongest predictor of MI or revasc in Cox models was a change in clinical stability of CAD. For Revasc, other predictors were BMI, total cholesterol, diastolic BP, and angina at entry; for MI, other predictors were resting ejection fraction and treadmill duration. SO status was not a predictor of either event. Conclusions: Long term rates of revasc and MI in pts with stable CAD managed with OMT were low, and comparable to the medical arms of the COURAGE and BARI2D trials. Our study extends those results to a longer follow up period. SO status per se was not a predictor of either revasc or MI, suggesting that pts with a previous recommendation for invasive management who sought a SO did not represent a higher risk sub-group.