We agree that the number of patients with mitral regurgitation during evolving infarction was small and that definitive statements on pathogenesis, prognosis, or treatment should not be made from this study alone. However, the highly significant 1-year mortality differences attest to the strength of association between regurgitation and death [1]. Moreover, in the same issue of Annals, Tcheng and colleagues [2] noted strikingly similar rates of unadjusted mortality in a larger cohort of patients with acute infarction.

The underlying mechanism relating mitral regurgitation and mortality remains perplexing. Most patients had mild regurgitation with no apparent effect on ventricular contractility, volumes, or filling pressures [1]. Our assessment of severity, however, occurred early; myocardial remodeling or other delayed processes may have further exacerbated valvular dysfunction to the point of late hemodynamic compromise and increased likelihood of death. Alternatively, mitral regurgitation may serve simply as a marker due to its association with a different predictive covariable. This covariable does not appear to be left ventricular dysfunction, given that mitral regurgitation showed an independent predictive power when assessed using a multiple logistic regression model. Indeed, regurgitation entered the stepwise model earlier than did ejection fraction. However, an important but unidentified covariable may yet exist. Further work is needed to better delineate the underlying mechanism before rational treatment can be developed.