To the Editor.
—We enjoyed reading Dr Levy's1 Controversies on surrogate markers in acquired immunodeficiency syndrome research and concur with much of his analysis. However, on a few occasions, he may have carried his argument too far. For instance, we agree that the relationship between low viral load and favorable clinical outcomes seen in studies of the natural history of disease does not necessarily imply that similar benefits will be achieved when viral load is reduced by medication. But there is no logical reason to conclude that "a beneficial clinical outcome cannot be expected unless drug therapies mimic the natural cellular immune responses that lower the viral RNA." No clinically effective antiviral drug does this. Nor is this desirable, since natural immune responses have proved incapable of neutralizing the human immunodeficiency virus (HIV) in the long run. The value of a treatment regimen should be judged by its clinical