Dear Editor We are gratified to see that many of our answers to Kittleson and colleagues' 1st letter apparently provided satisfaction as these points were not mentioned in the second. Moreover, these authors now “do not dispute that ACE inhibitors (ACEI) may have other beneficial effects” than cardiac effects, which is substantially different from their statement in their 1st letter that ACEI “have been extensively studied without evidence of such benefits.” However, we believe that the criticisms put forth by Kittleson and colleagues are erroneous or merely personal opinions. The conclusion by Kittleson et al that all 3 published studies found that “ACEIs do not prolong the time to onset of heart failure in any breed with MVD” is incorrect, as the 16 authors of Vetproof concluded that “results suggest that enalapril modestly delays the onset of CHF in dogs with moderate to severe mitral regurgitation (MR).”1 Moreover, the implication that, in our study, we tried to hide results is unfounded: the fact that no significant differences in the time of onset of CHF were found between the ACEI and untreated groups is clearly stated on page 909.2 The statement that all-cause mortality should not be evaluated in such clinical trials is incorrect. A therapy that would reduce cardiac death but increase overall death rate would provide no overall benefit. As we wrote in our 1st letter, all-cause mortality is a standard endpoint in human cardiology. We do not agree with the statement that sudden death is not relevant to study in dogs with MVD. Time to sudden cardiac death was used as an endpoint in the recent QUEST study.3 Similarly, in a 56-month study period involving 71 dogs with MVD, 4 sudden deaths were recorded and only 2 dogs underwent cardiac-related euthanasia.4 Consequently, sudden death does occur in MVD and is relevant. On one hand, Kittleson and colleagues acknowledge retrospective studies' “contributions … to the standard of medical practice are incontrovertible.” On the other hand, they criticize such studies because “uncontrolled confounding” factors can lead to invalid conclusions “specifically for retrospective studies of treatment efficacy.” The concerns of Kittleson and colleagues seem surprising, while they have themselves published several retrospective studies dedicated to clinical outcome according to treatments.5, 6 Limitations of retrospective studies, which by definition are not randomized and blinded, are true regardless of the research topic and the authors. Such limitations, including the number of dogs, are clearly stated in our paper (pp. 912–913).2 Kittleson and colleagues state that “there are countless examples of observational studies of treatment efficacy whose findings have been refuted by subsequent randomized clinical trials” and provide as support 1 single reference, the New York Times Magazine, which is, in our view, not an appropriate nor peer-reviewed reference for such a strong statement. In contrast, it has been recently published in the New England Journal of Medicine that little evidence exists to show that observational studies find stronger treatment effects than randomized, controlled trials, contrary to what has been claimed for many years.7, 8 This conclusion applies exactly in the case of our study,2 which provides results in agreement with 2 previously published prospective trials.1, 9 Additionally, the practice of using retrospective trials to supplement randomized trials is justified and one need look no further than the prestigious journal Circulation to find a retrospective study designed to answer questions regarding treatment efficacy left unanswered by prospective clinical trials.10 Kittleson and colleagues state that “clinically unaffected dogs” with MVD “and a normal left atrial size” with “a moderate to severe mitral regurgitation (MR)” are “pathophysiologically impossible.” Firstly, it is “pathophysiologically” possible in humans.11 Secondly, Kittleson and colleagues refer to one of our papers in which 450/617 MVD dogs were asymptomatic.12 By comparison, in 1 retrospective study published by Kittleson and colleagues only 5/17 dogs with MVD were “judged clinically to have mild MVD.”13 Four of these 5 dogs had a regurgitant fraction (RF) between 22 and 41% (ie, for some of them moderate), and the other dog had a RF of 73% (ie, severe MR), but was arbitrarily considered by the authors as a “statistical outlier.” The authors of the letter have actually made similar observations to us but may have failed to recognize them or chose to term them “outliers.” If Kittleson and colleagues had included more than 5 dogs with “mild MVD,” they would have probably found more “statistical outliers.” Furthermore had they measured pulmonary arterial pressure, their statement about stage 1 MR and pulmonary hypertension would carry more weight.
Read full abstract