The important question is whether the potential interaction is of clinical importance. The study by Aumégeat et al in this issue suggests that it is not. Most of their patients were treated with low-dose aspirin, which is in line with the theoretical considerations delineated above. However, the previously published data are, again, not completely consistent.19–24
It is important to note that all of the studies investigating the clinical importance on outcome in cardiovascular diseases were post hoc analyses of large trials or were retrospective cohort studies. Therefore, none of these studies was free from considerable imbalance of the baseline characteristics of patients who were and were not receiving therapy with aspirin. Thus, none of these studies is able to conclusively resolve the controversy, despite the fact that some of these studies included a very large number of patients.,19–21,23–24
The study by Aumégeat et al has the same limitations and included only 755 patients, thereby lowering the statistical power significantly. However, this study has some special features that may deserve consideration. Thus, it included consecutive, unselected patients from a single center. In contrast, most of the other studies19–21,23–24
were post hoc analyses of multicenter trials, excluding a significant proportion of patients. Moreover, the assessment of the patients was comprehensive, including coronary angiography, echocardiography, and ergospirometry in most patients. Importantly, this improves the value of the data because undetected confounding factors are much less likely to be present. Although it is not possible to record all of the potential confounders, the more completely they are recorded, the less likely it is that an important confounding factor remains undetected. The study by Aumégeat et al found that these potential confounders did not influence the results. Also, doses of ACE inhibitors and aspirin were recorded not only at study entry, but also later on. As delineated above, the dose of aspirin may be crucial in this regard. Additionally, the dose of ACE inhibition also may be important, since the inhibition of angiotensin II formation in the long term may be dose-dependent, being present with the use of high doses only.26
Less is known about the dose-response relationship of the non-angiotensin II-mediated effects of ACE inhibitors. However, the positive prognostic effect of ACE inhibition at doses without sufficient angiotensin II suppression may indicate that these effects occur at lower doses, although the beneficial effect is enhanced at high doses.27
Thus, the interaction may be, in theory, more important with low-dose ACE inhibition. The study by Aumégeat et al suggests that this was not the case in their CHF population.