Diabetes Mellitus in Patients with Myocardial Infarction Complicated by Heart Failure: A ‘Low Ejection Fraction’ Equivalent?

This editorial refers to ‘The inter-relationship of diabetes and left ventricular systolic function on outcome after high-risk myocardial infarction’ by A.M. Shah et al., published in this issue on page 1229–1237. Diabetes mellitus (DM) is a public health problem of massive and rising proportions; the estimated global prevalence of DM increased from 135 million in 1995 to over 220 million in 2009, and is projected to surpass 366 million by 2030.1 Compounded with the markedly increased risk of coronary heart disease (CHD) associated with DM, and with the high incidence of CHD-associated heart failure (HF) in patients with DM,2 the increase in prevalence of DM equates to more and more patients requiring concomitant management of DM, CHD, and HF. In this context, understanding the interactions among these conditions—from pathophysiology to epidemiology to clinical practice—has never been more imperative. More than a decade ago, it was postulated that diabetic patients with no history of CHD have a risk of myocardial infarction (MI) similar to that of non-diabetic patients with prior MI.3 Although the original study underpinning this concept had limited statistical power and precision, the concept that DM is a ‘CHD equivalent’ for long-term cardiovascular disease prognosis has contributed to more directed and intensive CHD risk mitigation efforts targeting the patient population with DM, for example, in specific guideline recommendations for more intensive management of lipids and blood pressure.4,5 More recent data from several large randomized trials suggest that the risk of CHD events in patients with DM, most of which comprised DM patients with prevalent CHD or clustered CHD risk factors, is significantly lower than the risk traditionally used to define ‘CHD equivalency’;6–9 that is, about 1% annualized CHD risk in most of the recent DM trials, contrasted with a 2% annualized CHD risk traditionally considered ‘CHD equivalent risk'. The lower risk observed in the DM trials likely reflects some degree of clinical success attributable to the more specifically targeted and aggressive risk intervention applied over the past decade. In a similar construct, the study by Shah et al. published in this issue of the European Journal of Heart Failure suggests that following an acute MI complicated by signs and/or symptoms of HF, DM is a ‘low left-ventricular ejection fraction (LVEF) equivalent’ with regard to subsequent clinical risk prediction.10 Analysing data from the Valsartan in Acute Myocardial Infarction (VALIANT) trial, a study of 14 703 patients with MI complicated by signs or symptoms of HF with or without compromised LVEF, DM was associated with an increased risk of adverse clinical events across the LVEF strata, including all-cause death, cardiovascular death, death or recurrent MI, and death or hospitalization for HF, such that non-diabetic patients with lower LVEF and diabetic patients with preserved LVEF had similar risks of adverse outcomes.10 Aside from potential implications for post-MI risk stratification, the concept of DM independent of EF as low-LVEF equivalent may prove useful with regard to clinical decision-making in the management of patients with DM and acute MI, much has been incorporated into guidelines and practice for the chronic ambulatory management of CHD risk in DM patients. For example, the long-term use of angiotensin-converting enzyme (ACE)-inhibitors post-MI is currently recommended in patients with low LVEF (<40%) as well as in patients with DM across the spectrum of ACS events, both in Europe and in the United States (class I, level of evidence A).11–14 However, ACE-inhibitors continue to be under-used post-MI especially in patients with DM when compared with non-DM patients.15 The present findings of Shah et al. underscore the imperative to more equitably use evidence-based therapies, including but extending beyond ACE-inhibitors, post-MI in patients with DM with the goal of eroding the residual risk associated with DM in this population. Furthermore, while most guidelines specifically indicate that ACE-inhibitors should be started within 24 h of admission for MI in patients with decreased EF,11–13 no such indication exists for patients with DM in the absence of decreased EF. Therefore, if DM is indeed a low-LVEF equivalent, perhaps early vs. later initiation of ACE-inhibitor therapy could yield incremental benefit in patients with DM. This concept is in fact supported by post hoc analyses of the diabetes subgroups from randomized clinical trials evaluating early initiation of ACE-inhibitors. In these sub-analyses, ACE-inhibitor therapy initiated within the first 24 h post-MI resulted in no overall benefits, but did result in improved survival in the subgroup of patients with DM: all-cause mortality at 6 months post-MI among DM patients was reduced from 16.7 to 13.5% with early administration of enalapril in the Cooperative New Scandinavian Enalapril Survival Study II (Consensus II),16 and from 16.1 to 12.9% with early administration of lisinopril in the Gruppo Italiano per lo Studio della Sopravvivenza nell'Infarto Miocardico-3 (GISSI-3) study.17 While the results of these sub-analyses from otherwise negative clinical trials are not definitive, they certainly support the hypothesis of differential therapeutic effect of such a strategy in the setting of DM. A similar concept of equivalency between DM and HF with decreased LVEF derives from the Eplerenone Post-acute Myocardial Infarction Heart Failure Efficacy and Survival Study (EPHESUS) trial,18 the results of which led to the current recommendations for the use of aldosterone receptor blockers after MI in patients who are already receiving ACE-inhibitors and beta-blockers, have low LVEF (≤0.40), and have either symptomatic HF or DM (class I, level of evidence B in Europe; class I, level of evidence A in the United States).11–14 A second key observation from the analyses by Shah et al. is the differential association (i.e. statistical interaction) between LVEF and DM for selected outcomes. While the incremental risk associated with DM was uniform across the LVEF categories for most of the endpoints analysed, for the composite endpoint of death or hospitalization for HF, the incremental risk associated with DM was higher in patients with preserved systolic function and lower in patients with low LVEF. One possible explanation for this observation is competing risk—that is, with low EF, outcomes are more affected by cardiac dysfunction than other risk factors, but with preserved EF, other risk factors and co-morbidities such as DM play a more prominent role in determining clinical outcomes. Similar observations have been made with decreased magnitude of risk associated with DM in older vs. younger patients; and in men compared with women (especially pre-menopausal women). Of course, other explanations for the observed interaction are also plausible, as discussed by Shah et al. Even in the setting of preserved EF, patients with DM have myriad abnormalities of cardiac structure, function, and metabolism that may directly or indirectly influence clinical outcomes,19 and likewise, may represent additional targets for future therapeutic interventions. Finally, one specific limitation noted by the authors warrants additional comment. Specifically, the omission of data on DM duration, severity, and treatment, and the inability to account for the use of thiazolidinedione (TZD) medications is important, as all these factors potentially increase the risk for HF decompensation.20 Though contraindicated in patients with HF, the TZDs are used at relatively high rates in patients hospitalized for HF,21 and may have contributed to the especially high rates of re-hospitalization for HF among patients with DM, with or without low LVEF, in the study. In summary, the study by Shah et al. highlights that it is imperative to aggressively apply evidence-based therapies in patients with DM and MI; the need for additional therapies to further improve outcomes in these high-risk patients; and the importance of further research aimed at understanding the special cardiovascular vulnerabilities of patients with DM.