Severity of Myocardial Infarction: new insights on an elusive construct

Cardiovascular disease constitutes a burden of epidemic proportion (http://www.who.int/healthinfo/global_burden_disease/2004_report_update¹) and understanding its determinants is essential to design effective interventions. Doing so requires the ability to track disease burden at the population level. In the US, without national registries community surveillance is the method of choice to do so.² Community surveillance studies are retrospective by design, rely on diagnostic codes for case finding and require a defined population in which events can be consistently and reliably captured and validated using standardized approaches. Applied to cardiovascular disease, community surveillance measures its burden in communities by tracking the incidence of events, their severity and their mortality, thereby enabling the appraisal of the components of cardiovascular diseases in one given population.² Due to the aforementioned methodological requirements, few surveillance studies exist in the US. These include the Atherosclerosis Risk in Communities (ARIC) study,³ the Minnesota Heart Survey (MHS),⁴ the Olmsted County Study,⁵ and the Worcester Heart Attack Study.⁶ The data from these studies indicate that, while deaths from coronary disease have declined, the incidence of myocardial infarction in the US remained mostly stable. Thus, the decline in mortality can be envisioned as reflecting improvement in survival, which may be mediated by a declining severity of myocardial infarction.

The paper by Myerson et al in this issue of Circulation ⁷ specifically addresses this important and understudied issue in ARIC between 1987 and 2002. They examined a large, multiracial population and relied on several indicators including the composite score PREDICT to conclude that the severity of infarction declined over time. Indeed, the proportion of infarctions with major electrocardiographic abnormalities, ST segment elevation and Q waves decreased as did biomarker values and the proportion of persons presenting with shock. The PREDICT score also improved over time. Hence, these findings demonstrated a consistent improvement across all severity indicators thereby suggesting that the declining severity of myocardial infarction contributed to the decline in coronary disease mortality. As the upper age limit of the persons in ARIC is 74 years, these data do not include a growing number of patients experiencing myocardial infarction. Conversely, a distinct strength of the ARIC data is its large and diverse cohort allowing the examination of trends in African Americans, which paralleled overall trends. These data are important for several reasons.

First, they support and extend previous reports of a decline in the severity of myocardial infarction as reported in studies that examined severity indicators in ARIC, Worcester and Olmsted County.^8–10 The ARIC data reported on in this issue of Circulation are the most contemporary and pertain to a large cohort, thereby providing robust evidence that the trends observed in earlier time periods and with smaller sample sizes are confirmed and sustained. Importantly, the ARIC findings are independent of changes in biomarkers as the recent shift to troponin is adjusted for analytically.

Second, these data provide the important opportunity of an in-depth reflection on the meaning and implications of measuring the severity of myocardial infarction, a complex construct reflecting the composite result of disease severity indicators, care seeking behaviors and processes of care. For the ease of discussion, the indicators of severity of myocardial infarction can be schematically envisioned as reflecting the initial presentation of infarction or its in-hospital course. Cardiogenic shock and electrocardiographic ST segment elevation reflects the initial presentation of myocardial infarction and its characteristics during the initial twenty-four hours and are unlikely to be affected by in-hospital treatment. The changes in the frequency of cardiogenic shock should be interpreted in light of the trends in out-of-hospital coronary disease deaths. Indeed, declining out-of-hospital coronary disease deaths may result in greater numbers of persons hospitalized who would have in the past died out of hospital. This, in turn, may change the case mix of hospitalized infarction, thereby confounding temporal trends in cardiogenic shock. These considerations notwithstanding, a favorable temporal decline in cardiogenic shock was noted in ARIC and other communities.¹⁰

In acute myocardial infarction, electrocardiographic changes are time dependent such that the time between symptoms onset and electrocardiographic recording (hence presentation to care) affects the ability to capture ST segment elevation on the electrocardiogram.¹¹ Thus, changes over time in the time between symptom onset and presentation to care can confound temporal trends in the proportion of patients with ST segment elevation and should be considered for their interpretation. In ARIC, the time from onset of symptoms to arrival at the hospital did not change over time with only one third of patients presenting less than two hours after symptom onset ⁷. This sobering finding underscores the challenge of reducing total ischemic time by influencing care seeking behavior¹² and the need for renewed efforts in this domain. Conversely, as the time from symptom onset and presentation to care did not change, it is unlikely to confound trends in the proportion of ST segment elevation infarctions. A striking finding in the present report is the decreasing proportion of ST segment elevation infarctions in the community, which has been consistent across studies ^7–9 independently of time to presentation. While the reasons for this observation can only be speculative, one possible explanation is the increasing use of medications for the primary prevention of coronary disease. Indeed, over time, patients presenting with myocardial infarction are more likely to be receiving “cardio-protective” medications such as aspirin and beta blockers before admission for the acute event.¹³ These agents may reduce the size and severity of the infarction.¹⁴ Irrespective of these hypothetical considerations, recognizing the temporal decline in cases with ST segment elevation is essential to optimize care. Indeed, as non-ST elevation infarctions now constitute the vast majority of infarctions in communities, interventions must be designed while recognizing this epidemiological reality. This is particularly important as non-ST elevation infarctions may be receiving less optimal care compared to ST segment elevation infarctions.¹⁵

Other severity indicators including biomarkers (creatine kinase and troponin) and the development of electrocardiographic Q-waves reflect the in-hospital course of infarctions and may be influenced by treatment, particularly reperfusion. Evaluating the severity of infarction from biomarker values is subject to measurement error and variations in the timing and frequency of the measurements. The variability of the assays and the frequency of confounding by comorbid conditions are such that recorded values may not accurately reflect the true levels, and longitudinal trends may be impacted by assay changes. Further, the recent change in biomarkers with preferential reliance on troponin generates considerable challenges for the longitudinal use of biomarkers to assess the severity of infarction over time. The interpretation of electrocardiographic Q-waves is equally challenging as their appearance, like ST segment elevation, is time dependent¹¹ and may be influenced by reperfusion. Nonetheless, the prognostic implications of the development of Q-waves have been reaffirmed in an ancillary analysis of the Danish Trial in Acute Myocardial Infarction-2 (DANAMI-2) ¹⁶ and thus Q-waves constitute a legitimate severity indicator given their outcome implications.

These considerations underscore the complexity of evaluating the severity of myocardial infarction. The concomitant use of several different indicators is an important strength of the present study, as the consistency across indicators provides reassurance of the robustness of the trends. Given the aforementioned influences of treatment, either primary prevention or acute care, on all severity indicators, it is likely impossible to untangle the “natural history” of myocardial infarction from its management. Further, one may also question the mere existence and relevance of the concept of natural history applied to one of the manifestations of coronary disease, which has been the subject of such intense therapeutic efforts over the past decades.

Pragmatically, the most important construct in this context is the improvement of the health of the population and patients. Hence, these results underscore the progress made over time in terms of decreasing severity and case fatality rates during a period characterized by major therapeutic advances, in turn identifying the vital importance of striving for optimal quality of care as measured by access, affordability, and alignment with the strongest evidence in primary prevention, treatment and secondary prevention of acute coronary syndromes.

Finally, the findings presented herein help interpret population trends in coronary heart disease and heart failure. Indeed, to appropriately contextualize this report and its relevance to our understanding of the evolving burden of cardiovascular disease, it is important to remember that the heart failure epidemic is often attributed to the growing number of infarction survivors.¹¹ Certainly, the absolute number of individuals living with myocardial infarction, which reflects the combined effect of the increase in the population, its aging and the decreasing case fatality rate of myocardial infarction, can be expected to increase the number of persons at risk for heart failure. Yet, these factors do not entirely explain the “heart failure epidemic” as the decline in the severity of infarction documented herein and in other studies would lead to hypothesize that less severe infarctions should result in less subsequent heart failure. Indeed, in several community studies, the incidence of heart failure after myocardial infarction most often remained stable or decreased ^17–19. Thus, the data on the declining severity of infarction and the trends in subsequent heart failure underscore the importance of evaluating other contributors to the heart failure epidemic.

In summary, the paper by Myerson and colleagues reminds us of the importance of community surveillance to understand cardiovascular disease trends and of the fundamental implications of community surveillance for patient care. To this end, indicators of myocardial infarction severity may to a large extent be a proxy for the quality of care and thus make contemporary trends in infarction severity challenging to dissect from a causality standpoint. Conversely, the pragmatic interpretation of these trends suggests that progress has occurred as infarctions are becoming less severe. This report also underscores that we must recognize the increasing predominance of non ST elevation infarctions to move forward with interventions to improve the care of those patients who now constitute the majority of those who present with infarctions.