In a large prospective study, we demonstrate that patients with type 2 myocardial infarction represent a heterogeneous population with high-cardiovascular risk. Furthermore, we were able to identify the most important predictors, which might help physicians to differentiate patients presenting with T2MI or T1MI (Figure 5).
Central illustration of T1MI and T2MI patients. T1M1, type 1 myocardial infarction; T1M2, type 2 myocardial infarction.
Type 1 and Type 2 Myocardial Infarction
Patients with suspected MI frequently present to the ED and require an early decision-making to either rule-out, rule-in, or observe. While the rule-out of MI based on high-sensitivity troponin provides a high level of confidence, there is less certainty regarding the rule-in of patients. Patients with symptoms suggestive of MI and dynamic changes of troponin would usually be grouped to rule-in and diagnosed with MI. The physician now should decide if an angiography needs to be performed, either urgently or electively. While this is recommended by current guidelines in most T1MI patients, urgent angiography might not be beneficial for T2MI patients. T2MI patients have heterogeneous causes of symptoms, which included arrhythmias, heart failure and severe hypertension in our study. Therefore, treatment of the underlying pathology is the primary goal of treatment in T2MI. As all patients present with symptoms of MI and have elevated troponin per definition of MI, the clinical differentiation is challenging. Earlier studies utilizing medium-sensitivity troponin were not able to differentiate between T1MI and T2MI based on this biomarker. Using high-sensitivity troponin assays, even lowest concentrations can be measured accurately. By utilizing a high-sensitivity troponin I assay, we could document distinct changes of troponin at admission, after 1 and 3 h for T1MI and T2MI patients. As expected, T2MI patients had lower hs-TnI concentrations. Nevertheless, a single cut-off resulted in an AUC of only 0.63. This shows that a single troponin cut-off is not able to differentiate between both entities. Therefore, we aimed to include additional clinical relevant parameters to identify T2MI patients. After multivariable selection, we identified female sex, typical radiating chest pain and a baseline hs-TnI ≤ 40.8 ng/L as the most important predictors. These parameters are easily accessible in the ED and can be obtained rapidly after admission. The combination of these parameters to a binary score improved the discrimination to an AUC of 0.71. These findings highlight that the distinction of T1MI and T2MI remains challenging even after this careful selection of diagnostic parameters. However, these identified predictors might help physicians to differentiate T1MI and T2MI patients.
What could be the clinical benefit of this differentiation? Patients with T2MI often have coexisting obstructive CAD and invasive angiography might be reasonable at some point in time. However, in our study only 38% of all T2MI patients underwent angiography and none received revascularization at the index admission or in the follow-up period. Therefore, T2MI patients should not undergo invasive strategies within the ED, but as discussed above require treatment of the underlying cause. In summary, we would emphasize the careful clinical evaluation of T2MI patients, which might (but not unalterable) include invasive angiography at some point in time. For T1MI patients, an urgent invasive strategy is clearly recommended. A high score could differentiate these different populations and identify T2MI patients.
Comparison to Previous Studies
Compared with previous studies, our study design differs significantly: prior studies with focus on T2MI were mostly performed in registers or cohorts with unselected determination of troponin.[10,20] In the BACC study, we prospectively included only patients with suspected MI. These patients would be addressed in the ED setting in order to decide whether a patient should undergo early invasive diagnostics. This setting is important with regards to the causes of T2MI. Earlier studies included patients with severe infection or anaemia who had dynamic troponin changes. These populations are barely represented in the BACC study, where most causes for T2MI were cardiac related including heart failure, hypertension, and arrhythmias. Earlier studies reported very high-mortality rates for T2MI patients of up to 50% after 4 years. In our present study, we could not validate these high-mortality rates. Again, this might be explained by the preselection of patients with suspected MI. Nevertheless, T2MI patients had a mortality rate that was comparable to T1MI patients. Interestingly, most T2MI patients in our study died from cancer or respiratory failure and barely from cardiac reasons. These findings are in line with prior publications with focus on T2MI. Furthermore, in patients with suspected MI we could recently show, that hs-TnI concentrations between 6 and 27 ng/L were associated with a higher mortality rate, while cardiac causes of death were rare. Taken together, these findings highlight the predictive value of elevated troponin concentrations for future events. Elevated troponin concentrations without dynamic changes can be caused by various conditions including chronic heart failure, chronic kidney disease or stabile CAD. This is labelled as myocardial injury and also associated with poor outcome. The differentiation of T2MI and myocardial injury is challenging and there is substantial overlap for several conditions (e.g. acute and chronic heart failure or stable CAD with borderline stenosis and new onset atrial fibrillation). Therefore, this topic remains a matter of debate[24–26] and needs to be addressed in future studies and guidelines.
Strengths and Limitations
One strength of our study is that we prospectively included patients with suspected MI, who represent a clinically important population. The adjudication of the final diagnosis has been performed carefully and blinded by two physicians, which is another strength. Nevertheless, the exact definition of T2MI is crucial and challenging even after blinded adjudication. Therefore, some cases could have been misclassified. However, during the follow-up period no MI and only one PCI were observed for T2MI patients. This could indicate that no T1MI was falsely classified to be T2MI. Nevertheless, in future studies more detailed imaging results should be investigated for T2MI patients. A second limitation is that our results are specific for high-sensitivity troponin I and therefore limited to this assay. Finally, external validation of the developed point score is required in order to show the robustness of the model.
Eur Heart J. 2017;38(47):3514-3520. © 2017 Oxford University Press
Copyright 2007 European Society of Cardiology. Published by Oxford University Press. All rights reserved.