A counterpoint paper: Comments on the electrocardiographic part of the 2018 Fourth Universal Definition of Myocardial Infarction endorsed by the International Society of Electrocardiology and the International Society for Holter and Noninvasive Electrocardiology

Abstract The Fourth Universal Definition of Myocardial Infarction (FUDMI) focuses on the distinction between nonischemic myocardial injury and myocardial infarction (MI), along with the role of cardiovascular magnetic resonance, in order to define the etiology of myocardial injury. As a consequence, there is less emphasis on updating the parts of the definition concerning the electrocardiographic (ECG) changes related to MI. Evidence of myocardial ischemia is a prerequisite for the diagnosis of MI, and the ECG is the main available tool for (a) detecting acute ischemia, (b) triage, and (c) risk stratification upon presentation. This review focuses on multiple aspects of ECG interpretation that we firmly believe should be considered for incorporation in any future update to the Universal Definition of MI.


| INTRODUC TI ON
The Fourth Universal Definition of Myocardial Infarction (FUDMI), published simultaneously in 2018 in numerous journals including Circulation, Journal of the American College of Cardiology and European Heart Journal, focuses mainly on the distinction between nonischemic myocardial injury and myocardial infarction (MI) and the role of cardiovascular magnetic resonance in defining the etiology of myocardial injury, with less emphasis on updating the parts related to the electrocardiographic (ECG) changes related to MI (Thygesen et al., 2019).
Evidence of myocardial ischemia is a prerequisite for the diagnosis of MI and the ECG is the main diagnostic tool for detecting acute myocardial ischemia, as stated in the document, viz "Myocardial ischemia in a clinical setting can most often be identified from the patient's history and from the ECG" (Thygesen et al., 2019). Therefore, the definitions of "ischemic changes" should be accurate. The International Society of Electrocardiology (ISE) and the International Society for Holter and Noninvasive Electrocardiology (ISHNE) focus on the ECG and have members that are expert in interpreting ECG changes detected during ischemia and infarction. This counterpoint review focuses on several topics related to the ECG that we believe should be considered to be modified and incorporated into future versions of the document.

| ME THODS
YB, MF, KN, JGN, and AB read the FUDMI and communicated about topics related to the ECG that should be discussed in this article. Those ideas were incorporated into an initial draft by YB that has been circulated to all co-authors for review and

| MYOC ARDIAL INFARC TION A SSO CIATED WITH CORONARY ARTERY BYPA SS G R AF TING ( T YPE 5 MYO C ARD IAL INFARC TION)
Although the FUDMI states "It is important that the postprocedural elevation of cTn values is accompanied by ECG, angiographic, or imaging evidence of new myocardial ischemia/new loss of myocardial viability," the thresholds for the ECG changes, especially ST deviation, are not mentioned. The document specifies that ST-T changes are common after coronary artery bypass grafting (CABG) due to epicardial injury and are not reliable indicators of myocardial ischemia in this setting. "However, ST elevation with reciprocal ST depression or other specific ECG patterns may be a more reliable finding of a potential ischemic event" (Thygesen et al., 2019).
1. While later on, ST-T changes can be secondary to pericardial/ epicardial inflammation and, thus, can be nonspecific (Borgaonkar & Birnbaum, 2019), soon after completion of cardiac surgery they are probably more specific (Liu & Birnbaum, 2019). At this stage, when the patient is still sedated and intubated, ST deviation can be an early marker of acute bypass failure or type 5 MI. Further studies are needed to evaluate the accuracy of routine 12-lead ECG after completion of surgery to detect ischemia/ infarction.

As the thresholds for cardiac troponin elevation for diagnosing
MI are different for type 5 MI (>10 times the 99th percentile of the upper limit of normal) than for the other types of MI, it might be that different thresholds of ST deviation should be used in this scenario. As mentioned above, further studies are needed to clarify this issue.
3. Reciprocal changes are commonly seen in STEMI with ST elevation in the limb leads (inferior or lateral STEMI) (Birnbaum, Sclarovsky, Mager, Strasberg, & Rechavia, 1993). However, they are less common in anterior STEMI, especially when the left anterior descending (LAD) is occluded after the first diagonal branch . In our experience, acute MI caused by anastomosis failure or distal embolization of a graft to the LAD (usually the insertion is distal) normally does not cause reciprocal ST depression. interpretation that we firmly believe should be considered for incorporation in any future update to the Universal Definition of MI.

| ELEC TROC ARDIOG R APHIC DE TEC TION OF MYOC ARDIAL INFARC TION
As stated, "The ECG is an integral part of the diagnostic workup of patients with suspected MI, and should be acquired and interpreted promptly," we disagree with the statement that "more profound ST-segment shifts or T wave inversion involving multiple leads/ territories are associated with a greater degree of myocardial ischemia…". In our experience and based on the literature, T wave inversion in leads with an isoelectric ST segment does not occur with acute ischemia. It can be seen in the subacute phase or after reperfusion and should be regarded as "postischemic changes" . Negative T waves in leads with ST elevation are also seen after reperfusion or in the subacute phase of infarction. Only when seen in leads with ST depression can negative T waves signify acute subendocardial ischemia, or changes reciprocal to ST elevation in opposing leads. Therefore, we believe that T wave inversion should not be considered a sign of active ischemia.
It is written that "ST-segment depression ≥ 1 mm in 6 leads, which may be associated with ST segment elevation in leads aVR or lead V 1 and hemodynamic compromise, is suggestive evidence of multivessel disease or left main disease." However, in our opinion this statement should probably be restricted to patients in the appropriate clinical situation and to those with a relatively normal baseline ECG Knotts, Wilson, Kim, Huang, & Birnbaum, 2013). In many patients with left ventricular hypertrophy, critical aortic stenosis, cardiomyopathy, left bundle branch block, or nonspecific intraventricular conduction delay, dynamic diffuse ST depression associated with ST elevation in aVR can be seen. These changes can be more pronounced in patients with tachycardia or increased afterload. Implementation of the original statement could lead to overdiagnosing NSTEMI in patients with positive cardiac markers secondary to hypertensive crisis or exacerbation of acute heart failure.
We think that the next statement is also questionable: "Prolonged new convex ST-segment elevation, particularly when associated with reciprocal ST-segment depression, usually reflects acute coronary occlusion and results in myocardial injury with necrosis." The traditional literature emphasizes the "convex" pattern. Probably in the pre-reperfusion era when patients presented late (often with T wave inversion in the leads with ST elevation), the ST was often convex.
However, nowadays, when patients present early, in a large percentage of patients, the ST is concave (Huang & Birnbaum, 2011). This is especially common in patients with anterior STEMI presenting early with ST elevation and tall positive T waves ( Figure 1).
As mentioned above, reciprocal changes are commonly seen in inferior or lateral STEMI with ST elevation in the limb leads.
However, they are less common in anterior infarct, especially when the LAD is occluded after the first diagonal branch. Yet, occlusion of a short LAD before the first diagonal branch is usually associated with ST elevation in aVL and reciprocal ST depression in the inferior leads. However, in the majority of patients with anterior STEMI, reciprocal ST depression is not seen (Figure 2). "Reciprocal changes can help to differentiate STEMI from pericarditis or early repolarization changes." Indeed, in both early repolarization and acute pericarditis, reciprocal changes are commonly seen only in aVR. Yet, in patients with left ventricular hypertrophy, cardiomyopathy, and/or LBBB, "reciprocal" changes are common (ST elevation in V 1 -V 2 with ST depression in I, aVL, V 5 -V 6 ) (Birnbaum & Alam, 2014).
We believe that the cutoffs of ST-segment elevation for the different leads (table 2 in the FUDMI) should probably be limited to patients with narrow QRS and without voltage criteria for LVH (Macfarlane et al., 2004). [I introduced some of this in JECG 2004;37(Suppl): 98-103 and did say at that time that the criteria did not apply in the presence of LVH.] The thresholds for ischemic ST elevation in patients with LVH or cardiomyopathies have not been established (Birnbaum & Alam, 2014) (Macfarlane et al., 2004).
Moreover, many patients with right bundle branch block (RBBB) display ST depression in leads V 1 -V 3 at baseline. There are no established guidelines on how to diagnose acute inferolateral STEMI (ST depression in the anterior leads) in patients with complete or incomplete RBBB (Wei et al., 2013). In addition, it is unclear whether lower thresholds should be used for ST elevation in the anterior leads for diagnosing anterior STEMI in patients with complete or incomplete RBBB (Wei et al., 2013).
It should be remembered that due to the way in which the six limb leads are derived, "reciprocal" changes within these leads are merely a function of the lead derivation, for example, since aVR = −½(I + II), if there is ST depression in Leads I and II, then by definition, there must be ST elevation in avR. Simple mathematical considerations also show that ST elevation in aVL is reflected in ST depression in III and aVF so in a way these "reciprocal" changes in limb leads are "automatic" changes.
While the text states that "upsloping ST-segment depression > 1 mm at the J-point in the precordial leads" can be a sign of "significant left anterior descending artery (LAD) occlusion," this pattern has not been included in table 2 of the FUDMI. Upsloping ST depression is commonly seen during exercise stress tests, and there is controversy regarding its significance. While initially it was considered a nonspecific pattern induced by tachycardia and not specific for ischemia, more recent studies have suggested F I G U R E 2 Anterior STEMI with ST elevation in I, aVL, V 2 -V 6 without reciprocal ST depression that upsloping ST depression can be a true indicator of ischemia (Polizos & Ellestad, 2006;Rijneke, Ascoop, & Talmon, 1980) (de Winter et al., 2019de Winter, Adams, Verouden, & de Winter, 2016). Upsloping ST depression with tall T waves in the anterior leads was described as an ECG sign of proximal LAD occlusion in patients presenting with chest pain (de Winter, Verouden, Wellens, & Wilde, 2008;Verouden et al., 2009). More recently, this pattern was described in 11 patients (0.2%) of 5,588 with suspected acute coronary syndromes whose ECG was transmitted by the field triage team. All of them had a culprit lesion in the proximal LAD (de Winter et al., 2019;de Winter et al., 2016). However, there are anecdotal descriptions of a similar pattern of upsloping ST depression with tall T waves in left circumflex ischemia (Alam, Nikus, Fiol, Bayes de Luna, & Birnbaum, 2019;Birnbaum, Wilson, et al., 2014;Misumida, Kobayashi, Schweitzer, & Kanei, 2015) and even right coronary artery ischemia (Tsutsumi & Tsukahara, 2018).
Thus, we think that the description be changed to "significant coronary artery occlusion," rather than LAD occlusion based on the current literature.
It should also be noted that normal limits of ST elevation are race-dependent (Macfarlane et al., 2014). African and Chinese males, for example, have higher normal limits of ST elevation compared with Caucasians particularly in precordial leads. We think that allowance therefore has to be made for interpretation of ST shift particularly in Africans and Chinese.

| APPLI C ATI ON OF SUPPLEMENTAL ELEC TRO C ARD I OG R AM LE ADS
The FUDMI recommends the use of "posterior leads at the fifth in- There is specific recommendation for recording these leads "in patients with high clinical suspicion of acute circumflex occlusion (e.g., initial ECG nondiagnostic or ST-segment depression in leads Cardiac MRI-ECG correlation suggested that it is correlated with the projection of the vector of inferior ischemia on the anterior-posterior plan (Jia et al., 2018).
In the acute phase of STEMI, the reciprocal changes of ST elevation with positive T waves are ST depression with negative T waves.
Therefore, in the acute stages of inferolateral STEMI with ST depression in V 1 -V 3 , the T waves are usually negative. Only after reperfusion, or with a more advanced stage of infarction, do the T waves become positive (reciprocal changes of ST elevation with negative T waves) (Porter et al., 1998).

| CONDUC TION DIS TURBAN CE S AND PACEMAKER S
The FUDMI states "In patients with LBBB, ST-segment elevation ≥ 1 mm concordant with the QRS complex in any lead may be an indicator of acute myocardial ischemia." Yet, the sensitivity of this sign for STEMI physiology (acute occlusion of an epicardial artery) is low. We suggest that it be stated that since detection of ischemia by the ECG in LBBB is difficult, decisions concerning urgent reperfusion therapy should be based mainly on symptoms and hemodynamic parameters. According to the 2013 ACCF/AHA STEMI guidelines "New or presumed new LBBB has been considered a STEMI equivalent." Most cases of LBBB at time of presentation, however, are "not known to be old" because a prior ECG is not available for comparison.
New or presumed new LBBB at presentation occurs infrequently, may interfere with ST-elevation analysis, and should not be considered diagnostic of acute MI in isolation" (Jain et al., 2011;O'Gara et al., 2013). The European guidelines, published in 2017, specify "In the presence of LBBB, the ECG diagnosis of acute myocardial infarction is difficult but often possible if marked ST-segment abnormalities are present. Somewhat complex algorithms have been offered to assist the diagnosis, but they do not provide diagnostic certainty.
The presence of concordant ST-segment elevation (i.e. in leads with positive QRS deflections) appears to be one of the best indicators of ongoing MI with an occluded infarct artery. On the other hand, ST depression (usually concordant) in V 1 -V 3 in LBBB is very specific but not overly sensitive (Sgarbossa et al., 1996). Patients with a clinical suspicion of ongoing myocardial ischemia and LBBB should be managed in a way similar to STEMI patients, regardless of whether the LBBB is previously known. It is important to remark that the presence of a (presumed) new LBBB does not predict an MI per se" (Ibanez et al., 2018).
The FUDMI states "New, or presumed new, RBBB without associated ST-segment or T wave changes is associated with thrombolysis in myocardial infarction (TIMI) 0-2 flow in as many as 66% of patients (compared with > 90% in those with ST-segment or T wave changes)." The 2017 ESC guidelines for STEMI also specify "Patients with myocardial infarction and RBBB have a poor prognosis. It may be difficult to detect transmural ischemia in patients with chest pain and RBBB. Therefore, a primary percutaneous coronary intervention strategy (emergent coronary angiography and percutaneous coronary intervention if indicated) should be considered when persistent ischemic symptoms occur in the presence of RBBB" (Ibanez et al., 2018). These recommendations are mainly based on a retrospective study by Widimsky et al. (2012). These authors   (Neumann et al., 2019). In our experience, diagnosing ST elevation in the inferior and lateral leads can easily be done in patients with RBBB. The problem, as discussed earlier, is how to diagnose inferolateral STEMI equivalent in patients with RBBB and baseline ST depression in V 1 -V 3 and whether the threshold for ST elevation in the anterior leads V 1 -V 3 should be reduced.

| Reperfused STEMI
Although the FUDMI does not directly deal with indications for acute reperfusion therapy, the entity of (spontaneously) reperfused STEMI is not mentioned in the documents nor in the STEMI guidelines (Ibanez et al., 2018;O'Gara et al., 2013). Patients with spontaneous reperfusion at presentation (improvement in symptoms and ST-elevation resolution compared with a previous ECG) have a good prognosis without primary percutaneous coronary intervention Dowdy et al., 2004). As they do not have ongoing ischemia with progression of necrosis, coronary revascularization is indicated to prevent re-ischemia/ re-infarction, rather than salvaging myocardium. This can be done urgently, as in high-risk NSTEMI, rather than by using the time frame of primary percutaneous cor- and persistent ST-segment elevation" (Ibanez et al., 2018;O'Gara et al., 2013;Steg et al., 2012). This statement does not emphasize "ongoing" symptoms and therefore includes patients in whom symptoms resolved but who continued to have ST elevation above the relevant threshold. We believe that thrombolytic therapy should not be administered to patients with spontaneous reperfusion despite having residual ST elevation, as the current recommendation states: "If timely primary PCI cannot be performed after STEMI diagnosis, fibrinolytic therapy is recommended within 12 hr of symptom onset in patients without contraindications" (Ibanez et al., 2018;O'Gara et al., 2013;Steg et al., 2012).

| CON CLUS ION
In conclusion, although the ECG has been used for the diagnosis and triage of patients with suspected myocardial infarction for many decades, many of the concepts and terminology currently used should be updated, as in other fields of cardiology and imaging. As new information is available, based on high-sensitive biomarkers, angiography, and especially cardiac magnetic resonance images, our understanding of the significance of the various ECG patterns continues to be refined. We recommend that these new concepts should be further studied and given due consideration for incorporation in any future guidelines.