Secondary electrocardiographic stratification of NSTEMI to identify an acutely occluded culprit artery

In the United States, approximately 720 000 adults will experience a myocardial infarction (MI) every year. The 12-lead electrocardiogram (ECG) is quintessential for the classification of a MI. About 30% of all MIs exhibit ST-segment elevation on the 12-lead ECG and is therefore classified as an ST-Elevation Myocardial Infarction (STEMI), which is treated emergently with percutaneous coronary intervention to restore blood flow. However, in the remaining 70% of MIs, the 12-lead ECG lacks ST-segment elevation and instead exhibits a motley of changes, including ST-segment depression, T-wave inversion, or, in up to 20% of patients, have no changes whatsoever; as such, these MIs are classified as a Non-ST Elevation Myocardial Infarction (NSTEMI). Of this larger classification of MIs, 33% of NSTEMI actually have an occlusion of the culprit artery consistent with a Type I MI . This is a serious clinical problem because NSTEMI with an occluded culprit artery have similar myocardial damage like STEMI and are more likely to suffer from adverse outcomes compared to NSTEMI without an occluded culprit artery. In this review article, we review the extant literature on NSTEMI with an occluded culprit artery. Afterward, we generate and discuss hypotheses for the absence of ST-segment elevation on the 12-lead ECG: (1) transient occlusion (2) collateral blood flow and chronically occluded artery and (3) ECG-silent myocardial regions. Lastly, we describe and define novel ECG features that are associated with an occluded culprit artery in NSTEMI which include T-wave morphology abnormalities and novel markers of ventricular repolarization heterogeneity.


Introduction
Annually, 720 000 American adults in the Untied States will experience a myocardial infarction (MI), which causes myocardial injury and increases the risk for adverse outcomes (Thygesen et al 2018, Tsao et al 2022. While there are multiple mechanisms that are known to cause myocardial injury and MI, the most profound and potentially lethal mechanism is coronary plaque rupture and erosion with coronary thrombus formation (type I MI). If not treated timely, a type I MI will starve the myocardium of necessary oxygen, causing irreversible necrosis (Thygesen et al 2018). Subsequently, the contractility and electrogenesis of the myocardium become altered, increasing the risk for adverse outcomes such as cardiogenic shock, recurrent MI, heart failure, and death (McManus et al 2011, Jenča et al 2021, Tsao et al 2022. Past studies estimate the annual incidence of adverse outcomes after a type I MI to be between 2% and 10% (McManus et al 2011, Jenča et al 2021). Given the time dependency for the treatment, the common phrase 'Time is Muscle', proposed by Eugene Braunwald in the 1970s, is used to remind clinicians that the longer the coronary artery is occluded the greater the amount of necrosis and the higher the risk of adverse outcomes (Maroko et al 1971).
Based on theprecedent of timely care, current clinical guidelines in the United States (proposed by the American Heart Association/American College of Cardiology) and Europe (European Society of Cardiology) call for the use of the electrocardiogram (ECG) to stratify a suspected MI (Amsterdam et al 2014, Ibanez et al 2016, Levine et al 2016, Collet et al 2021. According to current American and European guidelines, if STsegment elevation meeting pre-specified criteria is observed on the 12-lead ECG, the MI is stratified as a STelevation MI (STEMI), and it is treated emergently in less than 2 h with coronary angiography and possible percutaneous coronary intervention (PCI) to restore blood flow (Ibanez et al 2016, Levine et al 2016. In contrast, if ST-segment elevation is not observed on the 12-lead ECG but MI is still highly likely based on elevated and dynamic troponin, the MI is stratified as a non-ST elevation myocardial infarction (NSTEMI) (Chang et al 2012, Amsterdam et al 2014, Collet et al 2021. In cases of NSTEMI, findings on the 12-lead ECG are non-specific and include ST-segment depression, T-wave inversion, or, in up to 20% of patients, no remarkable ST-segment/T-wave changes (Turnipseed et al 2009, Chang et al 2012, Amsterdam et al 2014, Collet et al 2021. To determine timing to coronary angiography with possible PCI, current guidelines advise risk stratification based on clinical stability. In the most recent 2020 guidelines published by the European Society for Cardiology, NSTEMI may undergo coronary angiography with possible PCI within a 24 h window. Current European guidelines advocate that very-high risk NSTEMI patients, defined as those with hemodynamic instability, cardiogenic shock, recurrent and refractory angina, life-threatening arrythmias, mechanical complications of MI, acute heart failure related to MI, and ST-segment depression >1 millimeter in at least 6 leads plus STsegment elevation in leads aVR and/or V1 suggestive of advanced coronary artery disease receive immediate (<2 h) coronary angiography with possible PCI. In the same guidelines, high risk NSTEMI patients, defined as those with established NSTEMI diagnosis, dynamic new or presumably new contiguous ST-segment/T-wave changes, resuscitated cardiac arrest without ST-segment elevation or cardiogenic shock, and a GRACE score >140 received early invasive intervention within 24 h. Lastly, the guidelines state that those with no recurrence of symptoms and none of the very high-or high-risk criteria are to be considered at low risk of short-term acute ischemic events and therefore received a selective invasive approach.
However, an important caveat to risk stratification is that a significant proportion of all NSTEMI patients are found to have an acutely occluded artery during coronary angiography with possible PCI (Amsterdam et al 2014, Khan et al 2017, Hung et al 2018, Collet et al 2021. Herein, an occluded culprit artery refers to a type 1 MI with occlusive or thrombus in the coronary artery resulting in no or very limited distal flow (TIMI flow 0-1). While this type of occlusion may present as a STEMI, systematic reviews and meta-analyzes published by Hung et al 2018 and Khan et al 2017 suggest 33% of NSTEMI patients have an occluded artery and risk stratification with the GRACE score, for example, is insensitive to differentiating NSTEMI with and without an occluded artery. Moreover, NSTEMI with an occluded artery have similar myocardial damage as STEMI and have nearly double the odds of recurrent MI, cardiogenic shock, and death compared to NSTEMI patients without an occluded artery (Khan et al 2017, Hung et al 2018. Potential reasons for the lack of ST-segment elevation on the ECG include the transient nature of occlusion, collateral blood flow masquerading advanced two-or three-vessel disease and/or an unknown chronically occluded artery, as well as an occlusion in a region of the heart with poorer sensitivity on the ECG. Unfortunately, these areas of research are not well-developed, and must be prior to the development of a stratification system which aims to differentiate NSTEMI based on the presence or absence of an occluded artery as advocated for in recent articles (Smith 2001, Aslanger et al 2021, Tziakas et al 2021. The objective of this review article is to discuss the extant literature on NSTEMI with an occluded culprit artery to grapple with the current understanding of NSTEMI with an occluded culprit artery and generate new research questions to promote the development of improved stratification methods for the detection and recognition of an occluded artery. First, we review type I MI compared to types II-V as well as the origins of STEMI/NSTEMI classification. Following, we discuss the extant literature on NSTEMI and with an occluded artery. Afterwards, we discuss three hypotheses for the absence of ST-segment elevation on the 12-lead ECG despite an occluded culprit artery: (1) transient occlusion (2) collateral blood flow and (3) ECG-silent myocardial regions, and also describe novel ECG features that are associated with an occluded culprit artery in NSTEMI, which include T-wave morphology abnormalities and novel markers of ventricular repolarization heterogeneity.

Defining types of MI
To best bring together advances in diagnostic tools and understanding of the many underlying mechanisms of myocardial injury, a consortium, including the European Society of Cardiology, the American College of Cardiology Foundation, the American Heart Association, and the World Heart Federation, published the 4th universal Definition of MI to propose a classification system based strictly on etiology of MI. Most broadly, myocardial injury is defined by the 4th universal Definition of MI as at least one high-sensitivity troponin value >99th percentile upper reference limit derived from a normal reference population. As such, myocardial injury is a broad diagnostic category, under which multiple possible mechanisms are considered including MI. If myocardial injury is acute, represented as a significant elevation in high-sensitivity troponin concentration over serial measurements, and associated with symptoms of myocardial ischemia, signs of ischemia on the ECG (STsegment changes or the development of pathological Q waves), or evidence of a new regional wall motion abnormality, the diagnosis of acute MI is made. If myocardial injury is acute as defined, but without signs of ischemia on the ECG or evidence of a new regional wall motion abnormality, the diagnosis of acute myocardial injury is made. This is often observed in acute heart failure or myocarditis. If myocardial injury is chronic, in which high-sensitivity troponin concentrations are stable or change minimally over serial measurement, then this is consistent with chronic myocardial injury which is most prevalent in structural heart disease and also non-cardiac disease such as chronic kidney disease.
The sub-classifications of MI are dependent on the suspected pathophysiology. Type 1 MI, as prefaced in the introduction, is attributable to atherothrombotic plaque rupture or erosion. Of note, this may be occlusive or non-occlusive to the coronary artery and is the main pathophysiological mechanism of concern for STEMI/ NSTEMI. This has ben supported by evidence published Baro et al (2019) and Wang et al (2021) have demonstrated that regardless of STEMI or NSTEMI classification, type I MI causes considerably greater myocardial damage measured using Troponin T and microRNA, respectively. In contrast, Type 2 MI occurs to an acute imbalance in myocardial oxygen supply and demand without atherothrombosis. This imbalance may be attributable to reduced myocardial perfusion due to coronary atherosclerosis without plaque disruption, coronary artery spasm, microvascular dysfunction, coronary embolism, coronary dissection, or other systemic causes including hypoxemia, anemia, hypotension, or bradyarrhythmia. The imbalance may also be attributable to increased myocardial oxygen demand secondary to tachyarrhythmia or severe hypertension. While the 4th universal Definition of MI also identifies types 3 (sudden cardiac death without evaluation of high-sensitivity troponin) and types 4-5 (related to coronary procedural events), these classifications are not the focus of this review.
Importantly, the 4th universal Definition of MI state that ECG findings have to be integrated with the aim of classifying type 1 MI into STEMI or NSTEMI in order to establish the appropriate treatment according to current guidelines.

Origins of the STEMI/NSTEMI classification
The STEMI/NSTEMI stratification system is based on experimental and observational evidence which linked coronary artery occlusion, myocardial necrosis, and ST-segment elevation on ECG. Early animal research demonstrated that coronary artery occlusion led to myocardial injury and necrosis, a strong propagator for arrhythmia, pump failure, and death (Maroko et al 1971, Muller et al 1975, Radvany et al 1975, Irvin and Cobb 1977. Radvany, Maroko, and Braunwald (1975) reported that myocardial necrosis almost perfectly correlated with ST-segment elevation among 11 anesthetized open chest dogs with occlusion of the left anterior descending coronary artery or one of its major branches for 20 min, and ST-segment elevation resolved with restoration of coronary blood flow (Radvany et al 1975). Using an animal model, Irvin and Cobb (1977) reported that there was moderate correlation between the degree of ST-segment elevation and myocyte injury subsequent to arterial occlusion (Irvin and Cobb 1977). During the same period, Reimer et al (1977), describes 'wavefront phenomena' where among dogs with coronary occlusions, subepicardial ischemia and infarction occur but with viable myocardium present (Lee et al 2008). Among humans, this was recently confirmed by Pessah et al (2022) who reported significant differences in ST-segment deviations with balloon occlusion of major coronary arteries most notably the left anterior descending, left circumflex, and right coronary artery (Pessah et al 2022). This experimental evidence provided a strong physiological basis for using ST-segment elevation as a surrogate for coronary artery occlusion, which was later extrapolated upon in a number of trials.
In the 1980s and 1990s, a number of randomized controlled trials compared fibrinolytics with placebo using ST-segment elevation as surrogate for myocardial necrosis but with significant oversite in clinical situations. These large scale randomized controlled fibrinolytics trails reported an impressed reduction in short-term mortality but failed to appropriately identify those who had most benefitted from therapy (Schröder et al 1987, Trialists 1994, Tebbe et al 1998. The fibrinolytic therapy trialists' (FTT) meta-analysis pooled data from 58 600 patients who were enrolled in 9 randomized clinical trials comparing thrombolytics to placebo for the treatment of MI (Trialists 1994). However, not all studies included in FTT performed angiography so it is difficult to discern if patients enrolled had a true coronary occlusion (Trialists 1994). Rather, these trials randomized highrisk patients based on the presentation of acute chest pain and poorly defined ECG findings (Trialists 1994). In fact, only 4 of the 9 trials included randomized pre-defined ST-segment elevation, each with varying cutoffs and methods of measurement (Trialists 1994). For example, the comparison trial of saruplase and streptokinase (COMPASS) trial and the intravenous streptokinase in acute myocardial infarction (ISAM) trial specified 0.1 mV in 2 frontal plane leads or >0.2 mV in 2 precordial leads (Schröder et al 1987, Tebbe et al 1998. Moreover, these trials did not include other important ECG findings such as hyperacute T waves, axis changes, and pathologic Q waves which are associated with occlusion of a coronary artery (Menown et al 2000). While the FTT meta-analysis reported a 3% absolute reduction in short-term mortality, specifically among patients with STEMI, no reduction in mortality among presenting among NSTEMI was reported (Trialists 1994). Based on this data, it was concluded that ST-segment elevation was a surrogate for significant myocardial necrosis due to acute coronary occlusion and subsequently greater risk for mortality; therefore, greater emphasis on reperfusion therapies was devoted to STEMI (Schröder et al 1987, Trialists 1994, Tebbe et al 1998. The current STEMI criteria cutoffs were established in the early 2000s based on data collected from the FTT meta-analysis (Schröder et al 1987, Trialists 1994, Tebbe et al 1998, Menown et al 2000, Levine et al 2016. Under the current guidelines, ST-segment elevation for a STEMI is defined as a cutoff of >2 millimeters of ST-segment elevation in at least one of the anteroseptal leads, or 1 mm in any 2 consecutive leads (Menown et al 2000, Levine et al 2016. Menown et al (2000), using data from the data from the FTT meta-analysis, reported this criterion correctly classified 83% of patients as acute MI or non-acute MI with a 56% sensitivity and 94% specificity (Trialists 1994). However, in this study, MI was detected solely based on biomarkers (e.g. creatinine kinase-myocardial band) and not angiography. Therefore, the true presence and also degree of coronary occlusion of the culprit artery remained unknown (Trialists 1994). Thus, ST-segment elevation can only be considered a surrogate for an occluded culprit artery.

Shortcomings of the STEMI/NSTEMI stratification
Its known the STEMI/NSTEMI stratification approach lacks the adequate sensitivity in detecting an occluded culprit artery (Writing Committee Members et al 2021). A large meta-analysis (1989) with a pooled sample size of more than 24 000 patients reported the overall sensitivity and specificity of dynamic ST-segment changes on the 12-lead ECG to be 68% and 77%, respectively (Gianrossi et al 1989). The recently completed DIagnostic accuracy oF electrocardiogram for acute coronary OCClUsion resuLTing in MI (DIFOCCULT Study) reported that the true sensitivity of STEMI criteria to be lower at only 54.4% (Aslanger et al 2020). The degree of STsegment elevation relates to the degree of coronary occlusion, and the true sensitivity varies between 23% and 100% (Faramand et al 2021). Since ST-segment elevation lacks the adequate sensitivity to detect an occluded culprit artery, a considerable amount of NSTEMI are found to have a occluded culprit artery similar to STEMI during delayed angiography ( (2017), NSTEMI patients with an occluded coronary artery were treated with coronary angiography with PCI on average, 22 h later than STEMI patients (Khan et al 2017). In regard to outcomes, NSTEMI patients with occluded coronary artery were at nearly double the risk of adverse outcomes compared to NSTEMI without an occluded artery (Khan et al 2017, Hung et al 2018. In the following sections, we will discuss the study population characteristics, ECG features, and rates of coronary angiography with PCI and outcomes of patients presenting as NSTEMI but with an occluded culprit artery. Patients with an occluded culprit artery may benefit from an emergent invasive approach; yet the absence of high-risk features and lack of characteristic ECG patterns (e.g. ST-segment elevation) excludes them from such an approach, per guidelines. Thus, there is some discussion, given the high rate of occluded culprit artery in this high-risk population, to perform early angiography.

Sample characteristics
Patients with NSTEMI and an occluded culprit artery tend to be 62-65 years of age, and male. Among 7 studies, including a total of 40 777 patients, the mean age of NSTEMI with an occluded culprit artery was 62 years compared to the mean age of NSTEMI without an occluded culprit artery being 65 years (Khan et al 2017). These results are similar to the other published meta-analysis (Hung et al 2018). Of note, these characteristics are very similar to NSTEMI in general; the recently published CREDO registries (n = 3254) reported a mean age of 69.8 ± 11.6 years for patients with NSTEMI (Takeji et al 2021a, 2021b). The distribution of gender was also comparable across all studies with the average percentage of males across all studies in the meta-analysis being 70% male in the NSTEMI with an occluded culprit artery compared to 68% (Khan et al 2017, Hung et al 2018. Similarly, these results are reproducible in large NSTEMI registries with 72%-74% being male (Takeji et al 2021a(Takeji et al , 2021b.

Comorbidity burden
In terms of comorbidities, there does not appear to be significant differences between NSTEMI patients with and without occluded artery. Based on recent data, the most prevalent cardiac comorbidities affecting NSTEMI include diabetes mellitus, hypertension, and hyperlipidemia (Tsao et al 2022). In the systematic review published by Khan et al (2017), the reported ranges of diabetes mellitus were 22%-41% among those with a occluded artery and 20%-34% without an occluded artery; hypertension was 49%-81% among those with an occluded artery and 53%-83% without an occluded artery; and the prevalence of hyperlipidemia ranged from 15%-54.6% among those with a occluded artery and 15%-50% without a occluded artery (Khan et al 2017). Thus the prevalence of comorbidities appear to be similar irrespective of an occluded artery. Khan et al (2017)

Severity and infarct size
Previous reports have documented that NTEMI with an occluded culprit artery have a significantly larger and more severe infarct compared to NSTEMI without a occluded culprit artery. Hwang et al (2018), Meyers et al (2021a), and Baro et al (2019) have reported that NSTEMI patients with an occluded artery exhibited a higher peak creatine kinase-MB, troponin I, and troponin T value more similar to STEMI than NSTEMI without an occluded culprit artery. Furthermore, left ventricular ejection fraction is more likely compromised among NSTEMI patients with an occluded artery compared to NSTEMI without an occluded artery (

ECG changes
While previous large trials have not reported consistent ECG changes associated with an occluded culprit artery among patients with NSTEMI, some signature features are worth noting. Warren et al (2015) reported that among 1957 patients from the Platelet IIb/IIIa Antagonism for the Reduction of Acute Coronary Syndrome Events in a Global Organization Network trial, an increased incidence of ST-segment depression in NSTEMI patients with an occluded culprit artery compared to NSTEMI patients without an occluded artery (Warren et al 2015).  Hung et al (2018) reported that the pooled odds ratio for infarction of the posterior or lateral wall among NSTEMI with an occluded artery was 2.24 (95% 1.63-3.09) and the left circumflex artery was more likely  to be the culprit artery (pooled odds ratio 1.65%, 95% 1.15-2.37) when compared to NSTEMI without an occluded artery (Hung et al 2018). A very large study including 131 729 patients from the Polish National Registry reported that the most common culprit artery was the left circumflex artery among patients with a NSTEMI and an occluded culpriter (35.86% NSTEMI with an occluded culprit artery versus 14.09% NSTEMI without an occluded culprit artery) (Terlecki et al 2021). Interestingly, occlusion of the left circumflex artery was highly associated with a normal admission ECG when compared to infarction of other culprit arteries (Moustafa et al 2016).

Patient outcomes
Patient outcomes are poorer among NSTEMI patients with occluded artery when compared to NSTEMI patients without occluded artery. Khan et al (2017) reported that the risk of major adverse cardiac events as well as all-cause mortality was significantly higher among NSTEMI with occluded artery when compared to NSTEMI without occluded artery (Khan et al 2017). Similar results were reported by Hung et al (2018). In terms of major adverse cardiac outcomes, Terlecki et al (2021) reported that among 131 729 patients (16 209 with NSTEMI and an occluded culprit artery), NSTEMI patients with an occluded culprit artery were more likely than both NSTEMI without an occluded culprit artery as well as STEMI to suffer from cardiac arrest during angiography (Terlecki et al 2021). Ayad et al (2021) expanded upon this finding by showing that the incidence of in-hospital arrhythmia was considerably higher among NSTEMI with an occluded culprit artery compared to NSTEMI without a an occluded culprit artery (Ayad et al 2021). Given the connection between left circumflex artery occlusion and NSTEMI with an occluded culprit artery, one previous study reported that NSTEMI patients with an occluded culprit left circumflex artery were more likely to suffer cardiac arrest and even higher mortality compared to STEMI (Chyrchel et al 2022). However, when evaluating long term outcomes, Kos et al (2021), with a median follow-up of 4.7 years, reported that the presence of a an occluded culprit artery, regardless of STEMI or NSTEMI presentation, was not associated with a higher long-term mortality; however, NSTEMI was associated with a higher mortality rate compared with STEMI, independent of angiographic presentation (Kos et al 2021). This may be due to providers ability to quickly identify STEMI due to the signature features on the 12-lead ECG, but may also be related to greater comorbidity burden in NSTEMI patients. Similarly, Fernando et al (2021) reported that baseline comorbidities, not occlusion of a culprit artery, predicted long-term mortality (Fernando et al 2021). This is an important finding which is similar to other registry reports that the long-term outcomes of NSTEMI patients are worse than STEMI due to poor comorbidity management (Gouda et al 2021, Takeji et al 2021aTakeji et al , 2021b.

Novel ECG features of total occlusion in NSTEMI
In this section, we will discuss two potential novel ECG markers may be able to help clinicians detect occluded artery in NSTEMI on the standard resting 12-lead ECG. While there is a movement to replace STEMI/NSTEMI stratification with one solely focused on occlusive/non-occlusive MI, we will focus on ECG markers which may complement STEMI/NSTEMI stratification (Smith 2001, Aslanger et al 2021.

T-wave morphology heterogeneity
The T-wave morphology reflects ventricular repolarization dispersion, and may be helpful in detecting an Hypotheses for lack of ST-segment elevation in NSTEMI with an occluded culprit artery While novel ECG findings may be more sensitive for detecting arterial occlusion even in the absence of STsegment elevation, understanding the mechanisms or factors for why ST-segment elevation is missing remain underexamined. In this section, we discuss three hypotheses for why ST-segment elevation may be absent despite occlusion of an occluded culprit artery in NSTEMI.

Transient occlusion
Coronary occlusions can be dynamic so one possibility that ST-segment elevation is absent on the 12-lead ECG is due to spontaneous resolution of the occlusive thrombus in the infarct related artery (  repeat 12-lead ECGs to capture this dynamic process, a lack of ST-segment elevation on the admission ECG can lead to under triage and delays in repeat ECG , Frisch et al 2020. Previously published evidence has demonstrated that continuous 12-lead ECG among MI patients can capture dynamic fluctuations in coronary artery patency based on ST-segment changes, however this is not routinely used in practice (Drew et al 1996, 1998, Pelter et al 2003. In replacement of continuous ECG, guidelines recommend seriel 12-lead ECGs though previous papers have not reported on this. Pelter, Carey, and Drew (2003) reported that transient ischemia measured on continuous 12-lead ECG was an independent predictor of worsening outcomes among patients presenting with NSTEMI (Pelter et al 2003). More recently, Pelter et al (2016) reported that transient myocardial ischemic events predicted worsening outcomes and actually proceeded symptoms by almost 3 h, suggesting continuous ECG may be able to more accurately capture the occlusive/non-occlusive nature of NSTEMI (Pelter et al 2016). Newer computerized algorithms may be superior in identifying transient ischemia which may prompt a transition toward use in hospital settings . While the hypothesis of spontaneous reperfusion is feasible for the continuum of arterial occlusion (Faramand et al 2021), it is more likely to affect NSTEMI with a sub-occluded artery (TIMI flow 2-3). As mentioned above, those with an occluded culprit artery present critically ill with greater myocardial damage, lower left ventricular ejection fraction, and larger infarct zone so it may be less likely that their coronary artery will spontaneous re-perfuse

Collateral blood flow
Another possible hypothesis to explain the lack of ST-segment elevation on the 12-lead ECG among NSTEMI patients is that the infarcted area is still receiving some blood supply through collateral blow flow (Khan et al 2017, Hung et al 2018, Tziakas et al 2021. By collaterals, we specifically mean blood vessels that form to provide an alternative source of blood supply when the original vessel fails to provide sufficient blood flow, due to severe atherosclerotic disease (e.g. non-occlusive 2-vessel or 3-vessel disease) or a chronically occluded artery. It is well known that timely enlargement of collateral blood flow can avoid transmural infarction, and this may explain why NSTEMI patients with an occluded culprit artery do not have the same degree of myocardial damage measured by cardiac biomarkers compared to STEMI (Figueras et al 2018, Terlecki et al 2021, Meyers et al 2021a. Figueras et al (2018) reported that the NSTEMI patients with occluded artery had more collaterals compared to NSTEMI without occluded artery as well as STEMI. Furthermore, NSTEMI patients with an occluded artery had greater myocardial damanage than NSTEMI without an occluded artery but less myocardial damage than STEMI (Figueras et al 2018). Similarly, Ayad et al (2021) showed that NSTEMI patients with an occluded artery also had significantly more collaterals (Ayad et al 2021). Bahrmann et al (2011) reported that NSTEMI patients with an occluded coronary artery and absent collaterals had significantly higher risk-adjusted 6 month major cardiac adverse events than those with an occluded culprit artery and visible collaterals, likely because the collaterals provided sufficient flow (Bahrmann et al 2011). Therefore, it is possible that patients who present as NSTEMI are found to have an occluded artery actually have a chronically occluded vessel and well-developed collateral flow. Chronic occlusions are relatively common affecting approximately 25% of NSTEMI patients, and, compared to STEMI, NSTEMI patients have more advanced disease affecting mostly 2-or 3-vessels, thus more likely to have a chronic occlusion (An et al 2021). It is important to recall that these NSTEMI patients with an occluded culprit artery may be either an acute Type I MI (elevated troponin plus one additional criteria as described above) and have a suspected infarct-related vessel; as mentioned, NSTEMI patients presenting with an occluded culprit artery are likely to present with a lower left ventricular ejection fraction and demonstrate wall motion abnormalities reflecting the insufficiency of collaterals, and be consistent with Tpye I MI (Bahrmann et al 2011, Khan et al 2017, Karwowski et al 2017b, Hung et al 2018, Hwang et al 2018, Fernando et al 2021. However, NSTEMI may also be a type II MI (elevated troponin in the absence of other criteria). As such, NSTEMI which would be defined as type II MI may be more prevelant in individuals with chronic occlusions because the collaterals are insufficient. The CASABLANCA (Catheter Sampled Blood Archive in Cardiovascular Diseases) study, which enrolled patients undergoing coronary angiographic procedures with or withour vascular intervention, reported that nearly 60% of patients with Type II MI still had 50% atheroscloetic obstruction in at least two vessels, and other studies have reported a bimodal distribution in which patients with type II MI either have no atherosclerotic obstruction or severe obstruction (Gaggin et al 2017, Sandoval andJaffe 2019). This suggests that severe disease is common among these patients, even among those whom do not fit the type I MI criteria (Gaggin et

ECG-silent myocardial regions
It is well documented that the 12-lead ECG has decreased sensitivity in the inferior, lateral, and posterior myocardial walls due to electrode lead placement. In particular, the lateral wall is often supplied by the left circumflex artery, which is differentially affected in NSTEMI with an occluded culprit artery (Khan et al 2017, Hung et al 2018. Thus, it is quite likely that NSTEMI patients with an occluded culprit artery may represent a subset of STEMI patients not recognized by the 12-lead ECG. Acute total occlusion of the left circumflex artery may result in isolated posterior infarction with ST elevation which can be detected in posterior leads V7-V9 (McCabe et al 2013). European Society of Cardiology and American Heart Association/American College of Cardiology guidelines recommend acquiring posterior leads V7 through V9 for suspected cases of acute left circmflex occlusion, such as in patients with typical symptoms and an initially nondiagnostic ECG or STsegment depression (STD) in leads V1-3 (Amsterdam et al 2014, Ibanez et al 2016, Levine et al 2016, Collet et al 2021.et al (2022) recently published about a synthesized 18-lead ECG which utilizes computerized software to create the additional V7-V9 leads. To test their software, they restropectively examined NSTEMI patients with left circumflex occlusion and reported a sensitivity and specificity of 46.7% and 95.6%, respectively, largely limited by small sample size (n = 16) (Horie et al 2022). It is important to note that Horie et al (2022) report that approximately 1-in-4 patients, whom would be candiates to have V7-V9 recorded, did not have V7-V9 recorded emphasizing the potential utility of synthesized leads.
Another potential solution is body surface mapping with 80-leads which has improve sensitivity in the inferior, lateral, and posterior myocardial walls. Daly et al (2019) tested the effectiveness of 80-lead body surface mapping to further evaluate NSTEMI patients and noted body surface mapping revealed significantly more patients with ST-segment elevation primarily in the posterior, lateral, and inferior regions (Daly et al 2019). In this study, the most common culprit artery was the left circumflex artery (Daly et al 2019). Of note, the sensitivity and specificity of 80-lead body surface mapping is 76% and 92%, respectively (12-lead ECG comparison 45% and 92%, respectively) (Owens et al 2008). While promising, there are a number of limitations for body surface mapping including cost, complexity, and necessary skill of interpretation.
Using solely the 12-lead ECG, some investigators have suggested so-called STEMI-equivalents which represent coronary occlusion without meeting the ST elevation criteria for STEMI. These markers include but not limited to: de-Winter ST/T-wave complex, N-wave, T-wave precordial instability, and posterior myocardial infarction (Wiśniewski et al 2019, Meyers et al 2021b. However, two studies which have applied these additional markers have reported mixed results with Wiśniewski et al (2019) reporting no significant benefit while Meyers et al (2021b) reporting significant benefit in the accuracy of detecting an occluded artery. However, given the complexity of their features, it may be difficult for clinicians to quickly recognize them on ECG; thus, further reliability testing and high-quality computer automation should be developed for rapid identification.

Discussion
In this review article, we reviewed the extant literature on NSTEMI with an occluded culprit artery to understand the current state of the science and generate new research questions to promote the development of improved stratification methods. In summary, nearly one-third of all NSTEMI patients are found to have an occluded culprit artery that receives delayed percutaneous coronary intervention (Khan et al 2017, Hung et al 2018. These patients receive delayed percutaneous coronary intervention because of the lack of ST-segment elevation on the ECG, but still have significant myocardial damage as indexed by biomarkers and echocardiography. Based on the review of the literature, patients presenting as NSTEMI with an occluded artery have significantly delayed time to angiography or percutaneous coronary intervention (per Khan et al (2017) mean for NSTEMI with total occlusion 31.3 h; mean for NSTEMI without total occlusion: 34.2 h). Subsequently, this prolongs the period during which the myocardium does not receive adequate perfusion and consequently causes a larger infarct and is responsible for the near double odds of adverse outcomes (Khan et al 2017, Hung et al 2018. To address ECG-based solutions to differentiate NSTEMI based on the presence or absence of an occluded artery, this review article also identified novel ECG markers that require computerized interpretation and further testing. In this article, we discussed T-wave morphology heterogeneity and markers of ventricular depolarization heterogeneity. While these are exciting findings to help differentiate NSTEMI patients as those with and without an occluded artery, it is hard to expect clinicians be able to recognize such novel and often subtle findings. As seen in table 1, these features are not computerized and implementing them into computerized ECG inteprteation will be necessary in order to classify NSTEMI based on the presence or absence of an occluded artery (Al-Zaiti et al 2022). To truly implement secondary screening of NSTEMI, we believe these findings must be incorporated into a computerized algorithms that can notify the physician that the algorithm suspects an occluded artery even in the absence of ST-segment elevation, similar to STEMI. This would be very important in the pre-hospital environment in which STEMI care is robust and well-developed. As such, it prompts the physician to treat the NSTEMI in a more time-dependent manner.
Before advancing this into existing care, further research and development is needed, mostl to understand the reproducibility and reliability of these novel ECG findings. We advise the following 3 overarching research questions include: (1)What other patient characteristics (e.g. bundle branch block, pacemaker, etc) interfere with the presence and measurement of novel ECG features derived from the QRS/ST-segment and T-wave?
(3)What are the implementation protocols and procedures of novel ECG features and/or equipment (e.g. 18lead ECG, 80-lead body surface mapping)?
We propose these research questions given the known confounders of the QRS complex, ST-segment, and T-wave, which most novel ECG findings derive from, as well as the limitations of these studies which have generated these results. First, it is well known that confounders such as bundle branch blocks and pacemakers significantly alter the QRS complex, ST-segment, and T-wave so the reliability of novel ECG findings for differentiating NSTEMI patients with and without an occluded artery with these confounders is necessary. Unfortunately many of the studies reviewed here excluded such confounders like bundle branch blocks (Niu et al 2013). This is particularly needed as bundle branch block has been associated with an occluded culprit artery that requires timely percutaneous coronary intervention (Ibanez et al 2016). Moreover, bundle branch blockers are prevelant among the NSTEMI population (Takeji et al 2021a(Takeji et al , 2021b. Another consideration is obesity which, despite being a risk factor for acute MI, can inadvertently affect the amplitude of ECG and make it more challenging to interpret ST-segment elevation (Lee et al 2008). Additionally, the reproducibility of these novel ECG features must be assessed. As noted, occlusions may be transient in nature yet most of the papers reviewed here only examined the 10 s, 12-lead ECG. While this is the diagnostic standard, it is a significant limitation and does not abide by current guidelines which recommend examination of contiuous or seriel ECGs.This is an important question to examine in order to establish the reproducibility across multiple patients with NSTEMI but also within the same patient who may have a transient occlusion.
The studies reviewed here were retrospective in nature and use ECGs from practice. As mentioned, most not use continuous ECG nor seriel ECGs to examine changes over time, an important consideration given the transient nature of occlusions. While pragmatic to use existing data, it is known that lead misplacement is frequent in clinical practice, and this may compromise the reliability of the novel features mentioned in this review.
Lastly, given the emerging relevance of advanced computerization such as computerized leads and machine learning, implementation protocols and clinical procedures are needed during initial clinical testing as well as clinical trials. We strongly believe this to be necessary so clinicians can appropriately use these technological innovations to patients' advantage to detect an occluded culprit artery. Beyond just ECG, other approaches such as coronary imaging using computed tomography may be used rapidly to determine the occlusiveness of the coronary arteries, and, therefore, identify the best treatment pathway (Kofoed et al 2018a(Kofoed et al , 2018b. The recently completed VERDICT trial (very early versus deferred invasive evaluation using computerized tomography) demonstrated that computed tomography angiography to be equivalent to invasive coronary angiography in differentiating obstructive and nonobstructive coronary artery disease in NSTEMI patients (Kofoed et al 2018a(Kofoed et al , 2018b.
In conclusion, this review explores secondary screening of NSTEMI based on the presence or absence of an occluded artery as a means to improve NSTEMI outcomes. Moreover, we emphasize on the 12-lead ECG and propose new avenues of future research needed to create valid and reliable classification of NSTEMI patients based on an occluded artery.

Conclusions
Nearly a third of NSTEMI patients have an occluded culprit artery but do not have the classic ECG markers like ST-segment elevetion to initiate emergent cardiac care. Given the success of STEMI care and outcomes, further research is needed for NSTEMI patients to develop ECG markers targeted towards an occluded artery. In particular, the addition of derived ECG features and synthesized ECG leads to extend the 12-lead ECG into the posterior and anterior regions is warranted.