Should We Be Looking for ‘Occlusion MI’ Rather Than STEMI?

Every emergency physician and interventional cardiologist knows that when a coronary artery is blocked without collateral circulation, speedy treatment is needed to restore blood flow. Looking for ST-segment elevation on the ECG, though, is an imprecise way to figure out which patients need to be rushed to the cardiac catheterization laboratory.

Some clinicians say that testing for occlusion myocardial infarction (OMI) instead might be a better tool for making that determination.

“We think we should get rid of the term ST-elevation MI, because it focuses on ST-elevation and makes you think that ST-elevation is all that matters—when what matters is acute coronary occlusion without collateral circulation leading to imminent myocardial necrosis or death,” Stephen Smith, MD (University of Minnesota and Hennepin County Medical Center, Minneapolis), told TCTMD.

Research has shown that more than half of patients with an occluded artery will not show ST-segment elevation and that can lead to delays in care and worse outcomes.

There are other ECG features beyond ST-segment elevation that also can be used to find the patients with an acute coronary occlusion who require an immediate trip to the cath lab, Smith said, pointing to components of the QRS complex and the T wave. Not many clinicians, however, can identify these features on their own, he added. “It takes a lot of expertise, a lot of time learning it, and I think it also takes a certain talent in seeing subtle waveform differences.”

Artificial intelligence (AI) appears to be a solution. “It can see these waveform differences, just like it can recognize a face. It can recognize OMI on an EKG,” Smith said. “The only way to implement this on a wide scale is to use AI.”

Smith is a shareholder in Powerful Medical, the company behind the PMcardio STEMI AI ECG model, also called the “Queen of Hearts.” The model has been shown to be more sensitive than STEMI criteria for identifying occluded vessels. In one study, the AI model outperformed STEMI criteria in terms of accuracy (90.9% vs 83.6%) and sensitivity (80.6% vs 32.5%), with similar specificity (93.7% vs 97.7%). And that was with the first version of the technology, Smith noted.

The Queen of Hearts model was designated a “breakthrough device” for the detection of STEMI and STEMI equivalents from the US Food and Drug Administration earlier this year.

What matters is acute coronary occlusion without collateral circulation leading to imminent myocardial necrosis or death. Stephen Smith

Jacqueline Tamis-Holland, MD (Cleveland Clinic, OH), an author of the updated comprehensive ACS guidelines from the American College of Cardiology and the American Heart Association released in February, didn’t advocate for getting rid of the term STEMI but told TCTMD the OMI concept is “very clinically relevant” in the historical context of taking STEMI patients directly to the cath lab and treating patients with non-ST-segment elevation MI less emergently.

In daily practice, physicians often see patients with NSTEMI who have an occluded vessel, Tamis-Holland said. “Officially, they were non-STEMI and we didn’t have to worry about the time to treatment for registry purposes and for documentation,” she said. “But in actuality, they were the equivalent of a STEMI because they had a total occlusion. So rather than defining it as STEMI and non-STEMI, it might be nice to see if we can recognize the cohort of patients who are at risk for an acute occlusion.”

That way, Tamis-Holland said, “perhaps we could serve the patients better by getting patients to the cath lab who have an acute occlusion and should be opened sooner.”

Time to Ditch STEMI?

Not all physicians are ready to move on from STEMI. Andrew Sharp, MD (University College Dublin and Mater Misericordiae Hospital, Dublin, Ireland), questioned the need for a new concept to guide decisions about which patients should be taken for an immediate cardiac catheterization.

“I used to run ECG courses teaching fellows and junior doctors ECGs 20 years ago, and we spent that time emphasizing that the degree of ST-elevation is not necessarily proportionate to the diagnosis, and that if you have ongoing chest pain with ECG changes, just because they don’t meet arbitrary criteria set down in a clinical trial for defining ST-elevation MI doesn’t mean that the vessel isn’t blocked,” he told TCTMD.

Even back in 2002, “we taught that they should go straight to the cath lab. They shouldn’t be thrombolized even when it was a therapy back then. Because of the uncertainty around the diagnosis, they should go to the cath lab, have an angiogram, and receive primary PCI,” Sharp said, noting that diagnostic angiography is a very low-risk procedure.

He acknowledged that the technology for interpreting ECGs hasn’t been very good and said it would be helpful to have better systems with true AI backing.

Nonetheless, Sharp said, “the vast majority of the time, the patient has ongoing cardiac chest pain. And really the thinking should stop there. That’s where a cardiologist should be consulted, and a cardiologist should make their mind up on whether the patient goes for an angiogram.”

He suggested that the use of AI to provide a deeper look at the ECG in search of an occluded artery might be helpful if it could be performed in an ambulance, allowing paramedics to decide whether a patient needs to go directly to a PCI-capable center rather than somewhere else.

“But I don’t see that as being distinct from ‘ECG changes plus ongoing pain equals heart attack center,’ because that’s how we did it in London 20 years ago. That’s how we’re doing it now in Dublin. And I think rebranding it is an educational project rather than a clinical project, because people should know that those patients should go for angiography, whether or not there’s a subtle ST change or not,” Sharp said.

“If somebody is saying this isn’t a heart attack because there’s 0.8 millimeters of ST-elevation and not 1, or there’s 1.5 millimeters of ST-elevation and not 2, and yet the patient has the symptoms, that’s a failure of decision-making and education,” he said.

If you have ongoing chest pain with ECG changes, just because they don’t meet arbitrary criteria set down in a clinical trial for defining ST-elevation MI doesn’t mean that the vessel isn’t blocked. Andrew Sharp

Smith, for his part, disputed that this is a matter that can be addressed by education alone. Although practice guidelines already state that NSTEMI patients with persistent chest pain should go immediately to the cath lab, it can take hours to come to that diagnosis of myocardial infarction, he said.

“The problem is that when someone comes in with chest pain, it might be their esophagus, it might be their chest wall, it might be something else altogether,” he said, pointing out that waiting for serial measurements of troponin to diagnose MI can take a lot of time, during which the myocardium is already dying.

Moreover, studies have shown that physicians often aren’t adhering to the guidelines around the management of high-risk patients with NSTEMI, creating unnecessary delays and worsening outcomes, Smith said. Some aren’t acting quickly enough because there are trials—including TIMACS—indicating that early intervention is no better than waiting in the absence of ST-segment elevation. But those trials are flawed, Smith said, noting that the time to early treatment ranged from 5 to 16 hours. At the upper end of that range, “the damage is already done,” he said.

In addition, those studies excluded patients with persistent pain because that group was already recommended by the guidelines to go to the cath lab. “Those are not the patients we’re so worried about,” Smith said. “We’re worried about patients who have active chest pain and an EKG that shows occlusion on it. Those patients are not being treated rapidly and they’re suffering irreparable harm from it.”

A Broader View

Ajay Kirtane, MD (NewYork-Presbyterian/Columbia University Irving Medical Center, New York, NY), said the fundamental premise behind the OMI concept is wanting to identify patients with occluded arteries because they generally have the most to gain from an urgent trip to the cath lab. “From a conceptual standpoint, if there was a better way to identify those patients, then it makes sense because the standard 12-lead EKG is a little bit limited,” he explained.

“Personally, though, I think from a practical standpoint, it’s actually even broader than an occluded artery,” he added. “What I would like to be able to find out with high reproducibility and fidelity is who are the patients that need to go to the cath lab and benefit from it versus who are the ones that don’t?”

For instance, Kirtane said, an NSTEMI patient with multivessel disease who doesn’t have an acute occlusion but who is in impending cardiogenic shock needs to go to the cath lab. He noted that there are other types of patients without blocked arteries who would benefit as well.

“I would say, let’s find an occluded artery. Let’s find super high-risk anatomy—so, left main and/or multivessel disease in conjunction with hemodynamic instability. Add those things all together so that you then can say, ‘Hey, yes, go,’” Kirtane said. “Because ultimately what you would like these things to do is help you with your decision-making. Should I take a patient? Should I not take a patient? And there’s more to it than simply: is the artery occluded or not?”

He noted that physicians should be doing this already—incorporating not just STEMI criteria, but also other factors like patient presentation, hemodynamics, and time from symptom onset into their decision-making.

There’s more to it than simply: is the artery occluded or not? Ajay Kirtane

Indeed, physicians generally are not basing their decisions only on ST-elevation, Tamis-Holland said. “We all look at other features. I have tons of patients who have 0.9 millimeters of ST-elevations, but they have clear reciprocal changes and come on, they don’t meet the criteria, but let’s face it, they’re a STEMI and you take them to the lab, right? We do that in clinical practice all the time.”

OMI in the Coming Years

Smith said to get the OMI concept off the ground, “the people who are stuck in their ways have got to either move aside or start to learn what’s important.” Then, more data will be needed to persuade physicians, he added. “As the months and years go on, we accumulate more and more data showing that this is really a great approach.”

The PMcardio STEMI AI ECG model has received CE Mark approval in Europe, and Smith said researchers in Aalst, Belgium, began using it in practice last year. Unpublished data indicate that the time to intervention for patients with acute coronary occlusions has dropped, with a reduction in mortality and improvement in LVEF.

Many physicians, though, won’t be convinced until there’s a randomized trial showing that assessing OMI rather than STEMI makes a difference in patient outcomes, Smith predicted, noting that there is a cluster-randomized trial underway in Turkey and others in the planning stages.

Smith predicts that in the next 2 to 5 years, the field will start to shift toward looking for OMI—or STEMI equivalents, the term the FDA uses—rather than STEMI alone. FDA approval of the AI model would be a huge step, he said.

Some centers are already essentially using the OMI approach, Smith said. “They just call it STEMI, even though they don’t . . . go by ST-elevation criteria. They will activate the cath lab and do the emergent intervention if they think the EKG shows an occlusion, even if there is not an ST-elevation.” A look at data from the Minneapolis Heart Institute and other institutions involved in the Midwest STEMI Consortium, for instance, indicated that only 55% of cases that led cath lab activations met the ST-elevation millimeter criteria.

“They activate the cath lab . . . and they have very good outcomes because they really do practice the OMI paradigm, but call it STEMI or STEMI equivalent,” Smith said. The name isn’t so important, he added, “as long as we do the right thing, which is get these patients to the cath lab early.”

We need to still focus on STEMI because the majority of those patients have an acute occlusion or a threatened closure about to occlude. Jacqueline Tamis-Holland

Tamis-Holland said “we need to still focus on STEMI because the majority of those patients have an acute occlusion or a threatened closure about to occlude, and we just need to add [the OMI concept] to the armamentarium.”

That will take time, however, to allow for more standardization in defining OMI and the accumulation of more supporting data, she said. “Until we’ve done more testing of this strategy, I don’t think it should be used as the sole way of assessing somebody.”

It’s likely that with time “clinically, this is what we’re all going to start doing and moving towards,” Tamis-Holland said.

As for how this might play out, Sharp suggested the OMI concept could be incorporated by simply stating that “the degree of ST-elevation may sometimes be absent despite the presence of vessel occlusion. In such circumstances, clinicians should rely on history-taking and the presence of a pattern of ST-depression to suggest at least ongoing ischemia. And in such circumstances, patients should go to the cath lab.”

This could potentially come as an addendum to the guidelines, he said. “There’ll be lots of papers written on the specificity and sensitivity of various ECG criteria,” he predicted. “And then, in the end, what we’re doing is we’re teaching the basics of clinical medicine all over again—but with an extra paragraph in the guideline document.”