60 year old with vomiting, diarrhea, and syncope: is this Wellens? Is this type 2 MI?

 Written by Jesse McLaren, with edits/comments by Smith and Grauer

 

A 60 year-old patient with diabetes and ESRD presented with 24 hours of vomiting, diarrhea, weakness and then a syncopal episode. Vitals: RR 18, sat 98%, HR 103, BP 124/71 and temp 38.0. Here’s their ECG: is this Wellens?

 

 

There’s borderline sinus tach, normal conduction, normal axis, and low voltages in the limb leads. The anterior leads have loss of R waves, mild convex ST segments and primary T wave inversion. In the context of QS waves, T wave inversion indicates old or subacute infarct, or reperfusion after significant infarction. Below is the old ECG:

 

 

This confirms the anterior changes are new--but is this from type 1 or type 2 OMI? The patient was diagnosed and treated for sepsis and DKA, and the only interpretation of the first ECG by the emergency physician was that it didn't meet STEMI criteria.

Smith: 

1) Some type 1 MI only present with atypical symptoms, especially in diabetics. Such atypical symptoms include vomiting, but not diarrhea.  

2)  Type 2 MI requires a supply demand mismatch.  Low supply: hypotension, hypoxia, anemia, dyshemoglobinemias.  High demand: tachycardia, hypertension, dilated LV (wall stretch).  This patient has no evidence of supply demand mismatch.

3) The ECG suggests a nearly completed LAD infarct due to the QS-waves in V2 and V3.

 

The first troponin returned at 68,000 ng/L, which was attributed to type 2 MI (though this is not a simple dichotomy, as Dr. Grauer explains in his comments below.) The patient was referred to cardiology, who found an anteroseptal wall hypokinesis on bedside echo. The patient had a prior stress echo with preserved EF but inducible ischemia in the LAD territory. As they noted, "Not for urgent cath in context of septic/DKA picture clouding assessment for type 1 MI, but definitely requires cath and medical treatment for ACS." So they initiated on dual antiplatelets and heparin with a plan for angiogram the next day after treatment for sepsis/DKA.

Smith: the extremely elevated troponin also in consistent with a subacute and completed (transmural) or nearly completed infarct.

 

Unfortunately, 12 hours after arrival, the patient had a VF arrest. Here’s their post-cardioversion ECG.

 

 

Anterior STEMI(+)OMI, so cath lab activated: the LAD had a 100% occlusion but this was chronic, with collaterals from an intact RCA and circumflex that had a 90% stenosis.

Smith: The anterior wall has clearly been supplied through collaterals from the RCA and circ since there is a chronic LAD occlusion.  The circ has a tight stenosis, but is open.  Thus the anterior wall is susceptible to supply demand mismatch, but (as above) we don't have any evidence of that.  Just because the RCA and circ are open now does not mean they were open 20 hours ago.  And the angiogram frequently does not identify a culprit.  So I suspect that one or both of these were occluded within the past 24 hours, resulting in a large anterior infarct.

There was an unsuccessful attempt to stent the chronic occlusion. Peak troponin was 95,000, EF was reduced to 45% with anterior wall motion abnormality, and below is the discharge ECG showing shallow anterior T wave inversion:

 

 

 

As this post explains, Wellens syndrome describes

1.     patient: anginal symptoms which have resolved

2.     ECG: primary reperfusion T wave inversion in LAD distribution but intact R waves, indicating reperfusion before significant infarction

3.     Troponin: mildly elevated

4.     Angiogram: critical lesion in LAD which is now open, or reperfused by collaterals, but is at high risk of-reocclusion

 

But in this case

1.     patient presented with sepsis vs anginal equivalent: sepsis can cause diffuse ST depression and reciprocal ST elevation in aVR (which is not a “STEMI equivalent” but a sign of diffuse subendocardial ischemia), but should not cause focal ECG changes that mimic OMI

2.     ECG showed primary T wave inversion in LAD distribution, but also loss of R waves indicating significant infarction

3.     Troponin was massively elevated

4.   Angiogram showed chronic total occlusion (CTO) of LAD and limited collateral circulation, which was compromised

As the EXPLORE trial explained, "Concurrent CTO lesions are found in 10% to 15% of patients with STEMI...Because of the procedural complexity and below-average success rate, PCI is attempted only in 10% of all CTO lesions, commonly in an elective setting...In patients with STEMI and concurrent CTO, we did not find an additional benefit for CTO PCI in terms of LVEF or LVEDV. However, a subgroup analysis suggests that patients with CTO in the LAD may benefit from early additional CTO PCI." (Henriques et al. Percutaneous Intervention for Concurrent Chronic Total Occlusions in Patients with STEMI: the EXPLORE trial. JACC 2016)

For patients with CTO, collateral circulation is crucial. As another study found: "In CTO lesions, the antegrade blood flow is completely interrupted, leaving the myocardium entirely dependent on collateral flow...If the collateral to the collateral-dependent myocardium of the CTO area originates directly or distally from the IRA, the myocardium could be endangered in case of blockage during STEMI. This could result in an increase in infarct size and a decrease in LVEF leading to a higher mortality. In our cohort there was a trend for higher mortality in patients with well-developed retrograde collaterals to the CTO, originating from the IRA, which were blocked during STEMI."(Elias et al. Impact of collateral circulation on survival in ST-segment elevation myocardial infarction patients undergoing primary percutaneous coronary intervention with a concomitant chronic total occlusion. JACC 2017)

While this describes collateral circulation compromised by acute thrombus (type 1 MI), it can also be compromised by supply-demand mismatch (type 2 MI). See this post on type 2 posterior OMI secondary to pneumonia and severe anemia: "An emergency formal echo showed an inferolateral wall motion abnormality.  In the record, an old angiogram reported a chronically occluded obtuse marginal (OM).  The previous echo was normal, but a stress echo had shown induced inferolateral hypokinesis.  Thus, there was prior proof that this area was vulnerable to stress; the territory of this artery was reliant on collateral circulation for oxygen delivery."

 

Take home

    1. Wellens syndrome describes LAD type 1 OMI with spontaneous reperfusion, but

         loss R waves, especially QS-waves, imply there has already been a large infarct

    2. spontaneous reperfusion can occur from the artery opening or recruitment of

        collaterals, i.e. reperfusion can occur with arteries that remain occluded (acutely or

         chronically) if there is sufficient collateral circulation - which is tenuous

    3. Patients with CTO rely on collaterals, which can be compromised by type 1 MI

     (thrombus) or type 2 MI (supply/demand mismatch): treatment is directed at

     restoring supply/demand mismatch, reperfusing acutely occluded coronary arteries

     +/- revascularizing CTO

====================================

MY Comment, by KEN GRAUER, MD (8/19/2022):

===================================

Today's case by Dr. McLaren prompts consideration of the relationship between Acute Infection and Myocardial Infarction. A recent article by Musher et al with this exact title reviews a series of fascinating features of this relationship (Musher et al — N Engl J Med 380:171-176, 2019 — with Free-Access summary of KEY Points by Hayek in ACC: Latest in Cardiology, 2019).

• Today's case by Dr. McLaren recounts the hospital admission of a patient with sepsis and DKA — with a complicated course including VFib arrest, massively elevated troponin and ultimately an anterior STEMI.
• Among the points brought out by the above NEJM Review — is the fact that the association between acute infections and acute MI has only been appreciated during the past few decades. A growing body of literature now documents the relationship between viral and bacterial infections such as influenza, pneumonia, urinary tract infections, septicemia, and many others. The finding that risk of acute MI is greatest at the onset of infection — and is proportional to infection severity — strengthens the premise of a cause-and-effect relationship.
• 
  
• Surprisingly — the NEJM Review authors contend that demand ischemia (ie, Type 2 Acute MI) should explain no more than a minority of infection-related MI events. Instead — they feel the cause of most infection-related MIs is acute coronary occlusion ( = Type 1 Acute MI).
• 
  
• Mechanisms proposed to explain how acute infection may result in acute coronary occlusion include (among others): i) Destabilization of existing atheromatous plaque by acute inflammatory factors (ie, cytokines, interleukins, tumor necrosis factor); ii) Increased thrombogenesis with platelet activation that is associated with acute infection; iii) Gene expression linked to platelet activation, endothelial dysfunction and hypercoagulability that may be promoted by certain viruses.
• Mechanisms proposed to explain how acute infection may result in Type 2 MI (from demand ischemia) include: i) Impaired coronary perfusion despite increased metabolic needs from acute infection (ie, compensatory tachycardia shortens ventricular filling time during diastole — thereby reducing coronary perfusion); ii) Toxin-mediated vasoconstriction from acute sepsis; iii) Hypoxemia with ventilation-perfusion mismatch (especially with severe respiratory infections); iv) A direct myocardial depressor effect from circulating toxins liberated by acute infection; v) Cytokine "storm" — with sudden release of these substances that may provoke a life-threatening systemic inflammatory syndrome leading to multi-organ failure.

The message from this NEJM Review is clear — A variety of mechanisms account for depressed ventricular function and potential for inducing either Type 1 or Type 2 acute MI as a direct result from almost any serious acute infection.

• We need to remain alert to this possibility!

A man in his 30s with chest pain and a normal bedside echo, without wall motion abnormality

Submitted and written by Dominic Nicacio MD, peer reviewed by Meyers, McLaren, Grauer, Smith


A man in his early thirties with history of congenital aortic stenosis status post mechanical valve placement presented to the ED with acute onset chest discomfort starting at approximately 1800 while walking. His chest discomfort radiated to the jaw and was associated with shortness of breath and dizziness. The pain is constant and unrelieved by rest. No recent fevers, cough, congestion, nausea, sweats, or abdominal pain. He admits that he has not been taking warfarin as directed for his mechanical valve.

Triage vitals: BP 152/57 temp 97.6 F HR 62 RR 18 O2 100%

Triage ECG (1.5 hours after pain onset):

Three old ECGs on file (unfortunately the clinical history surrounding these ECGs was unavailable):

Prior ECG#1 (4 years ago):

Prior ECG#2 (12 years ago):

Prior ECG#3 (also 12 years ago):

Meyers comments on the ECG:

The triage ECG (at the top of the post) shows sinus rhythm, LVH, and STE in II, III, aVF, V3-V5. There is reciprocal STD in I and aVL. The concern until proven otherwise is for OMI involving at least the inferior leads. Comparison to Prior ECG#1 confirms the concern for the inferior leads, but interestingly Prior ECG#3 appears to have some similar but less dramatic characteristics. Nonetheless, these findings must be considered OMI until proven otherwise.

Exam: appears comfortable, unremarkable except mechanical systolic murmur

Bedside US:
No pericardial fluid, no dissection flap in descending aorta
Grossly normal LV/RV systolic function
No signs of RV strain
No WMA seen in limited parasternal long and high quality parasternal short views (no view of the apex obtained)
Collapsible IVC

Chest x-ray:
Mild central vascular congestion
Median sternotomy wires
Mildly increased size of cardiac silhouette
No pneumothorax
Mediastinum appears to be within normal limits



Based on history and the ECGs above, the cath lab was activated within about 15 minutes of arrival.

Cardiac Catheterization:
100% stenosed distal/apical LAD lesion with TIMI 0 flow, likely coronary thromboembolism from mechanical aortic valve with INR = 1.0. Type 2 MI due to coronary thromboembolism. Successful revascularization with balloon angioplasty with restoration of TIMI-3 flow. No disease seen in other coronaries. Distal LAD revascularized with balloon angioplasty with restoration of TIMI-3 flow.

Initial troponin = 10 ng/L (99% URL = 20 for men for this assay)

 
The patient was continued on heparin with transition back to therapeutic warfarin.


Repeat ECG after cath:

OMI findings resolving.

Repeat ECG next day:

OMI findings further resolving, and reperfusion T wave inversion beginning in lead III.

No further ECGs were recorded.

Repeat troponins:

2160 ng/L

8095 ng/L

7572 ng/L

Echo:

The left ventricular cavity size is moderately to severely dilated. Wall thickness is normal. Systolic function is normal. The ejection fraction is 67%. No segmental wall motion abnormalities.

Hospital course:
He was continued on heparin drip and transitioned to warfarin. Transthoracic echo showed moderately to severely dilated LV cavity, EF 67%. Aortic valve with severe paravalvular leak with normally functioning mechanical heart valve. Transesophageal echo showed 2 mobile echodensities concerning for thrombus or vegetation seen in the left ventricular outflow track and ascending aorta. Blood cultures showed no growth after 4 days. He was discharged home in good condition.

Learning Points:

Depending on the operator and interpreter, the ECG may or may not be more sensitive than echo for OMI.  In order to be fully sensitive, the echo must be done on all segments!  

Beware: echo, and especially bedside echo, can have false negatives!!

Even formal bubble contrast echo can have false negatives for ACS that needs emergent therapy!!

See this: What do you think the echocardiogram shows in this case?

And this: Wait until after the ECG to give Nitroglycerine

This case

Notice that this OMI had initial troponin within normal limits at approximately 1.5 hours after chest pain onset.

Type 1 AMI involves plaque rupture and thrombus formation as the cause of coronary ischemia. Type 2 AMI excludes plaque rupture, and includes essentially any other mechanism of supply-demand mismatch causing AMI. This case of coronary embolism is considered type 2 AMI.

The myocytes cannot "distinguish" between different etiologies of acute coronary occlusion. Any condition causing acute coronary occlusion and immediate cessation of blood flow can cause the OMI ECG progression. This case would be classified as "type 2" OMI. 

Lead aVL almost always shows reciprocal STD and or TWI when there are OMI findings in the inferior leads.

The inferior leads show findings from more myocardial territory than simply the inferior wall of the left ventricle. We believe the LV apex, inferior septum, and parts of the right ventricle also produce ECG changes in the inferior leads, in general. In this case, I believe that the infarcting LV apex was the source of the STE in this case. We specifically looked at the inferior wall of the LV on parasternal short axis in real time using bedside ultrasound, and found it to be contracting normally without WMA. We did not obtain a good view of the apex. Interestingly, the formal echo also did not show any WMA.

LVH can mimic OMI and conversely make OMI more difficult to identify. STEMI criteria are defined in the absence of LVH, meaning STEMI has no definition or criteria in cases of LVH. In significant LVH with repolarization abnormalities, the appropriately discordant STD in lateral leads and STE in right sided anterior leads frequently causes alarm for ischemia and false positive cath lab activations. 

Comparing current ECGs with prior ECGs has important limitations! Do not assume that "prior" ECGs are "baseline" ECGs. You must also know the events during the prior ECG to be able to compare it to the current ECG. The three "prior" ECGs available in this case are each quite different - highlighting the need to understand their context in better detail that we have available for this case.

Regardless of the ECG, this patient had classic ACS symptoms, ongoing ischemic pain, and rising troponin. Even if we didn't have the ECGs in this case, we must remember that the diagnosis of OMI can be made and supported by many other clinical features.

===================================

Comment by KEN GRAUER, MD (6/17/2022):

===================================

Superb case by Drs. Nicacio and Meyers — which emphasizes a number of important clinical points that are worthy of repetition:

• I've reproduced the initial ECG in Figure-1. Although I was not immediately convinced by this ECG that an acute OMI was ongoing — strong indication for prompt cath was nevertheless present, because the patient's history was typical for ACS — and his ischemic-sounding chest pain was unrelieved.
• 
  
• As helpful as availability of prior tracings can be for determining if ECG findings are "new" or "old" — I found the 3 prior ECGs on record for this patient to be confusing. Each showed a slightly different ECG picture, including: i) Change in frontal plane axis; ii) Differing QRS amplitudes; and, iii) Variation in ST-T wave morphology showing previous ST elevation on the more remote tracings — but fairly deep inferior lead T wave inversion on the more recent ECG from 4 years earlier. As per Drs. Nicacio and Meyers — not knowing the clinical circumstances present at the time these prior tracings showing different ECG findings were obtained — made it difficult to know what (if anything) the difference between the initial ECG in today's case, compared to this patient's prior ECGs meant.
• 
  
• PEARL: Over the years, I found it helpful in patients with multiple ECGs in their charts — to quickly skim through a representative sampling of their prior tracings. This can be done surprisingly quickly once you familiarize yourself with "the pattern" for that particular patient. For example, on occasion I'd see a patient with known heart disease show diffuse T wave inversion — which normalized on their next tracing — only to alternate over time between a pattern that looks ischemic, and one that does not. Awareness that such variation over time in ST-T wave morphology may occur in a given patient (similar to that seen in comparing the 3 prior tracings in today's case) — adds perspective for how to interpret your comparison with the newest ECG.

The Initial ECG in Today's Case:

I found the initial ECG in Figure-1 both intriguing and puzzling. As per Drs. Nicacio and Meyers — the presence of LVH can make the diagnosis of OMI much more challenging. Predominant posterior forces (from an enlarged left ventricle) will often reduce anterior r wave amplitude (if not result in frank anterior QS complexes). In addition — marked LVH that manifests with deep anterior S waves will often present with anterior ST elevation (ie, the "mirror-image" of LV "strain", that in lateral chest leads manifests as ST-T wave depression — as per My Comment in the June 20, 2020 post in Dr. Smith's ECG Blog).

• Although today's patient was a young adult in his 30s — he had a long history of heart disease (ie, with congenital aortic stenosis requiring valve replacement). So it was no surprise that ECG criteria for LVH were easily satisfied with the finding of exceedingly deep chest lead S waves (ie, 33, 26 and 31 mm in leads V2, V3, V4) — with delay in transition until lead V6 because of such marked precordial lead S wave predominance.
• 
  
• The finding of most concern in ECG #1 was ST elevation in each of the inferior leads — with reciprocal changes in high lateral leads I and aVL. That said, in view of how marked the increase in QRS amplitudes from LVH was — I thought the amount of inferior lead ST elevation was relatively modest (and a bit less, but otherwise not overly different from the ST-T wave appearance in the limb leads on prior ECG #3).

The "shape" of the ST-T wave in each of the 6 chest leads of ECG #1 is different — such that I wasn't sure how to interpret these findings. 

• The T wave inversion in lead V1 is not necessarily abnormal. But I thought the horizontal "ledge" for the ST segment in lead V2 (before finishing in an independently peaked T wave) looked strange — and not at all typical for the anterior lead appearance of LV "strain".
• I thought the ST-T wave appearance in leads V3 and V4 to be consistent with LV "strain" — although why with similar S wave depth, the T wave peaking in lead V4 should be so much more marked than in lead V3 was uncertain to me.
• The ST segment "takeoff" in lead V5 then straightened — but given the modest J-point ST elevation in this lead with a 19 mm S wave, I was again uncertain about distinction between LV "strain" vs the ST elevation from OMI. 
• The ST segment then flattened in lead V6 — but I wondered if transition from predominant precordial S waves to the relatively small R wave in lead V6 might account for this.

Figure-1: The initial ECG in the ED.

I saw 2 additional ECG findings of interest in Figure-1:

• There is definite RAD (Right Axis Deviation) — as determined from the small, but entirely negative QRS complex in lead I. The fact that the P wave in lead I is positive — with global negativity (of P wave, QRS and T wave) in lead aVR — means that this is not LA-RA lead reversal. Instead, from a "pure" ECG interpretation perspective — the presence of RAD in association with marked LVH voltage should immediately suggest the possibility of RVH. Suspicion of RVH provided yet one more reason for my uncertainty in assessing the acuity of this initial ECG. (To note that subsequent Echo revealed no sign of RV strain.).
• P wave morphology is highly unusual. All limb lead P waves are of extremely low amplitude — with a tiny bizarrely notched P wave in lead II. Yet the P wave in lead V1 is huge — with a tall, pointed initial positive component. Is this a sinus rhythm? An ectopic atrial rhythm? Or — despite the lack of prolonged P wave duration — Does the subtle P wave notching in multiple leads reflect some type of intra-atrial conduction defect?

In conclusion — our THANKS to Drs. Nicacio and Meyers for presenting today's case. As interesting as I found the initial ECG to be — the KEY point in today's case is that identification of suspicious (albeit indefinite) ECG findings in this patient with typical unrelieved chest pain merited timely cath (which was done within 15 minutes of ED presentation) — and this confirmed the need for prompt revascularization.

Collapse, pulse present, ECG shows inferior OMI. Then there is loss of pulses with continued narrow complex on the monitor ("PEA arrest")

An elderly woman was witnessed to collapse.  911 was called and when EMS arrived, she was unresponsive with shallow respirations, a GCS of 3, pulse of 70 and BP of 78/67 by cuff pressure.

3 prehospital ECGs were recorded:

There is an obvious inferior OMI/STEMI, right?

The patient then had a PEA arrest while on the cardiac monitor, and CPR was started.

On arrival, all the usual things were done for cardiac arrest.

Transthoracic cardiac POCUS was of low quality, and so after intubation a TEE probe was inserted.

Aside: here is a recent report of our experience with over 550 TEE exams in the Hennpin ED: Feasibility, utility, and safety of fully incorporatingtransesophageal echocardiography into emergency medicinepractice

Here is the video:

You can see on the left the RV, which is very large, and on the right is the LV, which has a tiny chamber with excellent function.  This is all but diagnostic of severe acute RV failure.  

Sudden severe RV failure can be due to either: 

Inferior AND right ventricular infarction. 

But there will usually be some STE in V1 when V2 is isoelectric.

Or:

 Acute pulmonary embolism.

Assessment:

This patient is in extremis and is going to die unless something is done immediately, and probably even if the correct steps are taken.   

"PEA" arrest only means that the providers cannot feel a pulse, not that there is no cardiac activity.

As you can see from the ultrasound, there is cardiac activity, but extreme hypotension.  This is the situation in the vast majority of PEA arrest.  There is a pulse; you just cannot feel it.  But you can see activity on a high quality ultrasound.

ECG: 

Acute PE may present with ST Elevation mimicking type I OMI if coronary perfusion is extremely poor.

Acute OMI may present as PEA arrest on the monitor if there is additional severe RV infarct.  This would usually manifest as STE in V1 as long as there is no STD in V2 (and there is not).  Most OMI cardiac arrest are ventricular fibrillation; when they develop witnessed PEA arrest (on the monitor!), they usually manifest LV dysfunction, not RV dysfunction (unless, of course, they are huge RV MI also).

There is one other complicating factor: GCS of 3 when the patient still has a pulse and blood pressure.  This must bring the possibility of sudden intracranial hemorrhage into the differential.

However:

1. Patients who have cardiac arrest, or near arrest, due to intracranial hemorrhage do not survive. 

     a.  Intracranial hemorrhage that causes arrest is massive

     b. The combination of high intracranial pressure from the bleed and the low blood pressure from the arrest results in a cerebral perfusion pressure of zero, and brain death.

So the differential is: 

1. OMI with RV MI

2. Pulmonary embolism with type 2 OMI (severe supply demand mismatch); this is strongly supported by PEA and the cardiac ultrasound.

3. Intracranial hemorrhage.

The first 2 can be treated with thrombolytics.  

The latter will die with or without thrombolytics.  

So the benefit/risk ratio is to give thrombolytics.  The faster the better.

Outcome:

100 mg of tPA was given in a bolus but the patient did not survive.