Thursday, August 30, 2018

Total eclipse of the heart

Written by Pendell Meyers

A male in his late 50s had sudden chest pain radiating to the left side and back, with nausea and diaphoresis as well as bilateral finger tingling sensation. He drove himself to the Emergency Department.

Here is his ECG on arrival at 11:33 am (unclear whether pain was persistent at this time, or if it had decreased since onset):
What do you think?








There is a small amount of STE in II, III, and aVF, but there is not clear STD or T-wave inversion in aVL. There is STD in V2 and V3 which must be assumed to be abnormal unless proven otherwise by a baseline ECG. The T-waves in V3 and V4 are not definitively hyperacute, but could be hyperacute if a baseline ECG showed smaller baseline T-waves. If this were true, then there would be ST depression and hyperacute T-waves, which constitutes de Winter morphology and would be concerning for occlusion or near-occlusion of the LAD in this distribution. If the T-waves are not hyperacute compared to baseline, then we are left with STD maximal in V2-V3 which would be concerning for posterior OMI. But it simply doesn't look like posterior OMI. 

Apparently his pain resolved. The cath lab was not activated at that time.

Approximately 40 minutes later, he had sudden recurrence of crushing chest pain and appeared critically ill.

Here is his ECG at that time, 12:08:

What do you see?



Red lines show the J-points in all leads. There is sinus bradycardia with massive ST segment deviations. STE is present in V3-V6, I, aVL, II, and aVF. STD is present in lead III, aVR, and V1. This morphology in the anterolateral leads is known as "shark fin" ST elevation and is created by QRS distortion plus slurring of the J-point and ST segment which obscure the perception of STE. 
The STE in V3-V6, I, and aVL implies a very large vessel which supplies anterior and lateral walls. The STD in V1 indicates posterior wall involvement. The ST segment in V2 is awkwardly isoelectric, pulled down by posterior involvement but pulled up by the massive STE of anterior OMI which is obvious from lead V3 onward. These findings alone would indicate occlusion of a very large vessel that supplies territories of both the LAD and LCX put together. And yet the STE extends even further, extending into the inferior leads as well. 

Notice what is happening in lead aVR: Due to widespread massive STE in essentially all other leads, aVR registers massive reciprocal STD.

If any ECG on this blog has ever shown evidence of actual, complete left main occlusion, this is it.

Regardless of the speculation about where the lesion is, this enormous territory OMI pattern with new bradycardia portends very high risk of imminent cardiac arrest. 

Another ECG was recorded 1 minute later:

This shows diffuse "shark fin" ST Elevation.



30 seconds later, the patient suffered cardiac arrest. Apparently it was alternating between VF and "PEA" with persistent enormous ST segment deviations when the shocks were briefly successful.

ROSC could not be consistently maintained.

Cardiology was not willing to take the patient to the cath lab in this condition.

Cardiothoracic surgery was called to bedside for initiation of ECMO. However, the surgeon elected to take the patient to the OR with ongoing mechanical CPR for emergent LVAD placement (this case did not happen recently, as ECMO would be the preferred strategy in most centers today).

Amazingly, LVAD placement was successful and systemic blood flow was restored mechanically.

He went immediately from the OR to the cath lab for emergent angiography.

The angiogram revealed:

"...a saddle thrombus in the distal LMCA involving the distal LAD and LCX with significant disease in the OM and also in the RCA."

The left main, LAD, and LCX were all 95% obstructed by the saddle thrombus at the time of cath. It is reasonable to conclude that the left main was fully occluded or almost fully occluded at the time of the second ECG above.

"Several passes were done with the thrombectomy aspiration catheter with improvement in blood flow. POBA was done to OM, LMCA/LAD, and LMCA/LCX. After discussing the case with Dr. _____  it was decided that we could not use at this time any type of stents because the patient was not a candidate for antiplatelet therapy (no aspirin or P2Y12 inhibitor). Therefore suboptimal results were obtained but TIMI flow 3 obtained in all vessels. A final revascularization approach will be decided in the future according to his status."


Troponin I peaked at 269.05 ng/mL (extremely large MI).

Here is an ECG after these interventions:

Extremely low voltage, which likely has to do with his LVAD being in place, as well as very small amount of remaining myocardium to conduct the action potential. ST segments resolved.


He had a meaningful baseline condition and family continued to pursue aggressive therapies. He had a very stormy course over the next month in the CTICU and ultimately passed away despite maximal therapies.



If none of our other posts have convinced you that STE in aVR does not represent left main occlusion, let this example convince you!



Learning Points:

A patient with true left main occlusion will usually suffer cardiac arrest very soon thereafter.  With some exceptions, only in a case like this, where the patient develops left main occlusion in front of providers, can an ECG be performed soon enough before cardiac arrest to show the ECG findings of left main occlusion.

ST elevation in lead aVR does not mean that the patient has left main occlusion. This case shows exactly the opposite: aVR in this case shows massive ST depression because it reflects the massive ST elevation in other leads. In contrast, lead aVR shows reciprocal elevation in the case of diffuse subendocardial ischemia from left main stenosis with diffuse STD and reciprocal STE in aVR.

See these other cases explaining lead aVR and left main ACS:

ST-Elevation in aVR with diffuse ST-Depression: An ECG pattern that you must know and understand!





Deep and widespread ST depression signifies high risk coronary lesion




-----------------------------------------------------------
Comment by KEN GRAUER, MD (9/1/2018):
-----------------------------------------------------------
Superb step-by-step discussion by Pendell Meyers, that convincingly illustrates the ECG findings of acute LMain occlusion (which as per Dr. Meyers, do not include ST elevation in aVR)! I have nothing to add to his words of supreme wisdom regarding deduction of the principal “culprit” artery in this case. Instead, I limit my comments to the initial ECG — which for clarity, I duplicate in Figure-1.
Figure-1: Initial ECG on the man in his 50s, who presented to the ED with CP (chest pain) occurring earlier. As per Dr. Meyers — it was unclear whether the patient’s CP had decreased or resolved at the time this tracing was done.
While obvious from the 2nd ECG (obtained 40 minutes later) that extensive acute OMI was in progress — it would be EASY to overlook the subtle abnormalities present in the initial tracing shown in Figure-1.
  • Dr. Meyers has noted the small amount of ST elevation in each of the 3 inferior leads, which also manifest small q waves. While tempting to attribute these findings to the relatively vertical frontal plane axis, and to “early repolarization” (especially given the absence of reciprocal ST-T depression in lead aVL) — the burden of proof is on us to rule out (rather than to rule in) an acute cardiac event, given that this patient presented to the ED with new-onset CP. In my opinion, NO definitive conclusion can be made regarding the possibility of an acute event from the appearance of the limb leads alone in Figure-1.
  • The main point made by Dr. Meyers that I wish to emphasize — is that the ST segments in leads V2 and V3 of Figure-1 are not normal — and, in association with the slight ST elevation seen in each of the 3 inferior leads — this finding must (as per Dr. Meyers) be assumed abnormal until proven otherwise. In general, there should be slight ST elevation in leads V2 and V3 of a normal ECG. While difficult to determine the precise “amount” of ST depression in these leads in Figure-1 — the shape of these ST segments “looks” depressed, and at the least, clearly lacks the usual slight upward-concavity ST elevation.
Clearly, the subtle findings on this initial ECG in Figure-1 do not portend or predict the distribution of dramatic ST-T wave deviations that soon after occurred. Whether they are the result of the associated severe RCA disease that was seen on cath, or were merely abnormal findings with little change from previous tracings is unknown. But, in addition to the insightful deductive process by Dr. Meyers for determining the presence of acute LMain occlusion — an equally important Learning Point in this case is appreciation of the subtle ECG abnormalities seen in Figure-1, that in the context of new-onset chest pain are clearly not “normal”.


Thursday, August 16, 2018

"Unstable Angina still exists", even in the Age of High Sensitivity Troponin

This case comes from a long term blog reader from Norway.

Case: 49 y.o male. No previous medical hx. 15-20 pack year smoking history. Parent with ACS in their 60's. No DM. No HTN, no dyslipidemia, normal weight. Brought to the emergency department due to chest pain. Patient stated that 20 hrs before presentation he experienced numbness of the left hand,  1-2 hours after which he developed retrosternal chest pain ongoing from 02:00 lasting until 08:00 with no pain free period. Went to bed at 08:00. Awoke 11:00 pain free. The day of presentation, the chest pain started again while doing light physical work.

Pain free upon admission after NTG, 300mg ASA and morphine 2.5 mg given prehospital.

Here is the first ECG 1

I (Smith) am excluding limb leads as they do not add much and take up too much space.
Note: this is recorded at 50 mm/sec, the way they do it in Norway.  
Below I have pasted the same ECG the way it would look at 25 mm/sec, which most of us are accustomed to:
Not very worrisome

"I could not see any ECG sign that had me particularly worried. I thought for a moment that the T wave in V3 seemed just a tad "fat"." 

"Physical Exam was unremakable. With a hx of ongoing chest pain for many hours with no pain free periods my concern for ischemia was dampened when the high sensitivity troponin T returned at 10 (ref value 0-14 ng/L)." 

Repeat ECG (2, not shown)  30 min after first ED ECG was also without change dynamic change.

"I admitted him for observation. He was put on telemetry. The second troponin drawn 6 hrs later also 10 ng/L.  He had no further episodes of CP.  Telemetry without arrhythmia/ST-T changes.
Next day, the patient was pain free." 

"Due to worrisome symptoms, I went forth with stress test." 

ECG 3 was recorded just before the stress test and was not significantly different.

Stress test dynamic ECGs follow:

You need to press the play, but there is no arrow.
There is no change during this short timespan.


This ECG video starts 4 minutes into the stress test.


Pay attention to leads V3 and V4.  
The entire video is 90 seconds, but you can speed ahead.
High ST Elevation develops.
The patient had no symptoms during this ischemia

This ECG was recorded immediately at the end of the stress test:


This was at 30 seconds after:


This was recorded at 1 minute:


This was recorded at 2 minutes into recovery:

This was at 10-15 minutes after end of stress test:


"The presenting ECG was without any obvious (at least to me) ischemic findings.  The hs TnT was normal, even at 6 hours.  But the stress test had impressive ST-elevation with rapid resolution.  It is uncertain why he had no symptoms during the stress test; perhaps due to short duration of demand ischemia?"

"I cannot however explain ongoing symptoms for many hours prior to admission with normal high sensitivity troponin.  Maybe enough blood flow as to not result in irreversible damage to myocytes?"

"Since the findings during stress test quickly resolved during rest, I felt quite confident angiography would reveal severe fixed proximal LAD stenosis, but I was not particularly worried about ruptured plaque." 

Comment: I (Smith) would be worried about ruptured plaque because the patient had presented to the ED with chest pain that was induced only by mild exertion and had also been present at rest.

"The patient was transferred to PCI-capable hospital same day. Troponin T after transfer was 16 ng/L. I thought they would find proximal LAD stenosis. They found severe mid-LAD stenosis distal to D2. The patient did well. Maximum trop was 69 ng/L after stent placement." 

Here are the angiogram images:



You can see the tight stenosis that is stented.  It is wide open after intervention.
Notice the stent itself before contrast is injected.

Comment

Here is a relevant paper: Sandoval Y.  Apple FS.  Smith SW.  High Sensitivity Cardiac Troponin Assays and Unstable Angina.  European Heart Journal: Acute Cardiovascular Care 7(2):120-128; March 2018.  Online published online on July 7, 2016 as doi:10.1177/2048872616658591
--> High sensitivity troponin: All of the studies of high sensitivity troponin (hs-cTn), whether of hs-cTnI or of hs-cTnT, have false negatives.  While negative predictive values for many combinations of either initial hs-cTn below the level of detection, or below some other very low value, or of 0 and -1, or -2, or -3 hour serial protocols, can exceed 99%, the sensitivity for MI rarely exceeds this threshold.  Moreover, with any combination of hs-cTn values, neither the NPV nor sensitivity exceeds 99% for 30-day adverse events (which includes unstable angina, MI within 30 days, coronary intervention, and more).

This is why we must always put high sensitivity troponin results into context.  Those with red flags (very worrisome history) must be taken seriously even when their hs-cTn protocol is negative.  The only protocols which reach greater than 99% sensitivity for 30-day adverse events use an accelerated diagnostic protocol which requires use of a risk score such as HEART or EDACS, in combination with a non-ischemic ECG and a single very low hs cTn or serial cTn below various thresholds and with a minimal increase.

All protocols also rely on adequate follow-up.

Furthermore, stress testing may still be useful in selected patients.   There are patients whom you are worried about and even if all the objective criteria are negative, they may still have ACS and it may be revealed with a stress test.  It is true that you must do them judiciously so that you don't have too many false positives.

Here is the European Society of Cardiology Troponin T 0/1 hour rule out MI algorithm (Circulation 137(23):2536.  This is for rule out MI, not for rule out 30-day adverse events
In this case, there were 2 hs-cTnT that were 6 hours apart and both were 10 ng/L.  That is even more reliable than a 0/1 hour protocol.
You can see the NPV for MI is 99.8%, and sensitivity 99.3%.

But the sensitivity for ACS or for 30-day adverse events is not that high.
  
Mokhtari et al. found this protocol to be only 87% sensitive for ACS!   However, in an extended protocol in which they only allowed patients without high risk history and with non-ischemic ECGs to be considered "negative," the sensitivity was 97.5% for ACS.  
Mokhtari A. et al.  A 1-h combination algorithm allows fast rule-out and rule-in of major adverse  cardiac events.  JACC 2016;67:1531-40.

The only 0/1 hour protocol that looked at 30-day MACE. N = 1038.
0/1 hour protocol alone (same as other 0/1 hour hs-TnT protocols)
–Rules out 65% (n = 682)
87% sens for ACS if one includes unstable angina as a 30 day MACE
0/1 hour protocol + ECG non-ischemic + Gestalt not high risk
–Gestalt risk = very low, low, or intermediate (not high risk)
–Rules out 60% (n = 625)
97.5% (92.9-99.5) sensitivity, 99.5% NPV (98.6-99.9) for 30-day MACE

Here is another study:

Thelin et. al.  Eur Ht J, Acute Card Care 2014; 4(5):403-9
Initial hs-cTnT in 478 patients presenting w chest pain as primary symptom
160 (33.5%) had values below 5 ng/L 
–remember: NPV 99.5% for MI in large meta-analysis few slides back
NPV of 100% for NSTEMI (70 NSTEMI; 37 UA)
NPV of only 94% for any ACS (includes Unstable Angina)
Sensitivity 91% (80-95)
–Missed 10 of 107 ACS as diagnosed by angiography; 8 PCI
Clinical judgment and pre-test probability of ACS is critical

[From powerpoint:
"High_Sensitivity_Troponin_Smith_with_hidden_slides" in chest pain lecture files/Troponin/]

Learning Points

1. You cannot rely only on hs-cTn to rule out 30-day adverse events.  You must also consider high risk presentation and ECG
2. There is still value to further testing after negative troponins, with careful patient selection.



Here is another post on this topic:

Unstable Angina: Dr. Braunwald asks if it is time for a Requiem



A Simplified Formula Discriminating Subtle Anterior Wall Myocardial Infarction from Normal Variant ST-Segment Elevation

Dr. Emre Aslanger has published a simpler formula for differentiating electrocardiographically subtle LAD occlusion from normal variant ST elevation.

Here is the paper:
Aslanger E et al.  A Simplified Formula Discriminating Subtle Anterior Wall Myocardial Infarction from Normal Variant ST-Segment Elevation. American Journal of Cardiology.  https://doi.org/10.1016/j.amjcard.2018.06.053
https://www.sciencedirect.com/science/article/pii/S0002914918314206

Here is the simplified 4-variable formula:
(R-wave amplitude in lead V4 + QRS amplitude in V2) – (QT interval in millimeters + STE60 in V3)

This eliminates the QT correction and eliminates the constants.

This new practical formula had an excellent area-under curve of 0.963 (95% confidence interval, 0.946 to 0.980, p<0 .001="" nbsp="" p="">
At a cutpoint of 12, it had a sensitivity, specificity and diagnostic accuracy of 86.9%, 92.3%, and 90.1%, respectively.

My more complicated 4-variable formula was this:
(1.062 x STE at 60 ms after the J-point in V3 in mm) + (0.052 x computerized QTc) - (0.151 x QRSV2) - (0.268 x R-wave Amplitude in V4 in mm).

My formula is based on this paper:  

A value greater than 18.2 is quite sensitive and specific for LAD occlusion. 

This 4-variable formula has an associated iPhone app: 

Wednesday, August 15, 2018

A Middle-Aged Man with Blunt Trauma and Hemopericardium

A Middle-Aged man had a single vehicle motor vehicle collision with significant energy.

He was hypotensive upon arrival.

A bedside ultrasound was done immediately.  Here are 3 clips.








There is pericardial fluid with echogenic material, diagnostic of hemopericardium with thrombus.


An ECG was recorded:
What do you think?














Sinus tachycardia.  There are Q-waves in II, III, aVF, and V3-V6.  There is ST elevation in the same leads.  The T-waves are inverted in V4-V6, and are beginning to invert in II, III, aVF.    This is diagnostic of subacute MI.

What do you conclude?


















This explains everything.  There was subacute MI with myocardial rupture and hemopericardium that resulted in the patient becoming hypotensive and losing control of the vehicle.

The initial troponin returned at 74 ng/mL.  This is very high and cannot occur acutely.   In other words, this cannot be due to trauma.   The ECG cannot occur with trauma either (Q-waves, ST Elevation, T-wave inversion).  This confirms your diagnosis of non-traumatic myocardial rupture.

The patient was taken to the operating room where he could not be saved.  

You might think that death is inevitable from myocardial rupture, but it is not!

In 1994, from our institution (Hennepin), Plummer et al. published this case series: 

Emergency department two-dimensional echocardiography in the diagnosis of nontraumatic cardiac rupture.  23(6):1333-1342; June 1994.


In it, Dr. Plummer describes 6 cases of an apparent acute STEMI who presented with chest pain (3), SOB and weakness, and one with profound hypotension.   Rather than immediately receiving reperfusion therapy (which, at the time, was tPA), all underwent bedside ultrasound which uncovered hemopericardium.  One presented very hypotensive, 4 with Systolic BP 80-90, and one with SBP of 140.  All ECGs showed subacute MI.  All went to the operating room emergently.  2 survived neurologically intact.


Non-traumatic Myocardial Rupture

Differential of peri-infarct pericardial fluid
The differential includes 1) pericarditis with effusion or 2) hemopericardium.
1) Pericarditis with effusion
   a) If 3 weeks after MI, then Dressler's syndrome (Dressler's syndrome is also known as post-myocardial infarction syndrome, post-cardiac injury syndrome and postpericardiotomy syndrome--see this case), which is a late post-MI autoimmune pericarditis occurring about 3-4 weeks after the MI.  Dressler's syndrome appears to be quite rare, according to Shahar and Lichstein.  
   b) Nonspecific pericarditis
2) Hemopericardium would be due to myocardial rupture, which could be due to:
   a) Rupture of a coronary artery due to PCI or
   b) Free wall Myocardial rupture (see below, next paragraph). 


Also, not all rupture is of the free wall:

Large Transmural STEMI with Myocardial "Rupture" of Ventricular Septum


Myocardial rupture is not uncommon. It is found on 1% to 3.5% of autopsies of patients who died of MI. It is associated with transmural MI; since most STEMI are aborted with reperfusion therapy, it is not as common as it once was. It is more common in women, and in patients who have a first MI and have a good EF, as it requires a pump force from the healthy myocardium to produce high pressure which ruptures the infarcted myocardium. The "rupture" is not an explosion, rather a small tract through the myocardium which leaks blood into the pericardium, and kills by tamponade.
.
Myocardial rupture is usually preceded by postinfarction regional pericarditis (PIRP). PIRP is indicated on the ECG by 2 findings: 1) persistenly positive (upright) T-waves at 48 hours, or 2) premature reversal of inverted T-waves to positive deflection by 48 to 72 hours after STEMI. In contrast to re-occlusion of the infarct-related artery, this reversal should be gradual. There should be QS-waves indicative of completed transmural MI.
.
Patients who present with chest pain or cardiac arrest and have an ECG diagnostic of STEMI could have myocardial rupture. Obviously, administration of heparin and/or lytics is hazardous. These patients may survive. In a report of 6 cases at our institution (Hennepin County Medical Center), 2 survived with cardiac surgery. 5 of 6 presented with chest pain and an ECG indicating reperfusion therapy, but were detected by bedside ultrasound.


For more information, and several cases, see chapter 28 of Smith's "The ECG in Acute MI," starting on page 273, at this link.


For more cases related to myocardial rupture, go to this link.



Recommended Resources