Monday, December 9, 2019

CO poisoning. Routine ECG recorded before hyperbaric therapy....Are they related?

This patient was working on his truck in a garage without ventilation and the vehicle was running. He felt dizzy and lost consciousness. The next thing he remembers is waking up in the ambulance. He was last seen normal at 09:00 AM and was found down at 10:30 AM. He denied chest pain, abdominal pain, SOB or any other symptoms, though he did endorse chest pressure when asked specifically.

He underwent a routine ECG prior to hyperbaric therapy for CO poisoning:
Obvious Massive STEMI, Proximal LAD Occlusion

His CO level returned at 34%.

Further h/o revealed that he had had a proximal LAD stent placed for a Non-Occlusion MI more than one year prior, and had been on clopidogrel for one year but was no longer on it.

The cath lab was activated.

The interventionalist told the patient that his ECG might be the result of CO poisoning only.

I shook my head because this would be nearly impossible from CO poisoning only.  It might be possible with an extremely severely tight stenosis in the proximal LAD, but such a STEMI is never the result of CO poisoning alone.

It was clear to me that this would be a proximal LAD occlusion, possibly at the site of the previous stent.

It was decided to go the cath lab first, and then to the hyperbaric chamber for hyperbaric oxygen.

Angiogram: Culprit Lesion (s): Thrombotic, in-stent occlusion of the proximal LAD

24 hour echo:
Regional wall motion abnormality-distal septum anterior and apex, akinetic.
Regional wall motion abnormality-anterolateral, akinetic.
Regional wall motion abnormality-anteroseptal, akinetic.

After intervention, the patient went for multiple rounds of hyperbaric oxygen therapy in our world class hyperbaric chamber (see the images at this link).  The patient is on 100% oxygen at 2 atmospheres pressure, so that the pO2 is may be over 1000 mm Hg.

The Hennepin County Medical Center hyperbaric chamber construction was organized by Dr. Cher Adkinson (retired HCMC Emergency Physician) and now run by our current outstanding HBO Emergency Medicine team, including HBO director Chris Logue.

48 hour ECG, after artery is opened and CO level is resolved:
This shows QS-waves and some persistent STE.  This is a bad sign, and suggests no reflow (downstream platelet fibrin aggregates obstructing small vessel flow) future aneurysm formation.

Poor Microvascular Reperfusion ("No Reflow"): Best Diagnosed by ECG

Peak Abbott troponin I was 401.00 ng/mL (this is one of the highest troponins I have ever seen or heard of).

Another case of CO, with ST depression: What is the treatment for this subendocardial ischemia?

Carbon Monoxide displaces oxygen from hemoglobin, and thus effectively decreases oxygen saturation, in this case by 34%, down to 66%.  But it also binds with cytochrome oxydase to inhibit ATP formation.  Thus, its effect is the same as ischemia.

We consider cardiac ischemia (on the ECG, or by elevated troponin), by itself, to be an indication for emergent hyperbaric oxygen (HBO) even if there are no other indications such as neurologic disability, loss of consciousness, level greater than 40%, pregnancy, or other indications.

We treat many chronic conditions with hyperbaric oxygen, but are open 24/7/365 for emergencies including CO poisoning, air embolism, decompression sickness, and central retinal artery occlusion.

Dr. Adkinson's research here at HCMC, published in JAMA, found that myocardial injury (as indicated by troponin elevation, but also by ischemia on the ECG) is common in carbon monoxide poisoning and is independently associated with an increased risk of mortality at 7-year followup (38% vs. 15%).  A subsequent publication in JACC reported that myocardial injury was not at all predicted by CO level.

However, STEMI is rare in CO poisoning, with some case reports. All reports show coronary occlusion as the etiology.

So is it related to CO toxicity?

Some authors believe that CO toxicity increases platelet aggregation and thrombogenicity.  So it is possible that the CO toxicity resulted in the in-stent thrombosis and STEMI.

But the final common pathway for STEMI in CO toxicity is coronary occlusion, not the CO toxicity itself.  Such patients should go to the cath lab first, and to the hyperbaric chamber only after.


Late stent thrombosis secondary to carbon monoxide poisoning

A Case of Acute Carbon Monoxide Poisoning Resulting in an ST Elevation Myocardial Infarction

Ikeda H, et al. Free oxygen radicals contribute to platelet aggregation and cyclic flow variations in stenosed and endothelium-injured canine coronary arteries. J Am Coll Cardiol 1994; 24: 1749–1756.

Marius-Nunez AL. Myocardial infarction with normal coronary arteries after acute exposure to carbon monoxide. Chest 1990; 97: 491–494.  This ECG showed diffuse ischemia, but no focal ST elevation. 

MY Comment by KEN GRAUER, MD (12/9/2019):
This case poses the intriguing question of whether CO poisoning by itself can be the cause of this patient’s acute STEMI?
  • Of note — this patient had a proximal LAD stent placed for non-occlusive MI more than 1 year prior.
  • He presented to the ED with loss of consciousness from acute CO poisoning (initial CO level = 34%). Although the patient denied “chest pain” — he later acknowledged “chest pressure" ...

I feel close scrutiny of the 2 ECGs shown in this case is worthy of additional discussion.
  • For clarity — I’ve reproduced these 2 tracings in Figure-1.
Figure-1: The 2 ECGs shown in this case (See text).

As per Dr. Smith — ECG #1 (which was the initial ECG obtained in the ED, prior to sending this patient for hyperbaric therapy) — showed an obvious, extensive STEMI, almost certainly the result of acute proximal LAD occlusion.

  • Question #— What are all of the specific ECG findings that YOU see in ECG #1?
  • Question #2  Why is ECG #1 likely to represent an acute proximal LAD occlusion?
  • HINT to Question #2: If you don’t see 5 reasons why ECG #1 is likely to represent acute proximal LAD occlusion — GO BACK and LOOK AGAIN.

ANSWERS: The following ECG findings support the likelihood that there is acute proximal LAD occlusion in ECG #1:
  • Ssegment elevation begins as early as in lead V1. When there is mid- or distal- LAD occlusion — precordial lead ST elevation typically begins later (ie, in lead V2 or V3). Considering the relatively small amplitude of the QRS complex in lead V1 — there is a substantial amount of ST elevation in this first chest lead. ST elevation then becomes extreme already by lead V2.
  • There is marked reciprocal Sdepression in each of the inferior leads. Reciprocal changes in the inferior leads are much less likely to be seen when acute anterior STEMI is due to mid- or distal LAD occlusion.
  • There is substantial Selevation in lead aVL — which is a near “mirror-image opposite” of the ST depression seen in lead III. Substantial ST elevation in lead aVL is less likely to be seen when acute anterior STEMI is due to mid- or distal LAD occlusion. There is also ST elevation in lead I — though lead aVL is a much more reliable indicator for suspecting proximal LAD occlusion.
  • There is an incomplete RBBB pattern in ECG #1 — as manifest by the terminal R’ in lead V1, in association with a narrow terminal S wave in lead V6. Clearly, the QRS complex is not wide enough to qualify as complete RBBB — and there is no terminal S wave in lateral lead I. But normally, the QRS complex in lead V1 should be predominantly negative, and it most definitely is not in ECG #1.
  • There is LAHB — as manifest by predominantly negative QRS complexes in each of the inferior leads. In association with the incomplete RBBB that we have just described — this constitutes acute bifascicular block — and new RBBB (or incomplete RBBB+ LAHB, in association with acute anterior STEMI is a characteristic lesion of prognostic importance (!) when there is acute proximal LAD occlusion. This combination of conduction defects is less likely when LAD occlusion occurs at a more distal level.

To Finish Your Answer to Question #1: What additional ECG findings do YOU see in ECG #1 that have not yet been mentioned?
  • There are already Q waves in multiple leads in ECG #1. Although the Q wave in lead aVL is very small and narrow (and it could be a “normal” septal q wave) — in the setting of obvious extensive acute anterior STEMI with ST elevation in lead aVL — one has to assume that this Q wave in aVL is an “infarction Q wave”.
  • In addition — there are small-but-definitely-present Q waves in leads V1, V2 and even V3! The Q wave in lead V1 is the reason this incomplete RBBB pattern manifests a QR’ rather than an rSR’. And, even though the Q wave is small in lead V2 and tiny in lead V3 — there should be no Q wave at all in these leads given the incomplete RBBB pattern. These anterior Q waves are all “real” findings, and in association with the dramatic ST elevation, they portend what is soon to follow in ECG #2.
  • There is marked T-QRS-D (Terminal QRS Distortion) in leads V2 and V3. Morphologically, T-QRS-D is also present in lead V4 of ECG #1 — although we define T-QRS-D as the absence of both a J-wave and an S-wave in either lead V2 or V3, and not in V4 (CLICK HERE for illustration and full discussion of the concept of T-QRS-D in My Comment at the bottom of this 11/14/2019 post in Dr. Smith’s ECG Blog).

Now COMPARE ECG #1 to ECG #2: Dr. Smith summarized the ECG changes between these 2 tracings in the legend of the 2nd Figure in his discussion above. But to HONE your ECG interpretation skills — What are ALL of the specific ECG changes between these 2 tracings?

  • Although the amount of ST elevation in ECG #2 has dramatically decreased since ECG #1 was done — there still is 2-3 mm of ST elevation in leads V2-thru-V5, as well as a relatively large amount of residual ST elevation in leads I and aVL (especially considering the diminutive size of the QRS in these high lateral leads).
  • Similarly, although the amount of inferior lead reciprocal ST depression is decidedly less in ECG #2 — a worrisome amount of reciprocal ST-T wave change remains.
  • The small anterior Q waves we highlighted in ECG #1 — have evolved into large QS complexes in leads V1-thru-V3 of ECG #2. These are followed by substantial loss of R wave amplitude in lateral chest leads V4-thru-V6, as well as in high lateral leads I and aVL. Subtle Point: Despite loss of R wave amplitude in lead aVL — the Q wave in this lead has deepened and widened.
  • The leads which showed ST elevation in ECG #1 — now show deep T wave inversion. This is an evolutionary change.
  • The incomplete RBBB pattern seen in ECG #1 has resolved.
  • Criteria for LAHB are no longer met (ie, Considering the larger area under the R wave in lead II of ECG #2, compared to the smaller area within the skinny S wave in this lead — the frontal plane axis is now no morethan -20 degrees — or not enough to qualify as LAHB).
  • Bottom Line: As per Dr. Smith — the finding in ECG #2 of large QS complexes + loss of R wave amplitude this amount of persistent ST elevation at a point 48 hours after successful PCI is a very concerning prognostic sign.

Our THANKS to Dr. Smith for presenting this interesting case!

Thursday, December 5, 2019

A young F is hyperthermic, delirious, and dry: Fever-induced Brugada? Diphenhydramine toxicity? Tricyclic?

Case 1. 

A 30-something with no medical history presented with 1 day of high temperature and delirium.

Her temperature was 106 degrees.

As part of the workup, she underwent an ECG.

I walked into the room and saw this:
What do you think?

There is sinus tachycardia and also a large R-wave in aVR.

V1 and V2 have type 1 Brugada pattern

To determine whether it is indeed Brugada or not, it is important to determine whether the QRS is actually prolonged or not.

If the QRS is prolonged, then the differential includes:
1. Hyperkalemia
2. Drug toxicity, especially diphenhydramine, which has sodium channel blocking effects, and also anticholinergic effects which may result in sinus tachycardia, hyperthermia, delirium, and dry skin.
----Other drugs with sodium channel blockade: Tricyclic antidepressants.  They result in a large R-wave in aVR, but also usually have a wider QRS.
----Other sodium channel blockers, such as flecainide.

However, if the QRS is not prolonged, then Brugada uncovered by fever is the likely culprit.

Here I have drawn a line from the end of the QRS in lead II across the bottom up to V1 and V2.
This shows that much of the apparent R'-wave in V1 and V2 is really a downsloping ST segment.  Thus, Brugada is the likely diagnosis

A very nice explanation of this is given in the document quoted below on current ECG criteria for Brugada pattern.

Bayes de Luna, A et al. Current electrocardiographic criteria for diagnosis of Brugada pattern: a consensus report. J Electrocardiology 45 (2012):433-442.

Full text at Researchgate

Isolated RBBBs

Type 1 Brugada and advanced RBBB are both characterized by a terminal positive wave and a negative T wave in V1-V2.  The distinction between the 2 conditions is straightforward because, usually, in advanced RBBB, the ST segment is not elevated in the right precordial leads, the terminal wave (r' or R') is synchronous with the broad S wave observed in leads I and V6, and the QRS is wider (120 milliseconds). 

In type 1 BrP, however, usually no wide S-wave is present in the left leads because the terminal forces of the ventricular complex in V1-V2 can be recorded only by electrodes placed in proximity of the site where the abnormal electrical activity occurs (the outflow tract of the right ventricle) and not from further away leads. Therefore the QRS in left leads is usually less than 120 milliseconds.

Type 1 Brugada syndrome, which may be baseline or uncovered by fever, and has a normal QRS duration but a wide appearing R'-wave in V1 and V2 because of the downsloping ST segment.

Clinical course:
The patient was diagnosed with pyelonephritis and her fever was reduced with antipyretics.  She was given fluids and she improved and became lucid.  She was able to state she had taken no meds, no drugs, no OTCs, no dietary supplements.

We repeated the ECG:
Brugada pattern is mostly resolved.

Further history later: This patient personally has no further high risk features (syncope / presyncope), but her mother had sudden cardiac arrest in sleep.  

Follow up the next AM:
Brugada pattern is resolved

Below is what the electrophysiologist recommended.

This is based on the Sieira et al, 2017, risk calculator, which gives a borderline risk score (2).    

--Genetic testing could be helpful to confirm the diagnosis and allow for screening of other at-risk family members.   
--EP study to further risk stratify her is recommended, with ICD placement depending on the results.  --Regardless of further evaluation, she should avoid bradycardia, AV nodal blockers, Na channel blockers, and fevers. 
--If she has fever, she should aggressively take acetaminophen and/or ibuprofen and, if not improving, will need to go to ED for further evaluation.  
--If there are any syncopal or presyncopal events, she should be evaluated immediately in the ED.  
--All 1st degree relatives should have 1 time EGG.  


Case 2.

A few days later, this patient presented with influenza:

70-something with fever of 38.0, and was diagnosed with influenza (5523467)
This was read by the computer and overreading physician as:
"Type 3 Brugada Pattern"

Is it?

No!  This is classic type 1 Brugada pattern.

Read more about the types of Brugada syndrome here.  Types 2 and 3 have been merged into Type 2 and look substantially different.

Is this Type 2 Brugada syndrome/ECG pattern?

This was recorded about 30 minutes later:

A previous ECG was obtained and was normal.

The patient denied any chest pain whatsoever, and a troponin at zero and 2 hours were both undetectable. A bedside cardiac ultrasound revealed grossly normal to hyperdynamic systolic function with no obvious areas of wall motion abnormalities. His lab workup was significant for positive influenza A rapid test and hyponatremia.

Patient was advised of the importance of fever control, follow up with primary care and also a cardiologist.

They did not verify resolution of Brugada pattern with control of fever.  This would be advisable.


Here are more cases of Fever and Brugada:

--Hyperthermia and ST Elevation 

    -- The syndrome of Brugada and Fever was discussed at length

--A Patient with Syncope 

     -- a very subtle case of syncope due to Brugada that was missed on the first presentation and not diagnosed until he had recurrent syncope 3 years later.

     -- in this case, the computer diagnosed STEMI but the patient had Fever with Brugada


Fever and Brugada--Important articles

The literature below shows that fever-induced Brugada is indeed a high risk for an arrhythmic event.

This article is on Brugada in general, not just fever-induced:

A score model to predict risk of events in patients with Brugada Syndrome

Here is full text:

This was a single center cohort of 400 patients with Brugada syndrome.  78 patients had spontaneous type I ECG.  Clinical presentation was aborted sudden death in 20 patients, syncope in 111, and asymptomatic in 269 patients.  Familial syncope or sudden death was found in 184 patients, 31 of these in first degree relatives under 35 years of age.  ICD was placed in 176.  During a mean follow-up of 80.7 months, 34 arrhythmic events occurred (event rate: 1.4% year). Variables significantly associated to events were: presentation as aborted SCD (Hazard risk [HR] 20.0), syncope (HR 3.7), spontaneous type I (HR 2.7), male gender (HR 2.7), early SCD in first-degree relatives (HR 2.9), SND (HR 5.0), inducible VA (HR 4.7) and proband status (HR 2.1). A score including ECG pattern, early familial SCD antecedents, inducible electrophysiological study, presentation as syncope or as aborted SCD and SND had a predictive performance of 0.82. A score greater than 2 conferred a 5-year event probability of 9.2%.

These are specifically on fever-induced Brugada

Arrhythmic events in Brugada syndrome patients induced by fever.  Annals of Noninvasive Electrocardiology 2019.

These authors analyzed all the reports of fever with Brugada which are available and found that the incidence of arrhythmic events in patients with active fever and Brugada was 38% (very high!).  (This has a high risk of selection bias, as all are reported cases.)

This study references experimental evidence that higher temperature alters Na conduction through the membrane in patients with SNC5A mutation.

Studies have shown a higher prevalence of BrS in febrile patients compared to nonfebrile ones.  Fever not only unmasks a Brugada-type electrocardiogram (ECG) but also increases the risk of ventricular tachyarrhythmias such as ventricular fibrillation (VF) or sudden cardiac death.  The underlying mechanism of this finding might involve a decrease of the sodium current during a rise in temperature which has been recently described using human cardiomyocytes from induced pluripotent stem cells.

They emphasize that further analysis has to be executed in order to find a common way as how to treat patients that present with fever-induced BrS as well as to establish the risk of arrhythmic events that might threaten those patients with next fever episode. 

Conclusion of this paper: Fever is a great risk factor for arrhythmia events in Brugada Syndrome patients. Patients with known fever-triggered Brugada syndrome should be surveilled closely during fever and be started on antipyretic therapy as soon as possible.  There are limited data regarding the impact of EP study in BrS patients triggered by fever.  An EP study might be helpful in symptomatic patients (Sroubek et al., 2016) in the presence of spontaneous BrS ECG or drug-induced ECG.

Eighty-eight asymptomatic patients with fever induced Type 1 Brugada syndrome had a mean age of 45.8 ± 18.7 years and 71.6% (67 of 88) were men. Twenty-one percent (18 of 88) had a family history of sudden cardiac death and 26.4% (14 of 53) carried a pathogenic SCN5A mutation. Drug-challenge was positive in 29 of 36 patients tested (80.6%). The risk of ventricular fibrillation in asymptomatic patients was 0.9%/y (3 of 88, 43.6 ± 37.4 months).

Fever-related arrhythmic events in the multicenter Survey on Arrhythmic Events in Brugada Syndrome.  Heart Rhythm 2018.  In 588 patients with known Brugada syndrome and first arrhythmic event, 35 were during fever.  Family history of sudden death,  history of syncope, and spontaneous type 1 Brugada ECG were noted in 17%, 40%, and 71% of patients, respectively. VF was induced at EP study in 9 of 19 patients (47%). An SCN5A mutation was found in 14 of 28 patients (50%). The highest proportion of fever-related AE was observed in the pediatric population (age less than 16 years) with a disproportionally higher event rate in the very young (age 0–5 years) (65%). 

MY Comment by KEN GRAUER, MD (12/5/2019):
We have previously discussed Brugada ECG patterns and Brugada Syndrome on multiple occasions in Dr. Smith’s ECG Blog (Sept. 8, 2019) — (Jan. 30, 2019) — many others ...
  • In today’s post — Dr. Smith emphasizes the potential for high fever to precipitate a Brugada-1 ECG pattern, with associated increased risk for an arrhythmic event.

Review of relevant literature is provided by Dr. Smith (above). I limit my comments to the following synthesizing thoughts:
  • The ECG picture of a Brugada-1 pattern (as shown in the initial ECG of this case) should be instantly recognized. Emergency providers need to be comfortable recognizing this ECG pattern.
  • Acute febrile illness may “unmask” previously undiagnosed Brugada Syndrome. Treatment of the underlying cause of the fever (which was acute pyelonephritis in Case #1 above) — may resolve (or greatly diminish) the abnormal ECG pattern (as it did in Case #1).
  • What is clear — is that patients with spontaneous or induced (including induced by fever) Brugada-1 ECG patterns are at greatly increased risk of a future malignant arrhythmia event IF there is: i) A positive family history of sudden death — and/or — ii) A spontaneous Brugada-1 ECG pattern in a 1st-degree relative — and/or — iii) A previous history in the patient of syncope, presyncope, seizures or nocturnal agonal respiration — and/or — iv) A positive sodium channel-blocker challenge test after resolution of the acute precipitating problem.
  • What is less clear — is how much risk an asymptomatic and otherwise healthy patient with a structurally normal heart faces, especially IF the Brugada-1 ECG pattern resolves after resolution of the precipitating event. KEY QUESTION: Does the previously healthy, 30-something young woman in Case #1, whose Brugada-1 ECG pattern largely resolved with treatment of her acute febrile illness — Does she truly need an ICD? (Implantable Cardioverter-Defibrillator), with long-term potential for device-related complications from the ICD, including inappropriate shocks?
BOTTOM LINE: Whereas decision-making is clear for symptomatic patients with a Brugada-1 ECG pattern — it is complex, and continues to be a bit of a “moving target” for asymptomatic individuals without persistent ECG abnormality. I'd be happy to refer these patients to EP cardiology for formal evaluation.

P.S. (which I am writing the next day, on 12/6/2019): I missed the "Further History gotten Later" at the time I wrote my initial comment on 12/5. This further History adds that although this patient personally had no syncope/presyncope — her mother had sudden cardiac arrest during sleep ==> Therefore, a positive Family History that for this young woman clearly merits full evaluation by EP cardiology!

Recommended Resources