Monday, September 9, 2024

"The dye don't lie" ...except when it does. Angiogram Negative, or is it?

Written by Willy Frick


See the fascinating discussion on X started by this post. Link below.  


A 40 year old woman was at home cooking when she developed chest pain. She took an oxycodone and called EMS. Her presenting ECG is shown below:

ECG 1
What do you think?













I sent this to Drs. Meyers and Smith and the response was: "I’m quite confident that this is OMI."

Here is the Queen's verdict with explainability:

She sees OMI with low confidence. As an aside, the explainability above came from using QOH in the Telegram bot environment. I had previously run this ECG through QOH in the PMcardio app environment and she reported mid confidence, shown below.


Although the QOH is exactly the same in both environments, extremely subtle changes in digitization from one analysis to the next can slightly change her confidence. The QOH output is a number between 0 and 1. A value of 0 would be "not OMI" with high confidence, and a value of 1 would be "OMI" with high confidence. The cut points for what constitute OMI and not OMI as well as confidence levels are calibrated to maximize sensitivity and specificity according to the receiver operating characteristic.

Queen of Hearts version 2 is in active development, and she is trained on a much larger data set than version 1. The image below shows output for QOH version 1 and QOH version 2. You can see that version 2 has a higher number than version 1, hence she sees the ECG as more OMI-like than version 1.

Version 2 of the Queen of Hearts
She is more confident than Version 1



Back to the case...

The ECG was interpreted as showing diffuse STD with some STE in aVR. The notes do not comment on V1 and V2. In fact, V1 and V2 both have STE, and V2 in particular has hyperacute T waves. Taken together with STD in the lateral precordial leads, this represents precordial swirl, and is therefore diagnostic of LAD OMI proximal to the first major septal perforator.

Even though QOH did not have access to the prior ECG, we do. This is shown below:

Prior ECG

As compared to the prior ECG, we can now feel completely confident that the V2 T waves are in fact hyperacute, and that the current ECG shows LAD OMI.

Initial hscTnI was 10 ng/L (ref. <14). The patient was thought to have low likelihood of ACS, and cardiology recommended repeat troponin, urine drug testing, and echocardiogram. There was no recommendation for repeat ECG. I strongly believe you should never trend troponin without also trending ECG. Troponin tells you what was happening hours ago, ECG tells you what is happening right now.

Repeat hsTnI two hours after the first increased from 10 to 8,512 ng/L. At that point, cardiology elected to treat for ACS. Bedside echocardiogram showed hypokinesis of the mid to distal anterior wall and apex. Angiography is shown below.

Shown below is a left anterior oblique view of the RCA with a little bit of cranial angulation ("LAO crani").


Here is a representative still showing RCA and the distal bifurcation into the PDA and a right posteralateral (RPL) branch. The diaphragm tells you this is a cranial shot (i.e., you are looking down from above).


Next is RAO cranial shot.


Below is a representative still showing the LAD and septal perforators. The diagonal branches are overlapping one another and not well visualized in this view. Although this view is not meant for it, there is fair visualization of the LCx. The diaphragm again tells you this is a cranial view.


So far, we haven't seen anything to explain LAD OMI, but let's keep looking. The next shot is anteroposterior ("AP") cranial, which typically visualizes the LAD and diagonal bifurcation.


Here is a representative still showing the LAD and septal perforators. The first and second diagonal branches are both sizable vessels, but unfortunately they're overlapping here, so they are not very well visualized. The LCx is not well visualized (nor is it meant to be in this view).


Again, not much going on so far. Certainly there is TIMI 3 flow everywhere. The next view is LAO caudal (sometimes called the "spider view"). This view is best for left main and the proximal LAD, LCx, and branch vessels.


In this representative still, we see the LADdiagonal branches, and LCx. All proximal vessels appear widely patent, and the septal perforators are not well visualized (due to overlap with LAD). There is no diaphragm since this is a caudal view.


The last shot is RAO caudal, which is good for proximal LAD and LCx.


And here is a representative still showing LADdiagonal branches, the ostium of the septal perforators, and LCx.


To most, this probably looks like completely normal angiography. And it could easily be mistaken as such. Eagle eyed readers might notice a slightly scooped out appearance of the proximal LAD in the above view, right at the head of the top left green arrow. The cardiologist called this 20% stenosis.

In a large proportion of cath labs, the operator would probably have ended the case at this point. You can easily imagine this patient getting one of several diagnoses -- vasospasm, MINOCA, pericarditis, or maybe even no diagnosis at all beyond "non-obstructive coronary artery disease."

Smith comment: a very high proportion of MINOCA are ruptured plaque with lysed thrombus. That plaque is at risk of thrombosing again.  It is worthwhile remembering that the majority of plaques which rupture are non-obstructive before they ulcerate and thrombose.  If the thrombus completely lyses, then there may be no visible evidence on angiogram.  An angiogram is a "lumenogram" and does not "see" the extraluminal plaque.  Most plaque is outside the lumen!!  And so other modalities may be necessary.

Fortunately, this operator used intravascular ultrasound (IVUS). (Another option would be to use Optical Coherence Tomography for Coronary Imaging).  This showed a "completely normal vessel throughout the LAD, except in the proximal/ostial LAD there was a ruptured plaque." With this in mind, ECG 1 showing precordial swirl makes perfect sense. In fact, there was transient occlusion of the LAD proximal to the septal perforators.

At the time of IVUS, there was no thrombus. The report describes extension of the plaque into the distal left main. The operator documented thoughtful consideration of risks and benefits of stent placement. Technically, there was a very narrow landing zone for the stent, and missing this could result in "jailing" the LCx, which is ideally avoided. Furthermore, the operator worried about the patient's adherence to dual antiplatelet therapy, in which case she would be at risk for catastrophic stent thrombosis. For this reason, and given that she had very robust spontaneous recanalization, they decided to attempt medical management with a plan for immediate repeat cath if she had any return of symptoms.

Repeat hsTnI just after cath was 36,029 ng/L. ECG below shows obvious LAD reperfusion.


Unfortunately, a few hours later the patient complained of recurrent chest pain. Repeat ECG showed progressive reperfusion without any evidence of reocclusion.


However, given the context, she returned for immediate angiography and received a stent to her proximal LAD. A few selected echo images show extensive, severe hypokinesis in the distribution of LAD.



Learning points:
  • OMI is not ruled out by non-obstructive, almost normal coronary arteries.
  • Troponin tells you what happened hours ago, ECG tells you what is happening now.
  • Recognize precordial swirl, a very important sign for diagnosing subtle proximal LAD OMI.

This case engendered interesting discussion on Twitter.








==================================
My Comment, by KEN GRAUER, MD (9/9/2024):
==================================
Today's superb presentation by Dr. Frick provides a great example of how expert ECG interpretation can help the interventionist. Knowing (as Dr. Frick points out) that the initial ECG in today's case strongly suggests Precordial "Swirl" (therefore diagnostic of a proximal LAD OMI) — prompted the interventionist to keep looking at the area predicted to be the site of the "culprit" artery, with awareness of the need for IVUS (IntraVascular UltraSound) to make the final diagnosis once the subtle cath finding of a scooped out appearance to the proximal LAD was seen.
  • For as good as QOH is in recognizing acute OMI (with even better performance anticipated once Version-2 is released) — QOH confidence for rating acute OMI was less-than-high for today's initial ECG.
  • I'll suggest that joint assessment of this initial tracing by QOH, together with clinical ECG assessment by a skilled interpreter — could have (should have) increased confidence in the diagnosis of acute OMI even before a prior ECG on this patient was found.

For clarity in Figure-1 — I've labeled the initial ECG and the previous tracing.,

Figure-1: Comparison between the initial ECG and a prior ECG on today's patient.

============================
NOTE: For review of 20 cases of "Swirl" vs Swirl "Look-Alikes" — Check out the October 15, 2022 post in Dr. Smith's ECG Blog. At the bottom of the page in this Oct. 15 post — I summarize in My Comment, a series of tips to facilitate recognition of Precordial "Swirl".
============================

My "Take" on Today's CASE:
The history in today's case immediately places this patient in a higher-risk group for an acute event — in that this 40-year old woman suddenly developed new CP (Chest Pain) severe enough to contact EMS.
  • Initial assessment of ECG #1 was without the benefit of a prior ECG. This initial ECG shows sinus rhythm — normal intervals and axis — and no chamber enlargement.
  • Regarding Q-R-S-T Changes — the QS in lead V1 is not necessarily abnormal given development of a tiny r wave by V2 — with normal R wave progression (ie, showing transition to a predominant R wave already by lead V3).
ST-T wave abnormalities — are present in virtually every lead, with the challenge being that until we are able to find a prior tracing for comparison — it is difficult to know which leads to focus on.
  • As per Dr. Frick — my attention was immediately captured in ECG #1 by the 2 leads within the RED rectangle. In both lead V1 and lead V2 — RED arrows highlight more J-point ST elevation than should be seen in these leads. In support that this ST elevation is likely to be acute in this woman with new CP — is the abnormal ST segment straightening in lead V1 — and the disproportionately-tall (hyperacuteT wave in lead V2 that is much larger than one would expect given modest depth of the S wave in this lead.
  • That the above-described ST-T waves in leads V1,V2 are clearly abnormal — is supported by the surprise finding of ST segment straightening with slight ST depression in neighboring lead V3 (whereas normally there should be slight upward sloping ST elevation in lead V3).
  • My "Go-To" Lead for confirming suspicious anterior lead ST-T wave changes — is lead aVL. Whereas I was not certain if the ST depression and T wave inversion in leads I,II,III,aVF was acute — given the likely hyperacute ST-T waves in leads V1,V2 — I interpreted the disproportionately "bulky" upright T wave in lead aVL (considering tiny size of the QRS in this leadas also hyperacute.

  • As emphasized in My Comment at the bottom of the October 15, 2022 post — after establishing the abnormal ST elevation in leads V1,V2 — the final finding for identifying Precordial Swirl — is, in a patient who does not have LVH — the recognition of a relatively flattened appearance to a depressed ST segment in at least lead V6, if not also in lead V5 (BLUE arrows in these leads).

The Previous ECG:
As per Dr. Frick — comparison in Figure-1 of today's initial ECG with the previous tracing that was found confirms that the chest lead ECG findings diagnostic of LAD OMI with "Swirl" are new!
  • That said — Which of the 4 complexes shown within the RED rectangle in ECG #2 is the correct one? The ST-T wave in "C" would seem to be the most worrisome — and "A" the least worrisome — but there is no way to know from ECG #2 which complex should be used. 
  • While not critical to assessment in today's case (because it is obvious even without assessing ST-T waves in artifact-filled leads V2 and V4 that the changes in ECG #1 are new) — the point to emphasize is that when assessing patients for OMI, the ECG should be repeated IF critical leads are rendered uninterpretable because of artifact (as is the case for this previous tracing on today's patient).

"Take-Home"
 Message:
 The diagnosis in today's case of acute LAD OMI with "Swirl" became obvious as soon as the previous ECG was found.
  • That said, given the history of severe new-onset CP — assessment of the initial ECG alone (as described above) — should have been sufficient to make this diagnosis with sufficient confidence to know prompt cath is indicated.
  • As per Dr. Frick — any doubt about this need for prompt cath could have been quickly alleviated simply by repeating the ECG within ~15-20 minutes after the initial tracing was recorded.


 


No comments:

Post a Comment

DEAR READER: I have loved receiving your comments, but I am no longer able to moderate them. Since the vast majority are SPAM, I need to moderate them all. Therefore, comments will rarely be published any more. So Sorry.