Tuesday, July 21, 2020

A case of misinterpreted troponins, in spite of a very suspicious ECG....

This 50-something male with previous history of MI presented for intermittent CP and SOB for 2 days. 

CP lasted for hours at a time, was described as pleuritic, without radiation, but relieved by nitro. 

He was given nitro and full dose aspirin by EMS.  Prehospital ECG was similar to first ED ECG.

Here is the ED ECG for ED visit #1:
It is very abnormal, with potentially ischemic downsloping ST depression

There were 3 ECGs during an ED visit for chest pain one month earlier.  Let's call that ED visit zero.

Here is the last EKG from ED visit zero:
There is minimal ST depression without the downsloping.

Here is the first EKG ED visit zero:
This looks more like the one at the top and is more suggestive of ischemia.

On ED visit zero, the patient had 3 troponins below the level of detection.  Contemporary troponin I, Abbott, LoD = 0.010 ng/mL, 99% URL = 0.030 ng/mL

So back to visit number 1.  This 2nd ECG was recorded.
It look substantially less ischemic than EKG 1.
Thus, there are dynamic ECG changes.

So the ECGs alone are all but diagnostic of ischemia in the setting of chest pain in a patient with known coronary disease

The initial contemporary troponin I was 0.023 ng/mL.  The second was 0.025 ng/mL.

The patient was admitted.

The inpatient team recorded these contemporary troponins over 15 hours.
As you can see, there was a rise and fall of troponin, with several values above the 99th percentile of 0.030 ng/mL.  
This is diagnostic of myocardial infarction by the 4th Universal definition.

The inpatient team decided she was "ruled out" for MI.  Why?  Here are some quotes:
"EKG stable x 2"
"chest pain no longer present"
"trop of 0.041 trended downward to normal range."
"As pt states he feels as though it improves with bowel movements or passage of gas, possible that pain is referred from upper abdomen 2/2 constipation, possibly causing diaphragmatic irritation." "Also possible that with slight leukocytosis, cough, and low grade fever x3 days, patient could have viral URI with mild viral myocarditis."
"ACS was ruled out and denied any CP on discharge."

Clinicians will go to extremes to explain away troponins that are above the 99th percentile, as they are commonly elevated in conditions that are not ACS.  But one must 1) find an explanation and 2) not have a rise and/or fall, or explain a rise and/or fall. 

A falling troponin is NOT reassuring. It means that the rise is acute and that one is seeing it fall after it rose.

Interestingly, we had some high sensitivity values due to ongoing research (far right column):
Assay on far right: Abbott Architect high sensitivity troponin, LoD = 1.9 ng/L (0.0019 ng/mL), with URL at 16 ng/L for women and 34 for men.

At 33.5 ng/L, this almost reached the 99th percentile URL for a male, but not quite. 
So, by definition, it is not a myocardial infarction.  
This does not mean, however, that it is not ACS. 
There was a minimal, but real, rise and fall, but it never went above the 99th percentile. 
Along with the ECG findings, even if you only had the high sensitivity troponins, it should be considered unstable angina, as the followup will prove (unstable angina is, of course, a form of ACS)  

The patient returned one month later with CP, onset during sleep, duration 2 hours

Similar, but more pronounced ST depression than last time

Here are the troponins for this visit (contemporary, not high sensitivity)

Contemporary trops are reported as ng/mL, not ng/L

The patient underwent immediate angiography and PCI because of uncontrolled pain:

LCx: Severe (99%) stenosis in mid CX prior to first obtuse marginal. This is the culprit.

Lesion on Mid CX: Proximal subsection. 99% stenosis 14 mm length reduced to 0%. Pre procedure TIMI III flow was noted. Post Procedure TIMI III flow was present. Good run off was present.  

Learning Points:

1. Learn these dynamic ECG findings which are highly suspicious for ischemia.
2. Troponin elevation intrpretation is very complicated, and getting moreso with high sensitivity troponins
3. A rise and fall with one value over the 99th percentile, in the context of ischemic symptoms, is Acute MI
4. A rise and fall of high sensitivity troponins, even under the 99th percentile, is highly suspicious for unstable angina.

MY Comment by KEN GRAUER, MD (7/21/2020):
There is an important lesson conveyed by this case — namely the importance of recognizing dynamic ECG changes — appreciating the significance of such changes, especially when they occur in association with symptoms  and, then acting accordingly.
  • The patient in today’s case is a 50-something man with a prior history of MI. He presented to the ED for a 2-day duration of intermittent chest pain with both typical and atypical features for a cardiac etiology.
  • A total of 5 ECGs are shown above by Dr. Smith. I focus my comments on 3 of these 5 tracings that I’ve numbered in the sequence that they were shown above (Figure-1).
  • NOTE: Dr. Smith’s excellent discussion above provides superb insight on the clinical application of serial troponin values in today’s case. My comments are aimed imagining that you do not have any of those serial troponin results — since I believe optimal clinical decision-making in today’s case could have been made even without troponin results.

Figure-1: The 1st, 4th and 5th tracings shown above in today’s case (See text).

My THOUGHTS on these 3 ECGs: As always — I favor systematic assessment of the initial ECG — then correlating this initial ECG to the clinical history — and then proceeding with lead-by-lead comparison of this initial tracing to subsequent ECGs that are done.

Beginning with ECG #1 = the initial ECG obtained in the ED on this 50-something man with a known coronary disease and a 2-day history of intermittent, not-completely-typical chest pain.
  • There is baseline artifact in the limb leads of ECG #1 — though not enough to impede interpretation. The rhythm is sinus at ~65-70/minute — intervals (PR/QRS/QTc) and the frontal plane axis are normal — and specific criteria for chamber enlargement are not present.
  • Large Q waves are seen in each of the inferior leads. Smaller q waves are seen in the lateral chest leads (although the Q in lead V6 looks a bit deeper and wider than expected, especially given the relatively modest R wave amplitude in this lead).
  • R wave progression is remarkable for abrupt transition from a predominantly negative rS complex in lead V1 — to a huge biphasic complex in lead V2, in which R wave amplitude attains 16 mm!
  • Regarding ST-T wave changes in the Limb Leads — The shape of the ST-T waves in leads III and aVF looks as if there might be slight ST elevation (with straightening of the ST segment takeoff in lead aVF), though these changes are subtle and difficult to quantify. There is slight sagging ST depression in high lateral leads I and aVL, each with a low-amplitude biphasic terminal T wave.
  • Regarding ST-T wave changes in the Chest Leads — The T wave in lead V1 is upright. This is followed by nonspecific ST-T wave flattening in lead V2 — which evolves into slight-but-real J-point depression with sagging ST segments in leads V3-thru-V6. This is similar in shape to the ST-T waves we noted in leads I and aVL.
  • It looks like there may be small U waves blending in with preceding T waves in leads V1-through V3.

My Impression of the LIMB Leads in ECG #1:
  • I’d LOVE to be able to see a prior (baseline) ECG on this patient! Given that this is not available — what we can say, is that there definitely has been an inferior MI at some point in time. I suspect this is old — but given the subtle-but-real findings described in the previous bullet (especially given the ST-T changes in leads I and aVL) — I can not rule out the possibility of a recent event (perhaps superimposed on previous inferior infarction).

My Impression of the CHEST Leads in ECG #1:
  • The abrupt early transition with surprisingly tall R wave already seen in lead V2 is unusual, and of uncertain significance. It might reflect previous posterior MI (associated with the large inferior lead Q waves) — vs some underlying structural disease — vs some technical error with lead placement.
  • Of more concern — are the ST-T wave changes seen in leads V2-thru-V6. While these do not particularly look acute — they could reflect ischemia in this patient with known coronary disease and 2 days of new symptoms.
  • PEARL #1 — The T wave in lead V1 may normally be negative, flat, or upright (in which case there should be no more than minimal ST elevation). T waves in adults are normally positive in other chest leads, with T wave amplitude relatively proportional to QRS wave amplitude in those leads. A useful concept to be aware of is known as “precordial T-wave balance — which states that the size of a positive T wave in lead V1 should not be larger than the size of the positive T wave in lead V6. When this occurs — ischemia is suggested (Tewelde et al — West J Emerg Med 18[4]:601-606, 2017).
  • PEARL #2 — The T wave in lead V1 of ECG #1 is not normal. Although subtle — the peak of this T wave is fatter-than-it-should-be. Applying the concept of “precordial T-wave balance” (from Pearl #1proves that the T wave in lead V1 of ECG #1 is not normal — since it is clearly larger (more positive) than the tiny T wave in lead V6.
  • The U waves I noted above in leads V1-to-V3 are small. Hypokalemia (and/or hypomagnesemia) is always a possibility when there are U waves (especially given some of the ST-T wave changes in other leads) — and a chemistry profile will undoubtedly be part of the ED evaluation.
  • That somewhat larger-than-expected Q wave in lead V6 might be part of the picture of the inferior Q waves we noted (ie, prior infero-lateral MI).

Bottom Line for ECG #1:
  • ECG #1 is not normal. That said — none of the abnormal findings I describe above are necessarily acute. And, we do not have a prior ECG for comparison ... I’d interpret this ECG as consistent with previous inferior MI (possibly infero-lateral MI), with ST-T wave changes in multiple leads consistent with ischemia of uncertain age, possibly recent. Clearly more information is needed to decide on optimal management of this patient.

My THOUGHTS on ECG #4: We are told that ECG #4 was the repeat ECG done on this patient whose 1st ECG is shown in Figure-1. We are told that this patient was admitted to the hospital, presumably on the basis of serial troponin values above the 99th percentile — but it was felt by the treating clinicians that the patient had “stable ECGs X2” + relief of chest pain + not high enough troponin values to rule in ACS.
  • KEY Point: Once you have systematically assessed the initial ECG in the detail I illustrate above for ECG #1— it becomes EASY to compare serial tracings. One simply does lead-by-lead comparison between the 2 tracings with clinical correlation to any change in patient symptoms at the time the follow-up ECG was done.
  • As best I can surmise from the history presented — the patient’s symptoms had decreased at the time ECG #4 was done.
  • ECG #4 — shows a bit more baseline artifact than was seen in ECG #1, but it is still definitely interpretable. The rhythm is again sinus at a slightly faster rate. I see no significant change in QRST morphology in the limb leads. However, there are subtle-but-real changes in most chest leads. Specifically — the T wave in lead V1 looks more positive (even accounting for increased QRS amplitude in lead V1 of ECG #4) — and, the subtle J-point depression with ST segment sagging in a number of chest leads is clearly less than it was in ECG #1.
  • Keeping in mind the concept of precordial T-wave balance that I described above in Pearls #1 and #2 — Isn’t the discrepancy in T wave positivity between leads V1 and V6 of ECG #4 much more striking!
  • Therefore — Since ST-T wave changes in multiple leads that were present during chest discomfort (at the time ECG #1 was done) have greatly improved at a later time after the patient’s symptoms have resolved (at the time ECG #4 was done) — there are dynamic ECG changes. This defines active ischemia.

PEARL #3 — The utility of comparing serial tracings lead-to-lead — is that ECG findings that may have initially appeared subtle will often show obvious change on serial comparison. And, if improvement in ischemic ECG changes is correlated to improvement in clinical symptoms — then you have demonstrated active ischemia.
  • Bottom Line after Seeing ECG #4: Even without the benefit of the elevated troponin values highlighted by Dr. Smith — cardiac cath should have been done on this patient prior to discharge from the hospital. Keep in mind that this 50-something man has known coronary disease — and, he presented to the ED with new symptoms — and, he has demonstrated dynamic ECG changes in ECGs #1 and #4 in association with resolution of his chest discomfort.
  • NOTE: For another example of detailed analysis showing dynamic ECG changes — CHECK OUT discussion by Dr. Smith and My Comment at the bottom of the page in the September 15, 2018 post on Dr. Smith’s ECG Blog.

My THOUGHTS on ECG #5: Approximately 1 month after discharge from the hospital — the patient returned to the ED for another new episode of chest pain that woke him from sleep. ECG #5 is his initial ECG on presentation in the ED. His chest pain was uncontrolled at the time ECG #5 was done.
  • Since the last ECG on file for this patient was ECG #4 (which was obtained at a time when the patient was pain-free) — We should compare the new ECG ( = ECG #5) with this last previous tracing ( = ECG #4).
  • Artifact is minimal on this newest tracing. The rhythm is sinus at a somewhat slower rate. I see no significant change in the limb leads between ECG #4 and ECG #5.
  • NOTE: Isn’t there a marked difference in R wave progression between the 2 tracings! I’m unsure what accounts for this difference ... — but it is important to appreciate this change, because such marked change in QRS morphology might influence resultant ST-T wave changes. That said — in this case, despite marked change in R wave progression between ECGs #4 and #5 — I think it is clear from lead-to-lead comparison of ST-T waves in the chest leads of both tracings, that there has been an obvious increase the amount of ST-T wave depression in ECG #5.
  • As per Dr. Smith — ST-T wave abnormalities in ECG #5 are even more pronounced than they were in ECG #1 at the time he first presented to the ED one month earlier. Therefore — once again, there are dynamic ECG changes indicative of active ischemia. This time prompt cath was done — and PCI was performed on a 99% stenosis of the mid-LCx.

Our THANKS to Dr. Smith for this insightful case!


  1. This ECG and clinical picture is very suspicious for Cx artery desease. Angio must had been done immediately.

    1. @ Sergey — THANK YOU for your comment. What you mention is the KEY teaching point for this case. The clinical history — serial ECGs that show dynamic ST-T wave changes in association with corresponding change in symptom severity + subtle-but-definitely-positive serial troponin values ideally would have (should have) prompted earlier cath in this case. This serves as the important lesson — :)

    2. The 2nd ECG has dynamic changes (as well as ECG 2 at ED month earlier), typical for LCx in my opinion - More positive T at 3 and aVF with reciprocal negative T at 1 and aVL plus reciprocal depression at the side precordial leads. Thank you for your work! From Russia.

  2. excellent, gentlemen...
    (sorry! I've been away for a while. )
    thank you both for this case...
    in ecg 1 of this last ER visit: the question for me I think is "when to pull the trigger", ie, Cath lab. yes, it's abnormal.
    suppose we hadn't the benefit of the ekgs of a month before. then... with U waves across the precordium, may this be all electrolyte abnormality?
    but, with chest pain, maybe rapid serial Ecg's, and electrolyte results an urgent PCI can be done.
    I hate waiting even two hours for the next trop.

    thank you both.

    1. Hi Tom. GREAT to see you back! As per the detailed description in My Comment (above) of ECG #1 ( = the initial ECG in this last ED visit) — My “Bottom Line” thoughts on this tracing were, “Consistent with previous inferior MI, with ST-T wave changes in multiple leads consistent with ischemia of uncertain age, possibly recent. Clearly more information needed to decide on optimal management of this patient.” I thought the subtle U waves might reflect electrolyte imbalance (chem profile presumably pending). But since I did not see definitive acute changes on ECG #1 — I thought NO emergent cath was needed at that time until more information was gathered.

      The 2nd ECG done in this last ED visit ( = ECG #4) showed definite improvement in chest lead ST-T wave changes. Since this apparently was correlated with reduction in chest pain — this defined the change that we saw in ST-T wave appearance as “dynamic ECG changes” in a patient with known coronary disease who presented with new symptoms. Therefore, even without the abnormal troponin values — I thought timely cath was indicated prior to discharge from the hospital. That wasn’t done …

      To address your question — IF the sequence of symptoms and ECG changes would have been reversed (ie, if this patient’s chest pain in that last ED vist after arrival in the ED would have gotten worse AT THE SAME TIME as ST-T wave changes were increasing) — then urgent cath for PCI would have been indicated. Often, “Ya gotta be there” to best judge this essential relationship between onset and severity of symptoms with each of the serial ECGs that are done.

    2. Thought I'd chime in because I took a different point from what Tom said. I believe his point is if you are ever reluctant to activate cath due to the possibility of Potassium/Mag imbalance: Getting an urgent electrolyte panel while you wait for troponin might allow you to activate faster than the normal thought process of "wait for trops".

    3. Thanks for your comment Kyle! We are all aiming for that “optimal balance” between not allowing value time ( = myocardium) to pass — while at the same time being as certain as possible of the correct diagnosis. MY thought — looking first at electrolytes that might cause ST elevation (I’ve seen this, though it is RARE with hyperCalcemia; — BUT — the peaked T waves of hyperKalemia are a COMMON potential mimic of deWinter T waves). With hyperCalcemia — in my experience, you need a VERY HIGH serum calcium level for their to be enough change to potentially mimic acute MI (ie, probably >13 mg/dL) — and in addition to that being a RARE situation, there will OFTEN be a clue in the history (ie, a patient with known lung or breast, multiple myeloma, leukemia or other cancer that is likely to produce hypercalcemia).

      With hyperKalemia — We’ve reviewed numerous such cases on Dr. Smith’s blog. One looks for peaked T waves (typically with a narrow base) — QRS widening — loss of P waves — unusual axis (often right axis) — unusual conduction defects AND often a clinical setting (such as renal failure or K+-retaining medications) likely to predispose to hyperkalemia. And IF hyperKalemia is suspected from the appearance of the ECG — faster than waiting for a “stat” lyte panel — Give IV Calcium !!! (minimal downside and VERY positive upside).

      And among those lyte disorder likely to cause ST flattening and/or depression — that usuallly will NOT be an indication for emergent cath UNLESS: i) You see dynamic ST-T wave changes (which by definition means that some time has already passed … = probably enough time to get lytes back) or ii) Your patient is having SEVERE refractory chest pain in the midst of an ECG that probably looks something like diffuse subendocardial ischemia (in which case you can contact Cardiology on call and make plans for cath activation (which should give you at least some time, which often will be enough to get lab results back). But without knowing statistics — I would BET there is a small MINORITY of cases in which you will have a single ECG on a patient with ST depression that mandates immediate cath. We WELCOME submission of any such cases! — :)

  3. Do you know of any studies into the length of time Troponins stay elevated after MI? I recently had a case where patient presented 2 weeks after STEMI with Troponin T 650 and then 620. Their Troponin T on their first admission was 6,000+. May this be his original Troponin still coming down or another acute event?

    1. Yes, that is a normal troponin profile. After large MI it stays high for a couple weeks.


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.

Recommended Resources