Chest pain one day after a negative stress test

 Written by Dominic Nicacio MD, with edits by Pendell Meyers and Steve Smith

A man in his early 50s with known CAD and prior stent presented to the ED via EMS one day after hospital discharge for a chest pain admission with a negative stress test. He experienced recurrent chest pain the night following his hospital discharge and called EMS. He was noted by EMS to be frankly diaphoretic. He was given sublingual nitroglycerin and full dose aspirin. EMS called in over the radio and reported patient was having chest pain with concerning findings in his anterior ECG leads. The ECG was transmitted to ED providers by EMS. 

But when we received this ECG from EMS, we had no knowledge of any of the history above, other than acute chest pain and diaphoresis.

ECG sent by EMS to ED Providers at 1752:

What do you think?

Interpretation:
- NSR
- Borderline LAFB, normal precordial R wave progression

- slight STE in aVL, 1 mm STE V1, 2-2.5 mm STE V2-V4

-baseline movement makes it hard to say if there is STD in III and aVF

- Very tall, asymmetric T waves in V2-V5, which do not have excessive area under the curve for their height; the differential for these T waves is: normal variant, hyperacute, or potentially hyperkalemic

- QT interval WNL

The differential is anterior OMI vs. normal variant vs. hyperkalemia.

Meyers note: Usually, the defining feature of hyperacute T waves is that they are abnormally fat and broad, as if being inflated with air from below, causing increased area under the curve of the ST-T waves. I teach that hyperacute T waves look like they are being inflated with air, while hyperkalemic T waves look like a tent being pitched with a pole. I also teach that hyperacute T waves resemble the US capitol building, compared to hyperkalemic T waves which look like the Eiffel tower. Hyperacute T waves are often quite tall, but not always. When I saw this ECG without any other context, I initially had a moment when I wasn't totally sure of OMI, in part because I thought these T waves weren't as fat and broad as most hyperacute T waves. So I sought out the patient's prior ECG. 

Using the formula:

STE60V3 = 3.0 mm

QTc (computer) = 419 ms

R wave V4 = 15 mm

QRS amplitude V2 = 19.5 mm

Result = 18.01 (barely less than highly specific cutoff of 18.2) 

As with all dichotomous rules: "The closer the score is to 18.2 (e.g. between 17.7 and 18.7) the more likely it is to be a false negative or false positive"

So this result of 18.01 is not reassuring.

I was surprised to find that his most recent ECG was from yesterday (this is when we discovered that he had just been discharged):

Barely 0.5 mm STE in V2-V4, normal T waves.

Compared to this "prior" (1 day ago) ECG, the EMS ECG above is obviously diagnostic of acute LAD OMI.

Based on this, we activated the cath lab prior to patient arrival.

On arrival to the ER he was diaphoretic and uncomfortable appearing, and he was immediately rolled into a room in our ED high acuity area.


First set of vitals: BP 153/109 mm Hg, HR 67 bpm, RR 18/min, O2 96% on room air

Physical Exam:
General: Diaphoretic, uncomfortable appearing
Cardiovascular: Regular rate and rhythm, 2+ radial and DP pulses, brisk capillary refill
Respiratory: Clear to auscultation in bilateral anterior lung fields with normal work of breathing
Abdomen: Soft, non-tender
Neuro: Alert and oriented, but with paucity of speech and eyes closed unless prompted



On arrival, EMS showed us all their ECGs over 30 minutes:

1747:

Very tough! without comparison to the prior ECG, I would not be immediately certain that this was early LAD OMI.

1752 (the one they transmitted, shown above)

Interpretation above. Again, compared to prior ECG yesterday, completely diagnostic of LAD OMI.

1809

Very similar to 1747 ECG above, meaning interval reperfusion since 1752.

1815

Obvious STEMI(+) OMI, re-occluded since 1809. Notice the RBBB beats with concordant STE in V1-V3.

The interventionalist did a small double-take when we told him about the negative stress test yesterday, but of course agreed to take the patient immediately to lab.

While awaiting cath lab readiness, we obtained some ultrasound images:

Interpretation:

- Normal appearing LV and RV systolic function
- No signs of acute right heart strain
- No pericardial effusion
- No obvious dissection flap in PSL view
- IVC is collapsible
- There may be an anterolateral WMA on limited PSS view

While the patient was in the cath lab, I was curious to review this admission from yesterday with chest pain, to see if this could have been predicted or prevented:

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

Prior Visit (one day prior):

Pt presented for sharp left chest pain lasting 20 minutes while walking at work, immediately called EMS and presented to ED within less than 1 hour of the episode, resolved upon ED evaluation. He stated that these symptoms were identical to prior MI (for which he received an LAD stent).

Here is his ECG on presentation:

I see a tiny bit of STD in V6, II, III, and aVF. Otherwise, I see no other convincing features of ischemia. There was no ECG on file before this one, nothing to compare to.

Initial high sensitivity troponin I was 6 ng/L (limit of detection less than 6, and URL for men 20 ng/L)

Second troponin was 9 ng/L.

Third troponin was 16 ng/L. NO MORE TROPONINS WERE ORDERED!!!!!!

He was admitted to medicine

Cardiology was consulted: "Episode of exertional chest pain with typical and atypical features. ECG demonstrates no acute ischemic changes. He ruled out for myocardial infarction by serial high-sensitivity troponin values....He has history of prior proximal LAD drug-eluting stent....He will be scheduled to have vasodilator stress myocardial perfusion study in the morning for evaluation of myocardial ischemia and LV systolic function. If he has recurrent chest pain or significant ischemia on stress myocardial perfusion study, he will be referred for left heart catheterization."

"Adenosine stress myocardial perfusion study performed and demonstrated normal myocardial perfusion with no ischemia or prior infarction. Gated SPECT demonstrated normal LV systolic function with no wall motion abnormalities and calculated LVEF 59%. There was no chest discomfort or diagnostic ECG changes noted during study. The patient may be discharged home with outpatient follow-up in our office in 2 to 3 weeks."

Meyers note: "ruled out for MI?" You can't just stop measuring rising troponins and declare that there was no AMI. A patient with clear ACS symptoms and rising troponin (not stable angina, suggestive of a ruptured plaque) should not be subjected to a stress test. This is risky and would never be done intentionally if providers suspected acute type 1 ACS with plaque rupture and risk of reocclusion or significant worsening of ischemia during stress. Luckily for the patient, he had no adverse events during the stress test.

 

Another ECG was obtained on the day of discharge:

Similar to the ECG above on initial presentation. STD in III and aVF is abnormal, but I do not see clear high lateral OMI causing reciprocal findings. Could it be the subtlest south african flag pattern? But by all accounts, the patient had no further pain at all, even during stress. And also no reperfusion (Wellens morphology), probably because the ischemia was too brief.

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

Clinical Course:

He was found to have a proximal LAD stent with acute thrombotic 99% in-stent restenosis, also non-culprit 90% proximal RCA, 40% distal stenosis of his LM, and otherwise mild diffuse CAD. He had PCI performed on his proximal LAD and RCA with TIMI-3 flow after intervention. Formal echocardiography after LHC showed and LVEF 58% with hypokinesis of the basal-mid inferolateral myocardium as well as the basal inferior and inferoseptal myocardium (I find this to be an unexpected area of wall motion abnormality considering the ECG and angiogram). He was started on dual antiplatelet therapy, did well post procedure, and was discharged the next day.

Troponins during our visit:

Initial ED troponin: 12 ng/L (!!! because so little time between onset of symptoms and blood draw)

282

500

2608

none further measured (this is a significant problem in our retrospective research, because we are unable to quantify this patient's peak troponin).




Case Discussion Points:

The patient’s ECG transmitted to ED providers by EMS showed T waves that are easily confused with the peaked T waves of hyperkalemia. Sometimes hyperacute T waves have this slimmer, taller appearance, and it can be difficult to distinguish them in these cases. The new STE in anterior leads helps to understand these as hyperacute. Also, other signs of hyperkalemia were absent, as there was no P wave flattening or QRS broadening, bradycardia, new AV blocks or NSIVCD, bizarre QRS morphology, or Brugada patterning. Additionally the clinical context was clearly ACS in this case, and if there were remaining question of hyperkalemia a point of care VBG could be performed to obtain a rapid K+ level.

Comparison to prior ECGs is an absolutely critical piece of information that will help you distinguish OMIs from normal variant STE. Serial ECGs are just as important or more so.

When comparing the present ECG to a prior ECG, it is important to understand the clinical context at the time the prior ECG was captured, as it may not represent a true “baseline ECG”

Rising troponins that have not yet crossed the upper reference limit does not mean that AMI has been ruled out!

Dr. Smith's data using a different high sensitivity troponin assay showed that MI is not ruled out with 100% sensitivity unless all values are less than the URL at 9 hours.  If the first 2 troponins have a very small delta (change ＜ 2-3 ng/L depending on the hs-cTnI assay), that tells you that the peak troponin will not continue to rise. But if the delta on the first troponins is＞2 (≥ 3), that implies that the peak could continue to rise until it is greater than the URL.  Moreover, any such larger changes below the URL are real changes (that is, they are not due to biologicial variation or laboratory imprecision), and need explanation.  So, if the 9 hour troponin were still below the URL (and the ECG were not diagnostic as it is!), it would be reasonable to do further testing with a stress test (but a CT coronary angiogram is far better).  

Even high sensitivity troponins will be frequently within normal range in the first few hours of symptoms. In both presentations, even during active LAD occlusion, the initial troponins in these presentations were both within normal limits.

Do not perform stress tests on patients who may have acute type 1 ACS with ruptured plaque and risk of reocclusion or further ischemia during stress.

Stress test data obviously does not apply to patients who are in the throes of active acute coronary syndrome, as this has never been studied and shouldn't be studied. Once Acute MI has been ruled out, a negative stress test can identify patients who have a lower incidence of ACS events in the future. HOWEVER, when such patients later return to the ED with subsequent acute chest pain, then their prior negative stress test has no protective value as to whether their new chest pain is ACS or not (see references below). The pitfall is that providers may believe that the prior negative stress test has negative predictive value for ACS during the new chest pain event. 

For some more discussion about the "recent negative stress test", here is an excellent excerpt from a very relevant post a few years ago, written by Salim Rezaie for Dr. Smith's ECG Blog:

Chest Pain, "Negative" Stress Tests, POCUS, & ECG Equations -- A Case from Salim Rezaie (R.E.B.E.L. EM)

"Teaching Point #1: Uselessness of “Negative” Stress Tests

Many emergency providers have taken care of patients with true acute coronary syndrome or even primary cardiac arrest despite having had a recent “negative” stress test.  I know I have.  Overreliance on “negative” stress tests can be a common reason for misdiagnosis or delays in patient care.  It is important to remember that coronary artery disease can arise from atherosclerotic lesions that are only mildly stenotic with unstable plaques that rupture and not picked up by standard stress testing. There is a huge misconception about “negative” stress tests in the health care industry and by laypersons. The sensitivities and specificities for stress testing are often reported between 65 – 90% depending on which study you read.  Here are two trials that stress this exact fact:

Study #1 [1]: This was a retrospective chart review of 164 patients with either a “negative” stress test (122 patients) or a “normal” indeterminate stress test (42 patients) over the past 3 years.  34 patients (20.7%) from the total cohort were determined to have significant coronary artery disease in the next 30 days. Significant coronary artery disease was defined as myocardial infarction identified by positive cardiac markers, subsequent positive stress test of any type, cardiac catheterization requiring intervention, CABG, or death due to medical cardiac arrest.  Here is the troubling part…8/34 (23.5%) had their most recent stress test within 1 month prior to admission 7/34 (20.6%) had their stress test between 1 – 3 months, and 11/34 (32.4%) had their stress test between 6mo – 1 year.  Of the total cohort of 164 patients, 13 patients (7.9%) had an AMI.

Study #2 [2]: This was a prospective evaluation of 186 patients who had been referred for coronary angiography for suspected stable angina.  All patients had a normal ECG at rest, none had undergone coronary revascularization, or have diabetes mellitus. 50% of women and 25% of men who had reversible perfusion defects on radionuclide stress imaging had completely normal exercise electrocardiographic findings.  ECG stress is very insensitive.

Teaching Point #1 Bottom Line: Stress testing is used to identify critical stenosis causing obstruction to coronary blood flow, however in the setting of acute myocardial infarction the underlying pathophysiology is plaque rupture and thrombus formation.  Coronary lesions may not have been significant enough to be detected on stress testing.  Therefore a prior “negative” stress test should not be used to determine the disposition of your patients. If you think they are having ACS, then disposition them appropriately regardless of the prior “negative” stress test."

References:

1. Walker J et al. coronary disease in Emergency Department Chest Pain Patients with Recent Negative Stress Testing. West J Emerg Med 2010. PMID: 21079714
2. Hoilund-Carlsen PF et al. Usefulness of the Exercise Electrocardiogram in Diagnosing Ischemic or Coronary Heart Disease in Patients with Chest Pain. Am J Cardiol 2005. PMID: 15619400
3. Marwick TH et al. Techniques for Comprehensive Two Dimensional Echocardiographic Assessment of left Ventricular Systolic Function. Heart 2003. PMID: 14594869
4. Smith SW et al. Electrocardiographic Differentiation of Early Repolarization From Subtle Anterior ST-Segment elevation Myocardial Infarction. Ann Emerg Med 2012. PMID: 22520989

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

MY Comment by KEN GRAUER, MD (10/3/2020):

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

Superb and comprehensive case by Drs. Nacacio, Meyers and Smith — which emphasizes the following points.

• Even high-sensitivity troponins will frequently be within the normal range during the first few hours after the onset of symptoms (as was the case for today's patient in both of his presentations).
• An increase in hs-troponin — even if still "within" the upper reference range — is not a "normal" finding. In a patient with suspected ACS — an increasing hs-troponin trend mandates continued evaluation. Unfortunately this was not done in today's case.
• Stress Testing is contraindicated during suspected ACS. If there has been spontaneous reperfusion of the "culprit" artery — then a Stress Test may result in a false-negative result. If the "culprit" vessel is still occluded — then a Stress Test could result in a much larger infarction. 

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

What About the Cardiac Rhythm?

Although not essential for diagnosis and management of today's case — I found the rhythm in the 4th EMS tracing to be fascinating. I focus my comment on in depth discussion of this tracing — because the KEY points are relevant to diagnosis and management of other cases that readers of Dr. Smith's Blog will encounter!

• NOTE: What follows below includes some advanced concepts on rhythm interpretation! Are YOU up for the challenge?

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

Technical Note: There are a total of 8 ECGs shown in the above discussion by Drs. Nicacio, Meyers and Smith. For clarity — I have numbered these ECGs in the sequence that they are shown above:

• I focus my comment on ECGs #1 and #6 — which I have put together and labeled in Figure-1.
• 
  
• ECG #1 — is the 1st ECG shown above in today's case. This is the 2nd prehospital ECG, which was recorded by the EMS team at 17:52.
• ECG #1 is the same tracing as ECG #4. I chose to show ECG #1 — because the resolution of this ECG is better than the black and white version of this tracing.
• 
  
• ECG #6 — is the 4th prehospital ECG, which was recorded by the EMS team at 18:09 (or 17 minutes after ECG #1). This is the very challenging tracing that I focus my attention on.

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

The CHALLENGE:

What is going on with the rhythm in ECG #6? The complete answer to this question entails several components — each important in its own right:

• HINT: The reason I added ECG #1 to Figure-1 — is that ECG #1 is essential for figuring out what is going on with ECG #6.
• There are 14 beats in the long lead II rhythm strip of ECG #6. How many beats are shown in the 12-lead ECG?
• What is the advantage of this type of ECG recording system? What might some disadvantages of this system be?
• WHY is QRS morphology changing in ECG #6?
• 
  
• By itself — Is ECG #6 diagnostic of acute LAD OMI? IF so — WHICH beats are most definitive for this diagnosis? 

Figure-1: The 1st and 6th ECGs in today's case (which correspond to the 1st and 4th ECGs obtained prehospital by the EMS team).

MY Thoughts on the Rhythm in ECG #1:

When presented with a tracing that has both a 12-lead ECG and a long lead rhythm strip — I favor starting with the long lead rhythm strip — at least to gain a quick initial impression of the rhythm.

• The underlying rhythm in the long lead II rhythm strip of ECG #1 is sinus — as upright P waves with a constant PR interval precede all but 2 of the beats on the tracing (ie, No P wave precedes beats #1 and 4).
• The QRS morphology of these 2 non-sinus beats (beats #1 and 4) is different! The question is why?

KEY Points:

To assess the etiology of non-sinus beats — it is invaluable to be able to view QRS morphology in other simultaneously-recorded leads. This is EASY to do with most monitoring systems that I am familiar with — because long lead rhythms strips are typically recorded simultaneous with the 12-lead ECG above it. But this is not what is happening in ECG #1 !!!

• Note that beat #4 is clearly not simultaneous with the 12-lead tracing above it — because no QRS complex is seen directly vertical above beat #4 in leads aVR, aVL and aVF! This means that the long lead rhythm strip was not recorded at the same time as the 12-lead tracing above it!
• It turns out that the type of ECG recording system used in ECG #1 (and in each of the 4 prehospital tracings posted above in today's case!) shows the last 3 beats in the long lead rhythm strip ( = beats #12, 13 and 14 in ECG #1) in each of the 4 lead groups. With this type of system — only 3 beats are seen in the 12 lead ECG!
• 
  
• As I noted earlier — 12 beats are seen in the long lead rhythm strip for ECG #1, so we do have ample opportunity to get a true "feel" for the regularity of the rhythm. But we are not able to look at simultaneously-recorded leads for 9 of these 12 beats. The obvious disadvantage of this type of ECG recording system — is that we have no idea of what beat #1 looks like in leads I,II,III — nor of what beat #4 looks like in leads aVR,aVL,aVF.

To EMPHASIZE — Clinically, this does not matter for ECG #1 — because the overall rhythm is sinus, and there are only 2 non-sinus beats.

• I interpreted beats #1 and #4 in the long lead II rhythm strip of ECG #1 as PJCs (Premature Junctional Contractions) — because they are not preceded by P waves, and the QRS of beats #1 and 4 is narrow (albeit somewhat different in morphology from sinus-conducted beats — probably due to some aberrant conduction).
• 
  
• Editorial Comment: Among the literally thousands of ECGs I've interpreted on international ECG forums over the years — I have not encountered this type of monitoring system often. I hate this system. I find it totally confusing — because it is usually not at all apparent which (if any) beats on the rhythm strip correspond to which beats on the 12-lead ECG. Simply stated — this system negates the invaluable assistance simultaneous recording of all beats on a rhythm strip can provide for determining the etiology of an arrhythmia.
• 
  
• So WHY make a big deal about this recording system? The answer to this question lies with interpretation of ECG #6, that is shown in Figure-1.

Regarding the Rhythm in ECG #6:

Starting with the long lead II rhythm strip in ECG #6 — it should be apparent that every-other-beat occurs early!

• Once again — it should be clear that the long lead II rhythm strip in ECG #6 is not a "simultaneous" recording. Given that there are at least 3 different QRS morphologies in this long lead II (some of which manifest QRS widening) — I initially had no idea as to what these beats were. For interpretation of ECG #6 — it is important to appreciate the type of ECG recording system being used!
• 
  
• Once again — it is only the last 3 beats in the long lead rhythm strip ( = beats #12,13,14 of ECG #6) — that are shown simultaneously in the 12-lead ECG. These same last 3 beats are actually seen in each of the 12 leads! That is — While the long lead II rhythm strip in ECG #6 shows 14 beats — only 3 of these beats (#12,13,14) are seen in the 12 lead tracing.

Applying this Information to ECG #6:

The most characteristic of the widened early beats is the 2nd QRS complex in lead V1 of ECG #6. This rsR' complex, with small initial r wave — followed by S wave that descends below the baseline — finishing with a slender taller "right rabbit ear" R' deflection is absolutely characteristic of supraventricular conduction with RBBB!

• To Emphasize — this 2nd QRS complex in lead V1 — corresponds to beat #13 in the long lead II rhythm strip! 
• The predominantly negative QRS of beat #13 in the long lead II rhythm strip — is consistent with LAHB conduction. That LAHB conduction is present — is confirmed, because the 2nd QRS complex in both leads II and III — and the 2nd QRS complex in lead aVF all manifest LAHB morphology.
• Additional support that a supraventricular pattern of RBBB conduction is present in ECG #6 — is forthcoming from noting that the 2nd QRS complex in lead V6 (which corresponds to beat #13 in the long lead II rhythm strip) — manifests the wide terminal S wave that is characteristic of RBBB.

Putting It All Together:

Knowing that the underlying rhythm in ECG #1 was sinus with some PJCs — helps us greatly in deducing what is probably going on in ECG #6.

• NOTE: Interpretation of the rhythm in ECG #6 is especially difficult from a technical perspective — because there is LOTS of baseline artifact — and, because the tracing is slightly slanted and therefore distorted (so that R-R intervals and coupling intervals are not consistent).
• 
  
• That said — the overall rhythm in the long lead II of ECG #6 manifests bigeminy (ie, every-other-beat occurs earlier than expected). Since we demonstrated that beat #4 in ECG #1 was a PJC — it is most likely that every-other-beat in ECG #6 is also a PJC. 
• 
  
• BOTTOM LINE: The rhythm in ECG #6 is most probably Junctional Bigeminy — in which every-other-beat manifests some combination of RBBB with/without LAHB aberration (accounting for the varying width and varying morphology of beats #1,3,5,7,9,11 and 13 in the long lead II rhythm strip).
• 
  
• Relevance to Today's Case: Given that all early beats are supraventricular (ie, PJCs) — we can reliably assess ST-T wave morphology of these beats. The marked ST elevation of the 2nd QRS complex in leads V2-thru-V5 (with T-QRS-D = Terminal QRS Distortion in V2,V3,V4) — is absolutely diagnostic of acute LAD OMI.