Article, Cardiology

Differentiating type 1 and 2 acute myocardial infarctions using the N-terminal pro B-type natriuretic peptide/cardiac troponin T ratio

a b s t r a c t

Purpose: Differentiation of type 1 (T1MI) from type 2 myocardial infarction (T2MI) is important as recommendED treatments for each differ. Patients with T2MI may have more/earlier cardiac wall stress resulting in an increased N-Terminal pro B-type natriuretic peptide (NT-proBNP)/cTnT generation 5 ratio (cTnT Gen 5).

Methods: Emergency Department (ED) patients presenting with symptoms suspicious for acute coronary syn- drome (ACS) were enrolled from 2013 to 2015. Baseline blood samples were collected within 60 min of a triage ECG, with additional draws at 30, 60 and 180 min. NT-proBNP and cTnT Gen 5 levels were measured later in an independent laboratory. Acute myocardial infarction (AMI) was adjudicated using the Third Universal Definition of Myocardial Infarction.

Results: 575 patients were enrolled with 44 (7.7%) having AMI [25 T1MI (59.1%) and 18 T2MI (40.9%)]. Patient characteristics showed very few AMI type differences so accurate clinical differentiation was difficult. The median NT-proBNP/cTnT Gen 5 ratios were significantly higher in T2MI when compared to T2MI at baseline and 30, 60 and 180 min later [7.3 v 53.0 (p = 0.003), 5.8 v 49.5 (p = 0.002), 6.3 v 47.5 (p = 0.003) and 4.3 v 33.7 (p =

0.016) respectively]. Conclusions: The clinical determination of whether an AMI is type 1 or 2 is difficult as the ED patient characteris- tics of each are similar. The NT-proBNP/cTnT Gen 5 ratio can aid in making this differentiation. Additional multi- center trials are needed to validate our results.

(C) 2018

Introduction

The Third Universal Definition of Myocardial Infarction describes 2 types of Acute Myocardial Infarction (AMI) that are commonly seen in Emergency Department (ED) patients. Type 1 (T1MI) is relat- ed to atherosclerotic Plaque rupture, ulceration, fissuring, erosion or dissection with intraluminal thrombus in one or more of the coro- nary arteries leading to decreased Myocardial blood flow or distal

* Corresponding author at: Emergency Department, Henry Ford Health System, 2799 West Grand Boulevard, Detroit, MI 48202, USA.

E-mail addresses: [email protected] [email protected] (R.M. Nowak), [email protected] (G. Jacobsen), [email protected] (R.H. Christenson), [email protected] (M. Moyer), [email protected] (M. Hudson), [email protected] (J. McCord).

platelet emboli with ensuing myocyte necrosis. Type 2 (T2MI) is re- lated to a condition other than coronary artery disease that contrib- utes to an imbalance between myocardial oxygen supply and/or demand resulting in myocardial injury with necrosis [1]. Examples of the latter include tachycardia/bradycardia, significant hyper/hy- potensive episodes, acute anemia or hypovolemia, exacerbations of congestive heart failure, pulmonary embolus, renal failure, sepsis and other conditions, often in the setting of stable coronary artery disease. T2MI has been reported to be present in approximately one quarter of patients diagnosed with AMI [2].

There is ongoing debate as to whether T2MI should not be called an AMI but rather be labelled myocardial injury as it is not caused by the occlusion of a coronary artery [3], Regardless, Emergency Medicine (EM) physicians will continue to see patients with a rise and/or fall of cardiac troponin (cTn) values above the 99th percentile reference of a normal population (AMI diagnosis) and therapeutic decisions will

https://doi.org/10.1016/j.ajem.2018.06.073 0735-6757/(C) 2018

need to be made. With the coming approval by the US Food and Drug Administration of additional higher sensitivity troponin assays it is ex- pected that there will be additional T2MIs diagnosed. While currently there are no formal guidelines available for treating T2MI most clini- cians would agree that therapy should be directed to the underlying eti- ology and to correct the altered variable within the existing supply/ demand imbalance. However if the clinician is unsure of the type of AMI present should some individuals also receive aspirin, nitrates, anti- thrombotic therapy, undergo coronary angiography etc. and at what time? A current “gold standard” for the diagnosis of T2MI remains unde- termined and so making this diagnosis, especially in patients with existing comorbidities (such as coronary artery disease), can be quite difficult.

There have been a few reports to-date that have detailed some of

the clinical characteristics differentiating T1MI from T2MI. T2MI pa- tients are reported to more likely to be female, older, have more comorbities including hypertension, congestive heart failure, im- paired renal function, anemia, atrial fibrillation and malignancy, present with tachycardia and have more abnormal but non- ischemic ECGs [4-6]. The presenting symptoms are very similar for each type of AMI and cTn values tend to be lower in those with T2MI [7]. However these studies have included ST-segment eleva- tion myocardial infarction (STEMI) patients as T1MIs in proportions ranging from 17.6% to 32.9% [4, 5]. From an EM perspective individ- uals with STEMI are diagnosed quickly by ECG changes and then rap- idly transported to the cardiac catheterization lab for immediate coronary angiography. The challenge for EM physicians is to as accu- rately as possible differentiate T1MI from T2MI within the diagnosis of non St-segment elevation myocardial infarction (NSTEMI). While there are no data available regarding differential shorter term NSTEMI type clinical outcomes (during the index hospitalization or at 30 days), patients with T2MI are reported to have a significantly higher mortality rate at 2 years of approximately 50% compared to 26% for those with T1MI [8].

An objective measurement that could help reliably distinguish between patients with T1MI and T2MI is needed. Patients with T2MI have a preexisting non AMI acute clinical condition present that causes a myocardial supply/demand mismatch which results in the rise and fall of cTn resulting in the clinical diagnosis of AMI. Natriuretic peptides, B-type natrurietic peptide (BNP) and its amino-terminal pro-peptide cleavage equivalent, N-terminal pro-B type natrurietic peptide do aid in the diagnosis of acute heart failure [9-11]. We hypothesize that patients with T2MI will generally have more and earlier cardiac wall stress resulting in higher levels of na- triuretic peptide release than those with a T1MI, resulting in an in- crease in the natriuretic peptide/cTn ratio.

The objectives of this study were to determine if the N-terminal pro B-type natriuretic peptide (NT-proBNP)/generation 5 Cardiac troponin T (cTnT Gen 5) ratio measured at baseline, 30, 60 and 180 min in The Rapid Evaluation of ACuTe myocardial InfarctiON in the United States (REACTION-US) trial could aid in the differentiation of T1MI from T2MI in ED patients in the United States (US).

Methods

Study design, setting and participants

The REACTION-US study was an observational trial that enrolled ED patients with any symptoms suspicious of possible acute coro- nary syndrome (ACS) at a single tertiary US hospital. The inclusion criteria were: patients N21 years of age presenting with any symp- toms suspicious for ACS and who were able to comply with all as- pects of the protocol and had a triage electrocardiogram (ECG) completed. Exclusion criteria included: patients with acute distress requiring immediate Life-saving interventions, cardioversion/defi- brillation or thrombolytic therapy within the prior 24 h, STEMI

leading to immediate reperfusion therapy, traumatic injuries, trans- fers from other facilities and patients who were pregnant or breast feeding. Study approval was obtained by the Institutional Review Board at Henry Ford Hospital and patients gave written informed consent prior to enrollment. The study was conducted in accordance with the principles of the Declaration of Helsinki and International Conference of Harmonization good clinical practice guidelines.

Subjects underwent clinical assessment by an EM physician which included a history, physical examination, review of ECGs, standard blood testing including hospital cTn measurements and chest X-rays. The timing and implementation of these tests, treat- ments and dispositions were decided by the treating physician.

Blood measurements

Study baseline blood samples were collected in ethylenediamine tetraAcetic acid (EDTA) tubes within 60 min of the completion of a triage ECG (within 10 min of ED presentation). Additional Blood draws were obtained at 30 (+-10), 60 (+-15) and 180 (+-30) minutes after the base- line sample. Each blood tube was centrifuged to obtain plasma and then stored at -80 ?C within 1 h of draw. After completion of the study stored samples were sent to a core lab at the University of Maryland for independent and blinded batch measurements of both NT-proBNP and cTnT Gen 5 at each time interval. The frozen samples were thawed a single time.

Nt-proBNP levels were measured using an ElecSys cobas e 601 system (Roche Diagnostics, Indianapolis, Indiana) in accordance with the manufacturer’s recommendations for all NT-proBNP mea- surements. The precision of this method, indicated by the total Coef- ficient of Variation was 3.0% at 150 pg/ml, 2.5% at 1000 pg/ml and 2.0% at 7000 pg/ml. CTnT measurements were conducted using the Roche Diagnostics Elecsys Troponin T Generation 5 assay on a Roche cobas e 601 system. This Generation 5 TnT assay is cleared by US-FDA.: the overall 99TH reference upper reference for both sexes is 19 ng/L; the assay’s limit of detection and the limit of quan- tification (LoQ) is 6 ng/L [12].

Outcomes

Adjudication of the final diagnosis of T1MI and T2MI was per- formed by a board certified cardiologist and EM physician with addi- tional review by another board certified cardiologist in the event of disagreement. AMI diagnosis was determined by following the Third Universal Definition of Myocardial Infarction [1] using the hos- pital reported cardiac troponin I assay values (Siemens Centaur sys- tem TnI Ultra assay on a Centaur XP analyzer, 99th percentile 40 ng/L) and all available clinical data for 30 days after ED presenta- tion. The adjudicating physicians were blinded to the study biomark- er measurements.

The specific etiology of the supply/demand mismatch causing a T2MI varied but was not recorded by the adjudicating physicians.

After discharge from the hospital follow-up by study personnel was done within 30-45 days and 12-18 months to determine if the patient had suffered a major adverse cardiac event: death, AMI, or re- vascularization (percutaneous coronary intervention or coronary ar- tery bypass surgery). Follow-up information was obtained from telephone call and/or subsequent medical record review. If subjects and/or family members were not able to be reached by telephone an electronic medical record review and death registry search was completed.

Statistical analysis

The baseline patient characteristics were compared between the T1MI and T2MI patients using two-sample t-tests for normally distribut- ed numerical data, Wilcoxon rank sum tests for non-normally

distributed numerical data, Chi-square tests for non-sparse categorical data, and Fisher exact tests for sparse categorical data. At each of the 4 evaluation time points (baseline, 30, 60, and 180 min), the NT-proBNP and cTnT Gen 5 levels and the NT-proBNP/cTnT Gen 5 ratio data was compared between T1MI and T2MI patients using the Wilcoxon rank sum test. For the comparison test results, p-values b 0.05 were consid- ered statistically significant.

At each of the 4 time points, a receiver operating characteristic

(ROC) curve was obtained to evaluate the NT-proBNP/cTnT Gen 5 ratio as a predictor of T2MI relative to T1MI. The area under each ROC curve was obtained along with the corresponding 95% confi- dence intervals. The optimal NT-proBNP/cTnT Gen 5 ratio cut point was calculated as the maximum sum of the sensitivity plus specific- ity. For each optimal cut point, the resulting sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were calculated along with their 95% confidence intervals.

There were no Power calculations done in advance of the analysis as this was an exploratory, preliminary study concerning the use of a novel biomarker ratio as a possible aid in the differentiation of T1MI from T2MI.

Results

Characteristics of study patients

From May 2013 to April 2015 1699 patients were screened for this study but not enrolled. Reasons for non-enrollment were: 338 could not be consented and have the baseline blood draw completed within 1 h of the triage ECG, 226 were impaired (substance abuse or mental health issues), 86 had been previously enrolled, 327 refused to partici- pate and 147 were not enrolled for other reasons. We thus enrolled 575 subjects and 569 had suitable available data for inclusion in the analysis (4 patients had no clinically ordered cTnI measurements and 2 subjects had been enrolled twice under different identities, in which case only the first enrollment was used).

The adjudicated final diagnosis for all patients with at least 1 hospital

contemporary cTn I level reported above the 99th percentile (116 or 20.4%) was AMI in 44 patients (7.7%). Of these 1 had a STEMI that devel- oped after presentation (2.3%) while 25 had NSTEMI T1MI (56.8%) and 18 T2MI (40.9%). There was good interrater reliability in regards to the diagnosis of NXSTEMI with concordant final diagnosis amongst the

Table 1

Acute myocardial infarction (AMI) patient clinical characteristics.

Characteristics

All AMI (n = 44)

T1MI (n = 26)

T2MI (n = 18)

p-Value

Age (years)

59.8 +- 9.3

60.6 +- 9.5

58.6 +- 9.3

0.490 (T)

Male gender

27 (61.4%)

18 (69.2%)

9 (50.0%)

0.198 (C)

African American race

36 (81.8%)

22 (84.6%)

14 (77.6%)

0.697 (F)

Medical history

Hypertension

41 (93.2%)

24 (92.3%)

17 (94.4%)

1.000 (F)

Diabetes

15 (34.1%)

9 (34.6%)

6 (33.3%)

0.930 (C)

Hypercholesterolemia

29 (65.9%)

17 (65.4%)

12 (66.7%)

0.930 (C)

Smoking

20 (45.5%)

11 (42.3%)

9 (50.0%)

0.614 (C)

Personal history CAD

28 (63.6%)

19 (73.1%)

9 (50.0%)

0.118 (C)

Family history CAD

24 (54.5%)

13 (50.0%)

11 (61.1%)

0.467 (C)

PCI

25 (56.8%)

18 (69.2%)

7 (38.9%)

0.046 (C)

CABG

5 (11.4%)

3 (11.5%)

2 (11.1%)

1.000 (F)

Myocardial infarction

24 (54.5%)

15 (57.7%)

9 (50.0%)

0.614 (C)

Congestive heart failure

11 (25.0%)

5 (19.2%)

6 (33.3%)

0.314 (F)

Onset of symptoms Hours to presentation

45.7 +- 99.9

63.3 +- 126.3

20.4 +- 25.9

0.193 (W)

Presenting vital signs Systolic BP (mm Hg)

150.3 +- 32.6

147.2 +- 28.7

154.8 +- 38.0

0.451 (T)

Diastolic BP (mm Hg)

89.9 +- 22.7

90.0 +- 20.0

89.8 +- 26.8

0.971 (T)

Heart rate (beats/min)

90.5 +- 25.5

86.4 +- 24.0

96.4 +- 27.1

0.204 (T)

Electrocardiogram

Atrial fibrillation tachycardia

3 (6.8%)

0 (0.0%)

3 (16.7%)

0.062 (F)

Sinus tachycardia

11 (25.0%)

7 (26.9%)

4 (22.2%)

1.000 (F)

Left ventricular hypertrophy

10 (22.7%)

5 (19.2%)

5 (27.8%)

0.716 (F)

ST segment depression >=1 mm

9 (20.5%)

6 (23.1%)

3 (16.7%)

0.716 (F)

T wave inversion

25 (56.8%)

12 (46.2%)

13 (72.2%)

0.086 (C)

Within normal limits

4 (9.1%)

3 (11.5%)

1 (5.6%)

0.634 (F)

Home medications Aspirin

34 (77.3%)

21 (80.8%)

13 (72.2%)

0.716 (F)

Anti-coagulant

3 (6.8%)

1 (3.8%)

2 (11.1%)

0.558 (F)

Diuretics

9 (20.5%)

3 (11.5%)

6 (33.3%)

0.128 (F)

ACE inhibitor

26 (59.1%)

15 (57.7%)

11 (61.1%)

0.821 (C)

Beta blocker

28 (63.6%)

17 (65.4%)

11 (61.1%)

0.772 (C)

Calcium channel blocker

15 (34.1%)

8 (30.8%)

7 (38.9%)

0.576 (C)

Nitrates

16 (36.4%)

11 (42.3%)

5 (27.8%)

0.325 (C)

Presenting symptoms

Chest pain primary

41 (93.2%)

26 (100.0%)

15 (83.3%)

0.062 (F)

Pressure/tightness/heaviness

30 (68.2%)

21 (80.8%)

9 (50.0%)

0.031 (C)

Crushing chest pain

7 (15.9%)

7 (26.9%)

0.0 (0.0%)

0.031 (F)

Left arm/shoulder pain

23 (52.3%)

18 (69.2%)

5 (27.8%)

0.007 (C)

Dizzy/lightheaded

18 (40.9%)

7 (26.9%)

11 (61.1%)

0.023 (C)

Shortness of breath

28 (63.6%)

13 (50.0%)

15 (83.3%)

0.024 (C)

Numerical data is given as mean +- standard deviation. Categorical data is given as frequency (percent) measurements. (T) = Two-Sample T-Test, (W) = Wilcoxon Rank Sum Test, (C) = Chi-Square Test, (F) = Fisher Exact Test. CAD = Coronary artery disease, PCI = Percutaneous coronary intervention, CABG = Coronary artery bypass graft, BP = Blood pressure, ACE = Angiotensin converting enzyme.

adjudicating cardiologist and EM physician with the need for a second cardiologist review required in only 2 cases (1.8%).

Analyses of T1MI and T2MI patient demographic/clinical/comorbid- ity/presenting vital signs/symptom characteristics are shown in Table 1. The few patient characteristics significantly differing between the 2 types of NSTEMI were increased history of revascularization, percutane- ous coronary intervention and symptoms of pressure/tightness/heavi- ness, crushing chest pain and left arm/shoulder pain in T1MI and increased symptoms of dizzy/lightheaded and shortness of breath in T2MI. Many other symptoms such as alternative descriptions/radiations of the chest pain/discomfort and symptoms of nausea, vomiting, palpi- tations, diaphoresis and feeling weak were not significantly different in the two types of NSTEMI. Of the 4 enrolled patients receiving hemo- dialysis 2 had a T1MI and two T2MI (p = 1.000).

Overall, of all study patients, 109 were not able to be contacted by telephone call for the 30 day follow up. All patients had their electronic Medical records reviewed. Another 77 had no Henry Ford Health Sys- tem visits or testing done within this time period and so they could have been considered as lost to follow up (LTFU). However in a later fol- low up at 12-18 months only 14 of these 77 patients (overall 14/569 or 0.02%) remained as LTFU. At the 30 day follow up cardiac death occurred in 1 of the 26 (3.8%) T1MIs and in none of the 18 (0.0%) T2MIs (p = 1.00). There were no additional cardiac deaths found between the 30 day and the 1 year follow-up in either group. There were no non- cardiac deaths within the 30 day or 1 year follow-up in both types of NSTEMIs.

Main results

The T1MI patients had higher cTnT Gen 5 values at all four sampling time points but were not significantly different from those values

reported in individuals with T2MI. The NT-proBNP median levels (pg/ ml) alone were significantly higher in T2MI than T1MI at baseline (1792.0 v 393.6, p = 0.022), 30 (1896.5 v 409.3, p = 0.026), 60

(1710.0 v 253.5, p-0.020) but not at 180 min (1047.1 v 362.4, p =

0.134).

The serial NT-proBNP/cTnT Gen 5 ratios at baseline and 30, 60 and 180 min are shown in Fig. 1. As shown at all serial time points the me- dian ratios were significantly higher in the individuals with a T2MI, al- though there was some overlap. In both T1MI and T2M2 patients at all blood draws there were a few outliers with NT-proBNP/cTnT Gen 5 ra- tios of 100 or greater, more often but not always having a T2MI. Addi- tionally the median NT-proBNP/cTnT Gen 5 ratio changes from 0 to 30, 30-60 and 60-180 min are shown in Fig. 2. These median ratio changes were not significantly different between T1MI and T2MI pa- tients between these time intervals. Overall the NT-proBNP/cTnT Gen 5 ratios were more accurate in differentiating between the two types of NSTEMI than the NT-proBNP levels alone at all 4 serial sampling time points.

Receiver operating curves (ROC) were plotted, the area under the curves (AUC) measured and optimal cut points determined for the abil- ity of the NT-proBNP/cTnT Gen 5 ratio to differentiate T1MI from T2MI. Additionally the sensitivity, specificity, PPV and NPV of these optimal cut points to differentiate these two types of NSTEMI were determined. These overall results are detailed in Table 2. The AUC at all time points (ranging from 0.745-0.777) was moderate for the differentiation of T1MI from T2MI.

Discussion

Our report is the first to show that there are very few differences in the ED clinical characteristics differentiating patients specifically with

Fig. 1. NT-proBNP/cTnT Gen 5 median ratios for T1MI and T2MI at all sampling periods.

Fig. 2. NT-proBNP/cTnT Gen 5 median ratio changes for T1MI and T2MI between all sampling periods.

type 1 from type 2 NSTEMI (other published reports have also included STEMI patients as part of the T1MI cohort and the subsequent analysis). We have reported on the usefulness of the NT-proBNP/cTnT Gen 5 ratio at 4 different time points in the ED as an additional objective aid in help- ing make this difficult differentiating diagnostic decision.

Age, gender, race, medical history/comorbidities (with the exception

of history of revascularization and percutaneous coronary interven- tion), home medications, initial vital signs and ECG changes were not significantly different in the 2 types of NSTEMI. Twenty two different patient chest discomfort/pain symptoms were recorded and of these only the more classic complaints of pressure/tightness/heaviness/ crushing pain chest and left arm shoulder radiation were found signifi- cantly more often in those with T1MI. Of 7 non chest related complaints, only complaints of dizzy/lightheaded and shortness of breath were sig- nificantly more common in those with T2MI. While giving some clinical direction as to whether an individual might have a T1MI or T2MI the pa- tient characteristics of each NSTEMI type did not reliably allow differen- tiation between them. Additionally there was no difference in the 30 day and 1 year follow-up cardiac or non-cardiac deaths between them. This is most likely related to the relatively low numbers of N STEMIs included in our study.

T1MI patients had increased cTnT Gen 5 measurements at all 4 blood draws but these were not significantly higher than those with T2MI, limiting the value of cTn alone in helping to make the correct NSTEMI type diagnosis in the ED within the first 3 h of presentation. Additionally the initial median creatinine was 1.0 mg/dL in those with a T1MI and

1.3 mg/dL in T2MI (p = 0.095) suggesting a trend toward higher renal dysfunction in T2MI patients. Patients with T2MI had significantly higher NT-proBNP levels for the first three blood draws only but these were less significantly different than the NT-proBNP/cTnT Gen 5 ratios were at the same time points.

The median NT-proBNP/cTnT Gen 5 ratios were significantly higher at all 4 blood draws in patients with a T2MI (Fig. 1). This indicates that an EM physician could have these biomarker results reported and the ratio calculated at any time within the first 3 h of ED presentation and they would all similarly aid in the determination of STEMI type. There was some overlap in the ratios for the type including a few outliers with ratios of 100 or greater. Further individual patient reviews are needed to try to determine why these outliers might have occurred. Whether it would be best for all patients suspected of having a possible ACS have both study biomarkers measured initially or only after a cTnT Gen 5 value above the 99th % is reported to the treating physician

Table 2

Receiver operating curve results using the NT-proBNP/cTnT Gen 5 ratio for predicting type 1 from type 2 AMI in the REACTION-US study.

Time

Area under the ROC Curve

Optimal cut

Sensitivity for Type 2

Specificity for Type 2

PPV for Type 2

NPV for Type 2

point

(95% CI)

point

(95% CI)

(95% CI)

(95% CI)

(95% CI)

Baseline

30 min

0.765

(0.611 to 0.919)

0.776

14.8

13.0

88.9

(65.3 to 98.6)

88.9

65.4

(44.3 to 82.8)

69.2

64.0

(42.5 to 82.0)

66.7

89.5

(66.9 to 98.7)

90.0

(0.627 to 0.925)

(65.3 to 98.6)

(48.2 to 85.7)

(44.7 to 84.4)

(68.3 to 98.8)

60 min

0.777

(0.621 to 0.933)

26.0

70.6

(44.0 to 89.7)

87.5

(67.6 to 97.4)

80.0

(51.9 to 95.7)

80.8

(60.6 to 93.4)

180 min

0.745

(0.568 to 0.922)

11.0

85.7

(57.2 to 98.2)

66.7

(43.0 to 85.4)

63.2

(38.3 to 83.7)

87.5

(61.7 to 98.5)

Note: The optimal cut points are defined as the maximum sensitivity plus specificity.

remains to be determined. The ratio changes from the baseline to 30, 30-60 and 60-180 min intervals were not significantly different (Fig. 2) from one another and so these interval ratio changes were not useful in differentiating the NSTEMI type diagnosis.

The AUC (moderate) of the 4 ROC curves and the optimal cut points at each of the four blood draws for predicting whether an NSTEMI is a T1MI or T2MI are shown in Table 2, as well as the respec- tive sensitivity, specificity, PPV and NPV values. These values, while imperfect, are similar to those reported for the ability of BNP (at a cut point of 100 pg/ml) [11] and NT-ProBNP (at a cut point of 300 pg/ml) [9] to aid in the determination of Acute heart failure as the cause of acute dyspnea in the ED (sensitivity 90%and 92.1%, specificity 76% and 79.6%, PPV 79% and 51.4%, NPV 89% and 97.7% respectively). Some of these cut point values have been em- bedded as Class 1, Level of evidence: A, in clinical practice guidelines [13]. The best clinical use of our NT-proBNP/cTnT Gen 5 ratio data ap- pears to be, as with the natriuretic peptide use for the rule-out of AHF, to exclude T2MI at blood draws from baseline to 180 min later (optimal cut points and NPVs at 0, 30, 60 and 180 min are 14.8,

89.5%; 13.0, 90.0%;. 26.0, 80.8%; 11.0, 87.5% respectively).

Limitations

The REACTION-US trial involved a single urban center with relatively lower numbers of patients studied and so our results will require valida- tion in a prospective multi-center US trial. Our data were obtained in an observational diagnostic study in which the cTnT Gen 5 results were compared against an adjudicated AMI diagnosis that relied on the use of a contemporary cTn I assay. Perhaps had the adjudicating physicians used the cTnT Gen 5 assay results to determine the presence of NSTEMI the results may have been different. There were 1124 patients screened for study enrollment but 337 (30.0%) refused to participate and so there may have been some patient selection bias. However once consented and enrolled our rate of patients unqualified for inclusion was very low (6 of 575, 1.0%).

As we excluded those with life-threatening illness, cardioversion or defibrillation within 24 h of presentation, STEMI patients requiring im- mediate reperfusion or those who were pregnant or breast feeding we do not know how any of our analyses would have changed had these patients been included. As the majority of our enrolled patients overall and those with NSTEMI were African American these results could be different when studied in other races. Lastly we have used NT-proBNP as the natriuretic peptide in this study and do not know whether the use of BNP as an alternative biomarker could make the ratio prediction of the NSTEMI type better or worse.

Conclusions

The recommended treatments and inpatient disposition of NSTEMI T1MI and T2MI are currently different and so accurate diagnosis is im- portant. The determination of whether a NSTEMI is a type 1 or type 2 using all current available ED clinical data is difficult and unreliable as the patient characteristics in each group are similar. The NT-proBNP/ cTnT Gen 5 ratio measured at baseline or after 30, 60 or 180 min can aid in the exclusion of T2MI when used in conjunction with all other available clinical information. Future multicenter trials are needed to verify these results and to determine how the NT-proBNP/cTnT Gen 5 ratio might be best clinically used in differentiating whether a NSTEMI is a type 1 or type 2.

Presentations at meetings

One abstract of the results was accepted for presentation (poster) at the 2018 American College of Cardiology meeting in Orlando, Florida

and a second for presentation at the 2018 Society for Emergency Medi- cine Meeting in Indianapolis, Indiana.

Funding sources

The REACTION-US study was funded and sponsored by Henry Ford Health System Detroit, Michigan. Roche Diagnostics (Indianapolis, Indiana) provided support for independent external measurement of N-terminal B-type Natriuretic Peptide and Cardiac Troponin T Genera- tion 5 (University of Maryland) as well as for data analysis (done by Henry Ford Health System).

Author contributions

RMN: Study design, implementation and over sight, data analysis, primary author.

GJ: Statistical analysis, manuscript review and input. RHC: Independent hs-cTnT measurements, manuscript review and input. MM: Study manager, data analysis, manuscript review and input. MH: Study design, data analysis, manuscript review and input.

JM: Study design, data analysis, manuscript review and input.

Conflict of interest disclosure

Richard Nowak MD has received research support as part of REACTION-US from and is a consultant for Roche Diagnostics.

Robert H Christenson PhD reports personal fees from Roche Diag- nostics outside the submitted work.

James McCord MD has received research support as part of REACTION-US from and is a consultant for Roche Diagnostics.

Gordon Jacobsen MS, Michele Moyer MS BSN and Michael Hudson MD have no conflicts of interest to report.

References

  1. Thygesen K, Alpert JS, Jaffe AS, Simoons ML, Chaitman BR. White HD: the writing group on behalf of the joint WSC/ACCF/AHA/WHF task force for the universal defini- tion of myocardial infarction. Circulation 2012. https://doi.org/10.1161/CIR. 0b013e31826e1058.
  2. Sandoval Y, Smith SW, Thordsen SA, Apple FS. Supply/demand type 2 myocardial in- farction: Should we be paying more attention. J Am Coll Cardiol 2014;63:2079-97.
  3. Nagele P. The case for a revised definition of myocardial infarction-resolving the am- biguity of type 2 myocardial infarction. JAMA Cardiol 2016;1:247-8.
  4. Lopez-Cuenca A, Gomez-Molina M, Flores-Blanco PJ, et al. Comparison between type 2 and type 1 myocardial infarction: clinical features, treatment strategies and out- comes. J Geriatr Cardiol 2016;13:15-22.
  5. Greenslade JH, Adikari T, Mueller C, et al. Characteristics and occurrence of type 2 myocardial infarction in emergency department patients: a prospective study. Emerg Med J 2017;0:1-8.
  6. Saaby L, Poulsen TS, Hosbond S, et al. Classification of myocardial infarction: fre- quency and features of type 2 myocardial infarction. Am J Med 2013;126:789-97.
  7. Alpert JS, Thygesen KA, White HD, Jaffe AS. Diagnostic and Therapeutic implications of type 2 myocardial infarction: review and commentary. Am J Med 2014;127:105-8.
  8. Saaby L, Poulsen TV, Diederichsen ACP, Hosbond S, et al. Mortality rate in type 2 myocardial infarction: observations from an unselected hospital cohort. Am J Med 2014;127:295-302.
  9. Januzzi Jr JL, Chen-Tournoux AAC, Christenson RH, et al, on behalf of the ICON- RELOADED investigators. N-terminal pro-B type natriuretic peptide in the emergen- cy department. J Am Coll Cardiol 2018;71:1191-200.
  10. Chen AA, Wood MJ, Krauser DG, et al. NT-proBNP levels, echocardiographic findings and outcomes in breathless patients: results from the ProBNP Investigation of Dys- pnea in the Emergency Department (PRIDE) echocardiographic substudy. Eur Heart J 2006;27:839-45.
  11. Maisel AS, Krishnaswamy P, Nowak RM, et al. For the breathing not properly multi- national study investigators. Rapid measurements of B-type natriuretic peptide in the emergency department. N Engl J Med 2002;347:161-7.
  12. https://www.accessdata.fda.gov/cdrh_docs/pdf16/K162895.pgf.
  13. CW Yancy, Jessup M, Bozkurt B, et al. ACCF/AHA guideline for the management of heart failure: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines. Circulation 2013;128: e240-327 (2013).

Leave a Reply

Your email address will not be published. Required fields are marked *