Copeptin for Early Diagnosis of Myocardial Infarction

Abstract

Background: Early identification of myocardial infarction is crucial in chest pain patients. Objective: Our goal was to evaluate if copeptin allows the ruling out of myocardial infarction (AMI) when used with cardiac troponin Ic (troponin) in the emergency department. Methods: Patients with suspected acute coronary syndrome (ACS) were enrolled in this pilot trial. Copeptin and troponin were determined at admission, and 6 h later for troponin. Results: A total of 125 patients with a mean age of 60 ± 15 years were enrolled. Chest pain was present in 94 patients (75%) and other symptoms compatible with cardiac ischaemia in 25 (20%). Four groups were identified: 15 patients in group 1 had positive copeptin and positive troponin, 29 patients in group 2 had positive copeptin and negative troponin, 7 patients in group 3 had negative copeptin and positive troponin, and 74 patients in group 4 had both copeptin and troponin negative. Patients in group 1, 2 and 3 had a high incidence of STEMI and NSTEMI. In group 4, no patient experienced adverse events. Negative predictive value was 100% for the combined measurement of troponin and copeptin. A 3-month follow-up was completed without adverse events for patients of group 4. Conclusion: Combined determination of troponin and copeptin on admission in patients presenting early (<3 hours but >30 minutes) after onset of symptoms suspect of AMI provides a high negative predictive value. This combination might improve early rule-out of AMI and makes copeptin a useful complement to conventional troponin in the emergency department.

Identification of acute myocardial infarction (AMI) is crucial in chest pain patients to quickly initiate invasive treatment. Diagnosis of AMI is based on clinical symptoms, electrocardiogram (ECG) findings, and biomarker levels. Markers of myocardial necrosis such as cardiac troponin and CK-MB are the gold standard in detection of cardiac necrosis, and their use is recommended by current guidelines [1]. Cardiac troponin Ic (troponin) provides high specificity. The delayed release of markers explains some weakness in the diagnostic performance of conventional troponin assays very early after chest pain onset [2]. Therefore, markers with pathophysiological background independent of cell necrosis might improve rapid diagnosis of AMI. Current management of acute chest pain patients includes electrocardiographic evaluation and sequential specific cardiac enzyme measurements. However, in as many as one third of presentations, ECGs do not allow diagnosis of AMI [3], and a delay of 3 hours after the onset of symptoms for ultrasensitive troponin, and 6 to 9 hours for classical troponin is required to achieve sufficient sensitivity [1].

Copeptin is a new, recently tested biomarker of cardiac ischaemia and necrosis. It is a precursor of vasopressin, secreted in various stress situations, such as sepsis, pneumonia, chronic obstructive pulmonary diseaase (COPD) decompensation, heart failure, and AMI [4]. Copeptin is stable, easily measurable, and not rapidly cleared [5].

During AMI, copeptin rises early after the onset of symptoms and normalises within 10 hours. On the contrary, classical cardiac biomarkers (troponin and creatine kinase [CK-MB]) rise later with peak values 14 to 16 hours after onset of symptoms [2]. The aim of the current investigation was to test prospectively whether copeptin adds diagnostic informations to troponin and whether the combination of copeptin and troponin is superior to troponin alone in the early management of patients with suspected acute coronary syndrome (ACS).

Method

All patients with suspected ACS presenting to the emergency department of our hospital were enrolled in this study until the predetermined number of patients had been reached. Patients older than 18 years and younger than 85 years of age with angina pectoris or any other symptoms compatible with the diagnosis of AMI, such as chest pain, dyspnoea, epigastric pain or other abdominal symptoms, were eligible for inclusion in the trial. Exclusion criteria were trauma or major surgery within the last 4 weeks, pregnancy, intravenous drug abuse, and anaemia (haemoglobin 10 g/dl). All included patients gave their oral consent, as the local ethics committee did not require written consent. Of 130 patients screened, 2 were excluded (1%) and 3 (2%) refused to participate.

Diagnosis of AMI was established according to the universal definition of AMI [1], which comprised elevated biomarkers of myocardial necrosis together with at least one of the following criteria:

Patients with ymptoms compatible with myocardial ischaemia, ECG changes, or identification of intracoronary thrombus by coronary angiography were categorised as having AMI. ECG changes on admission suggesting AMI were: ST-segment elevation ≥0.2 mV in at least two contiguous leads from V2 to V6, ST-segment elevation 0.1 mV in any other leads or newly documented left bundle-branch block, as well as ST depression ≥0.05 mV in two contiguous leads, and/or T inversion >0.1 mV in two contiguous leads.

Troponin measurements

Troponin Ic was determined immediately using a 1-step enzyme immunoassay based on electrochemiluminescence (Beckman). Cut-off value was 0.09 ng/l. It was measured again when needed according to the local protocol of chest pain patient management (i.e., one troponin measurement at least 6 hours after the beginning of the symptoms).

Copeptin measurements

For copeptin measurements, blood was drawn at admission, directly frozen and kept at –86 °C. It was measured in EDTA plasma by use of TRACE^® immunoassay (Kryptor, B.R.A.H.M.S AG, Hennigsdorf, Germany) as described in detail elsewhere [17,20]. The assay has an analytical detection limit of 0.9 pmol/l in a range of 0.9 to 2 000 pmol/l. Cut-off value was 12 pmol/l, as advised by the manufacturer (Thermofisher, B.R.A.H.M.S). The final diagnosis was made in the emergency department.

Follow-up

Follow-up was via phone calls, 1 and 3 months after inclusion; the occurrence of major adverse cardiovascular events (MACE, defined as cardiac death, AMI, coronary revascularisation and ischaemic stroke) and/or new hospitalisation was sought.

Any reported clinical event was reviewed by asking the patient or a close relative, and followed up by contact with the patient’s general practitioner or treating institution. Information regarding death and cause of death were obtained from the hospital register or family physician’s records. The primary endpoint was noncardiac death and MACE at the 3-month follow-up.

Calculation of sample size

We designed a preliminary trial with a sample size of 125 patients to determine the feasibility of a larger study in our centre. This number represents 1/8 of 1 002, the number of patients to be included in the trial. This is based on an error of 0.05, a sample error of 3% and a probability of event of 0.25. Taking into account a 25% attrition of patients, the sample size was calculated to be 1 002 patients.

Statistics

The qualitative variables are expressed as frequencies and percentages and the quantitative variables are expressed as means and standard deviations (SDs). Comparison of the qualitative variables was performed using the χ2-test and that of the quantitative variables was carried out using Student’s t-test (or the MannWhitney U-test if the distribution was not normal). The values of sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) were calculated with a cut-off value for copeptin at 12 pmol/l. A receiver operating characteristic (ROC) curve was constructed, the area under the curve (AUC) was calculated with a 95% confidence interval (95% CI) and the sensitivity, specificity, PPV and NPV for the different cut-offs were obtained. All analysis was conducted using SPSS, version 18.0 (SPSS Inc., Chicago, Illinois, USA).

Results

A total of 125 patients (83 males) with a mean age of 60 ± 15 years old were included. Eighty-four (67%) presented with chest pain, 36 (29%) with diffuse pain, 3 (2%) had abdominal pain, and 2 (2%) complained of dyspnoea. Demographic data of our middle risk population are presented in

Table 1. Baseline characteristics.

		All patients		AMI		Non-AMI
		n	%	n	%	n	%
Sex	M	83	66	14	67	69	66
	F	42	34	7	33	35	34
Cardiovascular risk factors	0	17	14	2	10	15	14
	1	44	35	8	38	36	34
	2	39	31	6	29	33	32
	3	18	14	5	24	13	12
	4, 5	7	5	0	0	7	7
Coronary disease	No	25	37	11	52	25	44
	Yes	43	63	10	48	32	56
Diabetes	No	98	85	13	65	85	88
	Yes	18	15	7	35	11	11
Hypertension	No	58	48	6	29	52	52
	Yes	63	52	15	71	48	48
Nicotine abuse	No	70	61	17	81	53	57
	Yes	44	39	4	19	40	43
Hypercholesterolaemia	No	40	44	6	46	34	44
	Yes	50	56	7	54	43	56
Family history	No	57	66	9	82	48	63
	Yes	30	34	2	18	28	37
		mean	SD	mean	SD	mean	SD
Age		60	15	71	10	57	15
Body mass index		27	4	30	3	26	4

.

Forty-six patients (37%) were admitted between 0 and 3 hours after symptom onset, 20 (16%) between 3 and 6 hours, 16 (13%) between 6 and 12 hours, 36 (29%) more than 12 hours. In 7 patients, the delay between symptoms onset and admission could not be determined. Of the total population, 21 (17%) had ST-segment elevation or non-ST–segment elevation myocardial infarction (STEMI or NSTEMI, respectively). Unstable angina pectoris was diagnosed in 19 patients (15%). Noncoronary cardiac disease (arrhythmia, hypertension or pericarditis) was diagnosed in 20 patients (16%) and noncardiac chest pain or symptoms were diagnosed in 65 patients (52%), including gastrointestinal disease, musculoskeletal disorders and pulmonary disease. Copeptin and troponin values were available for all patients at time of admission, and 6 hours later for troponin. Biomarkers of myocardial necrosis were significantly higher in patients with ACS (mean troponin 0.1 ng/l, SD 0.1, in patients without AMI; 7 ng/l SD 13, in AMI patients).

Delay between symptoms onset and biomarkers dosage was comparable amongst patients diagnosed with AMI and those without AMI (Table 1).

Blood sample analysis

In patients with STEMI and NSTEMI, troponin levels increased within 6 h after admission (p <0.001). In patients with noncoronary chest pain, troponin levels remained stable over the observed period. No trend could be observed for copeptin, as it was only measured on admission.

Copeptin and troponin in diagnosis of myocardial infarction

Patients were classified into four groups according to admission copeptin and troponin results: group 1 consisted of patients with both high values of troponin and cocpeptin (Co+/T+); patients in group 2 had Co+/T–; patients in group 3 had of Co–/T+, and patients in group 4 had negative troponin and copeptin (T–/Co–) as shown in

Table 2. Correlation between copeptin and troponin and diagnosis.

Co/T on admission	Number of patients and %	Time to onset of symptoms (h)					Diagnosis					MACE
		0–3	3–6	6–12	>12	Unclear	NSTEMI	STEMI	Angina pectoris	Cardiac non-coronary	Others
Group 1: +/+	15 (12)	2 (13)	1 (7)	2 (13)	7 (47)	3 (20)	12 (80)	0	0	3 (20)	0	12 (80)
Group 2: +/–	29 (23)	12 (42)	5 (17)	3 (10)	5 (17)	4 (14)	2 (7)	0	4 (14)	7 (24)	16 (55)	5 (17)
Group 3: –/+	7 (6)	2 (29)	0	1 (14)	4 (57)	0	5 (71)	2 (29)	0	0	0	7 (100)
Group 4: –/–	74 (59)	30 (41)	14 (19)	10 (13)	20 (27)	0	0	0	15 (20)	10 (13)	49 (67)	0

. Patients with STEMI or NSTEMI had at least one of the two markers positive. None of the patients with STEMI or NSTEMI were in group 4 (T–/Co–). Among the STEMI and NSTEMI patients, 12 had clear changes in their 12-lead ECG.

Among the total cohort, 15 patients (12%) were in group 1 (Co+/T+), 29 (23%) in group 2, 7 (6%) in group 3, and 74 (59%) in group 4.

In group 3, seven patients had negative copeptin and positive troponin, two of them had a STEMI, and five a NSTEMI. Four patients were admitted >12 hours after symptoms onset. Two were admitted between 0 to 3 hours after symptoms onset, and one between 6 to 12 hours. However, all had positive troponin at admission.

The negative predictive value of the combination of troponin and copeptin dosage to rule out STEMI or NSTEMI was 100% in our series. The negative predictive value of T alone was 98%.

The diagnostic accuracy of troponin in the diagnosis of AMI at admission as quantified by the area under the ROC curve (AUC) was 0.88 (95% CI 0.79–0.97), significantly higher than the diagnostic accuracy of copeptin at presentation (AUC 0.81; 95% CI 0.71–0.91; p <0.00001). Combination of the two markers significantly increased the diagnostic accuracy provided by troponin alone, with an AUC of 0.97 (95% CI 0.91–0.99; p <0.0001) for the combination of troponin and copeptin (p = 0.0153; Figure 1).

Biomarker concentration according to symptom onset

In patients who had a first positive troponin, we have only one troponin measurement. Copeptin was also measured only on admission. Thus, a true evolution in biomarker concentration cannot be shown. However, when classifying AMI patients according to onset of symptoms, we note that copeptin levels were higher in those who presented early after the onset of symptoms (>30 min, <3 h), as shown in Table 3. Troponin levels increased with time. Some patients with NSTEMI or STEMI had normal copeptin owing to a long delay (>12 hours) between symptoms onset and copeptin measurement.

Positive copeptin and positive troponin patients (group 1) (n = 15) (Table 2)

The highest MACE rate (primary endpoint) occurred in group 1 patients. Most of the patients (12/15, [80%]) had NSTEMI. Diagnosis in the three other patients was cardiac insufficiency, massive pulmonary embolism and hypertensive crisis.

Positive copeptin and negative troponin patients (group 2) (n = 29) (Table 2)

In this group, patients had elevated copeptin and normal troponin at admission. Two of these patients had elevated troponin 6 hours later and had NSTEMI. In the 9 other patients, elevated copeptin could be explained by other serious illnesses (1 pericarditis, 2 hypertensis crisis, 1 pulmonary oedema, 2 unstable angina, 1 aortic dissection, 1 haemopneumothorax, 1 cardiac insufficiency). The 18 others were considered benign cases and were discharged.

Negative copeptin and positive troponin patients (group 3) (n = 7) (Table 2)

In this group, all patients had AMI. Among them, four arrived more than 12 hours after the beginning of the symptoms, which explains the negative copeptin test. One presented 6 to 12 hours after the beginning of the symptoms, a period of time where copeptin might already have returned to its normal range. The two remaining patients either did not notice the exact onset of the symptoms, and might have already experienced myocardial infarction before they reported having developed the first symptoms, or the emergency department physician did not identify the onset of the symptoms as characteristic of AMI.

Table 3. Biomarker concentration according to symptom onset in AMI.

Symptom onset	0–3 hours		3–6 hours		6–12 hours		>12 hours		Unclear
	Co	T1	Co	T1	Co	T1	Co	T1	Co	T1
	44	0.7	73	2.2	51	0.2	6	2	337	0.1
	370	0.1	13	0.08	9	0.9	20	0.9	17	58
	2	0.4					32	7.9	148	0.1
	11	0.1					8	23	128	3
							25	14	45	13
							12	0.3
							9	7
							11	13
mean	107	0.3	43	1.1	30	0.6	12	8.5	135	15
median	27.5	0.26	43	1.1	30	0.6	15	7.5	128	3

Table 3. Biomarker concentration according to symptom onset in AMI.

Symptom onset	0–3 hours		3–6 hours		6–12 hours		>12 hours		Unclear
	Co	T1	Co	T1	Co	T1	Co	T1	Co	T1
	44	0.7	73	2.2	51	0.2	6	2	337	0.1
	370	0.1	13	0.08	9	0.9	20	0.9	17	58
	2	0.4					32	7.9	148	0.1
	11	0.1					8	23	128	3
							25	14	45	13
							12	0.3
							9	7
							11	13
mean	107	0.3	43	1.1	30	0.6	12	8.5	135	15
median	27.5	0.26	43	1.1	30	0.6	15	7.5	128	3

Table 4. Prognostic value of copeptin.

	Number of patients (%)	Copeptin value: mean	SD	median
MACE	27 (22)	74	107	25
No MACE	93 (74)	22	77	6
Lost to follow-up	5 (4)	5	3	7
Hospital admission	48 (38)	50	84	16
No admission	77 (62)	21	83	6

Table 4. Prognostic value of copeptin.

	Number of patients (%)	Copeptin value: mean	SD	median
MACE	27 (22)	74	107	25
No MACE	93 (74)	22	77	6
Lost to follow-up	5 (4)	5	3	7
Hospital admission	48 (38)	50	84	16
No admission	77 (62)	21	83	6

Negative copeptin and negative troponin patients (n = 74) (Table 2)

The second dosage of T in these patients was negative. No patient was diagnosed with AMI. Fihy-nine patients (80%) were discharged, 9 (12%) were admitted to the regular ward, and 6 (8%) to the intensive care unit. Three months after their initial admission to the ED, five were lost of follow-up. The others did not present any MACE.

Patients presenting early after the beginning of symptoms

Forty-two (91%) of the 46 patients presenting 0 to 3 hours after the onset of symptoms did not have AMI as final diagnosis. Among them, 30 (71%) were –/–. None of the patients had a MACE. Twenty-eight (93%) were discharged and only two (7%) were admitted in the internal medicine department. Twenty-nine (97%) had no MACE or new hospitalisation during follow-up and one (3%) was lost to follow-up.

Of the 46 patients presenting early after the onset of symptoms, four were diagnosed with AMI (9%). All had positive troponin. Only two had a positive copeptin value, which was higher that the copeptin of patients presenting later. This shows, as stated before, that copeptin does not add a true diagnostic value for ruling-in AMI patients.

Prognostic value of copeptin

In spite of a relatively small number of patients, the high value of copeptin levels was associated with more severe conditions, more MACEs, and more admissions to the hospital (Table 3).

Follow-up

Follow-up after 1 and 3 months confirmed the initial results since no patients of group 4 had cardiac events. Patients were screened for possible coronary disease and in-hospital admission.

Discussion

Early identification of patients at risk of AMI in a population with new onset of chest pain is essential, because these patients can benefit from an aggressive therapeutic approach. Cardiac troponin had a low sensitivity (below 50%) in detecting AMI within 3 hours following symptom onset. Copeptin had a higher sensitivity (70%) in this population. Copeptin ideally complements troponin in the first hours following chest pain onset because of its high sensitivity and its kinetics: it peaks early after the beginning of the symptoms and decreases as soon as 6 hours after, whereas troponin increases slowly over the first 6 hours following symptom onset, although this could not be verified in our cohort owing to the small number of patients. However, according to various studies showing a decreasing trend over hours after the onset of the symptoms, copeptin levels might correlate with the age of the AMI [2,3,17].

We showed that combining troponin and copeptin significantly improved the ROC performance within 3 hours after chest pain onset, with a negative predictive value of 100% in the overall population. In patients with longer duration of symptoms (>6 hours), copeptin already decreases, according to the literature, and could be falsely negative. In this situation, troponin was always positive and allowed the determination of symptoms to be of a cardiac origin.

We could not show any added diagnostic value of copeptin compared with troponin alone; in the Co+/T– group, only two patients (0.06%) were diagnosed with NSTEMI (T+ at 6 hours).

Cardiac nonischaemic disease like pericarditis, arrhythmias, hypertensive crisis or cardiac decompensation also might lead to elevated troponin levels. Serial troponin also rises in these situations and does not allow ischaemic heart disease to be definitively eliminated as the cause of symptoms. This lowers the positive predictive value of both troponin and copeptin in evaluation of patients with chest pain. Acute or chronic disease like aortic dissection, pulmonary embolism, biliary lithiasis, pancreatitis, mesenteric infarction, chronic bronchopulmonary obstructive disease, etc. also cause rises troponin and copeptin. In these situations, biomarkers of different origins than troponin like copeptin, creatine kinase, amylases, hepatic tests and d-dimers are mandatory in a real world population of patients with new-onset chest pain.

The use of copeptin to exclude AMI in patients with chest pain has been studied previously [3]. Results of several trials evaluating the role of copeptin in the emergency room are discordant, some reported use of copeptin was helpful in patients with suspected AMI [3,6,7,8,9,10,11,12,13,14,15,16,17], others did not show any benefit [18,19,20,21,22,23].

These differences can be partially explained by different inclusion criteria. Some of these studies included patients with STEMI, which could strongly modify sensitivity and negative predictive value [3,6,9,13,16,17]. Indeed, the copeptin value is positively correlated to infarct size [15]. The populations studied also differ. Patients known for ischaemic heart disease were the target population included in two of these trials [11,12]. Interestingly, patients with chronic ischaemic heart disease had higher copeptin values. Interpretation of augmented copeptin in this population is as difficult as it is in patients with medical conditions known to modify copeptin values, like diabetes, smoking, obesity and renal insufficiency [24]. The value of copeptin concentrations in the elderly is also questionable as mentioned by the authors of a study limited to patients older than 70 years [8]. The cut-off value should probably be different in this population.

Our cohort is an all-comer population of the emergency department, without any selection on the basis of age, risk or sex. This makes our results more representative of daily practice and avoids selection bias.

The use of high-sensitivity troponin also seems to improve the negative predictive value of the test. In this particular situation, the copeptin level is less valuable [18,19,21,22,23].

Finally, in some studies, copeptin dosage was used to rule out the diagnosis of unstable angina pectoris. As the release of copeptin is strongly correlated to the amount of myocardial damage, it is obvious that some patients had unstable angina pectoris with little or no myocardial necrosis, leading to a negative copeptin value. We cannot consider negative copeptin in this particular situation of unstable angina as a false negative test. The role of copeptin measurement is not to rule out coronary artery disease definitively but to help in the stratification of the risk of patients with chest pain. Patients with a negative copeptin measurement should not be considered as having no coronary artery disease. On the contrary, if clinical suspicion is high, they must be checked with noninvasive tests for ischaemic heart disease detection. In our cohort, 4 of 19 patients with unstable angina as final diagnosis (21%) had a positive copeptin value. They should be considered as high-risk patients, as they all had MACE. In triage of patients with suspected AMI, combined assessment of conventional troponin and copeptin seems to improve diagnostic performance, specifically in the first 3 hours after symptom onset. This may accelerate decision making and triage.

With the emerging use of ultrasensitive troponin, biomarkers such as copeptin might have less value, as ultrasensitive troponin already peaks after 3 hours, allowing early rule-out of AMI. However, it is difficult to predict if excluding AMI in patients presenting between 30 minutes and 3 hours after the onset of symptoms really leads to some benefit in cost and patient management.

Limitation

The main limitation of the present trial is its small number of patients. However, as mentioned previously, it represents preliminary data as a test of the feasibility of a large trial with the goal to demonstrate the potential benefit of the use of copeptin measurement in the emergency department.

Conclusion

Combined determination of troponin and copeptin on admission in patients presenting early (<3 hours but >30 minutes) after onset of symptoms suggesting AMI provides a high NPV. This combination might improve early rule-out of AMI and makes copeptin a useful complement to conventional troponin in the emergency department.