Article, Cardiology

Performances of the heart fatty acid protein assay for the rapid diagnosis of acute myocardial infarction in ED patients

a b s t r a c t

Objective: We sought to evaluate the added value of heart fatty acid protein assay (HFABP) for rapid diagnosis of acute myocardial infarction in a prospective cohort of emergency department (ED) patients with acute chest pain. Methods: High-sensitivity Cardiac troponin T (hs-cTnT; Roche Diagnostics, Meylan, France) and HFABP (Randox, Mauguio, France) were blindly assayed from venous blood samples obtained at admission. Diagnosis was made by 2 ED physicians using all available data and serial cardiac troponin I as the biochemical standard. diagnostic performances of HFABP combined with hs-cTnT were assessed using logistic regression. Analysis was conducted in all patients and in patients without ST-elevation myocardial infarction.

Results: A total of 181 patients were included (age, 61 +-17 years; male sex, 66%). Acute myocardial infarction occurred in 47 (25.9%) patients, including non-ST-elevation myocardial infarction in 31 (17.1%). The receiver operating characteristic area under the curve was 0.893 for hs-cTnT levels at presentation (95% confidence interval, 0.812-0.974) and 0.908 (95% confidence interval, 0.839-0.977) for the combination of hs-cTnT and HFABP, with no significant (P=.07). Adding HFABP to hs-cTnT increased both sensitivity and negative predictive value (NPV) for non-ST-elevation myocardial infarction diagnosis, with about 13% and 3% increase, respectively, leading to a sensi- tivity of 97% and an NPV of 99%.

Conclusion: The assessment of HFABP at ED admission adds incremental value to initial hs-cTnT. The increase of sensitivity and NPV without sacrificing the specificity and positive predictive value in patients with chest pain with noncontributive electrocardiogram could potentially allow safe and early rule out of acute myocardial infarction without the need for further serial Troponin testing.

Introduction

Identification and management of patients with suspected acute myocardial infarction (AMI) are common and difficult challenges for emergency physicians. It is clear that an early discrimination between AMI and non-AMI patients facilitates more rapid triage and improves treatment. Currently, the High-sensitive cardiac troponin (hs-cTn) I or T is the gold reference biomarker in the diagnosis of acute coronary syn- drome (ACS), but its elevation occurs 6 to 9 hours after the onset of is- chemia. Furthermore, numerous studies have highlighted that the high sensitivity of current sensitive cardiac troponins is at the expense of specificity [1]. In this context, the use of biomarkers in earlier stage than cTn is of great interest [2]. There is also growing evidence of the

? Conflicts of interest: The authors declare no financial or competing interests.

* Corresponding author. 371, ave du doyen Gaston Giraud, 34295 Montpellier Cedex 5, France. Tel.: +33 4 67 33 83 15; fax: +33 4 67 33 83 93.

E-mail address: am-dupuy@chu-montpellier.fr (A.M. Dupuy).

benefits of a multimarker strategy over the use of a single marker when evaluating patients with ACS.

heart-type fatty acid binding protein (HFABP), known to be released from injured myocardium, is one of the promising plasma markers for AMI diagnosis. The HFABP is released very quickly in the circulation and could be detected as early as 1 hour after the onset of chest pain, reaching a peak at 4 hours and returning to baseline level within 24 hours [2, 3]. Until now, three manufacturers (Hycult Biotechnology, Dainippon Pharmaceutical, Randox laboratories) have released labora- tory immunoassays with the first two based on enzyme-linked immu- nosorbent assay in format. Unlike the 2 other available tests that are based on enzyme-linked immunosorbent assay, the Randox HFABP assay is the only automated immunoturbidimetric assay enabling rapid testing for HFABP assay in clinical routine.

The aim of this study was to evaluate the analytical performances of the Randox HFABP assay on Roche Cobas8000 analyzer and the accuracy of this parameter alone or in combination with hs-cTnT for diagnosis of AMI in patients presenting with acute chest pain in the emergency department (ED).

Subjects and methods

Study design

Adult patients with chest pain and onset within 12 hours of presenta- tion to the ED were enrolled. Venous blood for investigational biomarker testing was drawn at presentation and collected into lithium heparin and EDTA-treated tubes. Blood samples were processed at the biochemistry laboratory, and plasma was stored at -80?C for later analysis. All diagno- ses were reviewed at 1 month by 2 independent ED physicians using all available data, including serial cTnI results and cardiology reports, and blinded to investigational biomarker results (hs-cTnT and HFABP), as pre- viously described [4,5]. Acute coronary syndrome and non-ACS diagnoses were distinguished and further categorized. Acute coronary syndrome, which refers to the constellation of symptoms manifesting as a result of acute myocardial ischemia (AMI), encompasses unstable angina , ST-segment elevation myocardial infarction , and non-ST- segment elevation myocardial infarction (NSTEMI). Patients with exclud- ed ACS were categorized as having clinical symptoms of stable angina pectoris (group AP), nonischemic cardiac symptoms or noncardiac symp- toms (group NCAD), and symptoms of unknown origin (group UO).

Sample collection was registered at the French Health Ministry (no. DC-2009-1052). All patients provided written informed consent. The study was performed according to the principles of the Declaration of Helsinki and was approved by the local ethics committee.

Hs-cTnT measurement

The Hs-cTnT assay was performed from frozen lithium heparin plas- ma samples. Samples were thawed just before analysis and run on the Cobas 8000/e602 analyzer (Roche Diagnostics, Meylan, France). The lowest concentration measurable at the 10% coefficient of variation (CV) level is 13 ng/L, and the 99th percentile among healthy individuals is 14 ng/L (confidence interval [CI], 12.7-24.9), as claimed by the manu- facturer. The Limit of Detection (LoD) is 5.0 ng/L [6].

Analytical performances of the Randox HFABP assay on Cobas8000

The HFABP levels were determined using an immunoturbidimetric method from Randox applied on the c502/Cobas8000. serum protein calibrator (ref. FB 3134), controls 2 levels (ref. FB 4026 et FB 4027), and system reagent for HFABP (ref. FB 4025) were from Randox. Re- agent and calibrator were used according to the manufacturer’s recom- mendations with analytic range from 0.747 to 120 ng/mL. Calibration was performed once a month and with change of lot.

Analytical performances, including linearity, imprecision, limit of quantification, and LoD, were assessed according to Clinical Laboratory and Standards Institute guidelines [7,8]. Three different lithium heparin plasma pools with HFABP concentrations greater than 50 ng/mL were diluted in distilled water down to the following final percentage: 100%, 50%, 31%, 20%, 10%, 5%, and 2%. Specimens were analyzed in triplicate, and recoveries were calculated. An average recovery within 10% of expected values was considered acceptable. Imprecision values (CVs) were assessed using 3 plasma pool samples and 1 level of the quality con- trol from Randox ranging from 2.5 to 26.1 ng/mL and were determined through 20 replicated analyses. Plasma pool samples were divided into al- iquots on 20 consecutive days. Aliquots were thawed just before analysis and assayed in duplicate at 2 separate times per day on the basis of a single calibration. The CVs of 10% and 20% were obtained by extrapolation from the imprecision data. The LoD was determined using 10 replicates of both the distilled water and low-concentration samples. Low- concentration samples were made using plasma pools diluted down to 3 and 4 times the assay’s sensitivity claimed by the manufacturer. The LoD was calculated as LoD = LoA + 1.645 ?S, where LoA is the value of 10 replicates of the A sample used as analyte free sample, ?S is the stan- dard deviation of the low-concentration sample measurements.

Statistical analysis

Continuous variables were presented either as mean and standard deviation or as median and interquartile range. Patient groups were compared using Mann-Whitney U test for continuous variables. Logistic regression was used to assess the performance characteristics of the hs- cTnT and HFABP assays and their combination for diagnosis of AMI. Overall diagnostic values were quantified by calculating the area under the receiver operating characteristic (ROC) curve (AUC) using the trapezoidal method, with 95% CI computed by using the Delong method. Comparison between AUCs was performed through the Delong test for correlated ROC curves. Optimal hs-cTnT and HFABP thresholds were determined using the Youden index. A bootstrap analysis (2000 bootstraps) was performed to determine 95% CIs for optimal thresholds. Sensitivity, specificity, and predictive values were calculated for each threshold, along with 95% CI based on binomial distribution. Statistical analysis was conducted in all patients with chest pain as well as in a sub- group of patients with the exclusion of STEMI. The significance level was set at 5% for all tests. Statistical analysis was performed by NK using R

3.1.0 (R Foundation for Statistical Computing, Vienna, Austria).

Results

Analytical performances of the HFABP assay on Cobas8000

The linearity was tested in the range of 1 to 50 ng/mL according to the LoD claimed by the manufacturer (b 1 ng/mL). The linear equation of linearity was y = 1.01x – 0.99, r² = 0.99, with a mean recovery (SD) percentage of 89 (8)%. Within and total run precision ranged from 1.5% to 15% and 5% to 20%, respectively. The functional sensitivity at a total imprecision of 20% (which corresponds to the limit of quanti- fication) was 1.85 ng/mL. The lowest concentration giving CV of 10% was 6 ng/mL. The LoD was 1.29 ng/mL in our conditions, close to the LoD claimed by the manufacturer.

Characteristics of study participants

A total of 181 patients were analyzed; and 47 (26%) patients were diagnosed with AMI, including 31 (17%) with NSTEMI. The mean (+- SD) age of AMI patients was 61 (+-17) years; 66% were male. The pa- tients’ characteristics are shown in Table 1. At admission, hs-cTnT and HFABP concentrations were higher in patients with AMI compared with all other diagnoses. Median hscTnT level was 49 (interquartile range [IQR], 18-128.5 ng/L) in AMI vs 6.5 (IQR, 1.5-10.1 ng/L) in all other diagnoses (P b .001). Median HFABP level was 10.2 (IQR,4.9-22.0 ng/L) in AMI vs 3.9 (IQR, 2.8-5.7 ng/mL) in all other diagnoses (P b

.001). The Hs-cTnT and HFABP levels according to final diagnosis are

Table 1

Baseline characteristics of patients at admission

	All	HFABP positive^c	Hs-cTnT positive^c
Patient number^a	181	65 (35.9)	61 (33.7)
Male sex^a	120 (66)	46 (70.8)	44 (72.1)
Age, y^b	61 (17)	68 (18)	70 (17)
Time onset, h^b	4 (3)	4 (3)	4 (3)
CKD-EPI, ml/min/1.73 m²^b	85 (26)	72 (32)	67 (29)
Group ACS^a STEMI ^a	16 (8.8)	9 (56.3)	10 (62.5)
NSTEMI ^a	31 (17.1)	25 (80.6)	26 (83.9)
UA ^a	24 (13.3)	5 (20.8)	5 (20.8)
Group non-ACS ^a AP ^a	9 (5)	2 (22.2)	1 (11.1)
NCAD ^a	101 (55.8)	24 (23.8)	19 (18.8)

^a Values are expressed as number of patients (percentage).

^b Values are expressed as means (SD).

^c Positive and negative values were determined using cutoff concentrations of 5.8 ng/mL for HFABP and 14 ng/L for hs-cTnT on Cobas8000 instrument.

shown in Fig. 1. The time course of HFABP for the 47 AMI patients is re- ported in Fig. 2. The maximum values of plasma HFABP (12.35 [IQR, 4.52-43.9]) were observed in the early hours at 3 to 6 hours and de- creased after 9 hours (7.90 [IQR, 4.31-25.08]).

Diagnostic performance of combined hs-cTnT and HFABP for AMI diagnosis

In the overall population, ROC AUC was 0.851 (95% CI, 0.775-0.926) for hs-cTnT levels at presentation and 0.793 (95% CI, 0.715-0.870) for HFABP. Combining HFABP with hs-cTnT improved ROC AUC to 0.864 (95% CI, 0.797-0.932) compared to hs-cTnT alone, albeit the difference was not significant (P=.134). Optimal threshold to distinguish AMI from non-AMI defined using the Youden index was 6 (95% CI, 4.02- 10.03) ng/mL. This value is close to the 99th percentile (5.8 ng/mL) de- termined by Eggers et al [9]. However, we used the 5.8-ng/mL cutoff for the present study, which allowed us to compare our results with those of most published studies. The chosen cutoff for hs-cTnT was 14 ng/L, which is the level used in clinical routine. Sensitivity, specificity, and negative (NPV) and positive predictive values (PPV) for AMI diagnosis at the marker thresholds are reported in Table 2. Adding HFABP to hs- cTnT increased both sensitivity (about 20% increase) and NPV (about 5% increase) over hs-cTnT alone.

Combined hs-cTnT and HFABP for NSTEMI rule out

Performance characteristics of both markers and their combination for diagnosis of NSTEMI were assessed in study patients without ST- segment elevation (STEMI excluded) (Fig. 2). Exclusion of STEMI pa- tients did not lead to significant change in ROC AUC. The ROC AUC was 0.893 for hs-cTnT levels at presentation (95% CI, 0.812-0.974) and 0.908 (95% CI, 0.839-0.977) for the combination of hs-cTnT and HFABP, with no significant difference (P = .07) (Fig. 3). Adding HFABP to hs-cTnT at 5.8-ng/mL and 14-ng/L thresholds, respectively, increased both sensitivity and NPV for NSTEMI diagnosis, with about 13% and 3% increase, respectively, leading to a sensitivity of 97% and an NPV of 99% (Table 2). Out of the 165 study patients with the exclusion of STEMI, 98 (59.3%) patients had negative results for both hs-cTnT and

P < .001

100

HFABP (ng/mL)

0-3h 3-6h 6-9h > 9h

Time between onset of symptoms and Blood draw

Fig. 2. Box plots (median, interquartile ranges) of concentrations of HFABP according to the time delay since the onset of symptoms among the 47 patients with AMI.

HFABP at admission using threshold values determined using the Youden index. All but one of these 98 patients had a diagnosis other than NSTEMI (99%). Diagnoses included UA (16.3%), AP (7.1%), NCAD, or chest pain of unknown origin (73.4%). Of the remaining 67 (40.6%) of 165 patients with positive results for at least 1 biomarker, 30 patients (44.7%) had a diagnosis of NSTEMI.

Discussion

Our results highlight the added value of HFABP in combination with hs-cTnT for early rule out of AMI, particularly NSTEMI, at ED admission

P < .001

100

1000

HFABP (ng/mL)

HS-cTnT (ng/L)

NCAD AP UA NSTEMISTEMI

Diagnosis

NCAD AP UA NSTEMISTEMI

Diagnosis

Fig. 1. High-sensitivity HFABP and cTnT levels measured at admission according to final diagnosis. A, HFABP levels (ng/mL). B, High-sensitivity cTnT levels (ng/L). Box plots depicting median levels and interquartile ranges. Biomarker levels are reported using a logarithmic scale.

100%

showed improved specificity but at the expense of sensitivity. As a re- sult, PPV was virtually not improved.

4.2. HFABP and hs-cTnT to rule out AMI

Marker

HFABP, AUC = 0.827 (0.742-0.912)

hs-cTnT, AUC = 0.893 (0.812-0.974)

Combination, AUC = 0.908 (0.839-0.977)

75%

50%

Sensitivity

25%

0% 25% 50% 75% 100%

1 – Specificity

To date, the added value of HFABP for diagnosis of AMI has been demonstrated in combination with standard troponin [10]. The hs-cTn assays allow earlier detection of AMI. However, even using high- sensitivity assays, concentrations below the 99th percentile reference limit do not exclude AMI, especially in the first hours after the onset of symptoms. In our study, the AUC was not statistically different when we found an improvement in sensitivity and NPV in combining these 2 markers. This could be explained by the fact that the AUC is a global measure of diagnostic accuracy and data on AUC state nothing about predicative values and about the contribution of the test in ruling in and ruling out a diagnosis. The impact of this combination seems more relevant to rule out AMI particularly NSTEMI because of the higher sensitivity and NPV [11].

We further show that the combination of both markers at admission with use of prespecified thresholds (14 ng/L corresponding to 99th percentile for hs-cTnT and 5.8ng/mL for HFABP) increased sensitivity to 97% compared with hs-cTnT alone. More interestingly, this early marker combination provided an NPV close to 99% alone in patients

Fig. 3. Receiver operating curve analysis of individual or combined hs-cTnT and HFABP for diagnosis of NSTEMI. n = 165 patients with chest pain.

in a prospective cohort of patients presenting to the ED with chest pain. Our results indicate that, although HFABP cannot be used alone or in combination with hs-cTnT for the positive diagnosis of AMI, this marker could be a useful tool for excluding AMI in ED.

Our study is the first to assess the diagnostic performances of HFABP, when combined or not to hs-cTnT, with testing performed on the Cobas8000 analyzer used in clinical routine. The developed HFABP assay on Cobas8000 clearly shows acceptable analytical performances with a detection limit of 1.29 ng/mL and 10% functional sensitivity at 6 ng/mL close to the 99th percentile. In addition, HFABP levels can be determined from lithium heparin-treated blood samples, thereby allowing a multimarker analysis including hs-cTnT using a single venous blood sample on the same instrument.

HFABP does not allow the positive diagnosis of AMI

Although HFABP can be accurately measured in clinical practice, it cannot be seen as an alternative to troponin or CK-MB. In our study, sen- sitivity and specificity of HFABP for the diagnosis of AMI at admission were lower than those of hs-cTnT. A more interesting approach would be the combination of hs-cTnT and HFABP. Combining hs-cTnT and HFABP for the positive diagnosis of AMI at admission is, however, not appropriate. Because HFABP and hs-cTnT convey similar information, that is, cardiomyocyte lesion, with a different kinetic, this association

with the exclusion of STEMI. More than half (59.3%) of 165 patients had negative test results to both hs-cTnT and HFABP and could thus be identified as potentially suitable for early discharge based on this early marker strategy. Clearly, this combination enhances the sensitivity and NPV. In consequence, the number of false-negative results was decreased, thus allowing safe exclusion of AMI diagnosis in many patients at an early stage.

Only few studies were conducted with high-sensitive cTn T or I as- says and different nonautomatized HFABP assays, showing conflicting results. Although the majority of studies published until 2004 showed HFABP to be superior early diagnostic marker, the results were not con- sistent across all studies. Effectively, several factors must be taken into account such as the selected population (from ED, coronary care unit [CCU], or cardiology department), the end point criterion (ACS, AMI), the delay between the onset of chest pain and the difference in analyt- ical performances of the HFABP assays, as well as the generation of cTn. All these factors have an impact in the calculation of the sensitivity, specificity, PPV, and NPV. The conflicting results between all these stud- ies and that reported here can be explained by all these factors. In our study (Table 3), we selected the studies that used the hs-cTn and that investigated the utility of HFABP in addition to hs-cTn regarding NSTEMI. Despite a severe selection of the recent studies, the results are discordant, which can be explained by the heterogeneity of the pop- ulation and the differences in performance of the HFABP assay. Effec- tively, among the 4 studies reported in Table 3, 3 used a different format from ours for the measurement of the HFABP [11,13,14]. In addi- tion, the analysis of Eggers et al [9] was performed in a pooled popula- tion of patients admitted to CCUs with a primary aim to assess the diagnostic utility of frequently measured cTnI levels. In the study of

Table 2

Diagnostic performance of individual hs-cTnT and HFABP and their combination for diagnosis of AMI or NSTEMI

Diagnosis	Decision rule	Sensitivity	Specificity	PPV		NPV
AMI	hs-cTnT (14 ng/L)	0.766 (0.62-0.877)	0.813 (0.737-0.875)	0.59	(0.457-0.714)	0.908	(0.842-0.953)
	HFABP (5.8 ng/mL)	0.723 (0.574-0.844)	0.746 (0.664-0.817)	0.5	(0.376-0.624)	0.885	(0.811-0.937)
	hs-cTnT >= 14 ng/L and HFABP >= 5.8 ng/L	0.574 (0.422-0.717)	0.866 (0.796-0.918)	0.6	(0.443-0.743)	0.853	(0.782-0.908)
	hs-cTnT b14 ng/L and HFABP b5.8 ng/L	0.915 (0.796-0.976)	0.694 (0.609-0.771)	0.512	(0.4-0.623)	0.959	(0.898-0.989)
NSTEMI	hs-cTnT (14 ng/L)	0.839 (0.663-0.945)	0.813 (0.737-0.875)	0.51	(0.366-0.652)	0.956	(0.901-0.986)
	HFABP (5.8 ng/mL)	0.806 (0.625-0.925)	0.746 (0.664-0.817)	0.424	(0.296-0.559)	0.943	(0.881-0.979)
	hs-cTnT >= 14 ng/L and HFABP >= 5.8 ng/L	0.677 (0.486-0.833)	0.866 (0.796-0.918)	0.538	(0.372-0.699)	0.921	(0.859-0.961)
	hs-cTnT b14 ng/L and HFABP b5.8 ng/L	0.968 (0.833-0.999)	0.694 (0.609-0.771)	0.423	(0.306-0.546)	0.989	(0.942-1)

Diagnostic performance for diagnosis of AMI or NSTEMI was assessed in all study patients (N = 181, upper panel) or in patients with the exclusion of STEMI (n = 165, lower panel), respectively. Performance values are shown along with 95% CI. Decision thresholds were determined using the 99th percentile of a general population (Hs-cTnT 14 ng/L and HFABP 5.8 ng/mL).

Table 3

Diagnostic performance of individual or combined high-sensitive cTn and HFABP for the diagnosis of NSTEMI according to different assay method

Study	Population Origin of patients	n	NSTEMI (%)	Troponin Method (cutoff ng/L)	Se	Spe %	PPV	NPV	HFABP Method (cutoff ng/mL)	Se	Spe %	PPV	NPV	Combination of hs-cTnT and HFABP Se Spe % PPV	NPV
Eggers et al [9] Aldous et al [11]	CCU ED	360 384	128 (36) 61 (16%)	Roche Hs-cTnT (14) Roche	78.9 90	74.6 79.6	63.1 45	86.5 97.7	Randox Evidence (5.8) Hycult Biotech (6)	39.1 50	94.8 89.8	80.6 47.6	73.8 90.7	79.7 74.6 63.4 90 73.5 38.6	86.9 97.5
Inoue et al [13]	ED	432	39 (9%)	Hs-cTnT (14) Roche	82.7	61.2	54.4	86.3	Dainippon Pharma	58.7	78.2	60.3	77	– – –	–
Charpentier et al [14]	ED	677	99 (15%)	Hs-cTnT (14) Bayer	56.1	98.6	–	–	(6.2) Hycult Biotech (7)	26.3	97.1	60.5	88.5	– – –	–
Our study	ED	181	31 (17%)	cTnI (100) Roche	84	81	51	96	Randox-Cobas (5.8)	81	75	43	94	97 70 43	99
				Hs-cTnT (14)

Inoue et al, the patients included had chest pain within 24 hours of ad- mission. Our results are different from those obtained by Eggers et al (2012) [9] in the first study using an hs-cTnT assay in patients admitted to CCU but are closely related to those of Aldous et al [11] obtained from an ED population (Table 3). Clearly, the availability of HFABP testing in automated immunoturbidimetric assay with improved analytical performances will facilitate its use in clinical practice and should help promote studies in various populations, but additional studies are necessary to obtain robust data.

A rule out strategy is essential for rapid exclusion of AMI in the ED, and a Multimarker approach in association with hs-cTn seems to be relevant.

4.3. Limitations and conclusions

Our study includes some limitations. This study was conducted from a prospective cohort of patients with chest pain included in a single cen- ter, and our results may not be applicable to different patient popula- tions or settings. However, our study population is an overall population of consecutive patients presenting to the ED with acute chest pain. Although HFABP is found abundantly in cardiomyocytes, it is also expressed in skeletal muscle, kidney, brain, lactating mammary gland, and placenta. The knowledge of cause of blood elevation of HFABP, as well as patient’s age, renal function, and body mass index, is an essential parameter to consider when assessing the potential use efficacy of this biomarker. The renal function might affect plasma HFABP levels. However, in our study, the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) was not statistically different in the AMI or non-AMI group. In addition, only a small study conducted on 12 healthy subjects explored the biological variation [12]. Because the cutoff and the kinetic for serial change are of great importance to determine the magnitude of change that is clinically relevant, studies on a larger scale are needed to determine the reference change value.

Further interventional studies should be implemented to evaluate the impact of a combined biomarker approach integrating HFABP in the initial evaluation of patients with chest pain in the actual ED setting. Benefits regarding shorter length of stay as well as direct discharge from the ED for patients with nondiagnostic electrocardiogram and negative results to both biomarkers on admission should be achieved without greater risk for undetected AMI and without an increase in the cost of care.

The Randox immunoturbidimetric HFABP assay is capable of quickly and accurately measuring HFABP in clinical routine from the same blood tube than that used for measuring troponin.

In summary, our results show that assessment of HFABP at ED admission adds incremental value to initial hs-cTnT. The increase of sensitivity and NPV without sacrificing the specificity and PPV in

patients with chest pain with noncontributive electrocardiogram could potentially allow safe and early rule out of AMI without the need for further serial troponin testing.

Acknowledgments

Reagents for the determination of hs-cTnT and HFABP assays used in this study were kindly provided by Roche Diagnostics and Randox, France, respectively.

The authors would like to thank the staff of “Centre des Collections Biologiques de l’Hopital de Montpellier” for management and conservation of the Biobank.

References

AJEM

Performances of the heart fatty acid protein assay for the rapid diagnosis of acute myocardial infarction in ED patients

a b s t r a c t

Introduction

Subjects and methods