Heart-type fatty acid binding protein in early diagnosis of myocardial infarction in the era of high-sensitivity troponin: a systematic review and meta-analysis

Abstract

Introduction

Heart-type fatty acid binding protein (HT FABP) is an emerging biomarker of ischaemic myocardial necrosis. While previous studies have demonstrated its additive value when compared to contemporary troponin assays in the diagnosis of acute myocardial infarction (AMI), its utility in the era of high-sensitivity troponin (hsTn) assays remains undetermined.

Methodology

A systematic review and meta-analysis of relevant studies was performed comparing the diagnostic performance of HT FABP both alone and in conjunction with hsTn in the early diagnosis and exclusion of AMI.

Results

A systematic literature search yielded eight eligible studies including 3395 patients. Of these, 716 patients (21.1%) were eventually diagnosed with AMI. The pooled sensitivity and specificity for hsTn on admission was 82.5% (95% confidence interval [CI]: 79.8–85.0%) and 89.7% (95% CI: 88.7–90.6%), respectively, while the area under the curve (AUC) for the summary receiver operating characteristics (ROC) curve is 0.92 (SE 0.02). While the pooled specificity (84.6%, 95% CI: 83.2–85.9%) of admission HT FABP is similar to hsTn for the early diagnosis of AMI (P = 0.07), its pooled sensitivity (63.5%, 95% CI: 59.9–67.1%, P < 0.001) is significantly worse. Accordingly, the AUC of the summary ROC curve for HT FABP (0.79, SE 0.03) is inferior to hsTn (P < 0.0001). The addition of HT FABP to hsTn resulted in no improvement in the sensitivity (P = 0.058) and worsened the specificity (P = 0.001) in the early diagnosis of AMI compared to hsTn alone.

Conclusion

HT FABP does not appear to improve the diagnostic accuracy of hsTn, and consequently its routine use currently cannot not be recommended.

Keywords

Heart-type fatty acid binding protein high-sensitivity troponin acute myocardial infarction

Introduction

The timely diagnosis and exclusion of an acute myocardial infarction (AMI) is critical in the management of patients presenting to hospitals with a suspected acute coronary syndrome (ACS), as the associated clinical and economic implications can be considerable. Current diagnostic algorithms relying on contemporary troponin assays are limited by their relatively low sensitivity and therefore negative predictive value (NPV). Consequently, early exclusion of AMI and risk stratification with contemporary troponins remain a clinical challenge, prompting the search for an alternative marker of ischaemic myocardial necrosis which may accelerate the diagnostic process of AMI and facilitate timely delivery of care.

Over the last decade, the heart-type fatty acid binding protein (HT FABP) has emerged as a potential early biomarker of ischaemia-related myocardial necrosis. The HT FABP is a relatively small (15 kDa) intracellular protein which participates in the fatty acid metabolism by transporting fatty acids from the cell membrane to mitochondria for oxidation.¹ Its potential role in the diagnosis of AMI was first demonstrated by Tsuji et al.² who showed a significant and early elevation of HT FABP concentrations within 6 h of symptoms of an AMI. Since then, many studies have been dedicated to develop HT FABP as an alternative or adjunctive tool in the diagnosis of AMI in patients presenting with clinically suspected ACS. Others have also demonstrated its prognostic value in the determination of long-term outcome.^3,4

Thus far, the evidence surrounding the utility of HT FABP as an early diagnostic marker of AMI has been mixed. In particular, as a standalone test, HT FABP seems to perform poorly compared to contemporary troponin assays.^5,6 Carroll et al.⁶ performed a systemic review and meta-analysis of 17 studies evaluating the role of HT FABP at presentation in the early diagnosis of AMI. They reported a pooled sensitivity and specificity of 81% (95% confidence intervals [CI]: 50–95%) and 80% (26–98%), respectively, for the quantitative assays, and 68% (11–97%) and 92% (20–100%), respectively, for the qualitative assays. Compared to an earlier meta-analysis,⁷ only studies using the Universal Definition⁸ of myocardial infarction were included in this analysis, although there remained significant heterogeneity in the data presented due to variation in the inclusion and exclusion criteria, timing of sample collection in relation to symptom onset, reference troponin assays and diagnostic threshold among included studies. The addition of HT FABP to contemporary troponins in four studies included in this analysis improved the overall sensitivity (42–75% vs. 76–97%) in the detection of AMI, although specificity was significantly compromised (94–100% vs. 65–93%). The authors concluded that the role of HT FABP in the early diagnosis of AMI remains uncertain.

In another recent meta-analysis performed by Lippi et al.,⁵ the diagnostic performance of HT FABP was directly compared to contemporary troponin across eight studies. HT FABP appeared to be a more sensitive (80% vs. 73%, P = 0.02) but less specific (83% vs. 94%, P < 0.001) predictor of AMI, while its overall diagnostic performance was significantly worse when compared with troponin alone (area under the curve (AUC) 0.784 vs. 0.820, P < 0.001). Similar to the findings by Carroll et al., the combination of HT FABP and contemporary troponin yielded higher sensitivity than troponin alone in the early diagnosis of AMI, although only at the expense of the overall specificity. The overall performance of the combination of biomarkers appeared to be better than troponin alone however (P < 0.001).

Comparisons so far between HT FABP and troponins have been limited to the contemporary troponin I and T. The contemporary troponin assays have superior sensitivity and specificity when compared with Creatine Kinase MB⁹ in the diagnosis of AMI and are better predictors of prognosis. However, the sensitivity and NPV of contemporary troponin on admission are suboptimal,^10,11 and its ability to detect small variations in the concentrations is further restricted by their imprecision with changes at or below the 99th percentile level.¹²

In recent years, high-sensitive troponin (hsTn) assays have superseded contemporary assays in clinical practice due to their superior sensitivity. Its use has been shown to improve the diagnosis of AMI¹³ and improve patient outcomes,¹⁴ notwithstanding a minor decline in assay specificity. Accordingly, hsTn has now been adopted by many around the world as the biomarker of choice in the diagnosis of AMI.

As the primary strength of HT FABP over contemporary troponin assays appears to be the improved sensitivity in the early diagnosis of AMI, it is not clear if it would remain a useful adjunct in the era of high-sensitivity troponin (hsTn). Herein, we present a review of the comparison between hsTn and HT FABP in the early diagnosis of AMI based on their sensitivity, specificity, NPV and positive predictive value (PPV), and the overall diagnostic accuracy using the area under the receiver operating characteristics (ROC) curve (AUC).

Methodology

A systematic literature search was performed in July 2014 using Ovid Medline, Embase and Cochrane Database with Systematic Reviews with no date restriction. The aim was to determine the most optimal early diagnostic strategy for patients presenting to hospitals with suspected AMI. A combined search strategy was employed using keywords including: ‘heart type fatty acid binding protein’ or ‘HT FABP’, ‘high sensitivity troponin’, ‘troponin’, ‘myocardial infarction’, ‘chest pain’, ‘acute coronary syndrome’, ‘ischemic heart disease’ and ‘diagnosis’. Further, the reference list of relevant studies was reviewed for additional studies addressing this subject. Only studies published in English language were included for analysis.

Studies meeting the following criteria were eligible for inclusion: randomized control trials or diagnostic cohort study where AMI were diagnosed based exclusively on the universal definition of the European Society of Cardiology (ESC)/American College of Cardiology (ACC) guidelines with either troponin T or I exceeding the prespecified 99% cut-off;⁸ adults (>18 years) presenting to hospital with clinical suspicion of AMI who had blood drawn on admission and assayed both for HT FABP (index test) and hsTnT or hsTnI either prospectively or retrospectively; studies containing raw data for retrieving directly or permitting indirect derivation of sensitivity, specificity, PPV, NPV, AUC, as well as the associated 95% CI and standard errors (SE). Studies were excluded if incomplete data were reported or if the patients included were not representative of the cohort of interests.

The quality of the retrieved citations was assessed against the Quality Assessment of Diagnostic Accuracy Studies (QUADAS) checklist criteria. The data were extracted independently by two of the co-authors (KL and SH) and summarized into 2 × 2 tables. Any differences in data collected were resolved by detailed discussion and review of the literature. Pooled sensitivity and specificity for the assays as well as their respective 95% CI were derived based on bi-variate random effect model. The pooled AUC generated were weighted derivations from the summary ROC curve (sROC). The assessment of interstudy heterogeneity was performed using the I-square statistics. Statistical analysis was performed and cross-checked with Review Manager ver 5.2.9 (Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration), MedCalc ver 12.2.1.0 (MedCalc Software, Mariakerke, Belgium) and Comprehensive Meta Analysis ver 2.2.064 (Biostat Inc, USA).

Results

The systematic search performed yielded a total of 245 records from the three electronic databases as well as from review of the study citations. One hundred and fifty-seven records remained after duplications were excluded. Among these, 148 studies were eliminated after a review of abstracts and/or full text based on established criteria. The remaining nine studies were subjected to detailed review of their methodology and quality of the data, and one was further excluded due to insufficient data for derivation of test sensitivity and specificity (Figure 1). Consequently, eight studies (Table 1) were included for quantitative analysis.^15–22

Figure 1.

PRISMA flow chart of the study selection process.

Table 1.

Summary of studies included in the meta-analysis.

Study	Design	Setting	Age (Median)		Male (%)		Inclusion criteria	Exclusion criteria	End point	Timing of blood sampling	Timing of studied samples	Assay types	Time from symptom onset
Study	Design	Setting	AMI	No AMI	AMI	No AMI	Inclusion criteria	Exclusion criteria	End point	Timing of blood sampling	Timing of studied samples	Assay types	AMI	No AMI
Aldous¹⁵	Single centre Australia Prospective	Emergency department	NR	NR	NR	NR	Within 4 h of symptoms suggestive of AMI	ST elevation on presentation	AMI	Admission	Admission	Laboratory	≤4 h
Cappellini¹⁶	Single centre Italy Retrospective	Emergency department	73.9*	70.9*	69.2	65.9	Consecutive patients presenting to ED with chest pain	NR	AMI	Admission, then 3 and 6 h from admission	Admission	Laboratory	61.5%: <4 h 23.1%: 4-8 h 15.4%: >8 h	39.0%: <4 h 22.0%: 4–8 h 26.8%: >8 h 12.2%: NR
Collinson¹⁷	Multicentre UK Retrospective	Emergency department	53.7		60		Patients aged 18 years or older, presenting with acute chest pain to the ED	High risk ACS, non-coronary pathology, >12 h since symptom onset, unable to consent	AMI	Admission, 90 min from admission	Admission + 90 min post admission	Point of care	Median: 220 min (interquartile range: 155–345 min) 33.4%: < 3 h 75.3%: < 6 h
Eggers¹⁸	Multicentre Sweden Retrospective	Coronary care unit	66.8		65.6		Patients with chest pain with ≥15 min duration within the last 8 h	ST elevation on presentation	AMI	Admission, then 8 time points from admission within the first 24 h	Admission	Laboratory	39.7%: < 4 h
Kagawa²⁰	Single centre Japan Prospective	Emergency department	66	66	33	45	Consecutive patients who presented to the emergency department with symptoms, such as chest pain and angina pectoris, that were suggestive of an acute myocardial infarction and in whom the onset or peak of symptoms had occurred within 12 h	Advanced kidney failure	AMI	Admission, then 4–6 h after presentation if indicated	Admission	Laboratory	48.9%: <2 h 19.2%: 2–6 h 31.9%: >6 h	46.3%: < 2 h 32.8%: 2–6 h 20.9%: >6 h
Reiter²¹	Multicentre Switzerland Prospective	Emergency department	73	62	72	66	Patients presenting to the ED with chest pain suggestive of AMI with onset or peak within the last 12 h before presentation	End stage renal failure, ST elevation on presentation	AMI	Admission, then 1, 2, 6 h after presentation if indicated	Admission + 1, 2 and 3 h after admission	Laboratory	NR
Ruff²²	Single centre USA Prospective	Emergency department	36% > 65 years		60		Patients >18 years of age who presented within 24 h of onset of non-traumatic chest pain suspicious for acute coronary syndrome (ACS)	Recent hospitalization for ACS, severe renal insufficiency, and known chronic or acute systemic inflammatory disease or infection; renal disease.	AMI	Admission, then 4–6, 12–24 and 36–48 h from admission	Admission	Laboratory	NR
Ilva¹⁹	Single centre Finland Retrospective	Emergency department	68.1		62		Consecutive patients who were admitted to the emergency room for evaluation of chest pain or equivalent symptoms suggesting myocardial ischaemia	Patients who reported an uncertain or more than 24-h delay from symptom onset to admission.	AMI	Admission, then 6–12 h and 24 h from admission	Admission	Laboratory	Median: 4.7 h (95% CI: 4.1–4.5)

Mean.

AMI: acute myocardial infarction; ACS: acute coronary syndrome; ED: emergency department; NR: not reported.

Overall, 3395 patients were included in the final analysis. Among these, 716 patients (21.1%) were eventually diagnosed with AMI according to the Universal Definition. Patients with ST elevation MI (STEMI) were included in four studies,^16,19,20,22 while one study exclusively enrolled patients at low cardiovascular risk.¹⁷ The trials were set in the Emergency Department in seven studies^{15–17,19–22} and the coronary care unit in one.¹⁸ Four studies had a prospective design,^15,20–22 while the rest performed retrospective analyses of frozen plasma samples. Three studies used hsTnI,^19,20,22 four used hsTnT^15,16,21 and one assayed both.¹⁷ The analyser for hsTnT was produced by Roche Diagnostics in all (99% cut-off 14 ng/L), while hsTnI was measured using the Beckman-Coulter AccuTnI Enhanced (cut-off 40 ng/L) in one study,¹⁷ Siemens TnI-Ultra assay (99% cut-off 40 ng/L) in two,^17,22 and the Abbott-Architect assay (99% cut-off: 28 and 32 ng/L, respectively) in the remainder.^19,20 Four studies analysed their HT FABP with Randox immunoturbidimetric assay,^16–18,22 one by enzyme-linked immunosorbent assay (ELISA) provided by Dainippon Pharmaceutical, Osaka, Japan,²⁰ one with QuickSenshFABP assay,²¹ one by Hycult Biotech,¹⁵ and the last with Innotrac Aiol Immunoanalyser.¹⁹ The prespecified 99% cut-off for HT FABP varied widely between studies (Table 2). Renal dysfunction was cited as an exclusion criterion in three studies.^20–22

Table 2.

Summary of the diagnostic performance of admission hsTn, HT FABP and combined HT FABP/hsTn for acute myocardial infarction.

Study	Patient number	Number of AMI	AMI subtypes		hsTn						HT FABP					HT FABP + hsTn
Study	Patient number	Number of AMI	STEMI	NSTEMI	Troponin assay	99% cut-off (ng/L)	Methods of assay	PPV	NPV	AUC	99% cut-off (µg/L)	Methods of assay	PPV	NPV	AUC	PPV	NPV	AUC
Aldous¹⁵	384	60	0	60	hsTnT	14	Roche Diagnostic	45	97.7	NR	60	Hycult Biotech	47.6	90.7	NR	38.6	97.5	NR
Cappellini¹⁶	67	26	6	20	hsTnT	14	Roche Diagnostics	53.8	82.1	0.81	3.49	Randox immunoturbidimetric assay	50.9	100	0.84	NR	NR	NR
Collinson¹⁷	803	63	0	63	hsTnT	14	Roche Diagnostics	63.6	98.1	0.92	3	Randox immunoturbidimetric assay	47.1	96.9	0.84	46.6	98.7	NR
					hsTnI(S)	40	Siemens cTnI Ultra	76.7	97.7	0.9						50.9	98.7	NR
					hsTnI(B)	40	Beckman-Coulter AccuTnI enhanced	79.2	97.2	0.92						48	98	NR
Eggers¹⁸	360	128	0	128	hsTnT	14	Roche Diagnostics	63.1	86.5	NR	5.8	Randox immunoturbidimetric assay	80.6	73.8	NR	63.4	86.9	NR
Kagawa²⁰	114	47	26	21	hsTnI	28	Abbott-Architect	73	85	0.89	5.8	Dainippon Pharmaceutical	41	59	0.59	NR	NR	NR
Reiter²¹	1031	166	0	166	hsTnT	14	Roche Diagnostics	42	98	0.94	4.2	QuickSenshFABP assay	41	94	0.84	NR	NR	0.88
Ruff²²	343	91	13	78	hsTnI	40	Siemens TnI-Ultra	80	97	0.96	4.2	Randox immunoturbidimetric assay	41	94	0.78	56	97	0.963
Ilva¹⁹	293	135	56	80	hsTnI	32	Abbott - Architect	100	84.5	0.93	10.4	Innotrac Aio! Immunoanalyzer	84.5	77	0.85	NR	NR	NR

AMI: acute myocardial infarction; STEMI: ST elevation myocardial infarction; NSTEMI: Non-ST elevation myocardial infarction; PPV: positive predictive value; NPV: negative predictive value; AUC: area under the curve; hsTn: high-sensitivity troponin; NR: not reported; HT FABP: heart-type fatty acid binding protein.

High-sensitivity troponin (Figure 2)

Overall, the studies were highly heterogeneous in relation to the diagnostic performance of admission hsTn assays (I²= 78.1% and 98.1% for sensitivity and specificity, respectively, P < 0.001). The pooled sensitivity and specificity for hsTn was 82.5% (95% CI: 79.8–85.0%) and 89.7% (95% CI: 88.7–90.6%), respectively, while the AUC for the sROC is 0.92 (SE 0.02) (Figure 3).

Figure 2.

Summary of the diagnostic performances of HT FABP and hsTn.

Figure 3.

Summary ROC (sROC) for HT FABP vs. hsTn.

When hsTnT is considered in isolation, the pooled sensitivity and specificity was 85.6% (95% CI: 81.9–88.7%) and 83.2% (95% CI 81.6–84.7%). The AUC was 0.91 (SE 0.03). With hsTnI, the pooled sensitivity and specificity were 79.2% (95%CI: 74.9–83.1%) and 97.0% (95% CI 96.1–97.8%) respectively, while the AUC was 0.93 (SE 0.03).

Heart-type fatty acid binding protein (Figure 2)

The studies were equally heterogeneous with respect to the diagnostic performance of admission HT FABP (I²= 90.5 and 97.6 for sensitivity and specificity, respectively, P < 0.001). The pooled sensitivity (63.5%, 95% CI: 59.9–67.1%) for AMI was significantly worse compared to hsTn overall (P = 0.001), while the pooled specificity was numerically inferior (84.6%, 95% CI: 83.2–85.9) though did not reach statistical significance (P = 0.07). The pooled AUC for HT FABP (0.79, SE 0.03) was significantly worse compared to hsTn overall (P < 0.0001) (Figure 3).

When HT FABP was compared specifically with hsTnT,^15–18,21 the pooled sensitivity (60.3%, 95% CI: 55.5–64.9%) was much worse (P = 0.001) though the specificity (86.9%, 95% CI: 85.4–88.3%) did not differ significantly (P = 0.34). The AUC of 0.86 (SE 0.04) is significantly worse than hsTnT (P < 0.01). In contrast, HT FABP performs similarly to hsTnI^19,20,22 in terms of its pooled sensitivity (68.2%, 95% CI: 62.9–73.1%; P = 0.068), although hsTnI significantly outperforms HT FABP in terms of its pooled specificity in the early diagnosis of AMI (86.0%, 95% CI: 84.0–87.9%; P < 0.001). Accordingly, the pooled AUC (0.77, SE 0.03) of HT FABP was significantly worse compared to hsTnI (P < 0.0001).

HT FABP + hsTn vs. hsTn (Figure 4)

Four studies^15,17,18,22 assessed the diagnostic performance of a multibiomarker strategy on admission by adding HT FABP to hsTn in the early diagnosis of AMI. A positive test in this instance is defined as elevation of at least one of the two biomarkers. This yielded a pooled sensitivity and specificity of 85.0% (95% CI: 81.5–88.2%) and 87.5% (95% CI: 86.2–88.6%), respectively, and an AUC of 0.92 (SE 0.02). The addition of HT FABP, however, resulted in no improvement on the sensitivity (P = 0.058) and slightly worsened the specificity (P = 0.001) compared with hsTn alone. The overall difference in the diagnostic accuracy by AUC remained statistically insignificant (P = 0.68) (Figure 5).

Figure 4.

Summary of the diagnostic performance of hsTn and the multibiomarker strategy with hsTn/HT FABP in four included studies.

Figure 5.

Summary ROC (sROC) for HT FABP vs. HT FABP+ hsTn.

Three studies^15,17,18 assessed the value of adding HT FABP to hsTnT. The pooled sensitivity and specificity were 83.7% (95% CI: 78.5–88.0%) and 84.0% (95% CI: 81.9–86.0%), respectively, while the AUC was 0.91 (SE 0.04). Compared to hsTnT alone, the sensitivity and specificity of the combination of HT FABP/hsTnT did not improve significantly (P = 0.49 and 0.082, respectively), while the difference in the overall diagnostic accuracy (AUC) also did not reach statistical significance (vs. 0.91, SE 0.04, P = 0.69).

Two studies^17,22 assessed the value of adding HT FABP to hsTnI. The pooled sensitivity and specificity was 86.6% (95% CI: 81.4–90.9%) and 90.0% (95% CI: 88.5–91.4%). The combination did improve the sensitivity of hsTnI alone (P = 0.038), although only at the expense of specificity (P < 0.0001). Overall, the AUC for the combined HT FABP/hsTnI was statistically inferior to hsTnI alone (AUC 0.93, SE 0.01 vs. AUC = 0.97, SE 0.01, P = 0.02).

Discussion

The present meta-analysis demonstrates that, overall, the admission hsTn is a more sensitive and equally specific predictor of AMI compared to HT FABP. Specifically, HT FABP is less sensitive and equally specific when compared with hsTnT, while it is equally sensitive but less specific compared with hsTnI. The combination of HT FABP and hsTn does not provide additional diagnostic value when compared with hsTn alone, although the addition of HT FABP to hsTnI may improve its sensitivity but only at the expense of specificity. On the whole, HT FABP does not appear to provide any added value in the early diagnosis and exclusion of AMI in the era of hsTn.

While the Universal Definition of myocardial infarction⁸ provides a consistent and uniform means of patient characterization in clinical practice and scientific trials, its role in the early exclusion of AMI is not clear. This in part is related to the need for serial biomarker measurements, which in turn is a result of the relatively low sensitivity of contemporary biomarkers for AMI particularly among early presenters. The value of any emerging biomarker of cardiac injury therefore resides in their ability to improve the existing diagnostic algorithm particularly surrounding the early phase of admission for patients with suspected AMI. Accordingly, the novel biomarker needs to be both tissue and mechanism specific and has a favourable release kinetics.

Among all the emerging biomarkers, HT FABP in particular has been extensively investigated in both an experimental and clinical setting. While it has been detected in small quantities in the skeletal muscles (173 ug/g), brain (16–40 ug/g) and the intestinal system (2.7–4.9 ug/g), the HT FABP is found predominantly within cardiac tissue (540–600–ug/g), thus conferring its cardiac tissue specificity.²³ Because of its small size, HT FABP is released very early following the onset of ischaemia and becomes detectable in the circulation within 20 min of the onset of ischaemia. Levels then peak at 3–4 h and return to normal within 24 h.¹

Our meta-analysis, however, suggests that when HT FABP is compared with hsTn, the overall diagnostic performance was at best equivalent if not worse when interpreted in isolation. When HT FABP was combined with hsTn, there was also no incremental value in the diagnostic performance of hsTn. Compared to previous reports,⁵ the introduction of hsTn in our analyses has diminished the observed advantage of HT FABP over contemporary troponin in terms of its early sensitivity for AMI, while preserving its overall specificity. Similarly, the benefit observed by previous authors^5,6 with the combination of HT FABP and contemporary troponin in terms of its improved sensitivity and diagnostic performance has been neutralized by the use of hsTn.

Several factors merit consideration. It has been postulated by many that cardiac troponin exists in two cellular compartments. Troponin that becomes detectable following MI has been traditionally believed to originate from the structural or the muscular pool. It signals cell death and is responsible for the persistence of troponin in the circulation despite its short half-life.²⁴ Based on studies on non-ischaemic causes of troponin elevation,²⁵ however, it has been speculated that troponin also exists within the cytosol, which leaks into the lymphatics and blood circulation when the permeability of the cellular membrane is increased by physiological or pathological stress.²⁶ Since both troponin and HT FABP exist within the same cellular compartment, their release following the onset of ischaemia may parallel each other. Consequently, as the sensitivity of the modern troponin assays increases, the diagnostic advantage offered by HT FABP may be lost.

Secondly, the release kinetics vary between troponin and HT FABP, and the variability in the time of blood sampling in relation to the onset of symptoms among trials included may also confound our results. While the diagnostic performance of hsTn improves with serial sampling,^17,19,21 the diagnostic performance of HT FABP seems to decline with time.²¹ This is because HT FABP concentration decreases with serial sampling²¹ and normalizes quickly due to rapid renal clearance.²⁷ Accordingly, the sensitivity of HT FABP begins to drop 6 h after symptom onset while the overall diagnostic performance also decreases in parallel.¹⁹ A significant proportion of patients in this study presented more than 6 h after their symptom onset. As a result, the sensitivity of HT FABP in these patients in comparison to hsTn may have been misrepresented.

Given its release kinetics, it may be argued that the primary role for HT FABP is in the diagnosis of AMI in early presenters. In comparison to hsTn, however, the diagnostic performance of HT FABP appears consistently poor irrespective of the time delay between symptom onset and presentation. Collinson et al.,¹⁷ Reiter et al.²¹ and Ruff et al.²² all demonstrated poor performance of HT FABP against hsTn even in early presenters (<3–6 h), while Eggers et al.,¹⁸ Kagawa et al.²⁰ and Ilva et al.¹⁹ found the diagnostic performance of HT FABP against hsTn to be equally poor between early and late presenters. The interaction between the timing of sampling and combined HT FABP/hsTn appears more complex. In one study,²² the addition of HT FABP significantly improved the NPV in early presenters (<6 h) compared with hsTn alone, although the overall diagnostic performance did not change. Another study,¹⁸ however, demonstrated no improvement in the diagnostic performance (sensitivity or specificity) over hsTn alone in either early (<4 h) or late presenters (≥4 h) with the combination strategy.

The interpretation of our results may be further confounded by other sources of heterogeneity among the studies included. Specifically, not all studies included patients with STEMI. Indeed, one study exclusively enrolled patients at low baseline cardiovascular risk.¹⁷ While HT FABP is a fairly specific marker of myocardial damage, it is unclear if its concentrations may be affected by critical illness or by the systemic effects of STEMI independent of cardiac ischaemia. Renal dysfunction, which falsely elevates both the troponin and HT FABP concentrations, was only cited by three studies as an exclusion criterion. An uneven distribution of renal function among patients included may influence the specificity of troponin assays in relation to HT FABP. It is similarly difficult to control for heterogeneity in HT FABP assay methodologies across studies and the cut-off thresholds. Indeed, a trend exists between a lower cut-off threshold and a higher NPV and lower PPV, and it would be difficult to know if the overall results would remain the same if the cut-off points were standardized across the board. By the same token, the inclusion of both hsTnI and hsTnT in pooled analyses may also influence the results as there appears to be inherent differences in their diagnostic performance.¹³ Efforts have been made, however, to ensure separate analyses were performed with hsTnI and hsTnT, although the number of studies and subjects are reduced and the results may not be robust.

Further consideration needs to be given to the confounding effects of unreported variables in the included studies. While factors such as age, sex, hypertension, body mass index, previous bypass surgery and arrhythmia^16,22,28 have been shown to influence the concentration of HT FABP, no studies have incorporated these in a multivariate model. The mode of sample storage and the thawing process, as well as its impact on assay accuracy for both hsTn and HT FABP is also unclear. Finally, the results of the analyses may be further skewed by the presence of numerical outliers. It is important to note that the outlying results belong to the two smallest included studies^16,20 and the possibility of statistical error need to be taken into consideration.

Conclusion

HT FABP both alone and in conjunction with hsTn does not appear to improve the overall diagnostic performance of hsTn in the early diagnosis and exclusion of AMI. As the evidence is currently influenced by the considerable heterogeneity among the studies, the results of this systematic review need to be interpreted with caution. Nonetheless, the currently available data do not support the routine use of HT FABP in the era of hsTn as an early risk stratification strategy in patients presenting with suspected AMI.

Footnotes

Acknowledgements

None.

Declaration of conflicting interests

None declared.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Ethical approval

As this was a review and meta-analysis, no ethics approval was sought.

Guarantor

SO.

Contributorship

KL conceived the idea. KL and SH performed the literature search and extracted the data independently from relevant studies. KL performed the statistical analysis and wrote the first draft of the manuscript. SO provided overall supervision. All authors reviewed and edited the manuscript and approved the final version of the manuscript.

References

Kleine

Glatz

Van Nieuwenhoven

. Release of heart fatty acid-binding protein into plasma after acute myocardial infarction in man. Mol Cell Biochem 1992; 116: 155–162.

Tsuji

Tanaka

Sohmiya

. Human heart-type cytoplasmic fatty acid-binding protein in serum and urine during hyperacute myocardial infarction. Int J Cardiol 1993; 41: 209–217.

Viswanathan

Kilcullen

Morrell

. Heart-type fatty acid-binding protein predicts long-term mortality and re-infarction in consecutive patients with suspected acute coronary syndrome who are troponin-negative. J Am Coll Cardiol 2010; 55: 2590–2598.

Kilcullen

Viswanathan

Das

. Heart-type fatty acid-binding protein predicts long-term mortality after acute coronary syndrome and identifies high-risk patients across the range of troponin values. J Am Coll Cardiol 2007; 50: 2061–2067.

Lippi

Mattiuzzi

Cervellin

. Critical review and meta-analysis on the combination of heart-type fatty acid binding protein (h-fabp) and troponin for early diagnosis of acute myocardial infarction. Clin Biochem 2013; 46: 26–30.

Carroll

Al Khalaf

Stevens

. Heart-type fatty acid binding protein as an early marker for myocardial infarction: Systematic review and meta-analysis. Emerg Med J 2013; 30: 280–286.

Bruins Slot

Reitsma

Rutten

. Heart-type fatty acid-binding protein in the early diagnosis of acute myocardial infarction: a systematic review and meta-analysis. Heart 2010; 96: 1957–1963.

Thygesen

Alpert

Jaffe

. Third universal definition of myocardial infarction. J Am Coll Cardiol 2012; 60: 1581–1598.

Anderson

Adams

Antman

. Acc/aha 2007 guidelines for the management of patients with unstable angina/non-st-elevation myocardial infarction: a report of the American College of Cardiology/American Heart Association task force on practice guidelines. Circulation 2007; 116: e148–e304.

10.

Keller

Zeller

Peetz

. Sensitive troponin I assay in early diagnosis of acute myocardial infarction. N Engl J Med 2009; 361: 868–877.

11.

Reichlin

Hochholzer

Bassetti

. Early diagnosis of myocardial infarction with sensitive cardiac troponin assays. N Engl J Med 2009; 361: 858–867.

12.

Thygesen

Mair

Giannitsis

. How to use high-sensitivity cardiac troponins in acute cardiac care. Eur Heart J 2012; 33: 2252–2257.

13.

Weber

Bazzino

Navarro Estrada

. Improved diagnostic and prognostic performance of a new high-sensitive troponin t assay in patients with acute coronary syndrome. Am Heart J 2011; 162: 81–88.

14.

Mills

Churchhouse

Lee

. Implementation of a sensitive troponin i assay and risk of recurrent myocardial infarction and death in patients with suspected acute coronary syndrome. JAMA 2011; 305: 1210–1216.

15.

Aldous

Pemberton

Troughton

. Heart fatty acid binding protein and myoglobin do not improve early rule out of acute myocardial infarction when highly sensitive troponin assays are used. Resuscitation 2012; 83: e27–e28.

16.

Cappellini

Da Molin

Signorini

. Heart-type fatty acid-binding protein may exclude acute myocardial infarction on admission to emergency department for chest pain. Acute Card Care 2013; 15: 83–87.

17.

Collinson PG, David; Goodacre, Steve. Comparison of contemporary troponin assays with the novel biomarkers, heart fatty acid binding protein and copeptin, for the early confirmation or exclusion of myocardial infarction in patients presenting to the emergency department with chest pain. Heart 2014; 100: 140–145.

18.

Eggers

Venge

Lindahl

. High-sensitive cardiac troponin t outperforms novel diagnostic biomarkers in patients with acute chest pain. Clin Chim Acta 2012; 413: 1135–1140.

19.

Ilva

Lund

Porela

. Early markers of myocardial injury: cTnI is enough. Clin Chim Acta 2009; 400: 82–85.

20.

Kagawa

Toyofuku

Masaoka

. Comparison of heart-type fatty acid binding protein and sensitive troponin for the diagnosis of early acute myocardial infarction. Int J Cardiol 2013; 166: 347–351.

21.

Reiter

Twerenbold

Reichlin

. Heart-type fatty acid-binding protein in the early diagnosis of acute myocardial infarction. Heart 2013; 99: 708–714.

22.

Ruff

Bonaca

Kosowsky

. Evaluation of the diagnostic performance of heart-type fatty acid binding protein in the bwh-timi ed chest pain study. J Thromb Thrombolysis 2013; 36: 361–367.

23.

Pelsers

Hermens

Glatz

. Fatty acid-binding proteins as plasma markers of tissue injury. Clin Chim Acta 2005; 352: 15–35.

24.

Antman

. Decision making with cardiac troponin tests. N Engl J Med 2002; 346: 2079–2082.

25.

Agewall

Giannitsis

Jernberg

. Troponin elevation in coronary vs. non-coronary disease. Eur Heart J 2010; 32: 404–411.

26.

Jaffe

. Troponin release – reversible or irreversible injury? Should we care? Clin Chem 2011; 58: 148–150.

27.

Al-Hadi

William

Fox

. Serum level of heart-type fatty acid-binding protein in patients with chronic renal failure. SQU Med J 2009; 9: 311–314.

28.

Niizeki

Takeishi

Takabatake

. Circulating levels of heart-type fatty acid-binding protein in a general Japanese population: effects of age, gender, and physiologic characteristics. Circ J 2007; 71: 1452–1457.