Fibrinogen-to-albumin ratio as a novel biomarker for risk assessment in clinically suspected acute myocarditis

Abstract

Background

Acute myocarditis typically follows a benign course, but fatal complications may occur in fulminant cases. Evaluation is based on clinical, laboratory, and imaging parameters. We aimed to assess the prognostic value of the fibrinogen-to-albumin ratio (FAR) in patients with clinically suspected myocarditis.

Methods

Echocardiographic, laboratory, and clinical parameters were obtained from patients with clinically suspected myocarditis. We assessed the correlations between troponin, albumin, fibrinogen, FAR, left ventricular ejection fraction (LVEF), and complications during the index hospitalization. Repeat echocardiography was performed in patients with reduced LVEF. Several potential factors were tested for their ability to predict LVEF recovery.

Results

A total of 118 patients were included (80 % male; median age, 35 years; mean LVEF, 55 %). Median troponin level was 2558 ng/L, and the median fibrinogen-to-albumin ratio (FAR) was 120. In 38 patients (32 %), LVEF was reduced; among them, 15 (39 %) had persistently reduced LVEF on repeat echocardiography. Troponin levels and FAR were inversely correlated with reduced LVEF (r = –0.28, p = 0.002 and r = –0.31, p = 0.001, respectively). FAR remained an independent factor after adjustment for age and sex. Neither troponin nor FAR was associated with persistently reduced LVEF on follow-up echocardiography. In addition, high FAR was not correlated with longer length of stay (LOS) (p = 0.07), readmissions (p = 0.29), or complications during hospitalization (p = 0.22).

Conclusion

FAR is associated with reduced LVEF and may reflect disease severity in clinically suspected myocarditis; however, it does not appear to be useful for long-term prediction.

1. Introduction

Acute myocarditis is an inflammatory disease that primarily affects the cardiac muscle but may also involve the pericardial cavity [1]. Globally, the disease affects approximately 4 to 14 individuals per 100,000 each year [2,3]. The majority of cases are viral; less common causes include non-viral infections, medications, vaccines, and systemic autoimmune diseases [[1], [2], [3]]. Among viral etiologies, adenoviruses, enteroviruses, parvovirus B19, human herpesvirus 6, and coronaviruses have been implicated in cardiac injury [4,5]. The true incidence is unknown, as the diagnosis is often based on variable or less standardized criteria. Currently, myocarditis is primarily diagnosed using cardiac magnetic resonance imaging (CMR), which has largely replaced conventional endomyocardial biopsy [6]. Endomyocardial biopsy has limited sensitivity, owing to the patchy nature of myocarditis, and is associated with potential complications. Given these limitations, CMR is currently regarded as the diagnostic modality of choice [7]. CMR is noninvasive, highly sensitive, and can differentiate myocarditis from other cardiac diseases as well as distinguish acute from chronic forms. The clinical presentation of myocarditis ranges widely, from asymptomatic or nonspecific symptoms such as chest pain, dyspnea, fever, syncope, and palpitations, to severe manifestations including cardiogenic shock or sudden death. The severity is determined primarily by left ventricular ejection fraction (LVEF) [[1], [2], [3], [4], [5],8]. Mild myocarditis is usually a self-limiting condition and requires no specific treatment other than antipyretics and analgesics. However, in rare cases of fulminant myocarditis, signs of heart failure (HF) and malignant arrhythmias may predominate, carrying a poor prognosis [[1], [2], [3]]. In most cases, LVEF recovers within weeks following the index event. Several biomarkers have been investigated for risk assessment in myocarditis, including troponin, C-reactive protein (CRP), natriuretic peptides, and interleukin-6. Additionally, novel non-coding microRNAs are strongly associated with myocarditis and can help differentiate it from acute coronary syndrome (ACS) [[9], [10], [11], [12]]. Fibrinogen is a protein that serves both as an indicator of coagulation status and as an inflammatory marker, functioning as an acute-phase reactant. In contrast, albumin levels are inversely correlated with the intensity of the inflammatory response [13,14]. The fibrinogen-to-albumin ratio (FAR) is an inflammatory marker that has been shown to correlate with outcomes in chronic CAD, ACS, and HF [15]. Given the established association between FAR and other conditions involving cardiac injury, coupled with the limited research examining its relationship with outcomes in clinically suspected acute myocarditis, we aimed to investigate the prognostic value of FAR in this setting.

2. Methods

2.1. Study population

This retrospective study was based on data from patients hospitalized with clinically suspected myocarditis at Galilee Medical Center between 2019 and 2023. Patients were identified using the terms “acute myocarditis,” “acute idiopathic myocarditis,” and “myocarditis unspecified.” The diagnosis was established based on clinical presentation (e.g., chest pain, dyspnea, palpitations, or syncope), the rise and fall of troponin levels, and characteristic echocardiographic findings, including global left ventricular (LV) dysfunction with or without pericardial effusion. Patients included in the study were classified as having “clinically suspected myocarditis”. Although CMR was recommended for all patients, it was performed in only one-third, confirming the diagnosis. We excluded patients with hemodynamic instability requiring mechanical circulatory support, as well as those with missing laboratory or echocardiographic data. When myocardial infarction was suspected, invasive coronary angiography (ICA) or cardiac computed tomography angiography was performed to exclude coronary plaque rupture. Takotsubo cardiomyopathy was considered unlikely in our patients, as the diagnosis is usually confirmed by a typical echocardiographic pattern (basal hypercontractility with apical akinesis). Pericarditis was diagnosed based on at least two of the following four criteria: characteristic chest pain, new pericardial effusion, pericardial friction rub, and typical ECG changes. Cases of pericarditis with significant myocardial involvement, defined by marked troponin elevation (myopericarditis), were also included. Treatment was tailored according to disease severity and clinical course. Mild symptoms were managed with antipyretics and analgesics, whereas oral or intravenous diuretics were administered in cases of volume overload. In patients with LV dysfunction and signs of HF, beta-blockers and renin–angiotensin–aldosterone system (RAAS) inhibitors were initiated.

2.2. Echocardiography

Patients underwent echocardiography (Philips EPIQ 7 system, equipped with an EPIQ X8-2 t transducer; Philips, Andover, MA, USA) within 1–2 days of admission. LVEF was calculated using Simpson’s biplane method from apical four-chamber and two-chamber views, according to the following formula:

LVEF = [(LV end-diastolic volume − LV end-systolic volume) / LV end-diastolic volume] × 100. Diastolic function was assessed by calculating the E/eʹ ratio, using the average of septal and lateral mitral annular tissue Doppler velocities. Other parameters included the left atrial volume index (LAVI), which was measured at end-ventricular systole and calculated as LAVI = left atrial volume / body surface area, and the left ventricular end-diastolic diameter (LVEDD), measured in the parasternal long-axis view. Patients with reduced LVEF during the index hospitalization underwent a follow-up echocardiographic assessment 4–8 weeks later.

2.3. ECG and laboratory parameters

Electrocardiographic (ECG) changes were recorded in all patients upon admission and repeated when necessary. Pathological ECG changes included ST-segment depression measured 80 ms after the J-point, or new T-wave inversion at the nadir. Atrial fibrillation and ventricular tachycardia, when documented during hospitalization, were defined as complications. Fibrinogen levels are expressed in mg/dl and albumin in gr/dl. High-sensitivity troponin I (Hs-TnI) levels were measured using the ARCHITECT c-TnI assay (Abbott). Abnormal Hs-TnI levels were defined as >20 ng/L for women and >30 ng/L for men. N-terminal prohormone brain natriuretic peptide (NT-proBNP) is expressed in pg/mL and was determined using electrochemi-luminescence Elecsys immunoassay (Abbott Laboratories, Dundee, UK)), in which abnormal levels are above 125 pg/mL. Additional laboratory parameters included C-reactive protein (CRP), hemoglobin, white blood cell (WBC) count, and renal function.

2.4. Outcomes

Clinical outcomes included length of stay (LOS), recurrent admissions (for chest pain or HF symptoms), and complications such as arrhythmias (primarily atrial fibrillation and ventricular tachycardia) and signs of HF, including pulmonary congestion or edema. Echocardiographic outcomes included LVEF during the index hospitalization and at 4–6 weeks of follow-up in patients with reduced LVEF.

2.5. Statistical analysis

Categorical variables are expressed as percentages, and continuous variables are expressed as mean with standard deviation (SD)or median with interquartile range (IQR), as appropriate. Kaplan Meier curve was used to describe readmissions. Correlation between continuous variables was evaluated by Spearman test. Correlation between fibrinogen, FAR, and troponin with readmissions was evaluated by Mann-Whitney test, and for albumin with independent t test. he association of variables with complications was assessed using the Mann–Whitney test for fibrinogen, FAR, and troponin, and the independent t-test for albumin. Performance of FAR in predicting lower LVEF was assessed by Receiver Operating Characteristic (ROC) curve. In patients with reduced LVEF, Mann-Whitney test was used to assess the ability of FAR in predicting recovery. Multivariate logistic regression was performed to adjust for different variables in predicting LVEF recovery. All statistical analysis were performed using IBM SPSS version 27.

2.6. Ethical considerations

The study was conducted according to the guidelines of the Declaration of Helsinki and approved by the local ethical committee of Galilee Medical Center. Informed consent was waived by the Institutional Review Board due to the retrospective nature of the study.

3. Results

156 patients were identified with clinically suspected myocarditis. Of them, 3 patients were excluded due to hemodynamic instability, 30 for missing laboratory data, and 5 for incomplete echocardiographic reports. A total of 118 patients (80.5 % male, median age 31 [IQR, 22–42] were included in the final analysis. The study flowchart is depicted in Fig. 1 and the baseline characteristics in Table 1.

Fig. 1.

Study flowchart. After excluding 38 patients (missing data, unstable patients), 118 were included in the final analysis. FAR, fibrinogen-to- albumin ratio.

Table 1.

Baseline characteristics of the study population.

N = 118
Age (median, IQR)	31 (22, 42)
Male, n (%)	95 (80.5 %)
Hypertension	12 (10.2 %)
Diabetes mellitus	10 (8.5 %)
Hyperlipidemia	28 (23.7 %)
Coronary artery disease *	5 (4.2 %)
Tobacco use	54 (45.8 %)
Chronic kidney disease**	23 (19 %)

^*Defined as any history of acute coronary syndrome or coronary intervention.

^**Defined as estimated glomerular filtration rate < 60 ml/min/1.73 m².

Vital signs, laboratory results, and echocardiographic findings of the study population are presented in Table 2. The mean LVEF was 55 %, and pathological ECG changes were documented in 46 patients (39 %). The median troponin and fibrinogen levels were 2558 ng/L and 508 mg/dL, respectively (maximal values during hospitalization), with a calculated median FAR of 120.

Table 2.

Vital Signs, laboratory findings, and echocardiographic parameters.

Systolic blood pressure(mmHg), Mean ± SD	120 ± 15
Heart rate (beats per minutes), Mean ± SD	78 ± 13
ECG changes, n (%)	46 (39 %)
Troponin (ng/L), Median [IQR]	2,558 [589, 8318]
CRP (mg/L), Median [IQR]	59 [24, 133]
Fibrinogen (mg/ dL), Median [IQR]	508 [361, 650]
Albumin (g/dl), Mean ± SD	4. ± 0.4
FAR, Median [IQR]	120 [88, 169]
Creatinine (mg/dl), Mean ± SD	0.9 ± 0.3
WBC (×103/ul), Mean ± SD	10 ± 4
Hemoglobin (g/dl), Mean ± SD	14 ± 1.6
LVEF (%), mean ± SD	55 ± 10
LVEDD (mm), mean ± SD	48 ± 6
LAVI (ml/m2)	25 ± 6
E/A ratio	9 ± 2

CRP, C-reactive protein; ECG, electrocardiogram, FAR, fibrinogen-to-albumin ratio; IQR, interquartile range; LAVI, left atrial volume index; LVEDD, left ventricular end diastolic diameter; LVEF, left ventricular ejection fraction; SD, standard deviation; WBC, white blood cells.

3.1. Disease course

Most patients were classified as having clinically suspected myocarditis based on clinical presentation, elevated troponin, and echocardiographic findings. CMR was performed in 38 patients (32 %) either during the index hospitalization or a few weeks later, whereas ICA was performed in 20 patients (17 %). The median length of stay (LOS) was 4 days. Clinical complications occurred in 30 patients (25.4 %), and 29 patients (24.6 %) were readmitted with chest pain or HF symptoms. Data are summarized in Table 3.

Table 3.

Summary of hospitalization course.

N = 118
CMR, n (%)	38 (32)
ICA, n (%)	20 (17)
LOS (days), Median [IQR]	4 [3,5]
Complications, n (%) Atrial fibrillation, n (%) Ventricular tachycardia, n (%) Pulmonary congestion, n (%)	30 (25.4) 5 (4.2) 3 (2.5) 22 (18.7)
Diuretics	22 (18.7)
Beta blockers	34 (28.8)
ACE inhibitors/ARB's	30 (25.4)
Mechanical ventilation, (%)	0 %
Mechanical circulatory support, (%)	0 %
Readmissions	29 (24.6)

ACE, angiotensin converting enzyme; ARB, angiotensin receptor blockers; CMR, cardiac magnetic resonance; ICA, invasive coronary angiography.

We conducted a comparison between patients with (32 %) and without (68 %) available CMR (Table 4).

Table 4.

Baseline characteristics of patients with and without CMR.

n	With CMR 38 (32)	Without CMR 80 (68)	P value
Male, n (%)	29 (90)	66 (83)	0.63
Age	33 (22, 39)	31 (24, 46)	0.57
Hypertension, n (%)	3	9	0.57
Diabetes Mellitus, n (%)	4	6	0.58
Hyperlipidemia, n (%)	10	18	0.65
Coronary artery disease, n (%)	1	4	0.55
Tobacco use, n (%)	22	32	0.07
Chronic kidney disease, n (%)	5	18	0.23
ECG changes, n (%)	18	29	0.25
LVEF (%), mean ± SD	56 ± 8	55 ± 11	0.62
FAR, Median [IQR]	92 (62,160)	114 (82, 183)	0.12

During 5 years of study period, 29 patients were readmitted, The Kaplan-MEIER curve for readmissions is illustrated in Fig. 2.

Fig. 2.

Kaplan-Meier curve for readmissions. Kaplan-Meier curve for readmissions following the index hospitalization. Patients were readmitted for recurrent chest pain or heart failure symptoms and signs.

A statistically significant inverse correlation was found between FAR (r = -0.313, p = 0.001), fibrinogen (r = -0.287, p = 0.002), troponin (r = -0.282, p = 0.002) and LVEF. Serum albumin showed no significant correlation with EF (r = 0.099, p = 0.286). FAR remains independent factor for predicting reduced LVEF after adjusting for age, gender, CKD, CAD, and diabetes. These relationships are illustrated in Fig. 3(A–C).

Fig. 3.

Correlation between troponin (A), Fibrinogen (B) and FAR (C) with EF. A negative correlation was demonstrated between troponin levels (A) and cardiac function (A). high troponin levels are well established as a marker of cardiac injury during acute myocarditis. Fibrinogen levels (B) also exhibits similar correlation as a marker of increased inflammation and more severe disease. FAR (C) has the most powerful correlation (r = -0.3) when compared to troponin and fibrinogen. EF, ejection fraction.

Using ROC analysis we defined the performance of FAR in predicting LVEF as depicted in Fig. 4. The classifier achieved an area under curve (AUC) of 0.69 (95 % CI, 0.58–0.78, p = 0.001).

Fig. 4.

ROC curve of FAR performance for predicting LVEF. Receiver operating characteristic (ROC) analysis showing the performance of FAR in predicting EF. A cut off value of 114 may be used to predict reduction in LVEF.

According to the analysis, FAR cut off value of 114 may be used for predicting LV dysfunction during the acute episode. High FAR was not correlated with longer LOS (p = 0.07), readmissions (p = 0.29) or to complications (p = 0.22).

4. Discussion

We studied the association between FAR and clinical outcomes in patients with clinically suspected acute myocarditis and found that FAR was correlated with reduced LVEF and may predict poor outcomes in the acute phase. While the relationship between troponin levels and LVEF in acute myocarditis has been extensively investigated, our findings regarding FAR are novel. In the context of acute myocarditis, LVEF is an indicator of disease severity and has been considered a strong predictor of both short- and long-term outcomes [16]. We hypothesize that, similar to other inflammatory conditions, fibrinogen and albumin levels and consequently FAR may be altered in clinically suspected myocarditis depending on disease severity. FAR is a relatively novel biomarker that can be easily calculated from a simple blood test and may be measured serially if required. High FAR levels were previously shown to predict poor outcomes in ACS and chronic CAD [17,18]. All biomarkers of myocardial damage in acute myocarditis, such as troponin, BNP/NT-proBNP, and CRP, have certain limitations [19]. CRP is a nonspecific marker of systemic inflammation and does not necessarily increase in acute myocarditis [1]. Elevation of NT-proBNP reflects increased myocardial wall stress, as seen in HF, and is therefore elevated in myocarditis due to cardiac dysfunction. It may also serve as a prognostic marker in cases of myocarditis with HF [20,21]. However, fulminant cases of myocarditis may be paradoxically associated with normal or modest elevation in NT-proBNP levels, probably explained by the “burned-out” heart in fulminant myocarditis with limited capacity for natriuretic peptide release [22]. In such patients, FAR may be more useful for assessing disease severity, since fibrinogen levels are not directly influenced by cardiac tissue damage. We found that high fibrinogen levels and an elevated FAR were correlated with reduced LVEF in the acute phase, whereas albumin was not. Albumin is considered a negative acute-phase reactant, and its levels tend to decrease during inflammatory conditions due to increased capillary permeability [23]. The half-life of albumin is approximately 20 days, and therefore fluctuations in its levels are not as sensitive as those of fibrinogen [14,24]. We did not find a correlation between FAR and LOS, although reduced LVEF reflects disease severity. This may be explained by the fact that LOS can be influenced by several personal and social factors beyond disease severity. Moreover, we included only hemodynamically stable patients with clinically suspected myocarditis in whom LOS is relatively short due to the absence of major complications. FAR was also not useful for predicting persistent LV dysfunction. As persistent LV dysfunction is multifactorial, it may be better assessed using a combination of parameters, including CMR findings, clinical presentation, and laboratory results. Several clinical and laboratory parameters have been used for risk assessment in clinically suspected myocarditis, among which LVEF is probably the most powerful single predictor [1]. Other parameters include, clinical presentation, older age, ECG changes (alterations in ST segments, T-wave changes, Q waves, atrioventricular block, and bundle-branch block), and the presence of arrhythmia particularly ventricular tachycardia or fibrillation on admission [25,26]. We recommend the routine use of FAR for risk stratification in clinically suspected myocarditis, alongside other clinical, echocardiographic, and laboratory parameters. The optimal assessment of acute myocarditis severity likely involves a combination of several biomarkers, together with the clinical course and imaging modalities, primarily echocardiography and CMR. The current study suggests a potential role for FAR as a marker of disease severity, particularly in predicting cardiac dysfunction. Future studies may focus on serial measurements of FAR to assess disease progression, measuring FAR levels in fulminant myocarditis, and probably measuring FAR one month following the acute event, ideally combined with the correlation with late gadolinium enhancement in CMR.

5. Limitations

Our study has several limitations. First, 68 % of patients were classified as clinically suspected myocarditis since CMR was performed in only 32 % of cases (either during the index hospitalization or a few weeks later). However, the diagnosis is strongly suggested when the clinical presentation, cardiac biomarkers, and echocardiography are indicative of myocarditis. Second, our study is single-center with a relatively small sample; nevertheless, the main patient characteristics (gender, age, comorbidities) are very similar to those reported in multicenter studies. Third, we included only hemodynamically stable patients; therefore, the results are not generalizable to fulminant myocarditis. Fourth, we did not perform serial measurements of FAR. Serial measurements during hospitalization may provide greater insight into the role of FAR in monitoring disease progression and predicting severity. Fifth, due to the retrospective design, adjustment for potential confounders may be necessary.

6. Conclusions

FAR is a novel inflammatory biomarker that may be useful for risk assessment and predicting LV dysfunction in clinically suspected myocarditis. However, FAR may not be reliable for predicting long-term outcomes or LV function recovery.

CRediT authorship contribution statement

Lihi Levi-Gofman: Writing – original draft, Project administration, Investigation, Formal analysis, Data curation, Conceptualization. Shaul Atar: Methodology, Investigation, Conceptualization. Gassan Moady: Writing – review & editing, Methodology, Investigation, Formal analysis, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

The data presented and analyzed in the study are included in the article. Further inquiries can be directed to the corresponding author.