Contribution of serum FGF21 level to the identification of left ventricular systolic dysfunction and cardiac death

Cardiovascular disease comprises all cardiac and vascular disorders, including coronary arterial disease and peripheral vascular disease, among which coronary atherosclerotic heart disease (CAD) is responsible for a large portion [1]. With improvement in the quality of life in developing countries, the incidence of CAD is increasing rapidly. In particular, left ventricular systolic function has gained attention as it can serve as a good indicator for the prognosis of patients with CAD [2]. The left ventricular ejection fraction as measured by two-dimension echocardiography is commonly used to reflect left ventricular systolic function. The first step in the prevention and treatment of CAD is finding an optimal indicator for the early identification of relevant risk factors. The second step is to develop novel therapies for the complications associated with chronic cardiovascular disease including ischemic vascular disease and heart failure. The first step deserves more attention [3].

In a previous cross-sectional study [4], we found that fibroblast growth factor 21 (FGF21), an important cytokine involved in glycol-lipid metabolism in vivo, was closely correlated with CAD as diagnosed by coronary angiography. FGF21 can also be an independent risk factor for CAD. Therefore, this study aimed to investigate the association of FGF21 in left ventricular systolic dysfunction and cardiac death.

The characteristics of the study patients at baseline are seen in Table 1. The average age was 66.32 ± 10.10 years old. The average BMI was 24.59 ± 3.46 kg/m². Among the patients, 34.4% (75 individuals) were women who were all postmenopausal. At baseline, there were 45.4% smokers, 40.4% of the patients with diabetes, 76.6% with hypertension and 89.9% with dyslipidemia. There were 42.7% of the patients with family histories of cardiovascular diseases.

We determined the relationship between left ventricular systolic dysfunction and FGF21, as well as the traditional indicator NT-pro-BNP, and the identification of these two variables in cases of cardiac death in both cross-sectional and follow-up analysis. We found that FGF21 and NT-pro-BNP were both correlated with left ventricular systolic dysfunction at baseline. Both higher FGF21 and higher NT-pro-BNP levels could predict cardiac death within five years. We therefore concluded that FGF21 was non-inferior to NT-pro-BNP at identifying both left ventricular systolic dysfunction at baseline and cardiac death within five years.

FGF21 is involved in the pathogenesis of cardiovascular diseases and metabolic disorders [11]. FGF21 is confirmed secreted mainly in liver and is characterized by their reduced binding affinity for heparin that enables them to be transported in the circulation to act as a hormone-like cytokine. Its action is not tissue specific but general in vivo [12]. In our previous study, patients diagnosed with coronary artery disease by coronary angiography had higher levels of FGF21. This relationship was independent of traditional risk factors [4]. As was reported in other studies, FGF21 was found to be significantly correlated with subclinical atherosclerosis and lower extremity atherosclerosis in type 2 diabetes patients [13, 14]. FGF21 was presented as a potential biomarker for cardiovascular diseases. In a clinical trial named the FIELD study [15], FGF21 was found to effectively identify the risk of cardiac events. Higher FGF21 levels at baseline could predict overall cardiac outcomes within five years. However, baseline FGF21 levels were not reported to be associated with CVD mortality in the FIELD study, probably because the subjects of that study were type 2 diabetes patients only. In addition, serum FGF21 was reported to be elevated in newly diagnosed T2DM, and positively correlated with carotid and iliac lesions in patients with subclinical atherosclerosis, especially in women [16]. Among youths at risk for metabolic syndrome, FGF21 levels were positively correlated with obesity traits, blood pressure, and lipid profiles [17], indicating its essential role in the metabolic system.

Few studies have reported the relationship between FGF21 and cardiac function as evaluated by echocardiography. Chou et al. [18] reported that FGF21 was associated with diastolic dysfunction in heart failure patients with preserved ejection fractions, including 95 patients with diastolic dysfunction and 143 controls. Multiple logistic regression analysis showed that FGF 21 was independently associated with diastolic dysfunction. Both FGF21 and NT-pro-BNP levels could indicate major adverse cardiac events after one year of follow up. In another study of 1668 patients with CAD reported by Li et al. [19], serum FGF21 levels at baseline were associated with all-cause mortality and with CVD mortality in a U-shaped pattern at a median follow-up of 4.9 years, indicating that patients with both higher and lower FGF21 levels could suffer greater risks of death. Both quartile 1 (120.8 pg/mL) and quartile 3 (357.0 pg/mL) patients had significant higher risks of CVD mortality. In our study, FGF21 was not inferior to NT-pro-BNP at identifying left ventricular systolic dysfunction at baseline. Further analysis showed that, at a median follow up of 5.0 years, patients with higher FGF21 levels (≥321.5 pg/mL) at baseline are at higher risk of CVD mortality, which was also not inferior to NT-pro-BNP. The corresponding cutoff of FGF21 in our study was similar to that of Li Q, et al. Our findings are also consistent with the results of Chou et al. and are more typical of the association between FGF21 and LVEF, as Chou et al. excluded patients with LVEF <45%.

Ventricular remodeling is an adaptive response caused by the hypertrophy of myocardial cells during stress. Mature myocardial cells are differentiated cells that cannot regenerate. Myocardial cells tend to be hypertrophic rather than differentiated during stress, resulting from increased amounts of extracellular matrix and fibroblasts in the heart [12]. FGF21, as a novel cytokine, is involved in this pathogenesis. In rodent models, the expression of FGF21 in cardiac cells can be up-regulated by the hypertrophy of myocardial cells, which is also regulated by the SIRT1/PPARα pathway in the liver [20]. The mass of the heart tends to be larger, with deteriorating cardiac function in FGF21 knock-out rats after continuous infusion of isoprenaline [21]. Moreover, FGF21 were also shown to play an important role in oxidative stress [22]; FGF21 deletion aggravates aortic remodeling and cell death probably via exacerbation of aortic inflammation and oxidative stress [23]. In our study, FGF21 can identify left ventricular systolic dysfunction, which may be due to the above-mentioned mechanism. The secretion of FGF21 via a compensatory response may contribute to these results.

There are some limitations to this study. First, the sample size of this study is small and we did not calculate the sample size at baseline. Our preliminary results should be validated in further studies with a larger sample size, which can provide sufficient statistical power. Second, the subjects enrolled in our study at baseline are all from the department of cardiology, and tend to be elderly. We also intend to recruit subjects from different age ranges to generalize our results. Finally, no relationship between FGF21 and age is observed at baseline, which is not consistent with another study [15]. Patients do not undergo additional FGF21 measurements during their follow up so that we cannot further clarify the exact relationship between FGF21 and age.

In conclusion, we may consider serum FGF21 levels to be correlated with left ventricular systolic dysfunction at baseline. Patients with higher FGF21 levels at baseline are at greater risk of CVD mortality. The role of FGF21 in both left ventricular systolic dysfunction and cardiac death is not inferior to that of NT-pro-BNP.