Serum growth differentiation factor-15 (GDF-15) is a biomarker of cardiac manifestations in children with COVID-19

COVID-19 infection can cause detrimental effects on the cardiovascular system and result in complications that include myocardial injury, arrhythmia, arterial and venous thromboembolism, myocarditis, cardiomyopathy, cardiogenic shock, and cardiac arrest [1]. As a result, grouping patients into risk categories for suitable early anticoagulant or fibrinolytic therapy is difficult [2].

GDF-15 is a cytokine produced due to inflammation, tissue damage, and mitochondrial stress [29]. It belongs to the Transforming Growth Factor β (TGFβ) family of proteins [3]. In normal physiological conditions, GDF-15 mRNA is expressed by a wide variety of cells and tissues [4]. Furthermore, GDF-15 is increased in many different inflammatory diseases and infections. Its high plasma level during the cytokine storm of COVID-19 greatly increases the incidence of cardiovascular injury and can thus be considered as a general biomarker of disease severity, which is increasing with advanced disease and associated with all-cause mortality [5].

As a circulating myokine, GDF-15 is increased as a metabolic response to muscle-specific mitochondrial stress and muscle dysfunction in humans [6]. GDF-15 is expressed weakly in heart muscle under normal physiology; however, the production of GDF-15 is upregulated in cardiomyocytes following ischemia–reperfusion injury. The mechanical stress and injury on the myocardium induce the secretion and release of GDF-15 [3]. Also, The basal expression of GDF-15 is very low in skeletal muscles compared to other organs, but GDF-15 expression and secretion can be induced by various stress conditions. GDF-15 is thus considered a myokine and cardiokine [7]. It is also known to play a role in regulating appetite and energy balance, making it a key factor in metabolic and immune responses [8‐10].

Recent research has shown that GDF15 levels increase in response to COVID-19 infection, particularly in adults [11], but its role in children is still not known. This study aimed to assess the levels of GDF-15 in children with COVID-19 as a biomarker of myocardial and muscular affection in those patients.

From December 2021 to March 2022, this observational cross-sectional study was conducted on 72 COVID-19 children and 72 healthy children (age and sex-matched with the study group) as a control group, at Children’s Hospital, Ain Shams University, Cairo, Egypt.

Inclusion Criteria; Patient: Children aged from 1 month to 18 years with confirmed COVID-19 infection using real time-PCR (RT-PCR) with clinical presentation and radiological findings of COVID-19 [11]. Convenience samples were collected from the hospitalized children in the isolation department. Controls: Pediatric children with no clinical or laboratory evidence of COVID-19 infection (negative RT-PCR). No history of contact with positive patients. No clinical evidence of a recent infection. Age and sex matched with the patients.

Patients were subjected to history taking for personal data (age and sex), contact with a COVID-positive patient, duration of hospital admission, disease comorbidities, and symptoms at the onset of the disease. Patients were stratified according to severity into mild, moderate, severe, and critically ill [12]. Clinical examination was done for vital data and weight (plotted on weight for age Z-scores [13]. Investigations were done including complete blood count (CBC), C-reactive protein (CRP), liver and kidney functions, D-dimer, ferritin, creatine kinase MB (CK-MB), troponin I, prothrombin time (PT), partial thromboplastin time (PTT), and the international normalized ratio (INR). Computerized tomography (CT) chest was graded according to the COVID-19 Reporting and Data System (CO-RADS) score from 1–6 [14] and echocardiography (ECHO) was done for all COVID-19 patients.

There was no statistically significant difference between patients and controls regarding age and sex, but there was a statistically significant higher GDF-15 level in patients with COVID-19 than in controls (p < 0.01) (Table 1).

Children who had fatigue and dyspnea had significantly higher serum GDF-15 levels (P = 0.031, 0.043, respectively) than those without fatigue and dyspnea. Additionally, its level was significantly greater in underweight children than in children with normal weight (P = 0.043) and in children whose hospital stays were longer than 10 days (P = 0.023) (Table 2). The COVID-19 patients who were admitted to the PICU, with severe clinical grading and those who died had significantly higher GDF-15 than those admitted to the ward, having mild to moderate grading, and those who survived (Table 3).

Serum GDF-15 had a significant positive correlation with ESR, LDH, D-dimer, PT, PTT, INR, AST, CK-MB, troponin I, and the number of lung lobes affected by CT chest. While it had a significant negative correlation with weight for age Z-score, oxygen saturation at room air, lymphocytic count, platelets count, and ejection fraction of the heart (Table 4).

ECHO findings show that there was valvular affection in 33.3% of patients which was mainly tricuspid regurge (20.8%). Cardiomegaly was found in 26.4% of patients presented mainly as left ventricular dilatation and hypertrophy (23.6%). Other less common findings were: pulmonary hypertension in 5.6%, and pericardial effusion in 4.2%. Also, there was a highly statistically significant difference in serum GDF-15 level among patients who had cardiac affection including; Echo abnormalities, presence of carditis, and low ejection fraction, compared to those who hadn’t (Table 5).

Regarding the validity of serum GDF-15 level to discriminate children with COVID-19 from controls, its sensitivity was 79.17%, and its specificity was 100.0%, at a cut-off > 272.8 pg/mL. While, regarding the validity of serum GDF-15 level to assess cardiac affection of COVID-19 children, its sensitivity was 92.31%, and its specificity was 95.74% at a cut-off > 873.16 pg/mL. Regarding the validity of serum GDF-15 level to discriminate patients with different clinical severity grading (mild to moderate from severe and critically ill), its sensitivity was 88.57%, and its specificity was 97.3% at cut-off > 688.5 pg/mL (Table 6). Using multivariate linear regression analysis, higher GDF-15 was associated with lower ejection fraction and higher INR in COVID-19 children (Table 7).

GDF-15 elevated alongside other inflammatory biomarkers in COVID-19 patients. Compared to other inflammatory markers including CRP, ferritin, and D-dimer, GDF-15 was more sensitive and specific in diagnosing the early stage of COVID-19 severity and admission to the ICU in seriously afflicted COVID-19 patients. Myhre et al. (2020) study on 123 adult patients found that the GDF-15 serum level was extremely sensitive, specific, and superior to other known inflammatory and cardiovascular biomarkers, and it is linked to ICU admission of patients with severe disease [15].

Fatigue and dyspnea are common symptoms related to COVID-19 and Long COVID that could last 12 weeks postinfection [16]. In our study, Serum GDF-15 was significantly higher in children with dyspnea (P = 0.043), and fatigue (P = 0.031). Myhre et al. [15] also found a positive correlation between serum GDF-15 levels and the presence of fatigue in COVID-19 patients, which strongly supports the role of GDF-15 as a biomarker of muscular and myocardial affection in COVID-19 patients and could suggest the potential role of GDF-15 as a therapeutic target for Long COVID fatigue.

The study of Alserawan et al. [11] on adult patients with COVID-19 found a significant correlation between GDF-15 and other inflammatory biomarkers of COVID-19 severity as CRP, D-dimer, and lymphopenia, which were similar to our results in the pediatric population. Also, in our study, ESR was elevated in 65.3% of patients and this correlated positively with the serum level of GDF-15. This agreed with other studies such as Tanrikulu et al. [17] and Wischhusen et al. [10] both found a positive correlation between GDF-15 and ESR.

Although increased ferritin is a poor predictor of COVID-19 outcome [5], we could not prove its correlation with the GDF-15 level in our study. This is in contrast with the studies of Luis et al. [18], and Alserawan et al. [11] which found a significant positive correlation between GDF-15 levels and ferritin (r = 0.334, 0.285; p = 0.006, 0.013, respectively).

LDH is an enzyme found in nearly all living cells and plays a role in energy production. It is released into the bloodstream when there is cell damage or destruction. Elevated LDH levels can be seen in various diseases and conditions, including heart disease, liver disease, muscle injury, and hemolysis [19]. GDF-15 and LDH are not typically directly correlated, but their levels can be influenced by similar underlying conditions or disease processes.

Coagulation abnormalities, including elevated D-dimer and prolonged PT, INR, and PTT, have been observed in some severe cases of COVID-19, both in children and adults. These abnormalities are indicative of a hypercoagulable state and an increased risk of blood clot formation. Rochette et al. [3] suggested the role of GDF-15 in causing microvascular injury and endothelitis that could explain its role in cardiovascular injury.

In COVID-19, severe cases can lead to acute respiratory distress syndrome (ARDS) and hypoxia, where the body's oxygen levels are significantly reduced. Hypoxia is a serious complication of COVID-19 and can lead to significant tissue damage and organ dysfunction. GDF-15 is predominantly located in endothelial cells and is upregulated by hypoxia. GDF-15 expression is induced by lung injury and is suggested to be a hallmark of tissue injury in many organs [20]. This study showed that serum levels of GDF-15 were significantly correlated with oxygen saturation in room air. There was a positive correlation between the GDF-15 level and the degree of hypoxemia, in addition to a significant increase in the GDF-15 level of patients who presented with dyspnea and those with more affected lung lobes in chest CT. This shows that GDF-15 may be able to detect silent hypoxemic patients, a COVID-19 sign that has been linked to significantly higher risk [15].

The present study showed that GDF-15 levels correlated positively with both CK-MB and troponin I levels (r = 0.570 and 0.639 respectively), and this agreed with the study of Wallentin et al. [21] and Eindhoven et al. [22]. Also, we found a positive correlation between GDF-15 level and cardiac affection. This agreed with Eggers et al. [23] who found that GDF-15 exhibited the strongest incremental value to cardiovascular risk indicators, both in terms of prognostic discrimination and reclassification of cardiovascular risks (95% CI, 0.71 (0.67–0.76)) in comparison to NT-proBNP (P value < 0.001 and 0.040, respectively) in COVID-19 pathophysiology [24]. Similarly, Myhre et al. [15] found that GDF-15 correlated positively with cardiovascular biomarkers, cardiac Troponin (cTnT), and pro-B- type natriuretic peptide (pro-BNP); with correlation coefficients (0.52 and 0.49 respectively) and p < 0.05 for both.

The study of Hacıoğlu et al. [25] found that GDF-15 was detected to be significantly increased in patients with subtle changes in diastolic blood pressure (DBP), end-diastolic volume (EDV), stroke volume (SV), and cardiomegaly. Similarly, Roenningen et al. [26] found that GDF-15 was valuable in detecting the incidence of atrial fibrillation in patients with left atrial structural remodeling and hypertrophy. In addition, Lockhart et al. [27] found that the production of GDF-15 is upregulated in cardiomyocytes following ischemia–reperfusion injury. Also, this high expression has been reported after MI leading to the suggestion that GDF-15 may be a biomarker for heart failure.

In our study, plasma GDF-15 levels were significantly higher in non-survivors in comparison to survivors indicating that it’s a good predictor of mortality. Similarly, the study of Luis et al. [18], Myhre et al. [15], Ahmed et al. [28], and Teng et al. [29] found that the GDF-15 level was significantly higher in COVID-19 patients who died, suggesting its potential role in evaluating the prognosis of patients. Dynamic changes in GDF-15 levels reflected disease progression, with high levels linked to symptom deterioration, followed by a dramatic decline in plasma GDF-15 levels at the time of clinical and radiological improvement and discharge. This study indicates that GDF-15 could be used as a predictor of the progression of the disease [30]. Our cut-off value to detect severity was > 688.5 which is much lower than that detected by Myhre et al. [15] in adult patients with COVID-19 (2252 pg/ml). Also, Hongisto et al. [31] found that ˃7000pg/ml is the cutoff of GDF-15 to predict mortality in adults with cardiogenic shock. Our cutoff level to detect cardiac affection (> 873.16 pg/ml) is also lower than Tantawy et al. [32] which suggested 1500 pg/ml for differentiating patients with heart disease from those without (the study included children and adults without COVID-19). So, more studies are needed to predict the prognostic cutoff value of GDF-15 in pediatrics and adults.

The limitations of our study: 11 cases (15.38%) had comorbid diseases such as renal disease, cerebrovascular disease, and diabetes, and these cases had higher GDF-15 than those with COVID-19 without comorbidities which can probably indicate more severe disease but the effect of the comorbid disease on GDF-15 could not be excluded. Also, our study included only hospitalized children with COVID-19 without including non-hospitalized children.

Based on these findings, GDF-15, which is easily available by commercial assays on large automated analytic platforms, may represent a clinically useful risk stratification tool that provides important pathophysiological insights in patients hospitalized with COVID-19. Thus, in real-life situations with hospitalized COVID-19 patients, serial measurements of GDF-15 can be an aid in patient triage and follow disease progression more efficiently.

GDF15 is emerging as a promising field for research in the context of children with COVID-19. Its role as a potential biomarker for disease severity and a therapeutic target should be explored to improve our understanding and management of COVID-19 in the pediatric population. Further research and clinical trials are necessary to determine the utility and safety of GDF15-related interventions in children with COVID-19.