Document Type : Original Article
Authors
Department of Paediatrics, University of Calabar Teaching Hospital (UCTH), Calabar, Cross-River state, Nigeria
Abstract
Keywords
Sickle cell anemia (SCA), a severe disease caused by a point mutation in the β-globin gene (HBB:c.20A>T) and characterized by two hemoglobin S (HbS) molecules in each erythrocyte, is the commonest hemoglobinopathy in Nigeria with a prevalence of 20 per 1,000 births [1][2]. The resultant effect of these two HbS molecules is the polymerization and subsequent sickling of the erythrocytes during stressful conditions like infections and dehydration. This leads to sickled erythrocytes’ adherence to and blocking of the vascular endothelium, resulting in tissue ischemia and painful conditions such as vaso-occlusive crises, which can be life-threatening [3][4].
The early life mortality rate among children with SCA has been estimated to be 50 – 90% [5]. Therefore, the World Health Assembly recognized SCA as a global health burden [6]. Nigeria has the largest number of SCA cases worldwide, and the number of newborns with SCA is expected to increase from 91,000 in 2010 to about 140,900 in 2050 [7]. This may worsen the Nigerian national health burden since more funds will be needed for disease treatment despite the projected increase in the future national income [7].
There are varied phenotypic expressions of SCA among affected children ranging from asymptomatic disease to multiple organ failure. These phenotypic expressions are largely influenced by geographical, environmental, nutritional, infectious, and genetic (haplotype) factors [8][9]. As a result of this conundrum, it has been difficult to predict the clinical severity of SCA despite recent efforts in haplotype mapping. Despite the clear differences in clinical features between the haplotypes, clinical variation among patients with the same haplotype exists [9].
The most common clinical feature in children with SCA is painful crises which may develop following worsened steady-state sickling-unsickling cycles of the erythrocytes by stimuli like infectious agents [3][10]. This usual asymptomatic baseline sickling-unsickling of the erythrocytes at the venular and arteriolar ends of the capillaries during steady state may be the reason for abnormal electrocardiogram (ECG) [3]. The sickling may lead to local cardiac microvascular obstruction (myocardial ischemia) which may be intra-myocardial (T-wave inversion), patchy sub-endocardial (ST depression), or severe transmural (ST elevation) [11][12]. This ischemia may be accompanied by the rise and fall of cardiac troponin T, implying myocardial injury by definition [13]. Electrocardiographic patterns like prolonged QTc interval have also been associated with mortality in SCA patients [14].
There is no existing universal system for SCA scoring or disease severity classification. Compared to the traditional use of clinical features in severity classification, the genetic-aided severity classification of SCA cases is clearer and more accurate [15]. However, efforts have been made to define and classify the clinical characteristics of SCA, especially in resource-poor regions where facilities for genetic tagging are scarce. In this context, a scoring format has been developed and validated for sickle cell disease severity classification in Nigeria following novel research in the Netherlands that utilized modern clinimetry [16][17]. This scoring format serves as a classification tool for SCA severity (mild, moderate, or severe) in children during the steady state of the disease [16].
There is a dearth of studies exploring the link between SCA severity and abnormal ECG, especially during steady state when there are essentially no symptoms. On the other hand, there are connections between some abnormal ECG patterns and mortality in SCA. Therefore, this study aimed to find the relationship between abnormal ECG parameters during steady state and disease severity of pediatric SCA.
This cross-sectional study was conducted in the Paediatric Haematology clinic at the University of Calabar Teaching Hospital (UCTH), Calabar, Nigeria from October 2017 to April 2018.
Patients attending UCTH with confirmed SCA cases were consecutively enrolled.
Inclusion criteria
Subjects 4 to 17 years of age and in a steady state of health were considered for this study. Steady state health was defined as when the child was not experiencing pain, had no painful episode in the previous four consecutive weeks, and had no blood transfusion during the last four consecutive months [18]. Children of 4 years and above can provide meaningful and objective self-reporting of pain when present or when it was last felt [19].
Exclusion criteria
Patients with well-known causes of ECG abnormalities such as features of renal disease, respiratory tract infection, severe malnutrition [Body Mass Index (BMI) < -3 standard deviation for age and gender] [20], fever, hypertension (≥95th percentile for age, gender, and height) [21], severe anemia (packed cell volume ≤ 15% or hemoglobin concentration ≤ 50g/L), cardiac disease, heart failure or shock were excluded from the study. Furthermore, Intake of drugs like digoxin, frusemide, captopril, or chemotherapy; presentation with any acute episodic event (crises); hypokalaemia, hyperkalemia, hypocalcemia or low glomerular filtration rate (GFR)-for-age-and-gender, or receiving chronic blood transfusion were also considered reasons for exclusion.
Sickle cell disease severity
Based on the scoring format [16] (Table 1), the following information was obtained from recruits: number of painful crises in the previous 12 months, number of blood transfusion during the last 12 months, previous stroke, acute chest syndrome, pneumococcal meningitis, avascular necrosis of the femur, gallstone, chronic leg ulcer, osteomyelitis or priapism and number of hospital admissions in the previous 12 months from their hospital records. Additionally, the liver and splenic sizes, packed cell volume (PCV), and total white cell count (TWBC) on physical examination were considered for scoring. Scores < 8 were graded as a mild disease; 8-17 were graded as moderate disease, while cases with scores > 17 were graded as severe.
Electrocardiography
The graphical recording of each subject’s cardiac cycle was performed with GE Marquette Hellige MAC® 1200ST ECG device (GE Medical Systems Information Technologies, Milwaukee, WI, USA). The electrodes were placed over the limbs and the anterior chest wall, with the subject lying calmly in supine position. Electrode gel was used to improve adhesion between the skin and the electrode. After a waiting period of 10 seconds to allow the polarization voltages to stabilize, the recording was initiated using the 12-lead mode. The ECGs were interpreted using Pediatric standards [22]. The criteria for defining each of the ECG findings include:
Sinus arrhythmia and first-degree atrioventricular block are typical findings in children of the study age group [23][24]. Hence, the other ECG findings above are abnormal.
Laboratory tests
A blood sample (5 ml) was drawn from each child without a tourniquet. Two milliliters (2 mL) of blood was analyzed for full blood count using Sysmex KX-21NTM Automated Haematology analyzer (Sysmex Corporation, Kobe, Japan), which gave PCV and TWBC for each subject. Sera of the remaining 3ml of blood was analyzed for potassium, total calcium, and creatinine using ion selective electrode, cresolphthalein complex one, and modified Jaffe kinetic methods, respectively. These analyses were consecutively done as the subjects were enrolled. A total of 64 subjects were initially considered for the experiment. However, due to the laboratory test, eight were excluded because of hypocalcemia. Those with normal potassium, calcium, and glomerular filtration rate (GFR) for age and gender were included.
Social class grouping
Subjects were classified according to their socioeconomic classes [25].
The required clearance and registration were obtained from the Health Research Ethical Committee of UCTH (NHREC/07/10/2012). Informed consent was also obtained from each parent/guardian of the subject. Additionally, verbal assent was obtained from those at least 7 years of age.
Data were analyzed using Statistical Package for the Social Sciences (SPSS) version 23 software. All numerical data except PCV and QTc interval had skewed distributions and were presented as the median and interquartile range (IQR). Proportions were compared using Fisher’s Exact Test. Kruskal Wallis H-test was used to compare variables with more than two groups, while One-way Analysis of Variance (ANOVA) was used in comparing PCV. Relationships between values were assessed with Spearman rank correlation. The significance level was set at <0.05.
Fifty-six subjects participated in the study with a median (IQR) age of 10.0 (7.0 – 13.8) years and male to female ratio of 1.2:1. The median (IQR) oxygen saturation (SPO2) and arm-span/height ratio were 97% (95-99%) and 1.03 (1.00-1.06), respectively. Most subjects (78.6%) had mild disease, while severe disease was observed in 7.1% of subjects. The median (IQR) severity score was 4.0 (3.0-6.5), ranging from 1 to 19. Most subjects were of the upper social class (73.2%). Association analysis showed that neither demographic nor general factors were significantly associated with SCA severity (Table 2).
At least a third of the subjects had sinus arrhythmia, prolonged P-wave duration, T-wave changes, or left ventricular hypertrophy. The highest prevalent abnormality was T-wave changes observed in 55.4% of the subjects. On the other hand, there was no ST-segment shift or tall P-wave. No statistically significant association between any ECG abnormality and SCA severity was observed (Table 3). The mean (SD) QTc interval was 0.433 (0.02) seconds.
The highest number of ECG abnormalities per subject was five, observed in three subjects. These abnormalities exclude sinus arrhythmia and first-degree heart block. Fifty-four subjects had abnormal ECG, with about 70% having two or more (multiple) ECG abnormalities. It is also worth noting that all four severe SCA cases experienced multiple ECG abnormalities. However, the relationships between the number of abnormalities and SCA disease severity were not statistically significant (Table 4).
Prolonged QTc interval had a weak negative but statistically significant correlation with SCA disease severity. Other variables had extremely weak (0 to 0.09) to weak (0.1 to 0.29) correlations with disease severity that were not significant (Table 5).
This study showed that mild SCA (78.6%) was more prevalent than moderate (14.3%) and severe diseases (7.1%). The relationship between the three levels of SCA disease severity and age, gender, social class, SPO2, anthropometry, blood pressure, and PCV were not significant. A similar relationship was observed between the ECG findings and SCA disease severity except for prolonged QTc interval, which showed a significant negative relationship with SCA severity. Multiple ECG abnormalities occurred in most of the subjects (69.6%).
The abnormal ECG findings observed include sinus tachycardia, prolonged P-wave duration, prolonged QRS duration, prolonged QTc interval, left ventricular hypertrophy, and T-wave changes, which were most prevalent. These abnormal ECG patterns were found across all disease severity levels and did not substantially occur in any disease severity level. However, their occurrence may be due to the background sickling-unsickling cycles of the hemoglobin S erythrocytes in the cardiac capillaries that cause ischemia-reperfusion injuries and alteration of local myocardial action potential conduction. These ECGs have also been shown to occur during steady state [13]. The current observations imply that abnormal ECG is not due to disease intensity but rather the disease itself. Therefore, severe SCA is not associated with abnormal ECG more than the other low severity levels.
A preponderance of mild disease was observed, with most cases experiencing abnormal ECG patterns. The predominance of mild disease could be due to the prevalent (86.9%) SCA Benin haplotype in Nigeria [26]. It has been shown that Benin/Senegal heterozygous haplotype is associated with lower frequencies of hospitalizations, blood transfusions, and painful episodes than other haplotypes [26]. This haplotype combination may explain the high frequency of mild disease in the current study population, since Senegal and Benin haplotypes are associated with high and intermediate fetal hemoglobin (HbF) levels [9]. The HbF resists sickling of erythrocytes, and higher levels are associated with mild disease [16]. The preponderance of mild disease in this study contrasts with findings from adults with SCA who expressed moderate disease (65%) more than the mild disease (23.3%) [27]. This may be explained by the fact that childhood is characterized by different phases of development as they grow with good adaptation to the environment and infections, unlike adulthood, which is static [28].
Interestingly, prolonged QTc interval was observed to have correlated inversely with levels of disease severity. This correlation implies that prolonged QTc interval occurred more in those with mild disease. However, the relationship between disease severity and prolonged QTc interval was weak. The QTc interval represents the propagation of action potential across the cardiac ventricles comprising depolarisation and repolarisation [22]. Sickled erythrocytes blocking of the cardiac microvasculature with resultant ischemia results in the prolongation of ST-segment, which is a component of QTc interval. This irregularity may cause refractoriness in some parts of the ventricles around which electrical activities can repeatedly pass, leading to ventricular arrhythmia [29]. Ventricular arrhythmia can cause death, and prolonged QTc intervals are associated with mortality [14]. Another typical effect of cardiac ischemia is T-wave changes, which was the predominant abnormal ECG in this study.
Age, gender, social class, anthropometry, and blood pressure did not significantly relate to levels of disease severity. This observation means that phenotypic and environmental factors may not substantially influence the rate of baseline repeated sickling of the erythrocytes and hence, clinical features of persons with a similar haplotype. This finding is similar to previous reports, which showed that gender and social class did not influence disease severity [16]. Furthermore, the weak age association with disease severity in this study was similar to the findings reported in the Netherlands [17] but contradicts another validation study done in Nigeria [17], where a strong association was found between age and SCA severity. This difference may be due to the larger sample size in the validation study (115 subjects) than that in the Netherlands (92 subjects) and this study (56 subjects).
Abnormal ECGs occur similarly in the three levels of SCA disease severity. However, prolonged QTc interval, which has been linked with mortality in SCA, was substantially associated with mild disease, the predominant disease severity type in the studied locality. Therefore, regular electrocardiography during the steady state is recommended in order to identify patients at risk of ventricular arrhythmia for closer cardiac-oriented management.
A larger sample size would have given more certainty to the findings of this study. Fetal hemoglobin measurement and SCA haplotype assignment can substantiate the conclusions of this study in addition to elucidating the genetic-phenotypic links in children with SCA.
Conflict of interest statement
The authors declared no conflict of interest.
Funding statement
The authors declared that no funding was received in relation to this manuscript.
Data availability statement
The authors declared that all related data are included in the article.