Cardiopulmonary Complications of Sickle Cell Disease in Children

Abstract

Sickle cell disease (SCD) is an autosomal recessive hemoglobinopathy leading to hemolysis, increased endothelial adhesion, inflammation, and vasculopathy. While most children with SCD have normal pulmonary function, lung capacity and expiratory flows are lower compared to age- and racially matched controls. Airway obstruction dominates in children, with restrictive ventilatory defects becoming more prevalent in adolescents and young adults. Decreased pulmonary function, physician-diagnosed asthma, airway hyperresponsiveness, wheezing, and sleep-disordered breathing are associated with more frequent episodes of acute chest syndrome and vaso-occlusive crisis. Chronic lung disease, thromboembolism, hypoxemia, and sleep-disordered breathing are associated with the development of pulmonary hypertension and ventricular dysfunction which carry significant morbidity and mortality risk in adults. Most treatments for cardiopulmonary complications of SCD are based on guidelines developed for the general population. Although most guidelines do not recommend routine screening of asymptomatic children, patients with cardiopulmonary symptoms should be monitored and treated by subspecialists in a multidisciplinary setting. Disease modifying treatments such as hydroxyurea are attenuating some of the cardiopulmonary complications in SCD. More studies need to be done to assess the effects of newer disease modifying treatments targeting hemolysis and decreasing endothelial adhesion.

1. Introduction

Sickle cell disease (SCD) is a hemoglobinopathy resulting from mutations in the β-globin gene and is inherited in an autosomal recessive pattern [1]. Public health measures in developed countries have helped improve infant morbidity and mortality; however, the long-term complications from this condition represent a global health burden [2]. The point mutation in SCD is a substitution of valine for glutamic acid at the sixth position of the β-globin gene resulting in “sickled” hemoglobin that is less soluble than normal adult or fetal hemoglobin (HbF). With deoxygenation, hemoglobin S becomes polymerized resulting in decreased erythrocyte flexibility forming the characteristic “sickled” shape. These changes alter cellular function, enhance endothelial adhesion molecule expression, impair microvascular flow, and promote hemolysis and vaso-occlusion leading to anemia, hemolysis, vasculopathy, and chronic inflammation [1]. Cardiopulmonary involvement is a major cause of morbidity and mortality in adults. There is growing evidence suggesting earlier onset of pulmonary function abnormalities, and predisposition to disorders, such as asthma, thromboembolism, sleep-disordered breathing (SDB), ventricular dysfunction, and pulmonary hypertension, in children with SCD [3, 4, 5]. It is the aim of this narrative review to summarize current literature describing cardiopulmonary complications of SCD in children and evolving treatments that may improve outcomes.

2. Acute chest syndrome

Acute chest syndrome (ACS) is a known complication of SCD with high morbidity and mortality. Newer data suggest that repeated episodes of ACS may be associated with decreased lung function longitudinally in children, and increased inflammation [6]. Key factors in the pathogenesis of ACS include pulmonary infection, pulmonary infarction, and fat embolism. Identification of exact infectious etiology in children varies and can be identified in over two-thirds of cases, with viruses predominating [7]. Of note, since the initiation of pneumococcal vaccination with PCV-13 (13-valent pneumococcal conjugate vaccine), one cohort in France demonstrated a 41.8% decrease in ACS over time [8]. Diagnosis is clinical, and criteria are the following: new infiltrate on chest X-ray, hypoxemia, and respiratory signs/symptoms [1]. Treatment involves supplemental oxygen, empiric antibiotics to cover Mycoplasma pneumoniae and encapsulated organisms, chest physiotherapy and incentive spirometry, pain control, and, when indicated, transfusion and ventilatory support [9].

3. Pulmonary function abnormalities

Pulmonary function abnormalities have been demonstrated in infants and children with SCD. Koumbourlis et al. [10] found evidence of lower airway obstruction and hyperinflation in 20 infants (aged 3–30 months) with SCD, particularly those with homozygous sickle cell disease (HbSS). Functional Residual Capacity (FRC) was elevated, maximum expiratory flow rates at FRC and time to reach peak expiratory flow were decreased. A cross-sectional study of 22 infants, 6–18 months of age, showed evidence of obstruction, independent of acute chest syndrome (ACS) or vaso-occlusive events (VOE). The investigators found normal lung function in 77% of infants with SCD but lower Forced Vital Capacity (FVC), lower forced expiratory flow at 0.5 seconds (FEF0.5), and forced expiratory flow at 25 to 75% of FVC (FEF25–75) compared with controls [11].

Despite the frequency of respiratory complaints in patients with SCD, longitudinal studies spanning progression of pulmonary function abnormalities from childhood to adulthood are limited. Cross-sectional studies reveal that 70–80% of school-aged children with SCD have normal lung function but values are reduced compared to age-, gender-, and race-matched healthy controls. Although international guidelines recommend the use of lower-limit-of-normal (LLN) criteria (which can vary by age, height, and gender) to define abnormal lung function, several studies on children with SCD use cutoffs (such as a forced expiratory volume in 1 second (FEV1)/FVC < 0.70 or % predicted values below 80%) to distinguish normal versus abnormal [5]. Airway obstruction is the most common abnormality found in children (16–34%). Restrictive defects, which are usually found in adults with SCD, have been described less frequently in children (2–11%) [4, 5, 12, 13, 14, 15]. It is not known if lung deficits in SCD are secondary to underlying airway inflammation or a result of pulmonary injury following ACS or allergic disease [6]. Data from the Sleep and Asthma Cohort (SAC) showed that lung function patterns (obstruction, restriction, and nonspecific ventilatory defects) were not associated with future pain or ACS episodes over a 4-year follow-up period [15]. In a meta-analysis of case-controlled studies in 1889 children from six countries, forced expiratory volume in 1 second (FEV1) % predicted (p < 0.00001) and FVC% predicted (p < 0.00001) were significantly lower than controls; while FEV1/FVC ratio, total lung capacity (TLC) % predicted, and carbon monoxide diffusing capacity of the lungs (DLCO) were not significantly different than controls. Pooled analysis also supported the worsening of pulmonary function with recurrent episodes of ACS [14]. Inflammation in ACS has been associated with lung function deficits. Analysis of data from children in a 2-year randomized control study to examine the effects of vitamin D on SCD found a significant decline in FEV1 (p = 0.015) and FEF 25–75 (p = 0.039), in patients with a history of ACS. IP-10 level, a marker of TH-1 inflammation, was negatively correlated with changes in FVC in ACS patients, suggesting a role of interferon gamma-inducible chemokine receptors with lung function change. In patients with ACS, elevated levels of Th-2 inflammatory markers (IL-4, IL-5, IL-13) and IL-6 (a marker of monocyte inflammation) were found to have a negative influence on FVC and FEV1 [6]. Investigators have emphasized the role of vascular factors on obstructive airway phenotypes in SCD. In a study of 25 children with HbSS and 25 age- and ethnic-matched controls, children with SCD had greater pulmonary capillary blood volume (p < 0.0001) and increased respiratory system resistance (R5% predicted, p = 0.0046, as measured by impulse oscillometry) compared to controls. Among the children with HbSS, pulmonary capillary blood volume was positively correlated with R5% pred and the ratio of residual volume/total lung capacity (RV/TLC) and was negatively correlated with FEV1 and FEF 25–75, suggesting that increased resistance leading to airway obstruction in SCD may be related to increased cardiac output and increased pulmonary blood volume in response to chronic anemia [16].

More sensitive pulmonary function measurements may detect pulmonary function abnormalities earlier than spirometry or plethysmography. A recent study compared lung clearance index (LCI), a measure of ventilatory heterogeneity derived from a multiple breath washout technique, to changes in spirometry and body plethysmography in children with SCD versus controls. LCI (p = 0.0001), intra-acinar ventilation inhomogeneity (S_acin) (p = 0.04), z-scores for FEV1 (p = 0.002), FVC (p = 0.002), and TLC (p = 0.002) but not FEV1/FVC were significantly lower in SCD patients compared to controls. More patients (29%) had LCI >95th percentile of control subjects compared to patients with abnormal spirometry (23% had FEV1 < 5th percentile of the reference population). Significant differences from control subjects in LCI and S_acin but not in conductive ventilation inhomogeneity, (S_cond) and normal FEV1/FVC ratio suggest that the lung function changes were due to patchy peripheral lung disease. The multiple breath washout technique may serve as an early marker of peripheral lung disease before abnormalities in spirometry and plethysmography are detected [17].

Longitudinal cohort studies on children with SCD reveal a variable and inconsistent rate of decline in pulmonary function over time. A single-center longitudinal study in 312 children with SCD found an average decline in FEV1: 2.93% predicted per year for males, 2.95% predicted per year for females; average decline in TLC was 2.15% predicted per year in males, 2.43% predicted per year for females. Although only 18% of children had abnormal lung function at 17 years of age, the predominant change from 8 years of age was an increase in the number of children with a restrictive pattern (2.6% at age 8 vs. 18.7% at age 17 years) [18]. A longitudinal study of two cohorts of SCD patients compared to controls revealed a 0.93% decline in FEV1 over a 10-year follow-up. A younger cohort with a higher incidence of ACS episodes had a greater decline in FEV1 (1.45%) over a 2-year period and proportionately more children developed restrictive defects over time [13].

Treatment with disease modifiers may attenuate pulmonary function decline. A retrospective chart review of 62 children with SCD treated with hydroxyurea (HU) followed from 2000 to 2017 compared their spirometry to that of a group of untreated controls. FVC significantly increased in HU-treated children, while it decreased in controls (7.2 ± 17.1 vs. -3.4 ± 18.2, p < 0.01). HU is a disease modifier in SCD, resulting in increased fetal hemoglobin and decreased episodes of hemolysis [19].

Although universal screening for pulmonary function abnormalities is not recommended by the National Heart, Lung, and Blood Institute (NHLBI) or the American Society of Hematology (ASH) for asymptomatic patients with SCD, it should be considered in children with symptoms of cough, wheezing, shortness of breath, and oxygen desaturation. Results should be interpreted, and patients managed based on the presence of underlying complications associated with SCD, such as sleep-disordered breathing, asthma/wheezing, and/or pulmonary hypertension [5].

4. Asthma and airway hyperresponsiveness

Prevalence of asthma in non-Hispanic black children in the USA was reported to be 12.6% in 2017 by the Centers for Disease Control (CDC) [20]. Asthma is considered a distinct comorbidity in SCD and is associated with the higher rates of ACS, vaso-occlusive/pain crises, and mortality [1, 5, 21, 22]. The reported prevalence of asthma in children with SCD ranges from 8 to 53%, with several studies reporting a higher incidence of asthma in SCD compared to African American children. Differences in reported prevalence vary based on the definition of asthma, such as parental report of asthma, physician diagnosis of asthma, or use of an asthma medication [5, 15, 21, 22]. Multivariate logistic regression analysis of patients in the SAC cohort revealed that the presence of a parent with asthma (p = 0.006), wheezing causing shortness of breath, and/or wheezing with exercise (p = .0001) was 100% sensitive in identifying SCD children with asthma (physician-diagnosed asthma requiring asthma medications) [23].

Asthma and SCD have common inflammatory pathways. Diagnosing asthma in SCD patients is challenging since symptoms of cough, wheezing, shortness of breath, and chest pain, which are consistent with asthma, may also occur with chronic pulmonary inflammation, vasculopathy, and hemolysis in patients with SCD. The hallmarks of asthma pathophysiology, airway obstruction, and airway hyperreactivity from airway inflammation may be found in SCD patients, independent of an asthma diagnosis. Elevated serum IgE and leukotriene levels may also be found in SCD patients who have not been diagnosed with asthma [5, 24, 25]. SCD patients have elevated serum IgE levels, a biomarker of allergic asthma. Serum IgE elevation may be secondary to nonspecific immune activation of TH-2 pathways after ischemia-reperfusion injury in the lung in SCD, leading to an increased risk for asthma [21, 23, 24]. Cysteinyl leukotriene levels are elevated in SCD patients at baseline and during acute pain and ACS episodes. They are most elevated in patients with physician-diagnosed asthma and may contribute to the pathogenesis of vaso-occlusive disease [25]. Asthma is more common in SCD patients with allergies than in those without. Forty-five percent of children and adolescents in the SAC with a physician diagnosis of asthma had at least two positive skin prick tests for common aeroallergens compared to only 15% in those without asthma [23]. Although the prevalence of atopy in SCD is similar to that of the general population, reactivity to environmental allergens has been associated with increased incidence of ACS, independent of asthma diagnosis [21, 23]. Endothelial activation is considered the major pathway by which sickled red blood cells (RBCs) contribute to vaso-occlusion and consequently increased levels of pro-inflammatory cytokines, such as IL-3, granulocyte-macrophage colony-stimulating factor (GM-CSF), prostaglandin 2 (PGE-2), tumor necrosis factor alpha (TNF-α), and IL-6. The heightened inflammation is associated with pulmonary function abnormalities and susceptibility to infection. Heightened TH-2 inflammation has also been demonstrated in SCD mice models sensitized to ovalbumin. Arginine deficiency is a common pathophysiologic feature of allergic asthma and SCD. In SCD, increased hemolysis of red cells leads to low arginine bioavailability and is associated with complications, such as pulmonary hypertension and vaso-occlusive pain episodes. In asthma associated with SCD, increased arginase released by hemolysis and increased pulmonary eosinophilic inflammation decreasing arginine uptake may have an additive effect [26].

Wheezing has been reported in 26% of SCD patients with ACS and may be a marker of SCD severity [24]. In a retrospective cohort study of 262 patients with SCD who presented to an urban emergency department over a 4-year period, 19% had at least one presentation in which wheezing was documented, although less than 50% underwent an asthma diagnosis. Wheezing on physical examination was associated with more than twice the incidence of vaso-occlusive crises and ACS episodes, independent of an asthma diagnosis [27].

Airway hyperresponsiveness has been demonstrated in 17–77% of SCD patients and signifies airway lability secondary to bronchial inflammation [21, 24]. Leong was the first to describe airway hyperresponsiveness by cold air challenge or bronchodilator administration in 64% of SCD patients, independent of an asthma diagnosis [28, 29]. Lactate dehydrogenase (LDH) is a marker of hemolysis and has been associated with hyperresponsiveness to methacholine, suggesting that chronic hemolysis leads to inflammation in SCD, independent of a diagnosis of asthma [21, 24].

A diagnosis of asthma has been associated with increased disease-related morbidity in children with SCD. It is postulated that: (1) asthmatic bronchoconstriction-induced ventilation-perfusion mismatch leads to local tissue hypoxia and subsequent red blood cell sickling and/or (2) increased inflammation from oxidative stress and increased endothelial adhesion molecule expression and vaso-occlusion trigger the cascade of pathophysiologic events leading to respiratory complications in SCD [3, 30]. A pooled analysis of three different cohorts of children with SCD, that is, Cooperative Study of Sickle Cell Disease (CSSCD), Silent Cerebral Infarct Transfusion (SIT), and SAC, resulted in a dataset of 1685 participants, a mean follow-up of 6.1 years, and a total of 10,216 patient years with 23.1% diagnosed with asthma. Using negative binomial regression, with covariates of sex, age, hemoglobin, and white blood cell count, a significant positive association was found between asthma diagnosis and pain (IRR = 1.34, p < 0.001) and ACS episodes (IRR = 1.89, p < 0.001) [21].

SCD patients should be screened for recurrent respiratory symptoms, including wheezing, shortness of breath, exercise limitation, a personal history of allergies/atopy, and a family history of asthma, particularly a parental history of asthma [5, 24]. A validated asthma screening tool such as the Breathmobile questionnaire was found to have 87% sensitivity and 85% specificity in detecting a clinical diagnosis of asthma in school-aged children with SCD [31]. Referral to a pulmonary specialist should also be considered in patients with two or more episodes of ACS, the first episode of ACS in a child <4 years of age, life-threatening episode of ACS requiring RBC exchange transfusion, and concerns for sleep-disordered breathing [29]. Positive screening warrants further evaluation and management, including pulmonary function testing, which may reveal reversible airway obstruction supporting an asthma diagnosis and/or allergy testing if symptoms suggest that sensitization to environmental allergens are exacerbating symptoms [5, 15, 24, 30, 32].

Treatment of the SCD patient with asthma includes optimization of SCD management with the use of disease modifying therapy such as hydroxyurea, which has been shown to reduce morbidity and mortality [24]. When the diagnosis of persistent asthma is suggested by history (repeated wheezing and cough with a bronchodilator response, and personal and family history of atopy and asthma in a first-degree relative), therapy is based on NHLBI and GINA (Global Initiative for Asthma) guidelines [24]. The mainstay of therapy for patients with persistent disease includes anti-inflammatory controllers such as inhaled corticosteroids (ICS) alone or in combination with long-acting bronchodilators (LABA) and leukotriene receptor antagonists. Oral steroids, which are usually used to treat asthma exacerbations, carry a risk for acute vaso-occlusive pain and hospital readmission in SCD patients and should be used cautiously. Inhaled corticosteroids have not been associated with increased vaso-occlusive risk. While leukotriene receptor antagonists may have a role specifically in the treatment of asthma and SCD, they are associated with increased behavioral side effects compared to ICS [5, 21, 24, 30]. Further understanding of the inflammatory pathways in SCD may help to determine more effective treatments specific to this disorder.

5. Sleep-disordered breathing

The spectrum of sleep-disordered breathing (SDB) encompasses obstructive sleep apnea (OSA), central sleep apnea (CSA), and sleep-related hypoventilation, which in combination with hypoxemia may significantly impact patients with SCD [4]. Though the underlying mechanisms are not clearly understood, it has been suggested that children with SCD are at increased risk for SDB, with an estimated prevalence of around 36% [32]. Children with SCD are at increased risk for OSA because of compensatory lymphoid hyperplasia and subsequent adenotonsillar hypertrophy following splenic infarction; however, the mechanisms behind CSA or sleep-related hypoventilation are less clear. In adults, CSA may be more prevalent because of chronic opioid use and comorbid congestive heart failure [4]. Sleep architecture and sleep efficiency may be disrupted because of SCD-associated pain, particularly that of VOE. Shapiro et al. found that in children with increased SCD-associated pain, both sleep duration and sleep quality decreased [33]. The reciprocal relationship, in which children with poor sleep quality are at increased risk for pain [34], has also been shown. This is in addition to data that have shown that nocturnal oxyhemoglobin desaturations and nocturnal hypoxemia are prevalent in children with SCD (40%) and are associated with increased VOE [35].

Screening for SDB starts with a comprehensive sleep history and may be further strengthened with validated screening tools, such as the Sleep-Related Breathing Disorder scale, a component of the Pediatric Sleep Questionnaire (PSQ-SRBD) [36]. The gold standard for diagnosis of SDB is polysomnography (PSG); however, ASH does not recommend universal screening for SDB with PSG in asymptomatic children with SCD [34]. Instead, ASH recommends in-laboratory PSG for children who present with snoring, witnessed apneas, obesity, excessive daytime sleepiness, unexplained daytime or nocturnal hypoxemia, and a history of pulmonary hypertension. While traditional risk factors for SDB, especially OSA, may be absent in patients with SCD, the optimal time for screening is unclear and further objective testing is necessary. A 2020 prospective cohort study utilized the I’M SLEEPY sleep apnea questionnaire, previously validated in pediatrics, to screen for SDB in 100 children with SCD [34, 35]. Nineteen patients had a positive screen and were referred for PSG, of whom ten completed a PSG and seven had OSA. The most common responses included snoring, difficulty concentrating, and excessive daytime fatigue. The sensitivity and negative predicative values of this screen were 82 and 85%, respectively.

OSA is the result of recurrent upper airway obstruction and results in fragmented sleep and intermittent hypoxemia [37]. Among the general population in the United States, it is prevalent in 1–5% of children; however, it is thought that OSA may be more prevalent among those with SCD with reported prevalence between 5 and 79% [37]. Feld et al. retrospectively evaluated a cohort of children with both SCD and asthma who underwent PSG and reported an OSA prevalence of 59% that was also associated with both a lower nocturnal oxygen saturation nadir and a lower median daytime oxygen saturation [38]. In the SAC study, Rosen et al. identified an OSA prevalence between 10% (with an obstructive apnea-hypopnea index (OAHI) cutoff of ≥5 events/hr) and 41% (with an OAHI cutoff of ≥1 event/hr) among 243 children (median age 10 years), most of whom had HbSS [37]. There appears to be some variability in OSA classification, with the SAC study reporting predominantly mild OSA compared to a higher percentage of severe OSA reported in other studies [39].

OSA may affect other common SCD comorbidities. Feld et al. looked at the effect of OSA on clinical outcomes in children with SCD and found that, regardless of OSA, there was no difference in reported asthma severity [38]. Katz et al. performed a retrospective review of nearly 650 children with SCD who underwent screening for OSA over an 11-year period and compared SCD-associated complications in those with OSA (n = 136) to a matched-control set of children without OSA [40]. This study described an increase in the rate of SCD complications, notably hospitalizations for both ACS and pneumonia. Neurocognitive complications are described in both OSA and SCD; however, the association between OSA and SCD with respect to neurocognitive complications is less understood [41]. A retrospective cohort study, utilizing the US Representative Kids’ Inpatient Database (KID), found that among 204,000 pediatric hospital discharges for SCD during the study period covering 15 years, < 2% carried a diagnosis of OSA. Those with OSA were more likely to have neurological complications (3.45 vs. 2.17%, p = 0.0014), particularly seizures (2.91 vs. 1.66%, p = 0.0003), and were more likely to have ACS (11.27 vs. 8.85%, p = 0.003) [42].

Since upper airway lymphoid tissue hypertrophy is the most common cause of OSA in children, the American Academy of Pediatrics recommends adenotonsillectomy (tonsillectomy and adenoidectomy (T&A)) as the preferred and first-line treatment for OSA [43]. Complications following T&A among the general population include postoperative hemorrhage and pulmonary edema, velopharyngeal insufficiency, and nerve palsy [44]. Though the benefits of T&A in children without underlying SCD have been studied previously, the extent to which T&A provides improvement in symptoms related to OSA in patients with SCD is less clear. Farrell et al. recently performed a retrospective review of 132 children (≤ 18 years of age) with SCD (the majority of whom have HbSS disease) who underwent T&A for OSA as diagnosed on PSG [45]. This study found a significant improvement in several PSG parameters postoperatively, including a lower mean apnea-hypopnea index (7.6 versus 1.3 events/hour, p = 0.0001), higher mean oxygen nadir (81.2 versus 89.3%, p = 0.0003), and a higher mean oxygen saturation (95.7 versus 97%, p = 0.016). Complications occurred in 11.4% of the population, with the most common complication being postoperative ACS. While the review of this cohort 12 months after T&A showed a significant decrease in the mean number of emergency department visits, there was no significant change in the frequency of VOE or ACS. A similar study by Liguoro et al. looked at postoperative data from a small cohort of children with SCD, 19 patients in total, who underwent T&A and found that though there were improvements in PSG parameters, including mean oxygen nadir as well as a decreased mean annual rate of ACS; there was no change in the incidence of VOE or hospitalization rates [46]. Based on these studies, it seems apparent that T&A is beneficial in patients with SCD with respect to OSA; however, long-term implications with respect to SCD-related morbidity (i.e., ACS, VOE) are less clear. The impact that hydroxyurea may have on OSA in children is less well established. A 2017 chart review focused on two children with OSA, one of whom had concurrent HbSS disease and the other HbSβ° thalassemia, who had shown both improvement in symptoms and resolution in OSA confirmed on PSG approximately 1 year after initiation of hydroxyurea [47]. Whether well-established therapies such as hydroxyurea or newer medical therapies, including crizanlizumab and voxelotor, are viable treatments for patients with concurrent SCD and OSA is yet to be determined.

6. Thromboembolism

Venous thromboembolism (VTE) and pulmonary embolism affect 25% of adults with SCD and carry increased risk for mortality but are not common in children [1]. A retrospective observational study of 1062 patients revealed an incidence of 0.2% primarily in children with central venous lines (CVLs) over a 12-year period [48]. Using the Pediatric Health Information System (PHIS) database with data from 48 participating US institutions from 2009 to 2015, 1.7% of patients developed a VTE. The median age at diagnosis was 15.9 years. CVL placement, chronic renal disease, history of stroke, female sex, length of hospitalization, intensive care unit (ICU) utilization, and older age were significantly associated with VTE [49]. VTE is often overlooked in SCD because limb pain is attributed to VOE or chest pain to ACS. In situ microvascular thrombosis occurs in these patients and large vessel thrombi may not be seen, unless specialized tests are ordered [1, 4].

All three aspects of Virchow’s triad (hypercoagulability, endothelial dysfunction, and hemostasis) create a thrombogenic environment in SCD. Patients with SCD have both traditional and SCD-specific risks for hypercoagulability and thrombus formation [1]. Traditional risk factors include frequent hospitalization and immobilization, use of CVL for red cell exchange or chronic transfusion therapy, need for orthopedic surgery for avascular necrosis of the hip or shoulder, cholecystectomy, splenectomy, and increased pregnancy-related VTE. SCD-related risk factors for VTE include impaired fibrinolysis and upregulation of cellular adhesion molecules. Alteration in sickled red cell structure leads to intravascular hemolysis and externalization of highly procoagulant phosphatidylserine on the red cell membrane. The sickled red cells are more adhesive to the endothelium and the capture of adhesive red cells, leukocytes, and platelets to the endothelial wall triggers vaso-occlusion [1]. SCD patients have dysregulation of factors that initiate and perpetuate hemostasis, as evidenced by decreased levels of natural coagulants, such as protein C, protein S, antithrombin III (ATIII), factors V, VI, IX, and XII. SCD patients have increased circulating levels of antiphospholipid antibodies, and elevated plasma levels of thrombin-antithrombin complexes and prothrombin fragment 1 + 2 (a marker of thrombin and fibrin generation as well as platelet activation). During vaso-occlusive crises, increased tissue factor expression, increased circulating fibrinogen, von Willebrand factor, and clotting factors VII and VIII may predispose patients to VTE. Pulmonary emboli (PE) have also been associated with ACS episodes [4].

A study of 22,631 children with SCD (median age 10.8 years (range: <0.1–20.9)) utilizing the PHIS database from January 2010 to June 2021 revealed a prevalence of hospitalization for PE of 0.3%. The median age was 17.4 years (range: 6.6–20.9 yrs.) at PE diagnosis. Patients with PE had longer hospitalization and more frequent intensive care unit admissions than patients without PE (p < 0.001). Risk factors significantly associated with PE on multivariable analysis included older age, history of CVL, ACS, and exchange transfusion. Mortality was not significantly different between those with and without PE [50].

The d-Dimer test has limited predictive value in diagnosing PE associated with SCD since it may be elevated due to the chronic activation of the coagulation cascade. Current guidelines recommend a compression Doppler for patients suspected of having lower extremity deep venous thromboses (DVTs). Ventilation-perfusion scans may identify PE. Computed tomography pulmonary angiography (CTPA) is currently the test of choice for suspected PE [1, 3].

While anticoagulants are usually prescribed for adults with ACS, children are not routinely given anticoagulants, unless a diagnosis of embolism is made. Anticoagulant prophylaxis in children is reserved for those with additional thrombosis risk, aside from SCD. Antithrombotic therapy for proximal DVT or PE follows 2016 American College of Chest Physician and 2018 ASH guidelines for the treatment of pediatric venous thromboembolism with 3–6 months of anticoagulation [1, 3, 4, 51, 52]. For patients with SCD-PH and VTE, indefinite anticoagulation may be considered in patients without a significant bleeding risk. Duration of therapy may be altered based on clinical status, for example, resolution of VTE on imaging, bleeding risk, or recurrence of VTE [1, 34].

7. Ventricular dysfunction

Sickle cell anemia has a profound impact on the cardiac anatomy and function in children with SCD. Cardiac dilation is well recognized in young children with SCD [53, 54, 55, 56], with dilation leading to eccentric left ventricular hypertrophy. In addition, the left ventricle (LV) can develop both systolic and diastolic dysfunctions in children with sickle cell anemia, with diastolic dysfunction being much more common than systolic dysfunction.

Cardiac dilation occurs in sickle cell anemia due to the state of chronic anemia, increased blood volume, and stroke volume [53, 57]. It has been found that the increased cardiac output in patients with SCD occurs due to this increased stroke volume, with a comparatively minimal increase in heart rate [53]. LV wall stress is directly increased by LV afterload and radius of the LV, and inversely related to wall thickness. As the radius of the LV increases with cardiac dilation, so does the LV wall stress. To compensate for this, the wall thickness increases, mostly to a degree to keep the LV wall thickness normal to mildly increased, which leads to eccentric hypertrophy [58]. Studies on children with SCD have shown that the majority (50–60%) of children have dilation of the left heart, with a smaller percentage (20%) having increased LV wall thickness, with left heart abnormalities correlating with severity of anemia [54, 59].

Systolic dysfunction of the left ventricle is not common in children with SCD. The prevalence of pediatric patients with decreased systolic function indicated by a decreased shortening fraction (SF) and ejection fraction (EF) is very low, ranging from 0 to 5% [54, 55, 59, 60]. In adults, the prevalence of systolic dysfunction is only 9% [61]. SF and EF are LV function measurements highly influenced by preload, afterload, and heart rate; therefore, analyses of patients with SCD with these measures may not represent the heart’s actual contractility. Some more recent studies have analyzed LV contractility and systolic function using load-independent measurements and have found differing results. Some have shown LV contractility to be preserved [56], whereas others found a decrease in contractility [62]. The use of speckle tracking echocardiography can be used to identify abnormalities of LV “twist” in systole, with the normal LV having clockwise rotation at the base and counterclockwise rotation at the apex in systole. In a study performed by Di Maria et al., the basal LV rotation in systole was significantly lower in children with SCD [63]. In adults and adolescents, Braga et al. found that patients with SCD had significantly decreased twist (sum of basal clockwise rotation and apical counterclockwise rotation) in comparison with normal controls [64].

Both invasive and noninvasive measures of LV diastolic function in adults with sickle cell anemia have shown a high prevalence of LV diastolic dysfunction (DD). Cardiac catheterization performed on adults with pulmonary hypertension (PH) shows that ~50% also have DD [65]. Other studies have demonstrated a 65% prevalence of DD in adults with SCD and PH, and 20% in those without PH [66]. Using Doppler echocardiography assessment in adults, Sachdev et al. found that 18% of patients had diastolic dysfunction (defined as mitral inflow E/A < 1, where the E/A ratio denotes the ratio of the early (E) to late (A) ventricular filling velocities) and this group had higher risk of mortality with a risk ratio of 3.5. This study also found a further increased risk of mortality with DD and PH with a risk ratio for death of 12 [61]. There is a growing collection of research on pediatric patients with SCD and the assessment of LV DD. Zilberman et al. demonstrated that echocardiogram (echo) markers of LV stiffness, the E/E’ ratios, were significantly higher in pediatric patients with SCD [67]. In 54% of pediatric patients with SCD, echo findings of elevated LV filling pressures (with septal E/E’ > 8) were identified by Olson et al. [68]. Alsaied et al. studied adults and children (age range 8–43 years, mean 21) with SCD and found a 30% prevalence of DD, which was associated with decreased exercise capacity [69].

Hankins et al. evaluated children with SCD and iron overload; 77% had low E’ (mean mitral annular velocity), and all had elevated E/E’, both measures indicating diastolic dysfunction; there was no correlation with diastolic dysfunction and iron deposition in the heart [70]. Johnson et al. studied children with SCD and found increased E/E’ ratios, and greater diastolic dysfunction was found in those with sleeping and awake oxygen desaturations [71].

More recent research utilizing speckle tracking echocardiography and cardiac magnetic resonance imaging (cMRI) has provided further insights into the cause of this diastolic dysfunction. As noted previously, the use of speckle tracking echocardiography can be used to identify abnormalities of LV “twist” in systole, as well as the “untwisting” in diastole. In a study by Di Maria et al., the peak untwisting rate was significantly lower in children with SCD [63]. Other newer methods of assessing diastolic dysfunction include assessing left atrial (LA) strain with speckle tracking. Jhaveri et al. studied children with SCD and found a significant reduction in LA strain that correlated with the degree in decreased hemoglobin—for every decrease of 1 g/dL of hemoglobin, there was decrease in LA strain by 3.2% [72].

Using cMRI along with echocardiogram findings, Desai et al. compared adults with SCD with normal controls and found that patients with SCD had an increased incidence of myocardial fibrosis, abnormal myocardial perfusion reserve index (both by cMRI), and DD (by echocardiogram), with 29% of patients having DD. The DD did not correlate with cardiac iron overload [73]. Niss et al. studied adults and children with SCD; 29% met echo criteria for DD and 42% had diastolic abnormalities that did not reach the threshold for DD. Patients with DD had a greater prevalence of diffuse myocardial fibrosis on magnetic resonance imaging (MRI) (measured by quantifying the myocardial extracellular volume fraction, ECV). Greater ECV was associated with statistically significantly higher N-terminal pro hormone B-type brain natriuretic peptide (NT-proBNP) levels and correlated with the degree of anemia [74]. Alsaied & Niss et al. demonstrated that left atrial (LA) dysfunction (as measured by LA stiffness), and therefore, LV diastolic dysfunction were related to diffuse myocardial fibrosis (again by MRI ECV), exercise impairment, and increased tricuspid regurgitation velocities (TRVs) [69, 74].

8. Pulmonary hypertension

Pulmonary hypertension (PH) is a leading cause of morbidity and mortality in adults with SCD [75]. PH is defined as a resting mean pulmonary artery pressure (PAP) over 20 mm Hg measured by right heart catheterization (RHC). Hemodynamically, PH is subdivided as precapillary PH if the pulmonary capillary wedge pressure (PCWP) or left ventricular end-diastolic pressure (LVEDP) at right heart catheterization is ≤15 mm Hg and as postcapillary PH if the PCWP or LVEDP is >15 mm Hg. About half of SCD-related PH patients have precapillary PH with potential etiologies of (1) a nitric oxide (NO) deficiency state and vasculopathy, consequent to intravascular hemolysis, (2) chronic pulmonary thromboembolism, or (3) upregulated hypoxic responses secondary to anemia, low oxygen (O₂) saturation, and microvascular obstruction [53, 76]. The remainder have postcapillary PH secondary to left ventricular dysfunction, which is common in patients with SCD, possibly related to ventricular dilation and concentric hypertrophy of the myocardium as a response to chronic anemia and relative systemic hypertension [75].

The initial test of choice to screen for PH is transthoracic echocardiography (TTE). In general, the tricuspid regurgitation velocity (TRV) measured during TTE with estimated right atrial pressure is considered a valid estimate of systolic PAP. Other signs of PH on TTE include right ventricular hypertrophy and/or dilation, systolic flattening of the interventricular septum, right atrial dilation, and increasing tricuspid regurgitation (TR). Once PH is diagnosed, further evaluation is needed to identify contributors to PH that might require focused treatment, such as hypoxemia due to sleep-disordered breathing including obstructive sleep apnea (OSA), venous thromboembolism, or portal hypertension. Because of the known association of OSA with PH in non-SCD populations, the clinical guidelines for diagnosis of PH in SCD recommend a formal sleep study for all SCD patients with an elevated TRV [77]. In addition, the evaluation for other contributors to PH typically includes laboratory testing (liver function tests, antinuclear antibody), pulmonary function testing, radionuclide ventilation-perfusion scan, and CTPA.

N-terminal-pro-hormone B-type brain natriuretic peptide (NT-proBNP) has been evaluated as potential screening tool for PH in the pediatric SCD population [78, 79]. In adults with SCD, the degree of NT-proBNP elevation correlated with mortality [80]. The cutoff level for NT-proBNP is not defined in children, although guidelines suggest NT-proBNP >160 pg./mL to be indicative of an elevated PAP. A retrospective study of children with SCD found a higher median NT-proBNP of 70 pg./mL in 8–14-year-olds than age-matched controls. NT-proBNP levels were associated with markers of hemolysis, that is, reticulocyte count (r = .25, p = 0.01) and LDH r = .47, p.001). A positive correlation was found between NT-proBNP and diastolic left ventricular size (r = 0.28, p = 0.047) [81]. Hence, NT-proBNP and TTE as noninvasive tools may play a role in identifying PH, prior to RHC. Of note, NT-proBNP measurements may be misleading in patients with renal insufficiency or left heart failure [77].

The reported prevalence of PH in the pediatric age group with SCD varies, with studies including older patient populations reporting a higher prevalence suggesting a disease progression from childhood into adulthood secondary to chronic hemolysis and systemic vasculopathy [82]. An analysis of pediatric studies combining almost 1200 children revealed a prevalence of elevated TRV (>2.5 m/s) in 25%, while moderate to severely elevated TRV (>3.0 m/s) was noted in 4% of children [83]. Children with more severe hemolysis, anemia, and acute chest episodes are at a higher risk for developing PH.

Universal screening for PH is not recommended by the NHLBI and ASH for asymptomatic patients with SCD. A screening for TTE is recommended in children with symptoms suggestive of PH (e.g., exercise intolerance, fatigue, peripheral edema, and chest pain) [84]. The American Thoracic Society (ATS) suggests a one-time TTE in asymptomatic children with SCD who are aged 8 to 18 years (sooner in those with severe hemolytic anemia) [85]. If the TRV is elevated (>2.5 m/s), the patient may require additional evaluation starting with an RHC. Once individuals with SCD reach adulthood (18 years of age), TTE is recommended every 1 to 3 years, using the shorter intervals for those with respiratory symptoms, TRV ≥2.5 m/sec on prior echocardiogram, greater frequency of pain episodes, prior thromboembolic events, or greater severity of hemolytic anemia [77].

A general approach to the management of SCD-related PH involves supportive care, use of SCD-specific therapies, and consideration of specific agents for PH. Examples of supportive care and treatment of comorbidities include oxygen therapy for those with low oxygen saturation, treatment of left ventricular failure in those with postcapillary PH, and anticoagulation for those with thromboembolism. SCD-specific treatments, such as hydroxyurea or chronic transfusion therapy, may be of benefit by raising the hemoglobin concentration, reducing hemolysis, and preventing vaso-occlusive events that cause additional increases in PAP. A recent study by Dhar et al. found that hydroxyurea therapy was associated with a decrease in TRV in children with SCD [55]. Another prospective study of 204 children with SCD found that hydroxyurea use was associated with an estimated 5% decline in TRV after 2 years [86]. A carefully selected subgroup of symptomatic patients with RHC-confirmed elevation in pulmonary vascular resistance (PVR) and normal PCWP should be referred to an expert in PH for close follow-up and consideration whether PH-targeted therapy should be attempted. Specific therapies for the treatment of PH include prostacyclin derivatives, endothelin receptor antagonists, and phosphodiesterase-5 inhibitors [87]. Bosentan, an endothelin receptor blocker, appeared to be well tolerated in a randomized, double-blind, placebo-controlled trial of patients with SCD and PH, although the small sample size precluded an analysis of its efficacy [88]. Another randomized, double-blind, placebo-controlled study designed to evaluate the safety and efficacy of sildenafil was prematurely halted after interim analysis showed that sildenafil-treated patients were likely to have more acute sickle cell pain crises (35%) compared with placebo-treated patients (14%) [89]. Therefore, response to PH-targeted therapies appears to be different in this patient population.

In adults with SCD, PH is associated with increased mortality [90]. While PH is not associated with a higher mortality in children with SCD, these children may be at increased risk of cardiopulmonary decline [91]. A comprehensive assessment for symptoms suggestive of PH should be performed at each visit, and if suspected, evaluation should begin with a TTE. Currently, there is not enough evidence to suggest that early identification and treatment of children with SCD and elevated TRV improve survival. Treatment and follow-up of patients with confirmed PH should be performed at a center with experience in the treatment of pediatric PH, and this should occur in close conjunction with the hematology team.

8.1 Therapy

Treatment options for sickle cell disease include disease modifying treatments such as hydroxyurea, voxelotor, crizanlizumab, and red blood cell transfusions as well as curative or transformative therapies such as bone marrow transplantation and newly approved gene therapies.

Blood transfusions are lifesaving for individuals living with SCD. There are several instances where transfusions may be administered. They can be used to treat or prevent acute and chronic complications. Acute complications for which transfusions are recommended include splenic sequestration crisis, stroke, acute chest syndrome, aplastic crisis, or multiorgan failure [92]. Transfusions are not routinely recommended for vaso-occlusive pain episodes, priapism, or avascular necrosis. Chronic transfusion therapy is usually reserved for primary or secondary prevention of stroke. Rarely, they are used in patients who continue to have recurrent episodes of vaso-occlusive pain and acute chest syndrome, despite being on hydroxyurea. It is important to consult with a sickle cell specialist if a blood transfusion is being considered as the volume to be given will need careful thought. Transfusion therapy may lead to immediate and/or delayed side effects including transfusion reactions—acute and delayed; increased viscosity, which may increase stroke risk; concern for fluid overload; iron overload; and alloimmunization. It is important to counsel families about these side effects so care may be sought immediately.

Hydroxyurea is the oldest Food and Drug Administration (FDA)-approved disease modifying treatment for individuals living with SCD. One of the main benefits of hydroxyurea is an increase in fetal hemoglobin, which prevents sickling. Additional benefits include improved RBC rheology and survival. The effects of hydroxyurea on SCD have been studied for several decades. It reduces and prevents vaso-occlusive episodes, including pain and acute chest syndrome, transfusion requirements, and inpatient admissions. It has been shown to decrease morbidity and mortality [93]. It can prevent strokes in individuals with abnormal transcranial Doppler velocities, who do not have severe vasculopathy [94]. It is safe to administer in young children [95]. Hydroxyurea may be started as early as 9 months of age in individuals with sickle cell anemia (HBSS, HbSβ° thalassemia) to prevent any complications [84]. Current studies about hydroxyurea are looking at personalized dosing options [96]. Hydroxyurea is very well tolerated; myelosuppression is a potential side effect that requires monitoring.

Voxelotor is a novel oral disease modifying treatment approved by the FDA for adults and pediatric patients 4 years of age and older with sickle cell disease. It decreases the polymerization of HbS, which is responsible for much of the pathophysiology in SCD. Voxelotor binds reversibly to the alpha-globin chain of Hb and increases its affinity for oxygen. It has been shown to increase baseline hemoglobin by >1 g/dL [97]. However, there are no current data thus far demonstrating reduction in the frequency of pain episodes [98]. It is generally well tolerated. The most common adverse reactions include headache and abdominal pain. They can interfere with measurement of hemoglobin subtypes via liquid chromatography, thus affecting laboratory interpretation of hemoglobin electrophoresis.

Crizanlizumab is an FDA-approved novel injectable therapeutic agent that is approved for adults and pediatric patients with SCD, regardless of genotype, aged 16 years and older. In vaso-occlusive episodes, there is significant upregulation of P-selectin on the endothelial cells and platelets. These cell-to-cell adhesions are responsible for the pathogenesis of sickle-related pain crisis in individuals with SCD [99]. Crizanlizumab is a humanized monoclonal antibody that binds to P-selectin and inhibits the interaction between P-selectin and glycoprotein ligand 1, thereby decreasing adhesion [99]. In a phase 2 randomized, placebo-controlled trial, high-dose crizanlizumab resulted in 35% reduction in pain-related episodes in participants taking hydroxyurea and 50% reduction in those not on hydroxyurea [99]. It is administered intravenously, starting every 2 weeks and then as monthly infusions. It is generally well tolerated and adverse effects include infusion reactions, arthralgia, diarrhea, pruritus, vomiting, and chest pain.

Hematopoietic stem cell transplantation (HSCT) has been utilized as a curative treatment for SCD for the past four decades. An analysis of 996 patients who underwent HSCT between 2008 and 2017 shows that best outcomes are achieved when the donor is a human leukocyte antigen (HLA)-match sibling and the HSCT is performed before 13 years of age [100]. All patients with stroke, abnormal transcranial Doppler velocities, frequent acute painful events, and recurrent acute chest syndrome should be offered HSCT as a treatment option [101]. The risks related to HSCT are graft rejection, graft-versus-host disease, infections, infertility, and death.

Many people living with sickle cell disease do not have HLA-match donor. Therefore, gene therapy using patient’s own hematopoietic stem cells is an attractive alternative. There are three main approaches to gene therapy in sickle cell disease: (1) gene addition, (2) gene editing, and (3) gene correction. In LentiGlobin trials, patients’ stem cells were transduced with a lentiviral vector encoding a modified β-globin gene, which resulted in the production of an antisickling hemoglobin, HbA^T87Q98. Thirty-five patients received treatment and engrafted. All of the 25 evaluable patients had complete resolution of severe vaso-occlusive events. Hemoglobin level increased from 8.5 to 11 gm/dl [101]. In another approach, CRISPR-Cas9 (Clustered regularly interspaced short palindromic repeats -CRISPR-associated protein 9) technology was used to target BCL11A (transcription factor B-cell lymphoma/leukemia 11A), a transcription factor that represses γ-globin expression and fetal hemoglobin [102]. One patient with transfusion-dependent thalassemia and one patient with sickle cell disease underwent this treatment, which resulted in pancellular distribution of HbF in red blood cells and led to transfusion independence in the thalassemia patient and resolution of pain crises in the sickle cell patient. The Food and Drug Administration (FDA) has approved both these treatments in December 2023 for patients with SCD ≥12 years with recurrent vaso-occlusive events.

9. Conclusion

Sickle cell disease (SCD) is characterized by hemolysis, increased endothelial adhesion, inflammation, and vasculopathy. Various cardiopulmonary complications may occur in children including ACS, pulmonary function abnormalities, asthma, sleep disordered breathing, ventricular dysfunction, thromboembolism, and pulmonary hypertension that may lead to more severe outcomes in adulthood. Repeated episodes of ACS may be associated with increased inflammation and decreased longitudinal lung function in children. Increased airway hyperresponsiveness, leukotriene levels, and airway inflammation are seen in SCD patients. Children with SCD are at increased risk for asthma, sleep-disordered breathing, and obstructive sleep apnea, which have been associated with an increase in the rate of SCD complications such as VOE or ACS. Diastolic dysfunction associated with chronic anemia in SCD patients has been associated with decreased exercise capacity. Evidence of pulmonary hypertension in children with SCD suggests a disease progression from childhood into adulthood, secondary to chronic hemolysis and systemic vasculopathy. While most guidelines do not recommend routine screening or testing for complications in children, patients with symptoms, such as increased incidence of VOE or ACS, wheezing, shortness of breath, sleep-disordered breathing, or hypoxemia, should be screened and followed by specialists preferably in a multidisciplinary setting. Disease modifying treatments such as hydroxyurea may attenuate some of the cardiopulmonary complications in SCD. More studies need to be done to assess the long-term effects of newer disease modifying and transformative therapies.

Conflict of interest

Dr. Banu Aygun receives research funding from Pfizer and Bluebird Bio. She has participated in advisory board meetings for Global Blood Therapeutics, Agios, Pfizer, and Bluebird Bio. Dr. Elizabeth Fiorino has served as a consultant for Boehringer Ingelheim, Simumetrix, and the France Foundation.