Published ahead of print on September 5, 2008, doi:10.1164/rccm.200708-1219OC
© 2008 American Thoracic Society doi: 10.1164/rccm.200708-1219OC
Longitudinal Decline in Lung Volume in a Population of Children with Sickle Cell Disease1 Division of Respiratory Medicine, 2 Child Health Evaluative Sciences, Research Institute, 3 Division of Hematology/Oncology, Department of Pediatrics, the Hospital for Sick Children and the University of Toronto, Toronto, Ontario, Canada Correspondence and requests for reprint should be addressed to Joanna E. MacLean, M.D., F.A.A.P., F.R.C.P (C), Department of Respiratory Medicine, Children's Hospital at Westmead, Locked Bag 4001, Westmead, New South Wales 2145, Australia. E-mail: joannam4{at}med.usyd.edu.au
Rationale: Sickle cell disease (SCD) results in significant morbidity and mortality attributable to pulmonary complications. The pattern of lung function change across childhood in SCD is not well delineated. Objectives: To determine if the pattern of lung function in SCD differs from race-matched, predicted values across childhood, to describe that pattern of change, and to examine the effect of clinical covariates on lung function. Methods: Lung function measurements for children with SCD, aged 8–18 years, from a single center were examined for inclusion. Mixed-model analysis was used to retrospectively review lung function in these children in comparison with those predicted by race-matched reference equations. The contribution of age, sex, Hb level, and β-globin genotype on longitudinal changes in lung function was examined. Measurements and Main Results: Children with SCD show significant decline in spirometric lung volumes across childhood that are concordant with the pattern of change in other measures of lung volume. The average decline for FEV1 and total lung capacity is 2.93 and 2.15% predicted/year for males and 2.95 and 2.43% predicted/year for females. β-Globin genotypes known to be associated with more severe disease showed a faster decline in lung function, whereas sex showed an inconsistent effect on lung function. Conclusions: Lung volumes in children with SCD decline with age. The pattern of decline begins in childhood, and supports a predominately restrictive defect.
Key Words: spirometry • lung volume • genotype • race matched
Sickle cell disease (SCD) is an autosomal recessive disorder that causes structural changes of Hb, leading to oxygen release, increased red cell fragility, hemolysis, and microvascular obstruction with secondary anemia and tissue ischemia. It is a multisystem chronic disease that results in a decreased life span, with a significant increase in mortality beginning in the third decade of life (1, 2). Mortality attributable to pulmonary disease is significant, with estimates ranging from 21 to 85% (2–4). Pulmonary disease in SCD has both acute and chronic components. Acute chest syndrome and pneumonic processes characterize acute lung involvement. Chronic lung disease is characterized by both parenchymal and vascular abnormalities, including abnormal lung function measurements, chronic hypoxia, pulmonary hypertension, diffuse interstitial fibrosis, and cor pulmonale with right ventricular hypertension (3, 5–11). Change in lung function measurements may be the earliest marker of sickle cell lung disease (3). Lung function studies in children with SCD to date have not identified a consistent pattern of change. Cross-sectional studies have shown a variety of patterns, including normal pulmonary function (6), a restrictive pattern (12–16), an obstructive pattern (17–23), and a mixed pattern (24–26). Little is known about the longitudinal change in lung function in children with SCD, as only two longitudinal studies have been published to date (3, 26). The first was restricted to patients already identified as having lung disease secondary to SCD. The second included measurements at only two specific ages rather than across childhood. This study describes the longitudinal changes in lung function in an unselected population of children with SCD across childhood using the results of retrospective annual lung function testing data. We hypothesized that lung function in children with SCD will decline at a greater rate than race-matched control data, and that factors that predict more severe SCD will also predict a faster decline in lung function. Some of the results of this study have been previously published in abstract form (27).
Subjects All children with SCD and lung function measurements were eligible for inclusion in this study. In order to obtain an unselected population of children with SCD, pulmonary function data obtained before 1989 were excluded to coincide with the implementation of routine annual lung function testing. All patients older than 7 years attending "well visits" to the sickle cell clinic underwent lung function testing on an annual basis as part of their routine clinical care. In order to be included in the analysis, children must have had at least two repeated measures available for spirometry or lung volume across the study period. The study was approved by the Research Ethics Board at the Hospital for Sick Children, Toronto, Ontario, Canada.
Lung Function Testing
Predicted Values
Covariates
Statistical Analysis Mixed-model analysis was used as an extension of the usual multiple regression analysis using SAS version 9.1. This type of analysis has previously been used to describe the change in lung function of children with cystic fibrosis (35). Additional information on mixed model analysis is provided in the online supplement.
Lung function was measured in 413 children with SCD between the ages of 8 and 18 years from January 1989 to January 2005. Spirometry results from 312 children and lung volume results from 294 children were included in this analysis. A total of 90% of children were of African descent, 4% were of Caucasian descent, with race not recorded or unknown for the remainder of children. Table 1 shows descriptive statistics of the sample and record distribution.
The pattern of change in standing height was examined relative to race-matched predicted values. Males in the sample showed fall off in z-score for height (decline in z-score/yr = 0.063; P < 0.0001), whereas females did not show a significant difference in z-score for height across age (decline in z-score/yr = 0.012; P = 0.34). The longitudinal analysis of pulmonary function in children with SCD demonstrated significant differences from predicted values for FEV1, FVC, and FEF25–75 (Table 2). The FEV1/FVC ratio did not differ from predicted values. The rate of decline in TLC showed a similar pattern to that of FEV1, FVC, and FEF25–75 (Figures E2 and E3). At 8 years of age, 96.5% of children had normal lung function, 0.9% showed an obstructive pattern, and 2.6% showed a restrictive pattern. At 17 years of age, 81.3% of children had normal lung function, none showed obstruction, and 18.7% showed a restrictive pattern. The mean RV/TLC ratio at 8 years of age was 0.30, and declined to 0.21 at 17 years of age. Limiting the analysis to those children with at least three measures of pulmonary function did not change the results of the longitudinal analysis for any of the pulmonary function variables. Results of lung diffusion capacity corrected for Hb did not show significant changes across age (see the online supplement).
In order to examine the influence of covariates on lung function, Hb level at the time of lung function testing, sex, and HbB genotype were added to the longitudinal model. The effect of age was consistent across all pulmonary function parameters (P < 0.0001). Hb level at the time of lung function testing did not show a significant effect, except for FEF25–75, which decreased with lower values of Hb (β estimate [± SE] = 0.0031 ± 0.0015; P = 0.05). The influence of sex was inconsistent, with a significant effect seen only for FVC, for which males showed a faster decline than females (β estimate [± SE] of age x sex for FVC = –0.017 ± 0.0075; P = 0.05). For each 1 year increase in age; FVC declined by 88 ml for males and 79 ml for females; FEV1 declined by 89 ml for males and 89 ml for females; FEF25–75 declined by 133 ml for males and 145 ml for females; and TLC declined by 151 ml for males and 135 ml for females. Next to the influence of age, HbB genotype had the strongest influence on the rate of change of pulmonary function (Figure 1). Hb SS, compared with Hb SC, shows a greater rate of decline for FVC, FEV1, FEF25–75, and TLC (Table 2). HbB genotype Hb SS was associated with a higher likelihood of ever showing a restrictive (Hb SS vs. Hb SC, 26.5 vs. 7.7%, respectively; likelihood ratio, 9.19; P = 0.002), but not an obstructive (Hb SS vs. Hb SC, 4.5% vs. 1.9%; likelihood ratio, 0.79; P = NS) pattern of lung function.
This is the largest reported longitudinal analysis of lung function in a population of children with SCD. The results of this study show that children with SCD have a progressive decline in pulmonary function across childhood. The rate of decline in lung function is similar to that of children with cystic fibrosis (2–3%/yr) (35, 36), and significantly greater than seen in children with asthma (0–0.88%/yr) (37, 38). This study also highlights the need for race match prediction equations in clinical practice, as estimate corrections for race may fail to identify patterns of change in pulmonary function over time. The limitations of this study include the use of reference equations to replace control data and the inclusion of only a small number of clinical predictors. The use of reference equations to replace control data was necessitated by the retrospective design. Studies of lung volume measurements across ethnicity support differences in predicted volumes because of variations in body proportions (39). The equations that were chosen for spirometry are derived from data collected as part of a national survey completed in the United States. African Americans and African Canadians may differ in their genetic background, and, therefore, these reference equations may not specifically reflect predicted pulmonary function in the subject group for this study. However, this difference should not affect the pattern of change across age. Non–race-matched reference equations were used for comparison for lung volume parameters, as no validated race-matched reference equations were available at the commencement of this study. Although the use of these equations is unlikely to affect the pattern or direction of change, the estimates of the amount of change may be inaccurate. The clinical predictors chosen for this analysis were selected because each has been linked to outcome, and each was collected as part of clinical care. Other factors that may influence pulmonary function, such as Hb F, history of vasoocclusive crises, or acute chest syndrome, were either not collected or could not be ascertained reliably retrospectively from the clinical records. Although there is no specific treatment for SCD lung disease, treatment changes over the course of the study period may affect lung function over time. However, any treatments aimed at ameliorating SCD would be expected to have a positive effect on lung function, and, thus, our study would be an underestimate of the deterioration in lung function due to SCD alone. The pattern of decline in both spirometric and lung volume parameters supports a restrictive pattern of disease with an increase in the degree of impairment over time. Although this pattern is consistent with recent cross-sectional data (10, 13), cross-sectional studies or those including only specific groups of children with SCD may not capture this pattern of change across age. Whereas only 18% of children showed abnormal lung function at 17 years of age, the predominant change from 8 years of age was an increase in the number of children showing a restrictive pattern. The present study results are in agreement with those of Sylvester and colleagues (13), who concluded that restrictive abnormalities become more prominent with age. The pattern of higher RV/TLC ratio at 8 years of age with a decline over time is supportive of early injury or inflammation resulting in progressive changes in lung volumes across age. Cross-sectional studies have reported significant airway hyperreactivity and asthma in children with SCD (17, 19, 20, 22, 40). Using FEV1/FVC alone, the present study did not find a high proportion of children with airway obstruction, which is consistent with previous cross-sectional studies (17, 19). Asthma, however, is a clinical diagnosis supported by bronchodilator reversibility and airway hyperresponsiveness. The present study did not include challenge testing, which may be more sensitive to detecting airway responsiveness. Although the present study did demonstrate a decline in FEF25–75 across childhood, suggesting peripheral airflow limitation, previous reports of airway obstruction, increased airway responsiveness, and asthma in SCD may overestimate this association because of small sample sizes, select subject groups, nonstandard definitions of airway responsiveness, and the use of questionnaire data for the identification of asthma. Future studies of children with SCD should include measures of airway responsiveness, such as exercise challenge or bronchodilator reversibility, in addition to lung function in larger populations to further define the airway pathology in SCD. Caution, however, should be used with this type of testing in patients with SCD, because of the risk of associated hypoxemia precipitating vaso-occlusive crises. HbB genotype was the strongest predictor of decline in pulmonary function next to age. Those with more severe genotypes, such as Hb SS, showed a faster decline in lung function than those with genotype Hb SC. Hb SS genotype does confer an increased risk of severe disease (7), and has previously been associated with greater impairments in pulmonary function (12, 18, 24). The association of Hb SS with an increased rate of decline in pulmonary function may simply reflect an association with more severe disease. Alternatively, this genotype may confer a greater risk of lung disease, or may correlate with an altered ability for repair after lung injury. Contrary to mortality data showing higher female mortality in early adulthood (2), the only significant sex effect demonstrated in this analysis showed males to have worse outcome than females. Males, but not females, showed a progressive decline in height across childhood. Children with SCD are at risk of multiple bony crises that will affect growth, not only of the long bones, but also the spine and thoracic cage. Singhal and colleagues (41) described a delayed pattern of growth in a Jamaican cohort of SCD with differences based on genotype, but no effect of sex, whereas Phebus and colleagues (42) did find a greater impairment of height in males compared with females. This effect on height is seen at an early age, but appears to have a proportional effect on standing and sitting height (13, 43). This suggests that any effect of height should also influence thoracic volume and, hence, lung volume. Although the current study did show a sex effect on height, sex did not influence the decline in TLC. Further study is needed to determine the effect of sex on pulmonary function in SCD. This epidemiological study cannot relate pulmonary function changes in SCD to a specific type of lung injury or disease. Radiologic and autopsy data confirm that pulmonary parenchymal injury is present (8, 9), but do not explain the mechanisms accounting for this injury. The factors that account for lung disease seen in association with SCD undoubtedly include those that contribute to the occurrence of acute injury, such as those associated with acute coronary syndrome, in addition to altered response to this injury. The mechanism may include a relative nitric oxide deficiency and/or excess arginase activity also proposed to be involved in asthma (44), cystic fibrosis lung disease (45), and the development of pulmonary hypertension in SCD (46). In conclusion, this study demonstrates that SCD is associated with a progressive decline in lung function across age, and demonstrates a restrictive pattern of impairment. This decline begins in childhood, but further research, including that on younger children, is needed to determine early markers of lung damage or disease in children with SCD. Acute as well as chronic lung injury likely accounts for this decline, but the mechanisms underlying this injury are poorly understood. Prospective studies that include measures of lung function and lung growth, as well as clinical and biochemical markers, are needed to better understand and manage lung disease associated with SCD.
The authors thank Karl Zdravko Lukic for his assistance with extracting the lung function data.
This article has an online supplement, which is accessible from the issue's table of contents at www.atsjournals.org Originally Published in Press as DOI: 10.1164/rccm.200708-1219OC on September 5, 2008 Conflict of Interest Statement: None of the authors has a financial relationship with a commercial entity that has an interest in the subject of this manuscript. Received in original form August 17, 2007; accepted in final form September 5, 2008
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
AT A GLANCE COMMENTARY
