Spirometry Interpretation in Children With Sickle Cell Disease: <scp>GLI</scp> Global Versus Race‐Specific Normative Equations

The American Thoracic Society (ATS) and European Respiratory Society (ERS) have recently recommended the use of race-neutral normative equations for the interpretation of spirometry testing, the Global Lung Function Initiative (GLI) Global equations [1]. Normative equations are imperative for spirometry interpretation in children, where age, height, and sex have a significant impact on what may be considered normal ranges. Previously, race-specific equations were used to account for differences in population averages for spirometry volumes based on race. A change in spirometry normative equations to race-neutral standardizes these differences across the entire population. This may have a significant impact on interpretation of pulmonary function testing (PFT) and subsequent patient management, particularly in at-risk populations such as patients with sickle cell disease (SCD), who are typically of African or Caribbean descent. The impact of GLI Global equations will differ across populations and racial groups. In a cohort of healthy controls and patients with previous bronchiolitis, Doumat et al. determined that in White children, the prevalence of obstruction increased by 21%; however, in Black children, the prevalence of possible restriction increased by 222% [2]. Compared with White, Black race was associated with an increased prevalence of impairments and increased severity. Forno et al. also found that when changing to GLI Global equations, Black children had approximately threefold higher odds of changing spirometry pattern compared with White children [3]. Race-neutral equations may therefore have a significant impact on patients with SCD. The American Society of Hematology recommends performing PFTs in symptomatic children with SCD [4]. With an extensive list of indications for testing, including common asthma-like symptoms such as wheeze and exertional dyspnea, as well as routine follow-up after acute chest syndrome episodes, PFTs have become a commonplace respiratory assessment in patients with SCD. Clinicians must therefore understand the potential implications for interpretation of results that may occur with race-neutral equations. We aimed to determine the impact of GLI Global versus race-specific equations on spirometric parameters and classification in children with SCD. This was a retrospective cohort study of pediatric patients with SCD who underwent spirometry testing at the Hospital for Sick Children in Toronto, from January 1, 2021, to December 31, 2024. The study received institutional Research Ethics Board (#3469) approval. Pediatric patients aged less than 18 years, with spirometry testing meeting ATS/ERS acceptability and reproducibility criteria, were included [5]. The first test performed by each patient during the study period was included. Clinical data were collected via chart review and included SCD genotype, clinical symptoms, prescribed medications, and hemoglobin level at the time of spirometry (±1 month). Spirometry data was interpreted using GLI Global and GLI race-specific equations [6, 7], based on lower limits of normal (z-score −1.645) [5]. Analysis was completed in two subgroups based on hemoglobin phenotype, that is, patients with SCD hemoglobin SS (HbSS), and those with hemoglobin SC (HbSC), given the significant differences in disease phenotypes. Demographic and clinical data are presented as medians (interquartile ranges) and means (standard deviations). Bivariate statistics were completed using paired t tests for comparisons between forced expiratory volume in 1 s (FEV1), forced vital capacity (FVC) and FEV1/FVC z-scores for each normative equation, or χ2 tests for categorical variables. Univariate logistic regression was used to assess the association between hemoglobin level and experiencing a change in interpretation of spirometry results. A total of 375 spirometry tests were completed during the study period for an indication of SCD. After the exclusion of 30 tests for an age of 18 or older and 78 tests for poor quality based on ATS/ERS criteria, 267 tests remained from 207 patients, of which the first from each patient was included. Mean age of the 207 (52% female) patients was 13.43 years (SD 2.51, range 7.3–17.9). Of these, 160 had a genotype of HbSS, and 47 HbSC (Table 1). In patients with SCD HbSS, hydroxyurea was prescribed to 120 (75%), and 1 (0.6%) was prescribed inhaled corticosteroids for asthma. Mean hemoglobin level at the time of spirometry was 9.16 g/dL (SD 13.5). Most patients had no reported respiratory symptoms at the time of spirometry (129, 92%). Switching to GLI Global normative equations resulted in lower mean FEV1 (p < 0.0001) and FVC (p < 0.0001) z-scores compared with race-specific equations. GLI Global resulted in a change in interpretation for 46 tests (29%), fewer spirometry tests classified as normal (65% vs. 45%, p = 0.0005), and more tests suggestive of restriction (18% vs. 36%, p = 0.0007) (Table 2). Figure 1 shows the changes in classification when moving from GLI race-specific to GLI Global equations. Most spirometry tests that changed from normal to abnormal switched to a restrictive (22 [66%]) or nonspecific (11 [33%]) pattern. In patients with a change from normal, three (9%) had reported respiratory symptoms. For patients with HbSS, each 1 g/dL increase in blood hemoglobin level at the time of spirometry was associated with 25% lower odds of experiencing a change in interpretation with GLI Global equations (OR 0.75, 95% CI 0.56–0.98, p = 0.039). Patients with HbSC showed similar findings. GLI Global equations resulted in lower mean FEV1 (p < 0.0001) and FVC z-scores (p < 0.0001) and in a change in interpretation in 13 tests (28%). There was a non-significant reduction in normal tests (88% vs. 68%, p = 0.12) and an increase in tests suggestive of restriction (2% vs. 23%, p = 0.019). In this study, we explored the impact of changing from GLI race-specific equations to GLI Global equations on spirometry interpretation in children with SCD. We found that GLI Global resulted in lower FEV1 and FVC z-scores compared with race-specific equations in both HbSS and HbSC. Additionally, GLI Global resulted in a higher proportion of tests suggestive of restriction and a lower proportion of normal test results. Most patients with HbSS who changed from normal to abnormal did not have respiratory symptoms at the time of spirometry. The impact of race and ethnicity on defining normative values is of particular importance for people with SCD, given the high prevalence of African and Caribbean descent and multiorgan involvement. For instance, in the assessment of kidney function, race adjustment has historically been used to estimate higher glomerular filtration rates for Black individuals to account for an apparent higher muscle mass [8]. New estimating equations without the race variable have been created; however, their clinical impact remains unclear. In Black Europeans and Black Africans, the performance continues to remain suboptimal compared with measured values, highlighting the importance of research in the area of race adjustment [9]. Our study aimed to investigate the impact of race-neutral equations in the context of PFT by spirometry. PFT has been shown to be an important measure of disease activity and response to therapy in SCD. McLaren et al. demonstrated that hydroxyurea treatment was associated with an improvement in annual lung function decline in children with SCD, suggesting that PFTs are a useful measure of end-organ response to treatment [10]. In both adults and children, non-myeloablative hematopoietic stem cell transplant has also been associated with a stabilization in PFTs over time, preventing the expected progressive decline [11, 12]. GLI Global normative equations result in changes in z-score values, which will lead to a significant impact on spirometry interpretation. A recent study by Ghazala et al. in a population of 100 children with SCD showed that GLI Global equations resulted in lower average FEV1 and FVC z-scores compared with race-specific equations [13]. We found similar decreases in FEV1 and FVC z-scores in our population. However, we also showed that these decreases resulted in changes in final spirometry interpretation, which is clinically relevant for health care providers, including hematologists responsible for ordering these tests. A decrease in values of FEV1 and FVC z-scores will reduce the number of tests interpreted as normal, as shown in our study with a 20% reduction in normal tests, and an increase in those suggestive of restriction. In our study, we found that a higher hemoglobin level was associated with a lower odds of a change in spirometry interpretation. Previous studies have shown a positive association between hemoglobin level and FEV1 and FVC [14]. It is likely that our finding is related to patients with a lower hemoglobin level being closer to the lower limits of normal, and therefore more likely to change from normal to abnormal with alternative equations. It is imperative that clinicians understand these implications and are keenly aware of their local pulmonary function lab practices. When defining normal values using race-neutral equations, despite apparent changes in z-score values and spirometry interpretation, the underlying physiologic characteristics of patients will not change. Although increased abnormal findings may occur, it is unclear if changes to patient management will lead to improved outcomes. Additionally, as race-neutral equations are not currently available for lung volumes and diffusion capacity, providers should be aware that differences between tests will exist. Further research in this area is crucial. In our study, only 9% of children who changed from normal to abnormal with GLI Global had reported respiratory symptoms at the time of testing. Therefore, treatment decisions solely based on spirometry interpretation may result in over-treatment. Until further research is completed to explore the impact of GLI Global equations on long-term respiratory outcomes, clinicians will need to balance the risks of over-investigation and over-treatment against potential benefits. Spirometry testing, and any subsequent abnormal findings, should be used in combination with clinical histories, physical examinations, and other testing to determine if patient management should change. For clinicians involved in the care of children with SCD, if questions as to the significance of spirometry findings remain, a referral to pulmonology may be warranted to determine if interventions beyond close monitoring are needed. Jacob McCoy: study conceptualization and development, data collection, manuscript creation and submission. David Wilson: study conceptualization and development, manuscript input and review. Melanie Kirby-Allen: study conceptualization and development, manuscript input and review. Isaac Odame: study conceptualization and development, manuscript input and review. Hartmut Grasemann: study conceptualization and development, manuscript input and review, supervision of project. The study received approval from our institution's Research Ethics Board (#3469). The authors declare no conflicts of interest. Data available on request from the authors.