Iron overload is prevalent, likely underdiagnosed and undertreated in sickle cell disease: A real-world report from the ASH research collaborative Free

Introduction

Despite the utility of red blood cell transfusion therapy for managing the acute and chronic complications in sickle cell disease (SCD), frequent transfusions inevitably result in iron overload. Iron overload is independently associated with mortality in SCD, and guidelines for SCD management recommend annual screening for the at-risk subsets, such as those on monthly transfusions. Our prior work has demonstrated persistent gaps in the diagnosis and management of iron overload in SCD; however, we were unable to examine acute care utilization in that work.

In this study, we determined the prevalence, screening, diagnosis and management of iron overload in SCD using the ASH Research Collaborative (RC) Data Hub, a multi-institutional contemporary cohort of individuals living with SCD.Methods

We conducted a cross-sectional analysis using the 2024 data from the ASH RC Data Hub. Adults (≥18 years) with confirmed SCD and ≥1 year of follow-up were included. Key exclusions were those with a history of transformative therapies, and all individuals from sites with values for ferritin in< 50% patients. As liver iron concentration (LIC) was not readily available in the Data Hub, we used the lowest annual ferritin value with cutoffs derived from our prior study as a surrogate for an elevated LIC (> 5 mg/g): unlikely (ferritin <1000 ng/mL or not reported), possible (ferritin 1000–2500 ng/mL), or probable (ferritin >2500 ng/mL). Limited 3-year carry-over imputation (with measurements back to 2021) was used to estimate 2024 ferritin values for individuals with missing ferritin measurements, given the inherent chronicity of iron overload. Outcomes included prevalence of iron overload, acute care utilization, SCD genotype, and confirmatory abdominal MRI in 2023 or 2024.Results

Of the 3723 confirmed individuals with SCD who met criteria for inclusion, 378 (10%) were classified as probable iron overload, 328 (8.8%) as possible, and 3017 (81%) as unlikely. Individuals with HbSS genotype (n = 2394, 64.3%) had the highest prevalence for possible and probable ('credible') iron overload at 281 (11.7%) and 347 (14.5%), respectively. There were site differences in prevalence for 'credible' iron overload, ranging from 7.9% to 29.7%. Individuals with 'credible’ iron overload had higher mean ED and hospital admissions during the period of study. Twenty (5.3%) of the probable group were reported deceased, compared to 8 (2.4%) and 31 (1.0%) of the possible and unlikely groups, respectively. MRIs were obtained in 105 (27.8%) of the probable group, 75 (22.9%) of the possible group, and 237 (7.9%) of the unlikely group. Chelation was prescribed in 166 (43.9%) of the probable, 104 (31.7%) of the possible, and 82 (2.7%) of the unlikely group.Conclusion

Despite recent guidelines recommending widespread screening for and management of iron overload, the ASH RC Data Hub confirmed that a quarter of contemporary patients with HbSS likely have iron overload. Yet, only a small fraction (27.8% of the probable group) have received confirmatory MRI in the prior two years, and less than 50% are prescribed treatment with chelation. These findings suggest ongoing underdiagnosis and suboptimal treatment of iron overload in SCD, with limited MRI assessments on even those with prescribed chelation. Iron overload is associated with higher healthcare utilization and short-term mortality, a key finding that warrants further examination as it may be confounded by disease severity, an indication for transfusions which may worsen iron burden. Site differences in iron overload prevalence may be due to iron overload management differences and variable transfusion therapy use. This study is limited by the cross-sectional analysis, selection bias (only included sites with >50% of patients with ferritin values), the use of ferritin as a surrogate measure of LIC, and potential ascertainment bias for covariates and outcomes. Imputation for missing ferritin values could lead to misclassification, albeit as we coded missing ferritin as unlikely iron overloaded, this approach may underestimate the prevalence. Follow up analyses will examine the longitudinal association of iron overload with clinical outcomes and disease-modifying therapy use. Overall, this study underscores gaps in knowledge regarding the prevalence, and uneven treatment, of iron overload in SCD, which may contribute to morbidity and mortality.