Hyperferritinemia in Pediatric Acute Lymphoblastic Leukemia: What Does it Mean?
As long-term survival of pediatric acute lymphoblastic leukemia (ALL) improves, emphasis on the recognition and treatment of late effects of therapy is increasing.1,2 Iron overload is a known cause of morbidity in chronically transfused patients, results from as few as 10 transfusions and has an overlapping toxicity profile with chemotherapy.3–5
We sought to assess red blood cell (RBC) transfusion requirements and evaluate ferritin levels and bone marrow aspirate (BMA) staining in the assessment of iron status in pediatric ALL. In total, 28 consecutive children treated for ALL between August 2010 and January 2014 at the Royal Hospital for Sick Children, Edinburgh were included for analysis. RBCs were administered if the hemoglobin level fell to <80 g/L at a standard pediatric dose of 15 mL/kg. Serum ferritin was measured 3 monthly up to 1 year, then 6 monthly up to 3 years. BMA iron was assessed by Perl Prussian blue, where bone marrow hemosiderin deposition is graded from 0 (no stainable iron) to 6 (very heavy), normal range is 1 to 3.6
Patient characteristics are summarized in table I. All patients required RBCs, with a median of 195 mL/kg (range, 45 to 570 mL/kg) of packed RBCs per child, corresponding to 13 units and a median iron burden of 130 mg/kg (range, 30 to 380 mg/kg). RBC requirements were significantly higher in children with high-risk disease compared with standard risk (281 vs. 150 mL/kg; P=0.01). RBC transfusion volumes were non-normally distributed, with 25% requiring 10 units and outliers receiving up to 50 transfusions.
Median ferritin at diagnosis was 629 μg/L (range, 76 to 2790 μg/L) and end-of-treatment ferritin values normalized in only 1 patient. A ferritin value >500 μg/L was recorded in 93% (26/28) during the course of treatment. In total, 11% of patients had a ferritin >10,000 μg/L recorded. Ferritin was significantly higher in the high-risk than standard-risk cohort (median, 2089 vs. 698 μg/L; P=0.006). There was no significant correlation between serum ferritin and BMA iron staining.
At day 28, median BMA iron stain was 3 (n=13, range 0 to 4). At day 14, 28 BMAs (52% of total) were aparticulate, preventing iron storage assessment. Twenty BMAs were available from patients at a later stage of treatment; 4 (20%) were aparticulate and median iron grade was 4 (range, 2 to 5). In total, 35% of BMAs assessed at a later stage of treatment demonstrated ≥4+ iron staining, consistent with iron overload. There was a significant correlation between iron transfused (mg/kg) and BMA grading (r=0.7, P=0.035).
Our findings indicate that patients with pediatric ALL receive a significant iron burden during therapy, placing them at risk of later morbidity and that serum ferritin is not a reliable indicator of iron status in this population. Hyperferritinemia has a wide differential aside from iron overload, including infection, inflammation, and hemophagocytic lymphohistiocytosis (HLH).7 Extreme hyperferritinemia in the context of ALL is not specific for a diagnosis of HLH although ALL is a recognized cause of secondary HLH.
As the focus in pediatric ALL shifts from survival to preventing long-term morbidity, the need to establish clear guidelines for screening and management of iron overload in the context of pediatric ALL is evident.