G446 Highly elevated ferritin levels are associated with haemophagocytic lymphohistiocytosis – are we missing treatable diagnoses? a retrospective service evaluation of diagnosis in patients with ferritin >10,000 μg/l

BackgroundHaemophagocytic lymphohistiocytosis (HLH) is a hyperinflammatory syndrome which can complicate sepsis, malignancy or autoimmune disease and may lead rapidly to critical illness and death. Early treatment reduces mortality but diagnosis requires a high index of suspicion and correct interpretation of laboratory results. Highly elevated ferritin levels (HEF) >10,000 µg/L are highly specific for HLH and should prompt consideration/exclusion of hyperinflammation. Diagnostic guidelines for HLH were published in 2004 requiring the presence of ≥5/8 criteria.AimTo assess recognition of HLH in a paediatric population with HEF.MethodsThis retrospective study was conducted at 8 UK centres. Biochemistry databases identified patients≤16 years with serum ferritin >10,000 µg/L between 01.04.2014 and 31.03.2017. A standardised proforma was used to collect data. Cases were assessed against the 2004 HLH criteria. Due to limited access to some laboratory tests, modified criteria using a threshold of ≥4/5 (excluding tissue haemophagocytosis, decreased natural killer cell function, increased soluble interleukin-2 receptor) were also applied.ResultsNinety-four patients (60.6% male) were identified. 36% were aged <1 year, 16% 1–5 years, 34% 6–12 years and 14% 13–16 years. Management was in neonatal/paediatric intensive care in 49.4%. Diagnoses included: infections (29.8%), malignancies (21.3%), rheumatological (13.8%), immunological (9.6%) and cardiac surgery (6.4%). A diagnosis of HLH was made by the treating clinical team in 30.9% and considered in the differential in a further 18.1%. Just 5 patients had complete data available to score against all 8 2004 HLH criteria whereas 49 patients had data to score against the modified criteria. Using all available data:14/94 (14.9%) met ≥5/8 criteria and all these patients were diagnosed with HLH by the treating team.33 (35.1%) met ≥4/5 criteria and 17 (51.5%) of these were diagnosed with HLH by clinicians. HLH was not documented as being considered in the differential in 11 (33.3%)Overall mortality was 33.0% (31/94) but was 17.2% (5/29) in those patients diagnosed with HLH during their admission.ConclusionAlthough HEF is highly specific for HLH, the diagnosis was made or considered in just half of paediatric patients with this laboratory result. Increased awareness of this potentially-lethal condition is likely to lead to earlier treatment and reduced mortality.AcknowledgementsThe following authors also contributed to this work – EL Long and F McErlane, Department of Paediatric Rheumatology, Great North Children’s Hospital, Newcastle upon Tyne, UK, K McLellan, Department of Paediatric Rheumatology, Royal Hospital for Children, Glasgow, UK, K Gallagher and P Bale, Department of Paediatric Rheumatology, Addenbrooke’s Hospital, Cambridge, UK, K Mahmood and C Pain, Department of Paediatric Rheumatology, Alder Hey Children’s Hospital, Liverpool, UK