The Performance of Diagnostic Criteria for Hemophagocytic Lymphohistiocytosis in Critically Ill Patients

Objective: The diagnostic criteria for secondary hemophagocytic lymphohistiocytosis (HLH) have not been validated in the critically ill adult population. We set out to evaluate the performance of diagnostic criteria and determine the ferritin cutoff in critically ill adults. Design: A retrospective single-center study. Setting and Patients: Patients admitted to intensive care unit between 2008 and March 2010. Data were collected on consecutive patients who had ferritin measured. Charts were reviewed for the diagnostic criteria of HLH and components of Hscore. Measurements and Main Results: A total of 445 patients had a ferritin level measured during the study period. A diagnosis of HLH was made for 10 patients. Having 5 of 6 criteria had a specificity of 97% and a sensitivity of 70%. Hemophagocytosis was found in 41 (47.1%) of 87 bone marrow biopsies. Two hundred thirty-one patients had a ferritin level above 500 ng/dL. When determining the odds of HLH being clinically diagnosed, the optimal cut point for ferritin was 1197 ng/dL. When determining the odds of HLH based on the Hscore, the best cutoff was 143.5 (sensitivity of 90% and specificity of 90%) and patients who had HLH in our study population had an Hscore of 203.8 +/- 64.9. Conclusion: In this cohort of critically ill patients, the HLH criteria are specific for HLH but not sensitive. Critically ill patients can have a higher incidence of hemophagocytosis without HLH. A higher ferritin cutoff in combination with 5 other clinical criteria is comparable to the Hscore for the recognition of HLH in the critically ill population.