Related Experiment Video
Updated: Jul 30, 2025
From a 2DE-Gel Spot to Protein Function: Lesson Learned From HS1 in Chronic Lymphocytic Leukemia
Published on: October 19, 2014
A predictive model for identifying secondary underlying diseases of hemophagocytic lymphohistiocytosis
Wei-Bo Gao1, Li-Juan Hu2,3, Xiao-Lu Ma1
1Department of Emergency, Peking University People's Hospital, Beijing, China.
Background:
Secondary hemophagocytic lymphohistiocytosis (HLH) is a rare, life-threatening disease of immune hyperactivation that arises in the context of infectious, inflammatory, or neoplastic triggers. The aim of this study was to establish a predictive model for the timely differential diagnosis of the original disease resulting in HLH by validating clinical and laboratory findings to further improve the efficacy of therapeutics for HLH.
Methods:
We retrospectively enrolled 175 secondary HLH patients in this study, including 92 patients with hematologic disease and 83 patients with rheumatic disease. The medical records of all identified patients were retrospectively reviewed and used to generate the predictive model. We also developed an early risk score using multivariate analysis weighted points proportional to the β regression coefficient values and calculated its sensitivity and specificity for the diagnosis of the original disease resulting in HLH.
Results:
The multivariate logistic analysis revealed that lower levels of hemoglobin and platelets (PLT), lower levels of ferritin, splenomegaly and Epstein-Barr virus (EBV) positivity were associated with hematologic disease, but young age and female sex were associated with rheumatic disease. The risk factors for HLH secondary to rheumatic diseases were female sex [OR 4.434 (95% CI, 1.889-10.407), P =0.001], younger age [OR 6.773 (95% CI, 2.706-16.952), P<0.001], higher PLT level [OR 6.674 (95% CI, 2.838-15.694), P<0.001], higher ferritin level [OR 5.269 (95% CI, 1.995-13.920), P =0.001], and EBV negativity [OR 27.656 (95% CI, 4.499-169.996), P<0.001]. The risk score included assessments of female sex, age, PLT count, ferritin level and EBV negativity, which can be used to predict HLH secondary to rheumatic diseases with an AUC of 0.844 (95% CI, 0.836~0.932).
Conclusion:
The established predictive model was designed to help clinicians diagnose the original disease resulting in secondary HLH during routine practice, which might be improve prognosis by enabling the timely treatment of the underlying disease.
