Related Experiment Video
Updated: May 24, 2025
Predictive Immune Modeling of Solid Tumors
Published on: February 25, 2020
Development and validation of a predictive model for diagnosing EBER-positive lymphoma-associated hemophagocytic
Yuhong Yin1, Wenzhi Zhang1, Lizhen Zhao1
1Department of Pediatrics, Union Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, 430022, People's Republic of China.
Purpose:
This study aims to identify distinguishing factors between EBER-positive lymphoma-associated hemophagocytic lymphohistiocytosis and non-neoplastic EBV-associated hemophagocytic lymphohistiocytosis. Additionally, we developed and validated a predictive diagnostic model based on these factors.
Methods:
To evaluate the early identification of individuals with EBER-positive lymphoma-associated hemophagocytic lymphohistiocytosis versus non-neoplastic EBV-associated hemophagocytic lymphohistiocytosis, we carried out a retrospective cohort research. The medical records system included 148 individuals' diagnoses of EBV-associated hemophagocytic lymphohistiocytosis between 2015 and 2023.
Results:
In this study, 148 patients were included, 75 of whom had non-neoplastic EBV-associated hemophagocytic lymphohistiocytosis and the remaining 73 had EBER-positive lymphoma-associated hemophagocytic lymphohistiocytosis. The highest AUC, with a good predictive value, was found for IL-10 > 39.87 pg/ml in separating EBER-positive lymphoma-associated hemophagocytic lymphohistiocytosis from non-neoplastic EBV-associated hemophagocytic lymphohistiocytosis. The diagnosis of EBER-positive lymphoma-associated hemophagocytic lymphohistiocytosis was influenced by platelets < 33.5*109/L, IL-6 > 20.79 pg/ml, and IFN-γ > 12.12 pg/ml as independent variables. These factors were combined with the predictive value of IL-10 > 39.87 pg/ml to establish the predictive model of the nomogram for diagnosis. The training set's and validation set's areas under the ROC curves were 0.825 and 0.812, respectively, showing that the model had good discrimination, a well-calibrated model, and a clinically valid model as indicated by the clinical decision curve.
Conclusion:
The results of this study showed that the prediction model based on platelets < 33.5*109/L, IL-6 > 20.79 pg/ml, IFN-γ > 12.12 pg/ml, and IL-10 > 39.87 pg/ml could more accurately distinguish between EBER-positive lymphoma-associated hemophagocytic lymphohistiocytosis and non-neoplastic EBV-associated hemophagocytic lymphohistiocytosis. This could aid clinicians in the early detection and convenient individualization of treatment for EBV-associated hemophagocytic lymphohistiocytosis.
