Related Experiment Video
Updated: May 18, 2026
Comparison of Predictive Performance of Three Lymph Node Staging Systems in Colorectal Signet Ring Cell Carcinoma Based on Machine Learning Model
Published on: April 18, 2025
A new diagnostic algorithm for Burkitt and diffuse large B-cell lymphomas based on the expression of CSE1L and STAT3
D Soldini1, C Montagna, P Schüffler
1Institute of Surgical Pathology, University Hospital Zurich, Zurich, Switzerland.
Background:
Aggressive mature B-cell non-Hodgkin's lymphomas (BCL) sharing features of Burkitt's lymphoma (BL) and diffuse large B-cell lymphoma (DLBCL) (intermediate BL/DLBCL) but deviating with respect to one or more characteristics are increasingly recognized. The limited knowledge about these biologically heterogeneous lymphomas hampers their assignment to a known entity, raising incertitude about optimal treatment approaches. We therefore searched for discriminative, prognostic, and predictive factors for their better characterization.
Patients And Methods:
We analyzed 242 cytogenetically defined aggressive mature BCL for differential protein expression. Marker selection was based on recent gene-expression profile studies. Predictive models for diagnosis were established and validated by a different set of lymphomas.
Results:
CSE1L- and inhibitor of DNA binding-3 (ID3)-overexpression was associated with the diagnosis of BL and signal transduction and transcription-3 (STAT3) with DLBCL (P<0.001 for all markers). All three markers were associated with patient outcome in DLBCL. A new algorithm discriminating BL from DLBCL emerged, including the expression of CSE1L, STAT3, and MYC translocation. This 'new classifier' enabled the identification of patients with intermediate BL/DLBCL who benefited from intensive chemotherapy regimens.
Conclusion:
The proposed algorithm, which is based on markers with reliable staining properties for routine diagnostics, represents a novel valid tool in separating BL from DLBCL. Most interestingly, it allows segregating intermediate BL/DLBCL into groups with different treatment requirements.
