2 December 2023 in Computer Methods and Programs in Biomedicine
In Diffuse Large B-Cell Lymphoma (DLBCL), several methodologies are emerging to derive novel biomarkers to be incorporated in the risk assessment. We realized a pipeline that relies on autoencoders (AE) and Explainable Artificial Intelligence (XAI) to stratify prognosis and derive a gene-based signature. AE was exploited to learn an unsupervised representation of the gene expression (GE) from three publicly available datasets, each with its own technology. Multi-layer perceptron (MLP) was used to classify prognosis from latent representation. GE data were preprocessed as normalized, scaled, and standardized. Four different AE architectures (Large, Medium, Small and Extra Small) were compared to find the most suitable for GE data. The joint AE-MLP classified patients on six different outcomes: overall survival at 12, 36, 60 months and progression-free survival (PFS) at 12, 36, 60 months. XAI techniques were used to derive a gene-based signature aimed at refining the Revised International Prognostic Index (R-IPI) risk, which was validated in a fourth independent publicly available dataset. We named our tool SurvIAE: Survival prediction with Interpretable AE. From the latent space of AEs, we observed that scaled and standardized data reduced the batch effect. SurvIAE models outperformed R-IPI with Matthews Correlation Coefficient up to 0.42 vs. 0.18 for the validation-set (PFS36) and to 0.30 vs. 0.19 for the test-set (PFS60). We selected the SurvIAE-Small-PFS36 as the best model and, from its gene signature, we stratified patients in three risk groups: R-IPI Poor patients with High levels of GAB1, R-IPI Poor patients with Low levels of GAB1 or R-IPI Good/Very Good patients with Low levels of GPR132, and R-IPI Good/Very Good patients with High levels of GPR132. SurvIAE showed the potential to derive a gene signature with translational purpose in DLBCL. The pipeline was made publicly available and can be reused for other pathologies.
3 September 2026 in European Journal of Haematology
The Lymph2Cx assay has played a key role in the clinical translation of the diffuse large B‐cell lymphoma (DLBCL) molecular classification for cell‐of‐origin (COO) determination, as mandated by the 2017 Revised WHO Classification. Applicable to routine formalin‐fixed paraffin‐embedded biopsies, Lymph2Cx has shown high concordance with gold‐standard gene expression profiling (GS) performed on fresh/frozen samples. However, conventional concordance metrics may overlook systematic bias. We first reassessed agreement between Lymph2Cx and GS using Bland–Altman analysis and Cohen's kappa. Subsequently, the prognostic impact of the cell‐of‐origin of diffuse large B‐cell lymphoma as determined by Lymph2Cx was compared with that as determined by GS. This was done by assembling large cohorts of publicly available, clinically annotated cases, analyzed by Lymph2Cx or GS, comprising a total of 2321 cases. Lymph2Cx consistently underestimated the frequency of the activated B‐cell‐like (ABC) subtype, misclassifying a subset of ABC cases as germinal center B‐cell‐like (GCB) or unclassified (UNC). While ABC DLBCL retained its adverse prognosis across cohorts, GCB and UNC groups defined by Lymph2Cx showed higher hazard ratios than their GS‐defined counterparts, consistent with redistribution of higher‐risk ABC cases. Overall, although Lymph2Cx remains a valuable tool for routine COO determination, our findings indicate the need for further refinement to improve categorical accuracy, particularly for the ABC subtype, to support optimal clinical risk stratification.