Targeted gene expression profiling as a tool for diagnostic cell-of-origin determination and prognostic stratification in diffuse large B-cell lymphoma.

Classification of diffuse large B-cell lymphoma (DLBCL) according to cell-of-origin (COO) distinguishes two main biological subtypes: activated B-cell-like (ABC) and germinal center B-cell-like (GCB). Although this distinction reflects different pathogenetic mechanisms, its prognostic impact diminishes in the context of evolving therapeutic strategies. Molecular subtyping of DLBCL, which is based on the spectrum of affected genes and aims to personalize treatment approaches, is currently gaining importance. In the study, we applied targeted gene expression profiling (GEP) using a custom Lympho-qPCR panel, enabling rapid and practically applicable ABC/GCB classification together with risk stratification of patients. RNA isolated from a cohort of 89 DLBCL tissue samples was analyzed using three GEP-based classification models. Model A compared the expression profile with immunohistochemical (IHC) COO determination and showed the expected lower correlation (62 %). Model B employed the expression scores of selected genes to predict COO regardless of IHC classification. Model C was developed as a new IHC-independent prognostic tool allowing patient stratification based on expected survival. Patients identified as high-risk by Model C had significantly worse outcomes, regardless of existing clinical prognostic indicators. In patients with early progression, parallel DNA sequencing analysis (integrative LYNX panel) confirmed complex chromosomal aberrations and defects in BCL2, TP53 and CDKN2A/B. Our results demonstrate that targeted GEP testing represents a robust, rapid, and clinically applicable method for COO determination and risk stratification in DLBCL patients. In the near future, the predictive value of ABC/GCB classification is expected to increase in relation to novel targeted therapeutic regimens. Integration of transcriptomic and genetic data will be essential for independent and individualized risk assessment in the molecular diagnostics of DLBCL.