Progression-free survival prediction in diffuse large B-cell lymphoma: where can we further improve?

Diffuse large B-cell lymphoma (DLBCL) is a biologically heterogeneous disease in which approximately one-third of patients experience primary refractoriness or relapse despite modern chemoimmunotherapy. Although clinical prognostic indices remain the cornerstone of risk stratification, they incompletely capture the biological complexity underlying disease progression. Accurate prediction of progression-free survival (PFS) is essential to support personalized treatment strategies.

This review summarizes advances in PFS prediction in DLBCL, including clinical prognostic models, quantitative PET imaging, molecular and transcriptomic classification, genomic classifiers, tumor microenvironment profiling, circulating tumor DNA (ctDNA)-based biomarkers, and integrated multimodal prognostic models. Relevant literature was identified through PubMed, with ClinicalTrials.gov searched for ongoing clinical trials; searches were updated through 3 October 2026. We discuss the strengths, limitations, and clinical applicability of these approaches.

Future prognostic models will increasingly rely on integrated multimodal frameworks combining clinical, imaging, molecular, immune, and dynamic response biomarkers. Among these, ctDNA has emerged as one of the most promising tools for measurable residual disease assessment. However, improving patient outcomes will require moving beyond prognostic accuracy toward predictive biomarkers capable of guiding personalized treatment selection across an evolving therapeutic landscape. Prospective validation, assay standardization, and demonstration of clinical utility will be essential for translating these approaches into routine practice.