# Glycolysis-Driven Immune Subtypes in the Tumor Microenvironment Determine Clinical Outcomes in Diffuse Large B-Cell Lymphoma

## TL;DR

This study shows that glycolysis patterns in tumors correlate with immune subtypes and affect treatment outcomes in diffuse large B-cell lymphoma.

## Contribution

The paper introduces a metabolism- and immune-based classification framework for DLBCL prognosis and treatment optimization.

## Key findings

- Four glycolysis-related patterns were identified, correlating with three immune subtypes in DLBCL.
- The immune-inflamed subtype was associated with better survival and higher sensitivity to doxorubicin.
- The immune-desert subtype was linked to poor survival outcomes, while the immune-excluded subtype showed no survival benefit.

## Abstract

Diffuse large B-cell lymphoma (DLBCL) is an aggressive cancer with profound heterogeneity. While standard immunochemotherapy is curative for the majority, treatment failure in 30–40% of patients due to refractory or relapsed disease remains a significant clinical challenge, emphasizing the need for better prognostic markers. We investigated the interplay between tumor glycolytic metabolism and the tumor immune microenvironment. We identified four distinct glycolysis-related patterns that were associated with three immune subtypes (known as immune-inflamed, immune-excluded, and immune-desert). Importantly, these patterns stratified clinical outcomes and predicted chemotherapy sensitivity. For example, tumors with an immune-inflamed subtype were more sensitive to doxorubicin. Collectively, our findings propose a metabolism- and immune-based classification framework for DLBCL that may facilitate more precise risk stratification and treatment optimization.

Background: Diffuse large B-cell lymphoma (DLBCL), the most common B-cell non-Hodgkin lymphoma, remains challenging because more than one-third of patients fail to achieve durable remission with frontline therapy; therefore, a systematic and comprehensive understanding of its underlying biology is essential to improve risk stratification and treatment outcomes. Methods: Based on 30 glycolysis-related genes selected from 72 candidates in the MSigDB database, we comprehensively evaluated glycolysis-related modification patterns in 562 DLBCL samples and analyzed their associations with immune cell infiltration and patient prognosis in the tumor microenvironment (TME). In vivo, we identified three immune subtypes in DLBCL mouse models and accessed their correlation with glycolytsis-related patterns and outcomes using histopathological evaluation (H&E) staining, Kaplan–Meier survival analysis, Western blotting, and immunohistochemistry. Results: Four distinct glycolysis modification patterns were identified, which correlated with three immune subtypes: immune-inflamed, immune-desert, and immune-excluded. The immune-inflamed subtype, characterized by prominent immune infiltration, was associated with favorable survival. In contrast, the immune-excluded subtype, defined by stromal activation, did not confer a survival advantage. The immune-desert subtype, marked by a lack of immune infiltration, was linked to poor survival outcomes. Finally, using the DLBCL mouse model, we validated these three immune subtypes and demonstrated their corresponding associations with glycolytic patterns and prognosis. Conclusions: Our work demonstrated that glycolysis-related modification patterns have correlations with the immune cell infiltration within the TME of DLBCL and that distinct glycolytic states are associated with different clinical outcomes. These findings provide a framework for improved prognostic stratification and may inform new therapeutic decision-making in DLBCL.