B Cell Depletion Marks High-Risk Hodgkin Lymphoma, Spatial Study Finds

Patients with classical Hodgkin lymphoma whose tumors lack non-tumoral B cells face a higher risk of their disease progressing, according to a new spatial transcriptomics study published in the Journal of Experimental & Clinical Cancer Research. The work, led by Benedetta Donati and colleagues with senior author Stefano Luminari, links the spatial arrangement of tumor cells and their surrounding immune environment at diagnosis to clinical aggressiveness, and introduces a gene-based score designed to predict which patients will relapse after standard first-line chemotherapy.

Classical Hodgkin lymphoma is currently risk-stratified mainly through clinical and radiologic parameters, yet outcomes remain heterogeneous even among patients receiving standard first-line treatment. The research team set out to examine how the interplay between Hodgkin and Reed–Sternberg cells, the malignant cells of this lymphoma, and the surrounding tumor microenvironment at diagnosis shapes clinical behavior. Their central observation is that a depletion of B cells, both within the tumor microenvironment and in the malignant cells themselves, marks a high-risk profile.

To obtain this spatial view, the researchers used spatial transcriptomics to profile three compartments separately: Hodgkin and Reed–Sternberg cells marked by CD30, non-tumoral B cells marked by CD20, and the remaining tumor microenvironment marked by CD45 positivity and CD30 negativity. They compared a matched series of patients, four with relapsed or refractory disease and four who remained relapse-free. The relapsed or refractory tumors showed loss of B-cell lineage identity within the malignant Hodgkin and Reed–Sternberg cells and a marked depletion of naïve B cells in the surrounding microenvironment.

The team then moved from this small matched discovery series to much larger patient cohorts. They analyzed a retrospective monocentric training cohort of 155 classical Hodgkin lymphoma patients and an independent multicentric retrospective validation cohort of 89 patients. In both cohorts, gene expression was measured using digital barcoding-based profiling of 770 immune-related genes, allowing the researchers to derive and validate a gene score predictive of progression-free survival.

In the training cohort, where the four-year progression-free survival was 80.4 percent (95 percent confidence interval 74.1 to 87.3), genes associated with better progression-free survival were enriched for B-cell programs. From this signal, the researchers built an 18-gene B-cell signature that stratified patients into three groups with significantly different progression-free survival, with a log-rank P value below 0.001. Using elastic net regression, they then condensed this signature into an eight-gene score, which they named the B-score.

The B-score predicted progression-free survival in the training cohort with an area under the curve of 0.73 (95 percent confidence interval 0.62 to 0.84). Importantly, the score remained independently associated with progression-free survival after adjustment for clinical and radiomic variables, with a hazard ratio of 0.39 (95 percent confidence interval 0.17 to 0.92). This means the score added prognostic information beyond what current clinical and imaging-based criteria already capture.

The association held up in the independent validation cohort, which had a four-year progression-free survival of 74.6 percent (95 percent confidence interval 67.0 to 83.1). In that cohort, the B-score achieved an area under the curve of 0.64 (95 percent confidence interval 0.49 to 0.78) and a hazard ratio of 0.30 (95 percent confidence interval 0.10 to 0.88). The authors note that the score provides a quantitative, reproducible and biologically grounded tool for risk stratification that improves the performance of current prognostic criteria.

The biological story emerging from the spatial data is striking: the malignant cells of relapsed or refractory cases had lost their B-cell lineage identity, while the supportive immune neighborhood around them had lost its naïve B-cell population. The B-score translates this spatial and transcriptional biology into a single numeric measure derived from a standard panel of 770 immune-related genes.

Several limitations deserve emphasis. The spatial transcriptomics discovery comparison involved only eight patients, four in each group, so the spatial findings should be considered hypothesis-generating despite their consistency with the larger cohort results. Both the training and validation cohorts were retrospective, and the validation cohort’s discrimination was more modest than in the training data, with a confidence interval for the area under the curve that approaches the null value of 0.5. Even so, the study offers a concrete step toward incorporating tumor-immune spatial biology into the risk assessment of classical Hodgkin lymphoma, a disease where better tools for identifying patients at risk of progression have long been sought.

Subject of Research: Spatial transcriptomic profiling of B-cell depletion as a prognostic marker in classical Hodgkin lymphoma

Article Title: Spatial transcriptomics identifies a B cells–depleted profile underlying high-risk classical Hodgkin Lymphoma

Article References: Donati, B., Lazic, T., Valli, R., Di Chiaro, P., Durmo, R., Ascione, C., Gennaro, A., Ruffini, A., Gotti, M., Paulli, M., Pozzi, S., De Lorenzo, S., Marino, D., Arcari, A., Liso, A., Botto, B., Patti, C., Vitale, C., Neri, A., … Luminari, S. (2026). Spatial transcriptomics identifies a B cells–depleted profile underlying high-risk classical Hodgkin Lymphoma. Journal of Experimental & Clinical Cancer Research. https://doi.org/10.1186/s13046-026-03864-7

Image Credits: AI Generated

DOI: 10.1186/s13046-026-03864-7

Keywords: classical Hodgkin lymphoma, spatial transcriptomics, B cells, tumor microenvironment, Hodgkin and Reed-Sternberg cells, progression-free survival, gene expression profiling, B-score, risk stratification, relapsed refractory lymphoma, biomarker, hematology