Integrated spatially guided transcriptomics and multiplex imaging identify a high-risk CD163+/CD11c-subgroup in...

Emerging data identify infiltrating immune cells in diffuse large B‑cell lymphoma (DLBCL) as prognostic biomarkers with tentative implications for therapeutic stratification. To elucidate the tumor immune microenvironment (TIME) and molecular features governing outcome, we integrated data from multiplex immunofluorescence imaging of CD163⁺ and CD11c⁺ cells with spatially guided transcriptional profiling of infiltrating CD3⁺ and CD163⁺ immune cells alongside CD20⁺ malignant cells across a cohort of 561 DLBCL cases. For quantification of cell densities, a multi‑step workflow was used, incorporating machine‑learning classifiers to improve accuracy, and the underlying biology was explored through cancer and immunefocused gene expression analysis. We identified a subgroup of patients with a CD163low/CD11chigh TIME and favorable prognosis. By contrast, a CD163high/CD11clow TIME exhibited poor outcome independent of established clinicopathological risk factors. To verify the clinical utility, we showed that these subgroups can be identified with a routine pathology dual IHC staining. The high-risk CD163/CD11c subgroup had abundant expression of M2‑like markers (MRC1, SIGLEC1, MARCO) and showed evidence of T-cell exhaustion, with higher levels of LAG3. Of tentative translational value, the low-risk TIME expressed the checkpoint inhibitor CTLA4. Multivariate modelling identified a high-risk gene signature with prognostic value in independent public datasets of CHOP and R‑CHOP-treated DLBCL, underscoring the robust and broad prognostic impact of CD163+ and CD11+ cells across clinical studies. The combined analysis of CD163 and CD11c have the potential for implementation in routine clinical use and provides independent prognostic information with implications for stratification to novel therapeutic regimens, potentially checkpoint inhibitors LAG-3 or CTLA-4 for high and low-risk groups, respectively.