Non-small cell lung cancer (NSCLC) remains a clinically heterogeneous disease with variable prognosis despite recent therapeutic advances. The tumor microenvironment (TME) is a key determinant of tumor behavior and patient outcome, composed of multiple immune and stromal cell types organized in distinct spatial patterns. This study assessed the global immune landscape and specifically the spatial architecture of excluded immune cell populations located in the stromal compartment of NSCLC tissues to evaluate their relationship with overall survival.
Tissues from 674 NSCLC patients were analyzed using multiplex immunofluorescence (mIF) to detect major immune cell lineages. Images underwent AI-based digital image analysis that (1) segmented tissue into tumor and stromal compartments, and (2) identified and classified individual cells based on combinations of mIF markers to define major immune phenotypes at single-cell resolution.
Quantitative measures of immune cell phenotypes were extracted from the stromal area only. These single-cell resolved stromal data were used to cluster patients according to their excluded immune signatures and the spatial architecture of those excluded cells. The clustered groups were then explored for associations with overall survival.
The stromal compartment of NSCLC tissues exhibited distinct immune compositions and spatial organizations with varying combinations of cell densities and intercellular distances. Among identified patterns, a mixed immune landscape combined with a stromal organization characterized by long distances to B cells and short distances to Helper T cells was associated with long-term overall survival.
Using the spatially resolved stromal data, the authors were able to reduce the prognostic signal to a compact signature involving three cell types and two variables. Specifically, the combination of short distances between Helper T cells and Regulatory T cells together with high B cell densities provided superior prognostic value compared with other patterns described.
The study highlights that not only the presence or density of immune cell types but also their spatial relationships in the stroma—particularly among excluded immune populations—carry prognostic information in NSCLC.
The authors report a simplified two-variable signature that captures prognostic information present in more complex stromal immune patterns. The signature comprises:
This combined pattern was associated with superior long-term overall survival in the analysed cohort. The source abstract indicates the signature provided robust prognostic value after reducing the more complex clustering results, but specific performance metrics, statistical values, cutoffs, and validation details were not provided in the supplied abstract.
Clinical implications:
Limitations and missing details from the supplied source text:
Because the PubMed record's abstract ends mid-sentence, readers should consult the full text for complete methods, statistical analyses, and all results before applying these findings or designing follow-up studies.
In this cohort of 674 NSCLC patients, quantitative single-cell spatial analysis of excluded immune populations in the stromal compartment identified distinct immune landscapes. A mixed immune pattern with long distances to B cells and short distances to Helper T cells correlated with longer overall survival. The authors further distilled these observations into a reduced prognostic signature where short Helper T cell–Regulatory T cell distances together with high B cell densities provided the strongest prognostic signal reported in the abstract.
The PubMed abstract provided here is truncated and does not include full methodological details, statistical measures, or the concluding sentence. Those missing elements were not reported in the supplied source text and should be reviewed in the full article for complete interpretation and potential clinical application.