ABSTRACT
In combination with cell-intrinsic properties, interactions in the tumor microenvironment modulate therapeutic response. We leveraged single-cell spatial transcriptomics to dissect the remodeling of multicellular neighborhoods and cell-cell interactions in human pancreatic cancer associated with neoadjuvant chemotherapy and radiotherapy. We developed spatially constrained optimal transport interaction analysis (SCOTIA), an optimal transport model with a cost function that includes both spatial distance and ligand-receptor gene expression. Our results uncovered a marked change in ligand-receptor interactions between cancer-associated fibroblasts and malignant cells in response to treatment, which was supported by orthogonal datasets, including an ex vivo tumoroid coculture system. We identified enrichment in interleukin-6 family signaling that functionally confers resistance to chemotherapy. Overall, this study demonstrates that characterization of the tumor microenvironment using single-cell spatial transcriptomics allows for the identification of molecular interactions that may play a role in the emergence of therapeutic resistance and offers a spatially based analysis framework that can be broadly applied to other contexts.
ABSTRACT
AIMS: Small cell lung carcinoma (SCLC) can be classified into transcription factor-based subtypes (ASCL1, NeuroD1, POU2F3). While in-vitro studies suggest intratumoral heterogeneity in the expression of these markers, how SCLC subtypes vary over time and among locations in patients remains unclear. METHODS AND RESULTS: We searched a consecutive series of patients at our institution in 2006-22 for those with greater than one available formalin-fixed paraffin-embedded SCLC sample in multiple sites and/or time-points. Immunohistochemistry for ASCL1, NeuroD1 and POU2F3 was performed and evaluated using H-scores, with subtype assigned based on the positive marker (H-score threshold >10) with the highest H-score. The 179 samples (75, lung; 51, lymph nodes; 53, non-nodal metastases) from 84 patients (74 with two, 10 with more than two samples) included 98 (54.7%) ASCL1-dominant, 47 (26.3%) NeuroD1-dominant, 15 (8.4%) POU2F3-dominant, 17 (9.5%) triple-negative and two (1.1%) ASCL1/NeuroD1 co-dominant samples. NeuroD1-dominant subtype was enriched in non-lung locations. Subtype concordance from pairwise comparison was 71.4% overall and 89.7% after accounting for ASCL1/NeuroD1-dual expressors and technical factors including <500 cells/slide, H-score thresholds and sample decalcification. No significant difference in subtype concordance was noted with a longer time lapse or with extrathoracic versus intrathoracic samples in this cohort. CONCLUSIONS: After accounting for technical factors, transcription factor-based subtyping was discordant among multiple SCLC samples in ~10% of patients, regardless of sample locations and time lapse. Our findings highlighted the spatiotemporal heterogeneity of SCLC in clinical samples and potential challenges, including technical and biological factors, that might limit concordance in SCLC transcription factor-based subtyping.