Abstract Trastuzumab-based therapies are the standard of care treatment for HER2-positive breast cancer. We previously demonstrated that mucin 4 (MUC4) expression is associated with trastuzumab resistance and is an independent biomarker of poor clinical response. The aim of this work was to refine the prognostic and predictive value of MUC4.Five HER2-positive breast cancer samples obtained at diagnosis and classified as either MUC4-positive/non-responder or MUC4-negative/responder to trastuzumab were used to analyze spatial expression of transcripts using the Visium platform from 10X Genomics. Libraries were sequenced on a NextSeq 550 and data were processed with Cell Ranger. To minimize batch effects and to optimize noise-to-signal ratio, data were integrated using different approaches: reciprocal PCA, canonical correlation analysis and Harmony. RPCA was selected for optimal batch correction and preservation of biological heterogeneity.Cell annotation included three approaches: manual annotation using cell specific markers, automated prediction using data from other five spatial transcriptomic experiments and a semi-automated method using Unicell. Cells were annotated when at least two of these methods coincided in cell annotation. Differential gene expression was performed with Seurat and limma, accounting for donor and cell type composition. Notably, downregulated genes included multiple immunoglobulin-related transcripts consistent with previously reported trastuzumab response signatures such as HER2DX.To reduce false positives arising from gene list variability, we applied an elastic net penalized logistic regression which refined the initial set of 267 MUC4-associated genes to 37 high confidence candidates. For each gene, we analyzed structural features, biological function, protein-protein interactions and relevance in cancer using resources such as The Human Protein Atlas, Gene Cards, UniProt and PubMed. The analyzed gene set showed strong enrichment for Gene Ontology Biological Process categories related to extracellular matrix remodeling, stromal activation, epithelial-to-mesenchymal transition, angiogenesis, cell invasion and migration, immune modulation and metabolic reprogramming. Spatial transcriptomics therefore identified 37 differentially expressed genes in MUC4-positive breast cancer associated with stromal and immune transcriptional programs that may contribute to trastuzumab resistance and refine MUC4 predictive and prognostic value improving the identification of patients most likely to benefit from trastuzumab treatment. Citation Format: Maria F. Mercogliano, Nadine Schrode, Kristin Beaumont, Roxana Schillaci. A stromal-immune spatial transcriptomic signature associated with MUC4 identifies potential biomarkers of trastuzumab resistance in HER2-positive breast cancer abstract. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 7732.
Mercogliano et al. (2026) studied this question.