Work

Longitudinal Spatial Proteomics of Breast Cancer Reveals Recurrent Phenotypic Clone States and Temporal Trajectories

Year: 2026

Type: article

Abstract: Intratumoral heterogeneity complicates treatment response and relapse in breast cancer, yet longitudinal spatial studies that preserve tissue architecture remain scarce. Here we integrated matrix-assisted laser desorption/ionization mass spectrometry imaging (MALDI-MSI) with region-resolved dia-PASEF proteomics to map recurrent spatial proteomic programs across serial breast cancer specimens. Forty-seven formalin-fixed paraffin-embedded tissues from 16 patients were analyzed, encompassing primary tumors, relapses, and metastatic lesions. Cohort-level MSI co-segmentation provided a common reference atlas of 10 recurrent spectral states, and pathologist-guided microextraction of 201 tumor-containing regions yielded 402 extraction-level proteomic datasets. Because no genomic lineage tracing was performed, we use the term recurrent spatial proteomic state to denote a spatially localized, recurrent proteomic phenotype and not a genetically proven clone. Recurrent spatial proteomic states were observed across hormone receptor-positive, HER2-positive, and hormone receptor-negative/HER2-low disease, indicating that similar spatial programs can recur across clinical subtypes. Longitudinal analyses of representative patients showed marked state turnover during therapy, with persistence of some states and emergence of others at relapse or metastatic progression. Feature-ranking models identified proteins associated with each spatial state and with clinical sampling stage, while exploratory enrichment analyses suggested broad shifts from earlier immune- and hormone-associated programs toward later inflammatory, invasive, and stress-adapted programs. Because the cohort is modest, repeated regions from the same patient are not independent, and no external validation cohort was available, all predictive and therapeutic interpretations are presented as hypothesis-generating. These data provide a spatial proteomic atlas of recurrent spatial proteomic states in breast cancer and a framework for studying constrained, but not deterministic, evolutionary trajectories during progression and treatment.

Source: Molecular & Cellular Proteomics

DOI: 10.1016/j.mcpro.2026.101698