Prediction of outcome from spatial Protein profiling of triple-negative breast cancers

In tumors, reciprocal spatial interactions between immune cells, their mediators, the extracellular matrix, and mutated neoplastic cells impact all aspects of treatment resistance. The operational mechanisms of these interactions are foundational for developing insights and targets for cancer therapy and prevention. Spatial quantification of the tumor microenvironment system from image data has untapped potential for patient stratification. Here, we present SparTile, a powerful computational approach for the analysis of multiplex proteomics images to reveal clinically relevant structural organization in the tumor microenvironment. SparTile enables robust and unbiased identification and characterization of tumor microenvironments based on spatial relationships among protein markers without the need for cell segmentation or classification. Applied to tissues of patients with triple-negative breast cancer (TNBC), an aggressive subtype of breast cancer, SparTile identifies repeatable microenvironments with specific cellular relationships. Several microenvironments are characterized by risk markers such as Ki67+ (p-value = 0.052) and vimentin+ (p-value < 0.01) tumor cells, which correlate with poor survival. Furthermore, myeloid markers in an MX1-positive tumor environment correlate with shorter survival (p-value = 0.04). Finally, the relative distance between tumor and myeloid cell markers is a strong prognostic risk factor in multivariate Cox models (p-value < 0.01). This distance metric is externally validated on two datasets of breast cancer multiplex images (p-values < 0.01). Our results show that unbiased protein-based and segmentation-free spatial analysis is effective for identifying clinically relevant biomarkers from multiplex tumor images and identifying predictive biology. Foroughi Pour et al. developed a segmentation-free method, called SparTile, for analyzing spatial proteomics data. They show relative spatial distance of tumor and myeloid cells is a prognostic marker of survival risk in triple-negative breast cancer patients. Cancer tissues contain different cell types that interact with each other. These interactions can affect how patients respond to treatment. We developed a computational method, called SparTile, that looks at detailed protein images of tumors to find important spatial arrangements. Using samples of patients with triple-negative breast cancer, an aggressive form of the disease, SparTile revealed repeating spatial patterns linked to poorer survival. It also showed that the distance between tumor cells and certain immune cells can help predict a patient’s outlook. This approach may help identify high-risk patients and discover clues for better cancer treatments.