SPARTIN: a Bayesian method for the quantification and characterization of cell type interactions in spatial pathology...

Area of Science:

• Digital pathology
• Computational biology
• Cancer research

Background:

• High-resolution digital pathology enables detailed analysis of the tumor microenvironment.
• Understanding immune cell spatial composition is crucial for cancer development and therapeutic strategies.

Purpose of the Study:

• To introduce SPatial Analysis of paRtitioned Tumor-Immune imagiNg (SPARTIN), a Bayesian method for quantifying immune cell infiltration in digital pathology images.
• To develop a novel measure, Cell Type Interaction Probability (CTIP), to assess local tumor-immune cell interactions and their uncertainty.

Main Methods:

• SPARTIN utilizes Bayesian point processes for spatial quantification of immune cell infiltration.
• The method calculates Cell Type Interaction Probability (CTIP) to measure local tumor-immune cell interactions.
• The R-package for SPARTIN is publicly available.

Main Results:

• SPARTIN accurately distinguishes cellular interaction patterns in simulations.
• Analysis of 335 melanoma biopsies revealed significant associations between CTIP and immune cell prevalence (e.g., CD8+ T-Cells, NK cells).
• CTIP scores differed across transcriptomic classes and were associated with survival outcomes.

Conclusions:

• SPARTIN offers a general framework for analyzing spatial cellular interactions in digital histopathology.
• Findings have implications for melanoma treatment and prognosis.
• The study highlights the importance of spatial immune cell infiltration patterns in cancer.