Author: Todd A Aguilera, Bassel Dawod, Sebastian Diegeler, Eslam Elghonaimy, Purva Gopal, Jiaqi Liu, Hao Peng, Arely Perez Rodriguez, Nina N. Sanford, Robert Timmerman, Megan B Wachsmann, Haozhao Zhang 👨🔬
Affiliation: Department of Radiation Oncology, UT Southwestern Medical Center, Medical Artificial Intelligence and Automation (MAIA) Lab, Department of Radiation Oncology, UT Southwestern Medical Center, UT Southwestern Medical Center 🌍
Purpose: This study pioneers the integration of CODEX (co-detection by indexing)-based spatial profiling and advanced computational techniques to investigate the tumor immune microenvironment (TIME) in patients with locally advanced rectal cancer (LARC). Specifically, we aim to validate a framework for identifying synergistic effects of integrating sotigalimab (Sotiga), a CD40 agonist, with short-course radiotherapy (SCRT).
Methods: Multiplexed spatial imaging was conducted using the CODEX platform to profile immune cell phenotypes in pre- and post-treatment samples. Cell-cell interactions were quantified by analyzing pairwise relationships between each cell type and its five nearest neighbors, identified using the cKDTree algorithm. Statistical enrichment/depletion of cell-cell interactions was assessed with Z-scores calculated against null distributions generated from 1,000 permutations. Spatial clustering of CD8+ T cells and macrophages in tumor regions was analyzed using normalized cross-K (nK) functions with Ripley’s isotropic correction to mitigate edge effects. Confidence invervals for nK functions were derived from random spatial permutations (n=999) to identify significant clustering beyond complete spatial randomness.
Results: The framework was successfully applied to pre- and post-treatment samples from two LARC patients undergoing SCRT with or without Sotiga. Initial results suggest that this approach reliably captures cell-cell interaction dynamics within the TIME. Observed divergence of the clustering of CD8+ T cells and macrophages aligns with prior findings linking these immune cells to treatment response and survival, supporting the validity of the methodology.
Conclusion: This study highlights the feasibility and potential of integrating CODEX-based spatial profiling with computational methods to study the TIME in LARC. By establishing a foundation for spatial and interaction analysis, this approach offers a valuable tool for exploring the immunological effects of novel combination therapies. Future research will extend this framework to larger datasets and explore integration with radiomics and adaptive radiotherapy modalities, such as PULSAR (personalized, ultra-fractionated stereotactic adaptive radiotherapy).