TME-analyzer: a new interactive and dynamic image analysis tool that identified immune cell distances as predictors for survival of triple negative breast cancer patients
Hayri E. Balcioglu,
Rebecca Wijers,
Marcel Smid,
Dora Hammerl,
Anita M. Trapman-Jansen,
Astrid Oostvogels,
Mieke Timmermans,
John W. M. Martens,
Reno Debets
Affiliations
Hayri E. Balcioglu
Laboratory of Tumor Immunology, Department of Medical Oncology, Erasmus MC Cancer Institute
Rebecca Wijers
Laboratory of Tumor Immunology, Department of Medical Oncology, Erasmus MC Cancer Institute
Marcel Smid
Laboratory of Translational Cancer Genomics, Department of Medical Oncology, Erasmus MC Cancer Institute
Dora Hammerl
Laboratory of Tumor Immunology, Department of Medical Oncology, Erasmus MC Cancer Institute
Anita M. Trapman-Jansen
Laboratory of Translational Cancer Genomics, Department of Medical Oncology, Erasmus MC Cancer Institute
Astrid Oostvogels
Laboratory of Tumor Immunology, Department of Medical Oncology, Erasmus MC Cancer Institute
Mieke Timmermans
Laboratory of Translational Cancer Genomics, Department of Medical Oncology, Erasmus MC Cancer Institute
John W. M. Martens
Laboratory of Translational Cancer Genomics, Department of Medical Oncology, Erasmus MC Cancer Institute
Reno Debets
Laboratory of Tumor Immunology, Department of Medical Oncology, Erasmus MC Cancer Institute
Abstract Spatial distribution of intra-tumoral immune cell populations is considered a critical determinant of tumor evolution and response to therapy. The accurate and systemic search for contexture-based predictors would be accelerated by methods that allow interactive visualization and interrogation of tumor micro-environments (TME), independent of image acquisition platforms. To this end, we have developed the TME-Analyzer, a new image analysis tool, which we have benchmarked against 2 software tools regarding densities and networks of immune effector cells using multiplexed immune-fluorescent images of triple negative breast cancer (TNBC). With the TME-Analyzer we have identified a 10-parameter classifier, predominantly featuring cellular distances, that significantly predicted overall survival, and which was validated using multiplexed ion beam time of flight images from an independent cohort. In conclusion, the TME-Analyzer enabled accurate interactive analysis of the spatial immune phenotype from different imaging platforms as well as enhanced utility and aided the discovery of contextual predictors towards the survival of TNBC patients.
