Multi-modal image analysis for large-scale cancer tissue studies within IMMUcan.

Area of Science:

• Cancer Research
• Computational Biology
• Biomedical Imaging

Background:

• Multiplexed imaging is crucial for characterizing the tumor microenvironment (TME) and its relation to patient prognosis in cancer research.
• The IMMUcan consortium gathers multi-modal imaging data from extensive cancer patient cohorts for spatial profiling.
• Standardized data generation is essential for reliable analysis of complex cancer tissue samples.

Purpose of the Study:

• To describe and compare two standardized workflows for multiplexed immunofluorescence (mIF) and imaging mass cytometry (IMC) within the IMMUcan consortium.
• To introduce the IFQuant software for user-friendly and reproducible analysis of mIF data.
• To optimize high sample throughput for IMC and improve cell-phenotyping accuracy.

Main Methods:

• Development and comparison of mIF and IMC workflows for standardized cancer tissue analysis.
• Implementation of the IFQuant software for web-based mIF data analysis.
• Optimization of IMC protocols, including robotic slide loading and classification-based cell typing, utilizing manually labeled single-cell data.

Main Results:

• Established standardized pipelines for mIF and IMC data generation within the IMMUcan consortium.
• Demonstrated that tree-based methods significantly outperform other cell-phenotyping tools using manually labeled single-cell data.
• Achieved high sample throughput for IMC through protocol optimization and automation.

Conclusions:

• The developed mIF and IMC pipelines provide a robust foundation for multiplexed image analysis in large cancer cohorts.
• IFQuant software enhances the accessibility and reproducibility of mIF data analysis.
• The study highlights key learnings from five years of development, emphasizing the importance of standardized workflows and advanced computational methods for cancer spatial profiling.