An unsupervised MVA method to compare specific regions in human breast tumor tissue samples using ToF-SIMS

Area of Science:

• Biomedical Engineering
• Analytical Chemistry
• Cancer Research

Background:

• Human breast tumor tissues require advanced imaging techniques for molecular analysis.
• Metabolic flexibility in tumors influences treatment response.
• Conventional imaging lacks the spatial resolution for detailed molecular profiling.

Purpose of the Study:

• To characterize lipid profiles associated with tumor metabolic flexibility and treatment response in breast cancer.
• To develop and validate an unsupervised method for isolating specific tissue regions using imaging ToF-SIMS and PCA.
• To compare molecular differences in stromal and cellular regions of pre- and post-chemotherapy breast tumor tissues.

Main Methods:

• Utilized imaging time-of-flight secondary ion mass spectrometry (ToF-SIMS) for high-resolution molecular imaging.
• Applied principal component analysis (PCA) to segment tissue regions (cellular vs. stromal) from ToF-SIMS data.
• Reconstructed representative spectra from PCA-defined regions of interest for comparative analysis.

Main Results:

• Developed an unsupervised method to effectively isolate stromal and cellular regions, reducing spectral data scatter.
• Demonstrated distinct chemical separation between pre- and post-chemotherapy breast tissue regions using negatively-charged ion species.
• Identified significant differences in fatty acid, monoacylglycerol, diacylglycerol, and vitamin E profiles between pre- and post-therapy tumor tissues.

Conclusions:

• Validated a novel unsupervised approach for analyzing biochemically distinct regions in cancer tissues with imaging ToF-SIMS.
• The method provides a robust framework for comparing diverse tissue samples, particularly those with section-to-section variability.
• The findings offer insights into lipid alterations related to metabolic changes and treatment response in breast cancer.