Regresar

MALDI imaging mass spectrometry and chemometric tools to discriminate highly similar colorectal cancer tissues

Abstract:

Intratumour heterogeneity due to cancer cell clonal evolution and microenvironment composition and tumor differences due to genetic variations between patients suffering of the same cancer pathology play a crucial role in patient response to therapies. This study is oriented to show that matrix-assisted laser-desorption ionization-Mass spectrometry imaging (MALDI-MSI), combined with an advanced multivariate data processing pipeline can be used to discriminate subtle variations between highly similar colorectal tumors. To this aim, experimental tumors reproducing the emergence of drug-resistant clones were generated in athymic mice using subcutaneous injection of different mixes of two isogenic cell lines, the irinotecan-resistant HCT116-SN50 (R) and its sibling human colon adenocarcinoma sensitive cell line HCT116 (S). Because irinotecan-resistant and irinotecan-sensitive are derived from the same original parental HCT116 cell line, their genetic characteristics and molecular compositions are closely related. The multivariate data processing pipeline proposed relies on three steps: (a) multiset multivariate curve resolution (MCR) to separate biological contributions from background; (b) multiset K-means segmentation using MCR scores of the biological contributions to separate between tumor and necrotic parts of the tissues; and (c) partial-least squares discriminant analysis (PLS-DA) applied to tumor pixel spectra to discriminate between R and S tumor populations. High levels of correct classification rates (0.85), sensitivity (0.92) and specificity (0.77) for the PLS-DA classification model were obtained. If previously labelled tissue is available, the multistep modeling strategy proposed constitutes a good approach for the identification and characterization of highly similar phenotypic tumor subpopulations that could be potentially applicable to any kind of cancer tissue that exhibits substantial heterogeneity.