Hyper-spectral mass spectrometric imaging to enhance breast cancer diagnostics
File(s)
Author(s)
Gurung, Dipa Kumari
Type
Thesis
Abstract
Breast cancer is a highly heterogeneous disease and one of the most prevalent forms of cancer in women worldwide. Manual histological evaluation of the tissue sections has been the gold standard practice to establish diagnosis. However, the accuracy of this diagnostic approach is often compromised as it is dependent on the pathologists’ subjective interpretation, resulting in observer-to-observer reproducibility. In contrast, the chemical analysis of major tissue components such as lipids, metabolites and proteins offer a more robust alternative. Lipids represent an important class of biomolecules fulfilling many biological functions. While many of their functions remain poorly understood, it is known that changes in lipid metabolism is one of the major hallmarks in cancer, hence analysis of tissue lipidome can potentially be used as means for cancer detection and diagnostics.
DESI-MSI (desorption electrospray ionisation mass spectrometric imaging) is an emerging mass spectrometry technique with great potential in tissue analysis, especially in histological settings. DESI–MSI enables to visualise spatial distribution of lipid species across tissue sections allowing direct correlation of the lipidomic information with the morphological features. Therefore, as a first step, samples from 165 patients undergoing breast surgery were analysed by DESI-MSI to understand the lipidomic and lipidomic differences between histologically normal and tumour samples. The results showed that DESI-MSI can be effectively used for discriminating between the malignant and histologically normal glandular tissues (93% for negative ion mode and 86% in positive ion mode). Lipids belonging to the phosphatidylethanolamine, phosphatidylcholine, phosphophatidylglycerol and phosphatidylinositol classes were found to be more abundant in the tumour carrying tissue samples compared to the normal ones. The tumour dataset was divided into molecular subgroups to correlate with clinical phenotypes. Based on the lipidomic profile, an overall classification accuracy of 76% was achieved for discriminating histological grades of tumour, 76% for estrogen and 70% for progesterone receptor status. Furthermore, an average Spearman’s correlation coefficient of 0.63 was achieved when the distribution of lipid species in the tumour samples was compared with the distribution of the estrogen alpha receptor in tumour samples. The DESI-MSI analysis was further extended to formalin fixed paraffin embedded (FFPE) breast samples - the tissue preservation method routinely used in hospitals world-wide. The results showed that DESI-MSI can be used for the imaging of fatty acids and lipids in FFPE tissue samples despite harsh FFPE processing steps. A tissue microarray slide containing 89 cores from 30 different cancer patients was analysed and DESI-MSI was able to discriminate between tumour and surrounding normal tissues with 96% accuracy, highlighting the potential for investigating diseases in archived FFPE tissues.
In conclusion, DESI-MSI has been validated for the purpose of lipidomic analysis of breast cancer. This thesis shows that DESI-MSI can be utilised in the objective diagnosis of breast cancer and could therefore can serve as a reliable tissue characterisation tool in clinics.
DESI-MSI (desorption electrospray ionisation mass spectrometric imaging) is an emerging mass spectrometry technique with great potential in tissue analysis, especially in histological settings. DESI–MSI enables to visualise spatial distribution of lipid species across tissue sections allowing direct correlation of the lipidomic information with the morphological features. Therefore, as a first step, samples from 165 patients undergoing breast surgery were analysed by DESI-MSI to understand the lipidomic and lipidomic differences between histologically normal and tumour samples. The results showed that DESI-MSI can be effectively used for discriminating between the malignant and histologically normal glandular tissues (93% for negative ion mode and 86% in positive ion mode). Lipids belonging to the phosphatidylethanolamine, phosphatidylcholine, phosphophatidylglycerol and phosphatidylinositol classes were found to be more abundant in the tumour carrying tissue samples compared to the normal ones. The tumour dataset was divided into molecular subgroups to correlate with clinical phenotypes. Based on the lipidomic profile, an overall classification accuracy of 76% was achieved for discriminating histological grades of tumour, 76% for estrogen and 70% for progesterone receptor status. Furthermore, an average Spearman’s correlation coefficient of 0.63 was achieved when the distribution of lipid species in the tumour samples was compared with the distribution of the estrogen alpha receptor in tumour samples. The DESI-MSI analysis was further extended to formalin fixed paraffin embedded (FFPE) breast samples - the tissue preservation method routinely used in hospitals world-wide. The results showed that DESI-MSI can be used for the imaging of fatty acids and lipids in FFPE tissue samples despite harsh FFPE processing steps. A tissue microarray slide containing 89 cores from 30 different cancer patients was analysed and DESI-MSI was able to discriminate between tumour and surrounding normal tissues with 96% accuracy, highlighting the potential for investigating diseases in archived FFPE tissues.
In conclusion, DESI-MSI has been validated for the purpose of lipidomic analysis of breast cancer. This thesis shows that DESI-MSI can be utilised in the objective diagnosis of breast cancer and could therefore can serve as a reliable tissue characterisation tool in clinics.
Version
Open Access
Date Issued
2019-03
Date Awarded
2019-09
Copyright Statement
Creative Commons Attribution NonCommercial NoDerivatives Licence
Advisor
Takats, Zoltan
Nicholson, Jeremy
Sponsor
Imperial College London
Publisher Department
Department of Metabolism, Digestion and Reproduction
Publisher Institution
Imperial College London
Qualification Level
Doctoral
Qualification Name
Doctor of Philosophy (PhD)