The quantification of metabolite concentrations in high resolution magic angle spinning (HR-MAS) NMR spectra from ex vivo prostate tissue

Detection of pre-metastatic prostate cancer enables curative intervention to be applied to an otherwise lethal and common disease, facilitating substantial improvement in the quality of life in patients whose untreated disease would have a poor prognosis. While encouraging progress has been made in the screening for and detection of prostate cancer, understanding and predicting the disease prognosis for a given patient remains difficult.;Additional biomarkers may add key data to elucidate the determination of prognosis. While magnetic resonance imaging has been shown to be quite effective at detecting suspicious lesions, the addition of metabolite information through MR spectroscopic imaging has further improved the diagnostic capabilities over imaging alone. As stronger field magnets (3T and higher) become available clinically, the higher resolution spectra of the soluble tissue metabolites further separate their frequencies, facilitating the identification of the metabolites. Studying tissue samples ex vivo at high field strength (11.74T) allows the identification of patterns of metabolites that may uniquely identify tissue type and cancer grade. Additionally, the non-destructive nature of NMR spectroscopy of the tissue allows histologic post processing of the same sample imaged to determine the precise characteristics of that same tissue. High resolution magic angle spinning (HR-MAS) NMR spectroscopy provides a method in which the concentrations of various metabolites in a given prostate tissue sample can be assessed, while preserving the integrity of that sample for later histological evaluation of the same sample.;A more automated, quantitative, and objective method for correlating these metabolite profiles to lesion properties and ultimately patient prognosis is desirable. In this dissertation, metabolite quantification techniques for HR-MAS NMR spectra from ex vivo prostate tissue are explored. First, a commercial software package designed for metabolite quantification of low field, in vivo spectra from brain is assessed for high field, ex vivo spectra from prostate. These results motivated the development of a modified approach specifically suited for the ex vivo prostate spectra. Finally, the classification of the prostate spectral data with a principal component analysis (PCA) approach is presented and discussed. In each section, differences between healthy glandular, healthy stromal, and prostate cancer tissue are characterized and reported.