| dc.description.abstract | Breast cancer is the second most common cause of death from cancer in women, accounting for more than 1 million deaths globally per year. Current detection is based on X-ray mammographic screening, which involves the use of ionising radiation with potentially detrimental effects, or MRI scans, which have limited spatial resolution. The presence of microcalcifications in breast tissue has been associated with malignant disease. Unfortunately, X-ray mammography and MRI scanning techniques are not able to discriminate between microcalcifications from a benign lesion and those from a malignant lesion. The aim of this project was to use optical techniques based on vibrational spectroscopy, such as Fourier Transform Infrared (FTIR) absorption and Raman scattering, which are non-destructive, label-free and chemically specific, to investigate the composition of microcalcifications in breast tissue for augmented diagnostics and improved outcome for the patient. This work involved the characterisation of mineral standards of the type that can be found in the breast, in order to identify the precise composition of the microcalcifications. A series of calcium hydroxyapatite (Hap) compounds was used for calibration of the micro-FTIR and Raman spectra. The ratio of carbonate-to-phosphate band intensity for each individual Hap powder was determined and the data were used to assess the level of carbonate substitution in each breast tissue biopsy. In parallel, the analysis of potential precursor mineral phases (namely octacalcium phosphate and amorphous calcium phosphate) revealed similar features to Hap in both FTIR and Raman spectra, which can be translated to the biopsy samples. The accessibility to diverse panels of breast tissue sections (frozen and paraffin-embedded) was a great opportunity to test different approaches. A deparaffinisation protocol was applied to a set of samples for Raman analysis and the process was found not to affect the microcalcification composition. The FTIR analysis of the frozen tissues provided information on the carbonate peak in the short wavelength range (1500-1400 cm-1), which normally contains a strong contribution from paraffin in standard histological specimens. The study of breast tissue sections showed the heterogeneity in composition of microcalcifications between different samples from the same stage of pathology in terms of protein, lipid - which is usually not observed in formalin-fixed paraffin-preserved (FFPE) sections - and carbonate content. The mineralisation of the MDA-MB-231 breast cell line induced by two osteogenic agents: inorganic phosphate (Pi) and -Glycerophosphate (G) was investigated using Raman micro-spectroscopy. The uptake of osteogenic agent induced a faster mineralisation for cells cultured with a medium supplemented in Pi (day 3) than G (day 11). A shift (± 3 cm-1) of the phosphate peak at 956 cm-1 in the Raman spectra was apparent when the culture medium was supplemented with G, indicating the presence of precursor phase (octacalcium phosphate) during Hap crystal formation. New IR technologies such as bright laser sources e.g. quantum cascade laser (QCL) open possibilities for the analysis of biological samples. They allowed us to achieve a better signal-to-noise ratio than Globar thermal sources used in traditional FTIR systems, particularly on optically dense samples such as calcifications. The ability of selecting specific incident wavelengths allows significant improvements in the acquisition time. This work illustrates for the first time the identification of microcalcifications using a QCL source in the long wavelength range coupled to an upconversion system with a silicon detector for efficient sensing. The upconverted images showed a good agreement with the micro-FTIR images. Vibrational spectroscopy has been shown to be a powerful tool for discrimination of mineral species in breast calcification. These techniques can provide complementary information for the pathologist to be able to classify breast pathologies - benign, ductal carcinoma in situ (DCIS) and invasive cancer - with higher accuracy. | en_GB |
