An x-ray fluorescence system with a synchrotron radiation source was used to quantify the levels of iron, copper and zinc in breast tissue specimens. Healthy tissue and breast carcinomas were investigated as an aid to understanding the mechanisms of breast cancer and as a possible complementary diagnostic tool. Eighty samples were measured in total. Twenty samples were matched pairs, i.e. 20 tumour samples with 20 corresponding healthy tissue specimens taken at a distance from the tumour. The remaining 40 samples consisted of 20 excised tumours and 20 healthy specimens from breast reduction surgeries. The levels of the elements of interest were quantified via calibration models constructed using the XRF response from standard solutions. The statistical analysis of the results indicates elevation of the levels of all three trace elements in the tumours. The effect is more prominent for copper and zinc while the contrast between healthy and diseased tissue is enhanced when comparing the independent specimens rather than the paired samples. Specifically, the ratio of mean tumour to mean healthy concentration for iron was 1.6 for the paired samples and 2.7 for the non-paired samples. The ratios describing copper content were 3.1 for paired and 3.6 for non-paired samples while for zinc they were 2.4 and 4.4 respectively.

This paper presents improvements on a previously reported method for the measurement of elements in breast tissue specimens (Geraki et al 2002 Phys. Med. Biol. 47 2327-39). A synchrotron-based system was used for the detection of the x-ray fluorescence (XRF) emitted from iron, copper, zinc and potassium in breast tissue specimens, healthy and cancerous. Calibration models resulting from the irradiation of standard aqueous solutions were used for the quantification of the elements. The present developments concentrate on increasing the convergence between the tissue samples and the calibration models, therefore improving accuracy. For this purpose the composition of the samples in terms of adipose and fibrous tissue was evaluated, using an energy dispersive x-ray diffraction (EDXRD) system. The relationships between the attenuation and scatter properties of the two tissue components and water were determined through Monte Carlo simulations. The results from the simulations and the EDXRD measurements allowed the XRF data from each specimen to be corrected according to its composition. The statistical analysis of the elemental concentrations of the different groups of specimens reveals that all four elements are found in elevated levels in the tumour specimens. The increase is less pronounced for iron and copper and most for potassium and zinc. Other observed features include the substantial degree of inhomogeneity of elemental distributions within the volume of the specimens, varying between 4% and 36% of the mean, depending on the element and the type of the sample. The accuracy of the technique, based on the measurement of a standard reference material, proved to be between 3% and 22% depending on the element, which presents only a marginal improvement (1%-3%) compared to the accuracy of the previously reported results. The measurement precision was between 1% and 9% while the calculated uncertainties on the final elemental concentrations ranged between 10% and 16%.

No abstract

Tumourigenesis in experimental models is associated with the formation of new blood vessels (angiogenesis). Recent studies have suggested that tumour angiogenic activity may be inferred in histological sections by measuring the density of the vasculature. The purpose of this study was to determine whether the transition from normal to dysplastic and neoplastic tissue in the oral mucosa is accompanied by quantitative or qualitative changes in the vascularity of the tissue, and how the estimate of vascularity is influenced by the vessel marker and method of assessment. A total of 100 specimens of normal oral mucosa, dysplastic lesions, and squamous cell carcinomas were examined. Sections were immunostained with the pan‐endothelial antibodies to von Willebrand Factor (vWF) and CD31, or with an antibody to the αvβ3 integrin, previously reported to be a marker of angiogenic vessels. Vascularity was quantitated by two different methods: highest microvascular density (h‐MVD) and microvascular volume, as determined by point counting (MVV). The results showed that vascularity, measured by the MVV method using antibodies to either vWF or CD31, increased significantly (P<0·0001) with disease progression from normal oral mucosa, through mild, moderate, and severe dysplasia to early and late carcinoma (76 paraffin‐embedded tissues examined). In contrast, h‐MVD did not discriminate between dysplastic lesions and carcinoma. A similar percentage of the total vessel volume (MVV) and density (h‐MVD) were positive for αvβ3 in 24 frozen tissues examined, including normal oral mucosa. It is concluded that there is a close association between vascularity and tumour progression in the oral mucosa. Morphometric analysis reflecting microvascular volume is more informative than the currently popular analysis of microvascular density. The expression of αvβ3 in the vasculature of oral tissues does not necessarily reflect the presence of angiogenic vessels. © 1997 by John Wiley & Sons, Ltd.