| dc.description.abstract | The goal of this work was to investigate and exploit the potential of chemometrics-assisted
energy dispersive X-ray fluorescence and scattering (EDXRFS) spectrometry towards rapid
diagnostics of cancer (and its severity) based on the correlative concentration, speciation and
multivariate alterations of trace elements in human body and animal model tissues. This is due to
the increasing realization that successful treatment of cancer depends on accurate diagnosis of
the disease at the local stage of development based on trace elements as biomarkers.
Utility of EDXRF spectroscopy to direct rapid trace elemental analysis in body tissues is
however at the moment limited by the complexity of the samples, extreme matrix effects and
problematic recovery of weak fluorescence signals from the normally enhanced spectral
background and overlapped fluorescence peaks. Further, analysis of elemental speciation, which
is essential in disease diagnostics via trace element biomarkers, is not possible in conventional
EDXRF spectroscopy. Energy dispersive X-ray fluorescence and scattering (EDXRFS)
spectrometry exploits both X-ray fluorescence and Compton scatter radiations obtained directly
and non-invasively from samples for correction of the matrix effects observed in the
deconvolution of the fluorescence intensities to concentration and thus has the potential to be
developed towards diagnostics of cancer at early stage of development. In this work, a method
has been developed for diagnosis of cancer based on the use of paraffin wax ‘standards’ spiked
with a multi-element stock solution of Fe, Cu, Mn, Zn and Se in the concentration range 11.5120
ppm, 5-32.5 ppm, 0.75-9 ppm, 5-195 ppm and 0.5-5.5 ppm respectively. Simulate tissue
samples containing Fe, Cu and Mn were also prepared in their respective oxidation states for
development of a protocol towards speciation analysis. Spectra were acquired using the
Shimadzu tube-excited EDX 800HS spectrometer for a live time of 50 s at 50 Kv and 1mA and
analyzed by exploiting multivariate chemometrics. PCA was used for reduction of spectral data
dimension and pattern recognition (classification). ANN and PCR were used to calibrate
strategies for direct quantitative analysis using fluorescence spectral regions of interest and
Compton scatter peaks. K-nearest neighbor (KNN) technique was used for quantitative
speciation analysis of Mn, Fe and Cu. The results underscore the role of trace elements in cancer
and indicate that Cu, Fe, Zn, Se and Mn can be used as trace biomarkers to identify the presence
and stage of cancer development in the human body. ANN calibration strategy gave the good
prediction results for trace element concentration analysis in tissues.
The indicative concentration ranges of Fe, Cu and Zn in cancerous tissues were 154.0±4.5191.2±9.5
ppm, 17.4±5.4-30.9±2.6 ppm and 74.5±26.4-103.2±3.8 ppm respectively with
quantifiable multivariate alterations observed in the concentrations of Mn and Se from early to
late stage of cancer genesis. The observed elevated concentrations of Fe and Zn in cancerous
tissues can be attributed to the significant role of these elements in cell growth and proliferation
as a result of rapid cell division in neoplastic tissues. Results obtained from cultured tissue
samples indicate that Fe increased gradually from 44.3 ± 4.6-169.7±1.1 ppm with staging of
cancer, which can be attributed to the role of Fe in blood supply to the neoplastic tissues. The
Pearson correlation coefficient between Fe and Zn was high (0.98) in cultured cancerous as
compared to 0.93 in cultured normal tissues, which can be linked to the increased metabolic
activity in the neoplastic tissues. The results of speciation analysis indicate the dominant role of
2
Fe , 2
Cu and 4
Mn during progression of cancer from early to late stage. The higher
speciation of Mn and Cu in cancerous tissues is probably due to the role of these elements in
fenton reaction where they generate free radicals essential for cancer carcinogenesis. Levels of
Cu, Fe, Se and Zn were thereafter used to successfully characterize real tissues as either
cancerous or non-cancerous whereas speciation levels of Mn, Fe and Cu were used to
successfully indicate the various stages of cancer. It may be concluded that chemometrics aided
EDXRFS spectrometry may be exploited towards developing a cancer diagnostic model for rapid
analysis of concentration levels and speciation alterations of trace elements (Fe, Cu, Zn, Se and
Mn) in human body tissues which have hereby been evaluated as suitable biomakers for
diagnosis of cancer (especially at the local stage of development). | en_US |
