Vibrational Spectroscopy and Cancer - Vibrational Spectroscopy for Tissue Analysis - page 61

Adenocarcinoma
A pressure-tuning FTIR spectroscopic study of carcinogenesis in human
endometrium was reported by M. F. K. Fung et al. The spectra of normal
tissues differed from the obtained grade I and grade III adenocarcinoma.
Changes in the spectra of malignant samples were observed in the sym-
metric and asymmetric stretching bands of the phosphodiester backbones
of nucleic acids, the CH stretching region, the C-O stretching bands of the
C-OH groups of carbohydrates and cellular protein residuals, and the pres-
sure dependence of the CH 2 stretching mode. These spectral changes in the
endometrium were reproducible. It was also found for the first time that the
epithelium in the normal endometrium exhibits unique structural proper-
ties compared with the epithelium of other normal human tissues [57].
Krishna et al. reported on application of Raman spectroscopy for diagno-
sis of cancers in the human uterine cervix [58]. Raman spectra of normal and
malignant tissues were recorded in the fingerprint region. It was recognised
that the main spectral differences between the two groups of samples are
observed in the amide I, amide III, and 853 and 938 cm −1 , which can be attrib-
uted to structural proteins such as collagen. In order to develop an objective
discrimination method, the group carried out elaborate data analysis using
principal components analysis (PCA). Standard sets for normal and malig-
nant were prepared and tested retrospectively and prospectively. The results
produced very clean clustering of normal and malignant spectra, and sensi-
tivity and specificity of 99.5% were achieved.
Maheedhar et al. assessed the efficacy of conventional Raman spectros-
copy in differentiation of normal and malignant ovarian tissues [59]. A total
of 72 Spectra from eight normal and seven malignant ovarian tissues were
recorded by conventional NIR Raman spectroscopy (excitation wavelength
of 785 nm). Spectral data were analyzed by PCA and it was shown that
malignant spectra exhibit a broader amide I band, a stronger amide III band,
a minor blue shift in the delta CH 2 band, and a hump around 1480 cm −1 com-
pared to a normal spectrum. The normal spectra show relatively stronger
peaks around the 855 and 940 cm −1 region. It was demonstrated that conven-
tional Raman spectroscopy and the statistical methodologies in discrimina-
tion of normal from malignant ovarian tissues are efficient.
The purpose of Mo et al.'s study is to explore the ability of NIR Raman
spectroscopy in the high wavenumber region (2800-3700 cm −1 ) for the
in vivo detection of cervical precancer [60]. A rapid NIR Raman spectros-
copy system associated with a fibre-optic Raman probe was used for the in
vivo spectroscopic measurements. Multivariate statistical techniques includ-
ing PCA and linear discriminant analysis (LDA) were employed to develop
the diagnostic algorithm based on the spectral data from 2800-3700 cm −1 .
Classification result based on PCA-LDA showed that high-wavenumber NIR