447
15 FtIR Imaging of tissues: techniques and methods of Analysis
environment is under review. At present all commercial hPv tests in diagnostic use
rely on the detection of hPv nucleic acids in clinical specimens.
the first report highlighting the potential of Ft-IR microspectroscopy as a promising tool for screening for cervical malignancies was provided by Wong et al.,
who have showed reproducible spectral differences between normal and malignant
cervical epithelial cells [91]. A similar classification has been shown by using PCA
decomposition of IR spectra from 272 patients with normal cervical cells and those
with varying degrees of dysplasia and hPv-infected cells [92]. this discrimination reveals characteristic spectral features of the cervical cells. Namely, spectra of
the normal tissue show high absorbance of the glycogen band and high ν s (Po 2
−
)/
ν as (Po 2
−
) ratio on the contrary to smear classified as abnormal epithelial cells. In
the latter, the ν as (Po 2
−
) band becomes more intense and broader due to increasing
content of nucleic acids. Changes in intensity of bands in the high-wavenumber
region was observed when applying pressure-tuning FtIR indicating the disorder
of methylene chains in membrane lipids of cervical neoplastic cells [92]. the PCA
classification was supported by histological results (Pap smear and biopsy), showing 91 and 87 % correlation for normal and cancer diagnosis, respectively [92].
Another IR biomarker has been proposed by mordechai et al., who have studied
formalin-fixed tissue from 20 patients with diagnosed histopathologically invasive cervical carcinoma [93]. the RNA/dNA ratios calculated for the 1121 and
1020 cm
−1
bands are higher in malignant areas than in normal areas of the tissue,
possibly indicating higher transcription during carcinogenesis. By using this parameter in conjunction with identified earlier changes in band intensity of glycogen
and phosphate groups, two distinct clusters normal and cancerous cervical tissues
were reported [93]. however, further studies have demonstrated that the observed
spectral changes might not be strictly correlated with the presence of dysplastic
cells, but can result from differentiation, maturation, a stage of the menstrual cycle, contamination of mucin, blood cells, and other factors [94–98]. Consequently,
these changes can also cause non-specific depletion of glycogen and variation of
phosphate group bands. FtIR mapping [99, 100] and imaging [101] of the cervical
tissue have given a new insight into overall biochemical characterisation of histopathologically normal and cancerous tissues delivered from biopsies of the same
patients. the FtIR mapping of normal and squamous CIS cervical tissues allowed
for their discrimination within the basal layer which is important as this region is
where cancer cells often originate [99]. the basal layer of the both normal and CIS
biopsies was identified by the presence of sharp bands at 1204 and 1283 cm
−1
. Apart
from those bands, a high-intensity band at about 1155 cm
−1
was found in the normal
basal tissue, whereas a broad band near 1240 cm
−1
became more prominent in the
precancerous tissue. thus, classification of normal and cancerous cervical tissues
relied on the ratio of the 1155 and 1240 cm
−1
bands. this ratio was 1.40 ± 0.91 and
0.29 ± 0.19 for the normal and CIS biopsies, respectively [99]. In turn, Wood et al.
have demonstrated spectral characteristics of cervical sections classified to mild and
severe dysplasia by using FtIR mapping of 52 different regions of tissue from 10
different patients [100]. the examination of such a huge data bank was performed
with the use of multivariate unsupervised hierarchical Cluster method (uhCA).
15 FtIR Imaging of tissues: techniques and methods of Analysis
environment is under review. At present all commercial hPv tests in diagnostic use
rely on the detection of hPv nucleic acids in clinical specimens.
the first report highlighting the potential of Ft-IR microspectroscopy as a promising tool for screening for cervical malignancies was provided by Wong et al.,
who have showed reproducible spectral differences between normal and malignant
cervical epithelial cells [91]. A similar classification has been shown by using PCA
decomposition of IR spectra from 272 patients with normal cervical cells and those
with varying degrees of dysplasia and hPv-infected cells [92]. this discrimination reveals characteristic spectral features of the cervical cells. Namely, spectra of
the normal tissue show high absorbance of the glycogen band and high ν s (Po 2
−
)/
ν as (Po 2
−
) ratio on the contrary to smear classified as abnormal epithelial cells. In
the latter, the ν as (Po 2
−
) band becomes more intense and broader due to increasing
content of nucleic acids. Changes in intensity of bands in the high-wavenumber
region was observed when applying pressure-tuning FtIR indicating the disorder
of methylene chains in membrane lipids of cervical neoplastic cells [92]. the PCA
classification was supported by histological results (Pap smear and biopsy), showing 91 and 87 % correlation for normal and cancer diagnosis, respectively [92].
Another IR biomarker has been proposed by mordechai et al., who have studied
formalin-fixed tissue from 20 patients with diagnosed histopathologically invasive cervical carcinoma [93]. the RNA/dNA ratios calculated for the 1121 and
1020 cm
−1
bands are higher in malignant areas than in normal areas of the tissue,
possibly indicating higher transcription during carcinogenesis. By using this parameter in conjunction with identified earlier changes in band intensity of glycogen
and phosphate groups, two distinct clusters normal and cancerous cervical tissues
were reported [93]. however, further studies have demonstrated that the observed
spectral changes might not be strictly correlated with the presence of dysplastic
cells, but can result from differentiation, maturation, a stage of the menstrual cycle, contamination of mucin, blood cells, and other factors [94–98]. Consequently,
these changes can also cause non-specific depletion of glycogen and variation of
phosphate group bands. FtIR mapping [99, 100] and imaging [101] of the cervical
tissue have given a new insight into overall biochemical characterisation of histopathologically normal and cancerous tissues delivered from biopsies of the same
patients. the FtIR mapping of normal and squamous CIS cervical tissues allowed
for their discrimination within the basal layer which is important as this region is
where cancer cells often originate [99]. the basal layer of the both normal and CIS
biopsies was identified by the presence of sharp bands at 1204 and 1283 cm
−1
. Apart
from those bands, a high-intensity band at about 1155 cm
−1
was found in the normal
basal tissue, whereas a broad band near 1240 cm
−1
became more prominent in the
precancerous tissue. thus, classification of normal and cancerous cervical tissues
relied on the ratio of the 1155 and 1240 cm
−1
bands. this ratio was 1.40 ± 0.91 and
0.29 ± 0.19 for the normal and CIS biopsies, respectively [99]. In turn, Wood et al.
have demonstrated spectral characteristics of cervical sections classified to mild and
severe dysplasia by using FtIR mapping of 52 different regions of tissue from 10
different patients [100]. the examination of such a huge data bank was performed
with the use of multivariate unsupervised hierarchical Cluster method (uhCA).