496
S. Biswas et al.
tween the cancerous epithelium and surrounding stroma for understanding early
biomolecular events, which control the progression of the cancer, our group developed a 3d co-culture model from the mCF10A mammary acini model,along with
a stroma simulated by primary human mammary fibroblasts, and co-culturing the
stroma with mCF7 tumor breast epithelial cells [57]. Ft-IR Spectroscopy has also
been used to study the relationship between the chemical composition of microcalcifications in the breast tissue and the grade of breast cancer [58].
16.4.2 Prostate Cancer
As has been discussed in detail before, automated histopathology has been done on
IR images of prostate tissues sections, to classify the different cell types—in order
to detect as well as grade prostate cancer tissues [11, 59–61] and also to differentiate
between ‘confined’ and ‘locally invasive’ prostate cancer [62]. R Spectra of prostate
cancer tissue have been correlated with gleason score and the grade of tumour as
well [63]. Ft-IR spectroscopy has been applied to study the effect of certain potent anticancer drugs on prostate cancer cells, in terms of chemical and metabolic
changes to determine whether they would be effective, presenting an interesting adjunct set of data for future clinical protocols involving imaging [64–67]. our group
has also attempted to develop a completely automated histopathology technique for
distinguishing between malignant and normal tissue. Ft-IR spectroscopic images
were recorded from an unstained tissue micro-array and then pattern recognition
technique was used to identify epithelial cells without user input, by combining the
IR and optical data. morphological features, extracted by optimizing a minimumredundancy-maximal-relevance (mRmR) criterion and sequential floating forward
selection (SFFS), were applied to distinguish between malignant and normal tissues
and showed almost perfect accuracy [59] (Fig. 16.11).
While the presence of disease can be determined, extensions of these concepts to
detecting susceptibility to disease have also been carried out. For example, efforts
have been made to extract chemical information from the absorption spectra of
prostate adenocarcinoma tissues (CaP) from the two different populations–population with low susceptibility to CaP as India and population with high susceptibility to CaP as u.K., to recognize a susceptibility-to-adenocarcinoma biochemical
signature [68] and differences between benign and cancerous tissue have also been
demonstrated [69]. Post-diagnosis, the decision making in the clinic is often along
the lines of determining risk based on grade and stage. Principal component-discriminant function analysis (PC-dFA) algorithm was applied to IR microspectroscopic images for CaP grading in which, using typical gleason grading and the tumour/node/metastasis system, it was discovered, that the combined sensitivity and
specificity of the three-band gleason score criterion diagnosis was 92.3 and 99.4 %
respectively. Ft-IR spectroscopy along with histological stains have been used to
analyse the biomolecular structures of metastatic CaP cells in the bone marrow, to
understand the inherent biochemistry behind the proliferation of the cancerous cells
at the metastatic sites, to help produce more effective anticancer treatments [70].
S. Biswas et al.
tween the cancerous epithelium and surrounding stroma for understanding early
biomolecular events, which control the progression of the cancer, our group developed a 3d co-culture model from the mCF10A mammary acini model,along with
a stroma simulated by primary human mammary fibroblasts, and co-culturing the
stroma with mCF7 tumor breast epithelial cells [57]. Ft-IR Spectroscopy has also
been used to study the relationship between the chemical composition of microcalcifications in the breast tissue and the grade of breast cancer [58].
16.4.2 Prostate Cancer
As has been discussed in detail before, automated histopathology has been done on
IR images of prostate tissues sections, to classify the different cell types—in order
to detect as well as grade prostate cancer tissues [11, 59–61] and also to differentiate
between ‘confined’ and ‘locally invasive’ prostate cancer [62]. R Spectra of prostate
cancer tissue have been correlated with gleason score and the grade of tumour as
well [63]. Ft-IR spectroscopy has been applied to study the effect of certain potent anticancer drugs on prostate cancer cells, in terms of chemical and metabolic
changes to determine whether they would be effective, presenting an interesting adjunct set of data for future clinical protocols involving imaging [64–67]. our group
has also attempted to develop a completely automated histopathology technique for
distinguishing between malignant and normal tissue. Ft-IR spectroscopic images
were recorded from an unstained tissue micro-array and then pattern recognition
technique was used to identify epithelial cells without user input, by combining the
IR and optical data. morphological features, extracted by optimizing a minimumredundancy-maximal-relevance (mRmR) criterion and sequential floating forward
selection (SFFS), were applied to distinguish between malignant and normal tissues
and showed almost perfect accuracy [59] (Fig. 16.11).
While the presence of disease can be determined, extensions of these concepts to
detecting susceptibility to disease have also been carried out. For example, efforts
have been made to extract chemical information from the absorption spectra of
prostate adenocarcinoma tissues (CaP) from the two different populations–population with low susceptibility to CaP as India and population with high susceptibility to CaP as u.K., to recognize a susceptibility-to-adenocarcinoma biochemical
signature [68] and differences between benign and cancerous tissue have also been
demonstrated [69]. Post-diagnosis, the decision making in the clinic is often along
the lines of determining risk based on grade and stage. Principal component-discriminant function analysis (PC-dFA) algorithm was applied to IR microspectroscopic images for CaP grading in which, using typical gleason grading and the tumour/node/metastasis system, it was discovered, that the combined sensitivity and
specificity of the three-band gleason score criterion diagnosis was 92.3 and 99.4 %
respectively. Ft-IR spectroscopy along with histological stains have been used to
analyse the biomolecular structures of metastatic CaP cells in the bone marrow, to
understand the inherent biochemistry behind the proliferation of the cancerous cells
at the metastatic sites, to help produce more effective anticancer treatments [70].