Breast cancer and non-malignant breast data: Quantification by FPT - Signal Processing in MRS with Biomedical Applications - page 310

broadenoma compared to the noninfiltrated tissue of that patient. Moreover,
most of the computed metabolite concentrations were at least one standard de
viation greater than the mean values for normal breast tissue. In contrast, the
computed concentration of myoinositol was nearly the same, i.e., 0.465 M/g
ww and 0.448 M/g ww for the fibroadenoma and for the noninfiltrated tis
sue, respectively, of the same patient and showed the lowest difference from
the mean for normal breast tissue (+ 0.52 M/g ww of the SD) [22]. These
analyses confirm that a rich “window” of information is provided by in vitro
MRS study of metabolite concentrations in malignant versus noncancerous
breast tissue.
Corroboration of some of the above findings from Refs. [21, 22] is provided
by Sharma et al. [398] who used in vitro 2D MRS to compare 11 involved and
12 uninvolved lymph nodes from patients with breast cancer. They reported
that the concentrations of PCho and GPC were significantly higher in involved
compared to noninvolved nodes. This was attributed to increased membrane
synthesis in cancer cells, suggesting that metastatic breast cancer cells were
present in the lymph nodes. There was also a highly significant difference
between the lactate concentrations in involved and noninvolved nodes [398].
The high levels of lactate indicate the presence of cancer cells whose energy
source is from the anaerobic glycolytic pathway.
There is also support from animal model studies of breast cancer concern
ing the importance of assessing the rate of glycolysis and lactate clearance
with respect to the diagnosis and prognosis of breast cancer [399]. Thus far,
however, neither onedimensional nor twodimensional clinical in vivo MRS
has included lactate as a metabolic marker of breast cancer. In our anal
yses of the data of Ref. [395] a number of the metabolite concentrations in
the malignant tissues showed significant correlation. However, alanine was not
correlated with phosphoethanolamine or with glycerophosphocholine, nor was
choline concentration correlated with those of several other metabolites in the
malignant tissues. Furthermore, principle components analysis revealed that
those metabolites with the strongest diagnostic accuracy did not consistently
load with the others.
In addition, we also analyzed the phosphocholine to glycerophosphocholine
ratio for the data from Ref. [395]. A significantly higher PCho/GPC ratio
was found in the breast cancer samples compared to the normal tissue from
the same patient. The breast cancer samples showed a significantly higher
PCho/GPC ratio compared to the normal, noninfiltrated tissue [25]. Our
analyses [25] corroborate human breast cell line research, indicating that ma
lignant transformation is associated with a “glycerophosphocholine to phos
phocholine switch” [307].
This is related to overexpression of the enzyme choline kinase, which is
responsible for phosphocholine synthesis [308, 397]. It also reflects altered
membrane choline phospholipid metabolism. The major steps in choline
metabolism in mammalian cells are through the cytosine diphosphate (CDP)
choline pathway [308, 309, 400]. It should be noted that therein are several