Page 44 - Zoo Animal Learning and Training

36 Section I: Diagnostics and Planning

A B

Figure 4.8 Transverse T1‐weighted postcontrast MRI (A) and MRS spectra (TE = 144 ms) (B) of a crossbred dog with presumed glioma in the right frontal
lobe. The spectra was obtained within the intraaxial mass, clearly seen on the postcontrast image. The spectrum shows an abnormal lactate peak (upper
arrow), markedly reduced NAA peak (asterisk), and elevation of the choline peak (lower arrow), all typical of a high‐grade glioma.

spatial resolution (nominal voxel size, 0.45 cm ) MRS imaging to technique may help to increase the likelihood of a biopsy targeting
3
grade gliomas determined that there were significantly higher total the tissue with the highest probability of being malignant, and in
NAA values, lower Cho values, and lower Cho/NAA ratios for aiding inclusion of all neoplastic tissue while sparing normal tissue
grade II gliomas compared with grade III tumors [74]. This study during tumor resection. This same laboratory performed a follow‐
also determined that chemical shift imaging (CSI) spectra from up experiment using CNI to compare contrast‐enhancing versus
grade III astrocytomas had significantly higher maximum Cr values noncontrast‐enhancing areas associated with gliomas that were
compared with grade III oligodendrogliomas and grade III oligoas- hyperintense on T2‐weighted images [86]. This study confirmed
trocytomas in human patients. This higher maximum Cr concen- the original data in that the CNIs of biopsy samples not containing
tration in grade III astrocytomas compared with grade II tumors were significantly different from those obtained within
oligodendrogliomas is theorized as a result of the increased energy tumors; tumors were distinguished from nontumor with 0.96 sensi-
metabolism and increased growth rate of the high‐grade tumors tivity but only 0.57 specificity. When a CNI of 2.5 was used, the
[74]. The Cho/PCr‐Cr (phosphocreatine–creatine) ratio is also a sensitivity and specificity were 0.9 and 0.86, respectively [86]. Half
reliable marker for tumor grade identification. The Cho/PCr‐Cr of the 42 biopsy specimens from contrast‐enhancing tumors that
ratio was 2.05 ± 0.18 in low‐grade astrocytomas, 2.58 ± 0.11 in were positively identified as tumor with histopathology came from
grade III, and 5.1 ± 0.89 in grade IV [66]. These are typical exam- nonenhancing portions of the lesions. Additionally, 36–45% of the
ples, where MRS studies have used the ratios of NAA, Cho and Cr hyperintense lesions that did not enhance on contrast‐enhanced
to diagnose and differentiate brain lesions. A study by Vuori et al. T1‐weighted MRI were suspicious for tumor cells, based on the CSI
[84] showed that the total Cho and Cr values helped differentiate spectra and CNI of greater than 2.5 [86]. Unfortunately, histopa-
astrocytomas from oligodendrogliomas and oligoastrocytomas, but thology was not available for confirmation. This robust tool for dif-
that none of the metabolite ratios (NAA/Cho, NAA/Cr, Cho/Cr) ferentiating tumor versus peritumoral edema, necrosis, and normal
helped differentiate the tumor type. The use of metabolite ratios can tissue is valuable when using MRS with brain lesions, because of the
result in loss of metabolic information, since a ratio does not change high sensitivity to metabolic changes.
when neither the numerator nor the denominator change, and the Only a few reports have studied the use of MRS in canine brain
ratio does not reveal the direction of change. Furthermore, the tumors [62,87,88]. NAA and Cr were decreased and lactate was
ratios may not take into account regional and age‐dependent differ- increased in brain tumors (Figure 4.8) compared with the internal
ences in metabolite signals. A study by McKnight et al. [85] used control (normal contralateral side), but Cho exhibited no signifi-
three‐dimensional CSI to establish a linear model of total Cho ver- cant difference between neoplastic and normal tissues [62,87].
sus NAA, in the form of a Gaussian z‐score, the Cho to NAA index
(CNI) indicating the number of standard deviations of difference General Assessment of Images from CT and MRI
between amount of Cho and NAA at a voxel location and mean Once images have been evaluated for positioning and diagnostic
values of control spectra. Spectra from normal, edematous, and quality, they can be critically assessed. Interpretation of images
necrotic regions will have CNIs close to zero, whereas those from CT and MRI is similar to that for radiography and based on
recorded from tumor are elevated. A CNI of 2.0 was used as a classical Röentgen signs (size, shape, number, alignment, margination)
threshold to determine tumor versus peritumoral tissue [85]. This plus signal intensity or tissue attenuation [47]. The comparison of