Fig 1.
J3T1 cells were subcutaneously injected into the flank area of BALB/c nude mice. After two weeks, the mice were sacrificed to obtain viable tumour fragments. Preimplantation, dogs were treated with immunosuppressive agents for two weeks, and tumour fragments from the mice were injected into the right frontoparietal region of the brain parenchyma. After one week, postimplantation MRI with diffusion-weighted and perfusion-weighted imaging was performed every week until sacrifice. Immunosuppressive agents were also administered to the implanted dogs until sacrifice. Fifteen canine models were successfully created, and brain images from eight and seven subjects were obtained at 1 week and 2 weeks, respectively. After the follow-up imaging study, the brains were extracted for histological analysis.
Fig 2.
Flow chart of the quantitative image analysis.
Flow chart of the quantitative image analysis. A region of interest (ROI) was manually selected in each section of the enhancing lesions and semiautomatically coregistered with the relative cerebral blood volume map (rCBV) and coronal contrast-enhanced T1-weighted images. The volume of interest is determined by the summation of the ROIs from each slice; the rCBV values and apparent diffusion coefficient (ADC) values for the entire enhancing lesion are obtained. The relative perfusion values and ADC values in the normal-appearing white matter and cerebellum are also derived. The normalized perfusion values and ADC values were calculated by dividing each tumour and normal-appearing white matter by the corresponding cerebellar values.
Table 1.
Canine experimental glioblastoma multiforme model formation.
Table 2.
Serial follow-up imaging characteristics of the canine GBM model.
Table 3.
Imaging characteristics of the canine GBM model.
Fig 3.
Longitudinal follow-up imaging in a canine brain tumour model.
(A) After one week of surgery, a tumour was identified at the right frontoparietal area (arrow) with peripheral oedema (arrowhead) on the T2-weighted image. (B) The implanted glioblastoma multiforme (GBM) shows a large necrotic mass with peripheral enhancement similar to that seen in humans (arrow) on the contrast-enhanced T1-weighted image. (C-D) The mass shows diffusion restriction at the peripheral enhancing portion of the tumour (arrowhead) on the diffusion-weighted image (C) and apparent diffusion coefficient image (D). (E) There was also increased CBV (arrowhead) in the peripheral enhancing portion of the GBM on the dynamic susceptibility contrast-enhanced perfusion-weighted image. (F-G) On the two-week follow-up MRI, the peripheral enhancing lesion persists (arrow) with perilesional oedema (arrowhead) on the T2-weighted image (F) and contrast-enhanced T1-weighted image (G), and diffusion-weighted imaging does not show definite restriction in the peripheral portion (H-I). (J) However, there was still increased cerebral blood volume on the relative cerebral blood volume map (arrowhead) on the dynamic susceptibility contrast-enhanced perfusion-weighted image.
Fig 4.
(A) The gross specimen shows a necrotic mass at the right frontoparietal white matter, the tumour implantation site. (B) There were scattered peripherally located tumour cells (arrows) and central necrosis (stars) on Harris’ haematoxylin solution and eosin Y (H&E)-stained images. (C) Multiple gemistocytic tumour cells are located between the necrotic and inflammatory portions. (D) On high-magnification view, the presence of glioblastoma multiform (GBM) tumour cells (arrow) was confirmed. (E) The characteristic microvascular proliferation in GBM tissue is also demonstrated in micrographs.