Page 185 - Read Online

Costa et al. Hepatoma Res 2018;4:35 I http://dx.doi.org/10.20517/2394-5079.2018.06 Page 3 of 11

Thirty male Sprague-Dawley rats weighing 300-400 g at 8 weeks of age were used. HCC secondary to NAFLD
was induced by high fat and choline deficient diets [35% of total fat, enriched with 54% of trans-fatty acids
(Rhoster Ltd., BR)], with a DEN (Sigma-Aldrich Chemical, St. Louis, MO, USA) dose of 100 mg/L in the
drinking water ad libitum for 16 weeks. After this period stimuli were suspended and the animals were
randomly split into 2 groups. The control group (n = 10) received 1 mL of saline solution (0.9%) daily by
gavage for 3 more weeks. The sorafenib group (n = 20) received 5 mg/kg/day of sorafenib (Bayer Healthcare
Pharmaceuticals, Cologne, GY) by gavage for 3 more weeks.

At 16 weeks, rats were anesthetized with ketamine 80 mg/kg (Cristalia, BR) and xylazine 10 mg/kg (Bayer,
BR) intraperitoneally and then submitted to abdominal ultrasound (US) to quantify, measure, and localize
hepatic nodules . Board-certificated radiologist (M.C.C) with 24 years of experience performed the
[31]
procedure using a Philips Ultrasound IU 22 system (Bothell, WA, USA) with a VL13-5 transducer. Only
nodules larger than 0.2 cm were considered and catalogued in the distribution of parenchyma: left lobe, right
lobe, medium lobe, and caudate lobe.

PET images with [ F]FDG were acquired at the end of treatment (3 weeks after initial sorafenib) in a small
18
animal PET scanner (LabPET4 Gamma Medica-Ideas, Northridge, CA, USA), using methods similar to
those outlined in Park et al. . Animals were anesthetized with isoflurane 5% in oxygen 100% for induction
[15]
and 2%-3% for maintenance. [ F]FDG was injected in the penile vein (37.7 ± 6.29 MBq) and the animals were
18
allowed to wake up after injection for better tracer distribution. After 45 min of tracer injection the animals
were anesthetized again and the image acquired for 30 min. Computed tomography (CT) images were also
obtained with 65 kVp, 165 μA in 512 projections and magnification of 1.3 for anatomic correlation.

PET images were reconstructed by the ordered subsets expectation-maximization 3D (OSEM-3D) method
[32]
with 20 interactions, 4 subsets, a transverse field of view of 100 mm, and a matrix of 240 × 240, for pixel
resolution of 0.42 mm × 0.42 mm. The CT images were reconstructed with a filtered back projection method,
a matrix of 512 × 512, and pixel resolution of 0.17 mm × 0.17 mm. CT images were used for attenuation
correction of the PET images.

Images were analyzed by the PMOD™ software (PMOD Technologies Ltd., Zurich, CH) to obtain quantitative
18
measures of [ F]FDG uptake in defined regions of interest in the rats in each group. The visualization interface
in the software allowed regions of interest (ROI) to be drawn specifically and entirely within tumor lesions, liver
tissue and muscle. The uptake values were expressed as using the dose-normalized parameter standardized
uptake value (SUV). SUV = radioactivity concentration (kBq/mL)/[injected dose (kBq)/animal weight (g)].
The maximum value of the SUV within a region of interest is expressed as SUVmax.

Three days after the PET scan (week 19), the rats were euthanized with dextroketamine (Cristalia, BR) 120 mg/kg
and xylazine (Bayer, BR) 10 mg/kg intraperitoneally. Liver samples of the right and left lobes and the larger
tumors evidenced in PET/CT were collected for histological analysis. The liver specimens were fixed in 4%
formaldehyde and stained with hematoxylin-eosin (HE). These samples were blindly scored by a veterinary
hepatopathologist (B.C) with 12 years of experience, using a modified classification standardized by
Kleiner et al. . The variables analyzed were steatosis (0-3), lobular inflammatory changes (0-3), hepatocyte
[33]
ballooning (0-2), fibrosis (0-4), and ductular reaction (0-3) through the NALFD activity score (NAS) . HCC
[33]
was diagnosed with characteristics defined by Thoolen et al. for rats and then classified by Schlageter et al. .
[34]
[35]
Histological classifications are considered the gold standard for assessment of HCC and are used in this study
to evaluate the performance of the non-invasive characterization using F-FDG PET.
18
Immunohistochemical analysis was performed for protein glutamine synthetase (GS), hepatocyte specific
antigen (HEP-PAR-1), and cytokeratin 19 (CK-19). HCC was considered for nodules with positive results for