Experimental animals and main reagents
This experiment was approved by the Committee of Animal Ethics of North Sichuan Medical College [approval number: Animal Ethics Review of North Sichuan Medical College (2023) 66]. The authors confirmed that the study is reported in accordance with ARRIVE guidelines and that all methods were performed in accordance with the relevant guidelines and regulations. The New Zealand White rabbits used for the experiments were supplied by the Centre of Experimental Animals in North Sichuan Medical College [license number: SYXK (Sichuan) 2018-076], 2.5–3.0 kg in weight, aged 4–6 months, and male only. The VX2 prototype rabbit was purchased from Hangzhou Huashu Biotechnology Co., Ltd. The rabbit VEGF enzyme-linked immunosorbent assay (ELISA) kits were purchased from Shanghai ZuoCai Biotechnology Co. Donafenib Tosilate Tablets were purchased from Zelgen Medicine Co., Ltd., Jiangsu, China.
The magnetic resonance (MR) imaging was performed using a 3.0 T magnetic resonance tomograph (Area, Siemens, Germany) with a dedicated animal coil; the scanning parameters are given in Table 1. The contrast medium was Gd-DTPA-BMA (Omniscan) at 0.3 mmol/kg, which was injected via the ear marginal vein at 1 ml/s with a high-pressure syringe. DCE-MRI was performed with a T1 twist sequence, in which 6 phases were acquired prior to contrast injection, and then 70 phases were scanned consecutively in dynamic images, taking 4.3 s per phase.
The computed tomography (CT) scanning device used was a Spiral 16-row CT apparatus (MX16, Philips, Netherland). The scan parameters were as follows: window width (WW) 250 HU, window location (WL) 35 HU, slice thickness 2.0 mm, kilovolt peak 120 kV, mAs 220, and pitch 0.984:1.
The MWA equipment used was a cooled-shaft microwave system (KY-2000, Kangyou Medical, Nanjing, China) and a 15G ablation needle antenna (KY-2450B-1, antenna length 18 cm, active tip length 1.1 cm).
The ELISA equipment was as follows: enzyme labeller (SpectraMAX Plus384 Meigu Molecular Instrument Co., Ltd.); constant-temperature water bath (HH-6 Shanghai Lichen Bangxi Instrument Technology Co., Ltd.); automatic plate washer (PW-480 Shenzhen Huisong Science and Technology Development Co., Ltd.); water purifier (SSY-II Sichuan ShuiSiYuan Environmental Protection Science and Technology Co., Ltd.); pipette gun (DaLong Xingchuang Experimental Instrument Co., Ltd.); and pipette tips (biosharp company).
The following surgical instruments were used: 20G coaxial puncture needle (MN2020, BARD, USA); 23G needle with a 5-ml syringe, 20G needle with a 10-ml syringe, 26G intravenous indwelling needle, gauze, pressure-sensitive adhesive tape, gelatin sponge, hair removal cream, scalpel, and forceps.
Transmission of VX2 tumors
After the VX2 generation rabbits were sacrificed, the posterior thigh skin was dissected to uncover the tumor tissues within the muscles and to observe the shape, size and necrosis of the lesion tissues. After stripping off the cancer mass (Fig. 1A), homogeneous fish-like tissues were removed, cut with a scalpel into tiny pieces, placed in normal saline, further crushed, milled and strained to make a suspension, which was subsequently loaded in a syringe (Fig. 1B). The posterior thigh was punctured in a normal New Zealand rabbit using a 20G coaxial puncture needle (Fig. 1C). After removing the needle stylet and aspirating with a syringe to confirm that there was no blood return, a small amount of suspension was dropped into the cannula with another syringe. The suspension was repeatedly pushed into the tissue with the needle stylet, a small amount of gelatin sponge was inserted to seal the needle channel, the puncture needle was quickly withdrawn, and local pressure was applied for 2–3 min to stop bleeding. Following the surgery, penicillin (400,000 U per day) was given intramuscularly for 3 continuous days to prevent infection. After 2–3 weeks, new VX2 tumors formed in the posterior thigh tissue, thus completing the transmission of the VX2 tumor.
CT-guided implantation of VX2 liver tumors
After 12 h of preoperative fasting without restriction of water intake, we weighed the normal New Zealand White rabbits. An indwelling needle was inserted into the ear marginal vein of each rabbit, 3% sodium pentobarbital (1 mL/kg) was injected for general anesthesia, and the rabbits were immobilized in the supine position on the experimental operating table. Preoperative CT scans revealed no discernible anatomical variations in the livers of any of the rabbits. After skin preparation, the plants were disinfected and wiped in the subxiphoidal area (Fig. 1D), we placed a 20G coaxial puncture needle along the predetermined puncture route. The preferred location for tumor implantation was the left lobe of the liver at the level of the upper gallbladder. Once the needle tip arrived at the predetermined position (Fig. 1E), aspirated with an empty syringe, confirming that there was no blood return, and then the syringe was replaced. A small amount of the suspension (prepared using the same method as in the transmission procedure) was injected into the needle cannula and driven to the liver by repeated insertion of the needle stylet. Then, a small amount of gelatin sponge was added to the needle cannula and pressed inside the liver to seal the needle channel. The puncture needle was swiftly removed, followed by the application of local pressure for 2–3 min to stop bleeding. After ensuring that no bleeding was occurring, the surgical towel was removed, and the wound was bandaged. CT scans were performed after surgery to detect the presence of significant liver injury. Penicillin (400,000 U per day) was given intramuscularly for 3 continuous days to prevent infection. Two weeks later, CT and MR scans were obtained to monitor the size of the implanted tumor (Fig. 1F). Single intrahepatic tumors between 1.5 and 2.5 cm in diameter were successfully created in rabbits as models of liver tumors, with no apparent tumors in the remainder of the liver, no apparent metastases in the neighboring or distal tissues, and no apparent serious complications such as ascites. Forty successfully modeled hormonal rabbits were divided into 3 groups according to the random number table method (insufficiently invasive group, n = 15; combined treatment group, n = 15; control group, n = 10).
Insufficient MWA was performed on rabbits assigned to the insufficient MWA group or the combined treatment group. After the rabbits were anesthetized, they were placed in the supine position on the surgical table. The location of the tumor was observed via MR (Fig. 2A,B) and CT before the operation; then, an experienced doctor in our department selected the appropriate percutaneous route. After preparing and disinfecting the skin at the puncture site, the ablation needle was gradually inserted following the predefined route. During CT-guided puncture, multiple scans were conducted to ensure that the needle tip followed a safe path and to minimize complications. The output power was 20–25 W, and the duration was typically 1–2 min (based on preliminary experiments, the maximum ablation diameter of a single ablation needle can reach approximately 1 cm to allow incomplete ablation) (Fig. 2C). During needle removal, heat was continuously applied for at least 5 s to prevent metastasis in the needle channel and bleeding. After the withdrawal of the needle, CT and MR scans were conducted to verify that the ablated area partially covered the tumor and that there were no serious complications (Fig. 2D–F).
DOLs were dissolved in normal saline to a concentration of 5 mg/mL. 15 mg/kg*d of the solution was given by gavage to each rabbit in the combined treatment group for 14 consecutive days after MWA11.
The experimental rabbits in each group underwent DCE-MRI scans before treatment (0 days) and at 1, 3, 7, and 14 days after treatment to examine residual tumor lesions (Fig. 3), the presence of local progression, distant metastasis, and complications. In this study, the ablation range of each lesion did not exceed half of the original lesion range, and the following conditions needed to be met during subsequent follow-up MR examinations. 1. Low T1-WI and high T2-WI (slightly higher) signals can be observed in the original lesion area prior to the enhancement scan. 2. Enhanced arterial phase scanning revealed enhanced nodules in the original lesion area. 3. Limited diffusion lesions can be observed in the original lesion area. Given that we used only low power, enhancing lesions in the arterial phase and diffusion restrictions in diffusion-weighted imaging (DWI) were observed at the edge of each ablation region, and all lesions were incompletely ablated to emulate the clinical situation of insufficient ablation.
DCE-MRI perfusion parameter extraction
Data processing was performed by Siemens Syngo. A postprocessing workstation was used. First, the images were motion calibrated to reduce the effect of respiratory motion and then combined with each sequence of images to determine the location of the tumor to place the volume of interest (VOI). The MR Tissue 4D Tofts model was selected to construct a pseudocolor map corresponding to the parameters of DCE-MRI, the solid portion of the tumor at the largest level was selected to outline the region of interest (ROI), avoiding blood vessels, calcifications, hemorrhages, and cystic degeneration areas, and the range of the ROI was 10–20 mm2. The DCE-MRI parameters [volume translate constant (Ktrans), reverse reflux rate constant (kep), and extravascular extracellular volume fraction (Ve)] were measured, and the tumor diameter was measured on T2W images at the same level. The positions at which each parameter was measured were kept as consistent as possible. Each parameter was measured three times, and the average value was taken as the final result.
Serum VEGF level measurement
Blood was drawn through the middle ear artery before treatment (0 days) and at 1, 3, 7, and 14 days after treatment. The drawn blood was centrifuged at low temperature, and the supernatant was aspirated and refrigerated for subsequent ELISA analysis. All the operations were carried out in strict accordance with the operating instructions of the kit. The absorbance (OD) was measured at 450 nm using an enzyme meter, and the concentration of the sample was calculated.
All the data were analyzed and plotted using SPSS 27.0 and GraphPad Prism 10.0. The serum VEGF levels, DCE-MRI parameters and tumor diameters at each time point were normally distributed and are expressed as the mean ± standard deviation. The correlation between each perfusion parameter of DCE-MRI and the corresponding serum VEGF levels at the same time points was assessed separately by Pearson correlation analysis. Two-way repeated-measures ANOVA was used to compare the differences in the serum VEGF concentration, DCE-MRI parameters and tumor diameter between the groups at different time points, and pairwise comparisons were made. In addition, when the data did not conform to Mauchly’s test of sphericity, corrections were made using the Greenhouse–Geisser test. P < 0.05 indicated statistical significance.