ABSTRACT

Image-guided locoregional therapies are central to the management of liver cancer. Large-animal models are valuable preclinical tools for advancing interventional oncology because they allow testing with standard clinical imaging platforms and devices. The Oncopig^TM is a genetically engineered porcine model that enables site-specific tumor formation through inducible expression of KRAS^G12D and TP53^R167H following exposure to Cre recombinase. This pictorial essay illustrates the technical methodology for liver tumor induction in Oncopigs and demonstrates the angiographic use of tumor-bearing animals for evaluation of transarterial therapies. Liver tumors were induced using an ex vivo transgene activation approach in which liver biopsy specimens were incubated with an adenoviral vector expressing Cre recombinase and injected percutaneously into the liver under ultrasound guidance. Tumor development occurred in 34/37 (92%) induction sites across 17 Oncopigs and was confirmed by contrast-enhanced computed tomography. Hepatic arteriography using standard clinical devices enabled selective catheterization of tumor-feeding arteries. This pictorial essay provides a technical guide for investigators studying liver-directed therapies in a translational large-animal model.

Image-guided locoregional therapies (LRTs), including transarterial therapies, are a cornerstone of hepatocellular carcinoma (HCC) management.¹ Innovations in catheter technologies, embolic materials, pharmacological agents, and combinations with systemic therapies represent promising strategies to further improve patient outcomes.^{2, 3} Preclinical evaluation of these emerging approaches requires large-animal models that replicate human hepatic anatomy and are compatible with clinical interventional devices.⁴ Pigs represent a valuable translational model due to their similarity to humans in terms of anatomy, physiology, and drug metabolism.^{5, 6} The Oncopig^TM is a genetically engineered porcine model that enables site-specific tumor development through inducible expression of porcine KRAS^G12D and TP53^R167H following exposure to Cre recombinase.⁷ Liver tumor induction in Oncopigs has been described in several prior studies evaluating liver-directed LRTs.^8-11 This pictorial essay illustrates the workflow for liver tumor induction in Oncopigs and demonstrates angiographic utilization of tumor-bearing animals for transarterial therapy. This technical guide provides a practical reference for investigators seeking to evaluate LRTs in a clinically relevant large-animal model.

Study animals

The results reported in this essay are based on tumor inductions in 17 male Oncopigs (age 3.5–4 months). Oncopigs were obtained from the University of Illinois Urbana-Champaign. All procedures were approved by the Institutional Animal Care and Use Committee (IACUC) (protocol code: ACC #20-110). Trained veterinary staff supported daily animal maintenance; pre-, peri-, and post-procedural care; procedural anesthesia; and euthanasia.

Animal preparation

Tumor induction procedures were performed under general anesthesia with perioperative analgesia. The Oncopigs received intramuscular (IM) injections of 2.2–3.3 mg/kg of xylazine and 4.4–5.5 mg/kg of telazol for anesthetic induction. Other anesthetic regimens may also be used, in accordance with institutional veterinary and IACUC recommendations. For analgesia, 0.4 mg/kg of meloxicam was administered subcutaneously immediately before the procedure and orally once daily for 3 days thereafter. The animals were placed in left lateral recumbency, and the abdominal skin was shaved and sterilized (Figure 1).

Reagent preparation

The reagents required for Oncopig liver tumor induction include adenoviral vectors expressing Cre recombinase (AdCre) and a sterile 2-M calcium chloride solution in phosphate-buffered saline (PBS). Each liver biopsy specimen is incubated with 10⁹ plaque-forming units of AdCre. Adenoviral vectors can be produced in house¹² or obtained commercially. The tumor inductions described herein utilized AdCre under the control of a cytomegalovirus promoter (Ad5CMVCre; University of Iowa Viral Vector Core, Iowa City, IA, USA). Because repeated freeze/thaw cycles reduce viral vector titers, aliquoting the AdCre into small volumes upon arrival and storage at –80°C is recommended. In accordance with biosafety precautions, AdCre handling and aliquoting must be performed in a certified Class II biosafety cabinet.¹³ A sterile 2-M calcium chloride solution in PBS can be prepared in advance and stored at 4°C.

Shortly before the procedure, the required volume of AdCre was thawed on ice. Then, PBS containing calcium chloride was added to each tube containing AdCre to a total volume of 1 mL, and the solution was mixed gently by pipetting or inverting the tube. The solution was incubated for 15 minutes at room temperature before being combined with the liver biopsy specimen.

Liver tumor induction

Under ultrasound guidance, liver biopsy specimens were obtained from anesthetized Oncopigs using an 18-gauge core needle biopsy device (Figure 2). Multiple biopsy specimens can be obtained in the same procedure setting. Each specimen was placed in a tube containing AdCre solution and incubated for 30 minutes (Figure 3). During this time, gelatin sponge was cut into small pieces (2–3 mm²) and placed in a 3-mL syringe. Gelatin sponge serves as a scaffold to facilitate retention of the biopsy–AdCre suspension at the injection site and to reduce leakage into surrounding tissues or the peritoneal cavity following injection.⁸ Following incubation, the biopsy–AdCre mixture was aspirated into a 3-mL syringe and mixed with the gelatin sponge by repeatedly passing it through a three-way stopcock to macerate the tissue and create a homogenous mixture of approximately 1-mL total volume (Figure 4). The resulting slurry was then injected percutaneously into the liver of the same Oncopig using a 16-gauge needle under ultrasound guidance (Figure 5). Up to four sites can be inoculated in each Oncopig liver, with two in each liver lobe. Injecting multiple sites provides flexibility for targeting accessible tumors using LRTs, allows assessment of potential abscopal effects in untreated tumors, and ensures that the animal can still be used for therapeutic testing if tumor induction fails at one site. Injection sites should be well separated, safely accessible, and deep enough to prevent leakage of the injected material into the peritoneum. In the experiments described, 14/17 (82%) Oncopigs received two induction injections each, and 3/17 (18%) received three induction injections each.

Biosafety precautions

As with all experiments involving viral vectors, appropriate biosafety precautions must be taken when handling AdCre,¹³ and institutional biosafety committee approval must be obtained prior to initiating experiments. Adenoviruses are classified as Risk Group 2 agents, which are associated with human disease that is rarely serious and for which preventive or therapeutic interventions are often available. Transmission by direct contact, ingestion, aerosol exposure, or percutaneous injury may cause mild respiratory illness or conjunctivitis in healthy adults. Work involving adenoviruses generally requires Biosafety Level 2 (BSL-2) containment and procedures during preparation, manipulation, and injection. These precautions include restricted laboratory access, appropriate signage and laboratory setup, personnel training, decontamination of waste prior to disposal, and personal protective equipment to prevent skin or mucous membrane exposure.¹³ Personal protective equipment for in vivo tumor induction procedures includes an N-95 respirator, a face shield, gloves, a disposable coverall or suit, and shoe covers. A 0.5% sodium hypochlorite solution (10% bleach) should be used to decontaminate surfaces and materials potentially contaminated with the adenovirus, including pipette tips and tubes. Animals that receive adenovirus injections should be housed under animal BSL-2 containment for 72 hours, followed by reclassification to animal BSL-1 housing after a complete cage change.

Tumor imaging

Two weeks after tumor induction, the Oncopigs underwent contrast-enhanced abdominal computed tomography (CT) to confirm tumor development (Figure 6a–c). This imaging interval was selected based on prior studies⁸ and our experience demonstrating reliable tumor development within approximately 2 weeks following induction. The animals were sedated with IM injections of xylazine (2.2–3.3 mg/kg) and telazol (4.4–5.5 mg/kg) prior to imaging. Out of 37 tumor induction sites, a total of 34 tumors were detected by CT, with a mean tumor diameter of 1.8 cm (range, 1.2–2.4 cm). If CT imaging is not available, ultrasound may alternatively be used to confirm tumor development (Figure 6d).

Tumor angiographic utilization

The animals underwent hepatic arteriography in a fluoroscopy suite 3–5 days following imaging (Figure 7a–d). Following anesthetic induction with IM xylazine and telazol, the animals were endotracheally intubated and maintained under 1%–3% inhaled isoflurane anesthesia throughout the procedure. Arterial access was obtained via the common femoral artery with placement of a 5-French sheath. A 5-French catheter was used to access the celiac artery, and a 2.4-French microcatheter was then advanced into the segmental arterial supply of the hepatic tumor. Catheter position can be confirmed using cone-beam CT (Figure 7e). Subsequently, LRTs such as chemoembolization (Figure 7f) and radioembolization can be performed. Herein, all Oncopigs (n = 17) underwent transarterial chemoembolization (TACE) with doxorubicin. After completion of the procedure and removal of all devices, hemostasis at the arterial access site was achieved with 10 minutes of manual compression.

Tumor response to LRTs can be assessed using cross-sectional imaging, similar to clinical practice. Contrast-enhanced CT performed 7 days after TACE demonstrated the absence of contrast enhancement within the treated lesion in all animals, consistent with complete response (Figure 8).

Tumor harvest and characterization

The timing of euthanasia should be determined according to the experimental objectives. In this study, the Oncopigs were euthanized 4 days after liver angiography and TACE using a commercial euthanasia solution containing sodium pentobarbital and sodium phenytoin, administered intravenously at a dose of 1 mL per 10 lb of body weight. Following euthanasia, midline and transverse incisions were made to expose the abdominal cavity, and the liver was removed. Tumors were harvested for downstream analyses (Figure 9a). Tissue sections for histologic evaluation were fixed in 10% neutral buffered formalin, and samples intended for genomic and transcriptional analyses were flash-frozen and stored at –80°C. Tumor development occurred in 34/37 (92%) inoculation sites. Harvested tumors were evaluated histologically and demonstrated poorly differentiated carcinoma accompanied by immune cell infiltrates (Figure 9c, d).

Discussion

Large-animal models are important for preclinical evaluation of novel LRTs. Although mouse models of HCC are widely available,¹⁴ their small size precludes the ability to test LRTs. The Oncopig is a transgenic porcine model that enables temporal and site-specific tumor induction. This pictorial essay describes the development and angiographic utilization of Oncopig liver tumors using an ex vivo transgene activation approach (Figure 10).

Tumor induction success rates have been high across studies using this approach. In this cohort, tumors developed in 34/37 (92%) induction sites in male Oncopigs. Although only male Oncopigs were utilized in this study to reflect the epidemiologic male predominance of HCC, similar success rates were reported by Nurili et al.,⁸ who observed tumor formation in 44/48 (92%) sites across 18 female Oncopigs. These findings suggest that tumor induction is robust across sexes.

In this study, AdCre was used for tumor induction. The AdCre dose utilized in this study was selected based on the previously published approach described by Nurili et al.⁸ Adeno-associated virus (AAV) vectors may represent an alternative approach, as they are less immunogenic than adenoviruses and require only BSL-1 animal containment. However, AAV use in this model has not yet been evaluated.

An alternative method for developing liver tumors in Oncopigs involves in vitro transgene activation in isolated hepatocytes followed by autologous implantation of transformed cells.^{15, 16} Although this approach enables hepatocyte-specific transgene expression and facilitates in vitro gene editing of Oncopig HCC cells,¹⁷ it is more labor intensive, resource demanding, and time consuming than the ex vivo approach described here.

Although the described approach enables rapid and reproducible tumor induction, the model has several limitations. These include tumor regression, which has been reported in Oncopig liver tumors as well as other tumor types developed in the Oncopig model.^{8, 18} Tumor regression is thought to be related, at least in part, to anti-tumor immune responses characterized by cytotoxic T-cell and inflammatory immune cell infiltration.¹⁹ Consequently, this model may be best suited for short-term interventional studies performed shortly after tumor induction, prior to the onset of tumor regression. In the present study, angiographic procedures were performed approximately 2 weeks following tumor induction to maximize the likelihood of treating viable tumors. Additional longitudinal studies with serial imaging are needed to more fully characterize tumor growth kinetics and regression patterns in this model and to further inform optimal timing for therapeutic interventions.

Additional limitations include the absence of liver cirrhosis, a common comorbidity in patients with HCC,¹ and the fact that induced tumors in this model demonstrate poorly differentiated histologic features and may not fully recapitulate human HCC histopathology.^{8, 18} Given that porcine hepatic arterial anatomy differs slightly from human hepatic arterial anatomy,²⁰ procedural planning and catheterization approaches may need to be adapted. Nevertheless, the porcine hepatic vasculature is sufficiently comparable for translational angiographic and catheter-based interventional studies, and published studies support the applicability of human angiographic devices in the Oncopig.^8-11 Despite these limitations, the Oncopig model represents a valuable translational platform for evaluating liver-directed LRTs in a clinically relevant large-animal setting.¹⁸