Percutaneous microwave ablation of a transgenic large animal porcine liver tumor model after intra-arterial embolization

INTRODUCTION

Primary hepatic tumors (HT), most commonly hepatocellular carcinoma (HCC) are the fourth leading cause of cancer mortality worldwide, resulting in about 782,000 deaths/year.^[1] Secondary metastatic neoplasms commonly appear in the liver. Colorectal cancer (CRC) is the most prevalent secondary, metastatic cancer to the liver, with nearly half of the patients diagnosed eventually developing metastasis.^[2] Secondary spread to the liver of neuroendocrine, breast, and melanoma is also common.^[3-5]

For patients with limited hepatic disease, surgical resection (SR) is the gold-standard treatment, improving overall survival.^[6] However, only 10–20% of patients with HT are surgical candidates due to comorbidities precluding SR such as underlying hepatic dysfunction, portal hypertension, or cardiopulmonary disease.^[7] Other factors precluding SR are tumor extent and location.^[8] Percutaneous ablation (PA) has been established as a comparable minimally invasive local therapy with less complications and faster recovery compared to SR for lesions up to 3 cm.^[9,10] A critical limitation to PA has been local tumor progression due to incomplete ablation zones (AZs) in larger HT. A clinical need exists to permit the successful treatment of larger HT with appropriate margins.

Radiofrequency ablation (RFA) induces thermal damage through high frequency alternating current, causing ionic oscillation and frictional heating. Constraints of RFA include a limited active heating zone measuring only a few millimeters, heat sink effect from adjacent vessels, and electrical impedance from tissue desiccation which limits thermal transmission.^[11] Microwave ablation (MWA) has been created to overcome these disadvantages. Compared to RFA, the electromagnetic field of MWA creates higher temperature, improving thermal conductivity and thus tissue penetration. MWA is also less affected by tissue impedance, carbonization, and heat sink.^[12,13] The addition of embolization in combination with MWA has been proposed as a means to potentiate PA by improving heat conductivity and counteracting the heat sink effect to create a larger and more homogeneous AZ suitable for treatment of HT larger than 4 cm. Several retrospective clinical evaluations of the combined therapy in small clinical cohorts have exhibited these principles and clinical benefit in patients.^[14-16] At present, limited preclinical data are available to assist clinicians in optimizing this combination therapy.

Employing a large animal transgenic oncopig model of a HT that has previously been shown to express vascular perfusion similar to HCC and CRC, this manuscript presents an evaluation of how different embolization techniques before MWA can affect the size of the AZ.^[17]

MATERIAL AND METHODS

The Institutional Animal Care and Use Committee approved all research procedures 2022-103088-USDA. Eight 9-week-old Oncopigs (transgenic pigs with Cre-inducible TP53R167H and KRASG12D mutations) with a mean body weight of 25 kg were obtained from Sus clinicals Inc. (Chicago, IL). The animals were allowed to acclimate to the animal facility for 5 days. Before any procedures, the animals were fasted for 12 h. Each pig was sedated with an intramuscular injection of solution containing ketamine hydrochloride, acepromazine, and atropine sulfate. Once the pig was anesthetized, an endotracheal tube was inserted, and anesthesia was maintained with isoflurane, nitrous oxide, and oxygen.

MVA

Simulating clinical PA, US guidance identified the HT and a safe percutaneous path. MWA was performed using the 2.45-GHz NEUWAVE System and a single 15-gauge NEUWAVE PRXT antenna (NeuWave Medical, Inc., Madison, WI). Control group underwent MWA without prior embolization. Experimental groups underwent MWA immediately after embolization. In all animals, MWA was performed for 4 min with the power of 65 watts. In total, 20 tumor sites were treated. At the end of all ablation procedures, the animals were euthanized with an intravenous sodium pentobarbital solution without recovery from anesthesia 30 min after the MWA.

RESULTS

Percutaneous MVA

All ablations were successful, and no adverse events related to arterial access, embolization, or ablation requiring secondary restorative procedure or termination of the experiment were encountered during or immediately after the procedure, and the presence of these was monitored by the proceduralist physician and veterinary staff [Figure 1].

Close

Figure 1:

Liver tumor ablation in transgenic oncopig after embolization. (a) Ultrasound of the liver tumor (red arrow) Tumor is measured with calipers on Ultrasound Image. (b) Angiography of liver segment containing the liver tumor before embolization. (c) Post-embolization angiography with thin black arrowing showing tumor blush and thick black arrow identifying stasis of vascular flow within the tumor. (d) Ex vivo evaluation of the ablation zone after embolization. Black arrow is the non-charted coagulative zone, black triangle is the charred central portion, and the red arrow is the ablated tumor.

Export to PPT

The volume (mean ± standard deviation) of the AZ after distal embolization with particle sizes of 40 µm was 17.48 cm³ ± 1.22 cm³ or 65% larger than the control (6.06 cm³ ± 2.02 cm³); 100 µm, 14.81 cm³ ± 0.43 cm³, 59% larger than control, and with 300–500 µm particles, 12.16 cm³ ± 0.8 cm³ or 50% larger than control group [Table 1]. The MWA volume of the 40 µm when compared to the AZ after embolization with 900–1200 µm and lipiodol AZ was significantly larger (P ≤ 0.001 and <0.001, respectively).

Table 1: Change in ablation zone size after pre-embolization with various material

Group	Ablation Zone (cm^3)		Length (cm)		Length (cm)
	Volume	p=	Long axis	p=	Short axis	p=
Control (n=6)	6.06±2.02		3.56±0.17		2.29±0.19
40 µm (n=8)	17.48±1.22	<0.001	4.5±0.47	<0.001	3.23±0.53	<0.001
100 µm (n=4)	14.81±0.43	<0.001	4.67±0.05	<0.001	3.17±0.11	<0.001
500 µm (n=3)	12.16±0.8	<0.001	4.5±0.1	<0.001	3.6±0.1	<0.001
1200 µm (n=4)	9.28±3.29	0.287	3.07±0.59	0.282	2.43±0.51	0.671
Lipiodol (n=3)	5.79±1.2	0.547	3.3±0.17	0.101	2.77±0.5	0.236
Manufacturer Insert	12.1		3.7		2.5

The largest particles which provide a more proximal embolization and the liquid embolic did not create an AZ that was significantly larger than the control (900–1200 µm; 9.28 cm³ ± 3.29 cm³ and lipiodol 5.79 cm³ ± 1.2 cm³, P = 0.287, P = 0.547).

The AZ of the oncopig in vivo control group, without embolization, was found to be significantly smaller (P ≤ 0.001) when compared to the proposed AZ from the NeuWave probe manufacturers’ product material for 4 min at 65 watts (12.1 cm³).

The AZ diameter along the long access was also significantly longer after embolization with 40 µm, 100 µm, and 300–500 µm particles, 4.5 cm ± 0.47 cm, 4.68 cm ± 0.06 cm, and 4.5 cm ± 0.1 cm (P ≤ 0.001, <0.001, <0.001), respectively, compared to the length in the control group (3.56 cm ± 0.17 cm).

DISCUSSION

PA is an accepted local therapy for HT and can be employed with either curative intent or as palliative therapy.^[6,18] A major limitation has been the maximum size of HT that can be completely treated, with studies reporting reliable local cure can only be achieved in HT <3 cm.^[19-22] With all available thermal ablation techniques, the maximal size of the AZ that can be achieved is limited by the ability to conduct heat through the tumor and in adjacent liver tissue.

The present study, for the 1^st time, attempted to combine transcatheter arterial embolization with MWA in a large animal tumor model, to specifically assess how the embolization particle size changes the AZ. A single ablation protocol of 65 watts and 4 min was performed in all HT independent of HT size; this protocol was chosen to compare the AZ size to the manufactures’ estimation; ensure the AZ would remain entirely within the hepatic tissue for the most accurate measurements; for an “apples to apples” comparison of the changes in the AZ for the different groups. Data obtained from HT embolization with particle size between 40 µm and 500 µm resulted in a significant increase in the size of the AZ when compared to MWA alone or to large particle embolization that generates a proximal, central blockage of the larger HA. This supports the hypothesis that distal embolization of the HT, at the tumor arterioles with smaller particles, results in an AZ volume, length, and width that is larger. Distal embolization results in the superior alleviation of the detrimental effects of heat sink at the periphery of the AZ, likely related to an inability of surrounding vessels within the adjacent hepatic segment to be recruited for passive dissipation of heat and energy conduction. When larger particles are used, a more proximal embolization occurs, similar to central arterial ligation, permitting hepatic segmental collateral vessels to be engaged, functioning to conduct heat away from the ablation probe. The size of the AZ volume after embolization with lipiodol is similar to no embolization, suggesting embolization with the liquid embolic lipiodol alone permits residual blood flow in the, incompletely obstructing the flow of blood through the HA when compared with particles.

Animal models are essential to the study of human tumors; the most commonly employed models are in small animals. Considerable limitations exist in the quality of data obtained when animals with major differences in genetics, anatomy, and physiology compared to humans are exploited for research, such as rodents. The transgenic oncopig model is a new orthotopic tumor model for human cancer. Pigs permit the use of the same equipment applied in clinical patients, and thus, the data are easily translatable to humans. The presence of a tumor within a liver is also an ideal imitation of clinical MWA as tumor tissue has a distinct characterization and matrix when compared to normal liver tissue; prior porcine studies have been in normal liver tissue. The HT vascularity in the oncopig model has previously been shown to produce some hypervascular and hypovascular tumors that are similar to either HCC or CRC liver metastasis and has been used to assess the transcatheter embolization procedure.^[17]

When assessing MWA, the phenomenon of thermal convection exists only in an in vivo model. The presence of thermal convection results in the need for more energy to be delivered to reach a similar volume of AZ when compared to an ex vivo evaluation.^[23] Testing the AZ in an in vivo model is essential to accurately evaluate the ablation procedure and translate these results into clinical practice. In this animal tumor model, an AZ that was significantly smaller was observed when compared to the manufactures’ package insert of expected AZ volume for 65 watts and 4 min (12.1 cm³); this is likely because most data obtained for studying PA AZ is performed in ex vivo organs without tumors.

Various tools are available to perform thermal PA, including RFA, cryoablation, and irreversible electroporation. PA with different modalities has previously been studied with or without embolization; however, this work has been in normal non-tumor-bearing animals. In addition, this study is crucial because it evaluated MWA whereas most previous experimental work has applied embolization with RFA. MWA has superior heating characteristics; thus, reliably achieving appropriate margins in HT >3 cm in combination with embolization was felt more probable.^[24] A limited number of other groups have reported on embolization with PA but have applied different techniques to achieve arterial or portal embolization, which have included open vascular ligation as well as transcatheter procedures. Takamura et al. using MWA found that portal venous flow was most important to the diameter of the coagulation zone.^[25] This is logical since the liver receives the majority of its blood supply from the portal vein, and thus, embolization of the portal vein would have a greater influence on the vascular heat sink compared to arterial embolization. The most common embolization technique studied has been using lipiodol and gelfoam followed by RFA, most of these studies found only a small, increase in the AZ, predominantly along the short axis diameter from the ablation probe; moreover, a shorter time to achieve the AZ was also observed.^[26-29] Tanaka et al. in a normal pig studied ablation with calibrated particles before and after RFA and found smaller particle size significantly influenced the size of the AZ when performed before PA, not after.^[30]

In an open procedure, dissimilar to clinical transcatheter embolization performed by interventional radiologists, ligation of the hepatic blood supply followed by RFA, Shibata et al. reported RFA produced a significantly larger and more spherical AZ.^[31] An additional study assessing open, surgical MWA after clinical transcatheter embolization using 100–300 µm was completed and reported a significant increase in the area and diameter of the AZ after embolization, up to 60% that was specifically related to an increase in the peripheral zone. This study is comparable with the data obtained in our experiments that show a 65% increase when using 40 µm and 59% increase with 100 µm particles.^[32]

Limitations to this study include the small number of ablations for each group. During the design of the examination, calculations were made to optimize the power of the study to achieve a statistically significant result, which was achieved; however, more data points would have improved the study and help to better characterize the specific changes certain embolization materials cause when combined with ablation. Due to certain logistical challenges posed by working with large animals at our institution, cross-sectional imaging was not obtained pre or post-ablation which would have provided in vivo data concerning tumor perfusion after embolization and an in vivo assessment of the AZ which is typically assessed when PA is performed clinically, also affording a comparison of the AZ measured on imaging to our ex vivo assessment. Animals were euthanized immediately after ablation, having more remote time points to assess the longitudinal evolution of the AZ after treatment would have also been beneficial. Our animal model produces HT with vascularity similar to those seen in human tumors such as HCC; however, underlying liver disease will affect energy transfer and thus ablation size, the liver in this model is in a normal liver without changes of chronic liver disease or cirrhosis which is identified in many patients with HT.

CONCLUSION

The combination of embolization with PA has been studied in both pre-clinical normal livers and clinical studies, revealing an increase in the size of the AZ. We report on this combined procedure for the first time in a pre-clinical large animal orthotopic porcine liver tumor model, specifically, evaluating the embolization technique by varying the size of the particles and thus location of the embolization. Distal embolization directly affected the size of the AZ, with particle size below 300 µm generating the largest increase in the AZ.

Given our study, we hypothesize that the combination of embolization with MWA, particularly with smaller particles and multiple probes, will be able to successfully treat HT larger than 3 cm in the liver with an appropriate margin to achieve complete local cure.