Acute and subacute effects of irreversible electroporation on normal common bile ducts in a rabbit model

This study aimed to evaluate the acute and subacute effects of irreversible electroporation (IRE) on normal common bile ducts (CBDs).

of hospitalization, lower overall costs, and lower morbidity; however, maintenance of stent patency is difficult because of tumor ingrowth or overgrowth. 5,6 Tissue ablation techniques such as radiofrequency ablation (RFA) can safely induce tumor necrosis and are successfully used in early stage cancers or patients unfit for surgery. Recently, endoscopic or percutaneous focal ablation has been piloted to prolong ductal patency and debulk ingrowing tumors. 7 Intraductal RFA with an endobiliary catheter (Habib EndoHPB; EMcision, London, UK) has been designed to potentially improve the stent patency and survival rates. 8 However, because of heat sink and thermal injury, its clinical use was confined by the risk of treatment-related complications and limited efficacy. 9 Irreversible electroporation (IRE) is a non-thermal ablation technique that was first introduced for tumor ablation in 2005. 10 In contrast to other thermal ablation techniques, IRE can achieve non-thermal ablation through a high-voltage pulsed electric field without destruction of the extracellular matrix (ECM). 11 Therefore, IRE holds promise as an endoscopic ablation option in patients with malignant biliary obstruction. Various retrospective studies have demonstrated the safety of peri-biliary tumor ablation with IRE, 12,13 and catheter-directed IRE in swine bile duct was designed for biliary malignancies. 14 However, the effects of IRE on the bile duct deserve further investigation to determine the optimal parameters, the potential thermal injury and the long-term safety and efficiency. Thus, the purpose of this study was to evaluate the incidence of treatmentrelated adverse events after common bile duct IRE ablation and the temperature distribution during the treatment period.

| In vitro IRE experimental protocols
Three human cholangiocarcinoma cell lines (HUCC-T1, HCCC9810, and QBC939) and the human bile duct epithelium cell line (HIBEpiC) were used to assess the effects of in vitro IRE. All cells were treated with 70 pulses of high pulsed electric field with 90 μs duration at varying voltages. The CCK-8 assay (Dojindo, Japan) was performed to assess the cell viability changes 24 hours after IRE treatment. Detailed cell culture methods and IRE parameter were described in Appendix 1 in supplemental materials.

| Computer simulation of in vivo IRE
A two-dimensional model of common bile duct wall containing the epithelium layer (EL) and sub-epithelium connective tissue layer (SL) was constructed in this study ( Figure 1A). Computer simulations were performed to estimate the distribution of electric field and identify the potential of thermal injury during IRE treatment with a plate electrode. Details of the model construction were described in Appendix 2 (supplementary materials).

| In vivo evaluation of bile duct IRE
The animal experiments were performed according to the protocol approved by the Institutional Animal Care and Use Committee of Xi'an jiaotong University in accordance with ARRIVE guidelines and the National Institutes of Health guide for the care and use of laboratory animals. 38 adult male New Zealand rabbits (weight: 2.0-3.5 kg, IRE group: n = 28, sham control: n = 10) provided by Experimental Animal Center of Xi'an Jiaotong University were included for bile duct IRE ablation. Considering the potential effect of gender differences on the results, only male animals were selected in this study.
All rabbits were housed in an animal facility with a 12hour light/dark cycle and free access to food and water. Animals were maintained in a fasted state for 24 hours preoperatively. Sedation was administrated via intramuscular diazepam (2 mg/kg, Qinghai Pharmaceutical Factory) and general anesthesia was maintained using intravenous pentobarbital sodium (30 mg/kg, Sigma-Aldrich). After anesthesia, animals were placed in the supine position on the operation table. The skin overlying the upper abdomen was shaved, cleaned, and sterilized with povidone-iodine. A 30-40 mm midline incision was made to expose the porta hepatis. IRE ablation was then conducted using a plate electrode (platinum tweezertrode, 7 mm diameter, Harvard Apparatus Inc.). The electrodes were placed in direct contact with the common bile duct ( Figure S1), and the distance between the electrodes was adjusted depending on the diameter of the common bile duct.
Pulses were delivered using a commercial pulse generator (BTX ECM 830, Harvard Apparatus Inc.). All treatments were performed with a single set of experimental parameters (1500 V/cm, 70 μs pulse, 90 pulses, 1 Hz) that were determined according to the results of the in vitro experiments. Once the procedure was completed, the abdominal wall was closed. Intramuscular penicillin sodium (200 000 U/kg, North China Pharmaceutical Group Co.) was administered prophylactically.
Animals were euthanized at 1 day (IRE group: n = 4, sham control: n = 2), 3 days (IRE group: n = 4, sham control: n = 2), 7 days (IRE group: n = 7, sham control: n = 2), 14 days (IRE group: n = 7, sham control: n = 2), and 28 days (IRE group: n = 6, sham control: n = 2) after the IRE procedure respectively using pentobarbital sodium excessive anesthesia (90 mg/kg, Sigma-Aldrich) for sample acquisition. Intravenous blood samples were obtained preoperatively and at each follow-up imaging time-point to evaluate liver function. Incidences of bleeding and postoperative complications were recorded. Bleeding was defined as massive bleeding from major blood vessels. A 24G indwelling needle was placed through cystic duct to perform cholangiography to evaluate the incidence of adverse events post-IRE, and iohexol injection (Omnipaque 300 mgI/mL, GE Healthcare, Milwaukee, WI, USA) was used as contrast agent.

| Gross and Histopathological analysis
The animals were euthanized on either post-IRE day 1, 3, 7, 14, or 28 via excessive anesthesia. The common bile duct was removed en bloc and dissected longitudinally to expose the mucosal surface. The lesions along with the adjacent normal tissue were fixed in formalin, and the slides were stained with hematoxylin and eosin (H&E) and Masson trichrome (MT) stains. Terminal-deoxynucleoitidyl transferase-mediated nick end labeling (TUNEL) assay was performed according to the manufacturer's instructions to identify cell apoptosis. The slides were scanned with a digital slide scanner (Pannoramic SCAN II; 3DHISTECH Ltd.) and analyzed with CaseViewer 2.0 (3DHISTECH Ltd.). Histopathologic analyses were performed by a pathologist with 5 years of experience.

| Statistical analysis
Data are reported as mean ± standard error (SE). Statistical differences were analyzed using the parametric or F I G U R E 1 The computer simulation model of bile duct ablation. A, two-dimensional model of bile duct wall (Dotted line: the midline of bile duct lumen). B, Electric field distribution during pulse delivery with plate electrode. C, Tissue temperature at the beginning of the first pulse. D, Tissue temperature at the end of pulse delivery. E, The change of the maximum temperature during pulse delivery non-parametric unpaired Student's t-test. SPSS version 22.0 for Windows (IBM Inc.) was used for the statistical analyses and a P-value < .05 was considered statistically significant.

| Sensitivity of different cell lines to the effect of IRE
Three cholangiocarcinoma cell lines and one normal bile duct epithelium cell line were treated with IRE at different electric field strengths. The cell viability decreased in a dose-dependent fashion according to the applied electric field strength (Table S1). However, different cells showed different susceptibility to the effect of IRE. Compared to HCCC-9810 and QBC-939, the normal bile duct epithelium cell line HIBEpiC seemed more resistant to IRE treatment, and HUCC-T1 cells were least sensitive to IRE. The IC50 value was 1085.0 V/cm, 951.0 V/cm, 953.8 V/cm, 1372.0 V/ cm for HIBEpiC, QBC-939, HCCC-9810, and HUCC-T1 respectively ( Figure 2). Therefore, the electric field strength 1500 V/cm was chosen for further computer simulation and in vivo evaluation.

| Distributions of electric field and temperature
The distributions of electric field and temperature during bile duct IRE ablation are shown in Figure 1. The electric field and temperature were distributed unevenly in the EL and SL due to the difference of electrical and thermal properties. The electric field strengthen was higher in the epithelium layer and the corner of the electrode-tissue interface ( Figure 1B). IRE ablation caused an increase in tissue temperature, and the maximum temperature after IRE procedure was observed at the corner of the electrode-tissue interface ( Figure 1C,D). However, the maximum temperature of the entire area remained below 43°C which was considered the threshold of thermal injury ( Figure 1E) during IRE ablation period.

| In vivo IRE ablation of normal common bile duct
Irreversible electroporation of the common bile duct was successfully performed in 28 animals without any interruption and without bleeding, or extravasation and dilatation immediately post-IRE. Pulse delivery was completed within 2 minutes in each treatment process, and all the animals survived the IRE procedure.
During the follow-up period, two animals died (one each in the 7-and 14-day groups) due to severe abdominal infections on post-IRE ablation day 6. During autopsy, no IRE procedure-related biliary adverse events were observed. Biliary dilatation without apparent stenosis was first identified on postoperative day 7 and in a total of four animals throughout the follow-up duration. Furthermore, four animals experienced biliary stricture, which was also first observed on postoperative day 7 (Table 1). Cholangiography confirmed these biliary complications and revealed leakage-free passage of contrast in the treated bile ducts of these animals (Figure 3). No perforation of the bile duct was observed. Laboratory values showed an early increase in serum alanine and aspartate transaminases (ALT and AST) on the first post-IRE day, and they rapidly returned to normal levels within 1 week. Similarly, gamma-glutamyl transferase (γ-GT) and alkaline phosphatase (ALP) levels were increased in the first week post-IRE and gradually returned to normal levels over the duration of the follow-up period. There was no significant increase in serum bilirubin levels after the ablation ( Figure S2).

| Histopathology
Histopathologic analysis revealed epithelial sloughing and complete transmural cell ablation of normal common bile duct wall following IRE ( Figure 4A). H&E staining revealed a sharp and clear boundary between the ablated area and the surrounding normal tissue day 1 post-IRE. Full-thickness cell necrosis with hyperemia and infiltration of inflammatory cells was observed in the treatment region ( Figure 4A,B). The extracellular matrix (ECM) within the treatment region demonstrated inflammation, and MT staining indicated intactness of ECM in the ablated area ( Figure 4C). In TUNEL assay, the presence of diffuse positive cells in the ablated area demonstrated apoptotic cell death in the common bile duct wall at day 1 post-IRE ( Figure 4D). Tissue samples in the control group had histologic appearances of normal bile duct wall ( Figure 4E-H). Samples on day 3 demonstrated similar appearance ( Figure 5A-D). At day 7 post-IRE, diffuse fibrosis was observed in the treatment region with the missing epithelium ( Figure 5E-H). Regeneration of the bile duct epithelium was completed between days 14 ( Figure 6A

| DISCUSSION
Irreversible electroporation has emerged as a promising nonthermal tissue ablation technique over the last few decades. Due to the advantages of achieving tumor ablation without heat sink and ECM destruction, IRE is drawing more and more attention for its potential in the treatment of tumors of hollow viscera or close to major anatomical structures. 15,16 In the present study, the effect of IRE on normal rabbit common bile duct was evaluated directly. In vitro cell experiments were conducted to determine the appropriate electric field parameter for simulation and in vivo bile duct IRE ablation. We found that bile duct IRE ablation was feasible without dramatic temperature increase using an appropriate parameter, however, the risk of biliary dilatation and stricture may reduce the biliary patency in the long-term run. Results from the in vitro experiments of this study confirmed the sensitivity difference of different cell lines to the effect of IRE. This observation was consistent with previous studies. Gianulis et al. found that the difference in LD 50 spanned as much as nearly 80-fold among different cell types. Besides, they also found that the different sensitivity showed no correlation with cell or nuclear size, cell morphology, and metabolism level. 17 Kodama et al. observed that cancer cell lines seemed more sensitive to cell death from IRE compared to normal human bronchial epithelial cell lines. 18 Our results demonstrated that the normal cell line HIBEpiC seemed more resistant to IRE treatment than the cancer cell lines HCCC-9810 and QBC-939. However, HUCC-T1 cancer cell was the least sensitive to IRE. Thus, the different sensitivity may not be determined simply by benign or malignant characteristics. The mechanism of IRE susceptibility needs to be further investigated to provide evidence for optimal parameter selection.
Previous studies have evaluated the effects of IRE on vital structures adjacent to the target ablation area. Kevin et al. found that IRE was safe and effective for liver ablation in the liver hilum without the presence of heat sink, and that major bile ducts and portal veins were more resistant to IRE ablation. 19 Furthermore, long-term increase in serum bilirubin or transaminases levels was not identified, which was similar to our findings. Choi et al. evaluated the chronic effects of IRE on bile ducts and found that 6 of 7 animals in the periductal (electrode-to-duct distance ≤ 2 mm) electrode placement group developed bile duct narrowing, but no bile duct narrowing was observed when the electrode was placed more than 2 mm from the bile duct. 20 However, the parameters used in these studies were based on the clinical experience, and the electric field distribution and temperature changes during IRE procedure were not evaluated. Our study demonstrated that the electric field strength should be above 1500 V/cm to achieve cholangiocarcinoma cell ablation. The simulation models showed the uneven distributions of electric field and temperature in different bile duct layers because of the difference in electric and thermal properties, 21 this may illustrate the resistance discrepancy of different tissues. In addition, computer simulation revealed that the temperature was not dramatically increased during the ablation process, which may not cause thermal injury.
As for catheter-directed biliary ablation, Ueshima et al. performed normal porcine common bile duct ablation with endoscopic IRE. Their results illustrated the feasibility and early safety within 1 week after catheter-directed IRE. 14 However, the chronic effects were not evaluated in their study. Srimathveeravalli et al. assessed the effects of IRE on normal porcine ureters through catheter-mounted electrode.
They found that the intraluminal catheter-directed IRE ablation was safe and effective in normal ureter ablation over a relatively short follow-up period. However, the ureteral strictures were observed on day 7 post-IRE and did not resolve by day 28. 22,23 In the current study, four animals experienced biliary stricture and four experienced biliary dilatation from 7 to 28 days post-IRE. An alternative explanation for the different manifestation of biliary complication may be that the biliary dilatation may be secondary to a prior biliary structure or stenosis. As the extrahepatic bile ducts are composed of mostly connective tissue with little or no smooth muscle, the prolonged distension of the bile duct can lead to permanent deformation, which may not normalize even when the biliary structure or stenosis is relieved. 24 Although biliary injury has been observed in several studies, the application of IRE ablation for the treatment of biliary malignant tumors has potential advantages. Patients with unresectable cholangiocarcinoma always need biliary drainage with endoscopic biliary stenting; however, stent deployment is technically complex when more than one stent is needed. [25][26][27] Furthermore, the recurrent biliary obstruction rate after stent deployment has been reported as 20%-55%, which needs reintervention. 5 In such cases, IRE can be applied with endoscopy as an adjunct treatment before stent deployment and is expected to reduce the difficulty of stent deployment and prolong the stent patency duration. Moreover, endobiliary IRE ablation, with suitable catheter-based electrodes, can also be achieved via the drainage route of preoperative biliary drainage such as endoscopic ultrasonography-guided biliary drainage or percutaneous transhepatic biliary drainage. 28,29 In addition, for patients who can't tolerate curative surgery, IRE can serve as an alternative option for tumor reduction. This will allow opportunity for further curative surgery such as pancreaticoduodenectomy or bile duct resection. 30 There are several limitations to this study. The parameters of the in vitro experiments only changed the voltage, the pulse duration and number should also be assessed to obtain optimal treatment parameter in future studies. Besides, the computer simulation models can partly reveal the temperature distribution, however the change of tissue temperature was not measured in real time. Thermal injury may also exist because of the complex environment of the in vivo study as reported in previous studies. 23 Furthermore, we performed our study on normal rabbit common bile duct rather than a tumor model, electrical properties and microenvironments are different between the tumor and normal tissue; tumor ablation with IRE may require different parameters and have different effects. Therefore, further studies are needed to optimize the treatment parameters of tumor ablation.
In conclusion, the normal common bile duct retains the lumen wall integrity following IRE with immediate periductal placement of the electrode. However, the risk of biliary dilatation and stricture reminds that the parameters of IRE need to be determined more precisely to ensure the treatment efficacy and reduce the risk of collateral damage.