Actual 5-year survival rate is approximately 10% for all stages of pancreatic cancer [1]. With advances in multimodal management and surgical techniques, the actual 5-year survival rate of surgically resected pancreatic cancer patients has improved from 1.5% to 17.4% [2]. However, up to 80% of patients present with locally advanced pancreatic cancer (LAPC) [3]. In such cases, systemic chemotherapy with or without radiotherapy is considered as the first-line treatment [4].

Irreversible electroporation (IRE) is a novel ablative technique involving pulsatile delivery of electrical energy through needle electrodes placed within and around the tumor [5,6]. Electrical pulses can create nanopores in cell membranes by alternating membrane potential, thereby disrupting cellular homeostasis and inducing apoptosis. Unlike traditional ablation techniques, IRE is a non-thermal technique that might be associated with less damage to surrounding vessels [5,6].

Implementation of neoadjuvant therapy (NAT) has demonstrated promising results, showing a response rate as high as 34%. It allows for curative resection in patients initially staged as LAPC [7]. However, it is difficult to accurately assess response to NAT based on conventional cross-sectional imaging due to similar appearances of fibrotic tissue and tumor mass [8,9]. Therefore, even in cases showing radiological features suggesting a stable disease on post-NAT imaging, trial dissection is advocated [4]. Nonetheless, following NAT, resection might not be feasible in approximately 40% of patients with LAPC due to extensive involvement of arterial and venous structures [10]. If the disease is found to be unresectable upon surgical exploration, IRE may represent an option for palliative, local therapy (“back-up therapy”).

The aim of this paper was to report an early experience of intraoperative use of IRE in patients with an initial diagnosis of LAPC undergoing trial dissection following NAT showing radiological response or stable disease.

Study design

This was a retrospective study analyzing a prospectively collected clinical dataset. This study was registered with the Manchester Royal Infirmary Audit Department (QI-69).

Statement of ethics

Study approval statement: This study protocol was reviewed and approved by the Manchester Royal Infirmary Audit Department, approval number Q1-69.

Consent to participate statement: All patients provided written informed consent before undergoing the intervention.

Data collection

Consecutive patients with initial diagnosis of LAPC who underwent NAT followed by trial dissection and simultaneous intraoperative treatment with IRE from July 2017 to September 2022 at a single tertiary hepatobiliary and pancreatic (HPB) unit in the United Kingdom were included in this study. IRE was undertaken as either a “back-up therapy” with a palliative intent if the tumor was found to be unresectable or for “margin accentuation” if the tumor was resected. A single surgeon (NDLC) operated on all patients, with IRE performed in conjunction with a single interventional radiologist (VN). LAPC was defined according to the NCCN criteria [11]. The same enhanced recovery pathway after surgery used for pancreatoduodenectomy was used for the entire cohort. Perioperative outcomes including length of in-hospital stay, complications, 90-day mortality and readmissions were prospectively collected. Postoperative complications were defined and graded according to definitions of the International Study Group on Pancreatic Surgery (ISGPS) and the Clavien-Dindo (CD) system [12-15]. Cumulative impact of each complications was calculated using the Comprehensive Complication Index calculator [16]. This retrospective cohort study was based on the Strengthening the Reporting of Observational studies in Epidemiology statement guidelines [17].

Perioperative assessment

Results of all patients were discussed among members of the Unit’s tertiary HPB multidisciplinary team (MDT). Diagnosis of pancreatic ductal adenocarcinoma was confirmed by cytology or histopathology for all cases. All patients underwent triple-phase computed tomography scan of the thorax, abdomen, and pelvis as part of the staging process. Additionally, positron emission tomography and/or magnetic resonance imaging (gadolinium-based) was performed if further assessment of local and/or distant staging was deemed necessary.

The decision-making process and algorithm for managing LAPC used at our institution have been previously reported [5]. All patients received NAT with or without chemoradiotherapy using either FOLFORINOX or Gemcitabine-based regimen. Restaging assessment was performed approximately 4 to 6 weeks after completing NAT (within one month prior to surgery) with repeat triple-phase CT scan and tumor markers. If patients were found to have partial response or stable disease on restaging, either trial dissection ± IRE or surveillance was proceeded following discussion among members of the local HPB MDT. The decision regarding using IRE in a particular patient was also made during local HPB MDT discussions. All patients underwent open trial dissection with exploratory laparoscopy performed in some cases at the surgeon’s discretion to exclude intra-abdominal metastasis. The suitability of resection was assessed based on proximity and encasement of major venous and arterial structures according to resectability criteria defined by NCCN 2016 guidelines [11]. If deemed resectable, pancreatoduodenectomy was performed along with IRE for margin accentuation (IRE-MA). If the mass was found to be unresectable, IRE was performed as a “back-up” palliative therapy while biliary or gastro-jejunal bypass was considered depending on tumor localization and patient features.

IRE procedure

IRE was performed using an AngioDynamics Nanoknife system (AngioDynamics). The number of required probes (2–4) and their placement were determined based on intraoperative assessment of the tumor by the operating surgeon (NDLC) with assistance of intraoperative ultrasonography performed by an interventional radiologist (VN) to assess tumor size and its relationship with neighboring vital structures. IRE probes (19-gauge needle) were placed around the tumor to achieve adequate electroporation of the whole neoplastic mass or the resection margin. These probes were placed under ultrasonography guidance to ascertain appropriate positioning and distance between probes. Following initial positioning, ten test pulses were delivered to allow the machine to evaluate the current. The active tip length of the probes was 1.5 cm. It was reduced to 1 cm if the lesion was small (< 2 cm). The distance between probes did not exceed 2.5 cm. An end treatment current between 30 A and 40 A was targeted to achieve effective IRE. The actual treatment consisted of two cycles of a minimum 90 pulses, with the length of electrical pulse between 70 μs and 100 μs and a voltage setting between 1,400 V/cm and 1,800 V/cm (maximum 3,000 V/cm). The first cycle was administered into a deeper portion of the tumor. Depending on the tumor size, a second cycle was delivered following pull back to treat the superficial tumor portion. In the cases of margin accentuation, following Kocherization of the duodenum, IRE probes were inserted along the posterior margin of the tumor to encase the superior mesenteric vein (SMV) and pancreatic neck. In cases where only IRE was performed, the entire tumor was bracketed before IRE was performed. In some cases, when the tumor was large, 3 probes were used, with one probe used in the center of the lesion. Outer probes were moved to gain adequate bracketing of the entire tumor.

Follow-up assessment

Standardized follow-up was performed for all patients. All postoperative complications were discussed in the unit’s weekly M & M meeting. They were graded according to the ISGPS and CD grading system as previously described [18]. Following discharge, patients underwent routine follow-up with triple-phase CT scan and routine biochemistry at one month and three months after surgery, and every six months thereafter.

Data collection & statistical analysis

Data related to patient characteristics, tumor characteristics, neoadjuvant management, operative and IRE-procedure details, histopathological evaluation, perioperative course (including length of stay, complication rates, and 90-day mortality), evidence of first recurrence following procedure, and overall survival (OS) were reviewed. Recurrence-free survival (RFS) was defined as the period between pancreatic resection and first radiologically confirmed or MDT-confirmed evidence of recurrence. Progression-free survival (PFS) was used for patients who were deemed unresectable at trial dissection. It was defined as the time period from IRE treatment to first radiologically confirmed or MDT-confirmed evidence of disease progression. OS was defined as the period from the start of NAT to the date of death. Descriptive statistics for the above-mentioned data points was undertaken using GraphPad Prism 9.5.0 (GraphPad Prism Software Inc.).

A total of 19 patients constituted the initial study population. Of these, IRE was successfully performed in 18 (95%) patients. In one patient, IRE treatment was initiated but abandoned due to technical issues with the IRE generator. Finally, 18 patients (67% males) were included in the analysis with a median age of 60.2 years (interquartile range, 35–78 years). All patients were diagnosed with LAPC. All patients underwent NAT, with seven (39%) patients additionally receiving chemoradiotherapy. Fifteen (83%) patients received FOLFIRINOX chemotherapy.

The tumor was located in the head of the pancreas (n = 16, 89%) or in the body of pancreas (n = 2, 11%). The mean baseline tumor diameter was 28.82 ± 9.63 mm. The tumor involved SMV in 11 patients, superior mesenteric artery (SMA) in nine patients, celiac axis in two patients, and splenic artery in one patient.

During trial dissection, nine patients were found to have non-resectable pancreatic mass. They underwent either “back-up” IRE alone (n = 7) or biliary bypass + "back-up” IRE (n = 2). These patients were deemed non-resectable due to either extensive unreconstructible venous involvement/occlusion (n = 4), solid tumor contact of 180 or more degrees with arterial structures (n = 2), or both (n = 3). For the remaining nine patients, pancreatic resection was successfully performed and IRE was undertaken for margin enhancement (Table 1).

Table 1 . Comparison between back-up therapy (BT) and margin accentuation (MA) IRE

	BT-IRE	MA-IRE
Age (yr)	61.56 ± 4.61	58.89 ± 14.32
Sex
Male	6 (67)	6 (67)
Female	3 (33)	3 (33)
NAT
FOLFIRINOX	7 (78)	8 (89)
GEM-CAP	2 (22)	0 (0)
Both	0 (0)	1 (11)
NAT radiotherapy
Yes	5 (56)	2 (22)
No	4 (44)	7 (78)
Vascular involvement
SMV	8 (89)	3 (33)
SMA	5 (56)	4 (44)
CA/SA	1 (11)	2 (22)
Surgery
PPPD/PD	0 (0)	7 (78)
DP/total	0 (0)	2 (22)
Abandoned	9 (100)	0 (0)
Median IRE probes	4 (2–6)	3 (2–6)
Mean tumor size (mm)	32.25 ± 6.79	25.78 ± 11.09
Resection margin
R0		5 (56)
R1		4 (44)
Abandoned	9 (100)
Postoperative chemotherapy
Yes	2 (22)	8 (89)
No	5 (56)	1 (11)
Complications rate
Yes	2 (22)	4 (44)
No	7 (78)	5 (56)
Mean LOS (day)	17.57 ± 17.95	13.44 ± 11.71

Values are presented as mean ± standard deviation, number (%), or median (range).

IRE, irreversible electroporation; SMV, superior mesenteric vein; SMA, superior mesenteric artery; CA, celiac axis; SA, splenic artery; PPPD, pylorus-preserving pancreatoduodenectomy; PD, pancreatoduodenectomy; DP, distal pancreatectomy; LOS, length of stay.

The median number of IRE electrodes used was 3 (range: 2–6). Electrodes were placed in the peri-tumoral region in all patients. In eight patients, they were also placed near SMA/SMV. R0 resection margin was achieved in five (56%) of the nine resected patients. The remaining four patients were found to have R1 resection margin (SMV margin, n = 4; SMA margin, n = 1). All patients were confirmed to have pancreatic ductal adenocarcinoma on histopathological evaluation.

Postoperative complications occurred in six (33%) of the 18 patients, including postoperative pancreatic fistula (POPF) (n = 3), delayed gastric emptying (n = 1), postoperative pancreatic hemorrhage (PPH) (n = 1), and chyle leak (n = 2). Of note, two patients (11%) developed severe complications (CD ≥ 3), including one patient who developed superficial surgical site infection requiring debridement under local anesthesia (CD3a) and another patient who developed POPF and PPH requiring a reoperation for a leak from gastrojejunostomy (CD3b). One (6%) patient who developed portal vein thrombosis (PVT) with subsequent ascites required a radiologically inserted drain (CD3a). There was no CD4 or CD5 complication in our cohort. The median length of stay was 8 days (range: 6.25–15.00 days). There was no early postoperative death (90-day).

The median RFS of resected patients was 10 months (range: 4.5–15.0 months), with four (44%) of the nine resected patients experiencing disease recurrence following the procedure. The remaining five patients did not develop any disease recurrence, with a median follow-up of 14 months (Table 2). On the other hand, the median PFS of the nine unresected patients with “back-up” IRE treatment was 7 months (range: 3–11 months), with six (67%) patients experiencing a disease progression, while one patient was lost to follow-up and two died. The median OS of our cohort was 22 months (range: 14.75–27.50 months). The median OS from the date of intervention was 9.5 months (range: 6–16 months).

Table 2 . Patient intraoperative and postoperative details

No.	Treatment group	No of IRE probes used	Tumor size (mm)	Resection margin	Adjuvant therapy	Adjuvant treatment (type)	Recurrence site	RFS (mon)	PFS (mon)	OS (mon)
1	Resection + MA-IRE	3	25	R1	Y	Gemcitabine	Local (vessel) + peritoneal	2	NA	23
2	Resection + MA-IRE	3	8	R0	Y	Gemcitabine	N	14	NA	19
3	Resection + MA-IRE	6	39	R1	Y	Gemcitabine	Local (vessel)	3	NA	13
4	Resection + MA-IRE	3	14	R0	Y	Gemcitabine	Local (pancreas) + lung	9	NA	21
5	Resection + MA-IRE	3	35	R1	N	N	N	10	NA	17
6	Resection + MA-IRE	5	32	R0	Y	Gemcitabine	N	6	NA	14
7	Resection + MA-IRE	2	14	R0	Y	Gemcitabine-capecitabine	N	16	NA	25
8	Resection + MA-IRE	3	32	R0	Y	Capecitabine	N	26	NA	40
9	Resection + MA-IRE	2	33	R1	Y	Gemcitabine-capecitabine	Lung + peritoneal	12	NA	34
10	BT-IRE	6	25	NA	N	-	-	NA	9	23
11	BT-IRE	4	31	NA	Y	-	-	NA	13	27
12	BT-IRE	3	NA	NA	Y	-	-	NA	2	44
13	BT-IRE	4	35	NA	N	-	-	NA	7	19
14	BT-IRE	5	22	NA	N	-	-	NA	2	14
15	BT-IRE	2	40	NA	N	-	-	NA	8	23
16	BT-IRE	2	42	NA	N	-	-	NA	6	15
17	BT-IRE	3	32	NA	N	-	-	NA	18	29
18	BT-IRE	4	31	NA	N	-	-	NA	4	12

IRE, irreversible electroporation; MA-IRE, margin accentuation IRE; BT-IRE, back-up therapy IRE; RFS, recurrence-free survival; PFS, progression-free survival; OS, overall survival; Y, yes; N, no.

This report presents our initial experience from a tertiary HPB center of utilizing IRE as either a “back-up” therapy or for margin accentuation in patients undergoing trial pancreatic dissection for LAPC after NAT. Based on our limited experience, IRE appears to be a safe and feasible modality. However, further evidence is needed to assess its potential to improve oncological outcomes. In our cohort, there were three severe complications. However, there was no 90-day perioperative mortality.

Although PVT is an uncommon complication in patients treated with IRE, it can have serious consequences. Based on previous studies, the incidence of PVT following IRE ranges from 4% to 8% [19-21]. Although the exact mechanism remains unknown, the most likely explanation is related to the triad of Virchow: hypercoagulable state due to pancreatic cancer, reduction or stasis in portal venous flow due to involvement or abutment by the tumor, and possible vascular wall injury related to IRE [22-24]. In our cohort, one (6%) patient developed PVT following treatment with IRE. During trial dissection in this patient who had previously failed an upfront pancreatic cancer resection at another institution in the UK, tumor was deemed not resectable due to significant SMV encasement at the confluence with portal vein as well as extensive fibrosis and adhesion around SMV and SMA. Therefore, IRE as a “back-up” therapy was undertaken with six electrodes placed around and within the tumor. On postoperative day five, a CT scan revealed PVT with a small amount of ascites, which progressed to SMV thrombosis and symptomatic ascites by postoperative day 11 requiring ultrasound guided drainage. The patient was treated with treatment-dose low-molecular weight heparin for six months. In addition to PVT, IRE can also be related to other severe complications such as severe acute pancreatitis, upper gastrointestinal perforation/bleeding (especially duodenal), POPF, or hemorrhage from adjacent vessels (especially SMA) [5,21]. Overall, our initial experience demonstrates that IRE can be safely performed for patients with LAPC.

The overall prognosis of patients with LAPC remains extremely poor, especially in those with unresectable disease. In our study, the median OS for patients was 22 months (range, 14.75–27.50 months). These findings are comparable to those of previous reports regarding the utility of IRE in pancreatic cancer [24,25]. A systematic review including 691 patients has found that the median OS from diagnosis or treatment in LAPC patients undergoing IRE ranges from 10 to 27 months [25]. Martin et al. [26] have assessed the largest cohort (n = 200) of LAPC undergoing IRE to date and reported that the median OS after treatment is 23 months (range: 8.3–36.3 months) in patients treated with resection + IRE and 18 months (range: 4.9–55.4 months) in patients treated with only IRE in situ. However, it is important to note that the median follow-up period in their study is almost twice that in the present report.

Another potential role of IRE in the paradigm of pancreatic cancer management is related to accentuation of resection margins. In our cohort, a total of nine patients were successfully resected, with R0 resection margin achieved in 55% of these patients. Kwon et al. [20] have examined IRE-MA in patients with LAPC and demonstrated a R0 rate of 70.45% with a median OS of 22.4 months. Similarly, Martin et al. [27] have recently presented their data comparing borderline resectable pancreatic cancer undergoing pancreatectomy alone or with IRE-MA. Although patients in the IRE-MA group had significantly greater preoperative disease stage, margin positivity rate (27% vs. 20%, p = not significant) and local recurrence rate were similar to those of the pancreatectomy alone group [27]. Studies assessing the role of IRE in pancreatic cancer have primarily focused on advanced stages of this disease. However, the above findings present a rationale for assessing its effect in patients with resectable disease. The MACPIE-I study assessing IRE-MA in resectable pancreatic patients has found a significantly lower margin positivity rate (20.0% vs. 51.6%, p = 0.013) in an IRE-treated group than in a non-IRE group, although morbidity rates were similar [28]. Based on our findings and available literature, utilizing IRE seems to be a promising approach in the management of patients with LAPC or resectable disease.

However, the present study has limitations. First, it was limited by a small number of patients included in the analysis. There was also a possibility of selection bias, as seen in most cohort studies, especially as IRE treatment was undertaken by only one of consultant surgeons at our center. Additionally, the follow-up period following treatment with IRE was short, shorter than those of previous reports. Lastly, we did not compare outcomes of the IRE-treated group with those of a control group of non-IRE-treated patients. This was primarily because the main aim of this study was to assess the safety and feasibility of IRE. Nonetheless, the current report adds to the existing literature on the utility of IRE in the management of patients with pancreatic cancer.

In conclusion, intraoperative IRE is a safe and feasible approach. It can be incorporated as either as a “back-up therapy” or a “margin accentuation therapy” in the management of patients with advanced pancreatic cancer. However, its effect in improving long-term outcome requires further assessment.

Collaborator Authors: Saurabh Jamdar, Thomas Satyadas, Rahul Deshpande, Santhalingam Jeegatheeswaran, Panagiotis Stathakis, Aali Sheen.

None.

Francesco Lancellotti is a recipient of the BOWA liver surgical fellowship. The fellowship is not related directly with the present research and did not influence its outcomes. Nicola de Liguori Carino received training in the use of IRE technique from AngioDynamics, Latham, NY, USA. Nicola de Liguori Carino did not receive any financial support from AngioDynamics. Other authors do not have any conflicts of interest to declare.

Conceptualization: AP, FL, AKS, NDLC. Data curation: AP, FL, AKS, VN. Methodology: AP, FL, AKS, NDLC. Visualization: AP, FL, AKS, NDLC. Writing - original draft: AP, FL. Writing - review & editing: AKS, VN, NDLC.