Abstract
Background/Aim: The potential benefits of pancreatectomy with major arterial resection have been studied in the past, but findings remain controversial. Pancreatic neck/body cancer (PNBC) involving arteries frequently requires combined resection of the pancreas, artery and portal vein. Patients and Methods: Nine prospectively-registered consecutive patients with PNBC were enrolled, all underwent pancreatoduodenectomy with common hepatic artery en-bloc resection (PD-CHAR). We investigated the safety of PD-CHAR by blood flow evaluation with intraoperative indocyanine green fluorescence imaging in reconstructed vessels/organs. Results: Among patients who underwent PD-CHAR, there was no severe morbidity. Artery/portal vein combined resection and reconstruction was performed in all patients. Four (44%) patients had pathological positivity for cancer cell invasion into the nerve plexus of artery at the site of radiographic artery involvement, although one (11%) was diagnosed with pathological artery involvement. Conclusion: PD-CHAR following neoadjuvant therapy might be feasible for PNBC without severe postoperative complications. Survival benefits in PNBC should be confirmed in further studies.
Pancreatic cancer treatment continues to have extremely poor results, therefore currently it may be considered a medical priority. Surgery remains the only effective treatment for resectable pancreatic ductal adenocarcinoma (PDAC) (1, 2), and pancreatectomy with vascular resection is increasingly performed for patients with borderline resectable pancreatic cancer (BRPC) and locally-advanced pancreatic cancer (LAPC). To obtain negative surgical margins, pancreatic surgeons tend to favor venous resection, but due to significantly increased rates of postoperative morbidity, arterial resection and reconstruction remain controversial (3-8). There is particularly insufficient discussion regarding the situation of part of PDAC being located in the pancreatic neck/body and involving only the common hepatic artery (CHA) or the splenic artery (SA), defined as UICC T3 (9).
Pancreatic neck/body cancer (PNBC) contacting with the CHA frequently involves the portal vein (PV) at the same time, so it requires careful operative planning and perioperative management. We previously reported pancreatic neck cancer to be a specific character entity which should be addressed with multimodality management and surgical planning (10). As for indication of PNBC, to avoid residual cancer, pancreatoduodenectomy with common hepatic artery en-bloc resection (PD-CHAR) (11), distal pancreatectomy with celiac axis en-bloc resection (DP-CAR) (12, 13), or pancreatoduodenectomy with splenic artery en-bloc resection (PD-SAR) can be indicated for surgical procedure based on the tumor position and arterial anatomical features (14). Several papers focusing on hepatic artery/celiac axis resections have been published in recent years. However, regardless of previous larger cohort studies, the safety of artery-contacting pancreatic neck/body cancer has not been widely evaluated in prospective studies.
Pancreatectomy combined with arterial resection has not been established as a safe and standard surgery. In pancreatectomy with arterial resection, whether surgeons could intraoperatively confirm blood flow through reconstructed arteries or organs that have blood flow from the remaining blood vessels is a critical point. We, therefore, investigated the safety of CHA-combined pancreatoduodenectomy for T3 PDAC in a prospective pilot study assisted by intraoperative indocyanine green fluorescence imaging (ICG) blood flow evaluation in reconstructed vessels or organs that have blood flow from the remaining blood vessels. We discuss indication, the pitfalls of the surgical technique, perioperative management, neoadjuvant therapy, and the clinicopathological features of artery-contacting PNBC. This trial is registered at UMIN000027898 Clinical Trials Registry.
Patients and Methods
Patient characteristics and surgical indication. Between 2017 and 2020, 131 patients with borderline resectable (n=50) and locally-advanced (n=81) PDAC underwent neoadjuvant chemotherapy with intention for surgery at our hospital (5). Of these, to assess the safety and surgical outcomes, we investigated nine prospectively-enrolled consecutive patients with PNBC that underwent PD-CHAR. Consideration for surgical procedure was assessed according to the indication criteria (Figure 1) and was based on the anatomical tumor position (Figure 2). Patients with PNBC that contacted the CHA, the proper hepatic artery (PHA) or the SA with/without portal venous invasion were indicated for PD-CHAR or PD-SAR. Arterial reconstruction was performed for patients who underwent PD-CHAR. Meanwhile, we did not perform arterial reconstruction in patients who underwent PD-SAR with monitoring of organ blood flow including to the liver, the stomach and the spleen. Indications for PD-SAR were previously described (14). Pathologic stages were diagnosed according to the Union for International Cancer Control (UICC) seventh tumor-node-metastasis (TNM) classification. Based on pathological diagnosis of the resected specimen, we examined microscopic surgical margin status (R0 or R1). R0-status was defined as the absence of tumor cells within 1 mm of the resection margin and R1 status was the presence of tumor cells within 1 mm of the resection margin (15). This study was approved by the Wakayama Medical University Hospital (WMUH) Institutional Review Board (No. 2034). These procedures have been registered at UMIN Clinical Trials Registry, UMIN000027898 (https://upload.umin.ac.jp/cgi-open-bin/ctr/ctr_view.cgi?recptno=R000031966).
Indication flow chart for surgical procedure. MD-CT: Multi detector-row computed tomography; EOB-MRI: magnetic resonance imaging with gadolinium ethoxybenzyl diethylenetriamine pentaacetic acid; PET-CT: 18F-fluorodeoxyglucose positron emission tomography; AR: arterial reconstruction; SMA: superior mesenteric artery; PV: portal vein; BR-PC; borderline resectable pancreatic cancer; LAPC: locally advanced unresectable pancreatic cancer; CA19-9: carbohydrate antigen 19-9; SA: splenic artery; PHA: proper hepatic artery; GDA: gastroduodenal artery; CHA: common hepatic artery.
Consideration for surgical procedure planning based on the tumor position. Patients who underwent PD-CHAR with tumor close to or abuts to proximal PHA/GDA (A, C), and DP-CAR with tumor close to or abuts to SA/CHA (B, D). CA: Celiac artery; CHA: common hepatic artery; SA: splenic artery; PHA: proper hepatic artery; GDA: gastroduodenal artery; T: tumor. Double arrows indicate artery division sites.
Preparation. For arterial reconstruction, the inner diameter of PHA and SA were measured on multi detector-row computed tomography (MD-CT) at the proposed resection site. Data is stored in an electronic medical record system. We also consider and measure alternative grafts, such as the second jejunal artery, the right gastroepiploic artery, and the saphenous vein. To avoid inflammation of the arterial wall, preoperative coil embolization for CHA is not performed in this series. To enhance systemic physical condition all patients planned for artery-combined pancreatectomy were enrolled for the intensive prehabilitation (16).
Neoadjuvant therapy. In our institution, patients with PDAC underwent neoadjuvant therapy as follows: between April 2014 and May 2015, patients with borderline resectable and locally-advanced PDAC underwent modified FOLFIRINOX (without bolus 5-FU and LV, also decreased dose of irinotecan; FIRINOX) (6, 17). From July 2015, patients with borderline resectable and locally-advanced PDAC underwent nab-paclitaxel plus gemcitabine for 2 and 8 cycles respectively (18). Patients who underwent neoadjuvant therapy at other hospital were eligible for enrollment in this study. Planned pancreatectomy was performed on patients with no progression of disease and CA 19-9 level under 100 IU/l.
Surgical approach. The surgical approaches and techniques for DP-CAR and PD-SAR were described previously (12-14), but we describe those of PD-CHAR here.
First, for observation of tumor extent for resectability, the gastrocolic ligament is fully opened to observe the entire pancreas and to ensure the splenic artery (SA) is tumor-free. If the SA root has cancerous invasion, a second jejunal artery reconstruction is adopted, (J2A)-PHA.
Second, after the SA root is encircled with vessel tape, the surgical field is moved to the hepatoduodenal ligament, and it is dissected to confirm whether the proximal side of PHA and right hepatic artery are free from tumor invasion. Owing to the high rate of simultaneous invasion to the spleno-portal confluence and CHA in PNBC, the planned transection site is usually the pancreatic body/tail.
Third, the SA is dissected circumferentially from the distal to proximal side. The root of CHA is sequentially exposed to encircle it with vessel tape to secure the hepatic blood flow until artery/portal vein reconstruction. The SA is dissected along 5-8 cm so that it can be anastomosed to the PHA without tension and with natural angle according to the case, it is then divided with a clip.
Fourth, from the root of the CHA, en-bloc dissection is performed of the distal CHA and adipose tissue in front of the crus of the diaphragm, exposing the celiac axis, and lymph nodes 7, 8a, 8p, 9, 12a are dissected to expose the PV wall. The left gastric artery can be preserved in all patients unless the celiac artery is resected combinedly (19). Maintaining the tumor en-bloc state with the CHA and the PV, the pancreatic head nerve plexus is resected from the SMA and the CA to create a state of a resected specimen connected with only the hepatic artery and PV.
Finally, we proceed to vascular resection and reconstruction. It begins with division of the hepatic artery and sequential artery reconstruction while maintaining portal venous blood flow. The SA is exposed at the distal side so that it reaches the proximal cut end of PHA (Figure 3A, C, E). The PHA is microscopically reconstructed with the SA by plastic surgeons with an Everting interrupted suture by 8-0 diadem (Kono Seisakusho, Japan) in end-to-end anastomosis (Figure 3B, D, F). Next, the portal vein and superior mesenteric vein are clamped for their division and removal of the specimen. The portal vein is reconstructed in end-to-end fashion. After reconstruction of the arteries and the portal vein, ICG fluorescence imaging is performed to monitor and confirm the blood flow in the proper hepatic artery, portal vein, and in organs such as the liver, the spleen, and the stomach according to the performed procedure. Indocyanine green (ICG, 1 ml) (2.5 mg/ml) (Diagnogreen; Daiichi Sankyo, Tokyo, Japan) is injected intravenously to evaluate blood flow by infrared observation camera system photodynamic eye (PDE) (pde-neo, Hamamatsu Photonics K.K., Shizuoka, Japan). PDE is fixed 5 cm apart from the site of anastomosis for 5 min and observed in real time on the monitor (Figure 4). Blood flow in the reconstructed artery is also assessed with intraoperative doppler sonography directly contacting with the arterial wall or organs.
The splenic artery was exposed at the distal side so that it reaches to the proximal cut end of proper hepatic artery (PHA). Double arrows indicate artery division sites (A: case 5; C: case 8; E: case 11). The PHA was reconstructed with the SA microscopically, and arrow heads indicate anastomotic sites (B: case 5; D: case 8; F: case 11). CA: Celiac artery; CHA: common hepatic artery; SA: splenic artery; PHA: proper hepatic artery; GDA: gastroduodenal artery; T: tumor.
The proper hepatic artery (PHA) was reconstructed with the splenic artery (SA) microscopically, and arrow heads indicate anastomotic sites (A). Surgical field where indocyanine green (ICG) fluorescence imaging was performed to monitor and confirm the blood flow in the proper hepatic artery, and in the liver (L) and stomach (S) (B).
Postoperative management after artery reconstruction. All patients with artery reconstruction underwent doppler ultrasonography daily for a week and MD-CT on postoperative day 4 to confirm that reconstructed arterial blood flow was intact. To prevent excess tension loading at the site of artery reconstruction, patients did not eat a solid meal until postoperative day 7. Bed rest was otherwise unrestricted from postoperative day 1. Low-molecular-weighted heparin, 2,500 units/day on postoperative day 1, 5000 units/day on postoperative day 2, followed by 5,000 units/day of normal heparin for 2 weeks, were administered intravenously after surgery for arterial reconstruction.
Definition of postoperative complications. We defined delayed gastric emptying (DGE) according to consensus and used the clinical grading of postoperative DGE proposed by the International Study Group of Pancreatic Surgery (ISGPS) (20). Definition of pancreatic fistula was set according to International Study Group of Pancreatic Fistula (ISGPS) guidelines (21). Post-pancreatic hemorrhage was as defined by the ISGPS (22). Postoperative organ ischemia in the liver or spleen was defined as a lower density area of the liver or spleen on enhanced CT images within seven postoperative days. Surgical site infections included surgical wounds or intra-abdominal abscesses with positive cultures. Intra-abdominal abscess including hepatic or splenic abscess was defined as intra-abdominal fluid collection with positive cultures identified by ultrasonography or computed tomography associated with persistent fever and elevation of the white blood cell count. Ischemic gastropathy was defined by gastroduodenal ulcer or perforation due to ischemic change of the gastric wall identified by endoscopy or surgery, as we reported previously (23). Postoperative endoscopy was performed as needed for observation of gastrointestinal symptoms. Mortality was defined as any surgery-related death.
Statistical analysis. Cumulative survival was calculated by Kaplan– Meier method. All survival times were evaluated from the first day of initial therapy. All analyses were performed using the statistical software package SPSS II (version 20.0; SPSS, Inc., Chicago, IL, USA).
Ethical approval. All procedures performed were in accordance with the ethical standards of the institutional (WMUH) and national research committees and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.
Informed consent. Informed consent was obtained from all individual participants included in the study.
Results
Patient characteristics and surgical outcomes. Table I shows the characteristics of the 9 analyzed patients with PNBC, 8 with borderline resectable and 1 with locally-advanced PDAC. Tumors abutted to both the CHA and the portal vein in all patients. Artery/portal vein simultaneous combined resection was performed in all 9 patients (100%), and artery reconstruction was performed in all 9 patients (100%) between the SA and the PHA. Median operative time and blood loss were 507 min and 465 ml. We investigated the relationship between respective/total ischemia time and the incidence of hepatic infarction. No correlation was found between the factors.
Patients characteristics and surgical results of PD-CHAR.
Feasibility of ICG fluorescence imaging. All patients underwent intraoperative ICG fluorescence imaging, and were confirmed to have rapid blood flow and distribution in the reconstructed vessels or organs that had blood flow in the remaining blood vessels. No patients had allergic reaction to ICG.
Histopathological results. Table II shows the histopathological results of all patients. One of the nine patients (11%) who underwent PD-CHAR was diagnosed with pathological common hepatic artery involvement. Four of the nine patients (44%) with radiographic common hepatic artery involvement had positivity for cancer cell invasion into the nerve plexus of the common hepatic artery. The rate of R0-status was 89%. Only one patient had R1-status at the pancreatic transection site, although intraoperative diagnosis of the frozen section was negative for cancer cell invasion.
Patient histopathological results (n=9).
Surgical outcomes and complications. Table III shows the surgical outcomes, including the incidence of operative morbidity, mortality, and the completion of postoperative adjuvant chemotherapy. There were no major complications of higher than grade 3 and no 90-day mortality. One patient had infective colitis with Clostridium difficile toxin and underwent medication therapy (Table III). Due to sufficient diameter of the splenic artery (24), long-term patency of arterial reconstruction was confirmed on contrast-enhanced MDCT performed every three months in all patients. Long-term patency of the reconstructed portal vein was also confirmed in all patients
Patient surgical outcomes (n=9).
Survival. Median follow-up time was 21 (8-37) months. The estimated 1-year, 2-year, and 3-year survival rates were 78%, 49%, and 0% respectively, and estimated median survival time (MST) was 22 months in all patients (Figure 5A). Local recurrence as a first recurrence site was not confirmed in this series. All other patients otherwise had recurrence with distant metastases including the liver (n=3), peritoneum (n=2), remnant pancreas (n=1), and paraaortic lymph node (n=1), with some duplication. Estimated recurrence-free survival time (RFS) was 12 (6-28) months (Figure 5B).
Discussion
This study investigated the perioperative management for safety of pancreatectomy with en-bloc artery resection with reconstruction for artery-contacting PNBC within the bounds of the clinical trial. It should be understood that the degree of nerve plexus infiltration may be a top diagnostic priority in surgical decisions for artery-contacting PDAC even after neoadjuvant therapy. Cancer invasion along periarterial nerve plexuses around the CHA would be an especially troublesome obstacle to R0 resection.
In the present study, the incidence of positive invasion to nerve plexuses around the CHA were 44%, and to adventitia of CHA were 11%. An important feature of radiographic common hepatic artery involvement in this study might be a significant indication of perineural invasion of common hepatic artery, even where pathologically the tumor cells do not infiltrate into the common hepatic artery wall. These positive rates may be helpful in considering the risk of periarterial cancer development in the nerve plexus (NPL) around the artery. However, it might be not enough to appropriate indication of combined arterial resection by pathological findings of cancer invasion to resected arteries only. From another surgical point of view, Okada et al. recently reported that the NPL around the CHA dissection may not have survival benefits, even in the absence of CT evidence of CHA contact (25). There is still a possibility that we could experience R1 surgical margin at the periarterial nerve plexus around dissected CHA surface in a patient that underwent artery combined pancreatectomy on final permanent histopathological specimen. The artery-combined extensive pancreatectomy such as PD-CHAR is still, therefore, one option for artery-contacting pancreatic cancer, even in the strategic setting of neoadjuvant therapy. Clarification of the surgical benefit with arterial resection for patients with cancer invasion along periarterial nerve plexuses around CHA within the setting of a clinical trial is still required. The artery en-bloc pancreatectomy might improve the R0 resection rate and could have the potential to promote the prognosis for carefully-selected patients in a longer neoadjuvant clinical setting (26, 27).
Various studies have concluded that pancreatectomy with artery resection remains controversial because of the significantly increased rates of postoperative morbidity. To mitigate postoperative morbidity, we used intraoperative ICG fluorescence imaging to confirm sufficient blood flow through reconstructed arteries or organs that have blood flow from the remaining blood vessels. Several studies have reported that intraoperative ICG fluorescence imaging study has made it possible to clearly visualize blood flow or tissue perfusion and intraoperative identification of technical errors may prevent the need for re-exploration surgery due to anastomotic occlusion of vessels or anastomotic leakage during gastrointestinal surgery (28-30). Assessment of blood flow by intraoperative ICG is visually effective for displaying anastomotic occlusion of vessels. On the other hand, although doppler US analysis makes it possible to assess liver blood flow, it might be hard to assess anastomotic patency directly. In this study, ICG fluorescence imaging study showed good anastomotic patency in all cases. However, when ICG fluorescence imaging study shows poor anastomotic patency, vascular anastomosis will have to be redone. ICG fluorescence imaging study could, therefore, alter intraoperative decision making of such patients and prevent the need for re-exploration surgery due to postoperative anastomotic occlusion. Partial infarction of the liver was sometimes encountered postoperatively in patients who underwent hepatic artery resection on postoperative MD-CT, but we have not experienced hepatic abscess, similar to in previous reports (23, 31). We found no correlation between ischemia time and the incidence of hepatic infarction, so it may depend on the development of the communicating artery in the liver in individual cases.
We also proposed several approaches for pancreatectomy with en-bloc artery resection for artery-contacting PNBC. According to the anatomical position and distances from proximal/distal artery branch, PD-CHAR or DP-CAR are indications for PNBC involving the artery. Arterial anatomical features depending on each case can make PHA or CA closer to PNBC, a critical factor in avoiding residual cancer. In particular, counterclockwise celiac axis deviation brings the loop of SA to be in contact with PNBC, and if the tumor invasion is free from the hepatic artery, PD-SAR should be taken into account as a surgical indication. PNBC involving the artery is an interesting entity because the second to third branching of the CA are probably involved. In the direction along the CA, the root of celiac artery is rarely involved, so it is possible the involved artery is at the most distal celiac axis after LGA branching, the proximal SA/PHA, and most frequently the CHA. Importantly, the decision on which side of the pancreas to resect with PD-CHAR or DP-CAR should be based on CT findings, specifically whether the surgical margin is sufficient or not in each procedure.
Regarding the period of preoperative therapy, even with increased high R0 resection rate under this surgical strategy with artery combined resection under 2-8 months preoperative therapy, the estimated overall survival has not been favorable. Histological response to neoadjuvant therapy has also been poor. In spite of locally aggressive surgery, most of the first recurrences still had distant metastasis in patients with artery-contacting PNBC. These data have motivated us to undertake a further prospective clinical study of neoadjuvant therapy with longer period and resection for PDAC with high risk major arterial involvement (UMIN000040792).
As for prognosis, we compared the patient-level OS of patients who received FOLFIRINOX in the setting of BRPC and LAPC (32) in our systematic review and patient-level meta-analysis (6). OS for both groups was clearly superior to OS for patients treated with FOLFIRINOX for metastatic pancreas cancer in the RCT conducted by Conroy et al. (33). Remarkably, the median OS of 22 months for BRPC patients in the study was similar to the 24 months in the LAPC setting in patients who did not undergo surgery. Otherwise, the estimated median OS/RFS of 22/12 months of the present study was not worse than those of pancreatic neck cancer without abutment of CHA 17/11 months in our previous study (10). Recently, Del Chiaro et al. (34) reported that pancreatectomy with arterial resection seems to be safe and feasible in well-selected patients and is associated with an advantage of survival compared to palliation in patients affected by locally-advanced pancreatic cancer.
There were several limitations in relation to this study being a single arm, single institution-based study. The number of cases in the pilot test was small, so we paid attention to the procedure in order for the data not to vary (35). To allow a definitive conclusion regarding the safety of this approach, a further prospective study with a larger number of patients is required. We were not able to conduct larger studies with an appropriate sample design due to lack of data from previous prospective studies in this area. It would be important to clarify the features and outcomes of patients compared to similar baseline characteristics of those who did not undergo surgery. We did not find a significant impact on survival, but this does not exclude the possibility of our approach to be used as an optional strategy. A randomized, controlled phase III trial investigating patients with chemo/chemoradiation therapy alone is therefore needed to validate the true benefit of the pancreatectomy with artery en-bloc resection for PNBC.
In conclusion, PD-CHAR following neoadjuvant therapy might be feasible for PNBC without severe postoperative complications. However, survival benefit in PNBC should be confirmed in further studies.
Acknowledgements
We would like to thank Benjamin Phillis at the Clinical Study Support Center, WMUH, for proofreading and editing the manuscript.
Footnotes
Authors’ Contributions
Study concept and design: KO, MK; Acquisition of data: MK, SH, SH, MM, YK, MU, RK, AM, YW, SA; Analysis and interpretation of data: KO; Drafting of manuscript: KO, MK, SH; Critical revision of manuscript: HY.
Conflicts of Interest
All Authors declare that they have no conflicts of interest.
- Received October 7, 2021.
- Revision received October 27, 2021.
- Accepted November 29, 2021.
- Copyright © 2022 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.