Analysis of Hepatic Artery Reconstruction in Advanced Pancreatic Head Cancer

Abstract

Background:

Hepatic artery (HA) reconstruction is occasionally required in pancreatic head cancer with arterial involvement. With recent advances in combination chemotherapy, conversion surgery has become feasible for selected patients. This study evaluates the surgical techniques, postoperative outcomes, and survival of patients who underwent pancreatic resection with HA reconstruction.

Methods:

A retrospective review was conducted of 14 patients who underwent pancreatic resection with HA reconstruction at Saitama Cancer Center from 2012 to 2024. Clinical variables included demographics, tumor pathology, patterns of arterial invasion, reconstruction methods, anastomosis time, and complications. Patients were classified into 2 groups: those with common/proper HA invasion and those with replaced right HA invasion.

Results:

The cohort included 9 men and 5 women (mean age, 71.9 y). Six patients had common/proper HA invasion, and 8 had replaced right HA invasion. The overall median anastomosis time was 18 minutes 16 seconds. No anastomosis-related complications occurred, and hepatic arterial flow was patent in all patients.

Conclusions:

Meticulous microsurgical HA reconstruction enables safe arterial resection in selected patients with advanced pancreatic cancer. When combined with appropriate preoperative therapy and multidisciplinary management, this approach allows oncologically curative resection and may contribute to improved long-term survival.

Takeaways

Question: How can hepatic artery (HA) reconstruction be safely performed in pancreatoduodenectomy for locally advanced or anatomically complex pancreatic cancer requiring arterial resection?

Findings: In this retrospective series, microsurgeons performed HA reconstruction after arterial resection using posterior wall–first suturing and graft techniques. Postoperative Doppler and computed tomography demonstrated 100% patency, and no anastomosis-related complications occurred. Selected patients achieved long-term survival.

Meaning: Microsurgical HA reconstruction enhances safety and expands curative surgical options for advanced pancreatic cancer.

INTRODUCTION

Pancreatic cancer is one of the most aggressive malignancies, and surgical resection is the only potentially curative treatment. In recent years, advances in combination chemotherapy and microsurgical techniques have expanded the surgical indications for locally advanced diseases. Patients who were once deemed unresectable owing to arterial involvement are now considered candidates for resection with combination chemotherapy and vascular reconstruction in carefully selected cases.^1–4

Among these procedures, hepatic artery (HA) reconstruction remains technically demanding, as it requires meticulous microsurgical anastomosis and a thorough understanding of hepatic arterial anatomy. Although in classical anatomy the right HA (RHA) arises from the proper HA (PHA), in some anatomical variants, the RHA originates from the superior mesenteric artery (SMA). This variant is referred to as a replaced RHA (R-RHA) and poses unique challenges that must be carefully considered during operative planning.^5–7 (See figure, Supplemental Digital Content 1, which shows classical HA anatomy [A] and common variants of R-RHA [B], (https://links.lww.com/PRSGO/E840.)

Although several reports have demonstrated the feasibility of arterial resection and reconstruction in pancreatic cancer, evidence specific to HA involvement and reconstructive strategies during pancreatic cancer surgery remains limited.^1,2,4 Therefore, further clarification of the indications, safety, and oncological outcomes of HA reconstruction is warranted. To address this issue, we retrospectively analyzed 14 patients who underwent pancreatic resection with HA reconstruction at our institution, focusing on invasion patterns, reconstruction methods, and clinical outcomes.

METHODS

We conducted a retrospective review of patients who underwent pancreatic resection with HA reconstruction at the Saitama Cancer Center between January 2012 and December 2024. Fourteen patients were identified and included in this study. Clinical data, including patient demographics, tumor characteristics, pathological findings, and details of vascular invasion, were collected from medical records. Upfront surgery was defined as surgical resection performed as the initial treatment without prior neoadjuvant therapy. Conversion surgery was defined as surgical resection performed after initial systemic chemotherapy in patients initially considered unresectable or borderline resectable. The study protocol was approved by the institutional review board of Saitama Cancer Center (approval number: 2154).

The type and location of HA involvement were determined based on operative findings and preoperative imaging, and patients were classified into the common hepatic artery (CHA)/PHA invasion group or the R-RHA invasion group. Reconstructive procedures were categorized as either direct end-to-end anastomosis or interposition grafting. Data regarding the recipient vessels were retrospectively collected from the medical records.

The main microsurgeon was usually positioned on the patient’s left side, and the operating microscope was placed on the left side at the cranial level. We prefer to tilt the operating table slightly to the left to improve anastomotic site visualization. All arterial anastomoses were performed by experienced senior microsurgeons under an operating microscope with 8-0 monofilament sutures using the back-wall technique. (See Video [online], which shows HA anastomosis in patient 4 [CHA/PHA invasion group].) After completion of the anastomosis, fat was grafted over the anastomosis site and secured with fibrin glue to control minor bleeding and prevent kinking. The arterial anastomosis time was measured using a recorded operative video. Continuous variables were compared between CHA/PHA and R-RHA groups using the Mann-Whitney U test. Portal vein resection and reconstruction were performed by hepatobiliary–pancreatic surgeons when tumor invasion was observed.

Video 1. This video demonstrates the complete mesotherapy treatment protocol for facial skin rejuvenation using a sodium hyaluronate composite solution, including pre-treatment disinfection, anesthesia, standardized injection operation, and post-treatment care.

Download video file ^{(72.5MB, mp4)}

Postoperative Doppler ultrasonography was performed immediately after the anastomosis and again on the morning of postoperative day 1. Contrast-enhanced computed tomography was performed within a week after surgery in all patients to confirm the patency of the reconstructed HA. Perioperative complications, particularly vascular anastomosis–related events, were recorded. Overall survival was defined as the interval between surgery and death from any cause or the last follow-up.

RESULTS

Fourteen patients (9 men and 5 women; average age, 71.9 y; range, 51–83 y) underwent pancreatic resection with HA reconstruction (Table 1). Twelve patients had invasive ductal carcinoma, and 2 had intraductal papillary mucinous neoplasm–derived carcinoma. Of the 14 patients, 6 underwent conversion surgery, 5 underwent upfront surgery, and the 3 most recent resectable cases received neoadjuvant chemotherapy (NAC). Gemcitabine plus nab-paclitaxel was administered to the conversion surgery group, whereas gemcitabine plus S-1 was administered to the NAC group. Subtotal stomach-preserving pancreaticoduodenectomy was performed in 13 patients, whereas standard pancreaticoduodenectomy was performed in 1 patient due to a prior operation.

Table 1.

Patient Summary

No.	Age, y	Sex	Tumor Location	TMN*	Pathology	Type of Surgery	Vascular Invasion	Hepatic Artery Anastomosis	Anastomosis Time	Portal Vein Reconstruction	Vascular Anastomosis–Related Complications
1	79	F	Ph	T3N0M0	IDC	Upfront surgery	CHA/PHA	CHA–PHA	NA	〇	None
2	72	M	Ph	T4N0M0	IDC	Conversion surgery	CHA/PHA	CHA–PHA	41:22	〇 (graft)	None
3	64	M	Ph	T3N0M0	IDC	Conversion surgery	CHA/PHA	CHA–PHA	24:51	〇	None
4	77	M	Pbh	T3N0M0	IDC	Conversion surgery	CHA/PHA	CHA–PHA	24:00	〇	None
5	72	M	Ph	T4N0M0	IDC	Conversion surgery	CHA/PHA	CHA–graft–PHA	23:17 17:29	〇 (graft)	None
6	77	F	Ph	T3N0M0	IPMN-derived cancer	NAC/surgery	CHA/PHA	CHA–PHA	16:04	〇 (graft)	None
7	61	M	Ph	T4N0M0	IDC	Upfront surgery	R-RHA	GDA–RHA	NA	〇	None
8	61	F	Ph	T3N0M0	IDC	Upfront surgery	R-RHA	GDA–RHA	18:16	〇	None
9	74	M	Ph	T3N0M0	IDC	Conversion surgery	R-RHA	GDA–RHA	16:29	〇	None
10	83	M	Ph	T3N0M0	IDC	Conversion surgery	R-RHA	GDA–RHA	22:42	〇	None
11	72	F	Ph	T2N0M0	IPMN-derived cancer	Upfront surgery	R-RHA	GDA–RHA	16:48	—	None
12	80	F	Ph	T3N0M0	IDC	Upfront surgery	R-RHA	RHA–RHA	29:18	〇 (graft)	None
13	83	M	Ph	T3N0M0	IDC	NAC/surgery	Replaced RHA	RHA–RHA	12:09	〇	None
14	51	M	Ph	T3N0M0	IDC	NAC/surgery	Replaced RHA	J2A–RHA	14:28	—	None

Rows are arranged according to the type of vascular invasion.

^*TNM classification in the 7th and 8th editions of the JPS.

IDC, invasive ductal carcinoma; IPMN, intraductal papillary mucinous neoplasm; J2A, second jejunal artery; Pbh, pancreatic body and head; Ph, pancreatic head.

The CHA/PHA invasion group comprised 6 patients, whereas the R-RHA invasion group comprised 8 patients. All R-RHAs originated from the SMA. In the CHA/PHA invasion group, 5 patients underwent direct end-to-end anastomosis between the CHA and PHA (Fig. 1), whereas 1 patient required an interposition graft because the resection gap was too long for direct anastomosis to be feasible. In this patient, a descending branch of the lateral circumflex femoral artery was used as the graft because the femoral vein was harvested for portal vein reconstruction. In the R-RHA invasion group, 5 patients underwent reconstruction using the gastroduodenal artery (GDA) (Fig. 2), 2 using the R-RHA itself because the resection length was too short, and 1 using the second jejunal artery because both the GDA and proximal R-RHA were ligated at their origin. Portal vein resection and reconstruction were performed in 12 patients.

Fig. 1.

Typical schema of anastomosis in the CHA/PHA invasion type. The pancreatic invasion involving the CHA and PHA was resected, leaving a short gap that allowed tension-free direct end-to-end anastomosis (A). The CHA and PHA were anastomosed under an operating microscope using 8-0 monofilament nylon sutures (B).

Fig. 2.

Typical schema of anastomosis in the R-RHA invasion type. The pancreatic invasion involving the R-RHA was resected (A), and the R-RHA was anastomosed to the GDA under an operating microscope using 8-0 monofilament nylon sutures (B).

Arterial anastomosis time could not be assessed in 2 patients because the operative video records were unavailable due to the age of the records. In the remaining 12 patients (13 anastomoses), the overall median anastomosis time was 18 minutes 16 seconds. Median anastomosis time was 23 minutes 39 seconds in the CHA/PHA invasion group and 16 minutes 48 seconds in the R-RHA invasion group. No significant differences were observed between groups.

Patency of the reconstructed HA was confirmed in all patients using postoperative Doppler ultrasonography and contrast-enhanced computed tomography. No vascular anastomosis–related complications, including hepatic arterial thrombosis or hepatic abscess, were observed. The median overall survival was 1046 days (34.4 mo). At the last follow-up, 9 patients remained alive.

DISCUSSION

In this series, the patency of the reconstructed HA was confirmed in all the patients, and no vascular anastomosis-related complications such as hepatic arterial thrombosis or hepatic abscess were observed. These findings suggest that HA reconstruction can be safely performed using meticulous microsurgical techniques. Historically, preoperative HA embolization has been used to promote collateral circulation or redistribute hepatic arterial flow.^8,9 At our institution, the plastic and reconstructive surgery team collaborates closely with the gastrointestinal surgery team and is frequently consulted regarding complex reconstructive procedures, including abdominal wall and perineal reconstructions and the management of refractory peristomal skin ulcers. Through such collaboration, our team provided microsurgical expertise to optimize surgical outcomes. Within this multidisciplinary framework, HA reconstruction was initially performed by reconstructive microsurgeons in selected hepatic cancer cases. This practice has recently been extended to patients with locally advanced pancreatic cancer as surgical indications have expanded.

With close collaboration between hepatobiliary–pancreatic surgeons and reconstructive microsurgeons, pancreatectomy with arterial resection can be performed reliably. Although some reconstructive microsurgeons and vascular surgeons prefer surgical loupes, the use of an operating microscope for delicate arterial anastomosis is a key factor in ensuring safety and reliability, thereby minimizing risks of thrombosis and ischemic complications.^7,10–12 Nevertheless, even under magnification, HA anastomosis remains technically demanding because of its deep location, frequent caliber mismatch, and arteriosclerotic vessel changes. In our practice, we usually perform back-wall–first end-to-end anastomosis using 8-0 nylon sutures.¹³ Additionally, microsurgical methods, such as the double-needle technique and modified back-wall–first anastomosis, originally developed in liver transplantation, may also be applied to HA reconstruction in pancreatic cancer surgery.^14–17 The overall median anastomosis time was 18 minutes 16 seconds, with the longest time being 40 minutes 46 seconds in patient 5. Considering preparation for anastomosis, such as vessel positioning or trimming, the liver ischemic time was longer than the pure anastomosis time. Previous experimental studies have shown that approximately 60 minutes of warm global hepatic ischemia results in reversible hepatocellular injury.¹⁸ HA-only ischemia with preserved portal venous inflow is expected to cause less severe hepatocellular damage.

The reconstructive strategy varied depending on the site of arterial invasion, which we classified into the CHA/PHA and R-RHA invasion groups. In patients with CHA/PHA invasion, the pancreatic head cancer often extends along the GDA, necessitating resection and reconstruction of the CHA and PHA. Direct end-to-end anastomosis is performed for short-segment invasion and remains the preferred strategy when the resection length is less than 2 cm and both arterial ends can be approximated without tension.¹⁰ Interpositional grafting was required for longer resections. In 1 patient, the descending branch of the lateral circumflex femoral artery was harvested as an interposition graft because the femoral vein was simultaneously used for portal vein reconstruction. Although the radial artery has traditionally been favored as a graft for HA reconstruction, our findings suggest that the descending branch of the lateral circumflex femoral artery may represent another viable alternative.^11,19 However, at our institution, resection is not attempted in patients with celiac axis or SMA involvement, given the extremely high surgical risk and limited oncological benefit.⁹

R-RHA has been reported to occur in approximately 10%–20% of cases. Due to its close anatomical relationship with the pancreatic head, it is particularly susceptible to tumor invasion, often necessitating arterial reconstruction during pancreaticoduodenectomy.^4–6 In this study, all R-RHAs originated from the SMA; however, they have occasionally been reported to arise from the GDA or celiac axis.^4,8 Therefore, precise preoperative imaging and careful surgical planning are essential for safe and successful resection. In patients with R-RHA involvement, the recipient artery was selected based on its anatomical feasibility. The GDA was the most commonly used, whereas direct R-RHA–R-RHA anastomosis was feasible when the resection margin was short. When no option was available, the second jejunal artery was selected as the alternative inflow source. Other arteries, including the right gastroepiploic, splenic, and middle colic arteries, have also been reported as useful recipient vessels for HA reconstruction.^20–22

In the CHA/PHA invasion group, arterial resection for locally advanced pancreatic cancer corresponds to borderline resectable or unresectable disease, where HA reconstruction is required to achieve radical tumor resection. In contrast, in the R-RHA invasion group, most cases were classified as resectable; however, because of the aberrant anatomy, HA reconstruction was required even in nonadvanced cases. In addition, 2 patients in this group had both aberrant hepatic arteries and invasion of a major artery (SMA), necessitating HA reconstruction. Clearly distinguishing these groups underscores that they represent fundamentally different entities in terms of oncological resectability.

In this series, 6 patients underwent conversion surgery, 5 underwent upfront surgery, and 3 recent resectable cases received NAC. The use of gemcitabine plus nab-paclitaxel reflects advances in chemotherapy that have enabled resection even in cases previously deemed unresectable.^23,24 Median survival of 1046 days (34.4 mo), with 9 patients alive at the time of analysis, demonstrates the potential for long-term survival. Some controversies remain regarding the safety and oncological benefits of pancreatectomy with arterial resection.²⁵ However, in appropriately selected patients with borderline resectable or locally advanced pancreatic cancer, pancreatectomy with arterial resection has been reported to provide a clear survival benefit compared with palliative surgery.^1,26 These findings suggest that HA reconstruction in carefully selected patients may contribute to improved prognosis in pancreatic cancer. Moreover, recent progress in combination chemotherapy has expanded the role of conversion surgery to include initially unresectable cases. Taken together, these findings highlight the critical role of coordinated, multidisciplinary care—integrating hepatobiliary–pancreatic surgeons, medical oncologists, and dedicated microsurgeons—in safely delivering complex arterial reconstructions and improving outcomes for patients with locally advanced pancreatic cancer.

CONCLUSIONS

Microsurgical HA reconstruction plays a pivotal role in enabling safe and curative pancreatoduodenectomy in selected patients with advanced pancreatic cancer. By using refined microvascular techniques, dedicated surgical teams can achieve reliable arterial patency with minimal anastomotic complications, even in anatomically complex cases. Our experience highlights that a structured microsurgical approach—combined with multidisciplinary planning—enhances technical safety and expands resectability, underscoring the value of integrating plastic and reconstructive microsurgeons into pancreatic cancer surgery.

DISCLOSURE

The authors have no financial interest to declare in relation to the content of this article.

ACKNOWLEDGMENT

The authors thank Mayumi Noguchi for drawing the figures and Editage (www.editage.jp) for the English language editing.