December 19, 2024
With improved survival after esophageal cancer treatment, the incidence of second primary malignancies (SPMs) is expected to rise. Patients who have undergone esophagectomy and later require pancreatoduodenectomy pose unique challenges. Gastroduodenal artery (GDA) ligation, often necessary during pancreatoduodenectomy, can compromise the vascular supply of a gastric conduit. Previous reports describe various approaches: GDA preservation, ligation with reliance on collateral circulation, ligation with revascularization, or complete sacrifice of the gastric conduit with colonic interposition for reconstruction.
A patient with a history of Ivor Lewis esophagectomy for esophageal adenocarcinoma presented with jaundice and was diagnosed with ampullary adenocarcinoma. Axial imaging and angiography revealed the gastric conduit’s sole vascular supply was the right gastroepiploic artery. While we were prepared for microvascular revascularization or colonic interposition if needed, we successfully performed a GDA-preserving pancreatoduodenectomy. Additionally, we performed a Roux-en-Y gastrojejunostomy instead of the standard loop to minimize conduit manipulation and ensure tension-free anastomosis. The patient recovered well and was eligible for adjuvant chemotherapy but declined this option.
Surgical management of SPMs in patients may be limited by prior therapies. In these situations, it is best for surgeons to consider conservative operative strategies for improved outcomes and timely additionally indicated treatments.
pancreatoduodenectomy; esophagectomy
gastroduodenal artery (GDA)
carbohydrate antigen 19-9 (CA 19-9)
esophagogastroduodenoscopy (EGD)
endoscopic ultrasound (EUS)
superior mesenteric artery (SMA)
superior mesenteric vein (SMV)
right gastric artery (RGA)
pancreatic ductal adenocarcinoma (PDAC)
cholangiocarcinoma (CCA)
pancreatoduodenectomy (PD)
length of stay (LOS)
overall survival (OS)
recurrence-free survival (RFS)
renal cell carcinoma (RCC)
squamous cell carcinoma (SCC)
intraductal papillary mucinous neoplasm (IPMN)
Standard pancreatoduodenectomy for periampullary malignancies involves division of the gastroduodenal artery (GDA). However, this approach poses challenges when a patient has undergone a prior esophagectomy with a gastric conduit dependent on the right gastroepiploic artery (a terminal branch of the GDA), requiring alternative surgical strategies. Here, we present a patient with a history of Ivor Lewis esophagectomy who subsequently developed resectable ampullary adenocarcinoma.
A 58-year-old male presented with jaundice, dark urine, and pruritis. Labs revealed elevated bilirubin, liver enzymes, and carbohydrate antigen 19-9 (CA 19-9). CT scan demonstrated intra- and extrahepatic biliary and pancreatic ductal dilation without identifiable masses or strictures. Esophagogastroduodenoscopy (EGD) revealed an ampullary mass, and endoscopic ultrasound (EUS) confirmed a 2 cm malignant-appearing mass without lymphadenopathy. Biopsy confirmed invasive adenocarcinoma (T1cN0).
Three years prior, the patient was diagnosed with esophageal adenocarcinoma (cT3N0). Treatment included neoadjuvant chemotherapy followed by a standard open Ivor Lewis esophagectomy. During this surgery, the left gastric and short gastric arteries were divided, and a conduit was fashioned from the greater curvature of the stomach, leaving the right gastroepiploic artery as the primary blood supply to the gastric conduit (Figure 1). Notably, pathology demonstrated a complete response to neoadjuvant therapy. The patient recovered well, requiring only occasional endoscopic dilations for dysphagia. At the time of presentation, he was two years post-esophagectomy with no evidence of recurrence.
Figure 1. Schematic Illustration of Gastric Conduit Arterial Supply Post-Ivor Lewis Esophagectomy.
Following a clinic discussion, the patient opted for surgical resection of the ampullary lesion. Preoperative angiography (Figure 2) delineated the gastric conduit’s vasculature, confirming exclusive reliance on the right gastroepiploic artery and demonstrating no contributions from the SMA or short gastric arteries, as suggested by prior axial imaging (Figure 3). The patient completed a bowel prep the evening before surgery, allowing for the possibility of colonic interposition.
Figure 2. Preoperative Diagnostic Angiogram of Gastric Conduit. Published with Permission
Figure 3. Preoperative CT Scan (axial view). Published with Permission
During the pancreatoduodenectomy, the gastric conduit was exposed, and the right gastroepiploic vessels were carefully isolated and protected. The gastroepiploic vein was traced to its superior mesenteric vein (SMV) insertion. The GDA was located in the porta hepatis, and dissection was carried distally, ligating superior pancreaticoduodenal branches. To address previous emptying issues and the tumor’s ampullary location, the decision was made to resect the duodenum, favoring this over a pylorus-preserving approach.
The GDA dissection was continued, revealing its bifurcation into the right gastroepiploic artery supplying the conduit. Intraoperative ultrasonography was repeatedly used to confirm anatomy and adequate blood flow to the conduit. Despite slight vasospasm, the pulse remained palpable, and the conduit was well-perfused. The gastroepiploic vessels were gently retracted leftward using a Penrose drain to protect the pedicle while tunneling between the pancreas and portal vein (Figure 4).
Figure 4. Intraoperative View of Intact Gastroepiploic Pedicle. Published with Permission
A Blumgart pancreaticojejunostomy was performed, positioning the gastroepiploic vessels posterior to the anastomosis. The cephalad position of the gastric conduit and significant adhesions limited mobilization, necessitating a Roux-en-Y gastrojejunostomy instead of the standard loop procedure.
Pathological examination revealed a moderately differentiated pancreatic adenocarcinoma originating in the uncinate process with direct ampullary invasion. Focally positive margins were noted on the uncinate, vascular groove, and posterior surface, along with involvement of six of fourteen lymph nodes (pT2N2).
The patient’s postoperative recovery was uneventful, and he was discharged on the fifth day. Following consultation with his local medical oncologist, he was deemed eligible for adjuvant multi-agent chemotherapy. However, the patient declined further treatment due to lingering side effects from previous chemotherapy regimens.
The rising incidence of esophageal adenocarcinoma and improved survival rates suggest clinicians are increasingly likely to manage patients who develop second primary malignancies after esophagectomy.2 In these cases, treatment approaches for resectable second primaries may face constraints due to the patient’s prior surgical history.
In fit patients with resectable disease, esophagectomy is often the preferred local therapy for esophageal adenocarcinoma.3 The most common procedures include transhiatal,4 Ivor Lewis,5 and McKeown esophagectomy.6 Alimentary continuity is typically restored by using the stomach as a neo-esophagus or creating a gastric conduit. This conduit relies primarily upon the right gastroepiploic artery (RGEA)—a terminal branch of the GDA—with minor contributions from the right gastric artery (RGA).7 As the GDA is routinely ligated during pancreatoduodenectomy, post-esophagectomy patients requiring pancreatic resection present a unique challenge.
Literature on pancreatoduodenectomy modifications for patients with prior esophagectomy is limited, though some technical adaptations exist to preserve the GDA and RGA.8 Our preferred approach, based on existing publications (Table 1), was GDA-sparing pancreatoduodenectomy.9‒17
Authors
|
Clinical History
|
Surgical Approach
|
Outcome
|
---|---|---|---|
Ikeda et al. |
Patient with previous esophagectomy with gastric conduit |
GDA-preserving pancreatoduodenectomy (PD)
|
Length of stay (LOS) = 31 days; received adjuvant gemcitabine
|
Patient with previous subtotal gastrectomy |
GDA-preserving PD with Roux-en-Y duodenojejunostomy
|
Overall survival (OS) = 3 years
|
|
Orii et al. |
Previous esophagectomy with gastric conduit |
GDA-preserving PD with Roux-en-Y gastrojejunostomy
|
LOS = 36 days |
Okimoto et al. |
Previous esophagectomy with gastric conduit |
GDA-preserving PD
|
LOS = 46 days
|
Addeo et al. |
Previous esophagectomy with gastric conduit |
GDA-preserving PD
|
LOS = 30 days
|
Fragulidis et al. |
Previous transhiatal esophagectomy with gastric conduit |
GDA-preserving PD
|
LOS = 15 days |
Kurosaki et al. |
Synchronous esophageal squamous cell carcinoma (SCC) and intraductal papillary mucinous neoplasm (IPMN) |
One-stage transhiatal esophagectomy and GDA-preserving PD
|
LOS = 43 days |
Uehara et al. |
Previous esophagectomy with gastric conduit |
GDA-preserving PD
|
N/A
|
Hirashita et al. |
Two patients, each with previous subtotal gastrectomy |
GDA-preserving PD
|
LOS = 18 and 29 days, respectively
|
Inoue et al. |
Previous esophagectomy with gastric conduit |
PD with microvascular reconstruction (GDA)
|
LOS = 56 days |
Okochi et al. |
Previous esophagectomy with gastric conduit |
PD with microvascular reconstruction (middle colic)
|
LOS = 2 months
|
Minagawa et al. |
Previous esophagectomy with gastric conduit |
PD with microvascular reconstruction (middle colic)
|
LOS = 90 days |
De Garcia de la Vega et al. |
Synchronous esophageal and ampullary adenocarcinoma
|
One-stage esophagectomy, total gastrectomy, colonic interposition, and PD
|
Received adjuvant therapy |
Belyaev et al. |
History of severe chronic pancreatitis |
One-stage esophagectomy, total gastrectomy, colonic interposition, and PD
|
LOS = 28 days |
Jayaprakash et al. |
Synchronous esophagogastric junction and ampullary adenocarcinoma
|
One-stage esophagectomy, total gastrectomy, Roux-en-Y esophagojejunostomy, and PD
|
LOS = 22 days |
Honig et al. |
Previous McKeown esophagectomy with gastric conduit |
PD with technically difficult duodenojejunostomy
|
LOS = 6 days |
Should GDA preservation prove impossible, our alternative plans were:
A single report describes a standard pancreatoduodenectomy with GDA ligation, relying on the RGA to supply the gastric conduit.24 However, this was not feasible for our patient due to minimal RGA flow on angiography.
The decision to utilize a Roux limb for the gastrojejunostomy adds a unique element to this case. While most published reports describe a standard loop technique, only two prior cases document the Roux limb approach.9,10 Given the anatomic location of the gastric conduit and associated adhesions, adequate mobilization for a loop gastrojejunostomy would have increased risk and potentially placed the anastomosis under excessive tension. A previous report where a standard loop technique was performed near the hiatus resulted in an anastomotic leak necessitating multiple reinterventions.24 Mindful of this risk, and anticipating likely adjuvant therapy,25 we prioritized minimizing potential complications for our patient. Notably, unlike many published case reports that detail postoperative issues or lengthy hospital stays, this patient fortunately avoided both.
This case demonstrates successful pancreatoduodenectomy after prior Ivor Lewis esophagectomy. Preservation of the gastroduodenal and right gastroepiploic arteries, along with a Roux-en-Y gastrojejunostomy, highlights the importance of surgical strategies that minimize morbidity in complex cases, facilitating timely adjuvant therapy.
As the incidence of secondary malignancies after esophageal cancer treatment increases, surgeons will likely encounter more cases where surgical approaches are influenced by prior therapies. Conservative operative strategies may improve patient outcomes and potentially expedite adjuvant therapy in complex scenarios. However, surgeons should always be prepared by developing contingency plans when operating on patients with complex surgical and oncologic histories.
Robbins KJa; Gauthier JMb; Kozower BDb; Fields RCa
Keenan J. Robbins, MBBS, MPHS
Department of Surgery
660 S. Euclid Avenue
Wohl Building, 9th Floor
St. Louis, MO 63110
Email: kjrobbins@wustl.edu
The authors have no conflicts of interest to disclose.
Research reported in this publication was supported by the Washington University School of Medicine Surgical Oncology Basic Science and Translational Research Training program grant T32CA009621, from the National Cancer Institute (NCI). The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.
Received: July 4, 2022
Revision received: October 27, 2022
Accepted: January 30, 2023