Tomohide Hori , Ryuhei Aoyama, Hidekazu Yamamoto, Hideki Harada,Michihiro Yamamoto, Masahiro Yamada, Takefumi Yazawa, Masazumi Zaima

Department of Surgery, Shiga General Hospital, 5-4-30 Moriyama, Moriyama City, Shiga Prefecture, 524-8524, Japan

Locally advanced pancreatic cancer located in the head or uncinate process (i.e., uncus) often invades the confluence of the superior mesenteric vein (SMV), portal vein (PV), and splenic vein (SV) [ 1 , 2 ]. Additionally, chronic pancreatitis easily occludes drainage flow via the SV [3] . These pancreatic diseases force surgeons to performenblocresection of the SV. Simple ligation of the remnant SV without venous resection results in sinistral portal hypertension (PH) (i.e., left-sided PH), gastrointestinal bleeding,splenic congestion, and hypersplenism over the long term [ 1 , 2 ].Postoperative sinistral PH is considered an intractable complication accompanied by refractory symptoms similar to those of PH due to liver cirrhosis [ 1 , 2 ]. Optimal management of the remnant SV is required during surgery [ 1 , 2 ]; however, intentional venous reconstruction for drainage flow of the SV is still controversial [ 1 , 2 , 4-8 ].We herein focus on sinistral PH due to occlusion of drainage flow via the SV, present actual characteristics in typical cases of pancreatic cancer and chronic pancreatitis, and discuss a strategic adaptation of the distal splenorenal shunt (DSRS) procedure.

From historical viewpoint, W. Dean Warren and signature DSRS are important for PH treatment. W. Dean Warren (1924-1989),Chairman of the Department of Surgery at Emory University(DeKalb County, GA, USA) and the University of Miami (Coral Gables, FL, USA), was a pioneer surgeon for PH and variceal bleeding [9] . Warren originally published a description of DSRS in 1967 as an operative procedure for bleeding esophageal varices [10] and realized a unique concept, the so-called “principle of steel” [9] .Briefly, the SV is detached from the SMV and PV and subsequently reattached to the left renal vein (LRV) [9] ( Fig. 1 ). His signature surgery (i.e., the Warren shunt) selectively reduces the pressure in esophageal varices and controls bleeding and clotting [9] . Warren documented the technical aspects for successful DSRS and stated that the SV should be anastomosed to the superior aspect of the LRV because a venous anastomosis to the ventral aspect of the LRV is easily kinked and occluded [10] . In the present case, the SV was anastomosed to the superior aspect of the LRV.

Candidate for SV drainage route has been suggested. Splenic congestion may result in splenic flow disorder and/or infarction [ 1 , 2 ]. Some surgeons have suggested the importance of the inferior mesenteric vein (IMV) as the drainage route of the SV [ 2 , 8 , 11 , 12 ], omitting venous reconstruction of the remnant SV if the IMV flowing not into the SMV but into the SV is preserved. However, whether intentional preservation of the IMV as a drainage route of the SV will prevent sinistral PH remains controversial [ 1 , 2 ]. Intentional preservation of the IMV on the remnant SV might not prevent splenic congestion [ 1 , 2 ].

Sinistral PH associated with extended pancreatic surgery remains a matter of debate. Pancreatic cancer that occurs in the head and uncinate process (i.e., uncus) invades the confluence of the SMV, PV, and SV [13] . Although distal pancreatectomy generally accompanies splenectomy, the spleen is preserved in pancreaticoduodenectomy. Pancreaticoduodenectomy accompanied byen blocresection of the SV may result in the development of sinistral (left-sided) PH. Splenic congestion may result in splenic flow disorder and/or infarction [ 1 , 2 ]. Although the SMV and PV are inherently reconstructed afterenblocresection to maintain splanchnic flow into the liver, whether intentional decompression of the spleen must be performed in patients requiring diversion of the SV remains controversial [ 1 , 2 , 4-6 - 8 ]. Nevertheless, sinistral PH is an intractable complication after pancreaticoduodenectomy and more frequently occurs when the remnant SV is not adequately reconstructed.

We have experienced cases of sinistral PH due to simple ligation of the remnant SV during pancreaticoduodenectomy even in patients with a drainage route of the SV via the IMV. Fig. 2 A and B show typical findings of splenic flow disorder and sinistral PH in a patient with simple ligation of the remnant SV. The remnant SV in this case also had a drainage pathway via the IMV. A 49-year-old woman was diagnosed with locally advanced pancreatic head cancer. Her tumor invaded the confluence of the SMV, PV trunk, and SV. This confluence was simultaneously resectedenbloc, and the SMV and PV trunk were anastomosed in an end-to-end fashion.The remnant SV was simply ligated during surgery because it had a drainage pathway via the IMV. Although we estimated that the SV flow drained via the IMV, severe congestion of the spleen and sinistral PH occurred postoperatively. Enhanced computed tomography was performed on postoperative day 2. Splenic congestion resulted in splenic infarction from the early postoperative period.Pleural effusion and ascites were also observed.

Fig. 1. “Principle of steel” for liver cirrhosis with PH. Patients with cirrhosis accompanied by PH and a systemic hyperdynamic state. The SV is detached from the PV and SMV and subsequently reattached to the LRV. The Warren shunt selectively reduces the pressure in esophageal varices and controls bleeding and clotting (dotted red arrow). DSRS: distal spleno-renal shunt; IMV: inferior mesenteric vein; IVC: inferior vena cava; LRV: left renal vein; PH: portal hypertension; PV: portal vein; SMV: superior mesenteric vein; SV: splenic vein.

Fig. 2. Typical findings of splenic infarction due to simple ligation of the SV and development of collateral vessels from the spleen due to chronic pancreatitis. A and B:Typical findings of splenic flow disorder and sinistral PH in a patient with simple ligation of the remnant SV are shown. The remnant SV in this case also had a drainage pathway via the IMV. C and D: Typical findings of collateral development in a patient with chronic pancreatitis are shown. Collateral vessels from the spleen ( arrows )obviously developed. IMV: inferior mesenteric vein; PH: portal hypertension; PV: portal vein; SMV: superior mesenteric vein; SA: splenic artery; SV: splenic vein.

Also, sinistral PH associated with chronic pancreatitis remains a matter of debate. Chronic pancreatitis sometimes causes formation of venous collaterals, cavernous transformation, extensive fibrosis,porto-mesenteric stenosis, or venous thrombosis [3] . Drainage flow via the SV is easily occluded, although splenic inflow is preserved.Hence, collateral vessels subsequently develop [3] . Both benign diseases (e.g., alcohol-induced or autoimmune pancreatitis) and malignancies (e.g., pancreatitis associated with mechanical obstruction of the main pancreatic duct) result in chronic pancreatitis. Fig. 2 C and D show typical findings of collateral development in a patient with chronic pancreatitis. Chronic pancreatitis due to a benign condition in a 45-year-old man caused venous obstruction of the SV,and development of collateral vessels from the spleen was clearly observed. Splenic inflow from the SA was well preserved even in the presence of chronic pancreatitis. Hence, chronic pancreatitis can easily obstruct the SV flow, and remarkable development of venous collaterals from the spleen may occur. Pancreaticoduodenectomy in patients with chronic pancreatitis is a challenging procedure because of the presence of well-developed venous collaterals [3] . In patients with chronic pancreatitis, the collateral vessels might disappear if drainage flow via the SV is improved. Therefore,the remnant SV should be optimally handled during surgery for chronic pancreatitis.

In terms of the postoperative state of pancreaticoduodenectomy,refractory symptoms due to intractable PH are unwelcome challenges for pancreatic surgeons. Intentional decompression of the SV is a matter of debate [ 1 , 2 , 4-8 ], although venous reconstruction of the SMV and PV is widely accepted. This raises a simple question: is venous reconstruction of the remnant SV using DSRS a high-risk procedure in non-cirrhotic patients? The answer is No.Although Warren originally introduced DSRS for cirrhotic patients,we consider DSRS safe and feasible for venous reconstruction of the remnant SV during pancreatic surgery in non-cirrhotic patients without PH.

The following question can also be asked: how do pancreatic surgeons dismantle inadequate pancreatic surgery without DSRS?We consider it important that skillful pancreatic surgeons routinely perform DSRS and continuously promote the inevitable performance of this advanced surgical procedure worldwide.

From future perspective, the remnant SV should be adequately reconstructed. The frontier of this procedure is large for motivated surgeons.

Acknowledgments

None.

CRediT authorship contribution statement

Tomohide Hori : Conceptualization, Investigation, Supervision,Validation, Visualization, Writing - original draft, Writing - review & editing. Ryuhei Aoyama : Formal analysis, Investigation, Resources, Visualization. Hidekazu Yamamoto : Investigation.Hideki Harada : Investigation. Michihiro Yamamoto : Investigation.Masahiro Yamada : Investigation. Takefumi Yazawa : Investigation.Masazumi Zaima : Conceptualization, Formal analysis, Investigation, Resources, Supervision.

Funding

None.

Ethical approval

This report was approved by the Institutional Review Board of Nagai General Hospital (Tsu, Japan). The patients involved in this study provided written informed consent authorizing the use and disclosure of their protected health information.

Competing interest

No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article.