Sang-Hoon Kim,Deok-Bog Moon ,Woo-Hyoung Kang,Dong-Hwan Jung,Sung-Gyu Lee

Division of Hepatobiliary Surgery and Liver Transplantation, Department of Surgery, Asan Medical Center and University of Ulsan College of Medicine, 88 Olympic-ro 43-gil, Songpa-gu, Seoul 05505, Korea

Portal vein thrombosis (PVT) is no longer a definitive contraindication in liver transplants (LTs) [1] . Complex vascular reconstructions such as cavoportal hemitransposition (CPHT) [2–5],renoportal anastomosis (RPA) [ 6,7 ],and use of sizable collaterals (pericholedochal varix [ 8,9 ],coronary vein,peripancreatic or perigastroesophageal varices [10],right superior colic vein [11],ileocolic vein [12],and left gastric vein [13] ),or combined liverpancreas-small bowel transplant [14] are required for portal inflow in patients with total portosplenomesenteric thrombosis. A deceased donor liver transplantation (DDLT) case of successful portal inflow restoration directly from the gastroepiploic vein (GEV) without the use of an interposition graft was introduced [ 15,16 ]. Here,we report a successful living donor liver transplantation (LDLT)case using GEV for portal inflow restoration in a patient with totally obliterated splanchnic vein.

A 52-year-old man with hepatitis B virus-related cirrhosis and a model for end-stage liver disease score of 20 had repeated esophageal variceal ruptures and massive ascites. Preoperative computed tomography (CT) revealed totally obliterated portal,splenic,and superior mesenteric veins with extensive portosplenomesenteric thrombosis and a dilated 8-mm diameter GEV without a large splenorenal shunt (SRS) and peri-choledochal varix.However,a considerable amount of splenic venous flow drained into the GEV at the splenic hilum,and then ran through the paraduodenal collaterals to the inferior vena cava (IVC) ( Fig. 1 ). The living donor was a 24-year-old man who underwent a right lobectomy. The graft weighed 710 g with a graft-to-recipient ratio of 1.04.

Fig. 1. A : Preoperative three-dimensional computed tomography scan visualized a large engorged 8-mm diameter GEV (arrow) which ran through the paraduodenal collaterals to the IVC with a totally obliterated splanchnic vein due to the chronic thrombosis of the portal,splenic,and superior mesenteric veins; B : intraoperative findings of the large engorged 8-mm diameter GEV (arrow). GEV: gastroepiploic vein; IVC: inferior vena cava.

During the bench surgery of the donor graft,tributaries of the middle hepatic vein were reconstructed using Hemashield(collagen-impregnated woven double velour polyester grafts) with one segment V and two segment VIII veins. Two inferior right hepatic veins were reconstructed as a single outflow passage after quilt-venoplasty [17] . The portal vein (PV) of the donor graft was fenced with the recipient’s bisected great saphenous vein (GSV) to make a funnel-shaped,enlarged,and elongated PV opening for feasible anastomosis with a fresh cadaveric IVC interposition graft.

During the recipient’s operation,an engorged 8-mm diameter GEV,which was insufficient for PV inflow,was freed from the great gastric curvature and isolated to a length of>5 cm in an undivided state. The posterior wall of the GEV was incised longitudinally at a length of 5 cm and anastomosed side-to-end to the cadaveric IVC grafts with continuous polypropylene 5-0 sutures for portal inflow while the isolated GEV was clamped at both ends( Fig. 2 A). The anastomosed cadaveric IVC interposition graft was then placed through a retro-gastric course for anastomosis to the donor’s PV with the bisected GSV ( Fig. 2 B).

Fig. 2. A : Side-to-end anastomosis between engorged GEV and cadaveric IVC interposition graft; B : retrogastric placement of the anastomosed IVC interposition grafts; C :portal vein reconstruction using GSV fence; D : anastomosis between reconstructed PV and IVC interposition grafts; E : postoperative abdominal schematic; F : reconstruction view of three-dimensional computed tomography scan visualizing the patency of the portal anastomosis site (arrow) with a more engorged 15-mm diameter GEV. IVC:inferior vena cava; PV: portal vein; GSV: great saphenous vein; GEV: gastroepiploic vein; SMV: superior mesenteric vein.

The implantation of the liver graft started with hepatic vein anastomosis. Portal reconstruction was performed by anastomosing end-to-end the opposite end of the cadaveric IVC interposition graft and the reconstructed donor’s PV with continuous 5-0 polypropylene sutures ( Fig. 2 C). Here,the funnel-shaped,enlarged,and elongated bisected GSV fence on the graft side and the cadaveric IVC on the GEV side served as an interposition graft connecting the portal inflow from the GEV to the donor’s PV ( Fig. 2 D). The right hepatic artery was used for arterial anastomosis,and biliary anastomosis was performed duct-to-duct. The anastomosis sites revealed good patency in the intraoperative Doppler ultrasound,and the flow was within normal limits. However,we interrupted the paraduodenal collaterals at the entrance of the IVC to avoid postoperative portal flow steal [ 18,19 ],which resulted in better portal inflow in the intraoperative Doppler ultrasound. Moreover,an intraoperative portogram revealed hepatopetal flow without portal flow steal through the interposing vascular graft between the GEV and donor’s PV [20] . The postoperative abdominal schematic is shown in Fig. 2 E.

The intraoperative cold ischemic time was 89 min and warm ischemic time was 58 min,respectively. The total operative time was 926 min,and 24 units of red blood cells were transfused.Meticulous isolation of long GEV,portal inflow reconstruction,and intraoperative portography prolonged operative time. The patient was discharged without any complications at postoperative day 27.Preoperative ascites amounting to 4500 mL was drained during surgery. Meanwhile,postoperative ascites was drained as follows:1750 mL 7 days after surgery,525 mL 14 days after surgery,and 170 mL 22 days after surgery.

One year after LDLT,the patient demonstrated normal liver function. CT scan revealed patency of portal anastomosis with a more engorged 15-mm diameter GEV compared to the pretransplant CT scan ( Fig. 2 F).

The surgical strategies for portal reconstruction depend on the classification of PVT,as proposed by Yerdel et al. [21] . Grade IV diffuse PVT (DPVT),defined as total obliteration of the PV and proximal and distal superior mesenteric vein,is the most complex thrombosis and requires non-anatomical reconstruction for portal inflow restoration [22] . Portal reconstruction in DPVT depends upon the presence of large spontaneous or previously surgically created portosystemic shunts [23] . In Asan Medical Center,multidisciplinary approaches for nontumorous PVT including surgical thrombectomy,intraoperative portogram,additional PV plasty,PV stenting,interposition graft,or interruption of sizable collaterals reduced PVT-related complication rate and showed almost same results despite difference of severity of PVT in LT patients [24] .For management of complex DPVT of LT patients,in case of presence of spontaneous shunt,renoportal anastomosis through large splenorenal shunt is considered first,and use of large pericholedochal varix or cavo-portal anastomosis is the other option for portal flow restoration. If no shunts are available for portal inflow,CPHT and reno-portal anastomosis are used for portal flow restoration.

There have been several reports of the use of large left gastric veins for portal flow restoration in DPVT [23] . However,portal revascularization using GEV through direct end-to-end anastomosis to PV were rarely reported in DDLT [ 15,16 ]. Unlike DDLT,it is important to properly block the collaterals flow in LDLT to prevent hyper-perfusion that causes small-for-size syndrome. To the best of our knowledge,our report is the first successful case of portal inflow restoration using an engorged GEV in LDLT.

Here,portal inflow was restored using a cadaveric IVC interposition from GEV,which may be the only available portal inflow.However,a few technical obstacles must be overcome for successful LDLT. First,the GEV’s diameter was only 8 mm,which was insufficient for a sound PV reconstruction,that usually requires at least 10 mm. Second,the continuity of the GEV should be preserved to avoid isolated splenic or mesenteric venous hypertension because the GEV is the only sizable communicating vein between the splenic and mesenteric venous systems. Third,the positioning instability of the GEV due to gastric motility might result in portal flow disruption related to kinking or traction of the PV anastomosis. Fourth,the paraduodenal collaterals communicating with the GEV and finally draining into the IVC may be a route of portal flow steal,resulting in graft failure postoperatively.

For successful portal flow reconstruction,a preoperatively 5-cm long fresh cadaveric IVC graft was prepared for interposition graft. Meticulous side-to-end anastomosis was performed between the GEV and cadaveric IVC to avoid anastomotic stenosis related to the small-diameter GEV after longitudinal incision along the running course of the GEV without disconnection. In addition,the GEV was widely mobilized from the gastric wall through the division of communicating vessels. The IVC interposition graft was positioned through the retrogastric course to minimize the instability of GEV and IVC interposition grafts,which is related to gastric motility. After the engraftment was completed,we interrupted the paraduodenal collaterals through kocherization to avoid portal flow stealing.

In conclusion,portal inflow restoration using a venous interposition graft from the engorged GEV in LDLT could be a feasible and safe surgical option in patients with diffuse portomesenteric thrombosis. Especially,in cases of no available sizable collaterals or portosystemic shunts for portal inflow,the surgical technique we proposed will expand the indication of LDLT.

Acknowledgments

None.

CRediT authorship contribution statement

Sang-HoonKim:Visualization,Writing – original draft,Writing – review & editing.Deok-BogMoon:Conceptualization,Project administration,Supervision,Writing – review & editing.Woo-HyoungKang:Data curation.Dong-HwanJung:Data curation.Sung-GyuLee:Conceptualization.

Funding

None.

Ethical approval

This study was approved by the Institutional Review Board of Asan Medical Center,University of Ulsan College of Medicine (approval number: 2022-0389 ).

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.