Fig. 1
The four major prerequisites for technically successful LDLT are the following: adequate graft volume to avoid SFS graft syndrome, sufficient PV inflow for liver graft regeneration, good HV outflow to prevent graft congestion, and a secure BD anastomosis to prevent sepsis. BD bile duct, HA hepatic artery, HV hepatic vein, LDLT living donor liver transplantation, PV portal vein, SFS small-for-size
Operative Procedure
Recipient Hepatectomy
The surgeons need to be aware of both recipient and donor anatomy and must also have a specific plan for how to manage any recipient anatomical problems including the hepatic artery (HA), portal vein (PV), hepatic vein (HV), and bile duct (BD). In addition, there must be a plan for how to maintain adequate portal inflow without portal flow steal in the presence of large portosystemic collaterals, massive splenomegaly with enlarged splenic artery, and small graft size with less than 1.0 graft-to-recipient weight ratio (GRWR).
At the time of surgery, the recipient’s abdomen is prepared and draped; simultaneously preparation and draping of the left groin and thigh may also be necessary to access the great saphenous vein (GSV) often used for reconstruction on the back table and also for graft implantation in the recipient.
Before beginning perihepatic dissection, a full laparotomy should be performed, and the decision made as to when the preoperatively planned interventions should be undertaken, i.e., splenic artery ligation, splenectomy, isolation of portosystemic collaterals, etc. Usually splenic artery ligation or isolation of splenorenal shunt causing portal flow steal is undergone at the initial stage because edematous changes of bowel and mesentery after liver graft implantation might hinder those procedures. Other tasks, such as interruption of portosystemic collaterals and splenectomy, are performed after engraftment.
Using electrocautery, both coronary and triangular ligaments are divided; the detachment should be performed along the avascular plane. In some cases, however, it is more effective to peel off the hepatic capsule by thorough cauterization of both diaphragmatic and hepatic surface and ligation of sizable diaphragmatic vessel so as to reduce intraoperative bleeding when perihepatic collateral vessels or dense adhesions are present, such as in salvage or re-transplantation. After adrenal detachment, hemostasis of the perihepatic area becomes easier, and there should be minimal bleeding before dissection of retrohepatic IVC is begun. The following step is often dissection of hepatic hilum prior to retrohepatic IVC dissection. This takes into consideration the possibility of significant bleeding and/or technical difficulty depending on the extent of caudate lobe hypertrophy, Budd-Chiari syndrome, etc. Early dissection and division of hilar structures can facilitate the following procedures with less bleeding, particularly in salvage LDLT patients who have undergone previous major hepatectomy or in secondary biliary cirrhosis patients who underwent repeated surgery for biliary problems. Venovenous bypass through the inferior mesenteric vein or PV or temporary portocaval shunt may be helpful to reduce bleeding and splanchnic congestion during total hepatectomy and the anhepatic phase.
The main technical principle of dissection of the hepatic hilum is to preserve implantation options by maintaining the length and integrity of all hilar structures. In particular, meticulous dissection of the hepatic artery to obtain a sufficient length and adequate diameter is very important in order to match with the small hepatic artery opening of the partial liver graft without tension and to avoid intimal dissection of the recipient hepatic artery, such as is often encountered during hilar dissection of a recipient with portal hypertension. Alternatives are not easily achievable because of the vessels’ small diameter of usually less than 3 mm, in contrast to the diameters of the BD and PV.
When cholecystectomy is performed, the cystic duct is divided close to the neck of the gall bladder in order to preserve as much of its length as possible in case the cystic duct might be necessary for duct-to-duct anastomosis of where two bile duct openings are present on the liver graft with a wide gap between them. When LL implantation is being planned, the left hepatic artery (LHA) is isolated first at the left hilum and then dissected up to the umbilical portion of the left hilum in order to get enough length and also to get the branch patch of hepatic segment 2 and 3 HAs to accommodate the often larger graft hepatic artery. When RL implantation is planned, this step is a type of insurance procedure, and it is important for a surgeon to proceed following dissection in a comfortable situation. Dissection of the middle and right HA should be performed with preservation of the periductal connective tissue encompassing the axial periductal microcirculation in order to avoid posttransplant biliary complications related to ischemia. The right hepatic artery (RHA) should be freed up to the anterior and posterior branches so as to overcome size disparity between the graft and the recipient HAs. Division of HAs without ligation in the recipient side is better than ligation of both sides in order to obtain longer hepatic arteries and to avoid intimal injury.
The division site of bile duct should be decided by size and number of ductal openings of the graft. Pre- and intraoperatively, there should be communication between the recipient and donor surgeons regarding the cholangiogram. If multiple ductal openings are expected and also situated widely apart in the graft, the Glisson tissue containing the duct in the recipient should be divided at a high level in the hepatic hilum in order to create multiple ductal orifices with wide distances between those of both corners.
The last structure in the hilum to be further identified is the PV. The PV is usually dissected toward the right and left bifurcation of the main PV. For dual-graft LDLT and/or obtaining autogenous vessel graft to use for graft implantation, the PV needs to be dissected toward and beyond the takeoff of the right anterior, posterior, and left portal branches up to the umbilical portion. As for the extent of PV dissection to the opposite side, the PV is mobilized down at least 2 cm in length from the portal bifurcation toward the superior margin of the head of the pancreas.
During the anhepatic phase in LDLT, portal bypass is usually not the preferred procedure because most recipients tolerate portal clamping without hemodynamic instability due to maintaining caval flow, and construction of hepatic and portal vein anastomoses requires less than 60 min. LDLT using a right lobe graft , excluding extended right lobe, left lobe, and dual-lobe LDLTs, does not require systemic bypass because the piggyback technique allows partial clamping of the vena cava with a side-biting clamp and without hemodynamic instability.
The next step is division of the gastrohepatic ligament. At the time of left liver or dual-graft LDLT, if a LHA arises from the left gastric artery, this should be dissected as long as possible for arterial reconstruction of the implanted liver graft. The caudate lobe of the liver is detached from the IVC using a left-side approach in order to enhance the retrohepatic dissection as much as possible.
Before the removal of the recipient’s liver, an autologous GSV is retrieved from the groin, most commonly on the left side, because the right side is usually used for placement of the femoral artery and vein cannulation by the anesthesiologist.
Considering the harvest time of donor graft and the duration of the bench procedures, the diseased liver is removed as late as possible to reduce anhepatic phase. Hepatic veins (HVs) are divided individually using a vascular clamp instead of an endovascular stapler because the recipient’s HV openings should be used for anastomosis with the graft HVs after venoplasty.
Anhepatic Phase
Recipient Side
After recipient hepatectomy, bleeding control should first be performed. A frequent bleeding site during this phase includes the retrohepatic dissection area, and particular care should be given to the inferior phrenic artery and adrenal gland because they are difficult to expose after engraftment, and bleeding from these sites often occurs during the postoperative course.
Optimal venous outflow is critical for the success of an LDLT. Making a wide HV orifice in the recipient to accommodate the corresponding donor HV is an essential preparatory step for the engraftment.
For anastomosis of reconstructed MHV tributaries of the modified right lobe (MRL) graft, the septum between the recipient’s middle and left HV is usually divided, and a single large opening is made using unification venoplasty. Several Allis clamps are placed at the end of the middle and left HV stump, after which the previously applied clamp is removed so as to facilitate deep placement of side-biting clamp to the right hepatic vein (RHV) and including the anterolateral wall of the IVC.
Inappropriate ventrodorsal matching of the graft-recipient RHV anastomotic sites was found to be a significant risk factor for the development of RHV stenosis (Hwang et al. 2010). For RHV venoplasty, a large side-biting clamp is placed on the RHV; inclusion of the anterolateral wall of the IVC is necessary. A longitudinal incision only toward caudal side or both longitudinal and transverse incisions of the RHV with the IVC wall toward caudal and ventral sides as well as a wide patch plasty using bisected autologous vessel grafts such as GSV, PV, or cryopreserved iliac vessels are performed. These methods result in acceptably low incidences of early onset RHV stenosis (0–2 %) (Hwang et al. 2010).
For reconstruction of large short hepatic veins (SHVs), deep and secure side clamping of the IVC is required in order to prevent unnecessary tension during the anastomosis. It is usually therefore necessary to extensively dissect greater than half of the suprarenal IVC, and some branches of the right adrenal veins to the IVC need to be divided (Hwang et al. 2012a).
Back-Table Procedures
Vascular plasty or reconstruction may be required at the back table as a preparation for engraftment. MHV tributaries of the RL graft should be reconstructed using autogenous vessels harvested from recipient, deceased donor iliac vessels, or synthetic vascular graft. The RHV of the RL graft and the trunk of the middle and left HV of the LL graft may require venoplasty using previously mentioned vessel grafts in order to prevent HV outflow obstruction. When single or multiple SHVs larger than 5 mm in caliber are present, venoplasty is performed according to the previously described guidelines (Hwang et al. 2012a). If two separate pig’s nostril-shaped PV orifices (right anterior and posterior branches) in the RL graft are present due to type III or II PV variants, the recipient’s Y-graft of PV bifurcation prepared from the recipient’s hilar dissection can be used to make a single PV opening with adequate length of the neck for simple and safe anastomosis during engraftment (Lee et al. 2003b). When the recipient’s native PV cannot be used for the Y-graft due to severe PV stenosis, or closely attached hepatocellular carcinoma, circumferential fencing of PV by using autogenous bisected GSV or Y-graft from cadaveric fresh iliac vein branches can be used as an alternative (Guler et al. 2013).
Likewise, if two arteries are present in the graft, arterial reconstruction can be performed using a recipient’s HA Y-graft including the proper, right, and left hepatic artery on the back table under optimal conditions (Marcos et al. 2001). However, two separate HA anastomosis under a microscope using mostly recipient’s lobar or sectoral HAs after reperfusion might be preferred in order to avoid recipient bile duct ischemia and size discrepancy between graft and recipient HAs. Therefore, the recipient HA is dissected as high as possible in order to obtain a long-length and size-matched HA. Rarely, the HA in the graft may be accidentally injured during harvesting procedure, be too short, or be located at the posterior side of the PV with a short stump, all making reconstruction of HA in the recipient’s side very difficult or even impossible. Under the microscope on the back table, repair of the injured HA or lengthening the HA using a previously dissected and size-matched recipient’s sectoral or segmental HA segment is possible.
If more than two separate orifices of the bile ducts are not too far apart or pig’s nostril-shaped orifices are present, unification ductoplasty can be performed to create a large single opening.
These back-table reconstructions may require up to 2 h due to complex anatomy of the HV, PV, and BD. The procedures should therefore be performed while the liver graft is submerged in a 4 °C cold, preservation solution inside an ice-packed container.
Graft Implantation
Hepatic Vein Anastomosis
To obtain optimal venous outflow, not only the anastomosis itself but also the positioning of the graft needs to be considered. A change in graft position due to regeneration can cause outflow problems. Venoplasty of the recipient’s HV extended to the IVC wall to make an oval-shaped wide orifice with adequate length of the neck may help to minimize the outflow complications, even though slight torsion of the anastomosis occurs and causes outflow stenosis. The recipient’s HV needs to be maximally incised longitudinally and/or transversely and then attached like a fence by using autologous vein patch with a thick wall. For proper alignment at the time of HV anastomosis, two 5–0 nonabsorbable sutures with double needle arms are placed at the cephalic and caudal ends of both the graft and the recipient’s HV. Venting of the liver graft on reperfusion is usually not necessary in LDLT when histidine-tryptophan-ketoglutarate (HTK) solution is used as it has low potassium content.
Portal Vein Anastomosis
As mentioned in Fig. 1, adequate portal flow is essential for successful LDLT. The PV anastomosis is performed using the recipient’s PV trunk or with PV bifurcation to avoid redundancy of the PV anastomosis. Occasionally, the recipient right or left PV branch is used due to better size match or more favorable alignment than the PV trunk. The PV anastomosis must be constructed without tension, redundancy, or twisting. The preferred suture material is 6–0 Prolene, and the anastomosis is generally performed in a running fashion, incorporating sufficient “growth factor” (Starzl et al. 1984).
In recipients with severe PV thrombosis who cannot undergo a thrombectomy and/or PV plasty to enlarge the diameter of PV, mesenteric or renoportal interposition grafts are necessary using a cadaveric iliac vein or a polytetrafluoroethylene (PTFE) vascular graft (Moon et al. 2011). Patients who require caval transposition or arterializations of the PV or both are at significantly higher risk of morbidity and mortality and are perhaps inappropriate candidates for LDLT because adequate portal inflow is mandatory for partial liver graft regeneration.
Prevention of Portal Hyperperfusion or Portal Flow Steal
Virtually all living donor liver grafts in adult recipients are relatively small-for-size (SFS) and require optimal portal inflow for immediate graft regeneration, particularly when GRWR is less than 0.7–0.8 %. Portal hyperperfusion can cause excessive shear stress against sinusoidal cell of a SFS graft, which is known to be the primary cause of the SFS syndrome (Troisi et al. 2005). Various remedial procedures such as splenic artery ligation, splenectomy, and small-caliber hemiportocaval shunt (HPCS) creation have been utilized to modulate portal inflow (Kiuchi et al. 2003). High portal pressure is related however to not only to portal blood flow, but also to liver graft outflow resistance. The safest and most effective way to manage portal hyperperfusion in a SFS graft is provision of good HV outflow and modulation of high portal venous flow (PVF) and pressure (PVP) by splenic artery ligation or splenectomy (Ito et al. 2003; Ogura et al. 2010). Splenectomy is usually not indicated to decrease portal hyperperfusion because of its inherent complications such as bleeding, pancreatic fistula, abscess formation, PV thrombosis, and serious postsplenectomy infections. HPCS may trigger portal hypoperfusion and result in encephalopathy, graft atrophy, and even allograft necrosis, which may occur during the first 2 weeks postoperatively due to portal flow steal (Moon et al. 2008); thus, permanent HPCS is not an appropriate choice for resolving portal hypertension. In the author’s experience, PVF ≥250 mL/min/100 g graft weight or early elevation of PVP ≥20 mmHg after reperfusion is not associated with poor outcomes in SFS grafts as long as perfect venous outflow is provided and portal flow steal is completely interrupted. Considering portal flow steal phenomenon, it is not an issue limited to small-for-size grafts having less than GRWR <0.8 % undergoing HPCS, but a common issue in the field of LDLT using partial liver grafts with more than GRWR ≥0.8 % and having sizable spontaneous portosystemic collaterals (Lee et al. 2007; Moon et al. 2008).
Intraoperative Doppler ultrasound (IOUS) has been used to ensure hemodynamics of the implanted liver graft. IOUS however has difficulty in evaluating correct anatomical and hemodynamics parameters of portosystemic collaterals. Even when IOUS showed adequate portal inflow during LDLT, portal flow steal syndrome can manifest a few days after LDLT. To overcome the limitation of IOUS, IOCP has been used to evaluate portal flow to liver graft and to detect stealing hepatofugal flow through persistent portosystemic collaterals (Moon et al. 2007) (Fig. 2). In addition, IOCP is therapeutically utilized to complete interruption of surgically inaccessible portosystemic collaterals by coil embolization and to treat PV stenosis interfering with hepatopetal flow by stent placement (Kim et al. 2007). Measurement of PVP and PVF is stopped after introduction of IOCP and MHV reconstruction. To properly manage the potential small-for-size graft syndrome that may develop, both modulation of graft hyperperfusion by excessive portal hypertension and abolishment of portal flow steal through spontaneous or surgically created portosystemic collaterals are equally important.
Fig. 2
IOCP may salvage liver graft from portal flow steal. (a-1) IOCP and (a-2) the schema after engraftment demonstrate persistent PV stenosis in the intrapancreatic portion (black arrowheads) and portal flow steal through persisting sizable collateral (white arrowheads). (b-1) PV stent was placed, and ligation of the collateral vein was performed under guidance of a guidewire through the inferior mesenteric vein during IOCP. Portal flow steal through the collateral vein was not shown (white arrowhead), but intrapancreatic PV stenosis was not completely relieved (black arrowheads). (b-2) Balloon dilatation (black arrowheads) of the remnant PV stenosis was performed additionally. (c) Follow-up 3D CT after 45 months post-LDLT revealed a patent PV stent without stenosis and disappearance of a large collateral with good liver graft regeneration. CT computed tomography, IOCP intraoperative cineportography, LDLT living donor liver transplant, PV portal vein, LRV left renal vein
Hepatic Arterial Anastomosis
In LDLT, arterial anastomosis is performed after reperfusion in most cases as the donor hepatic artery is thin, small, and short and the anastomosis is tedious and often a time-consuming work requiring great attention. The diameter of the donor hepatic artery particularly in Asians is less than 3 mm in more than three-quarters of the donors (Inomoto et al. 1996; Okochi et al. 2010). These small anastomoses are generally performed in an interrupted fashion with 9–0 or 10–0 Prolene sutures under operating microscope. Introduction of microsurgical technique instead of surgical loop magnification has resulted in a decreased HA complication rate (Inomoto et al. 1996). Selection of the recipient’s HA for the anastomosis is decided primarily by size matching with the donor HA. When there is size disparity, the branch patch technique using small branches of donor or recipient HA is useful for wide and tension-free anastomosis (Aramaki et al. 2006). The length, condition of the arterial wall, and feasibility of stable positioning during anastomosis are also all important factors for choosing the anastomotic artery.
In many cases of LL and a few cases of RL grafts, multiple donor hepatic arteries are present. Whether all accessory vessels require reconstruction remains debatable (Ikegami et al. 1996; Suehiro et al. 2002). All hepatic arteries, including replaced and accessory arteries, are however essentially necessary arterial inflows because hepatic arteries are end vessels that supply a specific area of the liver. In addition, it remains somewhat unclear what impact a smaller ligated artery in the presence of good pulsatile backflow has on the arterial blood supply to segmental bile ducts (Suehiro et al. 2002). Reconstruction of all hepatic arteries is therefore performed to reduce possible HA complications, particularly related biliary complications that remain the Achilles’ heel of LDLT.
When extended intimal dissection occurs after hilar dissection of the recipient, a destructed HA is present related to previous transarterial chemoembolization, or HA thrombosis occurs after LDLT; the right gastroepiploic artery can be commonly usable for alternative HA inflow (Ahn et al. 2005). The right gastroepiploic artery is straightforward to dissect, is free from limitation of length, is frequently enlarged as a compensatory mechanism, and is feasible to perform anastomosis with a sizable graft HA. The right gastric artery, gastroduodenal artery, left gastric artery, splenic artery, inferior epigastric artery, internal iliac artery, sigmoid artery, inferior mesenteric artery, radial artery, and saphenous vein are also useful interposition grafts for extra-anatomical HA reconstruction (Uchiyama et al. 2010). When saphenous vein graft for HA reconstruction in LDLT must be used, a saphenous vein harvested from the ankle area is preferred to reduce pseudoaneurysm formation because it has a thick and strong wall and its caliber is usually well matched to that of the graft HA. Occasionally, an interposition graft from the infrarenal aorta using a fresh cadaveric artery or GSV is necessary for the arterial reconstruction when the recipient hepatic artery is thrombosed or obliterated all the way to the origin of the celiac axis.
Biliary Reconstruction
Biliary complications adversely affect the recipient’s quality of life and can occasionally even cause graft loss and death (Liu et al. 2004). Careful management in the intraoperative and postoperative period to reduce or treat the expected biliary complications is essential to prevent poor outcomes.
Hepaticojejunostomy with stent or without stent was once the standard biliary reconstruction method. More recently duct-to-duct (D-D) anastomosis became the standard technique for its several advantages over hepaticojejunostomy (Shah et al. 2007). Regardless of technique used, a patient with multiple ductal openings has a higher incidence of biliary stricture (BS) than those with single duct. When a LL graft is used, bile duct reconstruction is generally straightforward because 88 % of cases have single ductal orifices. In contrast, nearly 50 % of RL grafts have two or three ductal orifices, and often two orifices are more than 1 cm apart.
When RL grafts are to be used, precise investigation of the donor’s biliary anomalies is of paramount importance to minimize the number of ductal reconstructions and to avoid injury to the donor’s BD near the hepatic duct confluence. To obtain united bile duct openings, the right hepatic duct should be divided accurately by localizing the division site using a rubber-band tagging method after near-complete parenchymal transection (Lee et al. 2002) (Fig. 3).
Fig. 3
Rubber-band tagging method for bile duct division. (a) Schema of rubber-band tagging method during right hepatic duct division in RL donor hepatectomy. (b) Intraoperative cholangiography shows two short segments of radiopaque rubber-band marker sutured transversely 2 mm apart on the presumed division site of bile duct, and three bile duct openings are expected to come out. RL right lobe
Ductoplasty is not suitable for ductal orifices that are further apart than the diameter of the larger ductal orifice. Inappropriate approximation of two ductal orifices under tension may cause ischemia, leakage, and subsequent stricture of anastomosis. Hepaticojejunostomy is a better option in this situation. During ductoplasty, simply joining medial walls of two ducts will narrow the longitudinal diameter of a new opening and may further increase the risk of BS. Here septoplasty to make a much larger orifice is necessary to facilitate reconstruction and reduce BS formation (Fan et al. 2002).
The confirmation of viability of the donor and recipient bile ducts before reconstruction of duct-to-duct anastomosis is important to reduce biliary complications; the viability is decided by the presence of pulsatile arterial bleeding from the cut ends of the BD (Dulundu et al. 2004). The recipient duct opening needs to be one and half times larger than the size of the fully expanded graft duct opening in order to reduce BS.
The role of stenting in biliary anastomosis creation is controversial (Liu et al. 2004). For a large ductal opening, a stent may not be necessary. For small openings, stents may help prevent occlusion of the anastomoses by edema in the early postoperative period or prevent technical error such as catching of the posterior wall during placement of the anterior row of sutures. Routine use of small external drainage tubes exiting via the anterior wall of the common hepatic duct is preferred for several reasons. Firstly, biliary drainage can prevent leakage by minimizing intraductal pressure at the anastomotic site. Secondly, external drainage tubes can help keep lumens open in the early postoperative period. This may be important particularly when dealing with a very small (<2 mm) accessory ducts or a spiral cystic duct. Thirdly, it allows collection of information about bile production and hence about graft function. In addition, cholangiography can be performed to determine occurrence of leakage or stricture (Hwang et al. 2006b; Kasahara et al. 2006). The overall incidence of BS in the Asan Medical Center LDLT using right hemiliver has been less than 10 % in single ductal openings, 14 % in a ductoplasty opening, 24 % in two ductal openings, and 70 % in three ductal openings. All BS have been managed nonsurgically except three adult LDLT patients. Expert and dedicated interventional radiologists and endoscopists are absolute prerequisites to tackle the biliary complications in adult LDLT programs.