Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Pediatric liver transplantation has driven technical innovations in surgery over the past 25 years. Early successful liver transplantation relied on the use of size-matched whole-liver grafts. This requirement tended to exclude small children of less than 10 kg from liver transplantation because of the lack of donors and the higher rate of technical complications such as hepatic artery thrombosis. In addition, as increasing numbers of children were being referred for treatment, waiting times and waiting list mortality began to rise. Reduction techniques based on the segmental anatomy of the liver were developed to reduce waiting list mortality and transplant smaller children. The ability to cut livers down to match a child’s abdomen significantly opened up the donor pool and led to an increase in pediatric transplantation. Liver reduction proved to be successful; however, it led to the discarding of the right liver with a fall in the number of grafts from young donors available for adult recipients. The concept of “splitting” one liver for two recipients was developed to use both lobes and is now an established technique associated with excellent outcomes.
The development and current status of liver splitting are reviewed, with emphasis on donor and recipient selection, surgical techniques, postoperative complications, and clinical outcomes.
The segmental anatomy of the liver of relevance to splitting was described by Couinaud in 1957. This study provided the rationale for developing liver resection surgery and subsequently liver reduction (producing a functional graft suitable for a smaller recipient). Initially livers were cut down to provide a left lateral segment (segments II and III) or a left lobe graft (segments I to IV or II to IV) for critically ill children with promising early results. These reduced-size liver grafts provided patient survival that was comparable with full-size grafts and had a lower incidence of vascular complications and rapidly became established in the treatment of children with liver disease. The development of new techniques of hepatic vein reconstruction for the implantation of left lateral segment grafts, was critical in enabling safe transplantation of the smallest of children (piggyback technique).
In 1988 Pichlmayr was the first surgeon to split a liver to provide two grafts, one for a child and the other for an adult (liver splitting). Emond et al published a series of 18 patients who received split liver grafts with patient and graft survival of 67%, and 50%, respectively. Early experience reported by several European centers appeared to be discouraging, in contrast to the results obtained using whole-liver grafts ( Table 52-1 ). Graft survival ranged from 40% to 60% at 1 year with patient survival of 50% to 75%. In 1993 Houssin et al reported ex situ splitting in 16 recipients; of these 12 were urgent, and 4 were elective operations. Their technique was based on an anatomical assessment of the liver using arteriography and cholangiography on the back table with patient and graft survival at 1 year of 75% and 69%, respectively. The recognition of anatomical variations, including graft size and quality, hepatic and portal venous anatomy, and biliary and arterial anatomy, was emphasized. A European workshop on split liver transplantation in 1993 analyzed clinical data from nine centers of 50 livers split to provide 100 grafts. This collective experience (1988-1993) was published in 1995, by which time it was clear that there was a learning curve to this technique that contributed to a higher rate of complications and graft loss. Part of the problem had been the selection of higher-risk recipients, particularly for right lobe grafts, which contributed to suboptimal outcome. Patients transplanted from intensive care were unable to tolerate technical complications. These complications included portal vein (4%) and hepatic artery (11.5%) thrombosis and biliary leak (18%). The overall retransplantation rate was up to 22% for elective adult recipients receiving a right split graft compared to 10% with a whole liver. Back-table cholangiography or arteriography did not appear to reduce the incidence of arterial or biliary complications. Overall 41 out of 100 grafts in this collected series were lost. Long cold ischemic time and technical shortcomings were thought to be important risk factors for early complications. However, at 6 months graft and patient survival rates were similar to those reported after whole-liver grafting by the European Liver Transplant Registry when patient groups were matched for severity of liver disease. Outcomes for emergency transplantation in adults remained poor, and it was concluded at this time that split liver grafts should be used only in elective cases capable of tolerating early complications.
Author | Year | No. | Patient Survival (%) | Graft Survival (%) |
---|---|---|---|---|
Emond et al | 1990 | 18 | 67 | 50 |
Broelsh et al | 1990 | 30 | 60 | 42 |
Langnas et al | 1992 | 10 | 50 | 50 |
Houssin et al | 1993 | 16 | 75 | 69 |
de Ville de Goyet et al | 1995 | 98 | 68 | 62 |
Subsequently Kalayoglu et al described six ex situ splits using a metal probe to define the arterial and biliary anatomy. They recommended excision of segment IV after observing ischemic complications in several right lobe grafts. Although follow-up was short, the five adult and seven pediatric cases had patient and graft survival of 91% and 75%, respectively. Their splitting procedure was designed to retain the main vessels and common bile duct with the right lobe graft. The left lateral segments grafts were revascularized using interpositional venous or arterial grafts.
Improvements in the selection of patients and refinement of surgical technique led to the widespread adoption of splitting. Subsequently, Azoulay et al described their experience of 27 split grafts with a 1-year patient and graft survival of 79% and 78%, respectively. The incidence of arterial and biliary complications was 15% and 22%, respectively, although none of these led directly to graft loss. Again, donor selection was recognized as key, and only optimal livers were considered for splitting. Right lobe grafts were used for elective adult recipients, and the liver parenchyma was divided through the middle of segment IV to try to overcome the problems of postoperative ischemia and cut surface bile leaks.
Rogiers et al, in 1995, described in situ splitting as a consequence of the development of left lateral segment living donor operations. They reported a 6-month patient and graft survival of 92% and 85% with no biliary complications. Segment IV was preserved in six out of seven cases, although one was subsequently resected after implantation and a second patient developed a cut surface abscess. A comparison with 19 ex situ split grafts performed over the same period showed a lower peak postoperative aspartate aminotransferase level. In 1997 Goss et al reported 15 in situ splits resulting in 28 grafts with patient and graft survival of 92% and 86% and expressed the view that in situ splitting was associated with better outcome. However, improving results from ex situ splitting were reported in 1998 by Rela et al of 44 split grafts (from 22 donor livers) with patient and graft survival at a median follow-up of 12 months of 90% and 88%, respectively, and by Mirza et al in 1998 of 24 split grafts with patient and graft survival of 78% and 68%, respectively. Busuttil and Goss in 1999 reviewed 349 ex situ and in situ split liver transplants performed in 10 centers between 1990 and 1998 with 72 in situ splits performed in their own institution between 1992 and 1999. Busuttil and Goss concluded that in situ splitting was the technique of choice because of the apparent superior results. However, in a comparison of the best results from a single center performing ex situ splitting, the outcomes were identical (90% patient survival and 88% graft survival).
Further modifications have combined the advantages of each technique with minimal in situ dissection of the hilum but completing the parenchymal division to secure cut surface hemostasis. The dissection is completed on the back table after standard perfusion and retrieval. This offers the advantage of good hemostasis and the ability to probe the biliary tree to delineate the anatomy carefully before division of the hilar structures. In situ splitting offers advantages in recognizing relevant anatomy, preserving the segment IV artery, ease of sharing grafts with other centers, and securing hemostasis before implantation. In contrast, ex situ splitting requires no additional resources or surgical expertise at the retrieval, and the liver can be split while preparing the recipients for implantation. The majority of active splitting centers are currently using ex situ splitting techniques.
The long-term results of split liver transplantation are now comparable to those of whole grafts in both adult and pediatric recipients ( Table 52-2 ) . Although this procedure has become routine in many centers and has been successful in reducing waiting times for children, it has not achieved universal acceptance, and splitting rates remain disappointingly low internationally (1.3% of total transplants are split livers in the United States). Initially it was hoped that splitting would provide grafts for children while maintaining adult transplant numbers; however, many transplant centers remain reluctant to use such grafts.
The need for predictable liver function and thus a good-quality liver was recognized early in the development of split liver transplantation. At that time a marginal donor was defined by age of greater than 50 years. Donors less than 40 years of age were arbitrarily defined as suitable for splitting to provide a left lateral segment graft for a child. A right lobe recipient with a graft weight–to–body weight ratio (GWBWR) of less than 0.8% is potentially at risk for developing small-for-size syndrome. Although donors should ideally be less than 50 years of age, older donors can be used. Mildly fatty livers (less than 20%) may be suitable if cold ischemic times are kept as short as possible. Donors with mildly abnormal liver function (aspartate aminotransferase level less than three times normal or if higher with a falling trajectory), intensive care stay of greater than 5 days, and significant vasopressor support can be used with caution. If several risk factors are present, then the liver should not be split. The use of marginal grafts (e.g., fatty livers) for split liver transplantation carries a higher and unpredictable risk for graft failure because of the reduced functional mass and the likelihood of significant ischemia-reperfusion injury.
Splitting livers from donors in donation after cardiac death (DCD) may be possible in livers from young donors (under 30 years); however, few cases have been reported, and outcomes are less than satisfactory. Short warm and cold ischemia times are important factors to be considered if early dysfunction or ischemic cholangiopathy are to be avoided. Simultaneous implantation of both grafts and careful recipient selection (to avoid long or difficult dissections) should be considered if these grafts are to be used successfully.
Vascular and biliary anatomical variations are not an absolute contraindication for splitting a donor liver for a child and an adult recipient, but they should be considered in the light of potential posttransplantation complications. A thorough understanding of the surgical anatomy of the liver and particularly that pertaining to the caudate lobe is essential before undertaking split liver transplantation.
The initial experience of split liver transplantation was gained by transplanting high-risk patients on intensive care, but with poor outcomes. It was quickly appreciated that if this technique was to match the outcomes of reduced-size or whole-liver grafts, a more selective approach to recipient selection was required. Left lateral segment split grafts are suitable for all children with acute or chronic liver disease with excellent short- and long-term survival. The use of a split right lobe graft in adults requires assessment of recipient size and complexity and when possible should be used in stable recipients. Patients unlikely to tolerate biliary complications should be avoided.
The left lateral segment is invariably allocated to a child and the residual right lobe to another child or adult ( Fig. 52-1 ). The volume of liver allocated to the child is calculated by the donor-recipient ratio of 10:1 (for example, a liver from a 70-kg donor could produce a left lateral segment to be used in a 7-kg child approximately). The surgical techniques are different for ex situ and in situ split.
Ex situ splitting is performed after the donor liver has been retrieved and perfused with preservation solution. It should be split in the transplant center on a back table prepared for this purpose. The liver anatomy relevant for liver splitting is assessed. The liver is maintained in University of Wisconsin (UW) solution and fastidiously kept at 4° C for the procedure. Crushed sterile ice should be replaced regularly in the container to maintain the desired temperature. A thermometer to check the temperature of the preservation fluid is useful. Regular drainage of water and replacement of ice helps to maintain the temperature at 4° C.
A systematic assessment of the following structures should be performed:
Hepatic veins: The left hepatic vein should join the vena cava separately from the middle hepatic vein. Identify if one or two veins are present. If two veins are present, back-table reconstruction will be necessary.
Hepatic artery: Check for the presence of left or right accessory hepatic arteries. Identify the left, right, and segment IV hepatic arteries.
Portal vein: Identify the bifurcation between left and right portal vein.
Biliary system: Avoid dissection of the common bile duct. The bile duct is divided in the hilar plate. As little dissection as possible should be performed between the bile duct and the hepatic artery to preserve the arterial supply to the bile duct.
The allocation of vascular structures depends on the liver anatomy and the transplant recipient. If conventional anatomy is present, the left hepatic vein is isolated at the level of the vena cava and divided transversely ( Fig. 52-2 , A ). The orifice created in the vena cava should be sutured transversely later. The portal vein is dissected to the bifurcation, and the tributaries to the caudate lobe are ligated and divided. The left portal vein is divided at its origin (see Fig. 52-2 , B ), and the defect in the portal vein is sutured transversely. The hepatic artery is dissected to the bifurcation. If there is a clear division between the left and right hepatic arteries, then the left hepatic artery is taken at its origin (see Fig. 52-2 , C ). If there are multiple arteries present, then allocation should be performed in such a way as to minimize the vascular reconstruction required. A fine metallic probe can be used to determine the bile duct anatomy. The identification of the bile ducts draining the left lateral segment and segments I and IV are key in avoiding posttransplantation biliary complications. The left bile duct is divided and the hilar plate oversewn with a fine PDS suture (see Fig. 52-2 , D ).
Once the vessels and bile duct have been allocated and divided, the dissection of the liver parenchyma starts at a point 1 cm to the right of the falciform ligament. The parenchyma is crushed gently with a hemostatic clamp revealing vascular structures, which are tied or clipped individually and then divided. This procedure is repeated meticulously until the liver is divided. The left lateral segment graft is then flushed with preservation solution through the artery and portal vein to test leakage at the cut surface. Further cut-surface sutures are often required to achieve satisfactory hemostasis.
The left lateral segment graft is then repacked until required for implantation. The vessels/structures for this graft are the left hepatic vein, hepatic artery (left hepatic artery or common hepatic artery), portal vein (left or main trunk), and left hepatic duct. Ideally the recipient hepatectomy should have been completed by the time the graft is ready for implantation to minimize cold ischemia.
The extended right lobe graft requires further preparation. Segment I (caudate lobe) is removed to help with implantation and to avoid kinking of the portal vein. We routinely resect segment IV to avoid ischemic liver and to reduce the incidence of cut surface leaks. The extended right lobe graft consists of the vena cava (with the left hepatic vein orifice sutured transversely), hepatic artery (right hepatic artery or common hepatic artery), portal vein (right branch of the portal vein or main portal vein), and the common bile duct (left hepatic duct remnant oversewn) ( Fig. 52-3 ).
The procedure is performed during the multiorgan retrieval and before the perfusion with preservation solution. The donor liver parenchyma is divided while maintaining an intact blood supply through the left lateral segment. The transection of the liver parenchyma is performed using the clamp crush/fracture, Cavitron Ultrasonic Surgical Aspirator (CUSA) dissector, or other techniques used in hepatectomy. The ligature of vessels needs to be meticulous on both cut surfaces to avoid excessive bleeding and hemodynamic compromise of the donor. Once the parenchyma has been divided, the perfusion of segment IV can be assessed. If compromised, it can be removed on the back table. The organ retrieval procedure then continues as normal, and the organs are perfused with preservation solution, the liver vascular structures and bile duct are divided as convenient, and both liver grafts are stored separately. These grafts are ready for implantation and can be delivered to different transplant centers.
The advantages of in situ split over ex situ split are as follows:
Better hemostasis on the cut surfaces
Reliable assessment of the viability of segment IV after parenchymal division
No need for bench work after retrieval
Better for organ sharing with other centers
Potential reduction in cold and warm ischemic times
The disadvantages of in situ split over ex situ split are the following:
Prolonged organ retrieval
Infrastructure requirements in donor hospital
Greater expertise required of the surgical retrieval team
Potential to compromise donor stability and impact on other organs
Become a Clinical Tree membership for Full access and enjoy Unlimited articles
If you are a member. Log in here