Matching donor to recipient in liver transplantation: Relevance in clinical practice

Abstract

Achieving optimum outcomes after liver transplantation requires an understanding of the interaction between donor, graft and recipient factors. Within the cohort of patients waiting for a transplant, better matching of the donor organ to the recipient will improve transplant outcomes and benefit the overall waiting list by minimizing graft failure and need for re-transplantation. A PubMed search was conducted to identify published literature investigating the effects of donor factors such as age, gender, ethnicity, viral serology; graft factors such as size and quality, recipient factors such as age, size, gender and transplant factors such as major or minor blood group incompatibility and immunological factors. We also report technical and therapeutic modifications that can be used to manage donor-recipient mismatch identified from literature and the authors’ clinical experience. Multiple donor and recipient factors impact graft survival after liver transplantation. Appropriate matching based on donor-organ-recipient variables, modification of surgical technique and innovative peri-transplant strategies can increase the donor pool by utilizing grafts from marginal donors that are traditionally turned down.

Key Words: Liver transplantation; Donor-recipient mismatch; Immunological mismatch; Viral serology mismatch

Core tip: Multiple donor and recipient factors impact graft survival after liver transplantation. In addition, interaction between donor, graft and recipient factors may significantly affect management and outcomes. Appropriate matching based on donor-organ-recipient variables can avoid wastage of liver grafts, improve outcomes and decrease graft loss. Modification of surgical techniques and innovative peri-transplant strategies can expand the donor pool by utilizing grafts from marginal donors that are traditionally turned down.

INTRODUCTION

Transplantation involves the transfer of vascularized organs between genetically disparate individuals. Each individual has his/her distinct innate and acquired characteristics. Often combinations of these factors interact to affect outcomes in liver transplant recipients.

In view of the widening disparity between organ availability and demand, there is increasing pressure to accept all available liver grafts for transplantation. The only absolute contraindications at present are current or recent malignancy (other than some cutaneous and primary brain tumors) or infection with human immunodeficiency virus. Within the cohort of patients waiting for liver transplantation (LT), better matching of the donor organ to the recipient will improve transplant outcomes and benefit the waiting list by minimizing graft failure and need for re-transplantation. The method of choosing a recipient for a deceased donor graft varies amongst allocation systems. In most systems, patients with acute liver failure (ALF) are given priority over patients with chronic liver disease (CLD). Among patients with CLD, some systems prioritise based on MELD score while others prioritise based on waiting time. Blood group compatibility, graft-recipient size match are other factors considered in matching grafts.

This review attempts to highlight interactions between donor and recipient factors that may affect LT outcome. Ways to tailor operative technique and peri-operative management to counteract these mismatched factors is also described. Use of donation after cardiac death (DCD) donor grafts and use of steatotic grafts is not discussed in this review.

DEMOGRAPHICS FACTORS

AGE

Increasing donor age has been consistently identified as a factor affecting transplant outcome adversely[1,2]. The donor risk index (DRI)[3] is a mathematical formula that predicts the risk of liver graft loss. It is calculated using several donor and graft factors including donor age, ethnicity, donor cause of death, type of graft, etc. Donor age greater than 40 years is considered a risk factor with grafts from donors older than 70 years having a 65% increased risk of allograft failure.

Liver grafts from young adults are usually of excellent quality and should be split to provide grafts for two recipients. This maximizes graft utilization and ensures access for paediatric patients[4]. Grafts from paediatric donors are of good quality and should be prioritized to appropriately matched paediatric recipients. This ensures transplantation of these grafts into recipients who theoretically will need the graft to function for the longest duration. Full size graft paediatric transplantation into children is technically straightforward and minimizes the risks of biliary complications seen with split grafts[5]. Pediatric transplant outcomes using grafts from donors less than 6 years have similar graft survival when compared to older donors[6]. Their use for adult recipients may not provide an adequate liver mass and may be associated with vascular complications and potential risk of graft rotation and outflow problems[7].

Liver grafts from elderly donors are more fatty, with an element of fibrotic change. They may not tolerate long periods of cold ischemia. In a retrospective study of 772 adult transplants, Hoofnagle et al[8] reported increased incidence of initial graft dysfunction and early and late graft loss in transplants using grafts from donors older than 50 years. Atherosclerotic changes in the hepatic artery of grafts from elderly donors can increase the risk of arterial complications[9]. Modification of surgical techniques such as avoiding the use of the donor carrel patch for anastomosis, accessory right artery reconstruction to the gastroduodenal artery has decreased the incidence of hepatic artery thrombosis[10]. Older grafts are particularly associated with poorer outcomes in recipients with liver disease due to hepatitis C virus (HCV). In a retrospective review of 111 patients transplanted for hepatitis C related CLD, Rayhill et al[11] reported that grafts from donors over 60 years of age were associated with severe recurrent HCV. Several authors have suggested that these should be avoided in HCV recipients[12,13]. Despite these issues, older donors remain a valuable source of organs[14-16]. The risk can be minimised by careful visual assessment, routine pre-transplant biopsy to rule out fibrosis and keeping the cold ischemia time to the minimum[17]. Their use can be considered in patients with hepatocellular carcinoma who usually have stable liver disease and can tolerate a marginal graft better[18,19]. Similarly, these grafts will provide adequate graft function for elderly stable recipients who otherwise may be disadvantaged in an organ allocation system where graft placement may be biased by a utilitarian viewpoint of maximum life-years gained.

Gender

In a UNOS database analysis of over 34000 transplants, Rustgi et al[20] reported that grafts from female donors transplanted into male recipients have a 20% increased risk of graft loss as compared to gender matched male recipients. Other authors have reported no such difference in graft loss rates. Teodorescu et al[21] analyzed the differences in liver grafts from male and female donors using a dataset of 28000 transplants from the UNOS database. Their analysis suggested that female donors were shorter, older and more likely to die from cardiovascular disease as compared to their male counterparts. Once these factors were adjusted for, graft outcomes from female-to-male transplantation were similar to other gender based donor-recipient pairings.

Female recipient gender has been reported to be a risk factor in the setting of transplantation for HCV. Female recipient status increases the risk of advanced fibrosis and graft loss[22]. They are also at higher risk for treatment failure for recurrent HCV[23].

Ethnicity

Majority of studies investigating the effect of donor race on LT have originated from the United States. Studies have previously reported poorer outcomes with transplantation of non-Caucasian donor livers into Caucasian recipients. African-american (AA) race is a factor included in the calculation of the donor risk index[3]. Molenaar et al[24] investigated the effect of donor race on outcomes of liver transplant in AA race HCV recipients. They reported poorer outcomes when Caucasian donor livers were transplanted into AA recipients with HCV. This effect disappeared in transplantation for non-HCV and in Caucasian HCV recipients irrespective of donor race. The authors reported that the effect was unrelated to donor factors such as age, cause of death, weight, cytomegalovirus (CMV) status and HLA mismatch.

Wallace et al[25] investigated the relevance of ethnicity in determining graft outcomes by analyzing OPTN data from 10874 transplants. They reported that while unadjusted data showed increased risk of graft loss in transplants with livers from AA, Asia-Pacific Indians (API) and “Others” donors, the difference disappeared in the former two racial groups when other confounding factors were adjusted. Only livers from “others” sub-group (i.e., non-Caucasian, non AA, non-API donors) livers were associated with a higher risk of graft loss. The authors suggested that as over 80% of the “others” sub-group were of Hispanic ethnicity who have a higher prevalence of fatty livers; this could explain the worse post-transplant outcomes. Similarly, Chen et al[26] suggested that socio-economic parameters such as median household income were more influential than the ethnicity in affecting recipient outcomes.

Graft-recipient size mismatch

Space for the new liver graft is created by recipient hepatectomy. The musculoskeletal cage formed by the lower right hemithorax, diaphragm and vertebral column limits the size of liver graft that can be transplanted. Space consideration is particularly important in the antero-posterior dimension, where the graft right lobe will be positioned. In practice, the height and abdominal girth of the individual provide an approximate measure of the liver size.

Implantation of a large liver in a small recipient is technically difficult. Compression after wound closure can compromise graft perfusion. Presence of significant pre-operative ascites may ease the situation due to chronic stretching of the abdominal wall. Graft reduction by a right posterior sectionectomy can improve the space constraint by decreasing the antero-posterior dimensions of the graft. However, this leaves a large cut surface and is not appropriate for routine use. Implanting a large graft can also cause poor alignment in the position of the inflow structures for the graft and recipient. In these cases, it is preferable to site the caval anastomosis as high as possible on the recipient cava (piggyback) to bring the graft and recipient portal structures in alignment.

Large graft size is also an issue in transplantation of babies. Reduced left lateral segment grafts and mono-segmental grafts have been used[27,28]. Hyper-reduction of grafts to provide liver grafts for very small babies has been reported with satisfactory results[29]. Mismatch in size of hepatic arteries of the graft and paediatric recipient is another factor encountered in this situation and imaginative ways to deal with this problem have been described[30]. Wound closure may be difficult in these babies due to the relative size of the graft. Use of temporary closure with synthetic material followed by delayed closure is usually feasible[27].

Transplanting a small liver into a large adult may be technically easier as there is adequate space for the liver to be rotated during implantation. Difficulty may occur during portal and bile duct anastomosis due to wide gap between donor and recipient structures if the liver is implanted using piggyback technique. A side-to-side cavo-cavoplasty will enable the surgeon to tailor the level of venous anastomosis thus bringing the donor hilar structures down to an appropriate level for safe anastomoses. Smaller livers also have the risk of torsion in the roomier upper abdomen increasing the risk of vascular outflow complications. Caval replacement technique may be used to minimize outflow problems. Firm fixation of the falciform ligament to the diaphragm also helps in minimizing this risk.

MISMATCH IN VIRAL SEROLOGY

CMV

The world-wide prevalence of CMV IgG positivity suggestive of previous CMV infection is around 50%. The prevalence is highest in the developing world where seropositivity rates can reach 90%[31]. New infection or re-activation of previously acquired CMV infection is a significant cause of morbidity in the post-transplant setting.

In current clinical practice, donor-recipient pairs are classified as high-risk and low-risk groups based on the recipient’s risk of developing CMV disease (Table 1). In addition, patients needing induction immunosuppression or steroid boluses for acute rejection are also considered high-risk. Use of CMV prophylaxis varies with some units favoring it for high-risk recipients only while others use it universally. Two strategies for CMV prophylaxis are available[32]. The more common strategy is oral prophylaxis for all high-risk transplants for three months. Use of universal CMV prophylaxis has been reported to decrease the risk of CMV disease in the early post-transplant period. While CMV infection and disease can occur beyond three months, the patient is on less intense immunosuppression and the risk of serious CMV disease is lower. An alternative means is monitoring of CMV titres using periodic assays for CMV viremia. Treatment for CMV disease is instituted when the titres reach a pre-determined level[33]. This is applicable in low risk recipients.

Table 1 Risk stratification and need for cytomegalovirus prophylaxis.

Donor CMV IgG status	Recipient CMV IgG status	Risk type1	Need for CMV prophylaxis
Positive	Positive	Low risk	Low
Positive	Negative	High risk	High
Negative	Positive	Low risk	Low
Negative	Negative	Low risk	Low

¹Patients needing induction with antithymocyte globulin and OKT3 or steroid boluses for acute rejection are considered high-risk irrespective of their CMV mismatch status. CMV: Cytomegalovirus.

Hepatitis B virus

Hepatitis B virus (HBV) related liver disease (ALF, CLD or hepatocellular carcinoma) is an important indication for LT. Outcomes of transplantation for HBV have greatly improved over the years[34]. Post-transplant graft re-infection with HBV depends on donor hepatitis B core antibody status, pre-transplant HBV DNA titre, indication for transplant (ALF or CLD) and the use of oral anti-viral therapy and hepatitis B immunoglobulin (HBIG) after transplantation[35]. Almost all these patients need life-long anti-viral therapy in the post-transplant period. HBIG is now primarily indicated for patients with high pre-operative HBV DNA titres[36].

Individuals who recover from an acute HBV infection develop antibodies for the hepatitis B core antigen. Use of the core antibody donor livers was previously avoided except in HBsAg positive recipients. With increasing demand for donor organs, grafts from core antibody positive donors are being used in many centres. This is particularly pertinent in countries where more than 50% of living donors are core antibody positive[37]. Safety of using core antibody positive grafts has been confirmed by several retrospective studies[38,39]. Development of de novo HBV infection is the main concern in this situation and the risk depends on the recipient’s prior exposure to HBV, immunization status and use of prophylaxis[40] (Table 2). The ideal prophylactic therapy is unclear with some centres using HBIG based regimens[41] while others have used oral antiviral-based regimens[42,43]. We maintain our patients on anti-viral therapy alone due to cost considerations. Use of these grafts needs a careful discussion with the potential recipient regarding the risk of de novo HBV infection and the cost of additional prophylaxis.

Table 2 Risk of de novo hepatitis B in recipients receiving grafts from hepatitis B core antibody positive donors.

Recipient status	Recipient HBVantibody status	Risk of de novo hepatitis
		No prophylaxis	With prophylaxis
HBV naive	HBcAb-, HBsAb-	58%	11%
No past infection, immunized	HBcAb-, HBsAb+	18%	2%
Past infection, immune	HBcAb+, HBsAb+	4%	3%
Past infection, not immune	HBcAb-, HBsAb+	14%	3%

Incidence in patients with or without prophylaxis is shown (Modified from Skagen et al[40]). HBV: Hepatitis B virus.

Another strategy to decrease the risk of de novo HBV is by active immunization with HBV vaccine[44]. Prospective studies have shown that both pre-transplant and post-transplant vaccination are effective in preventing de novo HBV infection, though additional doses of the vaccine may be required to induce an effective immune response[44,45]. Our current recommendation for patients with non-HBV related liver disease is to be immunized for HBV. This provides protection against new HBV infection before transplant and decreases the risk of de novo HBV infection if the patient receives a core antibody positive graft.

HBsAg positive donor grafts are not routinely used even in recipients with HBV related liver disease due to the risk of early graft damage. It is also contraindicated in individuals who have concurrent Hepatitis Delta virus infection. However these may be used in life threatening situations such as ALF or HBV related HCC where delay may make these cases untransplantable. Two small retrospective studies have reported the safe use of HBsAg positive liver grafts in patients with HBV related CLD[46] and HBV unrelated CLD[47] with satisfactory results. Both groups have suggested that long-term HBIG prophylaxis may not be effective and advised institution of double anti-viral therapy as prophylaxis. Careful assessment of graft quality (fibrosis and inflammation on biopsy and serum enzyme levels) is essential to avoid transplanting chronically damaged grafts in this setting.

Hepatitis C virus

HCV infection of the new liver graft after transplantation for HCV related liver disease is nearly universal. The rate and severity of graft damage due to HCV is however variable and has been found to depend on several donor and graft related factors. High viral titres, older donors, inflammation and fibrosis on graft biopsy, prolonged cold ischemia time and more intense immunosuppression have been associated with poorer outcomes in HCV patients[48,49].

Grafts from HCV seropositive donors are not routinely utilized due to concern regarding transmission of the infection to recipients. Use of grafts from HCV antibody positive donors has been suggested as a way to improve access to transplantation for HCV related liver disease patients. Several case-control studies have suggested equivalent results in terms of frequency and severity of HCV recurrence, graft and patient survival[50,51]. Though the patient numbers in these studies are small and they are all retrospective studies, the evidence is promising. It is unlikely that a clinical trial comparing outcomes with HCV infected or uninfected grafts can ever be organized for ethical reasons. When a HCV positive graft is considered for transplantation, a pre-transplant biopsy is necessary to ensure no significant hepatitis or fibrosis. A detailed discussion with the potential recipient is also mandatory. The significance of donor viral load and co-infection with two different genotypes of HCV on transplant outcome is presently unclear[52].

IMMUNOLOGICAL MISMATCH

ABO incompatible liver transplants

Most liver transplants are either between ABO identical or ABO compatible donor-recipient pairs. Earlier studies had reported increased risk of humoral and cellular rejection, arterial thrombosis and biliary complications after ABO incompatible liver transplants (ABOiLT)[53].

Measures such as peri-operative plasmapheresis[54], and rituximab[55] have been used to lower peri-operative recipient antibody levels and thereby decrease the risk of these complications. ABOiLT in the pediatric population has been used more frequently and outcomes similar to ABO compatible LT have been achieved[56]. Its role in adult transplantation is still unclear though it remains an option in desperate situations like fulminant hepatic failure when an ABO compatible organ is unavailable[57,58]. Re-transplantation will be required in some of these patients.

Use of blood group A2 donors

Around 10% of all blood group A individuals can be sub-typed as A2. A2 sub-group patients have lower expression of A antigen on their RBCs and are hence less likely to undergo immune mediated haemolysis on coming in contact with serum containing anti-A antibodies (present in serum of blood group B and O patients). This fact has been exploited in the use of A2 grafts for O and B recipients and use of A2B grafts for B recipients. While this strategy has been regularly used in kidney transplantation, the first large series in deceased donor LT has recently been published[59].

Minor ABO incompatibility

Blood group-O is considered as universal donor and O-group grafts are considered for patients of all blood groups in acute situations. Transplantation of grafts from O donors to A, B, AB blood groups has been reported to cause haemolysis. This occurs due to the passenger lymphocyte syndrome where donor lymphocytes transferred via the graft produce antibodies against A and B antigens on the recipient red cells[60]. These antibodies cause recipient RBC haemolysis by fixing complement. This is a self-limited phenomenon as the donor lymphocytes gradually die out and the haemolysis stops. This has been most commonly reported in O to A transplants though it can occur in O to B, O to AB or minor blood group incompatibilities[61].

Rhesus factor mismatch

Rh factor is usually not considered significant in matching organs for transplantation. Bryan et al[62] investigated the effect of Rh mismatch on the outcome of kidney transplantation. They reported poorer 7-year graft survival in cases of Rh mismatched transplantation. Ashkenazi et al[63] reported Rh mismatched transplantation as a significant risk factor for biliary complications after LT. Anecdotal reports of severe haemolysis or graft versus host disease have reported[64,65]. However, Rh-mismatch in LT is not taken into consideration for organ matching in most centres.

HLA matching

The role of HLA matching across the A, B, and DR loci in kidney transplantation is well established. Lymphocytic cross-match prior to LT is not routinely used. One reason is that cross-match takes 4-5 h to complete, which can increase the cold ischemia time of the graft.

Balan et al[66] investigated the effect of HLA mismatching on outcomes in 799 patients undergoing LT. They reported poorer 10-year survival for recipients receiving grafts with HLA-A locus mismatch. Similarly a mismatch at HLA-DR locus was found to increase recurrence of auto-immune liver disease. Lan et al[67] conducted a meta-analysis investigating the role of HLA matching in LT. They found that increasing number of mismatches was associated with increased risk of acute rejection though the graft survival rates were similar. In contrast, Muroetal[51] analysed data from 242 liver transplants and reported that matching at HLA-A locus increased the risk of graft failure. There is hence a lack of clarity regarding the relevance of HLA matching in LT. There is no current recommendation regarding its routine use in matching liver grafts either in the DDLT or LDLT settings.

Immunological pre-sensitisation

Liver graft has the capacity to absorb large quantities of antibodies and hence pre-sensitization, which is a risk factor for poor outcome in kidney transplantation is not relevant in LT. Hyper-acute rejection caused by preformed antibodies against donor antigens is very uncommon in LT. In fact, simultaneous liver or split LT along with a kidney transplant in a highly sensitized recipient protects the renal graft from immune damage[68].

HLA matching and graft versus host disease

Graft versus host disease (GVHD) is a rare complication after LT occurring in around 0.5%-1% of recipients[69]. It is associated with high mortality due to complications of bleeding, sepsis and multiple organ failure. Close HLA matching has been reported as a risk factor for GVHD in LT. Soejima et al[70] described six cases of GVHD after LT where the donor was homozygous at HLA A,B and DR loci with one haplotype match with the recipient leading HLA mismatch of 1000. This association of donor-dominant one-way HLA matching in the 3 loci of HLA-A, -B, and -DR with GVHD has also been confirmed from other studies[71]. This is of particular relevance in the setting of paediatric living donor LT where a parent might be donating to the child.

Glutathione S-Transferase T1 genotype mismatch and de novo auto-immune liver disease

Glutathione S-Transferase (GST) is an enzyme present in the liver and kidneys and is involved in drug metabolism. T1 genotype of this enzyme is absent in around 20% of Caucasian population. Patients who do not have the GSTT1 gene (GSTT1 null) can develop antibodies against GSTT1 when a GSTT1 positive donor liver is transplanted. In 2004, Aguilera et al[72] reported the relation between GSTT1 donor-recipient mismatch and the incidence of de novo auto-immune hepatitis (AIH). Further work by this group has found an association between high titres of anti-GSTT1 antibodies and development of de novo hepatitis. They reported that while anti-GSTT1 antibodies are present in 6.9% of patients with a mismatched graft, the incidence of de novo AIH in patients with antibodies was 60% at 36 mo post-transplant. The incidence of antibodies is lower in patients maintained on tacrolimus[72]. Its clinical significance in the pre-transplant setting is controversial. Knowledge of the mismatch may help in management of post-transplant immunosuppression as these patients may benefit from tacrolimus and long-term low dose steroids as part of their immunosuppression protocol.

CONCLUSION

Multiple donor and recipient factors impact graft survival after LT. Appropriate matching based on donor-organ-recipient variables can improve outcomes and decrease graft loss. Modification of surgical techniques and innovative peri-transplant strategies can increase the donor pool by utilizing grafts from marginal donors that are traditionally turned down.