Challenges in Living Donor L i v e r Tr a n s p l a n t a t i o n James F. Trotter,
MD
KEYWORDS Living donor liver transplantation Hepatocellular carcinoma ABO-incompatible transplant Living donor KEY POINTS Living donor liver transplantation (LDLT) is a procedure that accounts for approximately 3% of adult liver transplants in the United States. The enthusiasm toward this operation has waned in recent years. Although there is no apparent survival advantage for LDLT recipients with hepatocellular carcinoma, properly selected candidates may benefit from the shorter waiting time compared with deceased donor liver transplantation (DDLT). The publication of recent protocols with ABO-incompatible LDLT suggests that this barrier may be successfully overcome to expand the potential living donor pool.
INTRODUCTION
Adult-to-adult LDLT is a procedure that has evolved over the past 2 decades. First introduced in the United States in the 1990s, LDLT was primarily relegated to pediatric recipients until late in the decade. Then, a combination of factors contributed to a proliferation of cases. Waiting times for liver transplant increased in the late 1990s as the number of patients listed for transplant far exceeded the modest gain in deceased donors. In addition, important changes in the operative technique improved LDLT recipient outcomes. The initial experience with LDLT used the smaller left hepatic lobe. Although this small graft was adequate for diminutive (pediatric) recipients, initial results in adults were poor. In the late 1990s, selected centers demonstrated favorable recipient outcomes by transplanting the larger right hepatic lobe.1,2 As the advantage of right hepatic lobe LDLT became apparent, the popularity of the procedure increased and the annual number of LDLTs increased from fewer than 100 to more than 500 in 2002, accounting for approximately 10% of adult liver transplants in the United States.3 The application of LDLT over the past decade, however, has dropped substantially to fewer than 200 adult cases per year, representing only approximately
Disclosure: None. Department of Medicine, Baylor University Medical Center, 3410 Worth Street, #860, Dallas, TX 75246, USA E-mail address: [email protected] Clin Liver Dis 18 (2014) 651–660 http://dx.doi.org/10.1016/j.cld.2014.05.007 1089-3261/14/$ – see front matter Ó 2014 Elsevier Inc. All rights reserved.
liver.theclinics.com
652
Trotter
3% of all liver transplants. The reasons for the decline of LDLT are not entirely understood but are likely due to a combination of forces.4 As with any novel procedure, there is initial enthusiasm leading to rapid growth followed by a more measured approach as the full spectrum of risks and complications becomes apparent over time. Such is the case with LDLT. Over the past decade, there have been several publications highlighting complications in donors and recipients (discussed later), which has tempered interest in the procedure. The most important complication of LDLT, donor death, has received widespread media attention, although its occurrence is rare, at just over 1/ 500. Finally, federal regulators have placed transplant centers under increasing scrutiny for favorable outcomes. Consequently, transplant centers have become more risk averse and this may have had an impact on their decision to offer LDLT to their patients. The trend toward limited application of LDLT in the United States is largely reflective of the European experience, where living liver donor rates are approximately 1 donor per million (dpm) population. In some parts of the world, however, the procedure is thriving; most notable is South Korea, with 17 dpm, the highest rate worldwide, followed by Turkey (8 dpm), Egypt (5 dpm), and Japan (4 dpm). The Asan Medical Center in Seoul, South Korea, performs approximately 300 LDLTs per year surpassing the entire US volume by approximately 2-fold. This review focuses on 3 of the most important developments in LDLT in recent years: hepatocellular carcinoma (HCC), ABOincompatible transplant, and donor risk and its management. LDLT FOR HEPATOCELLULAR CARCINOMA
Compared with DDLT, LDLT offers the potential advantages of speed and timing, which can be particularly important for patients with HCC. The average living donor evaluation takes approximately 6 to 8 weeks; so, LDLT can often be performed faster than DDLT, where waiting times are months to a few years for HCC patients. Prolonged pretransplant waiting times increase the risk of tumor progression, which, in turn, increases the risk of removal from the DDLT list and posttransplant recurrence. Up to 20% of HCC patients are removed from the list due to disease progression while awaiting a transplant.5–7 Therefore, rapid procession to transplantation potentially offers a therapeutic advantage in the treatment of HCC. Despite the theoretic advantage of LDLT for HCC patients, however, 3 separate reports have each concluded that there is no survival advantage for LDLT patients. A study from Toronto, which has a robust LDLT program, reported no survival advantage with LDLT.8 It compared survival and HCC recurrence rates for 345 transplant recipients after LDLT, 58 (17%), and DDLT, 287 (83%), over a 16-year period. As expected, the LDLT recipients had significantly shorter waiting times compared with DDLT (3.1 vs 5.3 months; P 5 .003). There was no difference in 5-year HCC recurrence rates for LDLT (15%) and DDLT (17%) for the DDLT group (P 5 not significant [NS]). There was also no difference in 5-year survival rates for LDLT (75%) and DDLT (75%) (P 5 NS). Similar results were reported from a French group in 183 patients with HCC, with LDLT (n 5 36) and DDLT (n 5 147).9 At listing, patient and tumor characteristics were comparable in the 2 groups, whereas the mean waiting time was shorter with LDLT (2.6 months) compared with DDLT (7.9 months) (P 5 .001). All of the 27 (18%) of patients who dropped off the list, primarily for tumor progression, prior to transplant were listed for DDLT. There was no difference in posttransplant recurrence rates, however, between the 2 groups, at 13% each (P 5 NS). More important, there was no difference in survival on an intention-to-treat basis. Finally, the Adult-to-Adult Living Donor Liver Transplantation Cohort Study Group (A2ALL) has published a large (n 5 229) intention-to-treat analysis evaluating LDLT and DDLT in HCC patients who had at least 1 potential donor
Challenges in Living Donor Liver Transplantation
evaluated.10 The LDLT cohort included patients undergoing this procedure, whereas the remainder comprised the DDLT cohort and included patients whose donor was rejected or those receiving DDLT before completion of donor evaluation. Predictably, the LDLT patients had a shorter time to transplant, were more likely to have tumors exceeding Milan criteria, and had higher a-fetoprotein (AFP) levels. As a result, the unadjusted risk of HCC recurrence was more than 3-fold higher with LDLT (38%) compared with DDLT (11%, P 5 0.0004), as was the adjusted risk (hazard ratio [HR] 2.35; 95% CI, 1.04–5.35; P 5 .04). Perhaps the most important finding is the lack of a survival advantage with LDLT; the adjusted risk of death from the time of donor evaluation was similar for patients who received LDLT compared with other patients (HR 0.73; 95% CI, 0.36–1.45; P 5 .36). Two meta-analyses that evaluated the efficacy of LDLT compared with DDLT in HCC patients reiterated these results.11,12 How can these data be interpreted related to the selection of LDLT candidates?13 For advocates of LDLT, these data provide a basis for its continued application; the outcomes for LDLT recipients are as good as for DDLT. For selected HCC patients, LDLT could be their best option. Small patients (who have inherent difficulty securing a size-appropriate deceased donor) in high–Model for End-Stage Liver Disease (MELD) regions (with long waiting times for HCC patients) may benefit from the predictably shorter waiting times associated with LDLT. An additional consideration with LDLT, however, is the impact on the donor, who undergoes a major operation with no medical benefit. Therefore, a more critical assessment is that LDLT could be worse than DDLT for HCC patients. With equivalent recipient survival and potential harm to the donor, LDLT has more potential for medical harm with no apparent additional benefit. There is likely a middle ground, however, between these 2 extremes. One of the reasons for the absence of benefit with LDLT is the high priority awarded to DDLT candidates with HCC under MELD-based liver allocation, which was implemented in 2002. As a result, HCC patients could initially be transplanted in less than 90 days with a DDLT, often faster than an LDLT can even be arranged. Over the past decade, however, the waiting times for HCC patients have increased, even with additional MELD points awarded over time. At many US centers, the waiting times for DDLT exceed 1 year for HCC patients. Therefore, the lack of benefit for LDLT found in these studies (which evaluated patients transplanted from years ago) may not be applicable to current liver transplant candidates. Another reason for the lack of benefit with LDLT could be that HCC patients who are selected for LDLT have worse tumor biology than DDLT patients. The diagnosis of HCC creates understandable anxiety for patients, their families (of potential donors), and the transplant team. The possibility of offering an expedited LDLT places the transplant team under pressure to “make the transplant happen.” These powerful forces may have an impact on the selection of donors and recipients for LDLT. Data from the A2ALL study provide clear evidence that LDLT recipients have worse tumor biology than DDLT patients. LDLT recipients have significantly higher AFP, more advanced tumor stage, large tumor size, and higher percentage over Milan and University of California San Francisco criteria.10 Therefore, LDLT recipients may be fundamentally different from DDLT patients, thereby explaining their worse outcomes after transplant. In summary, transplant physicians need to be cautious in offering LDLT to HCC patients. In particular, patients with poor tumor biology (higher AFP and advanced stage disease) may be poor candidates for LDLT, despite the opportunity to offer a desperate patient a potentially life-saving operation. LDLT may be effective, however, in patients with a long anticipated waiting time, in particular, small recipients with favorable tumor characteristics listed in donor service areas with high MELD scores.
653
654
Trotter
ABO-INCOMPATIBLE LDLT
There have been several recent developments to overcome donation barriers related to ABO incompatiblity, including medical treatment protocols as well as innovative strategies of matching donors and recipients. Historically, liver transplantation with ABO incompatibility between the donor and recipient has been associated with poor outcomes.14 Recipients of ABO-incompatible donor livers have antibodies directed against ABO antigens that are expressed on many cells, including the endothelium of blood vessels in the donor liver. As a result, such recipients are prone to developing diffuse hepatic vasculitis leading to biliary complications and graft loss.15 Until recently, attempts to reduce the circulating antibodies with plasmapheresis, splenectomy, and infusional prostacyclin (to prevent hepatic inflammation with) have largely been unsuccessful.16,17 ABO-incompatible recipients have lower patient and graft survival rates with significantly higher biliary complications compared with ABO-identical or -compatible recipients. There have been several publications, however, of successful implementation of innovative treatments to reduce circulating antibodies.17–28 These protocols have all used Rituximab, which is an antibody directed against the cell surface marker, CD20, expressed on plasma cells, the source of the destructive antibodies. These novel therapies for ABO-incompatible LDLT have largely been pioneered in Japan, where DDLT is rarely performed. The Japanese are willing to pursue a more aggressive approach toward LDLT because it is the only transplant option for most patients. One of the most comprehensive reports is a multicenter experience from Japan, which underscored several important findings.29 First, the number of ABO-incompatible LDLT has increased by approximately 10-fold from before 2000, when fewer than 5 yearly cases were done, to approximately 50 cases per year in 2011. The increased number of cases is likely due to the newfound success of the procedure. Rituximab treatment is given between 1 and 66 days (usually 1 to 2 weeks) before the anticipated LDLT at a dose of 375 mg/m2 to 500 mg/m2. The administration of Rituximab was associated with a significantly lower incidence of antibody-mediated rejection (6%) than in the untreated group (23%), (P
MD
KEYWORDS Living donor liver transplantation Hepatocellular carcinoma ABO-incompatible transplant Living donor KEY POINTS Living donor liver transplantation (LDLT) is a procedure that accounts for approximately 3% of adult liver transplants in the United States. The enthusiasm toward this operation has waned in recent years. Although there is no apparent survival advantage for LDLT recipients with hepatocellular carcinoma, properly selected candidates may benefit from the shorter waiting time compared with deceased donor liver transplantation (DDLT). The publication of recent protocols with ABO-incompatible LDLT suggests that this barrier may be successfully overcome to expand the potential living donor pool.
INTRODUCTION
Adult-to-adult LDLT is a procedure that has evolved over the past 2 decades. First introduced in the United States in the 1990s, LDLT was primarily relegated to pediatric recipients until late in the decade. Then, a combination of factors contributed to a proliferation of cases. Waiting times for liver transplant increased in the late 1990s as the number of patients listed for transplant far exceeded the modest gain in deceased donors. In addition, important changes in the operative technique improved LDLT recipient outcomes. The initial experience with LDLT used the smaller left hepatic lobe. Although this small graft was adequate for diminutive (pediatric) recipients, initial results in adults were poor. In the late 1990s, selected centers demonstrated favorable recipient outcomes by transplanting the larger right hepatic lobe.1,2 As the advantage of right hepatic lobe LDLT became apparent, the popularity of the procedure increased and the annual number of LDLTs increased from fewer than 100 to more than 500 in 2002, accounting for approximately 10% of adult liver transplants in the United States.3 The application of LDLT over the past decade, however, has dropped substantially to fewer than 200 adult cases per year, representing only approximately
Disclosure: None. Department of Medicine, Baylor University Medical Center, 3410 Worth Street, #860, Dallas, TX 75246, USA E-mail address: [email protected] Clin Liver Dis 18 (2014) 651–660 http://dx.doi.org/10.1016/j.cld.2014.05.007 1089-3261/14/$ – see front matter Ó 2014 Elsevier Inc. All rights reserved.
liver.theclinics.com
652
Trotter
3% of all liver transplants. The reasons for the decline of LDLT are not entirely understood but are likely due to a combination of forces.4 As with any novel procedure, there is initial enthusiasm leading to rapid growth followed by a more measured approach as the full spectrum of risks and complications becomes apparent over time. Such is the case with LDLT. Over the past decade, there have been several publications highlighting complications in donors and recipients (discussed later), which has tempered interest in the procedure. The most important complication of LDLT, donor death, has received widespread media attention, although its occurrence is rare, at just over 1/ 500. Finally, federal regulators have placed transplant centers under increasing scrutiny for favorable outcomes. Consequently, transplant centers have become more risk averse and this may have had an impact on their decision to offer LDLT to their patients. The trend toward limited application of LDLT in the United States is largely reflective of the European experience, where living liver donor rates are approximately 1 donor per million (dpm) population. In some parts of the world, however, the procedure is thriving; most notable is South Korea, with 17 dpm, the highest rate worldwide, followed by Turkey (8 dpm), Egypt (5 dpm), and Japan (4 dpm). The Asan Medical Center in Seoul, South Korea, performs approximately 300 LDLTs per year surpassing the entire US volume by approximately 2-fold. This review focuses on 3 of the most important developments in LDLT in recent years: hepatocellular carcinoma (HCC), ABOincompatible transplant, and donor risk and its management. LDLT FOR HEPATOCELLULAR CARCINOMA
Compared with DDLT, LDLT offers the potential advantages of speed and timing, which can be particularly important for patients with HCC. The average living donor evaluation takes approximately 6 to 8 weeks; so, LDLT can often be performed faster than DDLT, where waiting times are months to a few years for HCC patients. Prolonged pretransplant waiting times increase the risk of tumor progression, which, in turn, increases the risk of removal from the DDLT list and posttransplant recurrence. Up to 20% of HCC patients are removed from the list due to disease progression while awaiting a transplant.5–7 Therefore, rapid procession to transplantation potentially offers a therapeutic advantage in the treatment of HCC. Despite the theoretic advantage of LDLT for HCC patients, however, 3 separate reports have each concluded that there is no survival advantage for LDLT patients. A study from Toronto, which has a robust LDLT program, reported no survival advantage with LDLT.8 It compared survival and HCC recurrence rates for 345 transplant recipients after LDLT, 58 (17%), and DDLT, 287 (83%), over a 16-year period. As expected, the LDLT recipients had significantly shorter waiting times compared with DDLT (3.1 vs 5.3 months; P 5 .003). There was no difference in 5-year HCC recurrence rates for LDLT (15%) and DDLT (17%) for the DDLT group (P 5 not significant [NS]). There was also no difference in 5-year survival rates for LDLT (75%) and DDLT (75%) (P 5 NS). Similar results were reported from a French group in 183 patients with HCC, with LDLT (n 5 36) and DDLT (n 5 147).9 At listing, patient and tumor characteristics were comparable in the 2 groups, whereas the mean waiting time was shorter with LDLT (2.6 months) compared with DDLT (7.9 months) (P 5 .001). All of the 27 (18%) of patients who dropped off the list, primarily for tumor progression, prior to transplant were listed for DDLT. There was no difference in posttransplant recurrence rates, however, between the 2 groups, at 13% each (P 5 NS). More important, there was no difference in survival on an intention-to-treat basis. Finally, the Adult-to-Adult Living Donor Liver Transplantation Cohort Study Group (A2ALL) has published a large (n 5 229) intention-to-treat analysis evaluating LDLT and DDLT in HCC patients who had at least 1 potential donor
Challenges in Living Donor Liver Transplantation
evaluated.10 The LDLT cohort included patients undergoing this procedure, whereas the remainder comprised the DDLT cohort and included patients whose donor was rejected or those receiving DDLT before completion of donor evaluation. Predictably, the LDLT patients had a shorter time to transplant, were more likely to have tumors exceeding Milan criteria, and had higher a-fetoprotein (AFP) levels. As a result, the unadjusted risk of HCC recurrence was more than 3-fold higher with LDLT (38%) compared with DDLT (11%, P 5 0.0004), as was the adjusted risk (hazard ratio [HR] 2.35; 95% CI, 1.04–5.35; P 5 .04). Perhaps the most important finding is the lack of a survival advantage with LDLT; the adjusted risk of death from the time of donor evaluation was similar for patients who received LDLT compared with other patients (HR 0.73; 95% CI, 0.36–1.45; P 5 .36). Two meta-analyses that evaluated the efficacy of LDLT compared with DDLT in HCC patients reiterated these results.11,12 How can these data be interpreted related to the selection of LDLT candidates?13 For advocates of LDLT, these data provide a basis for its continued application; the outcomes for LDLT recipients are as good as for DDLT. For selected HCC patients, LDLT could be their best option. Small patients (who have inherent difficulty securing a size-appropriate deceased donor) in high–Model for End-Stage Liver Disease (MELD) regions (with long waiting times for HCC patients) may benefit from the predictably shorter waiting times associated with LDLT. An additional consideration with LDLT, however, is the impact on the donor, who undergoes a major operation with no medical benefit. Therefore, a more critical assessment is that LDLT could be worse than DDLT for HCC patients. With equivalent recipient survival and potential harm to the donor, LDLT has more potential for medical harm with no apparent additional benefit. There is likely a middle ground, however, between these 2 extremes. One of the reasons for the absence of benefit with LDLT is the high priority awarded to DDLT candidates with HCC under MELD-based liver allocation, which was implemented in 2002. As a result, HCC patients could initially be transplanted in less than 90 days with a DDLT, often faster than an LDLT can even be arranged. Over the past decade, however, the waiting times for HCC patients have increased, even with additional MELD points awarded over time. At many US centers, the waiting times for DDLT exceed 1 year for HCC patients. Therefore, the lack of benefit for LDLT found in these studies (which evaluated patients transplanted from years ago) may not be applicable to current liver transplant candidates. Another reason for the lack of benefit with LDLT could be that HCC patients who are selected for LDLT have worse tumor biology than DDLT patients. The diagnosis of HCC creates understandable anxiety for patients, their families (of potential donors), and the transplant team. The possibility of offering an expedited LDLT places the transplant team under pressure to “make the transplant happen.” These powerful forces may have an impact on the selection of donors and recipients for LDLT. Data from the A2ALL study provide clear evidence that LDLT recipients have worse tumor biology than DDLT patients. LDLT recipients have significantly higher AFP, more advanced tumor stage, large tumor size, and higher percentage over Milan and University of California San Francisco criteria.10 Therefore, LDLT recipients may be fundamentally different from DDLT patients, thereby explaining their worse outcomes after transplant. In summary, transplant physicians need to be cautious in offering LDLT to HCC patients. In particular, patients with poor tumor biology (higher AFP and advanced stage disease) may be poor candidates for LDLT, despite the opportunity to offer a desperate patient a potentially life-saving operation. LDLT may be effective, however, in patients with a long anticipated waiting time, in particular, small recipients with favorable tumor characteristics listed in donor service areas with high MELD scores.
653
654
Trotter
ABO-INCOMPATIBLE LDLT
There have been several recent developments to overcome donation barriers related to ABO incompatiblity, including medical treatment protocols as well as innovative strategies of matching donors and recipients. Historically, liver transplantation with ABO incompatibility between the donor and recipient has been associated with poor outcomes.14 Recipients of ABO-incompatible donor livers have antibodies directed against ABO antigens that are expressed on many cells, including the endothelium of blood vessels in the donor liver. As a result, such recipients are prone to developing diffuse hepatic vasculitis leading to biliary complications and graft loss.15 Until recently, attempts to reduce the circulating antibodies with plasmapheresis, splenectomy, and infusional prostacyclin (to prevent hepatic inflammation with) have largely been unsuccessful.16,17 ABO-incompatible recipients have lower patient and graft survival rates with significantly higher biliary complications compared with ABO-identical or -compatible recipients. There have been several publications, however, of successful implementation of innovative treatments to reduce circulating antibodies.17–28 These protocols have all used Rituximab, which is an antibody directed against the cell surface marker, CD20, expressed on plasma cells, the source of the destructive antibodies. These novel therapies for ABO-incompatible LDLT have largely been pioneered in Japan, where DDLT is rarely performed. The Japanese are willing to pursue a more aggressive approach toward LDLT because it is the only transplant option for most patients. One of the most comprehensive reports is a multicenter experience from Japan, which underscored several important findings.29 First, the number of ABO-incompatible LDLT has increased by approximately 10-fold from before 2000, when fewer than 5 yearly cases were done, to approximately 50 cases per year in 2011. The increased number of cases is likely due to the newfound success of the procedure. Rituximab treatment is given between 1 and 66 days (usually 1 to 2 weeks) before the anticipated LDLT at a dose of 375 mg/m2 to 500 mg/m2. The administration of Rituximab was associated with a significantly lower incidence of antibody-mediated rejection (6%) than in the untreated group (23%), (P
