Many patients with end-stage liver disease have to undergo living donor liver transplantation (LDLT) in Japan because there is a shortage of cadaveric organ donation. It is often difficult for recipients to find a suitable ABO-identical or -compatible living-related donor for LDLT. An ABO-incompatible (ABO-I) living-related donor is often the only way to save a patient with end-stage liver disease (1). ABO-I liver transplantation was considered a relative contraindication because of severe anti-blood type isoagglutinin-mediated rejection that resulted in graft necrosis (2, 3). However, the outcomes of ABO-I LDLT have dramatically improved since the introduction of rituximab, an anti-CD20 antibody (4–8). A new protocol for ABO-I LDLT using rituximab, intravenous immunoglobulin (IVIG), and plasma exchange (PE) without graft local infusion treatment has been established and successfully applied with 100% graft survival (9), which was superior to the recently reported results (4). Despite the better short outcomes of the ABO-I LDLT patients, the mid- and long-term outcomes of ABO-I LDLT patients, the kinetics of anti-blood type isoagglutinin titers, the restoration of B lymphocytes, and the impact of this protocol on defense systems to various pathogen remain to be elucidated. The aim of this study was to retrospectively review the data of ABO-I patients to determine these factors.
Overall Outcomes of 15 Patients of ABO-I LDLT
All but 1 of the 15 patients were alive after ABO-I LDLT at the last follow-up (follow-up time ranging from 5 months to 5 years 6 months). One patient died of accidental drowning 9 months after LDLT. There was neither in-hospital mortality nor graft loss. None of these patients experienced any septic complications after ABO-I LDLT. High-dose IVIG was used in six patients with suspected anti-blood type isoagglutinin-mediated rejection (cases 2, 3, 4, 5, 14, and 15). Biopsy-proven acute cellular rejection occurred in four patients (case 1 at 1044 posttransplant days [postoperative day (POD)], case 2 at 650 POD, case 3 at 104 POD, and case 8 at 117 POD), all of which responded to steroid pulse therapy.
Kinetics of Anti-Blood Type Isoagglutinin Titers Before and After ABO-I LDLT
Fig. 1 and SDC 1 (see Figure,http://links.lww.com/TP/A515) show that anti-blood type isoagglutinin titers ranged between 23 and 212 before PEs, and these titers rapidly decreased after PEs. Anti-blood type isoagglutinin titers were successfully decreased to 26 or less in all but cases 14 and 15 at the time of LDLT. Immediate posttransplant rebound of these titers was observed in the cases 1 and 2 (rebound is defined as titer increasing up to or beyond the pretreatment titer). Rebound titers were observed for anti-donor blood type A isoagglutinin in the case 2 and for non-anti-donor blood type B isoagglutinin. Anti-blood type isoagglutinin titers after ABO-I LDLT were generally lower than each respective pretreatment titer. Both anti-blood types A and B isoagglutinin titers were measured in the cases 2, 13, 14, and 15 with blood type O. This revealed both anti-A and -B isoagglutinin titers generally lower than each pretreatment titer irrespective of the donor's blood type, although the anti-donor blood type isoagglutinin titers never reached 20. Anti-donor blood type isoagglutinin titers in the cases 2, 13, and 14 were constantly lower than each respective non-anti-donor blood type isoagglutinin titer. All but the hepatitis C patients maintained good liver function tests despite the sustained presence of anti-donor blood type isoagglutinin titers during the mid- and long-term posttransplant periods (hepatitis C patients had fluctuations in their liver function tests caused by recurrent hepatitis C).
Catheter-related sepsis occurred after PEs in the cases 4 and 14, and their scheduled LDLTs were postponed. The anti-blood type isoagglutinin titers rapidly returned to each respective pretreatment titer during the treatment of sepsis, and further PEs were needed to lower the anti-blood type isoagglutinin titers just before the rescheduled LDLTs.
Expression of Blood A/B Antigens on Pretransplant and Posttransplant Liver Biopsies
Time-zero biopsies revealed the respective blood type antigens for each donor (Fig. 1; see Figure, SDC 1,http://links.lww.com/TP/A515 and Fig. 2). Positive immunostaining was mainly observed in the endothelial cells and sinusoidal cells. Posttransplant liver biopsies continued to express the blood type antigens of the donor (Fig. 2).
Restoration of B Lymphocytes After Rituximab Treatment
Seventy samples were available for flow cytometry of the peripheral blood. CD19-positive lymphocytes rapidly disappeared in the peripheral blood after rituximab treatment (Fig. 3). Restoration of CD19-positive lymphocytes started approximately 6 months after the rituximab treatment.
High Incidence of Cytomegalovirus Antigenemia in ABO-I Living Donor Liver Transplant Patients
The incidence of cytomegalovirus (CMV) antigenemia (defined as antigen detection of 10 or more in 150,000 white cells) was higher in the ABO-I group (57.1%) than the control group (11.1%; P=0.02); however, none of these patients became symptomatic and were easily treated by ganciclovir.
Accelerated Hepatitis C Viremia in ABO-I Living Donor Liver Transplant Hepatitis C-Positive Patients
All seven hepatitis C virus (HCV)-RNA-positive patients developed biopsy-proven recurrence of hepatitis C (case 4 at 55 POD, case 5 at 95 POD, case 6 at 184 POD, case 12 at 36 POD, case 13 at 133 POD, case 14 at 90 POD, and case 15 at 39 POD) and underwent pegylated interferon and ribavirin therapy. Only case 4 patient achieved sustained virological response by the final follow-up. Both ABO-I and control patients showed significantly increased 1-month posttransplant HCV-RNA loads in comparison to the pretransplant viral loads. However, Fig. 4 shows that the increases of HCV-RNA (HCV-RNA load at 1 month after transplantation minus HCV-RNA load at pretransplantation) in ABO-I patients (median 2.85 [range 1.8–3.8] logIU/mL) were significantly higher than those in the control patients (median 0.6 [range −0.7 to 2.2] logIU/mL) (P<0.0001).
The anti-donor blood type isoagglutinin titers did not reach 20 even with the relatively long-term observation period, and the graft livers continued to express donor blood type A and/or B antigen. The scheduled LDLTs were postponed in the cases 4 and 14 because those patients developed catheter-related sepsis, and the interruption of PEs resulted in the rapid increase of anti-blood type isoagglutinin titers to the pretreatment levels. Liver grafting itself may cause the removal of circulating anti-blood type isoagglutinins. In other words, graft livers may absorb these antibodies. Both anti-blood type A and B isoagglutinins were analyzed in the cases 2, 13, 14, and 15 with blood type O. The cases 2, 13, and 14 showed anti-donor blood type isoagglutinin titers lower than non-anti-donor blood type isoagglutinin titers, which may indicate the graft livers were absorbing circulating anti-donor blood type isoagglutinins (10). Fig. 1, SDC 1 (http://links.lww.com/TP/A515), and Figure 2 show that there was persistent expression of donor antigen on hepatic grafts with normal liver function irrespective of the persistent presence of anti-donor blood type isoagglutinins, even several months after LDLT. Do these circulating isoagglutinins attack ABO-I hepatic grafts? Or do ABO-I hepatic grafts become resistant to these isoagglutinins (11)? Tanaka et al. (12) reported intragraft expression of recipient-type ABO blood group antigens. Tanabe et al. (13) recently reported the decrease of antigenicity of graft endothelia after ABO-I kidney transplantation. These intragraft changes may partially contribute to this phenomenon.
The timing of rituximab administration before ABO-I LDLT is crucial because it takes at least 3 weeks after administration to achieve the maximum effect of rituximab (14). Rituximab was administered only 3 days before transplantation in the two fulminant cases (cases 1 and 2), which may have resulted in incomplete elimination of peripheral B lymphocytes and caused the rebound of anti-blood type isoagglutinin titers immediately after transplantation. These observations are similar to those of Egawa et al. (15). Rituximab was administered approximately 3 weeks before scheduled ABO-I LDLT in the remaining 13 cases, and they did not experience any rebound of anti-blood type isoagglutinin titers. These results may simply represent good fortune because the rapid increase of anti-donor blood type isoagglutinin titers may lead to a graft loss caused by accelerated anti-blood type isoagglutinin-mediated rejection, and there were only two successful cases with fulminant hepatic failure. More cases are needed to standardize the use of rituximab administration for ABO-I LDLT in patients with fulminant hepatic failure.
The most serious concern in ABO-I LDLT is that infectious complications may occur more frequently in ABO-I LDLT cases than ABO-identical or -compatible cases because of the intense immunosuppression protocol (16). None of the 15 patients experienced septic complications after the application of the current protocol for ABO-I LDLT. The differences between the immunosuppression protocol for ABO-I LDLT and ABO-identical or -compatible LDLT are pretransplant rituximab administration, pretransplant mycophenolate mofetil (MMF) use, and PEs just before LDLT. Local graft infusion was used early in the series (case 1 and the other two patients without rituximab use) and was often associated with catheter-related complications. Local graft infusion therapy is no longer used in ABO-I LDLT because it was associated with serious catheter-related complications or sepsis. The major cause of death in ABO-I LDLT patients in the past was septic complications secondary to intense immunosuppression and catheter-related complications (17, 18). We believe that severe anti-blood type isoagglutinin-mediated rejection is mainly caused by accelerating the anti-donor blood type isoagglutinin production. Antigen-stimulated B lymphocytes differentiate into antibody-producing plasma cells (19). Rituximab inhibits the production of accelerating isoagglutinin by eliminating peripheral mature B lymphocytes. The use of rituximab with ABO-I LDLT may eliminate the need for graft local infusion therapy. The simplicity of the current protocol may allow patients with ABO-I LDLT to avoid septic complications.
There is also concern whether a splenectomy should be performed during ABO-I LDLT. Egawa et al. (4) described that prophylaxis with rituximab might take the place of splenectomy in ABO-I LDLT. Unlike cadaveric whole liver transplantation, adult-to-adult LDLT patients receive only a small partial hepatic graft, and there have been several reports regarding either persistent portal hypertension or pancytopenia caused by hypersplenism and the beneficial effects of splenectomy in adult-to-adult LDLT patients (20–23). We performed splenectomy in our LDLT cases to control portal venous pressure after reperfusion and to treat pancytopenia (24), not to induce the immunosuppressive effects associated with a splenectomy. However, all 15 patients eventually underwent simultaneous splenectomy during LDLT. The role of splenectomy in ABO-I LDLT remains to be elucidated.
Peripheral B lymphocytes in the blood disappeared for 6 months after rituximab administration and began to recover approximately 6 months later (Fig. 3). The recipients may be vulnerable to various pathogens during this period because of the lack of B-cell immunity. Although no symptomatic CMV infection occurred in the current series, there was a high incidence of CMV antigenemia in ABO-I LDLT cases. At the present time, there is no known relationship between the number of PEs or the amount of immunosuppression and the incidence of CMV antigenemia. Furthermore, a rapid increase of HCV-RNA viral loads was observed in the ABO-I LDLT patients in comparison to the control patients. All seven HCV-RNA positive ABO-I LDLT patients experienced recurrence of hepatitis C early after LDLT and were treated by pegylated interferon and ribavirin therapy. Only one patient had achieved a sustained virological response by the final follow-up. The cases 13 and 15 patients whose hepatitis C viral load steeply increased early after LDLT experienced fluctuating liver function test results with a rapid increase in the total bilirubin after LDLT. We believed that fibrosing cholestatic hepatitis might have occurred in these patients and obtained several biopsy specimens, which did not show any signs of fibrosing cholestatic hepatitis but did show many acidophil bodies. We therefore speculated that severe viremia itself caused the graft liver injury and initiated interferon therapy for these patients, which effectively stabilized their liver function tests. No conclusions can be drawn concerning the safety of this protocol for HCV-positive patients, because of the small number of HCV-positive patients and the short follow-up period.
Further cumulative case studies are required to determine the safe anti-donor blood type isoagglutinin titer levels at ABO-I LDLT. The initial assumption was that pretransplant anti-donor blood type isoagglutinin titers should be decreased to less than 25 by repeat PEs. However, two patients (cases 14 and 15) underwent ABO-I LDLT who had high anti-donor blood type isoagglutinin titers of 26 and 27, and they survived the early posttransplant period, although both patients received IVIG therapy under a suspicion of anti-blood type isoagglutinin-mediated rejection. The higher the anti-blood type isoagglutinin titers are at ABO-I LDLT, the higher the incidence of anti-blood type isoagglutinin-mediated rejection is likely to be. The pretreatment anti-donor blood type isoagglutinin titer was 23 in the case 9, and the patient did not require PEs. The results in the cases 14 and 15 suggest that pretransplant PEs may not be required if the pretransplant anti-donor blood type isoagglutinin titer is below 27.
Haga et al. (25, 26) suggested that C4d-positive staining or periportal edema and necrosis can be a hallmark of acute humoral rejection in ABO-I LDLT. The protocol using rituximab for ABO-I LDLT was only recently developed, and this treatment may affect histological features of anti-blood type isoagglutinin-mediated rejection by depleting mature B cells. Further biopsy data are needed to establish histological diagnostic criteria for anti-blood type isoagglutinin-mediated rejection after rituximab treatment. The only effective treatment for ongoing anti-blood type isoagglutinin-mediated rejection is high-dose IVIG. This treatment has no serious adverse effects in contrast to steroid pulse therapy. Therefore, patients were treated with high-dose IVIG when the clinical data suggested anti-blood type isoagglutinin-meditated rejection. The assessment of clinical data is currently more important than biopsy evaluations for diagnosing anti-blood type isoagglutinin-mediated rejection.
In conclusion, ABO-I LDLT with our protocol is considered to be a safe option when an ABO-identical or -compatible donor is not available. Further study must be carried out to determine the safe pretransplant anti-donor blood type isoagglutinin titers and whether the application of this protocol to HCV-positive patients is justified.
MATERIALS AND METHODS
Fifteen patients underwent ABO-I LDLT in our hospital between November 2005 and December 2010. The patients' characteristics are summarized in Supplemental Table 1 (see SDC 3,http://links.lww.com/TP/A517). The case 2 was presented in a previous case report (5). Informed consent was obtained from each recipient and each donor after gaining approval from the Institutional Ethics and Indications Committees for each ABO-I LDLT. The patients were treated with previously described surgical techniques (27, 28). All 15 patients had a simultaneous splenectomy. The surgical and patients' medical records were retrospectively reviewed. The data of 27 HCV-positive patients who underwent ABO-identical or -compatible LDLT between July 2008 and December 2010 and survived more than 100 days after LDLT were reviewed as a control group. All HCV-positive patients underwent splenectomy for later interferon treatment. Control patients were treated with a triple immunosuppression regimen of oral calcineurin inhibitor (tacrolimus [initial target trough level at 10 ng/mL] or cyclosporine A [initial target trough level at 200 ng/mL]), MMF (initial dose of 1000 mg/day), and steroid. All patients were prophylactically administered 600 mg/day of oral acyclovir. No prophylaxis was used for CMV infection in any of the patients.
Current Protocol for ABO-I LDLT
Elective cases are treated with rituximab (375 mg/m2) 21days before LDLT (see Figure, SDC 2, http://links.lww.com/TP/A516). MMF (1000 mg/body/day) is initiated 7 days before LDLT and continues several months after LDLT. Several sessions of PE using blood type AB fresh frozen plasma are performed to lower the anti-blood type isoagglutinin titers less than 24-fold just before LDLT. Simultaneous splenectomy is performed. One thousand milligram of methylprednisolone is administered just after graft reperfusion, and the daily doses of methylprednisolone are tapered to 20 mg/day over 7 days after LDLT. The administration of oral calcineurin inhibitor (tacrolimus [initial target trough level at 10 ng/mL] or cyclosporine A [initial target trough level at 200 ng/mL]) is initiated 1 day after LDLT. High-dose IVIG (600 mg/kg/day) is administered for 5 days when anti-blood type isoagglutinin-mediated rejection is suspected (e.g., rapid increase of total bilirubin and rapid decrease of platelet count) (9, 29).
The cases 1 and 2 patients could not receive rituximab 21 days before LDLT because of fulminant hepatic failure and it was administered 3 days before LDLT. The case 14 patient had sepsis caused by a catheter-related infection. The catheter was inserted for PE. Because of the severe infection, the planned LDLT had to be postponed, which resulted in an elevation of the anti-blood type isoagglutinin titers up to the pretreatment levels. Approximately 2 months later, the patient had to undergo several sessions of PE again. We were afraid that the retention of the catheter for PE for more than 3 days would cause bacterial sepsis again. We therefore planned three sessions of PE, and the catheter was removed within 3 days, although the isoagglutinin titer at LDLT was more than 24. As a result, the anti-blood type isoagglutinin titer at LDLT in case 14 was 27.
The case 15 patient had high pretreatment anti-blood type isoagglutinin titers, and we tried to lower the titers less than 24 by PE. However, the patient had a severe allergic reaction to fresh frozen plasma, and he could no longer tolerate any PE. We had to perform his LDLT while he still had the high anti-blood type isoagglutinin titers.
Histological Evaluation of Blood Type A/B Antigen Expression on Graft Livers
Time zero-biopsies were available in 12 cases of ABO-I LDLT, and a total of 19 posttransplant biopsies (obtained on the events of hepatitis C recurrence or acute cellular rejection) could be evaluated for the expression of blood type A/B antigen on graft livers. Monoclonal antibody 7LE for Lewisa blood group antigen and 2-25LE for Lewisb blood group antigen (Exbio Praha, Vestec, Czech Republic) were used for immunostaining according to the manufacturer's instructions. The expression of blood type was graded as follows: −, no positive immunostaining; +, positive immunostaining restricted to part of endothelial cells; ++, positive immunostaining for almost all endothelial cells; and +++, positive immunostaining for sinusoidal cells and endothelial cells.
Measurement of Anti-Blood Type Isoagglutinin Titers, the Number of B Lymphocytes, CMV Antigen, and HCV-RNA in the Blood
The anti-blood type isoagglutinin titers for IgG were serially measured, and those values were expressed as 20, 21, 22, 23, so forth (tests for IgM were not available in our hospital). The numbers of B lymphocytes in the blood were calculated as %CD19-positive lymphocytes using flow cytometry (FACSCanto; Becton Dickinson, Franklin Lakes, NJ). We used CD19 to follow the number of B lymphocytes because the administration of rituximab can interfere with the detection of CD20 by flow cytometry, especially when examined early after rituximab administration (14). CMV antigenemia (detection of pp65 antigen on white cells) was examined approximately every 2 weeks or when necessary. Previral loads of HCV before any treatment and postviral loads in the posttransplant periods were determined by real time-PCR.
Continuous variables were expressed as the median and range and compared between two groups by the Wilcoxon rank sum test. Fisher's exact test was used to compare frequencies. Statistical significance was defined as having a P value less than 0.05. All statistical analyses were performed using the NCSS 2007 software package (Hintze JL, Kaysville, UT).
1. Raut V, Uemoto S. Management of ABO-incompatible
living-donor liver transplantation: Past and present trends. Surg Today
2011; 41: 317.
2. Gugenheim J, Samuel D, Reynes M, et al. Liver transplantation across ABO blood group barriers. Lancet
1990; 336: 519.
3. Demetris AJ, Jaffe R, Tzakis A, et al. Antibody-mediated rejection of human orthotopic liver allografts. A study of liver transplantation across ABO blood group barriers. Am J Pathol
1988; 132: 489.
4. Egawa H, Teramukai S, Haga H, et al. Present status of ABO-incompatible living donor liver transplantation
in Japan. Hepatology
2008; 47: 143.
5. Ikegami T, Taketomi A, Soejima Y, et al. Successful ABO incompatible living donor liver transplantation
in a patient with high isoagglutinin titer using high-dose intravenous immunoglobulin. Transplant Proc
2007; 39: 3491.
6. Morioka D, Togo S, Kumamoto T, et al. Six consecutive cases of successful adult ABO-incompatible living donor liver transplantation
: A proposal for grading the severity of antibody-mediated rejection. Transplantation
2008; 85: 171.
7. Jordan SC, Vo AA, Peng A, et al. Intravenous gammaglobulin (IVIG): A novel approach to improve transplant rates and outcomes in highly HLA-sensitized patients. Am J Transplant
2006; 6: 459.
8. Vo AA, Lukovsky M, Toyoda M, et al. Rituximab
and intravenous immune globulin for desensitization during renal transplantation. N Engl J Med
2008; 359: 242.
9. Ikegami T, Taketomi A, Soejima Y, et al. Rituximab
, IVIG, and plasma exchange
without graft local infusion treatment: A new protocol in ABO incompatible living donor liver transplantation
2009; 88: 303.
10. Ishida H, Koyama I, Sawada T, et al. Anti-AB titer changes in patients with ABO incompatibility after living related kidney transplantations: Survey of 101 cases to determine whether splenectomies are necessary for successful transplantation. Transplantation
2000; 70: 681.
11. Hanto DW, Fecteau AH, Alonso MH, et al. ABO-incompatible
liver transplantation with no immunological graft losses using total plasma exchange
, splenectomy, and quadruple immunosuppression: Evidence for accommodation. Liver Transpl
2003; 9: 22.
12. Tanaka Y, Haga H, Egawa H, et al. Intragraft expression of recipient-type ABO blood group antigens: Long-term follow-up and histological features after liver transplantation. Liver Transpl
2005; 11: 547.
13. Tanabe T, Ishida H, Horita S, et al. Decrease of blood type antigenicity over the long-term after ABO-incompatible
kidney transplantation. Transpl Immunol
2011; 25: 1.
14. Genberg H, Hansson A, Wernerson A, et al. Pharmacodynamics of rituximab
in kidney allotransplantation. Am J Transplant
2006; 6: 2418.
15. Egawa H, Ohmori K, Haga H, et al. B-cell surface marker analysis for improvement of rituximab
prophylaxis in ABO-incompatible
adult living donor liver transplantation
. Liver Transpl
2007; 13: 579.
16. Tanabe M, Kawachi S, Obara H, et al. Current progress in ABO-incompatible
liver transplantation. Eur J Clin Invest
2010; 40: 943.
17. Egawa H, Oike F, Buhler L, et al. Impact of recipient age on outcome of ABO-incompatible
living-donor liver transplantation. Transplantation
2004; 77: 403.
18. Kozaki K, Egawa H, Kasahara M, et al. Therapeutic strategy and the role of apheresis therapy for ABO incompatible living donor liver transplantation
. Ther Apher Dial
2005; 9: 285.
19. Shapiro-Shelef M, Calame K. Regulation of plasma-cell development. Nat Rev Immunol
2005; 5 230.
20. Kim SH, Lee JM, Choi JY, et al. Changes of portosystemic collaterals and splenic volume on CT after liver transplantation and factors influencing those changes. AJR Am J Roentgenol
2008; 191: W8.
21. Jeng LB, Lee CC, Chiang HC, et al. Indication for splenectomy in the era of living-donor liver transplantation. Transplant Proc
2008; 40: 2531.
22. Yoshizumi T, Taketomi A, Soejima Y, et al. The beneficial role of simultaneous splenectomy in living donor liver transplantation
in patients with small-for-size graft. Transpl Int
2008; 21: 833.
23. Ikegami T, Soejima Y, Taketomi A, et al. Hypersplenism after living donor liver transplantation
2009; 56: 778.
24. Kishi Y, Sugawara Y, Akamatsu N, et al. Splenectomy and preemptive interferon therapy for hepatitis C patients after living-donor liver transplantation. Clin Transplant
2005; 19: 769.
25. Haga H, Egawa H, Shirase T, et al. Periportal edema and necrosis as diagnostic histological features of early humoral rejection in ABO-incompatible
liver transplantation. Liver Transpl
2004; 10: 16.
26. Haga H, Egawa H, Fujimoto Y, et al. Acute humoral rejection and C4d immunostaining in ABO blood type-incompatible liver transplantation. Liver Transpl
2006; 12: 457.
27. Soejima Y, Taketomi A, Yoshizumi T, et al. Biliary strictures in living donor liver transplantation
: Incidence, management, and technical evolution. Liver Transpl
2006; 12: 979.
28. Uchiyama H, Harada N, Sanefuji K, et al. Dual hepatic artery reconstruction in living donor liver transplantation
using a left hepatic graft with 2 hepatic arterial stumps. Surgery
2010; 147: 878.
29. Morioka D, Sekido H, Kubota K, et al. Antibody-mediated rejection after adult ABO-incompatible
liver transplantation remedied by gamma-globulin bolus infusion combined with plasmapheresis. Transplantation
2004; 78: 1225.