The critical role of therapeutic plasma exchange in ABO-incompatible liver transplantation

Cheng-Zuo Hn , b , Qing Wei , b, c , Meng-Fn Yng , b , Li Zhung , Xio Xu , b, c, ∗

a Department of Hepatobiliary and Pancreatic Surgery, The Center for Integrated Oncology and Precision Medicine, Affiliated Hangzhou First People’s Hospital, Zhejiang University School of Medicine, Hangzhou 310 0 06, China

b Institute of Organ Transplantation, Zhejiang University, Hangzhou 310 0 03, China

c National Center for Healthcare Quality Management in Liver Transplant, Hangzhou 310 0 03, China

d Department of Hepatobiliary and Pancreatic Surgery, Shulan (Hangzhou) Hospital, Hangzhou 310022, China

Keywords:Plasma exchange ABO blood group system Blood group incompatibility Liver transplantation Graft rejection

A B S T R A C T

Introduction

Liver transplantation (LT) has become the only treatment for patients with end-stage liver disease (ESLD) and acute liver failure. However, the shortage of donor liver restricts the patients with ESLD to receive LT. Increasing the utilization rate of expanded criteria donor (ECD), including ABO-incompatible (ABO-I) donor, is an important way to solve this problem. In the 1980s and early 1990s,the outcome of ABO-I LT was extremely poor, and the high incidence of acute rejection and vascular biliary complications led to low graft and patient survival [ 1 , 2 ]. With the development of plasma exchange, splenectomy, immunoadsorption, vascular perfusion, and other techniques to remove blood group antibodies, the survival of blood group incompatible LT recipients has been greatly improved. At present, ABO-I LT is mainly used in emergency transplantation to treat acute liver failure [ 3 , 4 ].

Anti-A/B antibody is the trigger of immune response to ABO-I LT graft injury. Plasma exchange can quickly reduce the titer of plasma antibody and effectively inhibit the effect of humoral immunity [ 5 , 6 ]. Therapeutic plasma exchange (TPE) can reduce the level of A/B lectin in recipient blood, but the control standard of lectin revision level is still controversial, the target titer varies significantly with different centers, and the standard target titer has not yet been established. In this review, we discuss the emergence and development of plasma exchange, the mechanism of TPE applied to ABO-I LT, and how TPE is used in different centers.

Emergence and development of plasma exchange

Anti-A/B antibody is the trigger of immune response to ABO-I LT graft injury. Plasma exchange can quickly reduce the titer of plasma antibody and effectively inhibit the effect of humoral immunity [ 5 , 6 ]. Plasma clearance was first proposed by Abel et al. in 1914 [7] . With the advent of the intermittent blood cell separator in the 1960s and the invention of the membrane plasma separation device in the late 1970s [ 7 , 8 ], modern plasma clearance technology has been continuously improved. At present, plasma exchange can not only fully separate the whole plasma, but also selectively separate a certain component in the plasma. In the specific process of clinical application, the separated plasma can be discarded and the new plasma or substitute can be injected into the body.Plasma exchange is the first separation technique used by ABO-I LT. At present, different techniques are divided into conventional plasma exchange and cascade plasma exchange according to their selectivity to target molecules. The main difference between these treatments is their selective level of antibody clearance.

The clinical status of TPE has been fully affirmed because it can quickly remove pathogenic factors and regulate immune function. However, TPE requires a large amount of plasma transfusion, which increases the risk of transfusion reaction (anaphylaxis) and/or transfusion transmitted infection. However, the use of semi-selective techniques, such as cascade plasma exchange,can give priority to the removal of pathogenic substances, so that higher plasma volume can be handled, potentially improving the treatment. Dual filtration plasma exchange is a membrane technique for selective removal of macromolecules by dual filtration of plasma. It has two kinds of membranes with different pore sizes, which are respectively used as plasma separator and plasma fractionator [9] . Compared with TPE, the advantage of cascade plasma exchange is that dual filtration plasma exchange not only minimizes albumin exchange, but also maximizes the selective removal of protein by plasma fractionator. Another advantage over traditional TPE is that they can handle a large amount of plasma(1.5-3.0 plasma volumes) and require less replacement fluid when using cascade plasma exchange for surgery.

Mechanism of TPE for ABO-I LT

In liver, ABO antigen is mainly expressed in biliary cells and vascular endothelium [10] . When ABO-I LT is performed, the antigens in the vascular endothelium and biliary epithelium of the donor liver will bind to the corresponding antibodies in the recipient body because of the presence of antibodies against ABO blood group in the recipient. After that, the classical pathway of complement cascade will be activated, resulting in the formation of membrane attack complex (C5b-C9). Membrane attack complex damages endothelial cells, including intrahepatic vascular endothelium and biliary endothelium, resulting in antibody-mediated immune rejection (antibody-mediated rejection, AMR) in the early stage after operation.

As an immune preferential organ, liver has a lower risk of AMR after transplantation than other organs, which is mainly due to the unique anatomy and function of the liver. Liver has dual blood supplies, portal vein and hepatic artery, and a large area of sinus and Kupffer cells can dilute and remove immune complexes. Liver also has strong regeneration ability [ 11 , 12 ]. Despite this, AMR is still one of the common refractory complications after LT. AMR can cause diffuse thrombosis and biliary stricture. Acute liver necrosis and refractory bile duct injury are the most common clinical manifestations, which indicate that AMR will lead to a serious adverse outcome, and may eventually lead to the loss of function of the transplanted liver [ 13 , 14 ]. It is the naturally produced antibodies against ABO antigens that are the key vectors of AMR and prevent LT across the ABO barrier. It has been reported that other complications of ABO-I living donor liver transplantation (LDLT), such as liver necrosis and intrahepatic biliary complications, are closely related to high perioperative titer of anti-A/B antibodies [15] . Therefore, most current efforts to improve ABO-I LDLT results aimed at reducing anti-A/B antibodies to a considered safe level before transplantation. After retrospective analysis of 441 cases of ABO blood group incompatible transplantation, Takahashi [16] summarized the characteristics of AMR as follows: most AMR occurred within 7 days after transplantation, and then the incidence decreased gradually, so we should actively prevent and treat the occurrence of rejection during this period.

TPE is the most commonly used method in different treatments of AMR, which can effectively remove donor specific antibody in peripheral blood and reduce its intensity. In addition, TPE also effectively improves liver function in patients with hyperbilirubinemia. Although TPE can remove antibodies from peripheral blood before ABO-I LT, TPE does not inhibit the production of new antibodies by existing plasma cells. Therefore, repeated plasma exchange is an effective treatment for patients with elevated hemagglutinin (HA) titer after ABO-I LDLT [17] .

Strategies of TPE for ABO-I LT

TPE can reduce the level of A/B lectin in recipient blood, but the control standard of lectin revision level is always controversial,the target titer varies significantly with different centers, and the standard target titer has not yet been established. Therefore, different centers have made different attempts on whether to use TPE,when to use TPE and how often to use TPE.

Setting target titers before and after surgery

Kim and colleagues from Korea reviewed 22 successful cases of LDLT from ABO-I donors [18]. All patients in their study received TPE before transplantation. The double-needle TPE procedure was performed using a Cobe Spectra Apheresis system consisting of a single-stage channel filler and a disposable TPE device (Gambro BCT, Lakewood, CO, USA). TPE was performed every other day before transplantation until the titer of IgM and IgG lectins corresponding to the donor ABO blood group was less than or equal to 1:8. The titer of anti-ABO isoagglutinin (IgG/IgM) was monitored daily until 2 weeks after transplantation. Initially, the target titer at transplant and the first 2 weeks after transplantation was 1:32. All patients received a single intravenous injection of rituximab (375 mg/m2body surface area) 2 weeks before LDLT. During LDLT, all recipients used basiliximab as an inducer.Most patients received infusions of prostaglandin E1, gabexate mesylate, and methylprednisolone. Tacrolimus, steroids, and mycophenolate mofetil are the main immunosuppressive drugs after LDLT.Twenty-two recipients were followed up for 3-21 months (mean 10 months). When a patient’s antibody titer exceeded 1:32, their protocol was to administer TPE with an immunobooster inhibitor,but not with an additional dose of rituximab. When acute rejection was suspected, a liver biopsy was performed. When acute rejection was confirmed by biopsy, they opted for intravenous methylprednisolone (500 mg/d) for 3 days, then the dose was reduced to 60 mg/d for 4 days thereafter. The survival rate of both patients and grafts was 100%. There was no AMR. This study determined how much plasma exchange was needed for successful transplantation and showed that rituximab and TPE were effective in obtaining sufficient transplant titers.

Another study conducted by Yoon et al. [19] reviewed the medical records of patients who received LDLT treatment for liver cancer at an institution in Korea. A total of 165 patients were treated with ABO-I LT and 753 patients with ABO-compatible LDLT for hepatocellular carcinoma. Recipients with ABO blood group incompatibility received desensitization therapy to overcome ABO blood group disorders, including pre-transplant plasma exchange and injection of rituximab (300-375 mg/m2body surface area). The frequency and timing of TPE depended on the HA titer, with the goal of reaching an antibody titer ≤1:8 before LT. A more detailed description of the ABO-I adult LDLT desensitization program at the hospital where the team was based has been published [ 20 , 21 ]. The titer of HA and the proportion of CD20 + lymphocytes were measured daily in the first week after LT. If the titer of HA was higher than 1:32, TPE was performed at the same time to strengthen the immunosuppressive regimen. No extra dose of rituximab was given. All patients in the ABO-I and ABOcompatible groups used the same immunosuppressive regimen, including tacrolimus, mycophenolate mofetil (500 mg, twice a day)and steroids. Steroids decreased gradually within 3 months after LDLT.

Setting target titers before surgery

A study by Lee et al. [22] established a new simplified protocol that did not require splenectomy and local perfusion and also omitted the routine use of intravenous immunoglobulin. The study included 19 patients who received ABO-I LDLT. Patients received a single dose of rituximab (375 mg/m2) 10 days prior to transplantation. TPE was performed to reduce the ratio of the recipient’s isoagglutinin titer to the target titer to 1:32. Tacrolimus(0.1 mg/kg/d), mycophenolate mofetil (10 0 0 mg/d) and prednisolone (20 mg/d) were given 7 days before transplantation. They did not undergo local perfusion or splenectomy. All patients were treated with basiliximab 20 mg on day 0 and day 4 after transplantation. They performed preoperative TPE on 16 recipients. One problem with the simplified scheme is the rebound of antibodies.Therefore, the titer should be reduced immediately after transplantation, which can prevent the occurrence of AMR. In their study,9 recipients received additional TPE after transplantation. Among them, 3 cases of isoagglutinin returned to 1:64, but responded well to TPE and no AMR occurred. After surgery, one died of pneumonia (5.3%). There were 4 cases of acute rejections (21.1%), and all of them were treated successfully with steroid pulse or antithymocyte globulin. AMR and graft failure did not occur. This study shows that ABO-I LDLT with the use of simplified protocol can be safely performed without increased risk of AMR and other complications. In fact, the use of basiliximab remains controversial. Previous studies [ 23 , 24 ] have shown that immunosuppressive regimens(including basiliximab) can reduce mortality and graft rejection in liver recipients compared with glucocorticoid induction. However, through network meta-analysis, Best et al. [25] ranked different induced immunosuppression schemes according to their safety and effectiveness, and found that most of the studies on induction schemes were based on small-scale trials with high bias risk and great uncertainty. Therefore, it is not possible to determine the effects of basiliximab on clinical outcomes.

Troisi and his colleagues set the preoperative antibody target titer at 1:16 [26] . They describe a new protocol, using Glycosorb ABO and four immunosuppressants including daclizumab and mycophenolate mofetil. Antigen-specific immunoadsorption (ASI) provides an efficient way to reduce HA titers. They use the immunization column (Glycosorb ABO, Glycorex Transplantation, Lund,Sweden, a low molecular carbohydrate column with A or B blood group antigen linked to a sepharose matrix specifically depleting anti-A or anti-B antibodies). A standard device for plasma exchange supporting these columns was used to filter the patient’s venous blood through a double-lumen central venous line. ASI sessions were started from the third preoperative day and continued until HA titers of 1:16 were reached. Daily posttransplantation immunoglobulin IgG and IgM HA titer measurements conditioned additional posttransplantation ASI sessions. All patients were treated with quadruple immunosuppresses. The spleen was preserved. The immunosuppressive protocol consisted of tacrolimus at a dose of 0.075 mg/kg 12 h before LDLT and at postoperative day 1, increased according to its trough levels with a target of 11–15μg/dL; mycophenolate mofetil at the daily doses of 2 × 1 mg orally, starting the day before LDLT and continued thereafter; daclizumab 1 mg/kg intravenously at postoperative days -1, 1, 15, 30, and 45; as well as methylprednisolone 15 mg/kg intravenously at reperfusion, then tapered to 1 mg/kg per day over a week, with a maintenance dose of 0.1 mg/kg during the first 6 months with the aim of further decreasing the dose and withdrawing it after 12 months. Antimicrobial prophylaxis consisted of 2 mg temocillin every 12 h and 400 mg teicoplanin per day. Antifungal prophylaxis was assured by 400 mg/d fluconazole and antiviral prophylaxis was obtained by giving 3 weeks intravenous gancyclovir (5 mg/kg/d). This study shows that ABO-I LDLT can be implemented using this adapted immunosuppress protocol.

Without TPE

There are also some centers that do not recommend the use of TPE before transplantation. They believe that the active clearance of antibodies with TPE may lead to the production of antibody from scratch after transplantation [ 21 , 22 , 27 , 28 ]. Some scholars believe that TPE should be avoided before transplantation because it may complicate citrate-induced hypocalcemia, nausea, hypotension and allergic reactions [29] .

Splenectomy can significantly reduce the number of ABO blood group antibody source cells. However, splenectomy is timeconsuming with large surgical trauma and many complications.Ikegami et al. [30] found that if ABO-I LT was administered 2-3 weeks after rituximab administration, CD20 was depleted simultaneously with the depletion of plasma cells. In ABO-I DDLT,due to the short time, emergency administration of rituximab before transplantation cannot fully eliminate B cells, so most centers choose combined splenectomy to prevent postoperative AMR reaction [ 31 , 32 ]. However, Lee et al. [33] found that splenectomy could not reduce the incidence of AMR.

A retrospective study by Yamamoto et al. [34] demonstrated the feasibility of rituximab in ABO-I LT without the need for any additional desensitization therapy. Egawa et al. [35] reported that in multivariate analysis, only the lack of rituximab prophylaxis was an important risk factor for AMR. However, Yamamoto et al. [34] believe that the cause of AMR may be the antibody produced after LT rather than the antibody that existed before LT. Takahashi et al. [36] pointed out that acute AMR was not caused by preexisting antibodies remaining after removal of pretransplant antibody, but evidently associated withdenovoantibodies produced after transplantation according to C4d deposition of renal graft by episode and protocol biopsies in ABO-I renal transplantation.

ABO-I deceased donor pediatric LT

There are differences in the results of LT between children and adults, and pediatric LT is more successful, which may be due to the low titer of HA and the immature complement system in young children. ABO-I LT has once again become another choice for patients waiting for transplantation. However, the treatment regimen for children receiving deceased donor ABO-I LT is not standardized. A small single-center study of children receiving transplants from living relatives has shown that increased titers of HA after transplantation may be a predictive risk factor for increased mortality and morbidity. Good results can be obtained by using TPE and rituximab to reduce the titer of anti-donor blood group antibodies [37–39] . Mysore et al. [40] from Houston implemented a new immunosuppressant program for children receiving deceased donor ABO-I LT based on the titer of pre-transplant immunosuppressant. Children with high HA titers ( ≥1:32) before transplantation received enhanced immunosuppressant regimens including plasma exchange, rituximab, intravenous gamma globulin and mycophenolate mofetil. Children with an HA titer of 1:16 were treated with steroids and tacrolimus. They initiated two TPE-based plasma exchanges before or during surgery, using plasma compatible with recipients and donors. After transplantation, they performed 1.3 plasma volume exchanges per day using plasma compatible with recipients and donors, up to 6 times a day, or until the associated HA titer could not be detected, whichever occurred first. They think the trends of HA are important, and the goal of TPE is to bring levels down to 1:2 or lower. The patients were given CD20 monoclonal antibody rituximab (375 mg/m2) after the first TPE treatment and high dose intravenous immunoglobulin (1 g/kg) after the last TPE treatment. Antimetabolic drugs such as mycophenolate mofetil (15 mg/kg, twice a day) were used in the first week after transplantation, depending on clinical manifestations,risk of acute infection, and age. They conducted a two-year retrospective evaluation of the results of ABO-I LT treatment in recipients of similar age and diagnosis. Ten children with a median age of 8.9 months received ABO-I LT, and 4 of them received immunosuppressants enhancement therapy because of their high HA titer.The incidence of complications (rejection, infection, complications of biliary tract and blood vessels) at 1 and 3 years after transplantation was comparable between the two groups. The graft function of ABO-I LT patients was good, the median follow-up time was 3.3 years, and the survival rate was 100%. This study shows that pediatric HA can be evaluated, listed, and tailored according to pretransplant ABO-I LT titers.

Challenge

Although TPE has several schemes to reduce antibody titers,there is a lack of clinical trials that provide standardized procedures. In addition to controlled clinical trials, further basic scientific research on blood group incompatible transplants is needed.The immunological mechanism of temporarily lowering the titer of ABO to allow organs to survive is still poorly understood. Although the presence of ABO antibodies can lead to hyperacute rejection after ABO-I LT, these antibodies can recover from low levels and the graft continues to maintain good function, which is called regulation. The mechanism of regulation is still uncertain, but it is assumed that it is due to the increased resistance of graft endothelial cells to injury mediated by the host immune system. In addition,further research and clinical trials should be conducted to determine the best regimen for TPE to remove ABO antibodies and prevent AMR. However, both TPE and titer determination are growing fields in which laboratories and blood transfusion physicians play a key and expanding role.

Acknowledgments

None.

CRediT authorship contribution statement

Cheng-Zuo Han: Writing - original draft. Qiang Wei: Conceptualization, Methodology, Writing - original draft. Meng-Fan Yang:Methodology, Writing - original draft. Li Zhuang: Methodology,Writing - original draft. Xiao Xu: Conceptualization, Funding acquisition, Supervision, Writing - review & editing.

Funding

This study was supported by grants from the National Natural Science Foundation of China (816250 03, 8180 0578, and 81930016), Key Research & Development Plan of Zhejiang Province(2019C03050 and 2021C03118), and Projects of Medical and Health Technology Program in Zhejiang Province (WKJ-ZJ-2120).

Ethical approval

Not needed.

Competing interest

No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article.