Prospects for Immune Tolerance


The extraordinary success of liver transplantation (LT) as a treatment for end-stage liver diseases has resulted in the need to redirect research priorities and clinical resources toward improving the health and well-being of what are now thousands of long-term surviving LT recipients. This requires the optimization of pharmacological immunosuppression to prevent immune-mediated liver damage while minimizing toxicity. The routine management of immunosuppressive drugs currently relies on the monitoring of serum liver biochemistry tests and calcineurin inhibitor (CNI) levels. This is despite evidence indicating that these parameters have a poor correlation with pharmacodynamic effects, are insensitive and nonspecific indicators of rejection and liver damage, and do not predict long-term outcomes. The limitations of the current standard of care are perfectly illustrated by the results of immunosuppression withdrawal trials demonstrating that up to 40% of selected liver recipients more than 3 years post-transplant can drastically reduce or even discontinue their immunosuppressive drugs. This phenomenon, known as spontaneous operational tolerance (SOT) , has been described in other solid organ transplant settings as well but is extraordinarily rare outside of LT, which reflects the unique immunoprivileged properties of this organ.

The Liver and the Immune System

The liver provides one of the first lines of defense between the host and external environment. It is exposed to a steady stream of antigens, but it has the unique ability to regulate the nature and intensity of its response. It receives vascular inflow from two separate sources, with approximately 20% being received from the hepatic arteries and the remainder from the portal vein. Upon entering the liver, the arterial and venous blood feeds into the sinusoidal bed. This is a unique capillary network lined by liver sinusoidal endothelial cells (LSECs), channeling blood from the portal tracts into the central vein. The vasculature network differs from other capillary systems because of its low oxygen pressure and a slower velocity of flow, which aids to maximize the contact between circulating molecules and the endothelium. The entering blood is rich in antigens from the native microbiome together with nutritional degradation products. However, despite the influx of these antigens, under normal circumstances, the liver does not mount cytopathic immune responses, but instead, the specialized microenvironment has evolved to exquisitely modulate how the immune system responds. For this reason, the liver has long been considered a tolerogenic organ.

The adaptive immune hyporesponsiveness is best demonstrated by the lack of response to lipopolysaccharide (LPS), which is a membrane component of gram-negative bacteria. In most tissues, toll-like receptors (TLRs) assist in pathogen recognition and removal by activating intracellular signaling cascades upon ligand binding, leading to the production of proinflammatory mediators. LPS is a pathogen recognized by TLRs and is associated with a highly inflammatory response, acting as a pyrogen in vivo and leading to the development of vasoplegic shock. The liver is exposed to high concentrations of LPS via the portal vein, but the concentrations drop over 100-fold between portal blood and peripheral blood. This clearance of LPS takes place without signs of inflammation. The continuous exposure to gut-derived LPS results in upregulation of TLR-signaling inhibitors blocking the production of proinflammatory cytokines and leading to the secretion of antiinflammatory mediators, such as interleukin-10 (IL-10) and transforming growth factor-β.

Both parenchymal (hepatocytes) and nonparenchymal (LSECs, Kupffer cells [KCs], resident lymphocytes, and dendritic cells [DCs]) liver cells are involved in regulating intrahepatic immune responses.

Liver Sinusoidal Endothelial Cells

In combination with KCs, LSECs are the most powerful scavenger system in the body. They are critical in pathogen detection and capture and act as nonprofessional antigen-presenting cells. They act to separate the underlying hepatocytes from the blood within the sinusoids. This endothelial layer does not have a basement membrane, which leads to the formation of a gap called the space of Disse. The LSECs form a fenestrated endothelium arranged in clusters known as sieve plates. The sieve plates serve as a portal to facilitate the bidirectional transfer of small substrates between the sinusoidal blood and the space of Disse. They also facilitate direct interactions between hepatocytes and T cells that are adherent within the sinusoid.

LSECs express a variety of pattern recognition receptors, which include the TLRs described above. They also constitutively express major histocompatibility complex (MHC) I and II and costimulatory molecules CD80 and CD86, together with adhesion molecules needed for interaction with lymphocytes.

Kupffer Cells

KCs are liver-resident macrophages adherent to the LSECs and are directly exposed to the flow of blood within the sinusoids. KCs express a multitude of molecules and receptors, including TLRs and costimulatory molecules, together with antibody and complement receptors, allowing them to detect, bind, and internalize pathogens. These receptors also in part drive the activation of KCs, leading to the production of cytokines and chemokines and allowing them to serve as an immune sentinel. The ability to catch bacteria in motion and their unique positioning within the sinusoids are their most important role and what makes them different from other macrophages. Together with pathogen clearance, they also act as antigen-presenting cells, expressing MHC I and II and the costimulatory molecules necessary for the activation of T cells. However, what has been demonstrated is that these cells are poor activators of the adaptive immune system, which makes them tolerogenic cells to the common gut-derived pathogens that flow through the portal venous system.

Dendritic Cells

These cells enter the liver via the portal vein and mature during their transit to the central vein. They also traverse the LSECs and enter the space of Disse and exit the liver through lymphatic drainage. They are professional antigen-presenting cells of the innate immune system promoting self-tolerance in the healthy steady state and can regulate both innate and adaptive immunity. They maintain an immature and nonimmunogenic phenotype and internalize antigens derived from the blood and transport them to a regional lymph node. Despite the strong potential to support T-cell-mediated immunity, the liver DCs appear less efficient at stimulating the activation of T cells. This is in part because of the immaturity of the cell, but also because of the local cytokine activity within the liver promoting antiinflammatory action with high IL-10 and low IL-12. The DCs themselves also produce additional IL-10, which induces antigen-independent hyporesponsiveness in T cells and promotes a shift from helper T cells (TH1)-type responses to TH2-type responses, which further suppresses cellular immunity and promotes the development of regulatory T cells.

Liver-Resident Lymphocytes

The sinusoids contain the largest populations of innate lymphocytes: natural killer cells (NK) and the natural killer T cells. NK cells respond to an abundance of cell surface ligands expressed by damaged or infected cells. After activation, the cells release cytotoxic granules together with a large number of cytokines and can shape and direct the immune response.


The hepatocytes are the main parenchymal cell type and make up approximately 80% of all liver cells, participating in both innate and adaptive immune responses. The hepatocytes express innate immune receptors, MHC class I molecules, and, under inflammatory conditions, MHC class II. There is evidence that these cells can act as nonprofessional antigen-presenting cells and present antigens to naive T cells. This is a type of antigen presentation that provides suboptimal T-cell priming as a result of lack of adequate costimulation by the hepatocytes and secretion of antiinflammatory molecules, together with upregulation of inhibitory receptors, such as programmed death-ligand 1 that engages the programmed death-1 receptor on activated T cells. The final process is the simultaneous suppression of effector T cells and the induction of regulatory T cells that can decrease immune activation and promote tolerance. Fig. 45.1 illustrates the complex interactions in the immune response.

Fig. 45.1

Diagram illustrating how the immune response is regulated by the complex interactions between the endothelium and the parenchymal and nonparenchymal cells along the sinusoid. LPS, Lipopolysaccharide, MHC , major histocompatibility complex.

In the setting of LT, the default environment of tolerance or immune hyporesponsiveness becomes all the more evident. This can be exemplified in a number of clinical conditions, such as the feasibility of performing an LT across a positive cross-match; irrelevance of human leukocyte antigen (HLA) matching for short- or long-term outcomes, the immunomodulating effect of the liver in cases of combined transplantation from the same donor (e.g., liver and kidney or liver and heart), reduced incidence of hyperacute rejection, occasional spontaneous recovery following a severe rejection episode, irrelevance of a single rejection episode in relation to later graft outcome, reduced incidence of chronic rejection and reversal of established rejection even when it has reached the ductopenic stage, and the ease with which a steroid-free immunosuppression regimen can be achieved. LT is also the only setting in which a sizeable number of patients have been shown to be able to discontinue maintenance immunosuppression without undergoing rejection. This phenomenon is known as SOT.

SOT is a state of immune unresponsiveness to an allograft, which, in clinical practice post-LT, is defined as stable graft function in a recipient who has been off immunosuppressive medications for 1 year or more and in whom no clinically significant detrimental immune responses or immune deficits are detected. It was in 1969 when tolerance was first demonstrated in experimental animal models (first in pigs and later on in rodents) by Calne et al. The mechanisms for tolerance are still not understood, but in rodents it has been observed that following transplantation, recipient lymphocytes infiltrate the graft and accumulate in the portal and central areas. These lymphocytes become activated and induce some degree of parenchymal damage. However, instead of destroying the liver, they are cleared. This process is at least in part controlled by so-called regulatory T cells (CD25 + CD4 + T cells), given that depleting this cell subset in the recipients by the administration of anti-CD25 antibody leads to rejection and loss of the graft.

Consequences for Clinical Practice

Short-term survival rates following LT have progressively improved throughout the past five decades because of the advances in surgical techniques, critical care, and the introduction of powerful immunosuppressants, in particular the CNIs. In contrast, the long-term survival rates of transplant recipients have not improved accordingly and remain suboptimal. This is in large part because of the toxicity associated with the long-term use of immunosuppression, ranging from recurrent infections to de novo malignancy (including posttransplant lymphoproliferative disease [PTLD]) and renal impairment, together with cardiovascular and metabolic complications. This discrepancy has motivated intense research into immunosuppression minimization strategies to maintain the health of the allograft at the lowest dose possible and, more particularly, those individuals who have demonstrated SOT.

It was in the 1990s when tolerance was first observed in humans by Starzl et al. at the University of Pittsburgh in a group of patients who had been noncompliant with their immunosuppression regimen. It was comprised of five tolerant individuals who had ceased immunsuppression 1 to 11 years post-transplant, with the longest duration of 13 years off any form of immunosuppression . The maintenance of stable graft function was based upon normal biochemistry rather than any histology. This initial observation led to further retrospective analyses and a number of small clinical trials. The first was by Tsakia et al. and included a pediatric population of seven individuals in whom immunosuppression had to be withdrawn because of clinical complications such as hepatitis C or PTLD. Out of the seven, four were female, the etiology of the liver disease leading to transplantation excluded any autoimmune cause, they had all been immunosuppressed with tacrolimus together with prednisolone, and four had had at least one episode of acute cellular rejection. Immunosuppression was stopped 0.5 to 1.3 years after transplant, and six out of the seven remained off for 1.3 to 2.8 years with no cellular rejection demonstrated on histology. The study also included the in vitro testing of immunocompetence in those individuals demonstrating tolerance.

In 1995, the first prospective study of both pediatric and adult patients by Ramos et al. evaluated 72 long-term LT patients with regards to candidacy for purposeful physician-led immunosuppression weaning. The criteria for inclusion were (1) more than 5 years post-transplant, (2) more than 1 year without an episode of rejection, (3) compliant, (4) have complications related to immunosuppression, (5) cooperation of the primary physician, and (6) baseline liver biopsy showing no evidence of rejection or hepatic disease. Of the 72 subjects evaluated, 13 were removed, leaving 59 with an age range of 12 to 68 years (20 were 12–20 years old, and 39 were 21–68 years of age). In terms of the immunosuppression regimens, they included prednisolone, ciclosporin, tacrolimus, and azathioprine (12 on triple therapy, 32 on dual therapy, and 15 on CNI monotherapy). In those with dual therapy, only 3 of the 32 did not include prednisolone. The weaning regimen was azathioprine first in those on a triple-drug regimen; otherwise, prednisolone was weaned first using a corticotropin stimulation test as a guide. The baseline agents were reduced in monthly decrements with weekly/biweekly blood tests.

In terms of their results at the time of publication, 16 patients had undergone complete weaning with 3 to 19 months of drug-free follow-up. There were 28 who were still in the progress of weaning, and 15 had failed, with 12 demonstrating histopathological evidence of rejection. Definitive conclusions could not be made because the study was still in progress, but it was observed that in terms of the etiology, the success rates appeared higher for those transplanted for biliary atresia and lower for those with primary biliary cholangitis or alpha 1 antitrypsin deficiency. There also seemed to be a higher rate of failure for those individuals on triple or double ciclosporin–based immunosuppression, and those on CNI monotherapy appeared to fair better. The conclusions made by the authors with regard to when to consider immunosuppression withdrawal were: (1) wait until 5 to 10 years post-transplant, (2) careful case selection, (3) close monitoring, and (4) prompt reinstitution of immunosuppression when necessary.

Starzl and colleagues noted that long-term LT recipients were donor-recipient chimeras, not only in the graft, but also in the blood, skin, lymph nodes, and other organs. The detection of multilineage chimerism, indicating migration and survival of donor cells from the allograft into multiple-recipient tissues, was more striking in liver recipients who managed to discontinue immunosuppression. It was suggested that macro- or microchimerism might be both a mechanism and biomarker indicative of permissive conditions for immunosuppression withdrawal. This constituted the rationale for attempting to intentionally induce tolerance by providing donor-type bone marrow or hematopoietic stem cells to LT recipients, a strategy that was not successful and is now considered ineffective in the absence of very aggressive recipient conditioning.

Following the original studies conducted in Pittsburgh, there were multiple additional single-center reports of immunosuppression withdrawal in both adult and pediatric LT (PLT) recipients, which are illustrated in Table 45.1 .

Table 45.1

Published Single-Center Trials of Immunosuppression Withdrawal in Both Adult and Pediatric Populations

Adult/Pediatric Cohort Inclusion Criteria Etiology of Liver Disease Time Since Transplant Immunosuppression regimen Liver Biopsy Outcome Comments
Both 95 (31 < 20 years old, 64 between 21 and 68 years > 5 years post-transplant
> 2 years without rejection
Cooperative local physician
Absence of acute or chronic rejection on baseline biopsy
Exclusion of vascular/biliary complications or recurrence of original disease
Chronic viral hepatitis, n = 13
Biliary atresia, n = 25
PSC/PBC, n = 16
Hepatoma, n = 4
Miscellaneous, n = 33
8.4 years ± 4.4
Range: 1.7–25 years
Adult mean: 9.13 yrs
Pediatric mean: 6.64 yrs
Aza/Pred, n = 13
Ciclo/Pred, n = 32
Ciclo/Aza/Pred, n = 24
Ciclo/Aza, n = 4
Ciclo, n = 11
Tac, n = 8
Tac/Pred, n = 1
Tac/Aza, n = 2
Variable weaning regimen initially was Pred, however, shifted to primary weaning of Ciclo/Tac
Yes, prewean 1997:
18/95 (19%) drug free for 0.84–4.8 years
2007 follow-up:
12 remain off IS with normal graft function
6 died of non rejection related causes
16 further patients had been successfully weaned
Included the previous study with the 59 from the Ramos et al. study
Two patients received a simultaneous liver kidney
Subjects on Pred and Aza, P = .0007, and those on Tac, P = .0031, were more likely to be off drugs at 1 year compared with those on Ciclo
King’s College
Adult 18 (23–69 years old) None stated PBC, n = 3
PSC, n = 3
Budd-Chiari, n = 3
HCV and AIH, n = 1
HCV, n = 1
Cystic fibrosis, n = 1
Acute liver failure, n = 3
ALD and HCV, n = 1
HCC and HCV, n = 1
Hepatic metastases from treated testicular teratoma, n = 1
Median: 7 years; range: 5-11 years Ciclo/Aza, n = 3
Ciclo/Aza/Pred, n = 15
Yes, pre- and postwean 1998:
5/18 (28%) drug free but not clearly stated the duration.
A liver biopsy was performed at 8–24 months after withdrawal
2005 follow-up:
2 remain off IS; the remaining three include acute rejection/chronic rejection/renal transplantation
Systemic chimerism was not associated with successful IS withdrawal
Viral or autoimmune etiologies of the underlying liver disease were more likely to experience failure of withdrawal from IS
Pediatric 63 patients
(26 elective, 37 nonelective)
> 2 years post-transplant
Good graft function
No rejection in the preceding 12 months
Not described At least 2 years Tac ± Pred Not stated 24/63 (38.1%) achieved complete weaning with median time 23.5 months. Range 3-69 months
Pediatric 115 patients (67 elective, 48 nonelective) > 2 years post-transplant
Good graft function
No rejection in the preceding 12 months
Not described At least 2 years Tac ± Pred Not stated 49/115 (42.6%) with 16 elective and 33 nonelective. Complete withdrawal in a median drug-free period of 21.9 months in elective and 36.8 months in nonelective. Range: 4 months–8 years)
Pediatric 581 patients (From 1990–2005) Pediatric patients
Normal liver function
Survived more than 2 years
No single episode of rejection in the preceding 12 months
Parental permission
Not described At least 2 years Tac ± Pred Not all 87/581 (15%) but duration not stated They began protocol biopsies in 2003.
Despite normal liver function, protocol biopsy demonstrated decrease in size and increase in number of the bile duct, together with fibrosis in patients who had completely withdrawn from IS.
Adult 9 patients > 2 years post-transplant None transplanted for viral hepatitis or autoimmune liver disease 62 ± 25 months Not stated Yes, prewean, and then at 2 months 3/9 (33%) 17–24 months off IS Endothelial cell chimerism was not associated with successful IS withdrawal
Pediatric 38 patients (19 patients with PTLD and 19 with EBV infection) All patients who had PTLD and those with EBV infection but no allograft rejection Not described PTLD: median: 235 days (range: 21–2743 days)
EBV infection: median: 298 days (range: 18–1658 days)
Not stated No 8/38 (21%) off IS for 1535.5 days ± 623 days
New Orleans
Adult 18 patients Had to be transplanted using the steroid-free IS protocol
Desire to be IS free
At least 6 months post-transplant without an episode of rejection
Tac monotherapy with trough levels < 5 ng/ml
Normal liver biochemistry
Carcinoid, n = 1
Crytogenic, n = 1
Fulminant, n = 1
Fulminant HBV, n = 1
Glycogen storage, n = 1
HCV, n = 2
HCV/HCC, n = 2
Haemochromatosis, n = 1
Laennec’s, n = 6
Laennec’s/HCC, n = 1
Patent ductus, venous n = 1
> 6 months Steroid-free IS
1.5mg/kg RATG while a hepatic and the second dose on D1 post-transplant
Commenced upon MMF and Tac and then switched to Tac monotherapy at 2 weeks
Liver biopsy performed for elevations in liver biochemistry over twice the normal range or elevation of bilirubin over 2 mg/dl.
Biopsy not performed preweaning
1/18 (6%) off IS for 12 months 12/18 required liver biopsies
Adult 104 patients
G1 underwent DBMC infusion, n = 45

G2 control, n = 59
At least 3 years post-transplant
Stable liver function
No acute rejection episode for at least 12 months
No history of autoimmune disease
HCV, n = 47
HBV, n = 16
Laennec cirrhosis, n = 16
Crytogenic cirrhosis, n = 16
Fulminant, n = 4
Budd-Chiari, n = 2
Trauma, n = 1
Alpha 1, n = 1
Neuroendocrine tumor, n = 1
G1: 3.63 ± 0.08 years

G2: 4.4 ± 0.19 years
G1: Tac, n = 42; Ciclo, n = 2; sirolimus, n = 1

G2 – Tac, n = 46, Ciclo, n = 13
Both liver and bone marrow biopsy were obtained at enrollment, 12 months, 24 months, and 6 months after successful withdrawal of IS 9 patients were removed because of noncompliance
20/95 (21%)
G1: 10 – Average off IS 2 years (1.1–3 years)
G2:10 – Average off IS 2.3 years (0.9–3 years)
2010 follow-up
7.27 ± 0.28 years
G1 received bone marrow infusions early post-transplant 1-5 infusions within 100 days
Some subjects were withdrawn from the study because of suspected rejection but without histological diagnosis
Tolerant group were 91% males
One patient had an episode of rejection 5 years after withdrawal
Tor Vergata
Adult 34 patients
25 male and 9 female
HCV RNA positive
> 12 months post-transplant
Biopsy-proven recurrent HCV
Normal graft function
All transplanted for HCV-related cirrhosis 63.5 ± 20.1 months At the beginning of the trial, all patients were on ciclosporin monotherapy Yes, preweaning and then yearly protocol biopsies 8/34 (23%) with mean follow-up 45.5 ± 5.8 months
Further follow-up 78 months
Independent predictors of IS-free state:
Low Ciclo trough level during first week post-transplant
Steroid-free initial IS
There was a slower fibrosis rate in tolerant individuals however this observation did not persist in the longer follow-up
Adult 26 patients
13 males and females
Free of rejection for minimum 2 years
On single or double IS
Transaminase levels < 1.5 times the ULN
Not directly stated but did include patients with an autoimmune etiology Not specifically stated On one or two agents Baseline liver biopsy 2/26 (8%) of IS after 1 year of follow-up Randomized to receive ursodeoxycholic acid versus placebo followed by IS withdrawal
No differences observed between groups
Adult 12 patients Stable graft function for more than 2 years
> 1 year without rejection; excluded viral disease, cancer, or autoimmune disease before transplant
Alcoholic cirrhosis, n = 9
Wilson, n = 2
Secondary biliary cirrhosis, n = 1
24–127 months On Ciclo-based IS Baseline liver biopsy
Biopsied for liver enzyme abnormalities and 2 months after withdrawal of IS
5/12 (42%) 10–30 months off IS None of the baseline factors (age, sex, treatment with steroids, Ciclo dose, Ciclo trough levels, liver function tests, history of previous rejection, or time since transplant) were associated with successful IS withdrawal
An increase in the frequency of CD4 + CD25 T cells and FoxPS transcripts was associated with tolerance
Pediatric 5 patients: 4 noncompliant and 1 with PTLD Very low Tac trough levels Biliary atresia, n = 2
Neonatal hepatitis, n = 1
HCC, n = 2
Median: 45 months (range: 14–60 months) Tac monotherapy at time of weaning None 5/5 (100%) off IS for 14–82 months 4 patients received a living donor graft and 1 cadaveric
Adult 24 patients Deceased donor graft
> 3 years post-transplant
Absence of an autoimmune liver disease
Absence of active viral hepatitis
Freedom from rejection in the last year
Normal liver function for the past 6 months
Required metabolic or malignant toxicity of IS or high risk of de novo malignancy
Main etiology:
Alcoholic cirrhosis
Median time was 112 months MMF, n = 7
Ciclo, n = 2
Tac, n = 5
Sirolimus, n = 2
Ciclo/MMF, n = 1
Tac/MMF, n = 5
Tac/Aza, n = 1
Ciclo/Aza, n = 1
Yes, preweaning 15/24 (63%)
Median: 14 months off IS (IR: 8.5–22.5 months)
Tolerant individuals had a longer median time since transplant: median 156 months versus 71 months for the nontolerant
Tolerant individuals had a lower median phytohemagglutinin stimulation index
Pediatric 15 patients; there were 16 but 1 dropout Age > 1 year at time of transplant with more than 2 years of follow-up or age < 1 year at transplant with more than 1 year of follow-up
Normal graft function with no evidence of rejection within 1 year
Metabolic liver disease, n = 4
Biliary actress, n = 7
Budd-Chiari, n = 1
HBV/HCC, n = 1
Hepatoblastoma, n = 1
Idiopathic liver failure, n = 1
Median: 2.03 years (range: 1.38–3.69 years) Tac monotherapy in 14/15
All weaned off steroids at 3 months
Antimetabolite used in patients > 8 years
Baseline biopsy
Yearly protocol biopsies
5/15 (33%)
Time off IS: 24.6–42.8 months
14/15 received a living donor graft
Comparing patient characteristics, metabolic liver disease, and shorter time post-transplant were associated with tolerance

Only gold members can continue reading. Log In or Register to continue

Stay updated, free articles. Join our Telegram channel

Feb 23, 2021 | Posted by in HEPATOPANCREATOBILIARY | Comments Off on Prospects for Immune Tolerance

Full access? Get Clinical Tree

Get Clinical Tree app for offline access