Trial of complete weaning from immunosuppression for liver transplant recipients: Factors predictive of tolerance


  • This work was supported by a grant from Research Projects of the Universidad de Navarra.

  • See Editorial on Page 933

Address reprint requests to J. Ignacio Herrero, M.D., Liver Unit, Clínica Universidad de Navarra, Pío XII 36, 31008, Pamplona, Spain. Telephone: +34 948-296637; FAX: +34 948 296500; E-mail:


Recipients of liver transplantation (LT) may develop immunological tolerance. Factors predictive of tolerance are not clearly understood. Transplant recipients with normal liver function tests and without active viral hepatitis or autoimmune disease who presented with side effects of immunosuppression or a high risk of de novo malignancies were selected to participate in this prospective study. Twenty-four patients fulfilled the inclusion criteria and, therefore, underwent a gradual reduction of immunosuppression. Tolerance was defined as normal liver function tests after immunosuppression withdrawal. Basal clinical and immunological characteristics, including lymphocyte counts and subpopulations (T, B, natural killer, CD4+, CD8+, and regulatory T cells) and the phytohemagglutinin stimulation index (SI), were compared for tolerant and nontolerant patients. Fifteen of the 24 patients (62.5%) were tolerant at a median of 14 months (interquartile range = 8.5-22.5 months) after complete immunosuppression withdrawal. Tolerant patients had a longer median interval between transplantation and inclusion in the study (156 for tolerant patients versus 71 months for nontolerant patients, P = 0.003) and a lower median SI (7.49 for tolerant patients versus 41.73 for nontolerant patients, P = 0.01). We identified 3 groups of patients with different probabilities of tolerance: in the first group (n = 7 for an interval > 10 years and an SI < 20), 100% reached tolerance; in the second group (n = 10 for an interval > 10 years and an SI > 20 or an interval < 10 years and an SI < 20), 60% reached tolerance; and in the third group (n = 7 for an interval < 10 years and an SI > 20), 29% reached tolerance. In conclusion, a high proportion of select LT recipients can reach tolerance over the long term. Two simple basal variables—the time from transplantation and the SI—may help to identify these patients. Liver Transpl 19:937–944, 2013. © 2013 AASLD.


alanine aminotransferase


alkaline phosphatase




aspartate aminotransferase


forkhead box P3


liver transplantation


natural killer




stimulation index


regulatory T cell

Although the survival of liver transplantation (LT) patients has improved since the early 1980s with refinements in immunosuppression therapy and surgical techniques, the morbidity and mortality rates of these patients are still greater than those of the general population.[1, 2] The quality of life and survival are still diminished in comparison with those of age- and sex-matched controls.[3, 4] As long-term survival after transplantation has improved, side effects caused by immunosuppression have become major causes of mortality. Cardiovascular complications, renal failure, diabetes, and de novo malignancies are the most frequent causes of late mortality.[5, 6] In a recent article, Pons et al.[1] reported that immunosuppression withdrawal was followed by improvements in renal function, hypercholesterolemia, hyperuricemia, arterial hypertension, and diabetes. Because of this, the transplant community has focused its efforts on studying the possibility of minimizing immunosuppression therapy.[7-9]

Immunological tolerance refers to the phenomenon by which some patients maintain normal allograft function as well as a normal immunological response in the absence of immunosuppression.[10-13] It was described 60 years ago,[11] but it has gained greater resonance only in recent years. Some authors have studied the potential of immunosuppression withdrawal to minimize immunosuppression side effects and improve patients' life expectancy and quality of life.[14-17]

According to different series, 20% to 40% of liver recipients show this potential,[14, 18, 19] and this is particularly true for patients with a longer time from transplantation[20, 21]; the proportion reached up to 60% in a series of pediatric recipients of parental liver grafts.[22]

Some studies have tried to investigate the mechanisms and predictive factors of tolerance, but their number in this field is low. Several of them have examined the differences between tolerant and nontolerant patients at the baseline and during immunosuppression withdrawal,[20, 23] whereas others have focused on the differences between the 2 groups after tolerance has been achieved.[21, 24] The differences at the baseline were not consistent in the 2 main studies, and the determination of some of these characteristics requires difficult or invasive procedures.[20, 23] Previously, the main characteristics that differed between tolerant and nontolerant patients were observed after immunosuppression withdrawal but not at the baseline.

The aim of this study was to find possible biomarkers of immune tolerance after LT and before the initiation of immunosuppression withdrawal so that they could be used as factors predictive of tolerance.


In this prospective study, 31 adult LT recipients with adverse effects of immunosuppression or with a history or high risk of de novo malignancies were initially screened for immunosuppression withdrawal. The investigational ethics committee of the University of Navarra approved the protocol for this study.

Inclusion Criteria

To be included in this study, patients needed to fulfill the following criteria: (1) deceased donor LT at least 3 years before inclusion; (2) absence of autoimmune disease before and after transplantation (ie, primary biliary cirrhosis and autoimmune hepatitis); (3) an absence of active viral hepatitis (in the case of a previous hepatitis C or B virus infection, a negative viral load in the last 6 months); (4) freedom from rejection in the last 12 months and persistently normal liver function tests in the last 6 months [defined as aspartate aminotransferase (AST), alanine aminotransferase (ALT), alkaline phosphatase (ALP), and serum bilirubin levels below the upper levels of normal at our laboratory (AST < 25 IU/mL, ALT < 29 IU/mL, ALP < 86 IU/mL, and bilirubin < 1.2 mg/dL); and (5) at least 1 of the following conditions: (a) relevant immunosuppression side effects such as diabetes, arterial hypertension, renal insufficiency, and dyslipidemia; (b) the presence of nonhepatic and noncutaneous malignancies or recurrent skin neoplasia before or after transplantation; and (c) a high risk of noncutaneous de novo neoplasia (including a history of active smoking and an age > 60 years).[25]

Study Design and Procedures

After patients signed the informed consent form, we performed a baseline assessment that included liver function tests (AST, ALT, ALP, and serum bilirubin) and liver biopsy in order to rule out any sign of rejection such as ductopenia or significant inflammatory infiltrates in more than one-third of the portal tracts. At the time, a baseline immunological analysis was performed, and it included lymphocyte subpopulations [T cells, CD4+ cells, CD8+ cells, B cells, natural killer (NK) cells, and regulatory T cells (Tregs)] and studies of lymphocyte proliferation.

Immunological Analysis

Analysis of Lymphocyte Subpopulations

Peripheral blood (100 μL) was stained with anti-CD3/fluorescein isothiocyanate (clone SK7), anti-CD56/phycoerythrin (PE; clone C5.9), anti-CD19/peridinin chlorophyll protein/cyanine 5.5 (clone SJ25C1), anti-CD4/PE/cynanine 7 (clone SK3), anti-CD8/allophycocyanin (APC; clone SK1), and anti-CD45/APC-H7 (clone 2D1) for 15 minutes at room temperature in the dark. After the staining, 2 mL of a fluorescence-activated cell sorting lysing solution (Becton-Dickinson, Erembodegem, Belgium) was added. After 10 minutes at room temperature, the samples were centrifuged and resuspended in 500 μL of phosphate-buffered saline. Leukocytes (1 × 105) were acquired with a FACSCanto I flow cytometer (Becton-Dickinson).

T cells were defined as CD3+; among them, CD4+ and CD8+ subpopulations were also registered. B cells were defined as CD19+, and NK cells were defined as CD56+.

For the analysis of Tregs, 100 μL of peripheral blood was stained with anti-CD3/fluorescein isothiocyanate (clone SK7), anti-CD4/peridinin chlorophyll protein/cyanine 5.5 (clone SK3), anti-CD25/APC (clone 2A3), and anti-CD45/APC-H7 (clone 2D1) for 15 minutes at room temperature in the dark. Then, cells were permeabilized and stained with anti–forkhead box P3 (Foxp3)/PE (clone PCH101) according to the manufacturer's instructions (Bioscience, San Diego, CA). The samples were acquired according to a double-acquisition protocol. In the first acquisition step, 3 × 104 leukocytes were acquired, and a live gate was drawn to select SSClo/CD4+ events. In the second acquisition step, 3 × 105 leukocytes were acquired, and gated events were recorded. Treg lymphocytes were identified as CD4+CD25hi+Foxp3+.

Data were analyzed with FACSDiva software (Becton-Dickinson). All monoclonal antibodies except for anti-Foxp3/PE (Bioscience) were purchased from Becton-Dickinson.

Studies of Lymphocyte Proliferation

Peripheral blood mononuclear cells were obtained by purification in a Ficoll gradient from heparinized blood. After purification, 2 × 105 cells per well were cultured in flat-bottomed, 96-well microtiter plates in Roswell Park Memorial Institute 1640 medium supplemented with 10% fetal bovine serum and 1% antibiotics in the presence or absence of phytohemagglutinin (2 μg/mL). One day later, tritiated thymidine (0.5 μCi/well) was added. The next day, cells were harvested, and thymidine incorporation was measured with a scintillation counter (TopCount, Packard, Meriden, CT). Lymphocyte proliferation was expressed as the stimulation index (SI), which is the ratio of tritiated thymidine incorporation by stimulated cells to tritiated thymidine incorporation by unstimulated controls.


Immunosuppression therapy was divided into 6 to 10 fractions, and it was gradually withdrawn every month until complete withdrawal or rejection (as evidenced by an alteration of liver function tests). In case of abnormality of liver function tests (AST, ALT, ALT or bilirubin above the upper normal level) less than 3 times the upper normal level at follow-up visits, there was no further reduction of immunosuppression therapy, and liver function tests were repeated 7 to 14 days later. If the liver function test results normalized, immunosuppression was progressively withdrawn. If the liver function test results remained at least twice the upper level of normal, the immunosuppression was increased until the test results normalized. In the absence of improvement, liver biopsy was performed to confirm allograft rejection. For acute rejection, we increased immunosuppression until it was resolved.

After complete immunosuppression withdrawal, patients came for follow-up every 6 weeks for 4 months. We defined tolerance as the absence of any immunosuppression therapy and the maintenance of liver function test results within normal limits with a minimum follow-up of 4 months after immunosuppression withdrawal.

Statistical Analysis

Age, sex, liver disease etiology, time from transplantation, immunosuppressive therapy, and immunological analyses at the baseline were compared for tolerant and nontolerant patients.

Differences in continuous variables between the 2 groups were compared with the Mann-Whitney U test. Differences in noncontinuous variables between the groups were compared with the chi-square test. Numerical variables are presented as medians and interquartile ranges, and categorical variables are presented as counts and percentages. A P value less than 0.05 indicates statistical significance. A multivariate analysis was not performed because of the low number of patients included in the study.

SPSS 15.01 (SPSS, Inc., Chicago, IL) was used for the statistical analysis.


At the time of the study, 138 patients were undergoing periodical follow-up at our center; 67 (49%) were potential candidates for this prospective study. We included 31 patients and screened them for immunosuppression withdrawal. The details of patient selection are shown in Fig. 1.

Figure 1.

Algorithm for selecting the patients.

Seven of the initial 31 patients (23%) had significant inflammatory infiltrates in screening biopsy samples and were considered screening failures. The baseline characteristics of the remaining 24 patients are shown in Table 1. The main indications for transplantation were alcoholic cirrhosis and hepatitis B. The median age at the time of transplantation was 57 years, and the median period of follow-up from LT to inclusion was 112 months.

Table 1. Baseline Characteristics of the Study Population
  1. a

    The data are presented as medians and interquartile ranges.

  2. b

    Wilson's disease, primary sclerosing cholangitis, and hepatitis C.

Patients (n)24
Age at study inclusion (years)a65 (60-70)
Time from LT to study inclusion (months)a112 (72-160)
Sex [n (%)] 
Female1 (4.2)
Male23 (95.8)
Indication for LT [n (%)] 
Alcoholic cirrhosis10 (41.7)
Hepatitis B virus cirrhosis10 (41.7)
Otherb4 (16.7)
Hepatocellular carcinoma [n (%)]7 (29.2)
Immunosuppression therapy at inclusion [n (%)]
Monotherapy16 (66.7)
Mycophenolate mofetil7 (29.2)
Cyclosporine A2 (8.3)
Tacrolimus5 (20.8)
Sirolimus2 (8.3)
Combined therapy8 (33.3)
Cyclosporine A + mycophenolate mofetil1 (4.2)
Tacrolimus + mycophenolate mofetil5 (20.8)
Tacrolimus + azathioprine1 (4.2)
Cyclosporine A + azathioprine1 (4.2)

Fifteen patients (48% of the screened patients and 62.5% of the 24 patients included in the study) were tolerant and maintained good allograft function at a median of 14 months (interquartile range = 8.5-22.5 months) after complete immunosuppression withdrawal. At the time of the analysis of the study results, only 2 patients had less than 6 months of follow-up, whereas 10 of them had been followed for more than 12 months after complete immunosuppression withdrawal; none of them developed graft rejection or had to start immunosuppressive therapy.

The remaining 9 patients (37.5%) showed abnormal liver tests in the follow-up, which prevented the complete withdrawal of immunosuppression. For 2 of these patients, liver function did not normalize after the increase in immunosuppression, so we performed liver biopsy. Biopsy samples revealed mild chronic rejection in the first patient and acute rejection in the second patient, who was controlled with a steroid treatment.

A comparison of tolerant and nontolerant patients is shown in Table 2. Tolerant patients had a longer median time since transplantation (156 months for tolerant patients versus 71 months for nontolerant patients, P = 0.003) and a lower median SI (7.49 for tolerant patients versus 41.73 for nontolerant patients, P = 0.01; Fig. 2). The rest of the comparisons of the 2 groups were nonsignificant.

Table 2. Comparison of Tolerant and Nontolerant Patients
 Tolerant Patients (n = 15)Nontolerant Patients (n = 9)P Value
  1. a

    The data are presented as medians and interquartile ranges.

Months since transplantationa156 (101-182)71 (51-88)0.003
Age at study inclusion (years)a65 (60-72.5)65 (60-66)0.59
Sex (n)  0.42
Indication for LT [n/N (%)]  0.42
Alcoholic cirrhosis5/10 (50)5/10 (50) 
Hepatitis B virus cirrhosis6/10 (60)4/10 (40) 
Other4/4 (100)0/4 (0) 
Hepatocellular carcinoma [n (%)]3 (42.9)4 (57.1)0.20
Single immunosuppression drug at inclusion [n/N (%)]  0.74
Yes12/16 (75)4/16 (25) 
No3/8 (37.5)5/8 (62.5) 
Lymphocytes T (/μL)a1497 (1049-1972)1607.0 (1360-2367)0.18
Lymphocytes T (%)a26.5 (22.6-31.2)32.6 (21.9-37.0)0.22
NK cells (/μL)a225 (132-292)131 (94-325)0.29
NK cells (%)a14.0 (9.9-20.7)8.3 (6.0-13.7)0.07
Lymphocytes T cells (/μL)a946 (682-1513)1276 (1064-1672)0.12
Lymphocytes T cells (%)a73.1 (61.3-80.6)77.2 (66.6-83.0)0.35
CD4+ cells (/μL)a618 (385-802)731 (506-833)0.35
CD4+ cells (%)a41.5 (34.9-47.0)41.7 (34.0-45.3)0.61
CD8+ cells (/μL)a363 (251-752)658 (359-978)0.07
CD8+ cells (%)a25.2 (23.1-34.6)37.0 (23.0-46.3)0.22
Tregs (/μL)a38 (25-67)47 (30-70)0.61
Tregs (%)a8.0 (5.8-9.1)5.7 (5.2-8.9)0.49
Lymphocytes B cells (/μL)a134 (71-212)227 (106-322)0.15
Lymphocytes B cells (%)a9.5 (6.8-12.1)9.9 (6.8-22.4)0.45
SIa7.49 (2.1-22.9)41.73 (18.8-64.9)0.01
Figure 2.

(A) Time from transplantation to study inclusion. The time from transplantation was longer for tolerant patients (median = 156 months) versus nontolerant patients (median = 71 months, P = 0.003). (B) SI for nontolerant and tolerant patients. Tolerant patients showed a lower basal SI (median = 7.49) than nontolerant patients (median = 41.73, P = 0.01).

Correspondingly, the patients were divided into groups according to the period between transplantation and immunosuppression withdrawal (n = 12 for <10 years and n = 12 for >10 years) and according to the SI (n = 12 for <20 and n = 12 for >20). Four groups with different magnitudes of success with the withdrawal protocol were established. In the first group (n = 7 for a period > 10 years and an SI < 20), 100% of the patients reached tolerance. In the 2 intermediate groups (n = 10 for a period > 10 years and an SI > 20 or a period < 10 years and an SI < 20), 60% reached tolerance. In the last group (n = 7 for a period < 10 years and an SI > 20), 29% of the patients were tolerant.


Operational tolerance has been reported in 20% to 40% of LT recipients in different published series[14, 18, 19]; however, the proportion of tolerant patients was as high as 60% in a series of pediatric recipients of parental living donor grafts.[22]

In an intention-to-treat analysis, 24 members of the total population with more than 3 years of follow-up after LT attempted weaning from immunosuppression (17.4%). Fifteen (10.9%) were tolerant.

We attempted weaning from immunosuppression in 24 patients: 62.5% were tolerant by the end (48% of those initially screened). Most importantly, we identified 2 easily obtained variables predictive of tolerance: a longer period since transplantation and lower stimulation lymphocyte reactivity (expressed as the SI).

The first of these 2 variables is in agreement with previous studies. Bohne et al.[20] and Martínez-Llordella et al.[21] also reported an association between tolerance and a longer time since transplantation. On the other hand, the higher SI could explain more powerful immunological reactivity in patients who do not achieve tolerance (even when they are on immunosuppressive therapy).

The phytohemagglutinin SI is a simple and reliable test when it is done under the same conditions.[26] It measures the cellular immune response. Until now, the monitoring of immune status with phytohemagglutinin in the transplant population has been used for predicting the risk of acute rejection or infection, but the results have not been consistent.[27-29] This test has also been used in other fields to monitor the immune response during and after treatment in, for example, oncological patients and patients with an inborn immunodeficiency.[30-32] To the best of our knowledge, this is the first study describing the SI as a possible biomarker for the prediction of tolerance.

In this series, the lower SI in tolerant patients versus nontolerant patients could indicate a global decrease in cellular immune responsiveness leading to tolerance. Similarly, these patients could have had a higher incidence of infections or neoplasia due to more profound immunosuppression, and this made immunosuppression withdrawal more important for them. In our series, the prevalence of posttransplant tumors, including cutaneous neoplasia, was 42% for patients with an SI < 20 and 25% for patients with an SI > 20 (P = not significant). The potential role of the SI as a measure of global immunosuppression could be evaluated in future studies.

One of the most surprising results of this study is the high rate of successful immunosuppression withdrawal in the study population. This may be partly related to a selection bias due to a combination of factors [eg, older patient age, longer period between transplantation and immunosuppression withdrawal, and LT indications (mainly alcoholic cirrhosis and hepatitis B)] differing from those in previous studies.

The average age of our study population exceeded that of most previous studies,[1, 18, 21, 23] but it was comparable to the ages for Tisone et al.'s series[17] and Bohne et al.'s tolerant group.[20] An older age has been associated with a lower risk of early acute rejection,[33, 34] and this corroborates the idea that older patients have a less reactive immune system. In previous reports, successful immunosuppression weaning was associated with an older age, as reported by Pons et al.[1] and Bohne et al. In the present study, the median ages of tolerant and nontolerant patients were the same.

Another factor is that the time from transplantation was also longer than that in previous series. Therefore, the high rate of tolerance in the present study could be partially associated with the longer time since transplantation. In fact, tolerant patients had a longer period between transplantation and immunosuppression withdrawal than nontolerant ones not only in our series but also in previously published series.[20, 21]

An additional factor that could be associated with this high rate of tolerance was the indication for LT. In this study, the majority of the patients underwent transplantation for alcoholic cirrhosis or hepatitis B, and in comparison with other series,[18, 20, 21, 24] the percentage of patients with hepatitis C was lower. This may have resulted in the selection of patients with a less reactive immune response.

Gómez-Manero et al.[33] found that patients who underwent LT for hepatitis B and alcoholic cirrhosis had the lowest risk of early acute rejection. We presume that these patients had a less reactive immune system, which made them more prone to tolerance. Unfortunately, the small study population does not allow a statistical analysis of this.

Our results must be evaluated with caution. As in other series, the patients included in this immunosuppression withdrawal study were carefully selected. In agreement with this, the tolerant patients represented less than 11% of the patients followed at our center.

The mechanisms of tolerance are not well understood; NK cell activity seems to play an important role in this phenomenon, but the exact mechanism of action is not well known. Bohne et al.[20] reported a higher baseline proportion of NK cells in tolerant patients as well as overexpression of NK-related genes. Martínez-Llordella et al.[21, 24] found an enrichment of NK receptors in tolerant patients after immunosuppression withdrawal. In our study, the difference in the percentages of NK cells in tolerant and nontolerant patients was not significant, but a higher proportion of NK cells was found in tolerant patients. However, this issue needs further investigation in order to explain the possible role of NK cells in the tolerance phenomenon.

Some series have found an increase in Tregs in tolerant patients during and after immunosuppression withdrawal but no difference in the total count and percentage of Tregs at the baseline.[20, 23] In this study, the basal counts of Tregs were similar for tolerant and nontolerant patients.

A recent study[35] compared the transcriptional and blood cell/phenotypic markers in tolerant liver and kidney transplant patients. The expression of total B cells and the overexpression of specific transcripts were found in tolerant kidney recipients. Those findings were absent in the tolerant and nontolerant liver recipient groups. A posterior analysis for identifying an increase in activated and memory B cell subpopulations in tolerant liver recipients failed. This is in agreement with the findings of the present study, in which there were no differences between the 2 groups in the total amount or percentage of B cells.

The small study populations present a limitation: in the different prospective studies published up to now, the median number of patients has been approximately 30,[14, 17, 18, 20, 23, 36-38] and this does not allow consistent conclusions. The present study has the same limitation. However, the previous reports and the present one suggest that a significant proportion of LT recipients could be tolerant. A large-scale, prospective, multicenter study is needed to validate these and other possible biomarkers of tolerance in LT.

Another possible limitation of the present study is the lack of postweaning liver biopsy for the tolerant group, despite recent recommendations.[39] We did not perform liver biopsy in the nontolerant group because their liver function test results improved with increases in immunosuppression therapy. Only 2 patients in this group persisted with abnormal liver tests, so biopsy was performed, and confirmed 1 case of chronic rejection and 1 case of acute rejection.

In conclusion, in select LT patients, immunosuppression withdrawal can be safely achieved. A combination of 2 simple characteristics—a long period since transplantation and a low SI—may be helpful in selecting patients with a potentially high rate of success with immunosuppression withdrawal.


The authors thank J. A. Pons, M.D., for his critical revision of this article and Juliane Chaccour for her medical writing services.