Current strategies for immunosuppression in liver transplant (LT) recipients include the design of protocols targeting a more individualized approach to reduce risk factors such as renal failure, cardiovascular complications and malignancies. Renal injury in LT recipients may be often multifactorial and is associated with increased risk of post-transplant morbidity and mortality. The quest for low toxicity immunosuppressive regimens has been challenging and resulted in CNI minimization protocols or CNI withdrawal and conversion to mycophenolate mofetil (MMF) and/or mammalian target of rapamycin inhibitor-based immunosuppressive regimens. Use of antibody induction to delay CNI administration may be an option in particular in immunocompromized, critically ill patients with high MELD scores. Protocols including MMF introduction and concomitant CNI minimization have the potential to recover renal function even in the medium and long term after LT. We review on hot topics in the prevention and management of acute and chronic renal injury in LT patients. For this purpose, we present and critically discuss results from immunosuppressive studies published in the current literature or presented at recent LT meetings.
biopsy-proven acute rejection
glomerular filtration rate
hepatitis C virus
Kidney Disease: Improving Global Outcomes
modification of diet in renal disease
model of end-stage liver disease
mammalian target of rapamycin
pre-end stage renal disease
risk, injury, failure, loss, and end-stage
Prevention and treatment of side effects of immunosuppressive drugs, namely calcineurin inhibitors (CNI), have gained an important role in the liver transplant (LT) setting. CNI are deemed to be associated with acute and chronic nephrotoxicity, de novo malignancies and cardiovascular or metabolic complications [1-3]. In addition, causes of structural renal damage before and after LT may include chronic liver diseases associated with concomitant kidney injury (e.g. glomerulonephritis in viral hepatitis) and cardiovascular risk factors (arterial hypertension, diabetes).
In acute CNI-induced nephrotoxicity, vasoconstriction in the afferent arterioles followed by a decrease in renal blood flow and reduced urinary excretion plays an essential role (Fig. 1) . Vasoconstriction is caused by an imbalance in the release of vasodilatory factors (prostaglandin E2, nitric oxide) and vasoconstrictive factors (thromboxane A2, endothelin 1, angiotensin II) and activation of the sympathetic nervous system. Renal function usually improves upon CNI cessation, whereas prolonged CNI-induced vasoconstriction may cause histological damage.
The incidence of chronic renal dysfunction characterized by arteriolar hyalinosis resulting in a variety of tubulointerstitial and glomerular lesions (Table 1) has been reported in up to 70% of patients in the long term after LT and varies widely depending on the length of follow-up, definition of chronic kidney disease and intensity of immunosuppressive therapy [5-7]. End-stage renal disease has been described to occur in 8.5% of patients during a follow-up of 11 years after LT .
|Etiology of chronic renal dysfunction|
|Calcineurin inhibitor-inducedArteriolar hyalinosis and sclerosisGlomerular sclerosis/fibrosisTubular atrophyInterstitial fibrosis||Concomitant risk factorsPre-transplant risk factors (Fig. 1)Post-transplant diabetes mellitusPost-transplant arterial hypertensionNephrotoxic drugsOthers|
|Classification of chronic kidney disease (KDIGO guideline) |
|Stage||Description||GFR (mL/min) per 1.73 m2||Related terms|
|1||Kidney damage with normal or increased GFR||≥90||Albuminuria, proteinuria, haematuria|
|2||Kidney damage with mildly decreased GFR||60–89||Albuminuria, proteinuria, haematuria|
|3||Moderately decreased GFR||30–59||Chronic renal insufficiency, early renal insufficiency|
|4||Severely decreased GFR||15–29||Chronic renal insufficiency, late renal insufficiency, pre-ESRD|
|5||Kidney failure||<15 (or dialysis)||Renal failure, uraemia, ESRD|
Strategies for avoiding acute renal dysfunction
Patients following LT are at major risk to develop acute kidney injury in the early post-operative phase. The reported incidence of replacement therapy-dependent acute kidney injury after LT has been reported to occur in 8 to 78% of patients [9, 10]. Such a wide range is attributed to the lack of a uniform definition of acute kidney injury. The Acute Kidney Injury Network developed a modified standard for diagnosing and classifying acute kidney injury. This classification system is a modification of the previously described risk, injury, failure, loss and end-stage (RIFLE) criteria [11, 12].
Several factors may trigger acute kidney injury in the post-transplant period. CNI nephrotoxicity plays a major role. New immunosuppressants and immunosuppressive approaches have been evaluated for preservation of renal function in de novo LT patients. Strategically, these trials have investigated 1) early minimization or early withdrawal of CNI treatment, 2) delayed CNI therapy and 3) CNI-free immunosuppressive regimens (Fig. 2).
Early minimization or withdrawal of calcineurin inhibitor therapy in de novo liver transplant patients
Mycophenolate mofetil and low dose calcineurin inhibitor treatment
A randomized controlled French multicentre trial by Boudjema et al.  (Table 2) compared tacrolimus (TAC) treatment at a standard dose (n = 100) with combined mycophenolate mofetil (MMF) and dose-reduced TAC therapy (n = 95) in de novo LT patients. Patients in both groups received corticosteroids at a descending dosage over a period of 6 months. The primary endpoints after 48 weeks' treatment were the onset of renal dysfunction, arterial hypertension, diabetes mellitus and acute transplant rejection. Transplant rejection occurred with a significantly higher frequency in the TAC standard therapy group than in the dose-reduced TAC/MMF group (46% vs. 30%, P = 0.024). This also applied to the event ‘renal dysfunction’ (42% vs. 24%, P = 0.004). There were no significant differences in the incidence of arterial hypertension or diabetes mellitus between both treatment arms. As was to be expected, changes in the blood count in terms of leucopenia or thrombocytopenia as well as diarrhoea were significantly more frequent in the TAC/MMF treatment group. In contrast to the ReSpECT study , evaluating TAC standard dose + steroids vs. TAC-reduced dose + MMF + steroids vs. daclizumab + delayed TAC therapy + steroids results from this French multicentre trial could demonstrate superiority of combined low dose TAC/MMF/steroid therapy over TAC standard dose/steroid treatment. Non-superiority of the low dose TAC group over the TAC standard dose group in the ReSpECT study might be because of the converging TAC concentrations (7.8 ± 2.7 ng/mL and 8.5 ± 3.7 ng/mL) at months 10–12.
|References||Design||Immunosupression||Acute rejection||Renal function (eGFR*), mean||Follow-up period||Comments|
|Boudjema et al. ||Randomized controlled|| |
A (n = 100): CNI+ ST
B (n = 95): rCNI + MMF +ST
P = 0.024
A: 78 ± 26
B: 90 ± 30
P = 0.004**
|12 months||rCNI +MMF+ST: Superior outcome of renal function and rejection rates|
|Fischer et al. ||Randomized controlled|| |
A (n = 102): CNI+anti-CD25 ± ST (4–8w)
B (n = 101): CNI+ anti-CD25 ± ST (4–8w)->CNI weaning+ EVL (8w)->switch to EVL
P = 0.684
A: 72.1 ± 24.5
B: 80.3 ± 26.4
P = 0.021
|11 months||Comparable incidence of graft loss, death and graft rejection. Superior renal function under EVL-regimen|
|Saliba et al.||Randomized controlled|| |
A (n = 243): CNI±MMF +ST (4w)
B (n = 245): CNI±MMF+ST (4w) -> rCNI+ST+EVL
C (n = 231): CNI±MMF +ST (4w)-> CNI tapering+EVL+ST->
(month 4) EVL+ST
P = 0.005
A: 70.3 ± 23.0
B: 80.9 ± 27.3
C: 80.8 ± 28.8
P < 0.001 (A vs. B)
|12 months|| |
CNI-free EVL arm discontinued caused by significantly higher rejection rate
Superior renal function under rCNI+ST+EVL
|Yoshida et al. ||Randomized controlled|| |
A (n = 76): CNI+MMF+ ST
B (n = 72): Anti-CD25 + delayed rCNI+MMF+ ST
P = 0.68
P = 0.038 at 6 months
P = 0.587 at 12 months
|12 months||Superior renal function under delayed rCNI only in the early post-transplant period|
|Bajjoka et al. ||Retrospective cohort study|| |
A (n = 80): CNI+MMF+ST
B (n = 118): ATG+delayed CNI+MMF+ST
P = 0.08
P < 0.001
|12 months||ATG induction with delayed CNI: lower incidence of early acute rejection and superior renal function|
|Soliman et al. ||Retrospective|| |
A (n = 129): CNI+ST
B (n = 262): ATG+
P = 0.001
A: 75 ± 21
B: 81 ± 23
P = 0.02
|12 months||ATG induction with delayed CNI: lower rejection rate and beneficial effects on renal function|
|Tepermann et al. ||Randomized controlled|| |
A (n = 137): MMF+CNI±antibody induction±ST
B (n = 139): MMF+CNI±antibody induction±ST -> (30–90 days) -> MMF+SRL
BPAR≥2 and RAI≥4:
GFR increase (%)
|12 months|| |
MMF+SRL: higher rejection rate and superior renal function;
study prematurely terminated because of SRL side effects and poorer outcome
|Garcia-Valdecasas et al. ||Randomized controlled|| |
Total (n = 250)
A: Anti-CD25 + Bela HD +MMF
B: Bela HD+MMF
C: Bela LD+MMF
|12 months||Study prematurely terminated caused by higher mortality and graft loss in the Bela groups|
In the multicentre randomized (1:1) controlled PROTECT study (CRAD001HDE10), de novo patients were treated with CNI [ciclosporin (CSA)or TAC] + basiliximab ± steroids for 4–8 weeks after LT and were then randomized to an everolimus (EVL)-based treatment arm or a CNI-based control arm . In the EVL-based treatment arm (n = 101), a 70% reduction of CNI (± steroids) was carried out over a period of 8 weeks, followed by treatment with EVL ± steroids. In the control arm (n = 102), treatment with CNI (standard dose ± steroids) was continued.
The primary endpoint, ‘difference in calculated glomerular filtration rate (GFR) of at least 8 mL/min between the two treatment arms using Cockcroft-Gault formula', could not be achieved by month 11, whereas using the modification of diet in renal disease (MDRD) equation, the endpoint could be achieved having a difference in calculated GFR of −7.8 mL/min (P = 0.02). The incidence of graft rejection, graft loss and deaths did not differ significantly between the two treatment arms.
In a multinational, randomized controlled licensing trial (CRAD001H2304), a TAC-free and a TAC-reduced regimen were compared with a TAC standard dose (TAC-C) regimen in terms of efficacy and safety in de novo LT patients  (Table 2). After a 30-day run-in phase on TAC-based immunosuppression (±MMF), patients were randomized to an EVL/prednisone/TAC-free group (TAC-WD) including TAC withdrawal at month 4 post-LT, an EVL/prednisone/reduced TAC group (EVL + rTAC) or a standard TAC control group (TAC-C). The primary combined endpoint comprised the events ‘biopsy-confirmed acute rejection’, ‘allograft loss’ or ‘death’. The secondary endpoint was ‘renal function’ at month 12. The TAC-WD arm had to be discontinued prematurely because of a significantly higher incidence of biopsy-confirmed acute rejections [19.9% (TAC-WD) vs. 4.1% (EVL + rTAC) vs. 10.7% (TAC-C)]. At month 12, a significantly higher proportion of patients in the TAC-C group had reached the combined primary endpoint compared with the EVL + rTAC group (9.7% vs. 6.7%, P < 0.001). Renal function was significantly better (P < 0.001) in the EVL + rTAC arm than in the TAC-C arm. The increased rejection rate in the TAC-WD group at month 4 may be because of the fact that, unlike the CRAD001HDE10 study, no induction therapy with basiliximab and no weaning of CNI over 8 weeks were carried out. Instead, CNI were stopped abruptly. Given the results of the CRAD001H2304 study, EVL as monotherapy will most likely not receive a licence in the LT setting, thus it can be assumed that recommended treatment with EVL according to the manufacturer will include TAC combination therapy.
Delayed calcineurin inhibitor treatment using induction therapeutic agents
In a recently published analysis of data from the United Network for Organ Sharing (UNOS), including 16,989 LT candidates, patient and transplant survival in non-HCV patients were not significantly different in those who received anti-thymocyte globulin (ATG) alone (n = 452), ATG + steroids (n = 1758), daclizumab alone (n = 683) or steroids alone (n = 14,005) . The treatment with IL-2 receptor antagonists in combination with reduced or delayed CNI therapy was associated with better renal function (mean difference in GFR: 6.29 mL/min; CI 1.66–10.91) and a lower incidence of post-transplant diabetes. The use of induction therapeutic drugs in the LT setting compared with kidney transplanted patients often varies between transplant centres. According to the manufacturer, the licence for daclizumab in the European Union was withdrawn in January 2009. In Germany, the IL-2 receptor antagonist basiliximab is commonly used in transplant centres employing induction therapeutic agents. As shown by previous studies, induction therapy definitely has its role in preserving and improving renal function in LT patients (18–20, Table 2).
Calcineurin inhibitor-free immunosuppressive regimen in de novo LT patients
As the immunosuppressive potential of mammalian target of rapamycin (mTOR) inhibitors and antimetabolites such as MMF and azathioprine is lower than for CNI, a CNI-free immunosuppressive regimen is associated with a higher risk of acute rejections in the early phase. Several studies have been stopped for these reasons thereby not allowing any conclusions about development of renal function in the long term after LT [21, 22].
Moreover, in view of feared increased incidence of infectious and vascular complications and wound healing problems, mTOR-inhibitors EVL and sirolimus (SRL) have not been administered in the immediate post-operative phase or have been used frequently no earlier than 4 weeks after LT [23-25]. Both drugs are licensed for kidney and heart transplant patients but not as yet for LT patients.
Mycophenolate mofetil and sirolimus therapy
In a prospective randomized, controlled multicentre study, patients with combined MMF and CNI therapy were randomized to receive MMF (2–3 g/day) and SRL (2–4 mg/day) or to maintain their current immunosuppressive regimen. Centres could decide individually whether antibody induction therapy and/or corticosteroids were administered additionally. Interim 12-month results of 139 MMF/SRL-treated and 137 MMF/CNI-treated control patients revealed that mean GRF increase was higher in the CNI-free group as compared to controls (22.1% vs. 5.25%), whereas biopsy-proven acute rejection grade ≥2 occurred less frequently in controls . This ‘Spare the Nephron Study’ was prematurely terminated as a result of mTOR inhibitor-induced side effects and poorer outcomes.
Calcineurin inhibitor-free ‘bottom-up’ immunosuppression
Schnitzbauer et al.  conducted a pilot study (Patron study) designed as a single arm study investigating a CNI-free ‘bottom-up’ immunosuppressive regimen. As main inclusion criterion, a creatinine level >1.5 mg/dL and/or GFR <50 mL/min should be present. Patients received MMF (1–2 g/day) and prednisolone (1 mg/kg body weight, tapered off over 6 months in accordance to the standard regimen), as well as basiliximab on day 0 and 4 (20 mg i.v. each). SRL (initial loading dose of 5 mg, further dosing with target levels of 4–8 ng/mL) was administered as bottom-up treatment no earlier than at day 10 post-operatively. A total of 27 patients have been enrolled in this study. The primary endpoint was the incidence of steroid-resistant acute rejection within 30 days after LT. Notably, no such event occurred in this study. Overall, safety assessments revealed an increased risk for acute rejections beyond day 30 after LT, and results suggest addition of CNI in particular in those patients who presented with an acute rejection episode under mTOR-based regimen (personal communication with A.A. Schnitzbauer).
Goralcyk et al.  pursued a very similar approach (CILT Study NCT00890253) using basiliximab on day 0 and 4, mycophenolate sodium (MPS) (alternatively MMF), prednisolone and EVL [from day 5, initial loading dose of 5 mg, then adapted to the target level (4–8 ng/mL)]. As in the Patron study, results have not yet been published.
CNI-free bottom-up immunosuppression is most likely to be suitable for critically ill patients with a high model of end-stage liver disease (MELD) score because they tend to require less immunosuppression than patients in lower MELD categories, and for patients in whom improvement of renal function is paramount. It is important to bear in mind the myelosuppressive side effects of MMF or MPS and mTOR-inhibitors which, if used in combination, not uncommonly necessitate dose reductions or treatment withdrawal.
Belatacept Study NCT00555321
Belatacept blocks the costimulatory signal by binding to CD80 and CD86 antigens, thus inhibiting the complete activation of T cells and promoting anergy and apoptosis . In a prospective randomized phase 2 trial, the incidence of rejection and transplant loss as well as mortality were evaluated including Belatacept-based immunosuppressive regimens (group 1, basiliximab + belatacept + MMF; group 2, belatacept (higher dosage) + MMF; group 3, belatacept (lower dosage) + MMF) vs. TAC-based immunosuppression (group 4, TAC + MMF; group 5, TAC) . Better renal function and a lower rate of cardiovascular, metabolic and neurotoxic complications were achieved in the groups of LT patients treated with belatacept. These results are in agreement with other studies in transplanted patients showing that metabolic end points were more favourable in belatacept versus CNI-based treatment groups . Two cases of post-transplant lymphoproliferative disease and one case of multifocal leukoencephalopathy occurred with belatacept. Mortality and transplant loss were also higher than with TAC/MMF-based and TAC immunosuppression, so that this study has meanwhile been halted because of the poorer outcome in the belatacept-based treatment groups (personal communication with the pharmaceutical company).
Sotrastaurin Study NCT01128335
Sotrastaurin, a novel oral protein kinase C inhibitor, blocks early T-cell activation. The efficacy, safety and tolerability of sotrastaurin/TAC-based vs. MMF/TAC-based immunosuppression in de novo LT recipients has been investigated in a randomized, controlled multicentre phase 2 trial. There are no interim results available so far.
Strategies for avoiding/improving chronic renal dysfunction
Emerging evidence suggests that acute kidney injury is accompanied with increased risk of progression of pre-existing chronic kidney disease to advanced stage renal disease and/or further impairment of kidney function because of CNI-induced nephrotoxicity (Table 1) [29, 30]. Despite CNI-related parenchymal damage, various clinical trials have demonstrated that CNI dose reduction or discontinuation results in an improvement of renal function in the majority of LT patients, thus suggesting a partly dose-dependent nephrotoxic effect and reversible functional kidney damage [31, 32].
Switch from calcineurin inhibitor to combined low dose calcineurin inhibitor and mycophenolate mofetil therapy
Combined MMF and minimal dose CNI treatment leads to a significant improvement in renal function without an increased risk of rejection (Fig. 3). The safety and efficacy of this approach has been proven in numerous prospective randomized studies [33-35].
In one of these trials, LT recipients were treated with MMF and reduced CNI (≥50% of initial dose) and compared with MMF-free controls in which CNI doses could be reduced up to 75% of the initial dose . In the MMF/low dose CNI group, renal function significantly improved at 1 year as compared to controls, and no rejection episode occurred.
In another prospective, randomized study, LT patients with renal dysfunction were randomized either to receive MMF followed by stepwise reduction of CNI with defined minimal CNI-trough levels (MMF group), or to continue their maintenance CNI dose (control group) . CNI dose was progressively tapered to achieve defined target trough levels as low as 2–4 ng/mL for TAC and 25–50 ng/mL for CSA. In the MMF group, renal function improved >10% in 62% of patients, was stable in 36% and impaired >10% in only 2% after 12 months compared with baseline values. In the control group renal function tended to deteriorate during this study period. Regarding cardiovascular risk profile, data were in agreement with other studies showing that CNI reduction or withdrawal is associated with significantly decreased systolic and diastolic blood pressure values [33, 37, 38].
Long-term follow-up data over 60 months from a study by Kornberg et al. demonstrated sustained renal response after conversion to reduced CNI and MMF therapy .
Mycophenolate mofetil and mammalian target of rapamycin-based immunosuppression
In a prospective randomized study, patients with standard-CNI dose (n = 75) were compared with patients (n = 75) who were switched to MMF monotherapy . During a five-year follow-up period only a trend towards a higher rejection rate was found in the MMF group. No chronic rejections were reported and patient and transplant survival, the incidence of cardiovascular, gastrointestinal and neurological side effects were comparable in both groups. During the observation period a significant improvement in renal function was observed in patients switched to MMF.
In a large-scale prospective randomized, controlled trial, in which 607 patients were either switched from CNI to SRL (n = 393) or maintained on treatment with CNI-based immunosuppression (n = 214) 6–144 months after LT, mortality at month 12 was comparable in both groups . Of note, the SRL-converted group had a higher rejection rate and a higher drop-out rate than the CNI-treated group at week 52.
In a retrospective study, CNI therapy could be discontinued within 12 months after the addition of EVL in 60% of patients . Four patients (1.6%) presented with rejection which was successfully treated by intensified EVL therapy (n = 4) or the addition of CNI. Renal function estimated by GFR had significantly improved at month 12 (64.2 ± 30 mL/min vs. 68.4 ± 32.5 mL/min, P = 0.007) upon switch to EVL.
The few published prospective randomized controlled studies investigating SRL- or EVL-based immunosuppression are flawed by small patient cohorts and rather short follow-up periods [43-45]. In a recent prospective study including 21 patients with chronic renal dysfunction, CNI elimination was feasible in 95.2% of patients who were converted to EVL . There were no rejection episodes after switch during a mean-follow-up of 20 months. Kidney function significantly improved at day 30, but failed to show an advantage after 3 and 6 months post-conversion. The lack of significant improvements in renal function can be attributed to delayed introduction of mTOR-inhibitors and to concurrent aetiological factors of renal dysfunction such as diabetic or hypertensive nephropathies.
To date, very few information is available from studies investigating combined MMF and mTOR-inhibitor therapy in long-term LT patients [47, 48]. In a retrospective study, efficacy, safety and development of renal function was investigated in LT patients switched from CNI to SRL monotherapy (Group A, n = 26) or SRL combined with steroids and/or MMF (Group B, n = 34) . Mean serum creatinine was 2.1 ± 0.7 mg/dl prior to conversion and decreased under CNI-free regimens (n = 60) to 1.6 ± 0.56 mg/dl at 6 months (P < 0.001). Seven (11.3%) patients had acute rejection (Group A, two; Group B, five; P=NS). Most episodes of rejection (5/7; 71%) and adverse events (45/54; 83%) occurred within 6 months of conversion. The authors concluded that both SRL monotherapy and SRL combined with steroids and/or MMF was efficacious and safe in LT recipients.
There is ongoing research to achieve optimal balance between efficacy and toxicity of immunosuppressive regimens. Reduction of modifiable risk factors including adaption of immunosuppressive therapy to overcome CNI-related adverse events is recommendable to prevent the development of serious complications and hence improve the outcome for LT patients. For this purpose, immunosuppressive regimens need to be individualized according to the patient's immunological risk and comorbidities.
In de novo LT patients, regimens including IL-2 receptor antibody induction and short-term CNI therapy followed by mTOR inhibitor-based immunosuppressive treatment and concomitant slowly CNI weaning have shown to provide adequate immunosuppression. New protocols using costimulation blocker Belatacept have provided safe and effective immunosuppression without renal or cardiovascular toxicities in kidney transplant patients but raised safety concerns in LT patients.
CNI, at least at low doses, with or without other immunosuppressive drugs, has been so far the cornerstone of immunosuppressive regimens in a substantial proportion of LT patients. Many studies have been conducted to evaluate new immunosuppressive agents to overcome CNI-related nephrotoxicity after LT. It seems that not a single drug but drug combinations compete with CNI for future primary immunosuppressive protocols. Further results from randomized controlled studies are therefore urgently warranted to determine which drug combinations are the most promising approaches for potential introduction of CNI-free immunosuppressive regimens.