Early use of renal-sparing agents in liver transplantation: A closer look


  • The authors wrote the first draft of the manuscript (James F. Trotter wrote the introduction and the section titled Avoiding or Limiting the Use of Calcineurin Inhibitors in the Maintenance Phase; Monica Grafals wrote the section titled Renal Dysfunction in Liver Transplant Candidates and Recipients, and Angel E. Alsina wrote the section titled Calcineurin Inhibitor Avoidance in the Induction Period). The authors' contributions were merged together by Danita Sutton, Ph.D. (Oxford PharmaGenesis, Inc.), who provided editorial assistance to improve the flow and eliminate any redundancies. Sutton also provided assistance with figure preparation and styling of the manuscript for submission. The final draft was approved by all the authors.

  • Funding for editorial assistance from Oxford PharmaGenesis, Inc., was provided by Novartis Pharmaceuticals Corp. The authors are fully responsible for all content and editorial decisions, and they received no financial support or other form of compensation related to the development of this article. The opinions expressed in this article are those of the authors, and Novartis Pharmaceuticals had no influence on the contents.

  • James F. Trotter is part of the speakers' bureau for Novartis and Salix and has advisory arrangements with Novartis. Monica Grafals has nothing to disclose. Angel E. Alsina is part of the speakers' bureau for Bayer and Novartis and has advisory arrangements with Bayer and Onyx.

Address reprint requests to James F. Trotter, M.D., Baylor University Medical Center, 3410 Worth Street, Suite 860, Dallas, TX 75246. Telephone: 214-820-8500; FAX: 214-820-0993; E-mail: james.trotter@baylorhealth.edu


Renal dysfunction is a critical issue for liver transplant candidates and recipients. Acute nephrotoxicity and chronic nephrotoxicity, however, are the compromises for the potent immunosuppression provided by calcineurin inhibitors (CNIs). To maintain the graft and patient survival afforded by CNIs while minimizing renal dysfunction in liver transplant patients, the reduction, delay, or elimination of CNIs in immunosuppression regimens is being implemented more frequently by clinicians. The void left by standard-dose CNIs is being filled by nonnephrotoxic immunosuppressants such as mycophenolates and mammalian target of rapamycin inhibitors. The results of studies of renal-sparing regimens in liver transplant recipients have been inconsistent, and this may be explained upon a closer examination of several study-related factors, including the study design and the duration of follow-up. Liver Transpl 19:826-842, 2013. © 2013 AASLD.


adenosine triphosphate–binding cassette B1


adverse event




blood pressure


biopsy-proven acute rejection


trough concentration


calculated glomerular filtration rate




confidence interval


chronic kidney disease


calcineurin inhibitor




creatinine clearance




cyclosporine A


cytochrome P450 C28




estimated glomerular filtration rate




glomerular filtration rate




Modification of Diet in Renal Disease


Model for End-Stage Liver Disease


mycophenolate mofetil


mycophenolic acid


mammalian target of rapamycin


not significant




Preservation of Renal Function in Liver Transplant Recipients With Certican Therapy


randomized controlled trial


relative risk





According to data from the US Scientific Registry of Transplant Recipients,[1] an important trend in immunosuppression over the past 10 years is the increasing use of regimens that avoid calcineurin inhibitors (CNIs). Biological agents are increasingly being used as induction immunosuppression to reduce CNI exposure (Fig. 1). The use of biological agents for induction has essentially doubled in liver transplant recipients over the past decade from approximately 15% in 2000 to 30% in 2009 (the most recent available data). Thymoglobulin, the most commonly used biological agent, is used more than all of the other biological agents combined, and it is followed by daclizumab and basiliximab.[1]

Figure 1.

Use of biological agents for induction immunosuppression from 2000 to 2009 (based on data from the US Scientific Registry of Transplant Recipients[1]). The induction agents included anti-thymocyte globulin, OKT3, thymoglobulin, daclizumab, basiliximab, and alemtuzumab.

According to the registry (which reports immunosuppression regimens at the time of discharge and at specific postoperative intervals thereafter), clinicians in recent years have been limiting patient exposure to CNIs during the induction period after transplantation (Fig. 2).[1] The proportion of liver transplant recipients receiving only a CNI before hospital discharge has decreased by half from 40% in 2000 to 20% in 2009. With this reduction in the use of CNI monotherapy, more patients are being treated with reduced-dose CNIs in combination with nonnephrotoxic drugs, including mycophenolates and mammalian target of rapamycin (mTOR) inhibitors. The proportion of liver transplant patients receiving a combination of CNI and mycophenolate mofetil (MMF) or mycophenolic acid (MPA) has nearly doubled from 40% in 2000 to currently more than 70%. The percentage of patients receiving an mTOR inhibitor in the induction period by 2009 was still relatively small (<10%). Trends in general immunosuppression strategies by year during the maintenance phase (ie, up to 2 years after transplantation) are shown in Fig. 3. The exposure to CNIs as monotherapy during this phase decreased from 70% in 1999 to less than 50% by 2009. Over the same period, the administration of non-CNI, nonnephrotoxic drugs (primarily the mycophenolates MMF and MPA) doubled from 20% to 40%. Finally, the proportion of patients maintained entirely without CNIs is small and approaches 10% during the maintenance phase of immunosuppression management, whereas the use of mTOR inhibitors for maintenance remains relatively low (<10%).[1] In this article, we review renal dysfunction in liver transplant recipients both before and after transplantation, describe studies investigating CNI avoidance in the perioperative period and maintenance phase of immunosuppression in liver transplant recipients, and take a closer look at the inconsistent results associated with these studies in an attempt to explain the findings.

Figure 2.

Immunosuppression regimen at the time of patient discharge from 2000 to 2009 (based on data from the US Scientific Registry of Transplant Recipients[1]).

Figure 3.

Maintenance immunosuppression regimens from 1999 to 2007 (based on data from the US Scientific Registry of Transplant Recipients[1]).


Renal dysfunction presents a significant challenge for clinicians who treat liver transplant candidates and recipients because renal disease is an extremely common problem both before and after liver transplantation. Before liver transplantation, many patients have chronic renal dysfunction due to age, diabetes, and hypertension, which may be exacerbated by acute tubular necrosis or hepatorenal syndrome. Among patients with cirrhosis, 44% and 17% have been shown to have acute tubular necrosis and hepatorenal syndrome, respectively.[2]

The pathophysiology of cirrhosis involves portal hypertension leading to splanchnic and systemic arterial vasodilation. The resultant primary systemic arterial vasodilation unloads the arterial stretch receptors in the carotid sinus and aortic arch. This baroreceptor response then triggers the compensatory activation of the neurohumoral axis with stimulation of the renin-angiotensin-aldosterone system, sympathetic nervous system, and arginine vasopressin.

Stimulation of the renin-angiotensin-aldosterone system, sympathetic nervous system, and arginine vasopressin contributes to the maintenance of blood volume through sodium and water retention and increased cardiac output. This compensatory neurohumoral activation also leads to renal vasoconstriction. In decompensated cirrhosis, renal vasoconstriction increases, as does sodium and water retention (ascites); cardiac output falls, and this results in diminished effective renal arterial blood flow. This resultant diminished renal function is of a functional nature. The resultant renal vasoconstriction in patients with cirrhosis predisposes them to acute tubular necrosis, especially if a second adverse event (AE) occurs, such as excessive diarrhea with lactulose treatment, a gastrointestinal hemorrhage, sepsis, or toxin exposure. With a second AE leading to acute tubular necrosis, there is often evidence of tubular dysfunction, as assessed by diminished tubular sodium reabsorption despite a fall in the glomerular filtration rate (GFR).[3, 4]

Patients with liver disease are also predisposed to other renal conditions. For example, patients with hepatitis C are known to be at increased risk of membranoproliferative glomerulonephritis and fibrillary glomerulonephritis as well as posttransplant diabetes.[5] Patients with hepatitis B and patients with ulcerative colitis have a higher risk of membranous nephropathy and membranoproliferative glomerulonephritis.[5] Patients with alcoholic cirrhosis are at increased risk of immunoglobulin A nephropathy; however, clinically evident nephrotic syndrome is very uncommon.[5, 6] The overall prevalence of these renal comorbidities in patients with liver disease is not well established; however, it is prudent to monitor patients with hepatitis C, hepatitis B, and cirrhosis closely for any deterioration in renal function.[5]

Acute renal failure complicates up to 60% of liver transplants,[7] increases morbidity and mortality,[8-11] and is also associated with significant implications in terms of financial costs (primarily associated with prolonged hospital stays) and clinical management.[7] O'Riordan et al.[7] showed with a multivariate analysis that acute renal disease (acute renal failure and acute renal injury) was associated with several conditions before liver transplantation, including hypertension, alcoholic liver disease, and increased inotrope and aminoglycoside use.

Chronic kidney disease (CKD) represents a critical complication that affects the vast majority of patients who survive beyond the first 6 to 12 months after liver transplantation.[12, 13] Mortality among liver transplant recipients is increased for those with CKD: the development of end-stage renal disease after liver transplantation increases patient mortality more than 40%.[7]

In comparison with other solid organ recipients, liver transplant recipients may be at increased risk for posttransplant renal impairment because of hepatorenal syndrome or physiological changes associated with hepatorenal syndrome. Many liver recipients have a hepatorenal syndrome physiology characterized by renal vasoconstriction and hormonal changes, which may persist into the posttransplant period. The additional vasoconstriction associated with the administration of a CNI might lead to chronic renal impairment in some patients.

Acute nephrotoxicity and chronic nephrotoxicity represent the compromises for the use of CNIs in immunosuppressive regimens. As reviewed elsewhere, the use of CNI-based immunosuppression is one of the most important determinants of post–liver transplant CKD, and CNI nephrotoxicity is the most common cause of end-stage renal disease.[14] Lee et al.[15] studied risk factors associated with CKD after liver transplantation in 431 recipients with hepatitis B virus infections. CNI nephrotoxicity was found in multivariate analyses to independently and significantly increase the risk of CKD by a factor of 4.24. Because normal renal function is associated with better graft and patient survival in liver transplantation, CNI minimization protocols have been developed to prevent renal dysfunction. As described in the following sections, current strategies to overcome CNI toxicity include reducing CNIs with or without the addition of MMF and switching to an mTOR inhibitor.[16]


The use of CNIs for immunosuppression in liver transplant recipients is associated with impaired renal function, whereas mTOR inhibitors such as sirolimus (SRL) and everolimus (EVR) may provide alternatives to preserve renal function.[17] Clinically meaningful renal dysfunction may occur as early as the induction period, and this emphasizes the importance of early intervention to prevent or reduce the development of CKD.[13, 18] Results of clinical trials evaluating CNI avoidance in the induction period are summarized in Table 1.

Table 1. Summary of Studies Evaluating CNI Avoidance in the Induction Period
StudyNStudy DesignImmunosuppressive TreatmentKey ResultsLimitations
Fischer et al.[17]20312-month OL RCT of conversion from CNI to EVR (starting 4 weeks after transplantation) versus CNI continuation• When CNIs were reduced to 70% of initial dose, EVR was added at 1.5 mg bid to target C0 of 5-12 ng/mL (8-12 ng/mL for patients on CSA); CNIs were stopped completely at week 8.• CNI-based regimen was continued. Basiliximab induction was used. CSs were optional in both groups.For EVR plus CNI d/c versus CNI continuation at month 11:• Mean MDRD cGFR improved by 7.8 mL/min (P = 0.021).• CG cGFR improved by 2.9 mL/min (P = 0.457).• Mortality: 4.2% versus 4.1%.• BPAR: 17.7% versus 15.3%.• Efficacy failure: 20.8% versus 20.4%.• Overall d/c: 49.5% versus 38.2%.• AE-related d/c: 27.2% versus 15.7%.Many patients in the CNI d/c group were still on some CNI for up to 14-16 weeks, and this may have skewed the results in favor of no difference. 46% of the patients failed randomization; it is difficult to broadly apply the results to all liver transplant recipients. Follow-up was short (12 months).
Masetti et al.[19]7812-month OL RCT of early conversion from CSA to EVR versus CSA ± MMFCSA was used for 10 days after transplantation, and then patients were randomized 2:1 to• EVR at 2.0 mg/day to target C0 of 6-10 ng/mL, which was increased on day 30 [when CSA (100 ± 25 ng/mL) was discontinued] to 8-12 ng/mL until month 6 (6-10 ng/mL thereafter).• CSA, which was adjusted to C0 of 225 ± 25 ng/mL until day 30, 200 ± 25 ng/mL until end of month 6, and 150 ± 25 ng/mL thereafter. CSA dose could be reduced to introduce MMF (loading dose = 1 g bid, maintenance dose = 500 mg 3 times daily)For EVR conversion versus CSA:• Month 12 mean MDRD GFR: 87.6 versus 59.9 mL/min/1.73 m2 (P < 0.001).• Patients with GFR < 60 mL/min/1.73 m2: 15.4% versus 52.2% (P = 0.005).• 1-year freedom from efficacy failure: 75% versus 69%.• 6-month and 1-year survival: 92.3% and 90.4% versus 92.3% and 88.5%.• BPAR: 5.7% versus 7.7%.• Hepatic artery stenosis/thrombosis: 1.9% versus 15.4% (P = 0.04).30% of the patients were not randomized because of complications.There were baseline differences in MDRD GFR (81.7 and 74.7 mL/min/1.73 m2 for the EVR and CSA groups, respectively).Follow-up was short (12 months).
Levy et al.[20]11912-month RCT of different EVR doses versus placebo in patients receiving CSA and CSs followed by 36-month OL extension of patients receiving EVRAll patients took CSA to C0 targets of 150-450 ng/mL for weeks 1-4, 100-300 ng/mL for months 2-6, and 75-300 ng/mL for months 7-12. Patients were randomized to receive• EVR at 1 mg/day.• EVR at 2 mg/day.• EVR at 4 mg/day.• Placebo. CSs were tapered to ≥5.0 mg/day by month 3, and then CSs were given at maintenance doses, tapered, or discontinued.For EVR at 1, 2, and 4 mg/day versus placebo at month 12:• BPAR: 32.1%, 26.7%, 25.8%, and 40.0% (P = NS).• Graft loss: 0, 13.3%, 3.2%, and 6.7%.• Deaths: 17.9%, 3.3%, 12.9%, and 16.7%.• d/c due to AEs: 32.1%, 33.3%, 32.3%, and 13.3%.• CrCl at month 12: 67.3 ± 22, 53.7 ± 14.7, 59.1 ± 13, and 59 ± 21 mL/min (P = 0.15). AEs (anemia, tachycardia, constipation, edema, elevated Cr, and agitation) were more common with EVR versus placebo (P = NS).d/c rates were high (43.3%-64.5% by month 12). Standard-dose CSA was used. EVR doses were fixed (C0 levels were not determined).
De Simone et al.[21]71924-month multicenter, prospective, OL RCT of reduced or eliminated TAC plus EVR versus standard-dose controls (results for 12 months reported)• Group A: TAC elimination. EVR was adjusted to maintain C0 of 3-8 ng/mL to month 4 and then 6-10 ng/mL; TAC was tapered to C0 of 3-5 ng/mL by week 3 and eliminated at month 4.• Group B: EVR plus TAC minimization. EVR was used at 3-8 ng/mL; TAC was tapered to 3-5 ng/mL by week 3, and this was maintained for the rest of the study.• Group C: TAC controls. TAC was used for C0 of 8-12 ng/mL to month 4 and 6-10 ng/mL thereafter.Group A terminated prematurely because of a higher rate of treated BPAR. For groups A, B, and C at 12 months:• Treated BPAR, graft loss, or death: 24.2%, 6.7%, and 9.7% (B noninferior to C).• Treated BPAR for B and C: 2.9% and 7.0% (P = 0.04). eGFR changes at 12 months were superior for B versus C (difference = 8.5 mL/min/1.73 m2, P < 0.001). AEs that were more common for B versus C were peripheral edema (17.6% versus 10.8%, RR = 1.63, 95% CI = 1.03-2.56), leukopenia (11.8% versus 5.0%, RR = 2.38, 95% CI = 1.24-4.55), and serious infections (13.9% versus 7.9%, RR = 1.76, 95% CI = 1.03-3.00).The control arm did not allow for use of MMF and consequently used higher TAC levels.
Boudjema et al.[22]19548-week OL RCT of reduced-dose TAC plus MMF versus standard-dose TAC controls• TAC was used at 0.040 mg/kg bid to C0 ≤ 10 ng/mL for 6 weeks, ≤ 8 ng/mL from week 7 to month 6, and ≤ 6 ng/mL from month 7 to month 12 along with MMF at 1.5 g bid for 6 weeks and then 1.0 g bid to month 12.• TAC was used at 0.075 mg/kg bid to C0 ≥ 12 ng/mL for 6 weeks, ≥ 10 ng/mL from week 7 to week 12, ≥ 8 ng/mL from month 4 to month 6, and ≥ 6 ng/mL from month 7 to month 12.For reduced-dose TAC plus MMF versus standard-dose TAC at week 48:• Acute graft rejection: 30% versus 46% (P = 0.02).• Overall survival: 96% versus 93% (P = 0.40).• Graft survival: 94% versus 92% (P = 0.65).• Renal dysfunction, arterial hypertension, or diabetes: 64% versus 80% (P = 0.021).• Renal dysfunction: 24% versus 42% (P = 0.004).• Leukopenia: 29% versus 11% (P = 0.001).• Thrombocytopenia: 14% versus 4% (P = 0.017).• Diarrhea: 23% versus 7% (P = 0.002).The standard-dose TAC group did not receive MMF. There was no antibody induction therapy. There was good baseline renal function (GFR ∼ 100 mL/min). Study duration was <1 year.
Neuberger et al.[23]52512-month RCT of patients with good pretransplant renal function comparing 3 groups: (A) standard-dose TAC plus CSs, (B) MMF plus reduced-dose TAC and CSs, and (C) induction with daclizumab plus MMF and then reduced-dose TAC plus CSs• Group A: TAC was used at 0.10-0.15 mg/kg/day orally or 0.01-0.05 mg/kg/day IV to C0 > 10 ng/mL for month 1 and then according to documented center practice; CSs were given according to local practice.• Group B: TAC was used at 0.05-0.10 mg/kg/day orally or 0.008-0.04 mg/kg/day IV to C0 ≤ 8 ng/mL for the study duration, CSs were given according to local practice, and MMF was given at 1 g bid IV for 5 days and then orally.• Group C: Group C followed the same regimen as group B, but with TAC initiation delayed to day 5 plus daclizumab at 2 mg/kg within 12 hours of transplant and at 1 mg/kg on day 7.For groups A, B, and group C at 52 weeks:• eGFR decreases from baseline: 23.61, 21.22, and 13.63 mL/min (A versus C, P = 0.012; A versus B, P = 0.20).• BPAR: 27.6%, 29.2%, and 19.0%.• Study d/c: 44.6%, 28.0%, and 27.4%.• Patient survival: 90.6%, 88.7%, and 93.5%.• Graft survival: 93.9%, 94.0%, and 92.9%.• Change in serum Cr from baseline: 0.23, 0.17, and 0.12 ng/mL.• Recurrence of liver disease: 9.7%, 14.9%, and 22.1%.• Hypertension AEs: 40.7%, 29.0%, and 27.2%.• Leukopenia: 6.6%, 10.7%, and 16.0%.• Renal insufficiency: 21.4%, 11.8%, and 7.7%.Standard-dose TAC is rarely used without MMF in clinical practice. Generally more healthy patients with good pretransplant renal function are not typical liver transplant recipients (low baseline MELD scores). Follow-up was short (12 months).
Yoshida et al.[24]14812-month RCT comparing daclizumab induction plus delayed low-dose TAC versus standard-dose TAC induction/maintenance dosing• Daclizumab was used at 2 mg/kg IV ≤4 hours after transplantation and at 1 mg/kg IV on day 4; starting on days 4-6, TAC was used at 1 mg bid to C0 of 4-8 ng/mL.• TAC induction was used to target C0 of 10-15 ng/mL for 30 days and then 4-8 ng/mL thereafter. Both groups received MMF at 1000 mg bid within 8 hours of transplant; CSs were tapered until d/c at month 3.For daclizumab plus low-dose TAC versus standard TAC at month 12:• Patient survival: 86.6% versus 92.9% (P = 0.21).• Acute rejection: 23.2% versus 27.7% (P = 0.68).• CG GFR at end of week 1: 110.7 versus 89.6 mL/min (P = 0.019).• MDRD GFR at end of month 1: 86.8 versus 70.1 mL/min/1.73 m2 (P < 0.001).• MDRD GFR at month 6: 75.4 versus 69.5 mL/min/1.73 m2 (P = 0.038).• MDRD GFR at month 12: no significant differences.Follow-up was short (12 months). TAC target C0 was the same for both groups after day 30; this possibly explains the few significant differences between the groups.
Calmus et al.[25]1992-year OL study (6-month RCT plus 18-month follow-up) of daclizumab induction plus delayed standard-dose TAC versus standard-dose TAC• TAC was started on day 5 at 0.075 mg/kg bid to C0 of 10-20 ng/mL for 4 weeks and 5-15 ng/mL thereafter, and daclizumab was used at 2.0 mg/kg at 12 hours and at 1.0 mg/kg between days 7 and 10.• TAC was started on day 0 at 0.075 mg/kg bid to C0 of 10-20 ng/mL for 4 weeks and 5-15 ng/mL thereafter. Both groups received MMF at 1 g bid from day 0 to day 14 and then at 0.5 g bid to day 56 and CSs at 15-20 mg/day for 1 month, then at 10-15 mg/day for month 2, and at 5-10 mg/day thereafter.For delayed TAC versus standard TAC at 6 months:• Serum Cr > 130 μmol/L: 22.4% versus 29.7% (P = NS).• BPAR: 17.5% versus 18.75%.• Dialysis required: 1 versus 6 patients.• Patient survival: 95.9% versus 95.9% (at month 24: 85.4% versus 87.3%).• Graft survival: 100% versus 99% (at month 24: 98.9% versus 96.5%).Standard-dose TAC was used in both groups (the treatment arms were identical after 7-10 days). There was good baseline renal function (serum Cr <180 μmol/L).
Otero et al.[26]1571-year OL RCT of daclizumab induction, MMF, and reduced-dose TAC (modified therapy) versus CSs plus standard TAC (controls)• Daclizumab was used at 2 mg/kg IV ≤6 hours after transplantation and at 1 mg/kg IV on day 7, MMF was used at 1 g bid ≤12 hours after transplantation, and TAC was used at 0.05 mg/kg bid to C0 of 5-15 ng/mL and then 4-8 ng/mL after week 4.• TAC was used at 0.05 mg/kg bid to C0 of 5-15 ng/mL; methylprednisolone was used at 15-20 mg/day to end of week 1 and was tapered from week 12 to week 24.For modified therapy versus controls at 24 weeks:• BPAR: 11.5% versus 26.6% (P = 0.017).• Shorter time to rejection with standard therapy (P = 0.044).• No significant differences in monthly mean serum Cr between treatments.• Study completion: 71.8% versus 70.9%.• Patient survival: 92.3% versus 94.9%.• Graft survival: 89.7% versus 88.6%.• Renal impairment AEs: 19.2% versus 27.8%.• Hypertension AEs: 20.5% versus 26.6%.• Thrombocytopenia: 12.8% versus 15.2%.• New-onset diabetes: 9.0% versus 13.9%.Follow-up was short (12 months).
Nashan et al.[27]601-year OL RCT of reduced-dose TAC versus standard-dose TAC (both with MMF and CSs)• TAC was used to C0 of 5-8 ng/mL.• TAC was used to C0 of 10-15 ng/mL. Both groups received MMF at 1 g bid and CSs according to local practice throughout the study.For reduced TAC versus standard TAC at week 26:• BPAR: 25.9% versus 17.2%.• Patient survival: 92.6% versus 96.6%.• Median estimated CrCl similar to screening values (78.6 versus 66.3 mL/min).• Serious AEs: 55.2% versus 44.8%.This was a small, preliminary study. The median estimated CrCl was higher at screening in the reduced TAC group versus the standard TAC group (78.2 versus 69.6 mL/min).

Early Use of Mammalian Target of Rapamycin Inhibitors

Two early EVR conversion studies have demonstrated the challenges in evaluating the effects of an mTOR inhibitor on renal function. In the Preservation of Renal Function in Liver Transplant Recipients With Certican Therapy (PROTECT) study [a 1-year open-label (OL) randomized controlled trial (RCT)], Fischer et al.[17] evaluated the renal protective effects of EVR started 30 days after transplantation as part of an early CNI discontinuation (d/c) regimen in liver transplant recipients with normal renal function or mild renal dysfunction [calculated glomerular filtration rate (cGFR) > 50 mL/min]. Patients were randomized 4 weeks after liver transplantation to either a continuation of CNIs (n = 102) or a conversion to EVR (n = 101). Basiliximab induction was used with or without corticosteroids (CSs). Eleven months after randomization, a significant improvement in GFR was demonstrated with the Modification of Diet in Renal Disease (MDRD) formula [improvement with EVR versus CNIs = 7.8 mL/min, 95% confidence interval (CI) = −14.36 to −1.19 mL/min, P = 0.02], but the primary study hypothesis that EVR was superior to CNIs [defined as a mean difference between treatments ≥ 8 mL/min in the (CG) Cockroft-Gault GFR at month 11] was not met. The mean difference with the CG formula was 2.9 mL/min (95% CI = −10.7 to 4.8 mL/min, P = 0.46). Rates of mortality, biopsy-proven acute rejection (BPAR), and efficacy failure (a composite of BPAR, graft loss, death, and loss to follow-up) were similar for both treatment groups, but infections, leukocytopenia, hyperlipidemia, and treatment d/c were more frequent in patients receiving EVR versus those receiving CNIs.

In a second study conducted by Masetti et al.,[19] EVR monotherapy after the early withdrawal of a CNI [cyclosporine A (CSA)] in de novo liver transplant recipients was associated with an improvement in renal function and no effect on the incidence of rejection or the development of major complications. During this randomized OL study, patients with an estimated glomerular filtration rate (eGFR) ≥ 29 mL/min/1.73 m2 at randomization were treated for 10 days with CSA, then were randomized to receive EVR plus CSA until day 30, and finally continued on either EVR monotherapy (n = 52) or CSA with or without MMF (n = 26). According to the MDRD formula, GFR at 12 months was significantly higher for patients receiving EVR monotherapy versus patients in the CSA group (87.6 versus 59.9 mL/min, P < 0.001); however, it should be noted that the mean MDRD value at randomization was higher for the EVR group versus the CSA group (81.7 versus 74.7 mL/min/1.73 m2, P = 0.30). At 12 months, the proportion of patients with stage 3 or higher CKD (GFR < 60 mL/min) was greater in the CSA group versus the EVR group (52.2% versus 15.4%, P = 0.005). Freedom from efficacy failure and patient survival rates for the 2 groups were similar.

First, it should be pointed out that both of these early EVR conversion studies reported only short-term (12-month) follow-up. Second, in both studies, large percentages of the patients were not randomized because of complications (46% for Fischer et al.[17] and 30% for Masetti et al.[19]). Third, a limitation of the PROTECT study[17] is that patients who were randomized to receive EVR received a median of 72 days of CNI treatment after the baseline assessment, and this means that many patients in this group were receiving CNI treatment at some level for up to 14 to 16 weeks after transplantation; this may have skewed the study results in favor of no difference between treatments and may explain the different outcomes for the PROTECT study versus Masetti et al.'s trial. Trials using EVR in the early posttransplant period will be hampered by exclusions. After a thorough review of the literature on early conversion trials, it is evident that conclusions of long-term benefits cannot be made at this point.

The safety, tolerability, and efficacy of EVR were investigated in a 2-phase study[20] (n = 119) in which 3 EVR dosage groups (1, 2, and 4 mg/day) of de novo liver transplant patients were compared with a placebo group. The first phase consisted of a 12-month multicenter, randomized, double-blind comparison of EVR doses with a placebo in patients concomitantly receiving CSA and CSs. For the second phase, which was an OL extension, patients originally assigned to receive EVR continued this therapy for up to 36 months. Patients receiving the placebo were discontinued at 12 months. Rates of AEs were higher in the EVR groups, especially among those patients who received 4 mg/day. There was no difference, however, between all groups in the incidence of thrombocytopenia or leukopenia. Among the EVR-treated patients, renal function, as determined by the serum creatinine (Cr) concentration and creatinine clearance (CrCl), remained stable from month 1 to month 36. Notably, this study was not powered to evaluate renal function, and this endpoint was not a primary or secondary aim of the trial. Moreover, this study used full-dose CSA and fixed-dose EVR, and dropout rates were high (65% in the 4-mg/day group); this further prohibits any meaningful conclusions about the effects of EVR on renal function.

A more recent and larger scale multicenter study by De Simone et al.[21] evaluated the efficacy and safety of using EVR to eliminate or reduce tacrolimus (TAC) in 719 de novo liver transplant recipients. Patients were randomized 30 days after liver transplantation to receive EVR plus reduced TAC, EVR plus TAC withdrawal (by the end of month 4), or standard-exposure TAC. Enrollment in the TAC withdrawal arm was stopped early because of a higher incidence of acute rejection. The EVR/reduced TAC arm was found to be noninferior to the standard-exposure TAC arm with respect to the primary efficacy endpoint (treated BPAR, graft loss, or death 12 months after transplantation) with a lower incidence of treated BPAR (2.9% versus 7.0%, P = 0.04). Twelve months after transplantation, the difference in eGFR favored the EVR/reduced TAC group by 8.5 mL/min/1.73 m2 (P < 0.001). No unexpected safety findings were reported. AEs that were more common in the EVR/reduced TAC arm versus the standard-exposure TAC arm included peripheral edema, leukopenia, and serious infections. These results suggest that the use of mTOR inhibitors may facilitate early TAC reduction and allow for better renal function without compromising efficacy. Further study is warranted to confirm that the long-term renal benefits associated with EVR in CNI-sparing regimens outweigh any additional AEs.

Tacrolimus Minimization Studies With and Without Mycophenolate Mofetil

In a multicenter, randomized study conducted by Boudjema et al.,[22] standard-dose TAC (n = 100) was compared with reduced-dose TAC plus MMF (n = 95) to determine whether reducing the TAC dose would decrease renal dysfunction during the induction and maintenance periods. Antibody therapy was not used, and patients in the standard-dose group did not receive MMF. Rejection rates were significantly higher in the standard-dose TAC group versus the group that received reduced-dose TAC and MMF (46% versus 30%, P = 0.02). Fewer patients in the reduced-dose TAC group developed arterial hypertension, although the difference did not reach statistical significance, and there were no between-group differences in the occurrence of diabetes mellitus. At 12 months, GFR was significantly higher in the reduced-dose TAC group (90 ± 30 mL/min) versus the standard-dose group (78 ± 26 mL/min, P = 0.02). The results reported in this study show an unacceptably high rejection rate in the standard-dose TAC group, which did not include MMF in its regimen.

In a second reduced-dose TAC trial (the ReSpECT study),[23] liver transplant patients were randomly assigned to group A, which received standard TAC and steroids (n = 183); group B, which received MMF, reduced-dose TAC, and steroids (n = 170); or group C, which received daclizumab, MMF, and TAC (which was delayed until the fifth day after transplantation; n = 172). At week 52, decreases in eGFR were significantly smaller in the group receiving TAC after a delayed introduction (group C; −13.63 mL/min) versus the group that received standard-dose TAC and steroids without MMF (group A; −23.61 mL/min, P = 0.01), although there was no difference in the changes in eGFR between groups A and B (−23.61 versus −21.22 mL/min).

The problem with the aforementioned studies is the avoidance of MMF in the control arms; this is an important point because the use of MMF is currently standard in liver transplantation (although the use of MMF in combination with TAC is not Food and Drug Administration–approved). That is, the comparator arms in both trials (TAC and prednisone without MMF) included an immunosuppressive regimen that is rarely used. Therefore, the results of TAC minimization studies with MMF in treatment arms may better reflect the actual clinical practice setting. In addition, both studies included patients whose renal function was much better than that found in most current liver transplant recipients. For example, the mean GFR in each trial was approximately 100 mL/min, and the average Model for End-Stage Liver Disease (MELD) score was approximately 15. The study protocols excluded individuals with serum Cr levels > 1.5 mg/dL, diabetes, and hypertension as well as all autoimmune patients (ie, ∼60% of typical contemporaneous liver recipients). Such patients are much more likely to tolerate the high TAC levels required in MMF-free control arms and to tolerate the expected drop in GFR of one-third or more in comparison with the average liver recipient, whose GFR is 60 mL/min.[28]

Other studies have found no differences in renal outcomes with reduced TAC exposure. In a multicenter, randomized trial similar to the ReSpECT trial (except that the control arm used MMF), Yoshida et al.[24] evaluated renal function in liver transplant recipients with 2 treatment arms: (1) daclizumab, MMF, steroids, and delayed, low-dose TAC (investigational arm; n = 72) and (2) standard-dose TAC, MMF, and steroids (control arm; n = 76). Although there were statistically significant improvements in eGFR at <6 months in the investigational arm, there was no significant difference at 12 months. In addition, there were no significant differences in BPAR rates at 12 months between the investigational and control arms (23.2% versus 27.7%, P = 0.68).

A multicenter, OL, randomized trial evaluated renal function at 6 months in liver transplant recipients who received either daclizumab and delayed TAC (n = 98) or standard-initiation TAC (n = 101).[25] Both regimens included MMF and steroids. At month 6, there were no significant differences in renal function (incidence of serum Cr levels > 130 μmol/L) between the delayed TAC and standard-initiation TAC groups (22.4% versus 29.7%). In addition, no differences in renal function were observed after 12 or 24 months of follow-up.

A third multicenter, OL, randomized study in liver transplant recipients (n = 157) compared a regimen of TAC (trough level = 5-15 ng/mL) and steroids with a regimen of daclizumab induction, MMF, and reduced-dose TAC [trough level = 5-15 ng/mL (adjusted to 4-8 ng/mL after 4 weeks)].[26] At 6 months, the incidence of AE reports of renal impairment was lower in the reduced-dose TAC group (19.2% versus 27.8%), but the difference was not statistically significant. Mean serum Cr levels were not significantly changed from the baseline in either treatment group. Fewer patients experienced BPAR in the reduced-dose TAC group (11.5% versus 26.6%, P = 0.02), but graft and patient survival rates were similar between the treatment arms.

In a smaller preliminary study of liver transplant recipients (n = 60) that compared reduced-dose TAC to standard-dose TAC (both with MMF and steroids),[27] no meaningful differences in the changes from the baseline in the median estimated CrCl were observed between the treatments for the duration of the study. The primary endpoint of BPAR requiring treatment or graft loss was reported for 25.9% of the patients in the reduced-dose TAC arm and for 17.2% in the standard-dose TAC arm. The authors concluded that although the between–treatment group safety and efficacy results appeared similar, more definitive studies were needed because of the exploratory nature of the trial.


In this section, results are reported from studies evaluating immunosuppression regimens using CNI avoidance during the maintenance phase (>6 months after transplantation). It should be noted that there is some overlap between the maintenance studies discussed and induction studies that were carried out to 12 months. In general, the effects of CNI avoidance in the maintenance phase are mixed. Results of clinical trials evaluating CNI avoidance in the maintenance phase are summarized in Table 2.

Table 2. Summary of Studies Evaluating CNI Avoidance in the Maintenance Period
StudynStudy DesignImmunosuppressive TreatmentKey ResultsLimitations
Pageaux et al.[29]5624-month OL RCT of patients who underwent liver transplantation >1 year earlier and developed CNI-related renal dysfunction: comparison of MMF plus reduced CNI versus no MMF• MMF was used at 500 mg bid up to 2000-3000 mg/day; at full MMF dose, patients were tapered to ≥50% of their initial CNI dose.• Previous CNI therapy was continued with option of reducing dose to 75% of previous CNI dose.For MMF plus reduced CNI versus standard CNI at 12 months:• Cr: 143.4 versus 181.6 μmol/L (P = 0.001).• CrCl: 51.7 versus 44.8 mL/min (P = 0.04).• Patient survival: 96% versus 100%.• Graft survival: 96% versus 100%.This was a small study. The study included a large proportion of patients with alcoholic liver disease, who are less prone to rejection than other populations.
Schmeding et al.[30]1425-year RCT of conversion from CNI to MMF in patients who underwent liver transplantation 6 months to 7 years before studyAll patients initially received CNI plus CSs (tapered over 3 months) and basiliximab induction (20 mg bid on days 0 and 4):• CNI was withdrawn over 3 months with introduction of MMF over 4 weeks (final dose = 2000 mg/day).• CNI-based regimen was continued.For MMF conversion versus CNI continuation at 5 years:• Patient survival: 90% versus 94%.• BPAR: 11.1% versus 2.9% (P = 0.055).• Defined renal complication: 0 versus 9 patients (P = 0.018).• Renal function improvement: 51.4% versus 3.5% (P < 0.01).• New-onset/worsening diabetes: 22.7% versus 52.6% (P = 0.054).More than 20% of the patients in both groups were on a regimen including a CNI and MMF at 5 years (up to 40% of the patients were unable to maintain their assigned study group).
Schlitt et al.[31]286-month RCT of conversion to MMF versus continuation of CNI in patients with suspected CNI toxicity after transplantation• MMF was used at 250 mg bid in week 1, at 500 mg bid in week 2, at 750 mg bid in week 3, and at 1000 mg bid thereafter; from week 5 onward, CNI was reduced by 25% each week and withdrawn after 2 months.• Previous low-level CNI treatment was continued.For MMF conversion versus controls at 6 months:• Mean decrease in serum Cr: −44.4 versus −3.1 μmol/L (mean difference = 41.3 μmol/L, 95% CI = 12.4-70.2 μmol/L).• Mean change in systolic BP from baseline: −11.2 versus −0.4 mm Hg.• Mean change in diastolic BP from baseline: −5.8 versus −0.8 mm Hg.• Mean change in serum uric acid from baseline: −80.6 versus 2.5 μmol/L.• Reversible acute graft rejection: 3 versus 0 patients.This was a small, unblinded study.Follow-up was short (6 months).
Orlando et al.[32]4224-month study of conversion from CNI to MMF in patients who underwent liver transplantation at a mean of 70.5 months before study• MMF was used at 250 mg bid; this was increased by 500 mg weekly to full daily dose of 1.5 g.• TAC or CSA was tapered by 25% of initial dose every month until a complete withdrawal.At 12 months:• Mean decrease in serum Cr from 1.8 mg/dL at baseline to 1.6 mg/dL (P < 0.05).• Mean increase in GFR from 48 to 58 mL/min (P < 0.05). At 24 months:• Mean decrease in serum Cr to 1.4 mg/dL.• Mean increase in GFR to 63.2 mL/min. No deaths or major toxicity requiring MMF d/c occurred. One patient required CNI reinstitution for acute rejection.This was a small, uncontrolled study.
Créput et al.[33]493-year study of conversion from CNI to MMF in patients who underwent liver transplantation at least 12 months before study• MMF was used at 500 mg bid; this was increased weekly to optimal dose of 1 g bid at 1 month after conversion.• After optimal dose of MMF was reached, TAC or CSA was tapered by 10%-20% stepwise every 4 weeks and withdrawn, if possible, within 6 months.• There were mean increases in CrCl from 43 mL/min at baseline to 49 mL/min at 1 year, to 50 mL/min at 2 years, and to 58 mL/min at 3 years (P < 0.0001).• No patient developed acute or chronic graft rejection after CNI reduction or interruption.This was a small, uncontrolled study.
Liver Spare the Nephron[34]2932-year OL RCT of MMF plus SRL versus MMF plus CNI in patients randomized 30-180 days after transplantation• MMF was used at 1-1.5 g bid plus SRL at 2-4 mg/day to C0 of 5-10 ng/mL with d/c of the CNI.• Current regimen of TAC or CSA plus MMF at 1-1.5 g bid was continued.For MMF plus SRL versus MMF plus CNI:• Mean increase in cGFR from baseline (month 12): 19.7% versus 1.2% (P = 0.0012).• BPAR, graft loss, death, or loss to follow-up: 16.4% versus 15.4%.• BPAR (month 24): 12.2% versus 4.1% (P = 0.0159).• Graft loss (month 24): 5.4% versus 9.7%.• Treatment failure (first 12 months): 48.6% versus 37.9% (P = 0.0845).• Hyperlipidemia AEs: 70.3% versus 50.0% (P = 0.0004).• New-onset diabetes: 14.2% versus 26.7% (P = 0.0084).Full study results have not been published.
Abdelmalek et al.[35]60712-month OL RCT of conversion from CNI to SRL in patients 6-144 months after transplantationPatients were randomized 2:1:• An SRL loading dose (10-15 mg), then 3-5 mg/day on days 2-6, 6-16 ng/mL, and subsequently 8-16 ng/mL were used.• Patients were continued on CSA (C0 50-250 ng/mL) or TAC (C0 3-10 ng/mL).MMF or AZA was allowed in both groups.For SRL versus CNI at month 12:• Graft survival: 93.4% versus 94.4%.• Deaths: 3.3% versus 1.4%.• BPAR: 11.7% versus 6.1% (P = 0.02).• Treatment failure: 48.3% versus 26.7% (P < 0.001).• Adjusted mean change in CG GFR from baseline: −4.5 versus −3.0 mL/min (P = 0.34).There were high d/c rates, enrollment difficulties, protocol amendments, less loss of renal function in the CNI arm than expected on the basis of historical data, and a wide range of times since transplantation.
De Simone et al.[36]14512-month study (6-month RCT plus 6-month follow-up) of EVR plus CNI reduction or d/c versus continuation of standard-dose CNI in patients with CNI-related renal impairment 12-60 months after transplantation• EVR was used at 1.5 mg bid on day 1; after week 2, it was adjusted to maintain C0 of 3-8 ng/mL (with concomitant CNI) and 6-12 ng/mL (when CNI was eliminated). CNI was reduced by 50% on day 1 and by 50% increments at subsequent visits until d/c. MPA and AZA were discontinued on day 1; CSs were unchanged.• There was a continuation of baseline CSA or TAC regimen ± MPA, AZA, or CSs; CNI dose could be reduced by ≤25%.For EVR conversion versus standard CNI at 6 months:• Mean change in CrCl from baseline: 1.0 versus 2.3 mL/min (P < 0.46).• BPAR: 1.4% versus 1.4%.• Deaths: 1.4% versus 0%.• No graft losses in either group.• Any AE: 95.8% versus 69.9% (P < 0.001).• Hypercholesterolemia: 13.9% versus 2.7% (P = 0.017).• Mouth ulceration: 26.4% versus 0% (P < 0.01).• AEs of increased hepatitis C virus loads, dry skin, eczema, and rash: all 6.9% versus 0% (all P = 0.028).77% of the patients in the control group reduced their CNI dose.The mean time after transplantation was long (>3 years).Follow-up was short (12 months).

A prospective, multicenter, randomized study was conducted by Pageaux et al.[29] to compare a reduced-dose CNI (≥50% of the initial dose) and MMF (n = 27) with CNI monotherapy (n = 29) in liver transplant recipients. Although the investigators reported no difference in rejection rates, the mean CrCl increased from 42.6 ± 10.9 to 51.7 ± 13.8 mL/min over 12 months with the reduced-dose CNI plus MMF but remained relatively unchanged with CNI monotherapy (changing from 42.8 ± 12.8 mL/min at the baseline to 44.8 ± 19.7 mL/min at month 12).

Similar reports by Schmeding et al.[30] and Schlitt et al.,[31] who compared converting to MMF to remaining on a CNI, demonstrated lower GFRs in one study but improved Cr concentrations in another. In the large-scale, prospective, randomized study conducted by Schmeding et al., 70 liver transplant recipients remained on standard CNI therapy, and 72 were switched to MMF monotherapy with a follow-up of 5 years. There was a trend toward a higher rate of rejection in the MMF group versus the CNI group (11.1% versus 2.9%, P = 0.06). Liver transplant recipients in the MMF group had lower mean GFR values throughout the study; however, no statistically significant between-group differences were observed, and patients in the MMF arm began the study with a lower mean GFR (59.2 versus 70.3 mL/min, P = 0.04). In contrast, Schlitt et al. reported that the replacement of a CNI with MMF in liver transplant recipients with renal dysfunction was associated with an improvement in renal function as measured by mean serum Cr levels after 6 months of the study drug (mean difference = 41.3 μmol/L, 95% CI = 12.4-70.2 μmol/L). This strategy, however, resulted in an increased rejection risk 3 to 4 months after study entry.

Two additional single-center, uncontrolled trials were conducted by Orlando et al.[32] and Créput et al.[33] Orlando et al. reported the successful conversion from a CNI (TAC, n = 5; CSA, n = 37) to MMF monotherapy. They weaned 41 of 42 patients from a CNI to MMF (≤1.5 g/day). Renal function improved in 89% of the patients from the baseline mean serum Cr and GFR values of 1.8 mg/dL and 48 mL/min to 1.6 mg/mL and 58 mL/min at 12 months, respectively (P < 0.05). No deaths or major toxicity requiring MMF d/c occurred, and only 1 patient required reinstitution of a CNI for acute rejection. Créput et al. converted 49 liver recipients on a CNI (TAC, n = 14; CSA, n = 35) with renal dysfunction to MMF to either reduce or withdraw the CNI. After the introduction of MMF, CrCl increased from 43 to 49 mL/min after 1 year, 50 mL/min after 2 years, and 58 mL/min after 3 years (P < 0.0001). There were no instances of rejection after the initiation of MMF.

Use of Mammalian Target of Rapamycin Inhibitors in the Maintenance Phase

The Liver Spare the Nephron study[34] (n = 293) showed that the combination of MMF (2-3 g/day) and SRL (2-4 mg/day) provided significantly greater improvements in renal function in comparison with the combination of MMF (2-3 g/day) and a CNI (TAC or CSA) 12 months after liver transplantation. Specifically, the increases in GFR from the baseline at that time were 19.7% and 1.2%, respectively (P = 0.0012), and this benefit was maintained at 24 months (13.5% versus −8.6%, P = 0.0053). The proportions of patients with biopsy-proven rejection, graft loss, death, or loss to follow-up at 12 months were similar for the treatment groups (16.4% for MMF/SRL and 15.4% for MMF/CNI, P = 0.9417).[37]

In a large-scale (n = 607), prospective, OL, randomized study of SRL, which was intended as a registration trial for SRL conversion, investigators evaluated liver transplant patients 6 months to 12 years after transplantation who were stable on their regimen with cGFRs between 40 and 90 mL/min.[35] After maintenance immunosuppression with a CNI since early transplantation, patients were assigned randomly to an abrupt conversion from the CNI to SRL (<24 hours; n = 393) or CNI continuation for up to 6 years (n = 214). The coprimary endpoints were GFR and patient and graft survival. The results of this study were negative (it did not achieve the primary endpoints) and led to a black-box warning by the Food and Drug Administration related to a higher rate of deaths (although not statistically significant) in the SRL arm and higher rates of AEs.

The patients in this study were older with a mean age of 55 years and a mean baseline GFR of 65 mL/min, and the average time for transplantation was relatively long at 4 years, with a substantial proportion of patients (approximately 50%) more than 3 years after transplantation.[35] The loading dose of SRL was high (10-15 mg), and patients received relatively high doses (up to 5 mg/day) for the first week of the study. At 12 months, there was no difference in the changes in GFR between the 2 groups (−4.5 versus −3.0 mL/min), and the graft loss/death rates were similar at 6.6% and 5.6% for the SRL and CNI arms, respectively.

In addition, the incidence of rejection was significantly higher in the SRL group versus the control group (11.7% versus 6.1%, P < 0.05).[35] SRL was also associated with significantly higher rates (all P < 0.05) of many AEs at month 12, including herpetic infections (8% versus 0.5%), stomatitis (24% versus 1%), mouth ulceration (10% versus 0.5%), hyperlipidemia (34% versus 7%), and cytopenias [anemia (20% versus 3%), leukopenia (13% versus 4%), and thrombocytopenia (13% versus 2%)]. In summary, SRL conversion in the maintenance phase of immunosuppression showed no demonstrable benefit with no differences in GFR, a trend toward a higher death rate, and significantly more AEs.

The conduct of this study, however, was problematic and should lead one to question the conclusions.[38] Although the study did indeed have the scientific and regulatory rigor associated with an RCT registration trial, flaws in the methodology of the study may have compromised the validity of its safety findings. In particular, the dose of SRL was up to 5 times the standard dose, and this may have increased the rate of AEs. The most common infection (herpetic) reported in the study was not documented by culture, and more importantly, 50% of the deaths were recorded as such with “missing data counted as event.”

Although the number of deaths in the SRL arm (n = 13 or 3.3%) was not significantly higher than the number in the CNI group (n = 3 or 1.4%),[35] this is cited in the SRL black-box warning. The full SRL black-box warning indicates that an increased susceptibility to infection and the possible development of lymphoma and other malignancies may result from immunosuppression [the reported rates of lymphoma/lymphoproliferative disease are 0.7% to 3.2% with SRL and 0.6% to 0.8% with azathioprine (AZA) and placebo controls].[39] When it is used in combination with CNIs, the risks of excess mortality and graft loss (22% with SRL plus TAC versus 9% with TAC alone) and hepatic artery thrombosis (7% with SRL plus a CNI versus 2% with controls) are elevated in liver transplantation.[39] Cases of bronchial anastomotic dehiscence have been reported in SRL-treated lung transplant recipients, and most have been fatal.[39]

A recent meta-analysis of 11 studies with SRL (3 RCTs and 8 observational studies) examined its efficacy in renal preservation and found a nonsignificant improvement of 3.38 mL/min (95% CI = −2.93 to 9.69) associated with the use of SRL.[40] Although SRL was not significantly associated with death [relative risk (RR) = 1.12, 95% CI = 0.66-1.88] or graft failure (RR = 0.80, 95% CI = 0.45-1.41) at 1 year, SRL was associated with statistically significant risks of infection (RR = 2.47, 95% CI = 1.14-5.36), rash (RR = 7.57, 95% CI = 1.75-32.70), ulcers (RR = 7.44, 95% CI = 2.03-27.28), and d/c of therapy (RR = 3.61, 95% CI = 1.32-9.89).

In a 6-month multicenter, randomized trial in liver transplant recipients receiving mTOR inhibitors in the maintenance phase, De Simone et al.[36] evaluated whether renal function (CNI-related renal impairment) would improve with a reduced CNI dose or CNI d/c. Patients began EVR with CNI reduction or elimination (n = 72) or continued to receive standard-exposure CNI therapy (n = 73). The primary study endpoint, a between-group difference in the CrCl change of 8 mL/min from the baseline to month 6, was not achieved in this study. These results may be attributed in part to the baseline CNI exposure, which was low for both groups and likely a reflection of previous reductions by physicians to improve renal function.

Rationale for Inconsistent Results From Calcineurin Inhibitor Avoidance Studies

If posttransplant end-stage renal disease is so prevalent and CNIs are an important contributing factor, then why are the results of studies of CNI avoidance studies mixed? Why have numerous studies failed to demonstrate a benefit in CNI avoidance or withdrawal? One reason is that many transplant recipients have fixed chronic renal dysfunction at the time of surgery. Therefore, they may have a smaller margin for improvement in comparison with those patients with a higher GFR. There is evidence to support this claim in that some of the studies with positive findings for CNI avoidance were selected to include patients without chronic renal disease. For example, reports by Neuberger et al.[23] and Boudjema et al.,[22] which were positive for CNI avoidance, had a mean transplant GFR of 100 mL/min and excluded patients with chronic renal dysfunction, whereas the patients in the negative studies of Nashan et al.[27] and Yoshida et al.[24] had a mean GFR that was 25% lower at 75 mL/min.

Another possibility is that the metrics used to evaluate the efficacy of CNI avoidance may be flawed. In each of these studies, the risk of CNI avoidance (rejection) was compared to the benefit to renal function (change in GFR). Immediately after CNI withdrawal, the immediate risk of rejection is approximately 10%.[35] The renal benefits of CNI withdrawal, however, may not be evident for many years. Because almost all of the randomized trials have evaluated outcomes at relatively short intervals (6-24 months), these studies are more likely to reveal side effects than improvements in renal function, which may take far longer. In addition, the risk of CNI withdrawal (ie, rejection) may be overstated. That is, a bout of rejection, especially if it is discovered quickly, may have no long-term ill effects, whereas the risk of chronic renal failure is 28% and may be even higher in patients selected for conversion. Therefore, equating rejection with a change in GFR may not be a useful comparison. Another important consideration is that there may be diminishingly less improvement in renal function with less TAC exposure. For example, the average TAC level in liver transplantation recipients 15 years ago was 11 ng/mL (1 year after surgery).[41] In comparison, a more recent survey by the Immune Tolerance Network found that the average TAC level at 1 year was only approximately 6 ng/mL.[42] Most practitioners would agree that the nephrotoxic effects of TAC at 6 ng/mL are far less severe than the effects at 11 ng/mL. However, there may be less improvement in renal function when TAC is weaned to a level < 6 ng/mL. Another factor to consider is the difficulty in maintaining patients on a specific immunosuppressive regimen over time. Many patients must discontinue specific agents because of side effects and medication intolerance. In fact, d/c rates from an initial immunosuppressive regimen are as high as 50% even in carefully monitored registration trials, and these rates may be higher in clinical practice. As a result, the attribution of specific drugs to outcomes, which may take years to observe, is very difficult.

There are 2 other specific factors that might alter a patient's response to changes in CNI dosing. One of these is pharmacogenomic factors associated with CNI renal insufficiency. TAC is metabolized primarily in the liver by the cytochrome p4503A4 pathway. There have been numerous studies evaluating the effects of p4503A4 genetic polymorphisms on TAC clearance as well as renal dysfunction. One of these studies by Fukudo et al.[43] reported a significantly higher incidence of renal dysfunction 1 year after transplantation associated with the recipient's cytochrome P450 3A5 genotype (*1/*1 and *1/*3 versus *3/*3: 17% versus 46%, P < 0.05). Another pharmacogenetic alteration associated with CNI-induced renal dysfunction occurs through the cytochrome P450 C28 (CYP2C8) pathway, which counters the vasoconstrictive effect of TAC through a reduction of the formation of epoxyeicosatrienoic acids by CYP2C8. In 41 liver transplant recipients, carrying the CYP2C8*3 polymorphism was associated with a higher risk of developing renal dysfunction (odds ratio = 16.67, 95% CI = 2.8-99.6).[44] The adenosine triphosphate–binding cassette B1 (ABCB1) transporter actively transports substrates, including TAC, out of the cell. Therefore, variations in the ABCB1 expression rate may alter the plasma and/or intracellular TAC concentration. In liver transplant patients, Hebert et al.[45] reported that 50% of patients with the ABCB1 11/22 haplotype (2677G,3435C/2677T,3435T) experienced renal dysfunction, whereas 31% of patients with the ABCB1 11/11 haplotype (2677G,3435C/2677G,3435C) and 11.2% of patients with the ABCB1 22/22 haplotype (2677T,3435T/2677T,3435T) did. Theoretically, patients with more favorable pharmacogenomic genotypes may be identified as more likely to respond to reductions in CNI exposure.

Another important factor affecting a patient's response to changes in serum TAC levels is the variation in the tissue concentration of the drug. Capron et al.[46] compared serum and tissue levels of TAC in 146 liver recipients. They found that TAC tissue levels correlated with the liver histopathological Banff rejection score, whereas blood levels did not. This suggests that serum levels may not be the most accurate measure of the biological exposure to TAC. Therefore, the alteration of the TAC level in specific patients may not correspond to similar changes in tissue levels. In such circumstances, a lowering of serum TAC levels might not be associated with a corresponding reduction in renal exposure.


Renal dysfunction is a common problem both before and after liver transplantation. Over the last few years, studies investigating the clinical utility of reducing or eliminating CNIs in the induction period and maintenance phase of immunosuppression in liver transplant recipients have shown inconsistent findings. This presents a challenge to clinicians in decision making, especially when they are weighing the potent immunosuppressive effects of CNIs against the potential to preserve renal function by reducing or delaying CNIs and filling the void with the mycophenolates (now standard) and/or mTOR inhibitors. Future studies addressing factors such as flaws in methodology and the duration of follow-up may provide more consistent results on which to base conclusions regarding the benefit of renal-sparing agents in liver transplantation.