SEARCH

SEARCH BY CITATION

Keywords:

  • Black box;
  • graft failure;
  • hepatic artery thrombosis;
  • patient death;
  • portal vein thrombosis;
  • rapamycin;
  • renal failure

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Disclosure
  9. References
  10. Supporting Information

We studied whether the use of sirolimus with reduced-dose tacrolimus, as compared to standard-dose tacrolimus, after liver transplantation is safe, tolerated and efficacious. In an international multicenter, open-label, active-controlled randomized trial (2000–2003), adult primary liver transplant recipients (n = 222) were randomly assigned immediately after transplantation to conventional-dose tacrolimus (trough: 7–15 ng/mL) or sirolimus (loading dose: 15 mg, initial dose: 5 mg titrated to a trough of 4–11 ng/mL) and reduced-dose tacrolimus (trough: 3–7 ng/mL). The study was terminated after 21 months due to imbalance in adverse events. The 24-month cumulative incidence of graft loss (26.4% vs. 12.5%, p = 0.009) and patient death (20% vs. 8%, p = 0.010) was higher in subjects receiving sirolimus. A numerically higher rate of hepatic artery thrombosis/portal vein thrombosis was observed in the sirolimus arm (8% vs. 3%, p = 0.065). The incidence of sepsis was higher in the sirolimus arm (20.4% vs. 7.2%, p = 0.006). Rates of acute cellular rejection were similar between the two groups. Early use of sirolimus using a loading dose followed by maintenance doses and reduced-dose tacrolimus in de novo liver transplant recipients is associated with higher rates of graft loss, death and sepsis when compared to the use of conventional-dose tacrolimus alone.


Abbreviations
CI

confidence interval

CNI

calcineurin inhibitor

DSMB

Drug Safety and Monitoring Board

HAT

hepatic artery thrombosis

ITT

intent-to-treat

LTx

liver transplantation

MELD

Model for End-Stage Liver Disease

MMF

mycophenolate mofetil

PVT

portal vein thrombosis

Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Disclosure
  9. References
  10. Supporting Information

Over the past two decades, outcomes after liver transplantation (LTx) have dramatically improved [1]. An important contributing factor to these improved clinical outcomes is the enhanced experience with calcineurin inhibitor (CNI)-based immunosuppressive therapy. However, CNI-based immunosuppression is associated with increased morbidity and important adverse effects including renal insufficiency, hypertension, diabetes, neurotoxicity and electrolyte abnormalities [2-5]. In this context, several hypotheses have been proposed to reduce CNI-related morbidity while avoiding higher risks of acute rejection, graft failure or patient death. One such hypothesis involves reduction in early exposure to CNI by the use of adjunct immunosuppressants.

Sirolimus, an mTOR inhibitor, has been used as an alternative or adjunct to CNI-based immunosuppression in solid organ transplantation. Sirolimus is indicated for the prophylaxis of organ rejection in patients aged ≥13 years receiving renal transplants in combination with cyclosporine [6]. However, outcomes of sirolimus-based therapy after LTx are less clear. Analysis of data submitted to the Organ Procurement and Transplantation Network suggests that sirolimus use may be associated with worse patient and graft survival in selected subgroups [7]. However, recent randomized controlled trials have not found a difference in patient and graft survival among persons converted to sirolimus-based therapy after initial CNI-based therapy [8-10]. In liver transplant recipients undergoing conversion to sirolimus-based therapy at least 6 months after LTx, use of sirolimus was associated with an increased risk of adverse events including infection, acute rejection and discontinuation of therapy [8, 11]. However, other multicenter observational studies have reported more favorable outcomes [12-15].

Specifically, evidence of the role of sirolimus initiated immediately after liver transplant is still lacking despite its use in several centers [13, 16-19]. Herein, we present the results of an open-label, randomized, active controlled trial that evaluated the role of sirolimus in combination with reduced-dose tacrolimus as compared to standard-dose tacrolimus alone in the immediate posttransplant period. Results of this study formed the initial basis of a black box warning against the use of sirolimus immediately after LTx, driven by an imbalance in adverse events and a numerically higher rate of hepatic artery thrombosis/portal vein thrombosis (HAT/PVT). Although the data presented here have been fully and publicly disclosed before, this is the first complete publication of this study of the role of sirolimus in the immediate posttransplantation period. These data still remain relevant for several reasons: First, given that this was the only de novo study examining the role of sirolimus along with reduced-dose tacrolimus and led to a black box warning in the United States and special warning in Europe, it is unlikely that further definitive evidence regarding the role of sirolimus in liver transplant recipients would be possible as this scenario would be discouraged by regulatory agencies. Furthermore, subsequent single-center observational studies (not clinical trials) comparing de novo sirolimus initiation at time of LTx have not shown an increased risk of graft failure or HAT as compared to other strategies, which underscore the importance of making the data of this multicenter study available to clinicians [15, 20]. Second, given emerging literature on the role of another mTOR inhibitor, everolimus, it is important to gauge whether similarities exist in the risk and benefit profile for these two medications [21-27]. Above all, transparency in the full presentation of the results of the study are important given that it may impact our current approach to management of immunosuppression after LTx.

The hypothesis was that use of sirolimus with reduced-dose tacrolimus, as compared to standard-dose tacrolimus, would be associated with a significant decrease in efficacy failure, defined as acute rejection, graft loss or death at 6 months after LTx. This study was terminated early due to an imbalance in the number of adverse events in the sirolimus-containing treatment group.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Disclosure
  9. References
  10. Supporting Information

This was a phase II international multicenter, randomized, open-label, active controlled and parallel-group trial conducted at 31 centers across the United States, Australia, Canada and Europe. The study was conducted from August 2000 to April 2003. The objective of the study was to evaluate the safety, tolerability and efficacy of sirolimus-based therapy. Two regimens were compared: (1) orthotopic liver transplant recipients assigned to receive conventional-dose tacrolimus and corticosteroids (tacrolimus arm) and (2) those receiving sirolimus, plus a reduced dose of tacrolimus and corticosteroids (sirolimus arm) immediately after LTx. The primary objective of the study was to compare the two groups for the prevention of efficacy failure at 6 months (biopsy-confirmed acute rejection requiring treatment within 48 h, graft loss or death). At the time of study design, it was expected that with prevailing regimens, approximately 30–70% of liver transplant recipients would experience one or more episodes of acute rejection [28] as compared to rates of less than 10% expected with sirolimus-based therapy [29].

Subjects

The planned enrollment was 300 subjects (150 in each arm), and the anticipated study duration was 24 months. With this sample size, there was 80% power (two-sided significance α = 0.05) to detect a 50% reduction in the rate of the primary composite end point in the sirolimus arm when the rate was estimated to be 30% in the tacrolimus group and 15% in the group receiving study drug.

All adult subjects undergoing deceased donor orthotopic LTx were eligible including those with prior histories of viral hepatitis, primary hepatic malignancy meeting established Milan criteria pretransplant and who had been treated with locoregional therapy but had not received systemic chemotherapy, or those with a history of treated skin cancer [30]. Subjects undergoing multiple organ transplantation, retransplantation or with a history of nonhepatic malignancy or use of systemic chemotherapy within 5 years, or evidence of systemic infection were excluded.

Subjects were randomly assigned (1:1) to one of the two treatment groups posttransplant with treatment initiated within 48 h. A random allocation sequence was generated by a computerized randomization/enrollment system with blocks of four per center stratification. Participants, care providers and those assessing outcomes were not blinded to randomized treatment assignment.

Medications were administered via intravenous route or nasogastric tube until oral intake was feasible. In the conventional-dose tacrolimus arm, tacrolimus was started at 0.06–0.10 mg/kg/day to achieve a trough of 7–15 ng/mL within the first 3 months and 5–10 ng/mL thereafter. Subjects not receiving tacrolimus for >3 consecutive days were withdrawn from the study. In the sirolimus arm, tacrolimus was started at 0.03–0.05 mg/kg/day to achieve a trough of 3–7 ng/mL within the first 3 months and 3–5 ng/mL thereafter.

Sirolimus

Sirolimus was started within 48 h posttransplant. A sirolimus loading dose of 15 mg was recommended for day 1. Dosing was initiated at 5 mg/day on subsequent days and the dose was adjusted to achieve a trough of 4–11 ng/mL. A dose of sirolimus up to 15 mg/day was permissible until the first level was available (maximum of 5 days). Samples were sent to three centralized labs for US, Australian and European centers. The typical turnaround time was less than 3 days but could have been longer.

Subjects not receiving tacrolimus or sirolimus for >7 consecutive days were withdrawn from the study. Time points when trough levels of tacrolimus and sirolimus were measured are detailed in the Results section.

In both study treatment arms, corticosteroids were started intraoperatively, reduced to 20 mg/day of prednisone or equivalent by day 7 and to 5–10 mg/day by end of the third month after which the medication was discontinued over 4–6 weeks. A taper of the dose of corticosteroids was started no sooner than 30 days after treatment of any potential acute cellular rejection. Steroids were not tapered if subjects required antilymphocyte antibody treatment, and had ≥2 episodes of rejection, autoimmune hepatitis or autoimmune disease prior to LTx requiring corticosteroids post-LTx.

Safety analysis

Patients were monitored for development of adverse events or serious adverse events using standardized definitions and protocol. Besides standard safety measurements, drug levels were monitored at various intervals. Given concerns of a potential increased risk of vascular thrombosis based on previous studies, subjects were monitored for development of HAT by Doppler ultrasound before hospital discharge (or by week 2) and at 4 weeks after LTx [17]. Angiographic confirmation of thrombosis was not required. All serious adverse events were reported to a sponsor medical monitor within 24 h. An external Drug Safety and Monitoring Board (DSMB), consisting of two transplant surgeons and a statistician, reviewed cumulative data for graft loss and patient death, HAT and PVT. All blood work results, including tacrolimus trough levels, were determined by local laboratories. Sirolimus blood concentrations were analyzed at a central laboratory for all samples.

Early termination

Enrollment started in August 2000. In May 2001, the DSMB recommended withdrawal of subjects who had received sirolimus as study drug for less than 30 days given an imbalance in the number of cases of HAT. These cases were determined not to be definitely related to the study drug, and therefore subjects already on treatment were continued, but no new subjects were enrolled. Due to a further observed imbalance in rates of graft loss and death, a decision was made by the study sponsor in May 2002 to terminate the study and to convert all subjects in both arms to standard of care regimen. A follow-up visit was conducted after study termination.

Statistical analysis

The primary end point was prevention of efficacy failure, defined as a composite end point of biopsy-proven acute cellular rejection requiring treatment within 48 h, graft loss or patient death at 6 months after LTx in all randomized patients based on an intent-to-treat (ITT) analysis. Graft loss was defined as patient death, receipt of a retransplant or listing for retransplant. The secondary efficacy outcomes were patient and graft survival, and biopsy confirmed acute cellular rejection (assessed at 6, 12, 24 months of treatment), renal function (at 12 and 24 months), efficacy failure (at 3 months), rate of posttransplant diabetes mellitus, incidence of HAT and PVT (all at 6 months) and development of infection or malignancy. Due to early termination of the study, only the following secondary end points were analyzed: patient and graft survival, acute rejection-free survival, incidence of HAT/PVT and renal function. All analyses were performed on the ITT population encompassing all randomized patients. The safety analysis was limited to patients that received at least one dose of the study medication. Fisher-exact test was used to determine differences in incidence of HAT/PVT across the groups. Kaplan–Meier analysis was conducted to estimate the time to event with significance determined by the log-rank test. For renal function, covariance analysis was performed based on all available data with treatment and baseline values as covariates in the subjects who had a baseline and a valid assessment at the visit.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Disclosure
  9. References
  10. Supporting Information

The study period was from August 2000 to April 2003. Patients were enrolled through March of 2001 and followed up until study termination in May 2002. A follow-up visit was conducted after termination. Figure 1 shows the flow of subjects through the study. A total of 234 subjects were enrolled in the study. Twelve subjects did not receive a transplant or withdrew prior to randomization. Overall 222 subjects were enrolled and randomly assigned to either arm and constituted the ITT population. Three subjects did not receive any study medication (2 in the sirolimus arm and 1 in the tacrolimus arm), and therefore 219 patients who took at least one dose of study medication were analyzed for safety information; of these 111 subjects received standard-dose tacrolimus and 108 subjects received reduced-dose tacrolimus plus sirolimus.

image

Figure 1. Study flow. #Main reason was initiation of mycophenolate mofetil; discontinued intervention relates to patients discontinuation of study drug and is inclusive for all reasons that are listed below. **Seventy-nine of 90 due to early termination by the Drug Safety and Monitoring Board (DSMB) (<30 days treatment) and the termination of the study by sponsor.

Download figure to PowerPoint

The mean age was higher in patients receiving sirolimus (52 vs. 49 years, p = 0.02). Overall, the distribution of gender, race and indications for LTx were similar across the two groups (Table 1). Donor characteristics (age, race and cold ischemia time) were also similar. Hepatic tumor found in the liver explants was higher in the tacrolimus arm (8% vs. 1%, p = 0.02). Given that the study was conducted in the pre-Model for End-Stage Liver Disease (MELD) era, MELD score data were not available.

Table 1. Baseline demographics
 n (%)/mean (SD)p-Value
Standard-dose tacrolimus, n = 111Sirolimus and reduced-dose tacrolimus, n = 108
  • CMV, cytomegalovirus.

  • 1

    More than one reason possible.

  • 2

    Data available only in 69 (tacrolimus) and 87 (sirolimus and tacrolimus) subjects.

  • 3

    Data available only in 86 (tacrolimus) and 70 (sirolimus and tacrolimus) subjects.

Recipient
Age in years49 (1)52 (1)0.02
Women29 (26)36 (33)0.3
Caucasian95 (86)92 (85)0.8
Hispanic ethnicity11 (10)7 (7)0.46
Weight in kg80 (1.8)81(1.8)0.7
Reason for liver transplantation1
Alcoholic liver disease41 (37)38 (35)0.88
Hepatitis C36 (32)36 (33)1.0
Hepatitis B16 (14)14 (13)0.85
Primary sclerosing cholangitis14 (13)8 (7)0.26
Malignant neoplasm2 (2)6 (6)0.17
Hepatic tumor at explant9 (8)1(1)0.02
Age in years44 (16)44 (16)0.97
Bilirubin, mg/dL4.3226.6720.93
Serum creatinine, mg/dL1.1431.0530.43
Donor
Donor race and ethnicity0.24
Black4 (3.6)3 (2.8)
Asian4 (3.6)2 (1.9)
Caucasian77 (69)80 (74)
Hispanic ethnicity10 (9)4 (4)
Other, unknown or not disclosed16 (14.4)19 (17.6)
Cold ischemia time in hours8.5 (2.6)9.2 (2.8)0.15
Donor CMV IgG77 (69)62 (57)0.07

Patient and graft survival

Table 2 summarizes the main findings of the study. Subject survival was significantly lower (p = 0.010) at 24 months in the sirolimus arm (deaths, n = 22, 20%, median time to death: 126 days, range 5–623) as compared to the standard-dose tacrolimus arm (deaths, n = 9, 8%, median time to death: 213 days, range 0–530; Figure 2A, Table S1). The fraction of deaths within the first 6 months was 12/22 (55%) in the sirolimus arm and 4/9 (44%) in the tacrolimus arm. The most common cause of death was sepsis or multiorgan failure in the sirolimus arm (n = 8). In the tacrolimus arm, two cases of death were due to infection. Reason for death was not recorded for five deaths across the two groups.

Table 2. Cumulative event rates at 24 months of selected outcomes in the intent-to-treat population
 Standard-dose tacrolimus, n = 112 (%)Sirolimus and reduced-dose tacrolimus, n = 110 (%)p-Value1
  • HAT, hepatic artery thrombosis; PVT, portal vein thrombosis.

  • 1

    Based on log-rank tests.

  • 2

    Subjects may be included in both patient death and graft loss categories therefore sum of these two rows may be higher than row labeled death or graft loss.

Death or graft loss14 (12.5)29 (26.4)0.009
Patient death29 (8)22 (20)0.01
Graft loss26 (5.4)12 (10.9)n/a
Graft loss due to HAT/PVT1 (0.9)6 (5.5)n/a
Episodes of rejection34 (30.4)29 (26.4)0.62
HAT/PVT3 (2.7)9 (8.2)0.065
image

Figure 2. (A) Patient and (B) graft survival in patients randomized to standard-dose tacrolimus (TAC) or sirolimus and reduced-dose tacrolimus arm (SRL). Closed circle: Sirolimus and reduced-dose tacrolimus. Open circle: Standard-dose tacrolimus.

Download figure to PowerPoint

Graft survival was significantly lower at 24 months (p = 0.009) in the sirolimus arm as compared to the tacrolimus arm (Figure 2B, Table S1). The incidence of graft failure (death, receipt of a retransplant or listed for retransplant) was higher in the sirolimus arm (n = 29, 26.4%, median time to graft failure: 97 days, range 2–623) as compared to the standard-dose tacrolimus arm (n = 14, 12.5%, median time to graft failure: 120 days, range 0–632). A majority of the graft failure was in the first 6 months (20/29, 69% for sirolimus vs. 7/14, 50% for tacrolimus); among these, graft loss was attributable to HAT/PVT in 7 cases (6 in the sirolimus arm and 1 in the tacrolimus arm).

Acute rejection

There were 34 and 29 episodes of acute cellular rejection in the tacrolimus and sirolimus arms, respectively, with a majority of the episodes (30 and 26, respectively) within the first 6 months. There was no significant difference (p = 0.6) in acute rejection-free survival across the two groups (Figure S1).

Adverse events

Incidence of HAT/PVT

The incidence of HAT or PVT was higher, though not statistically significantly different, in the sirolimus arm as compared to the tacrolimus arm (n = 9, 8.3% vs. n = 3, 2.7%, p = 0.065; Figure 3 and Table 3). HAT/PVT occurred within 18 days of transplantation in 10 of the 12 cases. Cases of HAT were not clustered at a particular center.

image

Figure 3. Incidence of hepatic artery thrombosis/portal vein thrombosis in patients randomized to standard-dose tacrolimus (TAC) or sirolimus and reduced-dose tacrolimus arm (SRL). Closed circle: Sirolimus and reduced-dose tacrolimus. Open circle: Standard-dose tacrolimus.

Download figure to PowerPoint

Table 3. Incidence of hepatic artery and portal vein thrombosis
CaseAgeGenderStudy dayReasonDetails
  • HAT, hepatic artery thrombosis; PVT, portal vein thrombosis.

  • 1

    Requiring retransplantation.

  • 2

    Tacrolimus given at least 1 day before sirolimus.

A. Standard-dose tacrolimus arm
140M14PVT
2153M228HAT
349M699HATOccurred in the retransplanted organ after primary lost due to recurrent hepatitis C
B. Sirolimus and reduced-dose tacrolimus arm
156M1HATRight HAT, had required suture of right HA which was erroneously closed at time of LTx, no further treatment required
21,260M1PVTHad PVT in native PV at time of transplant
31,255F2HAT/PVTPVT diagnosed before receipt of study medication
41,247M6HAT
5159M7HAT
6158F7HAT
748M14PVT
8136M18HATReceived four liver transplants with at least two with HAT (1st and 3rd). Cause of allograft failure unknown for 2nd and 4th
952M59PVT

All nine cases in the sirolimus arm occurred within the first 3 months (median 7 days). Of these, six were the cause for retransplantation. In one case that developed both PVT and HAT, the PVT was present before receipt of sirolimus. In one case in the sirolimus arm, a right HAT was attributed to erroneous ligation of the right hepatic artery at the time of LTx.

In the tacrolimus arm, one of the cases occurred at 14 days (PVT); one case of HAT occurred on day 228, and one case of HAT occurred in a patient that withdrew from the study due to retransplantation (recurrent hepatitis C) and on day 699 developed HAT in the second transplanted liver.

Other adverse events

Adverse events were reported by 98.2% of subjects in the tacrolimus arm and 100% of subjects in the sirolimus arm (Table 4). The most common reported events (>50%) were abdominal pain, fever, diarrhea, headache, peripheral edema, thrombocytopenia, hyperglycemia, tremor, pleural effusion and anemia and were similar across the groups. As compared to the tacrolimus arm, the sirolimus arm had significantly higher rates (p ≤ 0.05) of bile leak (11.1% vs. 1.8%), coagulation disorder (11.1% vs. 3.6%), acidosis (13.9% vs. 2.7%), abnormal healing (21.3% vs. 9.9%), elevated cholesterol (19.4% vs. 7.2%), dyslipidemia (33.3% vs. 16.2%), hemoptysis (6.5 vs. 0.9%), albuminuria (6.5% vs. 0.9%), unspecified eye disorders (6.5% vs. 0.9%), depression (18.5% vs. 9.5%) and “kidney failure” (9.3% vs. 2.7%). Leukocytosis (19.8% vs. 9.3%) was higher in the tacrolimus arm. Rates of malignancy (after exclusion of hepatic malignancies found in the explant) were similar. The incidence of sepsis was higher in the sirolimus arm (20.4% vs. 7.2%, p = 0.006).

Table 4. Selected adverse events in the safety population (n = 219) with rates of at least 5% in either group
 Standard-dose tacrolimus (n = 111)Sirolimus and reduced-dose tacrolimus (n = 108)p-Value
 n%n%
  • 1

    Biliary, urinary, respiratory, line.

Overall events
Abdominal pain7063.16257.40.4
Headache6861.354500.1
Fever4338.75550.90.08
Arrhythmia21.887.40.06
Hypertension5448.64642.60.42
Bile leak21.81211.10.01
Anemia6457.77165.70.27
Acidosis32.71513.90.003
Abnormal healing119.92321.30.03
Elevated cholesterol87.22119.40.01
Dyslipidemia1816.23633.30.01
Peripheral edema6760.46459.30.89
Albuminuria10.976.50.03
Depression1092018.50.05
Hemoptysis10.976.50.03
Skin ulcer87.254.60.57
Infections
Any infection7668.58679.60.07
Sepsis187.22220.40.006
Cellulitis10.965.60.06
Cholangitis65.487.40.59
Pneumonia119.998.30.82
Herpes zoster21.821.91
Herpes simplex76.365.61
Malignancies
Any malignancy76
Skin carcinoma43.621.90.68
Recurrent hepatocellular carcinoma02

Discontinuations

Overall, 213 patients withdrew from the study (Figure 1). The most common reasons for withdrawal were early termination of the study and sponsor/DSMB request (n = 110). Among these, in the sirolimus arm, 20 patients discontinued following the DSMB decision to cease sirolimus administration in patients within 30 days of transplantation. Rates of discontinuation for adverse events were higher in the sirolimus arm (36% vs. 27%). Among the adverse events, discontinuations due to infection were higher in the sirolimus group (4.6% vs. 0). Discontinuations due to protocol violations were higher in the standard tacrolimus arm (10% vs. 2%) and mostly due to initiation of mycophenolate mofetil (MMF).

Renal function

The baseline mean creatinine clearance (Cockcroft–Gault) was 115.1 mL/min in the tacrolimus arm and 112.2 mL/min in the sirolimus arm, for patients that had data available at 1 month. The mean creatinine clearance was significantly lower than at pretransplant baseline in both treatment groups at all time points. The adjusted mean decrease in creatinine clearance from baseline within each group at 12 months was significant at the 0.001 level (−28.3 [SE:3.94] mL/min [n = 61] in the tacrolimus arm and −21.8 [SE:4.39] mL/min [n = 49] in the sirolimus arm). There was no significant difference in the adjusted mean changes (decrease from baseline) between the tacrolimus and sirolimus arm except for the first month where the adjusted mean decrease from baseline was greater in the tacrolimus arm than in the sirolimus arm (−28.9 vs. −17.2; p = 0.027; Table S2).

Drug levels

Initial mean tacrolimus doses were approximately 4.8–8 and 2.4–4 mg/day in the tacrolimus and sirolimus arms, respectively, based on targeted mg/kg dosing and mean body weights in each treatment group of 80.5 and 81.2 kg, respectively. The mean tacrolimus dose was 10.6 mg/day (standard deviation [SD] = 5.1) in the tacrolimus arm and 6.25 mg/day (SD = 3.2) in the sirolimus arm at 1 month after the start of dosing. At 1 month, the mean tacrolimus trough was 10.9 ng/mL (SD = 4.6) in the tacrolimus arm and 6.6 ng/mL (SD = 2.9) in the sirolimus arm. The mean tacrolimus trough remained within the target ranges in both arms within the first 90 days. After 90 days, mean tacrolimus levels remained within the target ranges for the tacrolimus arm but were higher than the target trough range (3–5 ng/mL) in the sirolimus arm.

The mean sirolimus loading dose was 6 mg despite the recommended loading dose of 15 mg; the peak daily dose was 5.9 mg/day (SD = 2.4) at 14 days and 3.2 mg/day (SD = 1.8) by 12 months. The mean sirolimus level remained within the target trough range throughout the study (4–11 ng/mL). The mean trough was 8.04 ng/mL (SD = 4.68) at 14 days and 8.46 ng/mL (SD = 3.12) at 12 months. Between 7 days and 15 months, 15.1% of subjects were above the sirolimus target concentration range and 10.5% were below the target concentration range. Further details of tacrolimus and sirolimus dose and troughs at other time points are provided in Table S3 and Figures S2 and S3.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Disclosure
  9. References
  10. Supporting Information

In this international, multicenter open-label, randomized, active controlled trial, recipients of LTx assigned to receive sirolimus and reduced-dose tacrolimus as compared to standard-dose tacrolimus immediately after transplantation had a significantly higher risk of death and graft loss at 24 months after LTx. The study was terminated early due to a numerical imbalance in the number of HAT/PVT events and subsequent increased rates of death and graft loss in the sirolimus arm. Hence, the primary end point was not met.

Published experience of de novo use of sirolimus subsequent to this trial has been limited to observational and retrospective studies, which have not shown an increased risk of graft failure or HAT [15, 20, 31]. Rates of HAT seen in the sirolimus arm in this study (8%) were much higher than those in other single- and multicenter observational studies (approx. 1–3%) [12, 15, 31]. Lower doses of sirolimus have been employed in some of the studies reporting lower or similar rates of HAT.

The timing of sirolimus initiation may have played a role in the imbalance between adverse events noted in the two study arms. In the Spare the Nephron Liver trial (2005–2007), subjects maintained on MMF and CNI were randomized 4–12 weeks after transplantation to receive MMF and sirolimus or maintained on MMF and CNI [10]. Rates of rejection were higher but graft loss was lower among persons with MMF and sirolimus (3.4% vs. 8.3%). The time to graft loss or death, however, was dissimilar with later deaths (after 6 months) seen in persons maintained on MMF and CNI. HAT occurred in two persons in the MMF and sirolimus arm as compared to one patient in the MMF and CNI arm. There was a nonsignificant increase in discontinuation rates with MMF and sirolimus. Infection rates were higher among persons maintained on MMF and CNI rather than MMF and sirolimus. In another randomized trial examining late conversion to sirolimus (enrollment started in 2002, mean time to conversion 4 years after LTx with only 10–12% converting within 1 year after LTx), 1-year survival rates after conversion were similar [8]. Though graft loss rates were similar, a majority of graft deaths occurred within 20 months in the sirolimus arm (85%) versus CNI continuation where half occurred after 20 months. Treatment failure inclusive of acute cellular rejection was higher among persons converted to sirolimus (48.3% vs. 26.7%). Rates of HAT were not higher in the sirolimus arm. There was a higher rate of discontinuation of sirolimus-based therapy. Infection rates were higher among persons converted to sirolimus [8]. In a recent meta-analysis of smaller controlled studies and observational studies, sirolimus was not significantly associated with death or graft failure, though reporting was incomplete [11]. Once again, however, rates of discontinuation and rates of infection were higher among persons receiving sirolimus.

In the PROTECT study, subjects were randomized to either conversion to everolimus-based immunosuppression or continuation of CNI-based therapy at 4 weeks after LTx [26]. Rates of mortality, efficacy failure, renal function and biopsy-proven acute cellular rejection were similar. Though no increased rates of HAT were observed, there were increased rates of discontinuations and risk of minor infections in the everolimus arm. In a separate multicenter open-label study, de novo liver transplant patients were randomized at 30 days to everolimus plus reduced tacrolimus, tacrolimus (standard of care) or everolimus monotherapy (arm subsequently discontinued) [25]. Similar rates of graft loss or death and lower rates of acute cellular rejection were noted in persons receiving everolimus and reduced-dose tacrolimus. Once again, the risk of infection (relative risk of serious infection = 1.76) and discontinuation for adverse events was higher in the everolimus and reduced tacrolimus arm (26% vs. 14%). There were similar rates of wound healing and HAT.

Thus, everolimus was recently approved for use in LTx beyond the first 30 days in combination with reduced-dose tacrolimus and corticosteroids.

Though higher rates of infection and discontinuation have been noted in the registration trials as described above, higher rates of HAT, graft failure or rejection were not observed.

In addition to timing, dosing and degree of immunosuppression may have also played a role in the different outcomes. The recommended sirolimus loading dose and daily dose may have been excessive. Further, though sirolimus levels were available to guide dosing, the turnaround time for receiving sirolimus trough may have been 3 days or longer. Though the actual loading dose used in the trial was lower than the recommended loading dose, granular data on subjects receiving the higher loading doses and its potential association with events was not available. Immunoassays used at the time of the study may have underestimated sirolimus levels and reported lower blood levels than the later immunoassays used in current practice, leading to potentially higher dosing (Uwe Christians, personal communication). However, no access to stored samples is available to examine this possibility. Despite the protocol necessitating reduced-dose tacrolimus, about 25% of the tacrolimus levels in the sirolimus arm were above 7 ng/mL. Over-immunosuppression may have played a role given that tacrolimus was used in the sirolimus arm at doses considered adequate by itself [27, 32-34]. Indeed, salient features that may differentiate the favorable experience of single centers using sirolimus versus those reported in the current study may be a lower total dose of immunosuppression and/or amelioration of factors that may be associated with a thrombotic tendency (e.g. elevated hematocrit or timing of administration) [12, 18, 31].

Despite being conducted over a decade ago, we feel that our study holds relevance. The study has had a palpable impact on the design of subsequent studies examining the role of mTOR inhibitors after LTx. The study highlights the potential contribution of timing and dose of mTOR inhibitors. This is important given the potential incorporation of everolimus as a therapy to minimize early CNI exposure after LTx. Indeed, issues with dosing, timing and degree of immunosuppression in the present study may have had an impact on the timing and dosing utilized in everolimus registration trials. Finally, given that there is a black box warning based on the results of the study, discussion on data of the thrombosis risk attributed to sirolimus use is important. Though dosing, timing and amount of immunosuppression are put forward as putative factors leading to an increased rate of HAT, no formal evaluation of the causes of HAT in this study is possible so long after its completion. Given that this was the second consecutive study that revealed a higher rate of HAT/PVT among persons receiving sirolimus, discussion of the findings remains relevant. Subsequent observational studies involving de novo use of sirolimus have not reported a similar risk of thrombotic complications.

In patients with available data (89% of patients in the sirolimus arm, 73% of patients in the tacrolimus arm), there was no significant difference in mean change in creatinine clearance at 24 months between the two groups. Similar results have been seen in patients undergoing late conversion [8, 11]. An obvious limitation is missing data and incomplete ascertainment of renal function for both groups; moreover, given that sirolimus was discontinued early on, the equivalent renal function changes at 2 years may reflect that both groups likely received the same standard of care immunosuppression for a large portion of the 2-year evaluation period. Indeed, the difference in creatinine clearance was present at 1 month after therapy (favoring the sirolimus group) but not at later time points.

Our study has several strengths. This was a large multicenter international study that was able to thoroughly examine the safety and efficacy of sirolimus-based de novo therapy in a relevant patient population. Developments of adverse events were carefully monitored. However, there were several limitations. The study was open label and inherent biases of being aware of treatment assignment may have confounded the results, particularly for adverse event reporting. There may have been more aggressive reporting of adverse events in one study arm over the other given that subjects and site personnel were aware of the open-label treatment assignments. Further, site-specific practices may not have been captured; as an example, variations in early postoperative care may have partly driven the results. However, rates of HAT/PVT were not clustered around a select group of centers. Most of the difference in the two treatment arms occurred early and persisted through the trial. While donor age and cold ischemia time were similar across the groups, unrecorded factors, such as donor steatosis, and recipient factors such as pretransplant MELD score, may have affected some of the results.

In summary, in this study of liver transplant patients, a greater incidence of HAT/PVT, death and significant infections was observed in patients assigned to a de novo sirolimus-based regimen with reduced-dose tacrolimus. Transparency is paramount for the transplant community to independently evaluate the role of mTOR inhibitors. Though subsequent studies have not noted the same risk of thrombosis or graft failure, differences in study design and administration (timing and dosing) should be considered along with the data to assess the risk of use of mTOR inhibitors, particularly in the early time period after LTx.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Disclosure
  9. References
  10. Supporting Information

We sincerely thank all of the patients, site investigators, coordinators and staff for their assistance with the study. Other site investigators and study sites at time of enrollment: Prof. Raffaello Cortesini, Universita degli Studi La Sapienza, Roma, Italy; Dr. François Durand, Hôpital Beaujon, Clichy, France; Dr. Christophe Duvoux, Hôpital Henri Mondor, Creteil, France; Prof. Jonathan Fawcett, Queensland Liver Transplant Service, Woolloongabba, Australia; Thomas Fishbein, MD, Mt. Sinai Medical Center, New York, NY; A. Osama Gaber, MD, University of Tennessee Medical Center, Memphis, TN; John Lake, MD, University of Minnesota Medical Center, Minneapolis, MN; Dr. Leslie Lilly, Toronto General Hospital, Toronto, Canada; Prof. G. McCaughan, Royal Prince Alfred Hospital, Camperdown, Australia; Sue McDiarmid, MD, UCLA Medical Center, Los Angeles, CA; Dr. José Mir Pallardo, Hospital de la Fe Unidad Hepatica, Valencia, Spain; Prof. Remo Naccarato, Universita degli Studi di Padova, Padova, Italy; Prof. Peter Neuhaus, University Hospital Charity Rudolf Virchow, Berlin, Germany; Dr. Kevork M. Peltekian and Dr. Vivian McAlister, Queen Elizabeth II Health Sciences Centre, Halifax, Canada; Dr. André Roy, Service de Chirurgie Générale, Montréal, Canada; Dr. Charles Scudamore, Vancouver General Hospital, Vancouver, Canada; James R. Spivey, MD, Mayo Clinic Jacksonville/St. Luke's Liver Diseases and Transplantation, Jacksonville, FL; Dr. William Wall, London Health Sciences Centre, London, Canada. Did not enroll patients: Steve Bynon, MD, University of Alabama—Birmingham, Birmingham, AL; Steven Colquhoun, MD, Cedars-Sinai Hospital, Los Angeles, CA; Mr. Nigel Heaton, Kings College Hospital, London, UK; Mr. Neville Jamieson, University of Cambridge, Cambridge, UK; Prof. Paul McMaster, Queen Elizabeth Hospital, Birmingham, UK; Dr. Peter Metrakos, Royal Victoria Hospital, Montreal, Canada; Mr. Stephen Pollard, St. James University Hospital, Leeds, UK; Andreas G. Tzakis, MD, Jackson Memorial Hospital, Miami, FL. We thank Uwe Christians for his expertise on immunoassays used in measurement of immunosuppression levels. This study was conducted, monitored and paid for by Wyeth, which was acquired by Pfizer in October 2009.

Data access and responsibility: RHW and SKA had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. The manuscript represents valid work and neither this manuscript nor one with substantially similar content under our authorship has been published or is being considered for publication elsewhere. Writing assistance: none; Conception and design: RHW, JFT, GK, JR, NK, LT, JJF, JMM; Acquisition of data: RHW, JFT, GK, EK, EM, JR, NK, LT, JJF, JMM; Analysis: SKA, EK, EM, RHW, JFT; Interpretation of data: all authors; Approval of final manuscript: all authors.

Disclosure

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Disclosure
  9. References
  10. Supporting Information

The authors of this manuscript have conflicts of interest to disclose as described by the American Journal of Transplantation. Dr. Katz and Dr. Maller are employees of Pfizer, Inc. and hold stock and stock options in the company.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Disclosure
  9. References
  10. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Disclosure
  9. References
  10. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
ajt12543-sm-0001-SuppInfo.doc419K

Figure S1: Incidence acute rejection in patients randomized to standard-dose tacrolimus (TAC) or sirolimus and reduced-dose tacrolimus arm (SRL). Closed circle: Sirolimus and reduced-dose tacrolimus. Open circle: Standard-dose tacrolimus.

Figure S2: Tacrolimus dose and trough levels in patients assigned to the standard-dose tacrolimus arm (mean ± standard error).

Figure S3: Sirolimus and tacrolimus dose and trough levels (mean ± standard error) in patients assigned to the sirolimus and reduced-dose tacrolimus arm.

Table S1: Causes of death and graft loss.

Table S2: Renal function as measured by creatinine clearance (mL/min) in standard-dose tacrolimus arm and sirolimus and reduced-dose tacrolimus arm.

Table S3: Drug levels in subjects assigned to the sirolimus and reduced-dose tacrolimus arm.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.