Sirolimus in De Novo Heart Transplant Recipients Reduces Acute Rejection and Prevents Coronary Artery Disease at 2 Years
A Randomized Clinical Trial
Background— Sirolimus reduces acute rejection in renal transplant recipients and prevents vasculopathy in nonhuman primates and in-stent restenosis in humans. Its effects on rejection and transplant vasculopathy in human heart transplant recipients are unknown.
Methods and Results— In a randomized, open-label study, sirolimus was compared with azathioprine in combination with cyclosporine and steroids administered from the time of cardiac transplantation. We report 6-month rejection rates (primary end point), 12-month safety and efficacy data, and 6- and 24-month graft vasculopathy data in 136 cardiac allograft recipients randomly assigned (2:1) to sirolimus (n=92) or azathioprine (n=44). At 6 months, the proportion of patients with grade 3a or greater acute rejection was 32.4% for sirolimus 3 mg/d (P=0.027), 32.8% for sirolimus 5 mg/d (P=0.013), and 56.8% for azathioprine. Patient survival at 12 months was comparable among groups. Intracoronary ultrasound at 6 weeks, 6 months, and 2 years demonstrated highly significant progression of transplant vasculopathy in azathioprine-treated patients. At 6 months, a highly significant absence of progression in intimal plus medial proliferation and significant protection against luminal encroachment was evident in sirolimus-treated patients, and these effects were sustained at 2 years.
Conclusions— Sirolimus use from the time of transplantation approximately halved the number of patients experiencing acute rejection. The measurable development of transplant vasculopathy at 6 months and 2 years in patients receiving azathioprine was not observed in patients receiving sirolimus.
Received March 7, 2004; revision received June 3, 2004; accepted June 3, 2004.
Immunosuppressive agents are typically judged by their ability to reduce acute rejection without adding toxicity.1,2 However, acute rejection causes only 12% of deaths 1 year after cardiac transplantation, whereas coronary vasculopathy represents 17% of annual mortality beyond 3 years.3 The m-TOR inhibitor sirolimus (rapamycin, Rapamune) has been shown to slow progression of cardiac allograft vasculopathy in heart recipients with established angiographic disease,4 and everolimus has been shown to reduce both acute rejection and cardiac allograft vasculopathy.5
In heart-transplanted animals, sirolimus inhibited proliferation of endothelial cells and fibroblasts5 and inhibited transplant vasculopathy.6 Sirolimus monotherapy prevented vasculopathy in transplanted monkey aortic grafts and, when combined with mycophenolate mofetil, reduced aortic thickening in rat models with carotid injury.7
In renal transplantation, sirolimus reduced acute rejection, graft loss, and death compared with azathioprine8 and placebo.9 In a small dose-ranging study in cardiac transplant recipients, sirolimus used as primary treatment for grade 2 or 3a rejection was effective in a dose-dependent manner.
In humans, sirolimus-eluting stents reduced rates of in-stent stenosis from 20% to 0% at 4 months, with effects maintained through 24 months.10–13 Sirolimus also reduced the end points of death, revascularization, infarction, and angiographic progression in patients with cardiac transplant vasculopathy.4
The aim of this study was to determine the effect of sirolimus, used from the time of transplantation, on acute cellular rejection and graft vasculopathy in human heart transplantation.
Five cardiac transplant centers in Australia and New Zealand enrolled 136 first heart transplant recipients in a randomized, open-label study of 12 months’ duration with safety observations to 24 months.
Inclusion criteria were age ≥18 years, white cell count ≥4.0×109/L, platelet count ≥100×109/L, and contraception for women of childbearing age. Exclusion criteria were multiple organ transplants, serum creatinine >265 μmol/L, familial hypercholesterolemia, pretransplantation cholesterol >387 mg/dL, previous investigational agents, antibody induction, lung infiltrate, systemic infection, and solid-organ malignancy within 5 years.
Approval by the research and ethics committee of each institution was obtained, and all patients gave informed consent.
All patients received cyclosporine (Neoral) and steroids preoperatively and were randomly assigned (2:1) to sirolimus or azathioprine within 8 hours of transplantation. Within 24 hours after transplantation, azathioprine patients received a 3-mg/kg loading dose, then 1 to 2.5 mg · kg−1 · d−1. Sirolimus patients received a 15-mg loading dose, then a 5-mg/d maintenance dose, taken 4 hours after cyclosporine. Intravenous methylprednisolone 500 mg was given preoperatively (125 mg every 8 hours for 24 hours), then oral prednisolone 1 mg · kg−1 · d−1, reducing to 0.1 mg · kg−1 · d−1 by month 6. Cyclosporine dose was reduced over time to target trough levels. Sirolimus dose was adjusted to trough level of 1 to 30 ng/mL (high-performance liquid chromatography).
Halfway through enrollment, concerns about nephrotoxicity and wound healing led to a protocol dose amendment. The sirolimus loading dose was reduced to 10 mg and the maintenance dose to 3 mg/d, with new trough levels of 8 to 18 ng/mL. Cyclosporine target trough levels were reduced by 25% to diminish the cyclosporine/sirolimus pharmacokinetic interaction.14
Required medications included diltiazem 180 to 240 mg/d, pravastatin 40 mg/d, trimethoprim/sulfamethoxazole twice weekly, and intravenous ganciclovir 5 mg/kg per day 3 times weekly for 6 weeks in cytomegalovirus (CMV)-mismatched recipients.
Antibody induction, tacrolimus, mycophenolate mofetil, total lymphoid irradiation, plasmapheresis, photopheresis, agents that increase QT interval, and potent cytochrome P450 inhibitors or inducers were prohibited.
Guidelines for Managing Toxicity
Protocol guidelines for managing sirolimus toxicities involved reduction or temporary cessation of sirolimus if serum lipids or hematological parameters were affected. Hyperlipidemia was managed by changing HMG-CoA reductase inhibitor or adding a second agent before altering sirolimus dosage.
Patients were reviewed at weeks 1 to 4, 6, 8, and 10; months 3 to 8, 10, and 12; and then every 3 months in the safety phase. Endomyocardial biopsy was performed at each visit and at other times if clinically indicated. A pathologist at each center who was blinded to treatment assignment graded cardiac biopsies using International Society of Heart and Lung Transplantation criteria.15 Adverse events, laboratory abnormalities, rejection, death, infection, malignancy, hypertension, hyperglycemia, and hyperlipidemia were recorded.
Selective coronary angiography and intracoronary ultrasound (ICUS) were performed at 6±2 weeks, 6±0.5 months, and 2 years. Vasodilators were withheld for 24 hours before the procedure. Sublingual nitroglycerin (0.4 mg) was given before imaging of a selected coronary artery, performed with a 30- or 40-MHz transducer (Boston Scientific) using continuous motorized pullback at 0.5 mm/s. Follow-up data from identical angiographic projections were analyzed at the core laboratory by D.M. and S.F., who were unaware of treatment assignment. Proximal and mid vessel regions were defined for the interrogated artery, which was almost always the left anterior descending coronary artery. The point of maximum stenosis in the proximal and mid-vessel segments was determined, and 6 slices taken across a 2-mm interval spanning either side of the maximal stenosis were analyzed, yielding 12 slices per study. Intima and media were considered together.
Intent-to-treat analyses were performed for the primary end point and biopsy-confirmed acute rejection at 6 months. Reduction from 90% to 70% in the percentage of patients experiencing rejection was anticipated and considered clinically meaningful. With 136 patients enrolled, a 1-tailed Fisher’s exact test with a 5% significance level would have an 80% power to detect a difference between therapies. Kaplan-Meier analysis was used to evaluate patient survival with log-rank test for time-to-event variables; absolute numbers of events were analyzed by Fisher’s exact test, and differences in demographic and secondary end points were evaluated by use of ANCOVA, with treatment group as a factor and baseline measure of respective parameters as covariate. ICUS parameters were analyzed by paired t test for intragroup and unpaired t test for intergroup differences.
The primary end point was the incidence of first occurrence of biopsy-confirmed acute rejection (≥grade 3a) within 6 months of transplantation or any presumed clinical rejection resulting in treatment. Secondary end points were patient survival, renal function, safety and laboratory parameters, and ICUS parameters, which included maximal intima-plus-media thickness, mean intima-plus-media thickness, mean intima-plus-media area, mean lumen area (LUA), mean vessel area (VA), plaque volume, and plaque burden calculated by the formula VA−LUA/VA.
Of 136 patients enrolled, 34 received sirolimus 3 mg, 57 received sirolimus 5 mg, 43 received azathioprine, and 2 (1 azathioprine, 1 sirolimus 5 mg) never received study medication. The groups were well matched at baseline for all demographic parameters except weight, which was 10% higher in patients receiving sirolimus 3 mg.
All treated rejections were biopsy proven. At 6 months, the primary end point (acute rejection) occurred in 32.4% of patients receiving sirolimus 3 mg (P=0.027) and 32.8% receiving sirolimus 5 mg (P=0.013), compared with 56.8% receiving azathioprine. Only 2 patients in the azathioprine and sirolimus 5 mg groups and 1 in the sirolimus 3 mg group required antilymphocyte therapy.
The trial was not powered to detect differences in mortality; however, survival rates at 12 months did not differ significantly (85.3% sirolimus 3 mg, 86.2% sirolimus 5 mg, 90.9% azathioprine; log-rank P=0.746). One death in each sirolimus group was attributed to an outbreak of Mycoplasma hominis sternal infection in 1 center.16 Causes of death at 12 months are shown in Table 1. No deaths occurred as a result of coronary artery disease.
At 12 months, mean serum creatinine levels were significantly higher in the sirolimus 5 mg group than in the azathioprine group (Figure 1).
Safety and Laboratory Parameters
Intent-to-treat analysis showed a significantly higher incidence of CMV systemic syndrome in azathioprine-treated patients and a significantly higher incidence of pneumonia in both sirolimus groups (Table 2). Malignancies at 12 months were infrequent; 5 patients receiving azathioprine and 4 receiving sirolimus 3 mg developed squamous cell carcinoma of skin, whereas on sirolimus 5 mg, 1 developed prostate carcinoma and 1 lymphoma. Mean systolic and diastolic blood pressures did not differ significantly at 12 months. Total cholesterol and triglyceride levels in sirolimus-treated patients were generally well controlled and are illustrated in Figure 2. Pravastatin was used in 94% of patients (100% sirolimus 3 mg, 89% sirolimus 5 mg, 98% azathioprine).
Mean fasting glucose levels were similar among groups at month 12. Mean white blood cell counts were within normal range and were not significantly different among groups. Mean platelet counts were significantly lower to 3 months on sirolimus, yet within normal range. At all time points to month 12, mean hemoglobin was significantly lower in the sirolimus 5 mg group versus the azathioprine group but was significantly lower only at months 2 and 3 for sirolimus 3 mg.
The percentage of patients discontinued from the study at 12 months was 44% for sirolimus 3 mg, 32% for sirolimus 5 mg, and 40% for azathioprine (P=NS). The main reasons for discontinuation from azathioprine were equally adverse reaction and unsatisfactory response, whereas for sirolimus it was adverse reaction (largely thrombocytopenia).
Dosage and Levels
On sirolimus 5 mg, the mean dose administered at 6 months was 3.7±1.3 mg/d, on sirolimus 3 mg, 2.7±2.0 mg/d, and on azathioprine, 1.6±0.5 mg · kg−1 · d−1. Mean trough levels at 6 months were 18.6 ng/mL in the sirolimus 5 mg group and 11.2 ng/mL in the 3 mg group.
Investigator-Reported Treatment-Emergent Adverse Events
Table 3 shows the incidence of treatment-emergent adverse events. Arrhythmias and atrial fibrillation occurred more often in the azathioprine group, reflecting the higher incidence of rejection. Diarrhea and abnormal renal function, anemia, thrombocytopenia, and mouth ulceration were reported more often in sirolimus-treated patients. Delayed sternal healing occurred in 5 patients on the sirolimus 3 mg dosage and was restricted to 1 center, associated with overly high trough sirolimus levels and related to an outbreak of Mycoplasma hominis.16
Paired data were available for 60 patients (44%) at 6 weeks and 6 months. All patients studied remained continuously on their initial drug assignment. For the purpose of ICUS analysis, data for the 2 doses were combined, because trough levels at 6 months and doses (see above) were similar in the 2 sirolimus groups because of therapeutic drug monitoring. Patients undergoing ICUS were demographically similar to the study group as a whole. All parameters of transplant vasculopathy reflecting mean and maximal intima-plus-media thickness and area and plaque burden and volume increased significantly by 6 months in azathioprine-treated patients. This resulted in significant narrowing in the mean coronary lumen area. Conversely, no parameters increased in those receiving sirolimus, and the differences between groups were statistically significant, indicating that sirolimus prevented the development of vasculopathy (Table 4). The effect on coronary vasculopathy was similar with both doses of sirolimus. Progression of disease did not relate to the slightly higher degree of disease present in the azathioprine group at baseline, because the regression equation showed a poor relationship between plaque burden at baseline and change over time.
Investigator-requested ICUS was performed at 2 years in 21 azathioprine-treated patients and 36 sirolimus-treated patients (42% of enrolled patients). Similar to 6-month data, all parameters of transplant vasculopathy increased markedly at 2 years in patients receiving azathioprine compared with baseline. In contrast, sirolimus patients demonstrated dramatically less transplant coronary disease, with significant preservation of coronary artery lumen (Table 5).
This is the first randomized trial using sirolimus from the time of heart transplantation. The compelling relative reduction in acute rejection (42% sirolimus 3 mg, 43% sirolimus 5 mg) in the first 6 months after heart transplantation was achieved without an increase in diabetes mellitus or malignancy and with comparable survival.
Sirolimus has no inherent nephrotoxicity and has been substituted for calcineurin inhibition to successfully ameliorate chronic calcineurin nephrotoxicity17 and to avert nephrotoxicity in calcineurin-free regimens.18 The creatinine increase observed on sirolimus relates to its interaction with cyclosporine in terms of both blood levels and enhanced concentration of cyclosporine in renal tissue.14,19,20 In addition, diltiazem use in this trial increased bioavailability of sirolimus by up to 30% and also of cyclosporine by a similar amount.21 A reduced dosage of cyclosporine is required when cyclosporine and sirolimus are used together to avoid enhanced cyclosporine nephrotoxicity.14
Although common, hypercholesterolemia was readily controlled by lipid-lowering agents. Triglyceride levels, however, were generally higher in sirolimus-treated patients. The effects on blood parameters for sirolimus were similar to those observed in previous trials.8,9 As observed in renal transplantation trials, sirolimus slowed wound healing in some patients with high trough levels.
Substantial loss of coronary lumen diameter and time-dependent late distal disease occur early after transplantation in large epicardial vessels. Hence, the most compelling finding of the present study was prevention of graft vasculopathy not just at 6 months but also at 2 years in patients receiving sirolimus. Furthermore, prevention of intimal and medial proliferation occurred despite lipid elevations, which were readily controlled. Differences already present at the 6-week ICUS examination may actually reflect an early treatment effect. The progression of coronary disease in this study was shown not to relate to the degree of disease present at baseline. ICUS findings are consistent with the almost complete absence of neointimal proliferation seen with sirolimus-eluting stents in humans.10,11 In addition, <10% of all heart transplants for the period were not enrolled for medical reasons, indicating that >90% of all heart transplant recipients would be candidates for this therapy in practice.
Withdrawal rates on both sirolimus and azathioprine were similar to that seen in previous trials in cardiac transplantation.1,2 Patients withdrawn from either trial agent were switched to mycophenolate mofetil. However, only patients remaining on the initial allocated drug underwent paired ICUS studies.
The mechanism by which sirolimus prevents vasculopathy includes regulation of proliferation and migration of vascular smooth muscle cells, heightened production of nitric oxide, antiangiogenesis effects, and inhibition of extracellular matrix accumulation and fibrosis. A derivative of the parent compound sirolimus, everolimus, was studied contemporaneously in heart transplantation at 2 doses against azathioprine.5 Trial design differed somewhat from the present study in use of induction therapy in 50% of patients, with a composite end point determined at 12 months, yet the 2 studies of m-TOR inhibitors yielded strikingly similar reduction in acute rejection and CMV infection, with comparable patient survival and similar reduction in transplant vasculopathy to 24 months in the context of similar withdrawal rates. Serum creatinine, however, on low- and high-dose everolimus was 20% to 25% higher than on sirolimus at 12 months (related to enhanced cyclosporine nephrotoxicity and leading to a protocol amendment to reduce late cyclosporine target level), and hyperlipidemia was less well controlled.5
This trial was not powered to test survival. The trial drug was not blinded to clinicians or patients but was blinded to pathologists and ICUS core laboratory personnel. However, the open-label nature of the sirolimus study allowed dose-ranging to be explored, and therapeutic monitoring was identified as a critical component, maximizing benefits and minimizing toxicities. Analysis on an intent-to-treat dose basis is somewhat arbitrary, because for the latter half of trial, doses were adjusted according to level. Of all randomized patients, 42% underwent 2 years of ICUS study, although these patients shared the same demographics as the whole group. The doses of sirolimus selected may represent a degree of overimmunosuppression, with more toxicities noted at trough levels of >20 ng/mL. It remains to be determined whether the benefits of sirolimus on proximal to mid-coronary vascular disease at 2 years will translate into benefits on distal disease.
In conclusion, sirolimus therapy used from the time of cardiac transplantation markedly reduced the proportion of patients experiencing acute rejection, with comparable patient survival. Systemically administered sirolimus significantly reduced the progression of coronary vascular disease, with effects sustained at 2 years. Sirolimus 3 mg had efficacy similar to that of sirolimus 5 mg and was associated with lower toxicity. Sirolimus 3 mg/d, adjusted to trough blood levels, is the recommended dosage in this patient population.
The Australian and New Zealand Sirolimus Study Group
Alfred Hospital, Melbourne: Meroula Richardson, MD; Peter Bergin, MD; Kylie Waters, RN; Don Esmore, MD; Bronwyn Levvey, RN; David Kaye, MD. Green Lane Hospital, Auckland: Peter Ruygrok, MD; Helen Gibbs, RN; H. Arthur Coverdale, MD; Denise Reddy, RN; David Haydock, MD. Prince Charles Hospital, Brisbane: Andrew Galbraith, MD; Sarah Gray, RN; Scott Bell, MD; Carol Swan, RN. Royal Adelaide Hospital, Adelaide: Enzo de Angelis, MD. Royal Perth Hospital, Perth: Gerry O’Driscoll, MD; Clare Wood, RN; Anuj Patel, MD; Helen Hayes, RN. St Vincent’s Hospital, Sydney: Anne Keogh, MD; Peter Macdonald, MD; David Muller, MD; Phillip Spratt, MD; Maria Correas, RN; Marie Sprunt, RN; Steve Faddy, MSc Med.
This study was supported by a grant from Wyeth Research, Collegeville, Pa. We acknowledge Paul Taylor, Princess Alexandria Hospital, Brisbane, for performance of sirolimus high-performance liquid chromatography; Joseph A. Scarola, MD, and Bernadette Maida for study design guidance; Zhe Shang for statistical analyses; and Susan A. Nastasee and Cathy Borg for manuscript editing and preparation.
Taylor DO, Edwards LB, Mohacsi PJ, et al. The registry of the International Society for Heart and Lung Transplantation: twentieth official adult heart transplant report – 2003. J Heart Lung Transplant. 2003; 22: 6:616–624.
Mancini D, Pinney S, Burkhoff D, et al. Use of rapamycin slows progression of cardiac transplantation vasculopathy. Circulation. 2003; 108: 48–53.
MacDonald AS, for the Rapamune Global Study Group. A worldwide, phase III, randomized, controlled, safety and efficacy study of sirolimus/cyclosporin regimen for prevention of acute rejection in recipients of primary mismatched renal allografts. Transplantation. 2001; 71: 271–280.
Sousa JE, Costa MA, Abizaid AC, et al. Sustained suppression of neointimal proliferation by sirolimus-eluting stents: one-year angiographic and intravascular ultrasound follow-up. Circulation. 2001; 104: 2007–2011.
Sousa JE, Costa MA, Abizaid A, et al. Lack of neointimal proliferation after implantation of sirolimus-coated stents in human coronary arteries: a quantitative coronary angiography and three-dimensional intravascular ultrasound study. Circulation. 2001; 103: 192–195.