Calcineurin-Inhibitor-Free Immunosuppression Based on the JAK Inhibitor CP-690,550: A Pilot Study in De Novo Kidney Allograft Recipients


* Corresponding author: Stephan Busque,


This randomized, pilot study compared the Janus kinase inhibitor CP-690,550 (15 mg BID [CP15] and 30 mg BID [CP30], n = 20 each) with tacrolimus (n = 21) in de novo kidney allograft recipients. Patients received an IL-2 receptor antagonist, concomitant mycophenolate mofetil (MMF) and corticosteroids. CP-690,550 doses were reduced after 6 months. Due to a high incidence of BK virus nephropathy (BKN) in CP30, MMF was discontinued in this group. The 6-month biopsy-proven acute rejection rates were 1 of 20, 4 of 20 and 1 of 21 for CP15, CP30 and tacrolimus groups, respectively. BKN developed in 4 of 20 patients in CP30 group. The 6-month rates of cytomegalovirus disease were 2 of 20, 4 of 20 and none of 21 for CP15, CP30 and tacrolimus groups, respectively. Estimated glomerular filtration rate was >70 mL/min at 6 and 12 months (all groups). NK cells were reduced by ≤77% in CP-690,550-treated patients. In the CP-690,550 arms, there were modest lipid elevations and a trend toward more frequent anemia and neutropenia during the first 6 months. These data suggest that coadministration of CP-690,550 30 mg BID with MMF is associated with overimmunosuppression. At 15 mg BID, the efficacy/safety profile was comparable to the tacrolimus control group, excepting a higher rate of viral infection. Further dose-ranging evaluation of CP-690,550 is warranted.


The Janus kinases (JAKs) and their downstream transcription-regulation proteins, called signal transducers and activators of transcription (STATs), mediate ligand-specific signal transduction between cell membrane receptors and the nucleus. In particular, inhibition of one of the JAKs called JAK3 blocks the signaling of the gamma chain subfamily of cytokines (interleukins 2, 4, 7, 9, 15 and 21) (1), resulting in immunosuppression. CP-690,550 is an orally active JAK inhibitor that potently inhibits JAK3 and prolonged kidney allograft survival in cynomolgus monkeys without concomitant calcineurin inhibitor (CNI) (2–4). Here, we report the first clinical trial of CP-690,550 in de novo human kidney transplantation based on a 6-month pilot study with follow-up through month 12.

Materials and Methods

Study design

A3921009 was a phase 2A, multicenter, randomized, open-label study with an active control. Patients who completed 6 months of treatment in the parent study (A3921009) were enrolled in an extension protocol (A3921021) (Figure 1). This study was conducted in 16 transplant centers in the United States. Both trials were approved by local institutional review boards and conducted in compliance with the Declaration of Helsinki. Written informed consent was obtained from each patient. A data-monitoring committee periodically reviewed the aggregate safety findings through 6 months. At the time of database closing, all patients had at least 12 months of posttransplantation follow-up.

Figure 1.

Study design.

A3921009 was originally designed as two sequential enrollment stages: a pilot trial (Stage 1; N∼54) followed by a larger confirmatory trial (Stage 2; N∼195) of identical design, both of 6-month duration. It was prespecified that Stage 2 would begin if one of the following conditions were met in Stage 1: (a) all CP-690,550-treated patients had completed 3 months of treatment without any biopsy-proven acute rejection (BPAR) or (b) all CP-690,550-treated patients had completed 6 months of treatment with a similar BPAR rate to tacrolimus and no safety concerns. Stage 2 was not initiated due to evidence of overimmunosuppression in the higher-dose CP-690,550 group in Stage 1, as manifest by a 20% incidence of BK-virus nephropathy (BKN). Therefore, data herein include only Stage 1 of A3921009 and its extension through 12 months.

Study treatment

Patients who received first kidney transplants and had immediate graft function defined by a ≥15% decline in serum creatinine and urine output of ≥40 mL/h during the first 24 h postoperatively were eligible. Key exclusion criteria included nonkidney transplants, retransplants and immunologically high-risk patients (defined as African Americans who had current panel-reactive antibodies >30%). Eligible patients were randomized 1:1:1 to three treatment groups: CP-690,550 15 mg twice daily (BID) (CP15), CP-690,550 30 mg BID (CP30) and tacrolimus. After 6 months of treatment, eligible patients were enrolled in the extension study in their previously assigned treatments. The tacrolimus dose was adjusted to achieve 12-h trough levels of 7–14 ng/mL during the first 3 months and 5–12 ng/mL for months 4 through 12. CP-690,550 doses were reduced after 6 months to 10 mg BID in CP15 and 15 mg BID in CP30. All patients received induction with daclizumab (N = 1) or basiliximab (N = 60). Mycophenolate mofetil (MMF) was initiated at 2 g/day; African Americans were allowed to receive up to 3 g/day (N = 2). All patients received corticosteroids and were maintained on at least 5 mg/day of prednisone through month 12.

The A3921009 protocol was amended after all patients had been enrolled because 4 of 20 patients in the CP30 group developed BKN. MMF was discontinued over 2–4 weeks in this group, steroids were tapered more rapidly and monitoring of the BK viral load in plasma was added. At the time of the protocol amendment, five patients in CP30 had completed 6 months of follow-up. Cytomegalovirus (CMV) prophylaxis was required for at least 3 months in high-risk (seropositive donor, seronegative recipient [D+R]) patients, and Pneumocystis carinii prophylaxis (e.g. with sulfamethoxazole/trimethoprim) was required for at least 6 months. Use of erythropoiesis-stimulating agents was limited to the first 2 months posttransplant to avoid masking potential effects of CP-690,550 on hematopoiesis. Use of granulocyte growth factors was prohibited.

Study assessments

The primary efficacy end point was 6-month incidence of first BPAR, assessed by a central pathologist blinded to treatment assignment. Secondary efficacy end points included graft loss, patient death and severity of first BPAR. Estimated glomerular filtration rate (eGFR) at 6 months (calculated by the Nankivell equation (5)) was the primary safety end point. Secondary safety end points included adverse events (AEs) and laboratory parameters. The following peripheral lymphocyte subsets were evaluated using flow cytometry: T-lymphocytes (CD3+), B-lymphocytes (CD19+) and NK cells (CD56+ through month 6, CD3CD56+ at month 12).

Statistical methods

Study A3921009 was designed with at least 80% power to demonstrate superiority of at least one of the CP-690,550 groups over tacrolimus for Nankivell eGFR, based on the projected combined sample size of Stages 1 and 2. All analyses were based on the full analysis set, which included all patients who were randomized and received at least one dose of study medication. As applicable, all primary and secondary analyses had 80% two-sided confidence intervals (CIs) or a one-sided significance level of 10% (prespecified in the protocol). For continuous data (including eGFR), least squares means (LSMs) and CIs were calculated using a linear mixed model with repeated measures incorporating all available data, which accounts for missing data implicitly. Means (standard deviation [SD]) were used where applicable. Event rates, such as BPAR rates, were estimated using Kaplan–Meier curves, and CIs for the rate difference were formed using normal approximation to binomials based on the Kaplan–Meier estimates. Treatment comparisons were made between each CP-690,550 dose and tacrolimus without adjustments for multiple comparisons. As only Stage 1 was completed, the statistical analysis of this study should be regarded as exploratory only.


Patient baseline characteristics, disposition and immunosuppression

Patients were enrolled between June 2005 and December 2005. In total, 61 patients were randomized: 20 to CP15, 20 to CP30 and 21 to tacrolimus in A3921009. The three groups were comparable in baseline characteristics except more patients with pretransplant diabetes mellitus were randomized to the CP-690,550 groups (Table 1). Most (75%) were recipients of living-donor allografts, with 41% of them being pre-emptive transplants.

Table 1.  Patient demographics and baseline characteristics
 CP15 (n = 20)CP30 (n = 20)Tacrolimus (n = 21)
  1. 1Asian, Hispanic and Native American. CMV = cytomegalovirus; HLA = human leukocyte associated.

Sex, n
Age, years
 Mean (range)46.9 (31–61)44.3 (20–68)42.8 (27–64) 
Race, n
Medical history, n (%)
 Pretransplant diabetes mellitus6 (30)5 (25)2 (9.5)
Number of HLA mismatches
 Mean (standard deviation)3.4 (1.8) 3.7 (1.7) 3.3 (2.0)  
Donor age, years
 Mean (range)41 (6–63) 39 (12–68)36 (14–55)
Donor relationship, n (%)
 Living related9 (45)6 (30)5 (23.8)
 Living unrelated8 (40)7 (35)11 (52.4) 
 Deceased donor3 (15)7 (35)5 (23.8)
Cold ischemic time for deceased donor, hours
 Mean (range)9.1 (2–19) 17.5 (5.6–25)19.8 (7–31)
CMV match status, n (%)
 D+R+8 (40)5 (25)13 (61.9)
 D+R2 (10)5 (25)0
 DR+5 (25)3 (15)3 (14.3)
 DR5 (25)7 (35)5 (23.8)

Patient disposition is shown in Figure 2. More patients discontinued treatment prematurely in the CP30 group than in the tacrolimus or CP15 groups. At month 6, 3 of 20 discontinued in CP15 group (embolic stroke, acute rejection, CMV disease and anemia); 7 of 20 in CP30 group (acute rejection [three patients], CMV disease, administrative issues [patient ineligibility, withdrawal of consent, lost to follow-up]) and 2 of 21 in the tacrolimus group (anemia, lost to follow-up). After 6 months, an additional patient in the CP30 group discontinued due to ongoing BKN.

Figure 2.

Patient participation throughout the parent and extension studies.

The CP-690,550 doses in CP15 and CP30, the tacrolimus doses and trough levels in the control arm and the MMF and prednisone doses in all treatment arms at months 6 and 12 are shown in Table 2.

Table 2.  Mean ± SD CP-690,550, MMF, prednisone, and tacrolimus daily doses, and tacrolimus trough concentration at months 6 and 12
  1. MMF = mycophenolate mofetil; SD = standard deviation.

CP-690,550 daily dose (mg)
 Month 626.9 ± 8.9 50.0 ± 21.7-
 Month 1217.1 ± 5.8 23.1 ± 10.3-
MMF daily dose (gm)
 Month 61.8 ± 0.50.4 ± 0.71.7 ± 0.5
 Month 121.5 ± 0.70.2 ± 0.41.9 ± 0.4
Prednisone daily dose (mg)
 Month 65.4 ± 2.85.5 ± 2.76.1 ± 3.8
 Month 124.6 ± 1.34.6 ± 1.45.0 ± 2.4
Tacrolimus daily dose (mg)
 Month 66.3 ± 5.5
 Month 126.2 ± 5.6
Tacrolimus trough concentration (ng/mL)
 Month 68.5 ± 4.0
 Month 1210.2 ± 4.8 


Acute rejection:  Six patients developed BPAR by 6 months posttransplant (one in CP15 group, four in CP30 group and one in the tacrolimus group). The 6- and 12-month first BPAR rates and the severity of BPAR episodes are shown in Table 3. One BPAR in the CP30 group was diagnosed locally as BKN (based on light microscopy examination and SV40 stain), but the SV40 stain was not available to the central pathologist. Another BPAR in the CP30 group was diagnosed locally as recurrence of lupus, but the immunofluorescence stains were not available to the central pathologist. This patient subsequently developed a BPAR that was confirmed centrally and locally. Figure 3 shows the Kaplan–Meier curves for freedom from BPAR. Most of the BPAR episodes occurred prior to month 3.

Table 3.  Kaplan–Meier estimates of the first biopsy-proven acute rejection (BPAR) rate and severity of BPAR episodes through month 12
 CP15 (n = 20)CP30 (n = 20)Tacrolimus (n = 21)
  1. 1Kaplan–Meier estimated BPAR rate difference between each CP-690,550 group and tacrolimus.

  2. CI = confidence interval.

Kaplan–Meier estimated BPAR rate %
 6 months5.321.14.8
 12 months5.321.19.8
Kaplan–Meier estimated BPAR rate difference and 80% CI (%)1
 6 months0.5 (–8.4, 9.4)16.3 (2.9, 29.7)
 12 months–4.5 (–15.2, 6.2)11.3 (–3.4, 25.9)
Number of patients with BPAR at month 12142
Severity of first BPAR (Banff 97 schema)
Time of onset of first BPAR posttransplant4.5 months1.3, 1.5, 1.7, 2.1 months3 days, 6 months
Figure 3.

Kaplan–Meier curve showing probability of freedom from first biopsy-proven acute rejection (BPAR) through 12 months.

Through 12 months, BPAR was reported in one CP15 patient, four CP30 patients and two tacrolimus patients. One patient in the tacrolimus group developed a second episode of acute rejection approximately 8.5 months posttransplant. None of the CP-690,550 treatment patients developed repeat acute rejections through month 12.

Graft survival and patient survival

There was one death in the CP30 group. After reducing the CP-690,550 dose to 15 mg BID because of oral thrush, this patient developed a grade 2B refractory BPAR 5 months posttransplant. The patient returned to dialysis and died approximately 8 months posttransplant from arrhythmia associated with hyperkalemia. There were no other graft losses in the study.


Kidney function:  The Nankivell eGFRs were similar across the treatment groups at 6 and 12 months. The LSM eGFRs were 76.9, 72.8 and 77.4 mL/min for CP15, CP30 and tacrolimus groups, respectively, at 6 months; and 83.6, 77.6 and 73.3 mL/min for CP15, CP30 and tacrolimus, respectively, at 12 months (Figure 4).

Figure 4.

Least squares means and standard errors of estimated glomerular filtration rate calculated by the Nankivell equation.

Adverse events:  Through 12 months, serious AEs (SAEs) were reported in 50% of patients in CP15 group, 50% of patients in CP30 group and 38% of patients in the tacrolimus group. Acute rejection and infection episodes accounted for the most commonly reported SAEs. Two patients in the CP30 group developed malignancies. One patient developed basal cell carcinoma and one developed mixed carcinoma of the appendix; both were surgically excised without sequelae. No patient developed posttransplant lymphoproliferative disorder.

Through 12 months, the most common treatment-emergent AEs by system organ class were gastrointestinal disorders, infections and general disorders (Table 4). There were more nervous system AEs in the tacrolimus group and more acneiform eruptions and/or rash in CP30 group.

Table 4.  Most common treatment-emergent adverse events by system organ class and preferred term at month 12
 CP15 (n = 20)CP30 (n = 20)Tacrolimus (n = 21)
  1. ALT = alanine aminotransferase; AST = aspartate aminotransferase; CMV = cytomegalovirus; HZV = herpes zoster virus.

Gastrointestinal disorders, number of patients, n13 13 15 
 Lower abdominal pain321
 Abdominal pain221
Infections and infestations, n14 13 11 
 Urinary tract infection154
 CMV infection341
 HZV infection401
General disorders, n12 10 13 
 Edema peripheral457
Investigations, n714 9
 Blood creatinine increased272
 ALT increased262
 AST increased252
 Weight increased133
Metabolism and nutrition disorders, n13 10 5
 Hyperglycemia (including patients with pretransplant diabetes mellitus)410
Nervous system disorders, n6414 
Blood and lymphatic system disorders, n796
Skin and subcutaneous tissue disorders, n696

There was a higher rate of clinically significant infections in the CP30 group than in the tacrolimus group at 6 and 12 months (Table 5). BKN developed in four patients in CP30 group before month 6. Since this finding was confined to the CP30 group, the protocol was amended to discontinue MMF in that group. No further cases of BKN were reported after the protocol amendment. In addition to the four BKN cases in CP30, BK viremia was detected during the first 12 months in two patients in the CP15 group, none in the CP30 group and three in the tacrolimus group.

Table 5.  Clinically significant infections, cytomegalovirus (CMV) disease, incidence of BK virus nephropathy and new-onset infections through month 12
 CP15 (n = 20)CP30 (n = 20)Tacrolimus (n = 21)
  1. 1Rates in parentheses were estimated using Kaplan–Meier curves.

  2. 2One-sided p-value < 0.1 for comparing the rate difference estimated from the Kaplan–Meier curves between CP-690,550 and tacrolimus using the Wald test.

  3. 3Three of the seven patients who developed CMV disease were in the high-risk subset (D+R): one patient in the CP15 group and two patients in the CP30 group.

  4. HZV = herpes zoster virus.

Cumulative number and percent of patients with clinically significant infections, n (%)1
 6 months6 (30.0)11 (59.3)25 (24.4)
 12 months9 (47.5)11 (59.3)27 (35.2)
Cumulative number and percent of patients with CMV disease,3 n (%)1
 6 months 2 (10.3)2 4 (21.1)20 (0)
 12 months 3 (16.3)2 4 (21.1)20 (0)
BK virus nephropathy040
Extension phase only
 (n = 14)(n = 13)(n = 18)
New-onset infections between months 6 and 12, n757
 HZV infections401

The 6-month rates of CMV disease were 2 of 20, 4 of 20 and none of 21 for CP15, CP30 and tacrolimus groups, respectively, and two patients developed coinfection with BKV and CMV. Only one of the six cases of CMV disease (in CP30) was tissue invasive, as diagnosed by kidney allograft biopsy. All other cases were diagnosed by symptoms and laboratory parameters. At 12 months, one additional patient had developed CMV disease in the CP15 group. A case of CMV disease developed in the tacrolimus group approximately 13 months posttransplant. Between 6 and 12 months, herpes zoster infection developed in five patients (four in the CP15 group and one in the tacrolimus group).

There was no difference among the treatment groups in the incidence of new-onset diabetes mellitus. Wound dehiscence occurred in one patient in the tacrolimus group and none in the CP-690,550 groups. Posttransplant lymphocele or seroma occurred in one patient in the CP15 group and in three patients in the CP30 group.

Laboratory parameters:  More patients in the CP-690,550 groups had neutrophil counts <1000 cells/mm3 (one, two and none patients for CP15, CP30 and tacrolimus groups, respectively) and hemoglobin concentrations <10 g/dL (four, three and two patients for CP15, CP30 and tacrolimus, respectively) during the first 6 months. By 12 months, there was no apparent difference between the CP-690,550 groups and the tacrolimus group in hemoglobin concentration (13.5 vs. 14.0 vs. 14.0 g/dL), neutrophil counts (4.4 vs. 4.4 vs. 4.9 ×1000 cells/mm3) or platelet counts (254.0 vs. 242.0 vs. 245.3 cells/mm3) (Figures 5A–C). Total cholesterol, low- and high-density lipoprotein cholesterol and triglycerides were increased in the CP15 and CP30 groups by up to 34% and 44%, respectively, compared with the tacrolimus group through month 12. More patients received lipid-lowering medications in the CP15 and CP30 groups than in the tacrolimus group (9 of 14, 7 of 13 and 6 of 18, respectively) at month 12. Among patients who did not have proteinuria pretransplant, three patients in each treatment group developed proteinuria (≥100 mg/dL) on dipstick during the first 12 months.

Figure 5.

Mean (standard deviation) of hemoglobin, absolute neutrophil count, absolute platelet count and absolute NK cell count over time through 12 months.

Circulating NK cells were lower in both CP-690,550 treatment groups than in the tacrolimus group at months 6 and 12. At 12 months, mean CD3CD56+ counts were decreased by 67% in the CP15 group and by 77% in the CP30 group compared with the tacrolimus group (Figure 5D). No association between circulating NK cell counts and opportunistic viral infections was evident. There was no apparent difference in circulating CD3+ or CD19+ cell counts among the three groups through month 12.

Blood pressure:  Systolic and diastolic blood pressures were similar between the CP-690,550 groups and the tacrolimus group at 6 months. By 12 months, the mean systolic blood pressures in all groups were reduced from the pretransplant baseline. The mean ± SD decrease from baseline in systolic blood pressure for tacrolimus (−17.1 ± 30.5 mmHg) was larger than from both CP15 (−11.6 ± 30.6 mmHg) and CP30 (−11.2 ± 20.0 mmHg). Most patients required antihypertensive medications posttransplant (12 of 14, 9 of 13 and 14 of 18 at 12 months for CP15, CP30 and tacrolimus groups, respectively).


This is the first evaluation of CP-690,550, a JAK inhibitor, after de novo human kidney transplantation. This phase 2A study was planned to be conducted in two stages. In accordance with criteria prespecified in the protocol, the second stage was not initiated because four cases of BKN occurred in the CP30 group. Consequently, this study became an exploratory study without sufficient power to definitively address its primary and secondary objectives. Nevertheless, it provides preliminary information on the efficacy and safety of a new mechanism of clinical immunosuppression.

There are several important findings from these studies. The first is the demonstration of high immunosuppressive potency with CP-690,550 at the doses studied. There were higher rates of opportunistic viral infections in the CP-690,550-treated patients, particularly in the CP30 group, than the control group. There was a high incidence (4 of 20) of BKN when CP-690,550 30 mg BID was coadministered with MMF. CMV disease was observed only in the two CP-690,550 groups (six cases in 40 patients by 6 months, Table 5), including four primary infections and two reactivations. Although all high-risk CMV recipients (D+R) were randomized to the CP-690,550 groups, this observation may suggest increased susceptibility to CMV infection in CP-690,550-treated patients. Whether this represents the result of increased aggregate immunosuppression or an inherent risk of combining CP-690,550 and MMF is unknown.

In terms of efficacy, evidence of pharmacological activity was provided by the comparable BPAR rates in the CP-690,550 treatment groups to those reported for CNI-based regimens (6,7). The BPAR rate was similar between CP15 and tacrolimus. The paradoxically higher BPAR rate in the CP30 group may be a result of the small number of observations in this pilot study.

In the CP15 and CP30 groups, patients had dose reductions to 10 mg BID (with continuation of MMF) and 15 mg, respectively, between months 6 and 9. Despite the dose reductions, there were no new BPAR episodes after month 6. These findings suggest that, beyond month 6, CP-690,550 doses of ≤10 mg BID in conjunction with MMF or ≤15 mg BID without MMF may be sufficient to maintain efficacy.

Kidney function for all groups, including the tacrolimus control group, was excellent throughout the follow-up period and higher than reported in recent studies (8,9). This may be attributable to patient selection in this study, in which approximately 75% of patients received living-donor allografts, with 41% of them being pre-emptive transplants.

CP-690,550 cross-reacts to a limited extent with JAK2, which mediates the downstream effects of various hematological growth factors. An unintended effect on hematological parameters may result from inhibition of JAK2 and be further compounded by concomitant medications (e.g. MMF, valganciclovir) and CMV disease. More cases of neutropenia (neutrophils <1000 cells/mm3) or anemia (hemoglobin <10 g/dL) occurred during the first 6 months posttransplant in the CP-690,550-treated patients. With reduction of immunosuppression in the CP-690,550 groups, there were no differences in neutrophil count or hemoglobin by 12 months.

Serum lipid levels were elevated modestly in the CP-690,550 treatment groups compared with the tacrolimus group. Lipid-lowering agents were also used more frequently in CP-690,550-treated patients. Since known JAK3 expression is restricted to the hematologic system, the mechanism for this is unclear. The clinical significance of the moderate reduction in NK cell counts is unknown.

Limitations of this trial include the relatively small sample size (∼20 patients per group), the use of only two dose levels of CP-690,550 and the imbalance in distribution of CMV high-risk patients among treatment groups. The observed safety and tolerability profile should be interpreted in light of the dose-finding nature of this pilot evaluation. Trends observed in the data of this pilot trial should be interpreted as hypothesis-generating only.

In summary, inhibition of the JAK3 pathway with CP-690,550 resulted in relatively low rates of acute rejection in this pilot study, with evidence of overimmunosuppression when CP-690,550 30 mg BID was combined with MMF. The incidence of BKN in patients treated with 30 mg BID of CP-690,550 in combination with MMF was unacceptably high. CP-690,550 15 mg BID coadministered with MMF resulted in similar outcomes to the tacrolimus control group but was associated with modest lipid elevations and a higher rate of viral infections. Considering the known short- and long-term deleterious effects of CNI-based regimens, further evaluation of CP-690,550 is warranted.


This research was sponsored by Pfizer, Inc. Editorial support for this manuscript was provided by Dr. Dean Clarke of Complete Medical Communications and was funded by Pfizer, Inc.

The protocol and its amendments were designed collaboratively by Pfizer and the investigators. Data were collected by centers of the A3921009 and A3921021 study group. Coordination of study operations was provided by Pfizer, which held the data and performed data analysis. Data interpretation was performed by Pfizer and Drs. Busque, Leventhal, Brennan, Steinberg, Klintmalm, Hariharan and Vincenti. The manuscript was drafted by Drs. Busque, Brennan and Chan and critically reviewed by the other authors.

We gratefully acknowledge the contribution of the principal investigators from the A3921009 study group: Dr. T. Batiuk (Legacy Good Samaritan Hospital, Portland, OR), Dr. D. C. Brennan (Washington University, St. Louis, MO), Dr. J. Bromberg (Mount Sinai School of Medicine, New York, NY), Dr. S. Busque (Stanford University Medical Center, Palo Alto, CA), Dr. R. A. Fisher (Virginia Commonwealth University, Richmond, VA), Dr. S. Hariharan (Medical College of Wisconsin, Milwaukee, WI), Dr. G. Klintmalm (The Baylor Regional Transplant Institute, Dallas, TX), Dr. J. Leventhal (Northwestern University, Chicago, IL), Dr. S. Mulgaonkar (Saint Barnabas Medical Center, Livingston, NJ), Dr. V. R. Peddi (California Pacific Medical Center, San Francisco, CA), Dr. T. Shah (The National Institute of Transplantation, Los Angeles, CA), Dr. D. Slakey (Tulane University School of Medicine, New Orleans, LA), Dr. S. Steinberg (Sharp Memorial Hospital, San Diego, CA), Dr. D. Thomas (Weill Cornell Medical Center, New York, NY), Dr. F. Vincenti (University of California, San Francisco, CA) and Dr. A. Wiseman (University of Colorado Health Sciences Center, Denver, CO). numbers: Study A3921009 – NCT00106639; Study A3921021 – NCT00263328.

Conflict of Interest Statement

J. Leventhal, D. C. Brennan, S. Steinberg, J. Bromberg, G. Klintmalm and R. A. Fisher have been consultants for Pfizer. S. Busque, S. Hariharan, S. Mulgaonkar, T. Shah, V. R. Peddi, D. Slakey and F. Vincenti have no conflicts of interest for this study. G. Chan, N. Lawendy and C. Wang are employees of Pfizer and, as such, may be eligible to receive stock options. J. Leventhal is married to a Pfizer employee and also has a brother who is a Pfizer employee.