Despite improvements in acute rejection (AR) rates, long-term survival in kidney transplant recipients has not significantly improved, emphasizing the need for immunosuppressive strategies that improve long-term survival (1). Unfortunately, the renal and nonrenal toxicities of immunosuppressants such as calcineurin inhibitors (CNIs) may compromise long-term patient/graft survival (2–6). Belatacept, a selective costimulation blocker, is being assessed as immunosuppressive therapy for kidney transplant recipients. The specificity of belatacept is designed to provide effective immunosuppression while avoiding the renal and nonrenal toxicities associated with CNIs.
The clinical profile of belatacept was demonstrated in two pivotal phase III studies (7, 8). Each phase III study (BENEFIT and BENEFIT-EXT) assessed a more intensive (MI) and a less intensive (LI) belatacept-based immunosuppressive regimen versus a cyclosporine A (CsA)-based regimen in adults receiving a kidney transplant. Both studies were designed as 3-year trials with the primary endpoints assessed after 1 year.
Although the overall safety and tolerability of belatacept-based regimens were similar to CsA in BENEFIT and BENEFIT-EXT, there was an increased incidence of posttransplant lymphoproliferative disorder (PTLD), specifically central nervous system (CNS) PTLD, in the belatacept treatment groups. The risk of PTLD was highest in patients who were Epstein-Barr virus (EBV) seronegative pretransplant and in those who received the belatacept MI regimen. The data support the selection of the LI dose regimen as the preferred option and suggest that immunosuppression with belatacept in kidney transplant patients should be avoided in EBV (−) patients.
These 3-year studies allow continued assessment of the renal benefits of belatacept relative to CsA beyond 1 year and the impact of belatacept's clinical profile on long-term outcomes. This report details outcomes through 2 years of treatment in the BENEFIT and BENEFIT-EXT studies, with additional data on patients who were EBV (+) at baseline—the recommended treatment population. An accompanying article reports further details on the safety of belatacept across BENEFIT, BENEFIT-EXT, and a phase II study in kidney transplant recipients.
Six hundred sixty-six patients in the BENEFIT study and 543 in BENEFIT-EXT study who were randomized and transplanted comprised the respective intent-to-treat populations (Table 1). In each study, most of the discontinuations in each treatment group occurred during the first year.
Demographic and Baseline Characteristics
Transplant recipient demographic and baseline characteristics were well balanced among treatment groups in each study (7, 8). In BENEFIT, 577 of 666 (87%) patients (MI, n=194; LI, n=199; and CsA, n=184) were EBV (+) at baseline, whereas 493 of 543 (91%) patients (MI, n=169; LI, n=156; and CsA, n=168) were EBV (+) at baseline in BENEFIT-EXT.
Patients Surviving With a Functioning Graft
In BENEFIT, 94% (95% confidence interval [CI] 90.9–97.2), 95% (95% CI 91.8–97.6), and 91% (95% CI 86.6–94.4) of intent-to-treat patients in the MI, LI, and CsA groups, respectively, survived with a functioning graft by month 24 (Table 2). Twenty patients (3%) lost their graft (MI, n=7; LI, n=5; and CsA, n=8), and 28 patients (4%) died (MI, n=7; LI, n=8; and CsA, n=13) through month 24. Three patients (0.5%) lost their graft (n=3 MI), and 11 deaths (2%) occurred between months 12 and 24 (MI, n=1; LI, n=4; and CsA, n=6).
In BENEFIT-EXT, 83% (95% CI 77.1–88.1) of intent-to-treat patients in the MI group, 84% (95% CI 78.6–89.4) in the LI group, and 83% (95% CI 77.1–88.1) in the CsA group survived with a functioning graft by month 24; 60 patients (11%) lost their graft (MI, n=18; LI, n=20; and CsA, n=22), and 36 patients (7%) died (MI, n=13; LI, n=11; and CsA, n=12). Seven patients (1%) lost their graft (MI, n=1; LI, n=4; and CsA, n=2), and 15 deaths (3%) occurred between months 12 and 24 (MI, n=5; LI, n=6; and CsA, n=4).
Renal Function—Measured GFR
In BENEFIT, 80% (MI), 81% (LI), and 74% (CsA) of the intent-to-treat patients had measured glomerular filtration rate (mGFR) data available at month 24. At month 24, mean (standard deviation [SD]) mGFR was higher in the belatacept MI (65±27.2 mL/min) and LI groups (68±29.9 mL/min) compared with CsA (51±20.5 mL/min; P<0.0001 MI or LI vs. CsA; Table 2). An on-treatment analysis (MI, n=157; LI, n=170; and CsA, n=146) yielded similar results (MI 67±27.1 mL/min; LI 71±30.3 mL/min; CsA 51±21.1 mL/min; P<0.0001 MI or LI vs. CsA).
In BENEFIT-EXT, 61% (MI), 71% (LI), and 64% (CsA) of the intent-to-treat patients had mGFR data available at month 24. At month 24, mean (SD) mGFR was 52±22.2 mL/min in the belatacept MI group, 50±23.7 mL/min in the LI group, and 45±27.2 mL/min in the CsA group (P=0.028 MI vs. CsA; P=0.1080 LI vs. CsA). An on-treatment analysis (MI, n=107; LI, n=115; and CsA, n=104) yielded similar results (MI 53±22.9 mL/min; LI 53±24.2 mL/min; and CsA 46±30.1 mL/min; P=0.074 MI vs. CsA; P=0.07 LI vs. CsA).
Renal Function—Calculated GFR
In BENEFIT, 82% (MI), 84% (LI), and 74% (CsA) of the intent-to-treat patients had calculated GFR (cGFR) data available at month 24. The mean (SD) cGFR in the MI (70±18.8 mL/min), LI (70±19.7 mL/min), and CsA (53±17.1 mL/min) groups at month 24 were generally consistent with the mGFR results. At month 24, the estimated difference in cGFR between the belatacept and CsA groups was 16 to 17 mL/min, which was a larger difference than that observed at 12 months (15 mL/min/1.73 m2). Mean cGFR over time is depicted in Figure 1 (top). The slope of the cGFR curves from months 3 to 24 indicated an improvement of approximately 1.2 to 1.3 mL/min/1.73 m2/year in the belatacept groups, compared with a decline of approximately 2 mL/min/1.73 m2/year in the CsA group.
In BENEFIT-EXT, 68% (MI), 76% (LI), and 69% (CsA) of the intent-to-treat patients had cGFR data available at month 24. Mean cGFR in the MI (44±26.7 mL/min), LI (43±24.1 mL/min), and CsA (35±21.6 mL/min) groups at month 24 was consistent with the mGFR results. At month 24, the estimated difference in cGFR between the belatacept and CsA groups was 8 to 10 mL/min, compared with 7 mL/min at month 12. Mean cGFR over time is depicted in Figure 1 (bottom). The slope of the cGFR curves from months 3 to 24 indicated a smaller rate of decline (−0.6 to −0.8 mL/min/1.73 m2/year) in the belatacept groups compared with the CsA group (∼2 mL/min/1.73 m2/year).
Renal Function—Chronic Kidney Disease Stages
In BENEFIT, 15% of CsA patients were in chronic kidney disease (CKD) stages 4 to 5 at month 24 compared with 8% each in the MI and LI groups (Fig. 2, top), whereas over twice as many patients in the belatacept groups were in CKD stages 1 or 2 at month 24 compared with the CsA group.
Similarly, fewer belatacept patients in BENEFIT-EXT progressed to advanced CKD stages, but because extended criteria donor kidneys were used, more patients were at higher CKD stages in this study. More CsA patients were in CKD stages 4 to 5 at month 24 compared with the MI and LI groups (Fig. 2, bottom). Over twice as many patients in the belatacept groups were in CKD stages 1 or 2 at month 24 compared with the CsA group.
Clinically Suspected, Biopsy-Proven Acute Rejection
Most episodes of AR occurred within the first 90 days posttransplant in each study, and few AR episodes occurred in either study after month 12. In BENEFIT between month 12 and 24, eight patients had AR episodes (MI, n=4; CsA, n=4; all nonrecurrent), resulting in total AR rates of 24% (MI), 17% (LI), and 9% (CsA) from baseline to month 24 (Table 2). In BENEFIT-EXT between months 12 and 24, there were three patients with AR episodes (MI, n=0; LI, n=1; and CsA, n=2; all nonrecurrent), resulting in total AR rates of 17% (MI), 18% (LI), and 15% (CsA) from baseline to month 24. Among patients who experienced clinically suspected, biopsy-proven AR, 5 of 53 patients in the MI group, 6 of 39 in the LI group, and 0 of 20 in the CsA group experienced graft loss or death by month 24 in BENEFIT, whereas 4 of 32 (MI), 7 of 32 (LI), and 6 of 28 (CsA) patients experienced graft loss or death in BENEFIT-EXT. AR was the adjudicated cause of graft loss in two patients in each treatment group in BENEFIT and in three, three, and six of patients in the MI, LI, and CsA groups, respectively, in BENEFIT-EXT by month 24. In all treatment groups, there was a decline in GFR after AR.
Donor-specific anti-human leukocyte antigen antibodies (DSAs) were present in 4% (MI), 3% (LI), and 7% (CsA) of patients by month 24 in the BENEFIT study. Among patients who had an AR episode, the frequency of DSAs was higher in CsA patients (15%) compared with the MI (4%) and LI (5%) regimens. Similarly, DSAs were present in 5% (MI), 5% (LI), and 12% (CsA) of patients by month 24 in the BENEFIT-EXT study. Among patients who had an AR episode, the frequency of DSAs was notably higher in CsA patients (22%) compared with the MI (6%) and LI (6%) regimens.
Mean systolic blood pressure was 5 to 6 mm Hg lower in the belatacept groups versus CsA in BENEFIT and was 8 to 10 mm Hg lower in BENEFT-EXT. Mean diastolic blood pressure at month 24 was 2 mm Hg lower in the belatacept groups versus CsA in BENEFIT and was 4 to 5 mm Hg lower in BENEFT-EXT. Fewer belatacept patients in each study used multiple (≥3) blood pressure medications (BENEFIT: MI, 30%; LI, 29%; and CsA, 39%) (BENEFIT-EXT: MI, 44%; LI, 38%; and CsA, 50%).
At 2 years, the improved lipid profile associated with belatacept in both studies was sustained versus CsA (Fig. 3). In both BENEFIT and BENEFIT-EXT, the incidence of new-onset diabetes after transplant (5%–8% BENEFIT; 3%–7% BENEFIT-EXT) trended lower in the belatacept groups compared with CsA (11% BENEFIT and 9% BENEFIT-EXT).
Outcomes in EBV (+) Patients
The increased risk of PTLD with belatacept was highest in EBV (−) patients, supporting the EBV (+) population as the recommended patients to be treated with belatacept. These findings, therefore, merit an assessment of outcomes in EBV (+) patients (BENEFIT, n=577; BENEFIT-EXT, n=493). The outcomes in patients who were EBV (+) at baseline were generally reflective of overall population (Table 3). In BENEFIT, 94% (MI), 96% (LI), and 90% (CsA) of EBV (+) patients survived with a functioning graft, whereas composite patient/graft survival in BENEFIT-EXT was 82% (MI), 85% (LI), and 83% (CsA) in EBV (+) patients at month 24. Table 3 shows the survival differences between each belatacept and CsA regimen and the associated 97.3% CIs.
The renal function benefit associated with belatacept compared with CsA was maintained in EBV (+) patients in each study (Table 3). The mean mGFR at month 24 among EBV (+) patients in BENEFIT was 70 mL/min (MI), 72 mL/min (LI), and 53 mL/min (CsA). The mean mGFR at month 24 among EBV (+) patients in BENEFIT-EXT was 53 mL/min (MI), 50 mL/min (LI), and 46 mL/min (CsA). Similarly, the cGFR in each study favored belatacept with mean cGFR values at month 24 of 65±25.1 mL/min (MI), 66±23.9 mL/min (LI), and 48±22.6 mL/min (CsA) in BENEFIT; and 44±27.1 mL/min (MI), 43±23.8 mL/min (LI), and 34±21.7 mL/min (CsA) in BENEFIT-EXT among EBV (+) patients. The rate of AR episodes by month 24 was similar in EBV (+) patients in BENEFIT and in BENEFIT-EXT to the respective overall study populations (Table 3).
In BENEFIT through July 2009, 7 (3%), 9 (4%), and 14 (6%) of patients died in the MI, LI, and CsA groups, respectively. The most common cause of death was infections (MI, n=3; LI, n=2; and CsA, n=5). The overall incidence of serious adverse events (SAEs) was 55% to 59% in the MI and LI groups, compared with 62% in the CsA group. The most common SAEs are listed in Table 4. A similar proportion of patients in each group (6%–7%) discontinued because of SAEs.
In BENEFIT-EXT through June 2009, 19 (10%), 13 (7%), and 16 (9%) of patients died in the MI, LI, and CsA groups, respectively. The most common cause of death was infection (MI, n=5; LI, n=5; and CsA, n=6). The overall incidence of SAEs was 78% to 80% across treatment groups, and the most common SAEs are listed in Table 4. A similar proportion of patients in each group (14%–17%) discontinued because of SAEs.
A similar proportion of patients in each treatment group experienced infections in BENEFIT (MI, 79%; LI, 80%; and CsA, 79%) and BENEFIT-EXT (MI, 80%; LI, 82%; and CsA, 80%). In BENEFIT, 30%, 30%, and 31% of patients in the MI, LI, and CsA groups, respectively, experienced a serious infection. In BENEFIT-EXT, 39% of patients in the LI group experienced a serious infection compared with the MI (48%) and CsA (45%) groups.
The frequency of malignancies in BENEFIT was lower in the LI (4%) and CsA (5%) groups compared with MI group (8%). The most common malignancies were basal cell carcinoma (MI, n=4; LI, n=1; and CsA, n=4) and squamous cell carcinoma (MI, n=3; LI, n=0; and CsA, n=3). In BENEFIT-EXT, the frequency of malignancies was similar across groups: LI (8%), CsA (7%), and MI (9%). The most common malignancies were squamous cell carcinoma (MI, n=4; LI, n=1; and CsA, n=3) and basal cell carcinoma (MI, n=3; LI, n=2; and CsA, n=3).
An increased risk of PTLD associated with the belatacept-based regimens was previously reported based on 1-year data from BENEFIT (8), BENEFIT-EXT (7), and a phase II study (9). The combined data indicated that both EBV (−) and EBV (+) patients were at an increased risk for PTLD with belatacept, specifically CNS PTLD. The highest risk was in patients who were EBV (−) and patients who received the MI regimen. Aside from the previously reported PTLD cases (7, 8), there was a one new CNS PTLD case in BENEFIT-EXT that occurred after month 18. This case occurred in the belatacept LI group in a 62-year-old EBV (+) male with over 4 years exposure to belatacept. The patient died of sepsis with a functioning graft approximately 2 months after the PTLD diagnosis.
At 2 years posttransplant, belatacept-based immunosuppressive regimens were associated with a comparable proportion of patients surviving with a functioning graft, better allograft function, and an improved cardiovascular/metabolic risk profile compared with a CsA-based regimen in both BENEFIT and BENEFIT-EXT. There were few additional episodes of AR during the second year in either study, and no new safety signals were identified.
The renal function benefits demonstrated with belatacept were sustained at 2 years compared with CsA in these phase III studies, consistent with long-term data from a phase II study of belatacept in kidney transplant recipients (9). In fact, the renal function benefit widened over time with belatacept compared with CsA in the BENEFIT study, because of decreasing renal function over time in the CsA group and improving renal function in the belatacept groups. The decreasing renal function in the CsA group is consistent with reports on the impact of CNI-related nephrotoxicity (10–12). The cause of the improvement in renal function over time experienced by belatacept patients in BENEFIT is uncertain, but it may suggest that belatacept enables compensatory hypertrophy in the transplanted kidney, as is seen in the remaining kidney after donor nephrectomy (13, 14).
AR in the belatacept groups occurred early, tended to be single episodes, and was not typically associated with DSAs, whereas AR in the CsA group tended to be more frequently associated with the development of DSAs. There were few episodes of AR in any treatment groups after month 12. The results agree with long-term results using other regimens showing AR as primarily an early event for most patients (15). The lack of association between the AR episodes in the belatacept groups and the development of DSAs is important, because early rejection that is not associated with development of DSAs is generally less detrimental to long-term graft survival (16, 17). In all treatment groups, patients with rejection had lower GFR than those without at both 1 (7, 8) and 2 years. This impact was evident early after the AR episode and was generally stable over time. Multiple factors contribute to this change in renal function, including the episode of AR itself, immunologic factors, switching from belatacept to CNI after AR (among those who changed immunosuppressive therapy following AR), and choice of therapy for the AR episode.
The proportion of EBV (+) patients in BENEFIT and BENEFIT-EXT at baseline reflects the general adult population (18). Overall, the results with belatacept in patients who were EBV (+) at baseline were consistent with overall study results, including a high proportion of patients surviving with a functioning graft and a persistence of the renal benefit over time. By month 24, the overall renal function benefit had reached a 16 to 17 mL/min advantage in cGFR in BENEFIT and an 8 to 10 mL/min advantage in BENEFIT-EXT. The particularly increased PTLD risk in EBV (−) patients confirms EBV seronegativity as a PTLD risk factor and underscores the importance of pretransplant EBV serologic screening and avoidance of belatacept use in EBV (−) patients (19, 20).
The 2-year results of BENEFIT and BENEFIT-EXT show the overall favorable benefit/risk profile of belatacept in kidney transplant recipients. Early risk, in the form of AR (within the first year) and PTLD (mostly within the first 18 months), is counterbalanced by clinically meaningful and durable benefits in renal function, reductions in the incidence of new-onset diabetes, and improvements in blood pressure control and lipid parameters. This is supported by the number of patients experiencing graft loss or death in each study. No new safety signals were observed after the reporting of the first year results. The net impact on long-term survival remains to be seen; however, the available data show that numerically fewer patients on belatacept LI had death or graft loss compared with CsA. Considering that progressive declines in GFR and chronic allograft nephropathy lead to graft loss and 41% of kidney transplant recipients have evidence of diabetes by 3 years posttransplant, and 30% to 40% of kidney transplant recipients who die with a functioning graft do so because of cardiovascular disease, clinical profile of belatacept provides a promising approach to immunosuppression (21, 22).
Results from the 3-year BENEFIT and BENEFIT-EXT studies will continue to be assessed to ascertain whether the favorable benefit/risk profile of belatacept is maintained and whether the results in EBV (+) patients—the recommended treatment population—will continue to demonstrate the utility of belatacept as a basis for immunosuppression in kidney transplant recipients.
MATERIALS AND METHODS
Most methodology pertaining to BENEFIT and BENEFIT-EXT has been described previously (7, 8). Both are 3-year, randomized, partially blinded, active-controlled, parallel-group studies in adult patients aged 18 years or older. Primary outcomes were assessed at 12 months posttransplant. The BENEFIT study included living donor or standard criteria deceased donor kidney transplants with an anticipated cold ischemia time of less than 24 hr (8). The BENEFIT-EXT study used extended criteria donors with an expanded United Network for Organ Sharing definition (7).
Patients in each study were randomized to receive an MI regimen of belatacept, a LI regimen of belatacept, or a CsA regimen for primary maintenance immunosuppression (7, 8). Patients were also treated with basiliximab induction, mycophenolate mofetil, and corticosteroids.
At month 24, each study assessed the proportion of patients surviving with a functioning graft, allograft function, the incidence and characteristics of AR, safety, and secondary outcomes including those that contribute to cardiovascular and metabolic risk. The proportion of patients surviving with a functioning graft, graft loss, and the cause of graft loss and death were adjudicated by committees blinded to treatment assignment. Renal function was assessed by the mGFR using subcutaneously administered cold iothalamate (23), by the cGFR using the Modification of Diet in Renal Disease equation (24, 25), and by the proportion of patients reaching a defined CKD stage at month 24.
The incidence, characteristics, and treatment of clinically suspected, biopsy-proven AR were assessed as described previously (7, 8). Scheduled (protocol) allograft biopsies were performed at implantation, at month 12, and in patients with clinical suspicion of AR. However, protocol biopsies were not performed at month 24. The decision to treat AR was based on a local biopsy reading, and the diagnosis and grade of AR were confirmed by a central pathologist blinded to the patient assignments. The presence of DSAs was performed as described (7, 8).
All analyses at month 24 were conducted on the intent-to-treat population, which was defined as randomized patients who received a transplant. The proportion of patients surviving with functioning graft up to year 2 was summarized using point estimates and corresponding 95% CIs within treatment groups and using two-sided 97.3% CIs for the difference between each belatacept and CsA regimen.
mGFR at month 24 was analyzed using an analysis of variance model with factor for treatment to assess the difference between each of the belatacept treatment groups and CsA. mGFR and cGFR were also assessed using an on-treatment analysis. Safety was assessed using all available data up to a database lock on July 7, 2009 (BENEFIT) or June 30, 2009 (BENEFIT-EXT), corresponding to a median exposure to belatacept of approximately 2.4 years. In addition, outcomes for patient/graft survival, mGFR, cGFR, and AR were assessed descriptively in patients who were EBV (+) at baseline.
The authors thank Brian Atkinson, Ph.D., Bristol-Myers Squibb for writing and editorial assistance.