ABSTRACT
Background
We previously reported excellent efficacy and improved safety aspects of rapid steroid withdrawal (RSWD) in the randomized controlled 1-year “Harmony” trial with 587 predominantly deceased-donor kidney transplant recipients randomized either to basiliximab or rabbit antithymocyte globulin induction therapy and compared with standard immunosuppressive therapy consisting of basiliximab, low tacrolimus once daily, mycophenolate mofetil and corticosteroids.
Methods
The 5-year post-trial follow-up (FU) data were obtained in an observational manner at a 3- and a 5-year visit only for those Harmony patients who consented to participate and covered clinical events that occurred from the second year onwards.
Results
Biopsy-proven acute rejection and death-censored graft loss rates remained low and independent of RSWD. Rapid steroid withdrawal was an independent positive factor for patient survival (adjusted hazard ratio 0.554, 95% confidence interval 0.314–0.976; P = .041).
The reduced incidence of post-transplantation diabetes mellitus in RSWD patients during the original 1-year study period was not compensated by later incidences during FU. Incidences of other important outcome parameters such as opportunistic infections, malignancies, cardiovascular morbidity/risk factors, donor-specific antibody formation or kidney function did not differ during FU period.
Conclusions
With all the limitations of a post-trial FU study, the Harmony FU data confirm excellent efficacy and beneficial safety aspects of RSWD under modern immunosuppressive therapy over the course of 5 years after kidney transplantation in an immunologically low-risk, elderly population of Caucasian kidney transplant recipients.
Trial registration: Clinical trial registration number: Investigator Initiated Trial (NCT 00724022, FU study DRKS00005786)
Keywords: Harmony study, long-term follow-up, mortality, rapid steroid withdrawal, renal transplantation
KEY LEARNING POINTS.
What was known:
Short-term results of rapid steroid withdrawal (RSWD) 1 year after renal transplantation demonstrate excellent efficacy and improved safety aspects compared with the current gold standard of immunosuppressive therapy consisting of basiliximab induction followed by corticosteroids, mycophenolate and tacrolimus (Harmony study).
Hereby, RSWD combined with induction therapy using either basiliximab or low-dose thymoglobulin led to equivalent 1 year results. RSWD allowed almost a bisection of the rate of posttransplantation diabetes mellitus but did not lead to an increase of biopsy-proven acute rejections (BPAR) compared with continuous corticosteroid use.
In another study with relatively young and predominantly living donor–related transplant recipients, long-term results of RSWD were equivalently effective and safe compared with corticosteroid treatment.
This study adds:
The observational 5-year follow-up (FU) investigation of our large Harmony trial shows translation of the equivalent efficacy and improved safety aspects of RSWD compared with the corticosteroid treatment arm from the 1-year results to the 5-year period.
During 5-year FU, BPAR, death-censored graft loss as well as donor-specific antibody rates remained low and independent of RSWD. The reduced incidence of post-transplantation diabetes mellitus in RSWD patients during the 1-year study period was not compensated by later incidences during FU.
With all the limitations of a post-trial FU study, the study data suggest RSWD as an independent positive factor for patient survival, which may relate to the elderly recipient population with a long time on dialysis exposed to predominantly elderly deceased-related donors, all of which are risk factors for patient survival after renal transplantation.
Potential impact:
The long term results of the Harmony post-trial FU study demonstrate RSWD as a safe procedure and suggest even improved patient survival, supporting the RSWD approach to become at least a promising new alternative to the current corticosteroid-continuing gold standard treatment in an immunologically low-risk, elderly population of Caucasian kidney transplant recipients.
This remarkable result likely relates to this typical elderly recipient population at risk, which has become the standard population in most European countries. These results fit with the long-term results of another RSWD study [16] demonstrating equivalent efficacy and safety in a much younger, living donor–related renal transplant population.
Taken together, these data should secure and stimulate renal transplant centers to make use of the RSWD approach minimizing corticosteroid-induced side effects without losing long-term efficacy or safety for their patients.
INTRODUCTION
Corticosteroids are potent immunosuppressive agents that are still part of the current standard of immunosuppressive therapy after renal transplantation. Based on the extensive corticosteroid side effect profile such as accelerated cardiovascular risk factors (diabetes mellitus, hyperlipidemia, hypertension, weight gain), one-third of all kidney transplant recipients in the USA, but much fewer in Europe, have corticosteroids withdrawn mostly within the first week after transplantation [1].
Nevertheless, due to a scarcity of randomized controlled trials, the KDIGO recommendation for a rapid corticosteroid withdrawal in renal transplant recipients with low immunological risk is only based on evidence level 2B (2: suggested, B: moderate quality of evidence) [2]. Recent registry data are all influenced by a selection bias but suggest that steroid-withdrawal protocols lead to reduced post-transplantation diabetes mellitus (PTDM) rates as well as superior long-term graft and patient survival [3, 4]. Conversely, corticosteroid-free or -withdrawal regimens were frequently associated with increased T cell–mediated acute rejection rates [5, 6], while long-term data on the frequency of HLA-sensitization and humoral rejections are rare.
In the original investigator-initiated, randomized, multicenter Harmony trial, 587 kidney transplant recipients with low immunological risk profile were equally assigned to three different treatment arms to evaluate which of the two induction agents (basiliximab or rabbit ATG) was most efficacious and safe at permitting rapid steroid withdrawal (RSWD) in a prolonged release tacrolimus, mycophenolate mofetil (MMF) and prednisolone-based immunosuppressive therapy (control arm) within the first year after kidney transplantation [7]. Both induction therapies combined with RSWD were equivalent to permanent corticosteroid control patients for the prevention of biopsy-proven acute rejection (BPAR) within the first year after transplantation. Patient and death-censored graft survival rates as well as most safety parameters in the first year were excellent and similar in all groups independent of RSWD. In contrast, RSWD in both groups (24%/23%) markedly and equivalently reduced the incidence of PTDM compared with control patients (39%).
The 5-year post-trial follow-up (FU) data of the Harmony trial reported here were obtained in an observational manner to evaluate how these excellent short-term surrogate parameters will translate into hard long-term outcome parameters like patient death/graft loss or will be complicated by increased rates of immunosensitization and humoral rejections.
MATERIALS AND METHODS
Study design and medication
This investigator-initiated, observational post-trial FU study was implemented to compare the 3- and 5-year safety and efficacy of basiliximab- or rATG-induced RSWD in immunologically low-risk kidney transplant recipients receiving low-dose tacrolimus and MMF/mycophenolic acid (MPA). The study was approved by the institutional review boards of all sites participating and written informed consent had to be obtained again from patients participating in the post-trial FU study (FU study DRKS00005786).
The original Harmony trial (NCT 00724022) was an investigator-initiated 1-year study in 21 German transplant centers performed as prospective, randomized, open-label, multicenter study in three parallel study arms of adult renal transplant recipients.
In the post-trial FU study, all patients alive after the Harmony trial and recruited by centers participating in the FU investigation were eligible. Transplant recipients were followed in a non-interventional manner according to center standard. FU data were collected at 3- and 5-year visits for patients who consented to participate and covered clinical events that occurred from the second year onwards. In addition, deaths occurring after the original trial but before informed consent to FU investigation were included in the present analyses.
Medication, inclusion and exclusion criteria as well as study endpoints are described within the Supplementary material [8–12].
Statistical analysis
In this observational trial, all analyses were performed in a purely descriptive manner. As in the original trial, no imputation methods were applied for missing data caused by any reason. Time-to-event endpoints were analyzed for the whole intention-to-treat population with all events and censoring time points reported either in the original Harmony study or the FU period. Otherwise, analyses were based on the subset of intention-to-treat population participating in the FU trial. Categorical variables were summarized as counts and percentages, and continuous variables as means with standard deviations or median and interquartile range. Categorical data were analyzed by Fisher's exact test, and continuous variables by Kruskal–Wallis H Test or analysis of variance, as appropriate. Time-to-event endpoints, i.e. BPAR, graft loss or death, were calculated from date of transplantation to event or censoring time point. These endpoints were investigated by Kaplan–Meier method and tested by log-rank test. After checking the proportional hazard assumption, Cox regression analysis was used to investigate risk factors and estimation of hazard ratios (HR) for events. Adjusted models were obtained by forward selection of independent covariates. A significance level of 0.2 was used for inclusion of variables in the model.
If not otherwise stated, P-values for comparison of three arms are presented. By analogy with the original study, the impact of RSWD was investigated by comparing Arm A with Arms B and C. A P-value of <.05 was considered statistically significant. In this descriptive FU analysis, no adjustment of type I error for multiple comparison was planned as the confirmatory analysis was conducted in the original Harmony trial. Analyses were performed using SAS software 9.4. A FU for data collection was done whenever possible with patients who were prematurely eliminated from the original study.
RESULTS
Patient population characteristics
From originally 587 Harmony study participants, at the 1-year time point 144/133/126 patients in Groups A/B/C finished the full 12-month study period, respectively [7]. A total of 135/113/111 transplant recipients (including deceased participants) representative of the original Harmony study population in Groups A/B/C provided data for the observational FU analysis (Table 1). By year 5 after transplantation, 11 (6/3/2) recipients were lost to FU in the different groups, respectively (see flow chart in Supplementary data, Fig. S1).
Table 1:
Baseline characteristics and main outcome data at 12 months of Harmony FU patients.
| Variable | Category | Arm A: basiliximab/steroids | Arm B: basiliximab/RSWD | Arm C: rATG/RSWD | FU-ITTa | Total-Harmony |
|---|---|---|---|---|---|---|
| Number | Evaluable | 135 | 113 | 111 | 359 | 587 |
| Age (years) | Mean ± SD | 55.0 ± 11.0 | 55.0 ± 12.5 | 53.6 ± 11.8 | 54.6 ± 11.7 | 54.1 ± 12.2 |
| Male sex | % | 68 | 65 | 66 | 66 | 66 |
| Cause of end-stage renal disease | ||||||
| Hypertension/large vessel disease | % | 38 | 38 | 34 | 37 | 37 |
| Glomerulonephritis | % | 29 | 29 | 26 | 28 | 27 |
| Polycystic kidney disease | % | 22 | 18 | 20 | 20 | 19 |
| Diabetes | % | 12 | 8 | 10 | 10 | 11 |
| Interstitial nephritis | % | 7 | 11 | 6 | 8 | 7 |
| Type of donor | ||||||
| Cadaveric Do-type | % | 87 | 92 | 87 | 89 | 87 |
| Donor with expanded criteria | % | 42 | 41 | 39 | 41 | 44 |
| Donor age (years) | Mean ± SD | 53.5 ± 15.3 | 54.7 ± 14.9 | 50.9 ± 16.1 | 53.1 ± 15.5 | 54.0 ± 14.7 |
| Antigen mismatches: A/B/DR | Mean | 0.7/1.0/0.8 | 0.9/1.0/0.8 | 0.9/1.1/0.9 | 0.8/1.0/0.8 | 0.8/1.0/0.9 |
| No panel reactive antibodies before Tx | % | 84 | 90 | 89 | 88 | 88 |
| Previous transplants | % | 5 | 3 | 4 | 4 | 4 |
| Diabetes mellitus before NTx | % | 18 | 12 | 16 | 15 | 15 |
| CMV serologic high-risk status | % | 24 | 27 | 29 | 26 | 25 |
| Endpoints at 1 year | ||||||
| BPAR excl. borderline | 1 year rate and 95% CI | 11.1 (6.7–18.0) | 11.2 (6.5–18.9) | 11.2 (6.5–18.9) | 11.1 (8.2–15.0) | 10.8 (8.5–13.8) |
| Overall survival | 1 year rate and 95% CI | 91.3 (84.8–95.1%) | 96.2% (90.1–98.6) | 95.2 (88.9–98.0) | 94.1 (90.9–96.1) | 96.1 (94.1–97.5) |
| Death-censored graft survival | 1 year rate and 95% CI | 98.4 (93.9–99.6) | 99.1 (93.6–99.9) | 99.0 (93.2–99.9) | 98.8 (96.9–99.6) | 95.8 (93.6–97.2) |
| Graft loss or death | 1 year rate and 95% CI | 9.4 (5.4–16.0) | 4.7 (2.0–11.0) | 4.8 (2.0–11.1) | 6.5 (4.3–9.7) | 7.6 (5.6–10.2) |
| PTDM acc. ADA criteria | % | 38.7 | 20.0 | 21.5 | 27.3 | 28.8 |
| CMV infection | % | 22.2 | 18.6 | 25.2 | 22.0 | 20.4 |
| BKV infection | % | 10.4 | 5.3 | 13.5 | 9.7 | 10.4 |
| EBV infection | % | 3.7 | 1.8 | 5.4 | 3.6 | 2.6 |
Baseline characteristics described for participants of post-trial FU study. Hereby, non-survivors of original trial or before 3-year FU visit were included to avoid bias of characteristics by positive selection of survivors only.
rATG: rabbit ATG; ITT: intention-to-treat; Tx: transplant; CI: confidence interval; EBV: Epstein–Barr virus; Do-type: donor type; CMV: Cytomegalovirus; BKV: BK polyomavirus; PTDM: Post-transplantation diabetes mellitus; ADA: American Diabetes Association; BPAR: Biopsy-proven acute rejection; RSWD: Rapid steroid withdrawal.
Baseline characteristics of the three study arms excluding all patients lost to FU study were well balanced and representative of the original Harmony study, as indicated in Table 1.
Measurements of HLA-antibody activity was not part of the original Harmony trial and therefore not available at the 1-year time point. At the 1-year time point, 90%/16%/17% of patients of the original Harmony study were receiving corticosteroids in Groups A/B/C.
At the 3- and 5-year FU investigations, the different steroid withdrawal frequency was still statistically significant, but only a minority of 38.5% of Group A patients remained on corticosteroids, while in Group B only 22.8% and in Group C 21.9% of patients received corticosteroid therapy (P = .012; see Table 2). More than 80% of the study patients remained on tacrolimus and MMF/MPA immunosuppressive treatment at either FU time point, equally distributed in all treatment groups. Only 4 patients received azathioprine (1.3%), 12 patients cyclosporine (3.9%), 15 patients mTOR inhibitors (4.9%) and 3 patients belatacept (1.0%) therapy during FU period (see Table 2).
Table 2:
Safety endpoints at 5 years after transplantationa.
| Arm A: basiliximab/steroids | Arm B: basiliximab/RSWD | Arm C: rATG/RSWD | FU-ITTa | ||
|---|---|---|---|---|---|
| Variable | n = 113 | n = 101 | n = 98 | n = 312 | P-valuee |
| Infections | |||||
| Severe bacterial infections, hospitalization required | 29/112 (25.9) | 35/101 (34.7) | 14/98 (14.3) | 78/311 (25.1) | .004 |
| More than one severe bacterial infection | 14/112 (12.5) | 19/101 (18.8) | 8/98 (8.2) | 41/311 (13.2) | .086 |
| Invasive opportunistic infection | 4/112 (3.6) | 7/101 (6.9) | 8/98 (8.2) | 19/311 (6.1) | .366 |
| Any CMV infection | 9/112 (8.0) | 10/101 (9.9) | 14/98 (14.3) | 33/311 (10.6) | .348 |
| Any BKV infection | 3/112 (2.7) | 2/100 (2.0) | 5/98 (5.1) | 10/310 (3.2) | .476 |
| BKV viremia | 3/112 (2.7) | 2/100 (2.0) | 5/98 (5.1) | 10/310 (3.2) | .476 |
| BKV nephropathy | 1/112 (0.9) | 0/100 (0.0) | 3/98 (3.1) | 4/310 (1.3) | .164 |
| Malignancies | |||||
| All malignancies | 15/109 (13.8) | 11/97 (11.3) | 10/96 (10.4) | 36/302 (11.9) | .773 |
| Skin malignancy | 12/109 (11.0) | 9/97 (9.3) | 8/96 (8.3) | 29/302 (9.6) | .831 |
| Organ malignancy | 4/109 (3.7) | 2/97 (2.1) | 2/96 (2.1) | 8/302 (2.6) | .741 |
| PTLD | 0/109 (0.0) | 0/97 (0.0) | 0/96 (0.0) | 0/302 (0.0) | |
| Allograft function | |||||
| Creatinine clearance (mL/min) (Cockroft–Gault formula) | n = 91, 60.3 ± 22.2 | n = 88, 60.3 ± 24.6 | n = 83, 64.0 ± 26.8 | n = 262, 61.5 ± 24.5 | .529 |
| eGFR (mL/min/1.73 m2) (CKD-EPI formula) | n = 94, 48.7 ± 17.4 | n = 92, 48.1 ± 19.6 | n = 87, 52.8 ± 22.3 | n = 273, 49.8 ± 19.8 | .226 |
| Proteinuria/creatinine (mg/g) at 5 year visit (median and IQR) | n = 63, 112 [34–245] | n = 65, 117 [66–232] | n = 65, 87 [46–164] | n = 193, 108 [46–189] | .394 |
| Proteinuria ≥1000 mg/g creatinine | 7/82 (8.5) | 4/76 (5.3) | 6/77 (7.8) | 17/235 (7.2) | .739 |
| Albuminuria/creatinine (mg/g) at 5 year visit (median and IQR) | n = 58, 60 [15–120] | n = 55, 30 [13–102] | n = 60, 29 [13–135] | n = 173, 37 [14–117] | .446 |
| Albuminuria ≥300 mg/g creatinine | 8/70 (11.4) | 9/65 (13.8) | 12/70 (17.1) | 29/205 (14.1) | .614 |
| Anti-HLA antibodiesb | |||||
| Screening performed | 54/113 (47.8) | 44/101 (43.6) | 45/98 (45.9) | 143/312 (45.8) | |
| All anti-HLA antibodies | 7/54 (13.0) | 11/44 (25.0) | 15/45 (33.3) | 33/143 (23.1) | .049 |
| De novo Anti-HLA antibodies | 3/54 (5.6) | 6/44 (13.6) | 9/45 (20.0) | 18/143 (12.6) | .087 |
| Donor-specific anti-HLA antibodies | 4/54 (7.4) | 5/44 (11.4) | 5/45 (11.1) | 14/143 (9.8) | .825 |
| De novo DSA | 0/54 (0.0) | 3/44 (6.8) | 1/45 (2.2) | 4/143 (2.8) | .072 |
| Diverse | |||||
| Any cardio- or cerebrovascular event | 11/109 (9.7) | 14/97 (13.9) | 7/96 (7.1) | 32/302 (10.3) | .281 |
| Anemia | 14/110 (12.7) | 16/97 (16.5) | 16/96 (16.7) | 46/303 (15.2) | .671 |
| Anemia requiring ESA | 3/110 (2.7) | 10/97 (10.3) | 10/96 (10.4) | 23/303 (7.6) | .041 |
| Any fracture event | 3/109 (2.7) | 2/97 (2.0) | 3/96 (3.1) | 8/302 (2.6) | .910 |
| Weight, relative changec (mean ± SD) | 1.0 ± 0.13 | 1.0 ± 0.13 | 1.0 ± 0.11 | 1.0 ± 0.13 | .190 |
| BMI, relative changec (mean ± SD) | 1.0 ± 0.1 | 1.1 ± 0.1 | 1.0 ± 0.1 | 1.0 ± 0.1 | .203 |
| PTDMd | 5/58 (8.6%) | 7/76 (9.2%) | 4/67 (6.0%) | 16/201 (8.0%) | .802 |
| Current immunosuppressive medication | |||||
| Corticosteroids | 42/109 (38.5) | 23/101 (22.8) | 21/96 (21.9) | 86/306 (28.1) | .012 |
| Tacrolimus | 87/109 (79.8) | 85/101 (84.2) | 81/96 (84.4) | 253/306 (82.7) | .626 |
| MMF/MPA | 87/109 (79.8) | 82/101 (81.2) | 83/96 (86.5) | 252/306 (82.4) | .444 |
| mTOR inhibitors | 7/109 (6.4) | 5/101 (5.0) | 3/96 (3.1) | 15/306 (4.9) | .602 |
| Cyclosporine | 4/109 (3.7) | 3/101 (3.0) | 5/96 (5.1) | 12/306 (3.9) | .760 |
Safety endpoints during FU reported for patients who gave informed consent for FU study and did not experience death by the first FU visit. Data presented as absolute and relative frequencies, mean ± SD or median and IQR.
Anit-HLA antibodies were considered positive with detectability.
Computed as value at 5-year visit divided by value at baseline (value 1 means constant weight).
Among patients without diabetes mellitus at baseline and without PTDM during original Harmony study (203 patients in total; Arm A 59 patients, Arm B 76 patients, Arm C 68 patients).
P-values calculated for comparison of Arm A vs B vs C. Fisher's exact test, analysis of variance or Kruskal–Wallis H test, as appropriate.
rATG: rabbit ATG; ITT: intention-to-treat; eGFR: estimated glomerular filtration rate; CKD-EPI: Chronic Kidney Disease Epidemiology Collaboration; SD: standard deviation; IQR: interquartile range; RSWD: Rapid steroid withdrawal; CMV: Cytomegalovirus; BKV: BK polyomavirus; PTLD: Post-transplant lymphoproliferative disorder; CKD-EPI: Chronic Kidney Disease Epidemiology Collaboration; DSA: Donor-specific antibody; ESA: Erythropoiesis-stimulating agents; BMI: Body mass index; PTDM: Post-transplant diabetes mellitus; MMF/MPA: Mycophenolates.
Efficacy endpoints
From transplantation up to the 3/5-year visits of FU, the cumulative incidence of the primary endpoint BPAR (excluding or including borderline) was low and similar among the three study arms (Fig. 1A, Table 3). Severity of new acute rejections was also similar: in Group A, two Banff 1A and one acute antibody-mediated rejection (ABMR) grade III, in Group B three Banff 1A, and in Group C two ABMR grade II and one chronic active antibody-mediated rejection were noted. Only two steroid-resistant rejections (Banff grade IA/borderline) were documented during FU period, both of which occurred in Group B. From day of transplantation to year 5, a total of 99 patients with rejections including borderlines were equally distributed in the three treatment Groups A (36), B (31) and C (32). Hereby, 10 new borderline rejections in seven patients were counted during the FU period, while zero events were counted in Group A patients, three events in two Group B patients and seven events in five Group C patients.
Figure 1:
(A) BPAR. Arm A: induction therapy with basiliximab and immunosuppression with tacrolimus extended-release once daily, MMF and prednisolone. Arm B: induction therapy with basiliximab and immunosuppression with tacrolimus extended-release once daily, MMF and RSWD within the first week after transplantation. Arm C: induction therapy with rabbit ATG and immunosuppression with tacrolimus extended-release once daily, MMF and RSWD within the first week after transplantation. Borderline acute rejections were excluded from the analysis of BPAR. Randomization took place before kidney transplantation after receiving informed consent from the patient. The analysis includes all patients in the intention-to-treat population of the original Harmony trial (n = 587). For the Kaplan–Meier estimates, data of patients who completed the study before having an event were censored at the time of their last visit. (B) Patient survival. Arm A: induction therapy with basiliximab and immunosuppression with tacrolimus extended-release once daily, MMF and prednisolone. Arm B: induction therapy with basiliximab and immunosuppression with tacrolimus extended-release once daily, MMF and RSWD within the first week after transplantation. Arm C: induction therapy with rabbit ATG and immunosuppression with tacrolimus extended-release once daily, MMF and RSWD within the first week after transplantation. Randomization took place before kidney transplantation after receiving informed consent from the patient. The analysis includes all patients in the intention-to-treat population of the original Harmony trial (n = 587). For the Kaplan–Meier estimates, data of patients who completed the study before having an event were censored at the last day they are known to be alive. (C) Death-censored graft survival. Arm A: induction therapy with basiliximab and immunosuppression with tacrolimus extended-release once daily, MMF and prednisolone. Arm B: induction therapy with basiliximab and immunosuppression with tacrolimus extended-release once daily, MMF and RSWD within the first week after transplantation. Arm C: induction therapy with rabbit ATG and immunosuppression with tacrolimus extended-release once daily, MMF and RSWD within the first week after transplantation. Randomization took place before kidney transplantation, after receiving informed consent from the patient. The analysis includes all patients in the intention-to-treat population of the original Harmony trial (n = 587). For the Kaplan–Meier estimates, data of patients who died or completed the study before having an event (graft loss) were censored at the corresponding study day. (D) Patient and graft survival. Arm A: induction therapy with basiliximab and immunosuppression with tacrolimus extended-release once daily, MMF and prednisolone. Arm B: induction therapy with basiliximab and immunosuppression with tacrolimus extended-release once daily, MMF and RSWD within the first week after transplantation. Arm C: induction therapy with rabbit ATG and immunosuppression with tacrolimus extended-release once daily, MMF and RSWD within the first week after transplantation. Randomization took place before kidney transplantation, after receiving informed consent from the patient. The analysis includes all patients in the intention-to-treat population of the original Harmony trial (n = 587). For the Kaplan–Meier estimates, data of patients who completed the study before having an event (graft loss or death) were censored at the last day they are known to be alive.
Table 3:
Efficacy endpoints at 3 and 5 years after transplantation.
| 3 year ratesa | 5 year ratesa | |||||||
|---|---|---|---|---|---|---|---|---|
| Variable groups | Category | Arm A: basiliximab/steroids | Arm B: basiliximab/RSWD | Arm C: rATG/RSWD | Arm A: basiliximab/steroids | Arm B: basiliximab/RSWD | Arm C: rATG/RSWD | P-valueb |
| BPAR | % (N + new events)c | 14.7 (23 + 3) | 12.4 (20 + 1) | 13.2 (19 + 2) | 14.7 (26) | 14.5 (21 + 2) | 13.2 (21) | .932 |
| BPAR including borderlines | % (N + new events)c | 19.4 (34 + 2) | 17.8 (28 + 2) | 19.6 (27 + 5) | 19.4 (36) | 20.8 (30 + 3) | 20.7 (32 + 1) | .987 |
| Patient survival | % (N + new events) | 89.2 (11 + 7) | 95.1 (4 + 3) | 93.4 (6 + 3) | 84.7 (18 + 7) | 89.4 (7 + 6) | 90.4 (9 + 5) | .178 |
| Death-censored graft survival | % (N + new events) | 95.8 (8 + 0) | 96.6 (6 + 0) | 95.3 (8 + 0) | 94.9 (8 + 1) | 93.0 (6 + 3) | 91.4 (8 + 5) | .477 |
| Patient and graft survival | % (N + new events) | 85.8 (18 + 7) | 92.7 (10 + 2) | 89.5 (13 + 3) | 80.7 (25 + 8) | 84.0 (12 + 9) | 83.2 (16 + 10) | .448 |
Estimated by Kaplan–Meier method.
P-values of log-rank test calculated for comparing Arm A vs B vs C.
New events = number of patients with first BPAR event during corresponding FU period with or without counting borderline rejections.
rATG: rabbit ATG; RSWD: Rapid steroid withdrawal; BPAR: Biopsy-proven acute rejection.
In total, 52 deaths were observed from transplantation up to the full 5-year FU period, of which 25 occurred in Group A, 13 in Group B and 14 in Group C patients (Table 3, Fig. 1B, log-rank test for A vs B vs C P = .178). Median observation time was 4.7 years and similar in all study arms. In analogy with the original study, the impact of RSWD was investigated by comparing Arm A with Arms B and C. This numerical difference of RSWD groups vs Arm A almost reached statistical significance (log-rank test for A vs B + C P = .064) and relates to an overall survival rate of 89.2%/84.7% in Group A, 95.1%/89.4% in Group B, and 93.4%/90.4% in Group C after a 3- and 5-year FU period, respectively. Death-censored graft survival rates were similar in all three groups. From transplantation until 5 years of FU, 31 graft losses (9 in Group A, 9 in Group B, 13 in Group C) were observed, of which only 9 (1 in Group A, 3 in Group B, 5 in Group C) were new events during FU period (Table 3, Fig. 1C). Patient and graft survival rates for the three groups at the 5-year FU time point (33 events in Group A, 21 events in Group B, 26 events in Group C) were not different (Table 3, Fig. 1D).
Graft function was similar in all treatment groups at any time point (Table 2). Interestingly, the trend of an improved graft function in Arm C started early on during the first months after transplantation and remained at a comparable difference throughout all time points examined up to 5 years.
Safety endpoints
During the FU period, monitoring of anti-HLA antibodies was introduced gradually in clinical routine of the transplant programs. From year 1 on during the full FU period, any anti-HLA antibody screening was done in 45.8% (143/312) of patients, whereas 23.1% (33/143) tested positive. The detection of any or de novo anti-HLA antibody tended to be more frequent in the steroid withdrawal groups, although not statistically significant, compared with the control group (Table 2). The donor-specific occurrence of anti-HLA antibodies (DSA) was similar, with de novo occurrence particularly evident in study Arm B (basiliximab/RSWD) during the FU period (Table 2).
Metabolic profiles
In analogy with the original study, de novo occurrence of PTDM was defined according to the American Diabetes Association (ADA) criteria [11]. Nevertheless, no single oral glucose tolerance testing was performed during the FU period. A novel event of PTDM during FU was therefore based on former absence of diabetes/PTDM diagnosis followed by center diagnosis/judgement of new PTDM or diabetic blood values including fasting glucose and HbA1c measurements at years 3 or 5. Altogether, only 16 new cases of PTDM were seen during the FU period from years 1 to 5. Five cases were observed in Group A, seven in Group B and four in Group C, confirming that the significant reduction of PTDM with rapid steroid withdrawal (RSWD) within the first year after transplantation is not at all compensated by later PTDM events up to 5 years post-transplantation (Table 2).
Most of the other metabolic profile parameters such as blood pressure, lipids or weight/body mass index (BMI) did not differ substantially among the three study arms as described further in Supplementary results.
Cardiovascular disease, bone disease
The FU incidence of any cardio- or cerebrovascular disease event was around 10% and equally distributed in all study arms (Table 2).
FU incidence of bone fractures reported was low (2.6%) and similar in all study arms (Table 2).
Infections, anemia and cancer
During FU period from year 1 to 5, the overall incidence of any serious bacterial infection requiring hospitalization was decreased in Arm C compared with Arms A and B (Table 2) with similar differences at the first and second 2 years of FU. In addition, Arm C also showed a significant reduction in the incidences of repetitive bacterial infections. The incidence of any invasive opportunistic infection was similar in the three groups as was the overall incidence of cytomegalovirus (CMV) infection (Table 2). During the full 5-year FU period, the detected incidence of new BK virus (BKV) infection according to either polymerase chain reaction measurement or histological criteria for BKV nephropathy were similar in all groups (Table 2).
While within the first year of the Harmony trial, the frequency of anemia was significantly higher after RSWD compared with control patients, during the 3- and 5-year FU time, anemia rates were low (around 15%) and similar in all groups (Table 2). In contrast, anemia requiring erythropoiesis-stimulating agent (ESA) therapy was still significantly more frequent in both steroid withdrawal groups compared with the control group during 5-year FU (Table 2).
Cancer development occurred at similar rates in all groups during the FU study. During FU, 36 patients encountered 37 new malignancies (15 in Group A, 11 in Group B, 10 in Group C) since year 1, of which 29 were skin and 8 were solid organ-related (Table 2). No post-transplantation lymphoproliferative disease (PTLD) occurred in any patient during the FU period (year 1–5).
Multivariable analysis of patient survival
To determine whether certain factors such as age, sex, RSWD, induction therapy, donor with expanded criteria, diabetes mellitus, cardiovascular disease, arterial hypertension, anemia, elevated cholesterol and hyperlipidemia at inclusion exert an independent impact on patient survival/death, a Cox regression analysis was performed evaluating the complete study period. No relevant collinearity was revealed among the independent risk factors. Hereby, affiliation of transplant recipients to RSWD Groups B + C compared with continued corticosteroid Group A recipients revealed as an independent protective factor for patient death (adjusted HR 0.554, 95% confidence interval 0.314–0.976; P = .041), but not the type of induction therapy. Multivariable analysis confirmed that recipient age, time on dialysis and diabetes mellitus, as well as cardiovascular disease at inclusion were independent risk factors for patient death (Table 4).
Table 4:
Cox regression analysis for time to death from any cause.
| Unadjusted models | Adjusted modela | ||||||
|---|---|---|---|---|---|---|---|
| Risk factor | HR | 95% CI | P-value | HR | 95% CI | P-value | |
| Age (per 10 years) | 3.143 | (2.163–4.567) | <.001 | 3.296 | (2.094–5.187) | <.001 | |
| Sex | Female vs male | 1.099 | (0.617–1.958) | .750 | |||
| Rapid steroid withdrawal | Yes vs no | 0.575 | (0.330–1.003) | .051 | 0.554 | (0.314–0.976) | .041 |
| Induction therapy | rATG vs basilix. | 0.773 | (0.411–1.456) | .426 | |||
| Time on dialysis (per 6 month) | 0.999 | (0.958–1.043) | .981 | 1.047 | (1.000–1.095) | .048 | |
| Donor with expanded criteria | Yes vs no | 2.757 | (1.529–4.970) | <.001 | |||
| Diabetes mellitus at transplant | Yes vs no | 4.818 | (2.731–8.498) | <.001 | 2.968 | (1.617–5.447) | <.001 |
| Cardiovascular disease at transplantb | Yes vs no | 3.391 | (1.922–5.981) | <.001 | 1.876 | (1.028–3.426) | .041 |
| Renal anemia at transplantb | Yes vs no | 0.913 | (0.484–1.720) | .778 | |||
| Arterial Hypertension at transplantb | Yes vs no | 0.994 | (0.358–2.764) | .991 | 0.461 | (0.159–1.339) | .155 |
| Hyperlipidemia at transplantb | Yes vs no | 1.792 | (1.027–3.129) | .040 | |||
| Elevated cholesterol at transplantb | Yes vs no | 1.333 | (0.746–2.381) | .332 | |||
| Elevated triglycerides at transplantb | Yes vs no | 1.363 | (0.749–2.482) | .311 | |||
Effect selection by forward selection. Significance level for entering a predictor into the model is 0.20.
According to centre definition and local laboratory references.
DISCUSSION
This 5-year FU data of the randomized Harmony study comparing induction with rabbit ATG or interleukin-2 receptor antibody treatment describes one of the largest prospectively collected dataset to achieve RSWD in an immunological low-risk kidney transplant population with modern immunosuppression (low-dose once daily tacrolimus and MMF). At the low-dose of rabbit ATG used, this intervention did not show superiority over basiliximab induction for the prevention of BPAR after RSWD within 1 year after renal transplantation [7]. The original 1-year trial also demonstrated that RSWD generally could be safely performed with either induction agent without compromising efficacy with a tacrolimus/MMF-based regimen. Further, it leads to an improved safety profile, especially by reducing the incidence of the important cardiovascular risk factor PTDM by 40%.
The Harmony FU study demonstrates that 1-year trial results are translated to the 5-year FU period confirming that steroid-free therapy remains safe and efficient over the course of 5 years after kidney transplantation. To interpret the 1- to 5-year Harmony FU study results, it is important to keep in mind that this study comprised an almost 100% Caucasian, elderly recipient population (mean by 55 years) with long dialysis vintage. Additionally, these recipients received in 87% of cases deceased and elderly donor kidneys (average donor age 53 years), almost 50% of whom met the expanded criteria. During the FU period, about 80% of transplant recipients remained on a tacrolimus/MMF/MPA combination therapy and about 80% in the RSWD arms remained corticosteroid-free. In the steroid control group, the percentage of steroid-free transplant recipients increased from 10% by year 1 to 63% of patients by year 3/5, indicating partially a late steroid withdrawal (LSWD) approach in the majority of FU patients.
During the complete FU period, only very few BPARs were diagnosed via indication biopsies in all treatment groups demonstrating great efficacy of a once daily tacrolimus and MMF/MPA-based immunosuppressive therapy independently of ongoing corticosteroids or RSWD. In addition, considering that two-thirds of all control patients were weaned off corticosteroids between years 1 and 3, it demonstrates that the LSWD approach was also not accompanied by a considerable risk for acute rejections as indicated by earlier results [13].
Multivariate analysis identified RSWD as an independent positive factor for patient survival within our study cohort. Nevertheless, due to the limitations of this study given its observational, post-trial FU design, this survival result should not be interpretated as RSWD superiority compared with corticosteroid use, but at least secure non-inferiority. The overall patient and death-censored graft survival rate (equivalent in all groups) within this study is excellent, especially when the elderly recipient and donor population with a very long waiting time on dialysis is being considered. SWD regimens have shown reduced cardiovascular as well as infectious events, and improved survival when large registry data sets were examined [4]. While large registry data are always complicated by selection bias, no randomized SWD study so far was large and long enough to show differences in survival. Although we could not find any evidence of selection bias in the present study, it can never be ruled out with certainty. Additionally, this study was not designed to detect a difference in graft or patient survival. In the largest RSWD study to date by Woodle and coworkers [14–16], RSWD appeared to be equivalently effective and safe to corticosteroid treatment. Five-year survival was around 94% and even after 10 years independent of SWD. In comparison with the aforementioned study [14–16], our patient population is characterized by a 9 years older mean age and a mean time on dialysis before transplantation of 64 ± 42 months, which is markedly higher than the usual times in the USA. In our cohort, multivariate analysis also identified recipient age, time on dialysis, diabetes mellitus and cardiovascular disease at inclusion, as well known independent risk factors for patient death, supporting the value of the new factor RSWD. Additional differences relate to the donor pool, since in their US study 57% of transplantations were living donor–related but only 13% in the Harmony trial (>40% marginal donor organs). These substantial differences between the two study populations may also be critical for the different results regarding the influence of RSWD on the relevant cardiovascular risk factor PTDM [16]. While RSWD had some limited beneficial influence on PTDM in a young, predominantly living donor–related transplant population [14–16], in the Harmony trial incidence of PTDM at the 1-year time point was generally higher and almost bisected by RSWD [7]. Irrespective of a substantial fraction of patients experiencing an LSWD as in our control patient group, this marked reduction of PTDM frequency via RSWD was not countervailed by later events suggesting that an early time point of steroid withdrawal is critical for beneficial effects on survival or PTDM in an elderly recipient population.
Safety with respect to CMV, BKV as well as invasive opportunistic infections did not differ significantly in all three groups during FU. Unexpectedly [17–19], in the rabbit ATG Group C, the incidence of any serious bacterial infection requiring hospitalization or repetitive bacterial infections during FU was reduced by about 50%. In comparison with many other studies using rabbit ATG, the very low cumulative mean dose of 4.6 mg/kg body weight may have played a role for this beneficial effect. Since the basiliximab induction Group B did not show this marked reduction in severe bacterial infections compared with control patients, no relation to the RSWD management can be established. Additionally, no PTLD and no difference regarding the development of cancers was noted in all treatment arms during the 5-year FU period.
In the original 1-year study, both RSWD groups were associated with a significantly higher incidence of anemia and ESA-requiring anemia [7]. Surprisingly, the ESA-requiring anemia effect at a much lower level was still translated into the FU period. Despite LSWD until year 3 in more than 60% of all control patients, this effect is most likely still attributable to corticosteroids, since kidney function during FU was not different in all groups.
Since corticosteroids are potent immunosuppressive agents, one major concern of corticosteroid withdrawal compared with corticosteroid continuation is the possibility of increased immunosensitization and its consequences, but no firm conclusions can be drawn from the transplant studies available so far [20]. According to our 5-year FU results, the frequency of neither rejection-related graft losses nor chronic cellular/humoral rejections nor donor-specific antibody formation increased in the RSWD groups. Nevertheless, some higher immunosensitization attributable to RSWD cannot be excluded, since the frequency of total or de novo anti-HLA antibody formation or borderline rejections tended to be increased (at a low level) compared with the control group but remained without any consequences for the trial outcome within the study period.
Our study has several limitations. The observational manner of the FU study with changes in medication according to center standard and patient course may introduce some bias and limits the equality of observation, but relates to real life treatment results. For interpretation of these 5-year FU study results, the frequent LSWD beyond year 1 in the control patient group needs to be considered. Measurement of anti-HLA antibody formation was not yet regularly established at every transplant center and was only available in up to half of patients. The lost to FU of patients from the original 1-year study potentially contributing to some selection bias relates to the fact that this FU study was not planned from the beginning due to financial restrictions. We also interpret our largely beneficial results of RSWD being restricted to an immunologically low-risk, elderly population of Caucasian recipients.
In conclusion, the 5-year FU period of our investigator-initiated Harmony trial confirms the excellent efficacy and beneficial safety aspects of RSWD under modern tacrolimus/MMF/MPA immunosuppressive therapy as already seen in the original 1-year study. A steroid-free therapy remains safe, efficient and in regard to survival at least comparable to a steroid-containing therapy over a course of 5 years after kidney transplantation. With all the limitations given in an observational 5-year post-trial FU study, the results shown here support the RSWD approach to be more frequently used as an equivalent alternative to the current mostly corticosteroid-continuing therapy in an immunologically low-risk, elderly population of Caucasian kidney transplant recipients.
Supplementary Material
ACKNOWLEDGEMENTS
O.W. is supported by an unrestricted grant of the Rudolf-Ackermann-Stiftung (Stiftung für Klinische Infektiologie).
Contributor Information
Julian Stumpf, University Hospital Carl Gustav Carus at the Technische Universität Dresden, Department of Internal Medicine III, Division of Nephrology, Dresden, Germany.
Oliver Thomusch, Albert-Ludwigs University Freiburg, Department of General Surgery, Freiburg, Germany.
Mirian Opgenoorth, University Hospital Erlangen, Friedrich-Alexander University Erlangen-Nürnberg, Department Nephrology and Hypertension, Erlangen, Germany.
Michael Wiesener, University Hospital Erlangen, Friedrich-Alexander University Erlangen-Nürnberg, Department Nephrology and Hypertension, Erlangen, Germany.
Andreas Pascher, University Hospital of Münster, Westfälische Wilhelms-University Münster Department of General, Visceral and Transplant Surgery, and Charité-Universitaetsmedizin Berlin, Campus Virchow/Mitte, Department of Surgery, Berlin, Germany.
Rainer Peter Woitas, University Hospital of Bonn, Department of Internal Medicine I, Division of Nephrology, Bonn, Germany.
Barbara Suwelack, University Hospital of Münster, Westfälische Wilhelms-University Münster, Department of Internal Medicine D, Transplantnephrology, Münster, Germany.
Markus Rentsch, University Hospital of Großhadern Munich, Ludwig-Maximilian University Munich, Munich, Germany.
Oliver Witzke, University Hospital Essen, University Duisburg-Essen, Department of Infectious Diseases, West German Centre of Infectious Diseases, Essen, Germany.
Thomas Rath, Westpfalz Klinikum, Department of Nephrology, Kaiserslautern, Germany.
Bernhard Banas, University Hospital Regensburg, Division of Nephrology, Regensburg, Germany.
Urs Benck, Medical Faculty Mannheim, Heidelberg University, Department of Medicine V, Mannheim, Germany.
Claudia Sommerer, University Hospital Heidelberg, Department of Nephrology, Heidelberg, Germany.
Christine Kurschat, Faculty of Medicine and University Hospital Cologne, Department II of Internal Medicine and Center for Molecular Medicine Cologne, Cologne, Germany.
Kai Lopau, University Hospital, Julius-Maximilians-University of Wuerzburg, Würzburg, Germany.
Julia Weinmann-Menke, Medical Center of the Johannes Gutenberg University of Mainz, Mainz, Germany.
Bernd Jaenigen, Albert-Ludwigs University Freiburg, Department of General Surgery, Freiburg, Germany.
Evelyn Trips, Coordination Centre for Clinical Trials, Faculty of Medicine Carl Gustav Carus, Technische Universität Dresden, Dresden, Germany.
Christian Hugo, University Hospital Carl Gustav Carus at the Technische Universität Dresden, Department of Internal Medicine III, Division of Nephrology, Dresden, Germany.
FUNDING
The Harmony FU trial had been financially supported by Astellas and Sanofi.
CONFLICT OF INTEREST STATEMENT
The following authors of this manuscript have conflicts of interest do disclose as described by the American Journal of Transplantation: C.H. received grant support, consulting fees and lecture fees from Astellas, Sanofi, Genzyme, Roche Pharma, Novartis and Wyeth/Pfizer. C.K. received consulting fees and lecture fees from Astellas, Chiesi, Novartis and Sanofi-Genzyme. J.W.-M. received grant support, consulting fees and lecture fees from Vifor, GSK, Novartis, AstraZeneca and Boehringer Ingelheim. O.W. has received research grants for clinical studies, speaker's fees, honoraria and travel expenses from Amgen, Alexion, Astellas, Basilea, Biotest, Bristol-Myers Squibb, Correvio, Chiesi, Gilead, Hexal, Janssen, Dr F. Köhler Chemie, MSD, Novartis, Roche, Pfizer, Sanofi, Takeda, TEVA and UCB. B.B. received grant support, consulting fees and lecture fees from Astellas, Bayer, Chiesi and Novartis. J.S. received lecture fees from Astellas, Boehringer-Ingelheim and Fresenius Medical Care.
All other authors of this manuscript have no conflicts of interest to be disclosed.
DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.
REFERENCES
- 1. Annual Data Report—USRDS.org. The 2015 Annual Data Report. Available at: http://www.usrds.org/adr.aspx (15 June 2016, date last accessed).
- 2. Kasiske BL, Zeier MG, Chapman JRet al. KDIGO clinical practice guideline for the care of kidney transplant recipients: a summary. Kidney Int 2010;77:299–311. 10.1038/ki.2009.377 [DOI] [PubMed] [Google Scholar]
- 3. Opelz G, Döhler B, Laux G. Long-term prospective study of steroid withdrawal in kidney and heart transplant recipients. Am J Transplant 2005;5:720–8. 10.1111/j.1600-6143.2004.00765.x [DOI] [PubMed] [Google Scholar]
- 4. Opelz G, Döhler B. Association between steroid dose and death with a functioning graft after kidney transplantation. Am J Transplant 2013;13:2096–105. 10.1111/ajt.12313 [DOI] [PubMed] [Google Scholar]
- 5. Vincenti F, Schena FP, Paraskevas Set al. A randomized, multicenter study of steroid avoidance, early steroid withdrawal or standard steroid therapy in kidney transplant recipients. Am J Transplant 2008;8:307–16. 10.1111/j.1600-6143.2007.02057.x [DOI] [PubMed] [Google Scholar]
- 6. Knight SR, Morris PJ. Steroid avoidance or withdrawal after renal transplantation increases the risk of acute rejection but decreases cardiovascular risk. A meta-analysis. Transplantation 2010;89:1–14. 10.1097/TP.0b013e3181c518cc [DOI] [PubMed] [Google Scholar]
- 7. Thomusch O, Jaenigen B, Rentsch Met al. Rabbit-ATG or basiliximab induction for rapid steroid withdrawal after renal transplantation (Harmony): an open-label, multicentre, randomised controlled trial. Lancet 2016;388:3006–16. 10.1016/S0140-6736(16)32187-0 [DOI] [PubMed] [Google Scholar]
- 8. Loupy A, Haas M, Solez Ket al. The Banff 2015 Kidney Meeting report: current challenges in rejection classification and prospects for adopting molecular pathology. Am J Transplant 2017;17:28–41. 10.1111/ajt.14107 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Cockcroft DW, Gault MH. Prediction of creatinine clearance from serum creatinine. Nephron 1976;16:31–41. 10.1159/000180580 [DOI] [PubMed] [Google Scholar]
- 10. Levey AS, Stevens LA, Schmid CHet al. A new equation to estimate glomerular filtration rate. Ann Intern Med 2009;150:604–12. 10.7326/0003-4819-150-9-200905050-00006 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Sharif A, Hecking M, de Vries APJet al. Proceedings from an international consensus meeting on posttransplantation diabetes mellitus: recommendations and future directions. Am J Transplant 2014;14:1992–2000. 10.1111/ajt.12850 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Alloway RR, Woodle ES, Abramowicz Det al. Rabbit antithymocyte globulin for the prevention of acute rejection in kidney transplantation. Am J Tranplant 2019;19:2252–61. 10.1111/ajt.15342 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Pascual J, Quereda C, Zamora Jet al. Steroid withdrawal in renal transplant patients on triple therapy with a calcineurin inhibitor and mycophenolate mofetil: a meta-analysis of randomized, controlled trials. Transplantation 2004;78:1548–56. 10.1097/01.TP.0000140969.43761.1F [DOI] [PubMed] [Google Scholar]
- 14. Woodle ES, First MR, Pirsch Jet al. A prospective, randomized, double-blind, placebo-controlled multicenter trial comparing early (7 day) corticosteroid cessation versus long-term, low dose corticosteroid therapy. Ann Surg 2008;248:564–77. 10.1097/SLA.0b013e318187d1da [DOI] [PubMed] [Google Scholar]
- 15. Pirsch JD, Henning AK, First MRet al. New-onset diabetes after transplantation: results from a double-blind early corticosteroid withdrawal trial. Am J Transplant 2015;15:1982–90. 10.1111/ajt.13247 [DOI] [PubMed] [Google Scholar]
- 16. Woodle ES, Gill JS, Clark Set al. Early corticosteroid cessation vs long-term corticosteroid therapy in kidney transplant recipients: long-term outcomes of a randomized clinical trial. JAMA Surg 2021;156:307–14. 10.1001/jamasurg.2020.6929 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Gurk-Turner C, Airee R, Philosophe Bet al. Thymoglobulin dose optimization for indication therapy in high risk kidney transplant recipients. Transplantation 2008;85:1425–30. 10.1097/TP.0b013e31816dd596 [DOI] [PubMed] [Google Scholar]
- 18. Kho MML, Bouvy AP, Cadogan Met al. The effect of low and ultra-low dosages thymoglobulin on peripheral T, B, and NK cells in kidney transplant recipients. Transpl Immunol 2012;26:186–90. 10.1016/j.trim.2012.02.003 [DOI] [PubMed] [Google Scholar]
- 19. Dadhania D, Snopkowski C, Ding Ret al. Epidemiology of BK virus in renal allograft recipients: identification of steroid maintenance therapy and rabbit anti-human thymocyte globulin induction as independent risk factors for BKV replication. Transplantation 2008;86:521–8. 10.1097/TP.0b013e31817c6447 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. O'Leary JG, Samaniego M, Crespo Barrio Met al. The influence of immunosuppressive agents on the risk of de novo donor-specific HLA antibody production in solid organ transplant recipients. Transplantation 2016;100:39–53. 10.1097/TP.0000000000000869 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.