Objectives: Posttransplant anemia affects 30% to 45% of kidney transplant recipients and is associated with increased morbidity. However, there is lack of evidence about safe hemoglobin levels after erythropoietin treatment. Studies are needed to better understand the potential benefits and risks, as well as to define safe target hemoglobin ranges in these patients.
Materials and Methods: In this single-center exploratory, open-label randomized controlled trial, kidney trans-plant recipients with anemia 3 months posttransplant were either treated with epoetin beta to a hemoglobin target level of 11.5 to 13.5 g/dL (n = 28) or given no treatment (n = 27). Treatment effects on graft function and health quality of life were assessed.
Results: After 2 years, hemoglobin concentrations were significantly higher in the epoetin beta treatment group than in the no treatment group (12.3 ± 0.18 vs 9.99 ± 0.22 g/dL; P < .0001). Estimated glomerular filtration rate, calculated by Modified Diet in Renal Disease 7, declined by 1.7 mL/min (interquartile range, -6 to 4.24) in the epoetin treatment group and by 4.16 mL/min (interquartile range, -12.42 to 2.78) in the no treatment group (P = .32). Rate of progression, determined by estimated glomerular filtration rate slope, was not significantly different between groups (-0.09 ± 0.1 vs -0.12 ± 0.15 mL/min for treated vs not treated; P = .78). Moreover, we observed no significant differences in proteinuria and blood pressure. Treated patients had greater improvements in the vitality and mental health domains of the Medical Outcomes Short Form Health Survey quality of life scores.
Conclusions: Treatment of anemia in kidney transplant recipients to a hemoglobin level of 11.5 to 13.5 g/dL with erythropoietin improves some quality of life scores. The treatment was safe and not associated with adverse outcomes. There were no changes in rate of decline of graft function.
Key words : Hemoglobin, Mental health, Renal transplant
The incidence of posttransplant anemia is high in kidney transplant recipients (KTR), with 30% to 40% of patients experiencing anemia at any one time.1-5 In addition to the traditional causes of anemia in chronic kidney disease (CKD), KTRs are exposed to other factors that may exacerbate or cause anemia, including immunosuppressive medication, higher risks of infection, and prolonged inflammatory milieu.
In both CKD and KTR patients, untreated anemia is associated with poor quality of life, left ventricular hypertrophy, and adverse cardiac outcomes.6-9 Specifically, in KTRs, a hemoglobin level of < 11 g/dL is independently associated with graft failure and mortality,10 and anemia per se negatively impacts quality of life with increased fatigue and reduced cognitive function and exercise capacity.3 Despite these negative effects, recognition and treatment of anemia in KTRs are often ignored.4 In addition, randomized trials have shown that treatment of anemia in CKD patients resulted in benefits in rate of progression of kidney failure.9,11,12 However, similar evidence in KTRs remains limited and inconclusive.
In this open-label randomized controlled trial, our aim was to determine whether treatment of anemia in KTRs to target hemoglobin levels of 11.5 to 13.5 g/dL reduced graft dysfunction, as shown by reduced decline in estimated glomerular filtration rate (eGFR) and reduced proteinuria. Secondary outcomes included effects of treatment on blood pressure, left ventricular hypertrophy, and quality of life.
Materials and Methods
This was an open-label randomized controlled trial performed in a single kidney transplant center in the United Kingdom. Ethical approval was obtained from the East London and City Research Ethics Council. The trial was approved by the Medicines Health Regulatory Authority (06/Q0603/126, EudRact 2006-003502-26) and was registered with the International Standardised Randomized Controlled Trial database (ISRCTN41687085). The trial was sponsored by Barts and the London NHS Trust (now Barts Health). The protocol conformed to the guidelines of the 1975 Helsinki Declaration. All trial participants provided fully informed and written consent.
All KTRs between 18 and 80 years old who received transplant procedures more than 3 months before start of study were screened. Patients meeting the following criteria were enrolled: hemoglobin level between 9 and 11.5 g/dL, iron replete (ferritin ≥ 100 mg/dL, transferrin saturation ≥ 20 %), and not already receiving an erythropoiesis-stimulating agent.
Eligible patients were randomized in a simple 1:1 ratio through simple computer randomization to 1 of 2 treatment groups: the epoetin beta (EB) treatment group, in which patients were treated with subcutaneous epoetin beta (NeoRecormon; Roche Registration Limited, Welwyn Garden City, UK) to achieve a target hemoglobin level of 11.5 to 13.5 g/dL, or the no treatment (NT) group. Rescue treatment with epoetin beta was initiated if hemoglobin levels fell below 9 g/dL in the NT group. Treatment was initiated at a starting dose of 50 U/kg once per week using prefilled syringes. The dose was titrated gradually to maintain the hemoglobin level in the target range. Clinical trial visits were every 4 weeks for the first 2 months and then every 2 months for 4 months and 3 monthly thereafter. Additional trial visits were conducted if required. Patients were followed up for a maximum of 24 months unless they withdrew from the study or were lost to follow-up. Intravenous iron (Cosmofer, Pharmacosmos UK Limited, Reading, UK; 100 mg/wk for 3 wk) was administered if patients became iron deficient. Antihypertensive medication was used to maintain a target blood pressure of < 140/90 mm Hg.
The primary endpoint was the effect of recombinant human erythropoietin (rHuEPO) treatment on graft function in KTRs as measured by change of kidney graft function using the Modified Diet in Renal Disease 7, estimation of glomerular filtration rate (eGFR), and changes in proteinuria measured by protein-to-creatinine ratio. Secondary endpoints included the effect of rHuEPO treatment on blood pressure and quality of life (assessed by the Medical Outcomes Short Form Health Survey; SF-36) and changes in left ventricular mass index (measured by transthoracic echocardiogram).
Retrospective data collected at our center showed that incidence of progression (decline of eGFR > 2 mL/min/year) in anemic KTRs was 60%. However, when treated with rHuEPO, this rate declined to 25%. Our study was designed on an intention to treat principle. We calculated that a sample size of 30 patients in each group would show a 35% reduction in the incidence of progression (> 2 mL/min/y) with 95% confidence and a 25% dropout rate. We calculated that a sample size of 60 patients in each group would show a 25% decrease in the incidence of progression (> 2 mL/min/y) with 99% confidence and a 25% dropout rate. Quantitative data were analyzed using GraphPad Prism 5.03 (GraphPad Software, San Diego, CA, USA), and parametric data were analyzed with t tests (mean ± standard error of the mean). Nonparametric data were analyzed with Mann-Whitney tests (median plus interquartile range [IQR]). Frequency was measured by chi-square test (number and percentage). The SF-36 surveys (version 2) were scored using the Quality Metric Health Outcomes Scoring Software 2.0 (QualityMetric, Lincoln, RI, USA).
From 852 screened patients, 55 were enrolled into the study. Of these, 28 patients were randomized to the EB group and 27 patients were randomized to the NT group (Figure 1). We observed no significant dif-ferences in age, sex, and ethnicity between the 2 groups and no significant difference in cause of CKD. Age of the current graft was similar, and there were comparable numbers of previous transplants and episodes of rejection and delayed graft function (Table 1). We also observed comparable height, weight, and body mass index (BMI) between groups (Table 2). Blood pressure was similar in both groups. Baseline hematologic and biochemistry parameters, including urine protein-to-creatinine ratio, were similar between groups (Table 3). Median follow-up was 23.34 months (IQR, 13.24-26.45) in the NT group and 23.12 months (IQR, 17.34-25.72) in the EB group. Five patients (9%) were withdrawn from the study. All 5 patients were in the NT group (18%); these patients were withdrawn for the following reasons: violation of study protocol by primary physician (n = 2, 7%), noncompliance (n = 2, 7%), and pro-longed hospital stay (n = 1, 3.5%).
At baseline, the groups were on similar types of medication (Table 2), receiving a similar number of antihypertensive medications and a similar dose of oral ferrous sulphate. We observed no differences in percentage of participants on an angiotensin-converting enzyme inhibitor or angiotensin receptor blocker. However, the NT group was receiving a higher dose of mycophenolate mofetil (1667 ± 145.7 vs 1235 ± 114.4 mg in the NT vs EB group; P = .02).
The baseline mean hemoglobin levels were 10.42 ± 0.12 g/dL and 10.37 ± 0.12 g/dL in the EB and NT groups, respectively (P = .81). Levels rose to 12.30 ± 0.18 g/dL in the EB group but decreased to 9.99 ± 0.22 g/dL (P < .001) in the NT group by end of study (Figure 2). The mean EB dose in the EB group was 3607 ± 137 U/week at initiation of study and 4643 ± 137 U/week at end of study.
Two patients (7%) required initiation of EB treatment in the NT group. Eleven patients (39%) in the EB group and 7 patients (26%) in the NT group required iron sup-plementation by end of study.
Renal function and estimated glomerular filtration rate (primary outcomes)
Baseline eGFR using Modified Diet in Renal Disease 7 was 30.9 ± 1.93 versus 34.7 ± 2.7 mL/min in the EB versus NT groups (P = .34). At end of study, eGFR results were similar in both groups (27.1 ± 2.1 vs 31.5 ± 5.2 mL/min for EB vs NT groups; P = .31) (Figure 3). We observed no differences in absolute change in eGFR different from baseline (-1.7 [IQR, -6 to 4.24] mL/min vs -4.2 [IQR, -12.42 to 2.78] mL/min for EB vs NT groups; P = .31). Rate of eGFR decline was also not different between the groups (-0.2 ± 0.08 vs -0.09 ± 0.13 mL/min for EB vs NT, respectively; P = .44).
At 1 year, the percentage of patients with pro-gression (decline in eGFR > 2 mL/min) was similar in both groups (45% in the EB group and 44% in the NT group; not significant). Similarly, at 2 years, the percentage of patients with progression was similar in both groups (39% vs 33% for EB vs NT groups; not significant). Two patients in the NT and none in the EB group progressed to end-stage renal disease requiring dialysis.
Proteinuria (primary outcome)
We observed no differences in urine protein-to-creatinine ratio at the beginning (18.5 [IQR, 9.3-33.5] mg/mmol vs 13.4 [IQR, 6-45] mg/mmol; P = 0.35) or end (33.9 [IQR, 11.3-70.3] mg/mmol vs 40 [IQR, 10.25-49] mg/mmol; P = .64) of the study in the EB and NT groups, respectively.
Blood pressure (secondary outcome)
Blood pressure remained stable in both groups (Figure 4). However, by the end of the study, more patients in the EB group were on 2 or more antihypertensive agents versus that shown in the NT group (2.5 ± 0.23 vs 1.9 ± 0.27; P = .07).
Quality of life (secondary outcome)
We observed significant improvements in quality of life in the vitality domain of the Medical Outcomes Short Form Health Survey (SF-36) in the EB group versus that shown in the NT group (6.25 [IQR, 3.12-12.5] vs 3.12 [IQR, -3.1 to 6.24] in the EB vs NT group, respectively; P = .2). There was no statistical differences in the other domains measured (Figure 5A). Significantly more patients also achieved a Minimal Clinically Important Difference in the EB group versus results shown in the NT group (50% vs 16% for EB vs NT group; P = .04) in the mental health domain (Figure 5B).
Adverse events (secondary outcome)
Five severe adverse events were reported in the study: 3 in the NT group (hospital admission for bowel obstruction, upper gastrointestinal bleed, and investigation of visual loss) and 2 in the EB group (hospital admission for severe anemia secondary to Parvovirus infection and hospital admission for investigation and treatment of bowel cancer). One patient developed headaches related to increased blood pressure, which responded to antihypertensive medication and lower EB doses. Two patients (7%) died in the EB group. The causes of death were severe pneumonia and newly diagnosed bowel cancer. Both deaths occurred within 6 months of initiation of the trial and were not attributed to the treatment. No other adverse effects due to treatment were reported. There were no differences in left ventricular mass index at the beginning or end of the study, which showed 107.3 ± 8.1 g/m2 versus 103.2 ± 7.1 g/m2 in the EB versus NT group (P = .46) and 114.4 ± 12.8 g/m2 versus 96.92 ± 6.4 g/m2 in the EB versus NT group (P = .54), respectively.
This study demonstrated that treatment of anemia with rHuEPO in KTRs to a target hemoglobin level of 11.5 to 13.5 g/dL improves the vitality and mental health domains in quality of life SF-36 scores. The treatment was safe, well tolerated, and not associated with adverse changes in cardiovascular outcomes or increased cardiovascular morbidity or thrombotic events. However, treatment did not reduce the rate of decline in graft function.
This trial is an important addition to the literature in the treatment of anemia with rHuEPO. Limited existing data have suggested that rHuEPO use is efficacious13-15 and safe for KTRs,16 but the long-term effects of rHuEPO treatment in KTRs have not been determined. The Correction of Anemia Post-Kidney Transplant Reduces Progression of Allograft Nephropathy (CAPRIT) study17 is the only other published randomized study that investigated the long-term benefits of rHuEPO in KTR. In this study, 128 participants between the age of 18 and 80 years who had received a kidney allograft more than 1 year previously and had anemia defined as hemoglobin < 11.5 g/dL were included. After randomization to receive EB to achieve either complete (hemoglobin level of 13-15 g/dL) or partial (hemoglobin level of 10.5-11.5 g/dL) correction of anemia, patients were followed for 24 months. In contrast to our study, there was a significant reduction in the rate of decline of renal function in the complete correction group compared with the partial correction group. There was also significant improvements in number of quality of life measures compared with baseline when assessed using the SF-36 in the complete correction group, including general health (P = .01), vitality (P = .01), physical function (P = .04), physical role (P = .02), mental health (P = .04), and social function (P = .05). Although our study was not powered to detect benefits in quality of life, we found benefits in the vitality domain and a higher proportion of patients achieving the Minimal Clinically Important Difference in the mental health domain in the EB group, with a trend to improvement; these results may show significance in a larger study.
There are important differences in both patient demographics and design between the CAPRIT and our study. We included participants from 3 months posttransplant compared with 1 year in the CAPRIT study. In addition, we compared treatment with no treatment rather than complete or partial correction of anemia. The mean erythropoietin dose required to achieve the targets were higher in the CAPRIT study, and they also had higher systolic blood pressures throughout the study compared with that shown in participants in our study. We found a lower rate of progression in the NT group of 4.16 mL/min (IQR, 12.42 - 2.78 mL/min) compared with the result of 5.8 ± 1.1 mL/min in the lower target group in the CAPRIT study.
It is possible that KTRs require a higher hemoglobin or erythropoietin dose to gain the benefits in graft progression and health quality of life shown in the CAPRIT study and that the benefits of rHuEPO and a raised hemoglobin are greater in older grafts as more chronic changes develop. Increased complications such as graft rejection and early infection in the first year may have affected the primary outcome results in our study. The lower systolic blood pressures in the NT group compared with that shown in the lower target group (group B) in the CAPRIT study may have confounded the outcomes, particularly with respect to progression of graft function. Progression of renal function in KTRs is complex and multifactorial and includes both immunologic and nonimmunologic factors. These findings suggest that additional factors such as good blood pressure control may exert a similar effect to achieving normal hemoglobin in preserving renal function in kidney allografts.
We found the use of rHuEPO to be safe and well tolerated; similar to that shown in the CAPRIT study, we did not see an increase in cardiovascular or thrombotic events. Our results are different from the results reported in 2 large randomized controlled trials in CKD patients: CREATE18 and CHOIR.19 This suggests that these risks may be different in the kidney transplant population. Potential explanations are that the dose of rHuEPO required in both transplant trials were less than in the CKD trials, which may have lessened cardiovascular and thrombotic risk, or that there was less erythropoietin resistance and potentially a less inflamed milieu in the transplant population.
Limitations of our study are the small study sample size and single center design, compared with other studies on rHuEPO treatment. Therefore, the apparent safety profile that we showed needs to be treated with caution because of the small study size. We found the rate of decline in the NT group to be lower than predicted from our retrospective data. The rate of decline was also lower than in the lower target group of the CAPRIT study. Therefore, a larger study group would be necessary to detect any differences in this particular population. Our follow-up period of 24 months, although comparable to CAPRIT and other anemia studies, may also not be long enough to assess for adverse cardiovascular effects of rHuEPO. The measurement of glomerular filtration rate by plasma or urine clearance of ideal filtration markers such as inulin or iohexol provides the most accurate measure of allograft function in KTRs. However, these are not used routinely in practice; therefore, validated calculations to estimate glomerular filtration rate were used. A further limitation is the unblinded nature of the study; this may have affected the health quality of life assessments.
Only 1 other trial has been published assessing the long-term effects of rHuEPO treatment on KTRs with anemia.17 Although this study does not confirm the benefits of targeting a hemoglobin level of 11.5 to 12.5 mg/dL to reduce the rate of graft dysfunction in KTRs shown in the CAPRIT study, it does suggest benefits in health in some quality of life domains. We found rHuEPO to be well tolerated, and no significant cardiovascular or thrombotic events occurred. More studies are needed to investigate the benefits of this potentially important outcome measure and to determine appropriate targets for KTR with anemia.
DOI : 10.6002/ect.2018.0283
From the 1Translational Medicine and Therapeutics, William Harvey Research
Institute, Queen Mary University of London, London, United Kingdom; and the
2Department of Renal Medicine and Transplantation, Barts Health NHS Trust,
London, United Kingdom
Acknowledgements: The authors have no conflicts of interest to declare. This study was supported with funding from Roche Products Limited. Roche Products Ltd. was not involved in the preparation, drafting, or editing of this manuscript. Roche Products Limited has conducted a factual accuracy check on the final article, but any decisions to incorporate comments were made solely at the discretion of the authors. The Clinical Trial Notation for this work is ISRCTN41687085.
Corresponding author: Taryn Pile, Renal Unit, Transplant, Renal and Urology Directorate, 6th Floor, Borough Wing, Guys Hospital, Great Maze Pond, United Kingdom SE1 9RT
Phone: +44 20 7188 5708
Figure 1. Enrollment, Randomization, and Study Completion
Figure 2. Change in Hemoglobin Over Time
Figure 3. Change in Estimated Glomerular Filtration Rate Over Time
Figure 4. Change in Blood Pressure Over Time
Figure 5. Changes From Baseline in the SF-36 Health Domain and Component Summary Measures
Table 1. Demographics and Baseline Characteristics
Table 2. Baseline Biometric and Blood Pressure Readings
Table 3. Baseline Hematologic and Biochemical Data