Posttransplantation anemia in kidney transplant recipients: A retrospective cohort study.

We sought to assess the frequency and predictors of early and late posttransplantation anemia (PTA). In addition, we aimed to assess the outcomes of patients with anemia and to assess the impact of anemia on mortality, graft function, and graft failure.Patients who underwent kidney transplantation in a single center during a 4-year period were included. Predictors associated with the development of anemia at 6 months (early PTA) or 2 years (late PTA) were evaluated in a univariate and multivariate analyses. The effects of anemia and other variables on mortality and graft function were assessed.A total of 266 kidney transplant recipients were included. The prevalence of PTA at 6 months (early PTA) was 51.3% and at 2 years (late PTA) was 36.6%. Female sex was significantly associated with early PTA. Patients with early PTA proceeded to late PTA. Patients with both early and late PTA had a higher mortality rate at 4 years compared to patients without anemia. On multivariable analysis, lower Hb at 2 years posttransplantation (hazard ratio [HR] 0.716, 95% confidence intervals [CI] 0.541-0.948, for every increment of 1 g/dL) was significantly associated with mortality. Patients with late PTA suffered a decline in eGFR compared to patients without anemia (P = .026). Furthermore, a lower Hb at 2 years posttransplantation was also associated with graft failure (HR 0.775, 95% CI 0.619-0.969, for every increment of 1 g/dL).Post-transplantation anemia is significantly associated with late mortality, with a decline in graft function and with an increased incidence of graft failure.

Introduction

Anemia frequently complicates chronic kidney disease (CKD) and has been shown to be associated with increased mortality in dialysis patients.[ Anemia may potentially be corrected by kidney transplantation, yet posttransplantation anemia (PTA) is estimated to be in the range of 20% to 40% of patients.[

PTA has not been uniformly defined in terms of timing after transplantation and degree of anemia. Some authors suggest distinguishing between early PTA (up to 6 months) and late PTA (after 6 months and up to 5 years).[

The most common reason for early PTA is iron deficiency.[ Iron deficiency may be caused by depletion of iron stores before transplantation and perioperative blood loss.[ In addition, slowly increasing levels of the newly graft-produced-erythropoietin may contribute to the anemia.[ The occurrence of late PTA has been associated with impaired graft function and the development of renal insufficiency.[

Several studies have shown that PTA may be associated with increased mortality[ and decreased graft survival,[ and of de-novo congestive heart failure.[

We sought to assess the frequency of early and late PTA, and to find predictors, in order to help define subgroups at risk for the development of anemia. In addition, we aimed to assess the impact of anemia on clinical outcomes—mortality, graft function, and graft failure.

Materials and methods

Study design and patients

This is a single-center retrospective cohort study using the Rabin Medical Center (RMC) transplantation department registry. This registry prospectively collects data regarding the recipients’ characteristics and outcomes. Adult patients (age >18 years) who underwent kidney transplantation in our center during a 4-year period between 1 January 2008 and 31 December 2011 were included. Due to the reported association between diabetes mellitus (DM) and anemia,[ only patients without DM before the transplantation were included in the analysis. The study was approved by the RMC institutional review board.

Definitions

Patients were classified as having PTA based on the Hb levels at 6 months (defined as early PTA) and at 2 years (defined as late PTA). Anemia was defined according to World Health Organization (WHO) and American Society of Transplantation guidelines as hemoglobin levels <13 mg/dL in men and <12 mg/dL in women.[

Data collection

Data relevant to the diagnosis of anemia and anemia workup were evaluated at 3 time points: before transplantation (maximum 6 months prior), 6 months, and 2 years following transplantation. Anemia workup included mean corpuscular volume (MCV), red cell distribution width (RDW), iron, transferrin, ferritin, percentage of hypochromic red blood cells (RBCs), vitamin B12, folic acid, thyroid-stimulating hormone (TSH), erythropoietin level, reticulocyte count, haptoglobin, lactate dehydrogenase bilirubin, and vitamin D.

Pretransplantation characteristics were collected. Post-transplantation laboratory variables included estimated glomerular filtration rate (eGFR) values, calculated according to the chronic kidney disease epidemiology collaboration (CKD-EPI) 2009 equation,[ and the number of patients with CKD, defined as eGFR <60 mL/min, at 6 months and 2 years after transplantation (the time of anemia workup). The occurrence of new onset diabetes after transplantation (NODAT) is defined according to the American Diabetes Association criteria for the diagnosis of DM.[

The use of maintenance immunosuppressive therapy, lipid-lowering medications, angiotensin-converting enzyme (ACE) inhibitors, or angiotensin receptor blockers (ARB), antiplatelets, and use of trimethoprim–sulfamethoxazole administered as pneumocystis jiroveci prophylaxis were documented.

Outcomes

The primary outcome of the study was all-cause mortality at the end of follow-up. We compared all-cause mortality between those who developed PTA and those who did not in order to evaluate the effect of anemia on mortality. Secondary outcomes included graft failure (defined as re-establishment of long-term dialysis therapy, the need for retransplantation or death), change in kidney function (defined as the difference in eGFR between eGFR at 2 years and at 6 months), and major adverse cardiovascular events (including myocardial infarction, revascularization, stroke, and acute coronary syndrome).

Statistical analysis

The statistical analysis was generated using SAS Software, Version 9.4. Continuous variables were presented by mean ± standard deviation, and categorical variables were presented by (N, %). T-Test was used to compare the value of continuous variables between study groups. The χ2 (for >2 groups) or Fisher exact (for 2 groups) tests were used to compare the value of categorical variables between study groups. Two-sided P values <.05 were considered statistically significant. Characteristics of patients who developed early and late PTA were compared in both univariate and multivariate analyses in order to find predictors for PTA. Time-to-event outcomes (mortality and graft failure) were compared using Kaplan–Meier survival analysis using the log-rank test. We evaluated the effect of the above variables on overall survival and graft failure in a univariate analysis. Variables found to be significantly associated with mortality (P <.05) were entered into a Cox regression multivariate analysis. Hazard ratios (HRs) with 95% confidence intervals (CI) were calculated.

Results

Study population, anemia prevalence, and follow-up

During the study period, 398 individuals underwent kidney transplantation at RMC. Of these, 132 were excluded due to pretransplantation DM. A total of 266 kidney transplant recipients were included in the study, of which 261 could be evaluated for anemia at 6 months and 254 could be evaluated at 2 years. Baseline characteristics of the cohort, according to development of PTA at 6 months and at 2 years are shown in Tables 1 and 2. The prevalence of anemia pretransplantation was 71% (184/259) with a mean pretransplantation Hb of 11.6 ± 1.6 g/dL. The prevalence of PTA at 6 months (early PTA) was 51.3% (134/261) with a mean Hb of 12.54 ± 2.05 g/dL, and the prevalence of PTA at 2 years (late PTA) was 36.6% (93/254) with a mean Hb of 13.18 ± 2.47 g/dL. Mean follow-up time for the cohort was 5.46 ± 1.21 years.

Table 1

Clinical characteristics of patients with anemia at 6 months (early PTA).

Table 2

Clinical characteristics of patients with anemia at 2 years (late PTA).

Risk factors for early and late PTA

Baseline characteristics pretransplantation associated with early PTA on univariate analysis included young age, low weight, and low cholesterol level. More patients with anemia had a history of a cerebrovascular accident. Hematological parameters included a higher percentage of hypochromic RBCs pretransplantation and at 6 months posttransplantation, as well as a higher RDW at 6 months posttransplantation. Other factors posttransplantation associated with early PTA included a lower eGFR, the presence of CKD, a lower vitamin D level, and proteinuria (Table 1). These factors, in addition to gender, were entered into a multivariate analysis, and those that remained significantly associated with early PTA were female sex (HR 3.13, 95% CI 1.180–8.284), lower eGFR (HR 0.97, 95% CI 0.95–0.99, for every increment of 1 mL/min/1.73 m2), and hypochromic RBCs (HR 1.053, 95% CI 1.001–1.1107).

Baseline characteristics pretransplantation associated with late PTA on univariate analysis included female sex, low weight, a lower triglycerides level, and a lower percentage of a living donor as the transplantation source (Table 2). Hematological parameters included a higher RDW and a higher B12 level pretransplantation and at 6 months posttransplantation, a lower Hb level 6 months posttransplantation and a lower iron level 2 years posttransplantation. Other factors posttransplantation associated with late PTA included lower eGFR and the presence of CKD, at 6 months and 2 years, proteinuria at 6 months, and the use of cyclosporine (Table 2).

On multivariable analysis, the hematological factors which remained significantly associated with late PTA included: a lower iron level at 2 years posttransplantation (HR 0.986, 95% CI 0.974–0.998, for every increment of 1 mg/dL), and a lower Hb level at 6 months posttransplantation (HR 0.601, 95% CI 0.453–0.799, for every increment of 1 g/dL). Other factors included a low body weight pretransplantation (HR 0.957, 95% CI 0.925–0.989, for every increment of 1 kg), living donor as the transplantation source (HR 0.343, 95% CI 0.142–0.831), and a low triglycerides level pretransplantation (HR 0.993, 95% CI 0.986–0.998, for every increment of 1 mg/dL).

Outcomes of patients with anemia

Patients with early PTA had a higher mortality rate at 4 years (8.96% vs 2.36%, P = .022) and a higher percentage of graft failure (18.66% vs 6.3%, P = .003) (Table 3).

Table 3

Outcomes of patients with anemia.

Patients with late PTA also had a higher mortality rate at 4 years (8.6% vs 1.24%, P = .004) (Fig. 1) and a higher percentage of graft failure (23.66% vs 3.11%, P <.001).

Figure 1

Overall survival of patients according to the presence of anemia at 2 years. Blue curve—patients without anemia, red curve—patients with anemia.

In patients with late PTA, there was a decline in kidney function. The eGFR declined with time, with a difference between eGFR at 2 years and at 6 months of 2.26 mL/min/1.73 m2 (SD 19.29) in patients with anemia, in contrast to a rise of 3 mL/min/1.73 m2 (SD 14.74) in patients without anemia (Table 3). Hence, there was a significant difference of 5.3 mL/min/1.73 m2 in eGFR between patients with and without anemia (P = .026).

There was no difference in the frequency of cardiovascular events between patients with early and late PTA and those without them (Table 3).

Risk factors for mortality

Fifteen patients died during follow-up (5.6% mortality rate). Risk factors for 4-year mortality on univariate analysis included: older age at transplantation, a lower Hb level at 6 months and 2 years, a higher RDW and a lower iron level at 6 months, a lower folic acid level at 6 months posttransplantation, and a lower iron level at 2 years posttransplantation. In addition, living donor as the transplantation source was associated with a less mortality. All these variables in addition to gender were included in the multivariable model. On multivariable analysis, the factors that remained significantly associated with mortality were older age (HR 1.123, 95% CI 1.040–1.213, for every increment of 1 year) and a lower Hb at 2 years posttransplantation (HR 0.716, 95% CI 0.541–0.948, for every increment of 1 g/dL) (Table 4).

Table 4

Multivariable analysis for mortality.

Risk factors for graft failure

Risk factors for graft failure at 4 years on univariate analysis included: older age at transplantation, the following parameters at 6 months posttransplantation: a lower Hb level, lower iron level, a higher RDW and a higher TSH, and a lower Hb level at 2 years posttransplantation. In addition, the presence of CKD and lower eGFR at 6 months and 2 years were associated with graft failure.

On multivariable analysis, the following factors were found to be significantly associated with graft failure: male sex (HR 0.453, 95% CI 0.212–0.969), lower Hb at 2 years posttransplantation (HR 0.775, 95% CI 0.619–0.969, for every increment of 1 g/dL), lower iron level at 6 months posttransplantation (HR 0.986, 95% CI 0.974–0.998, for every increment of 1 mg/dL), lower eGFR at 2 years (HR 0.927, 95% CI 0.900–0.956), and an elevated TSH level (HR 1.265, 95% CI 1.134–1.411).

Discussion

In this single-center retrospective cohort study of 266 renal transplant recipients, we attempted to define the prevalence of early and late PTA, to find predictors for anemia and to evaluate their effect on patient and graft survival.

The prevalence of PTA in our cohort at 6 months (early PTA) was 51.3% and at 2 years (late PTA) was 36.6%. These results are consistent with previously reported prevalence rates that show that at 6 months posttransplantation, approximately one-half of the patients are anemic[ and during the first 5 years posttransplantation, the frequency declines and only approximately one-third are anemic.[

The predictors found to be significantly associated with early PTA were female sex, lower eGFR at 6 months, and hypochromic RBCs. Previous studies have also found female sex to be an independent predictor for anemia at 6 months as in our study[ and at 12 months posttransplantation,[ as well as in patients with CKD.[ A higher eGFR was shown to confer protection against PTA at 12 months.[ In renal transplant recipients, low renal function was found to be associated with reduced synthesis of erythropoietin by the allograft, thereby resulting in PTA.[ Interestingly, hypochromic RBCs, which are a marker for iron deficiency, were found to be an independent risk factor for mortality in renal transplant recipients.[

A novel observation in this study, compared to previous studies, was that the presence of anemia at 6 months posttransplantation (early PTA) predicted late PTA. This fact possibly suggests that better prevention and treatment of early PTA might prevent late PTA. Other predictors for late PTA in our cohort included a low iron level at 2 years posttransplantation, a low triglyceride level as well as low weight before transplantation. Iron deficiency, which is one of the common causes of PTA may be due to postoperative losses, frequent phlebotomy, utilization of the iron stores for enhanced erythropoiesis, and nutritional deficiency.[ A low level of triglycerides has been shown previously by our group to be an independent predictor for emergence of anemia in women in a large cohort of 10,000 adults attending a screening center.[ Of note, in the same cohort, a low basal metabolic index (BMI) was associated with anemia in women, yet only in a univariate analysis. In another retrospective study of renal transplant recipients, anemic patients had a significantly lower BMI, as in our series.[ Low BMI has been shown to be associated with anemia in populations other than renal transplant recipients, such as institutionalized elderly patients.[

Although iron deficiency is suggested by our findings as the main cause for anemia, chronic inflammation may be an alternate hypothesis. We found no difference in ferritin levels in the anemic versus the nonanemic patients, yet the ferritin levels were relatively high. The iron levels at 2 years, although significantly lower in anemic patients, are not sufficiently low to be responsible for the anemia. Moreover, as ferritin is an acute phase reactant, this may suggest that inflammation has a role in the pathogenesis of anemia. This hypothesis could possibly be strengthened by demonstrating an elevated hepcidin level in the anemic patients, yet this test is currently not available in clinical practice. Interestingly, patients treated with trimethoprim–sulfamethoxazole had a lower frequency of anemia at 6 months and 2 years and this finding was statistically significant. A possible explanation may be that trimethoprim–sulfamethoxazole actually had a protective role for anemia by its antimicrobial and anti-inflammatory qualities. This again may lend support to the inflammation hypothesis.

A novel finding that was not previously reported in renal transplant recipients is the decline of eGFR with time in patients with anemia. There was a difference of 5.26 mL/min/1.73 m2 in eGFR, between 6 months and 2 years. The eGFR in anemic patients declined by 2.26 mL/min/1.73 m2, whereas the eGFR increased in the nonanemic patients. Hb level has previously been shown to be an independent risk factor for doubling of serum creatinine or for end-stage renal disease (ESRD) in patients with diabetes in the RENAAL study.[ Hb was incorporated into a risk score predicting ESRD alone or ESRD with death, together with albuminuria, hypoalbuminemia, and serum creatinine level.[ Similar results were shown in a cohort of Japanese patients with CKD. A rapid decline of eGFR was strongly associated with anemia.[ To the best of our knowledge, the finding that Hb is associated with GFR decline was not shown so far in renal transplant recipients. As low serum iron was significantly associated with late PTA, and hypochromia was a risk factor for early PTA, iron deficiency anemia may be partial responsible for the decline in GFR. These observations suggest that early and aggressive treatment of anemia after kidney transplantation might reduce the rate of GFR decline or even the decline in graft function. However, due to the retrospective design of our study, we cannot assume causality between anemia and GFR decline, and this might be only an association. Moreover, CKD is known to be associated with reduced erythropoiesis and anemia, and decreased renal function is associated with reduced synthesis of erythropoietin by the allograft.[

Mortality and graft failure rate were higher in patients with early and late PTA. The important findings in our study are that PTA at 2 years was an independent predictor for 4-year mortality, with every increment of 1 g/dL in Hb level, reducing mortality by 22% (HR 0.784, 95% CI 0.616–0.998). In addition, PTA at 2 years was an independent predictor for graft failure, with every increment of 1 g/dL in Hb level, reducing graft failure by 22% (HR 0.775, 95% CI 0.619–0.969, for every increment of 1 g/dL). These findings are consistent with previous reports. In a prospective cohort study of 938 kidney transplant recipients in a single center in Hungary, PTA was also found to be associated with mortality and graft failure at the end of a 4-year follow-up,[ with a HR of 1.69 (95% CI 1.115–2.560) for mortality. In a large retrospective analysis of 2031 transplant recipients in Austria, anemia was significantly associated with mortality and graft failure, at a median follow-up of 6 years, with every increment of 1 g/dL in Hb level, reducing mortality by 18% (HR 0.82, 95% CI 0.76–0.89).[ In another retrospective series of 626 transplantation patients in a single-center cohort from Pennsylvania, PTA at 12 months was associated with an increased risk of death (with a HR of 3.0, 95% CI 1.3–6.7).[

Nevertheless, other studies have shown conflicting results regarding long-term mortality. In a retrospective European study of 825 renal transplant recipients, anemia was not associated with all-cause mortality with a follow-up of 8 years.[ In another multicenter prospective cohort of 2102 transplant recipients, Hb levels were not associated with any effect on cardiovascular morbidity and mortality at a follow-up of 5 to 6 years.[ In a retrospective Chinese study of 887 renal transplantation recipients, PTA at 12 months was not associated with mortality.[ Yet, in all of these studies, anemia was significantly associated with graft loss, as in our study.[

Anemia has been associated with mortality in various clinical settings. In a population-based cohort study, anemia was found to be an independent risk factor for all-cause mortality in the general population.[ Anemia is also associated with increased mortality in the setting of heart failure,[ acute coronary syndrome,[ perioperative cardiac surgery,[ and malignancy.[ Importantly, anemia is associated with mortality in CKD.[

Several mechanisms may explain the association between anemia and mortality in renal transplant recipients. Similarly to the CKD population, cardiovascular factors are the leading cause of death in renal transplant recipients.[ Anemia has been associated with left ventricular hypertrophy and congestive heart failure in renal transplant recipients, and both of these factors, as well as anemia itself, were independent predictors of mortality.[ Iron deficiency, independently of anemia, was also found in a prospective study to be associated with mortality in renal transplant recipients.[ In our study, markers of iron deficiency (percentage of hypochromic RBCs) were significantly associated with early PTA, and low serum iron was associated with late PTA, which in turn was associated with mortality. Hence, the link between anemia and mortality may possibly be through iron deficiency.

The strengths of our study is the adherence to the acceptable WHO definitions for anemia, which were endorsed by the American Society for Transplantation and differ for men (Hb <13 g/dL) and women (Hb <12 g/dL).[ Many prior studies have used a single threshold for both men and women (usually <11–12), thus possibly underestimating the frequency of anemia. In addition, this study assessed PTA at 2 time points after transplantation, unlike most previous studies that studied only 1 time point.

Our study has several limitations. First, this is a retrospective cohort study and as such the information was collected from computerized medical files only and from a single-center database. Therefore, care should be taken in generalizing the results to the entire kidney transplantation population. Second, data regarding erythropoiesis-stimulating agents (ESAs) after transplantation were too scarce to analyze, and this may have possibly affected patient outcomes. However, randomized controlled trials (RCTs) that examined the effect of ESA compared with placebo on graft function did not find a beneficial effect.[ We also lacked data regarding administration of oral or intravenous (IV) iron posttransplantation. In contrast to the abundance of data from RCTs on the efficacy of iron in the CKD population,[ similar studies in transplant recipients are scarce. Our group reported on elevation of Hb levels with IV iron administration after transplantation in a small retrospective study.[ In contrast to our experience, an RCT comparing oral with a single dose of IV iron after transplantation did not show a difference in time to anemia correction.[

Implication for practice and research

The predictors that we found for PTA suggest that there are several risk groups in which a closer follow-up for early identification of anemia might be warranted—women and decreased eGFR for early PTA, a low triglyceride level, low body weight, a low iron level, and a low Hb level posttransplantation for late PTA. The fact that early PTA predicted late PTA in renal transplantation recipients and late PTA was associated with decline in eGFR, graft failure and mortality have important clinical and therapeutic implications. As anemia is a relatively easily modifiable risk factor, preventable causes for anemia such as iron deficiency should be sought and anemia should be treated as early as possible. Future RCTs should focus on correction of anemia using either iron, ESAs, or both in an effort to improve graft function, graft survival, and patient survival after transplantation. Future research should better elucidate the link between PTA and mortality.