Anemia is a common complication of chronic kidney disease (CKD) and is associated with myocardial hypertrophy, congestive heart failure, and death (1, 2). Although use of erythropoiesis-stimulating agents (ESAs) has significantly reduced its severity in end stage renal disease (3), anemia is common at the time of transplantation. Although graft function seems to significantly impact hemoglobin (Hb) level, it does not entirely account for posttransplant anemia. Other factors, including donor and recipient factors, medications, and acid-base status have been associated with anemia (4–8). Prior studies suggest an association between anemia and adverse transplant outcomes, including acute rejection, graft failure, and death (4, 5, 8).
Although ESA and iron therapy are cornerstones of anemia treatment in CKD, their use seems to be limited after kidney transplantation, with ESA use reported in 5% to 30% of anemic transplant recipients (5, 6, 8, 9). Although the impact of ESA on transplant outcomes is not well studied, a recent analysis suggested that targeting Hb more than 12.5 g/dL with ESA is associated with increased mortality risk (4).
The studies of anemia posttransplantation suffer from major limitations and have not provided convincing evidence for the negative impact of anemia on outcomes. These studies generally examine a single Hb value at various times posttransplantation and do not control for important confounding factors. The lack of strong evidence linking anemia and transplant outcome likely contributes to its suboptimal management.
We sought to examine anemia prevalence, ESA and iron supplementation use, and factors associated with Hb level posttransplant and to evaluate the association of Hb, measured at uniform time points as a time-varying factor, with long-term graft and patient survival. The results were adjusted for a number of time-dependent and independent factors, including graft function and immunosuppressive regimen.
During the study period, there were 530 renal transplant recipients eligible for inclusion in this analysis. Patient- and transplant-related characteristics are summarized in Table 1. Patients were predominantly male (63.2%), and 45.8% were African American (AA). The mean age was 52.3 years, and the majority received deceased-donor (DD) transplants (67.2%).
The gender-specific mean values of the Hb at different time points are shown in Figure 1a, showing an increase in Hb during the first 9 months and stable values afterward. We adopted a more conservative definition of anemia than the definition by World Health Organization (WHO) (10). For men, grades 0 to 4 included Hb values ≥12, 11 to 12, 10 to 11, 9 to 10, and <9 g/dL, respectively. Corresponding values for women were ≥11, 10 to 11, 9 to 10, 8 to 9, and <8 g/dL, respectively. The point prevalence of different degrees of anemia is depicted in Figure 1b. Overall, the prevalence of higher grades of anemia decreased during the 2 years posttransplant, with less than 3% of the patients with grade 4 anemia after the first year. Using the WHO definition, the overall prevalence of anemia was 89.4% at the time of transplant, 49.2% at 1 year, and 44.3% at 2 years posttransplant.
Overall, use of ESA decreased from 25.6% at 1 month to 8.23% at 24 months (Fig. 1b); its use in those with severe anemia (grades 3 and 4) varied from 30.9% to 51.2%. Iron supplementation occurred in 10.9% of participants in early posttransplant period and remained relatively stable at 21.0% to 29.2% afterward.
To identify the factors associated with Hb, we used generalized estimating equations, with age, gender, race, diabetes, hepatitis C, retransplantation, donor type, delayed graft function (DGF), induction agent, estimated glomerular filtration rate (eGFR), tacrolimus (TAC) levels, mycophenolate mofetil (MMF) and prednisone doses, serum bicarbonate, and use of ESA, iron supplements, and angiotensin-converting enzyme inhibitor/angiotensin receptor blocker (ACEi/ARB), as time-varying and fixed predictor covariates. Factors that were independently associated with higher Hb level included male gender (β=0.64, P<0.001, confidence interval [CI]: 0.45–0.82), eGFR (β=0.21 per 10 mL/min/1.73 m2, P<0.001; CI: 0.16–0.27), serum bicarbonate level (β=0.4 per 10 mmol/L increase, P<0.001; CI: 0.31–0.85), and use of ACEi/ARB (β=0.36, P<0.001; CI: 0.16–0.55), and those independently predictive of decreased Hb were AA race (β=−0.34, P=0.001, CI: −0.54 to −0.14), iron (β=−0.28, P=0.003, CI: −0.47 to −0.09) and ESA use (β=−0.73, P<0.001, CI: −0.93 to −0.52), and prednisone (β=−0.46, P<0.001, CI: −0.71 to −0.22 for >10 mg/day compared with none).
Anemia and Study Endpoints
During the 31.0±14.1-month follow-up, 64 patients returned to dialysis or underwent retransplantation. We examined the association of the degree of anemia and other factors listed in the tables with graft failure individually. Factors not associated with death-censored graft loss included age, gender, type of induction, diabetes, previous transplantation, use of ESA, and iron supplementation. However, AA race (hazard ratio [HR]: 1.84, P<0.02, CI: 1.1–3.0), receipt of a DD transplant (HR: 3.8, P<0.001; CI: 1.8–8.0), DGF (HR: 3.2, P<0.001, CI: 1.94–5.20), eGFR (HR: 0.93, P<0.001; CI: 0.91–0.94), TAC level (HR: 0.82, P<0.001; CI: 0.75–0.90), MMF dose (HR: 0.45, P=0.05, CI: 0.2–1.0 for ≥2000 vs. <1000 mg/day), prednisone dose (HR: 8.4, P<0.001, CI: 3.3–21.0 for >10 vs. 0 mg/day), serum bicarbonate (HR: 0.82, P<0.001, CI: 0.77–0.88), and grades 2 to 4 anemia (HR: 3.5, P=0.001, CI: 1.6–7.4 for grade 1; HR: 5.9, P<0.001, CI: 2.7–12.8 for grade 2; HR: 8.4, P<0.001, CI: 3.4–20.4 for grade 3; and HR: 29.2, P<0.001, CI: 13.7–61.9 for grade 4, all compared with grade 0) (Fig. 2) were all associated with graft failure. There was no interaction between degree of anemia and ESA use. To evaluate the independent association between anemia and the primary endpoint, we included all these predictor covariates in a single model. After adjustment for the potential confounding covariates, grade 4 anemia was independently associated with graft failure (HR: 5.8, P<0.001, CI: 2.2–15.4, compared with grade 0). To remove death with functioning graft as a competing factor for primary endpoint, we used a competing-risk regression model. Controlling for the listed covariates, the hazard of graft failure for grade 4 anemia was 5.25-fold higher, when compared with grade 0 (P=0.005, CI: 1.7–16.7).
There were 55 deaths during the follow-up period. We evaluated the predictive role of previously described variables for death. Among them, factors associated with this outcome included age (HR: 1.05, P<0.001, CI: 1.02–1.07), DD source (HR: 5.4, P<0.001, CI: 2.2–13.6), DGF (HR: 2.7, P<0.001, CI: 1.6–4.7), eGFR (HR: 0.97, P<0.001, CI: 0.96–0.98), serum bicarbonate (HR: 0.88, P=0.001, CI: 0.82–0.95), MMF dose (HR: 0.3, P=0.008, CI: 0.1–0.7 for ≥2000 vs. <1000 mg/day), use of ESA (HR: 1.8, P=0.004, CI: 1.2–2.8), and degree of anemia (HR: 3.4, P=0.002, CI: 1.6–7.5 for grade 2; HR: 6.0, P<0.001, CI: 2.5–14.3 for grade 3; HR: 8.4, P<0.001, CI: 3.4–20.8 for grade 4, all compared with grade 0) (Fig. 3). After controlling for these covariates, degree of anemia remained significant independent predictor of death (HR: 2.2, P<0.1, CI: 0.9–5.6 for grade 2; HR: 3.9, P=0.009, CI: 1.4–10.8 for grade 3; and HR: 4.8, P=0.08, CI: 1.5–15.4 for grade 4, all compared with grade 0). In this model, ESA was not independently associated with risk of death.
In this study, we examined the prevalence and severity of anemia within the first 2 years posttransplantation and evaluated the association between anemia and transplant outcomes. In contrast to other published data, we examined Hb value as a time-varying predictor variable for outcomes, measured at multiple time points, and controlled for major time-dependent and independent confounding factors in a large cohort of kidney transplant recipients (Table 2).
Our data provide further evidence that anemia is common after transplantation, and its prevalence declines during the first year posttransplant. An important finding in our study was the low rate of ESA use in anemic patients. Even with severe anemia, less than 50% of the patients received ESA. Moreover, we demonstrated a significant association between degree of anemia and graft and patient survival. Controlling for major confounders including eGFR and immunosuppressive drug exposure, we observed a significant increase in the risk of graft failure (HR: 5.3) with grade 4 anemia, Hb<9 g/dL in men and 8 g/dL in women. Furthermore, we found a graded association between anemia and risk of death, with increasing death risk with worsening anemia among patients with moderate-severe anemia (adjusted HR: 2.2–4.8).
Reports of anemia prevalence posttransplantation have demonstrated varied results. This variance is attributable to differences in definition of anemia, study design, that is, cross-sectional versus longitudinal, time of assessment, and patient population examined. Prevalence of anemia, using the WHO definition, decreased from 88.6% at 1 month to 49.3% at 12 months and 44.3% at 24 months in our cohort. Others have reported anemia in 19.3% to 35.5% of the patients at 6 months and among 12% to 25% of the recipients at 12 months posttransplant (7, 11–14) (Table 3). Mix et al. (11) retrospectively evaluated 240 recipients and found that prevalence of anemia, defined as hematocrit less than 36%, rose from 21% at 1 year to 36% at 4 years. Yorgin et al. (12) observed an increase in anemia, defined as hematocrit less than 33%, from 12% at 1 and 2 years posttransplant, to 26% at 5 years among 128 patients. In cross-sectional studies, 4.9 to 8.5 years (mean) posttransplant, anemia has been reported in 30.8% to 45.6% of the patients (6, 8, 15–17). More recently, in a retrospective analysis of 1023 renal transplant recipients by Chhabra et al. (18), 13% had mean Hb less than 11 g/dL after 3-month posttransplant.
In our cohort, we found that the factors independently associated with Hb included gender, race, graft function, acid-base status, and medications including ACEi/ARB, prednisone, ESA, and iron supplements. We did not find predictive roles for MMF or TAC. The majority of these associations are consistent with the previous reports. In the study by Shah et al. (6), the significant predictors of Hb were age, gender, and eGFR, whereas the impact of ACEi/ARB and immunosuppressive drugs were negligible. Winkelemyer et al. (19) found that hematocrit value, measured 7.7 years (mean) posttransplant, was associated with gender, serum creatinine (SCr), and use of TAC and MMF. Contrary to our observation, they reported a dose-response inverse relationship with ACEi but not ARB. Chadban et al. (17) also reported independent association between anemia and age, gender, and eGFR, with no significant effect of MMF dose on Hb. Although ACEi/ARB use was associated with lower Hb in women, there was no significant effect in men. On the contrary, Fernandez Fresnedo et al. (14) reported higher likelihood of anemia with MMF use and no difference with ACEi/ARB. In a retrospective study of 626 recipients, Imoagene-Oyedeji et al. (20) found donor age, and 3 months SCr and anemia as independent predictors of anemia at 12 months.
Similar to our findings, Yorgin et al. (12) reported a significant correlation between serum bicarbonate and anemia. The higher prevalence of anemia in the general population among AAs has been well recognized (21). Similar to our finding, Shibagaki and Shetty (7) reported an independent association between AA race and anemia.
We have observed an association between ACEi/ARB use and higher Hb level, which is at odds with some other studies (6, 7, 20). It is unclear whether this is a random observation as opposed to a true association. Also there was a paradoxical association between ESA/iron use with Hb, which is most likely due to patients with lower Hb were more likely to receive these medications.
The important issue that we tried to address was the association of anemia with graft and patient outcomes. In a cohort study by Kolonko et al. (5), among 385 kidney transplant recipients, both persistent anemia and late-onset anemia were associated with increased risk of graft loss (HR: 4.1). In the study by Kamar and Rostaing (22), allograft loss and death were more prevalent in patients with anemia. However, it was not independently predictive of graft loss. In this analysis, anemia was not further investigated by severity, and the total number of allografts lost was small. Winkelmyer et al. (23) examining 438 kidney transplant recipients 4.4 years (median) posttransplant did not find any significant association between Hb less than 10 g/dL with endpoint. However, they reported an independent association between percentage of hypochromic RBC with mortality. Chhabra et al. (18) reported an independent association between mean Hb less than 11 after 90 days posttransplant and increased risk of death (HR: 3.2) and graft loss (HR: 2.7). Moreover, they observed an increased risk of acute cellular rejection in anemic patients (HR: 1.8). Molnar et al. (15) studied the predictive value of baseline anemia and assessed 55 months (median) posttransplant on outcomes in 938 patients. They observed that Hb and anemia were independently associated with mortality (HR: 1.01 and 1.7, respectively) and graft failure (HR: 1.02 and 2.5, respectively).
Our study was not designed to evaluate the safety and efficacy of ESA after kidney transplantation. However, we did not observe any association between its use and graft failure. Although ESA was found to be associated with higher risk of death, this was not an independent association.
The major faults with the published studies regarding posttransplant anemia include use of a single measurement of Hb, and in only one study use of average value, and variable time of measurement after transplantation. In this study, we have used Hb value as a time-varying predictor variable, measured at multiple uniform time points in a large sample size from a single center transplanted during a relatively short time period. We have also assessed major time-dependent factors including eGFR, immunosuppressive agents, serum bicarbonate, and use of ESA, iron supplementation, and ACEi/ARB at the same time points and have controlled for these covariates and other time-independent factors to examine the independent association of Hb level with graft and patient survival after kidney transplantation. However, we do not intend to make any causal relationship between anemia and transplant outcomes. The fact that severe degree of anemia is predictive of graft failure may be reflective of association of anemia with an unidentified causal factor. For instance, inflammation has been associated with anemia and could also lead to progressive graft injury (24–26). However, we did not check any marker of inflammation, such as C-reactive protein, during the study period.
The findings of our study, albeit more robust compared with prior studies, should be interpreted with caution. We did not have data regarding other risk factors for study endpoints, for example, blood pressure and baseline coronary artery disease; similarly, we did not know the causes of death or graft failure. Therefore, we cannot exclude residual confounding. Furthermore, treatment of anemia was not uniform, at the discretion of caring nephrologist. Therefore, we found a wide variation in use of iron supplementation and ESA in patients with anemia.
In summary, in this cohort study, we observed a decline in prevalence of anemia during the first year after transplantation, but it still remained an important problem afterward, with 5% of the patients suffering from severe anemia. Furthermore, our data suggest that even in the modern era of transplant care, treatment of anemia with ESA and iron supplementation remains limited. Importantly, we observed an independent association between Hb level, measured at multiple time points, and graft failure, and mortality, after adjustment for several time-dependent and independent confounding factors, including eGFR, and immunosuppressive regimen. These findings raise important questions about whether more aggressive correction of severe anemia with iron supplementation and ESA would be beneficial in improving graft or patient survival posttransplantation. Further studies are needed to examine the causal relationship between anemia and graft and patient outcomes.
MATERIALS AND METHODS
This is a retrospective cohort study of adult renal allograft recipients who were transplanted at our center from January 2004 to June 2006. All patients received induction therapy with thymoglobulin or basiliximab. Patients received methylprednisolone intraoperatively, followed by tapering oral prednisone. Prednisone was tapered off over 3 weeks in majority of the cases. For patients with high panel-reactive antibody (PRA), previous transplants and those with early acute cellular rejection prednisone were generally tapered to 10 mg/day at 3 months and 5 mg/day at 6 to 12 months. All patients were started on MMF 2 g/day postoperatively, with subsequent dose adjustment as needed. TAC target 12-hour trough level was 8 to 9 ng/mL during first year and 6 to 8 ng/mL afterward.
After hospital discharge, patients were followed up according to the recommended guidelines (27). Patients with DGF or unexplained increase in SCr were investigated by graft biopsy, graded according to Banff ’97 criteria (28).
The primary outcome of interest was graft failure, defined as return to dialysis or retransplantation, and the secondary endpoint was death. Retrieved data included recipient age, race, gender, pretransplant diabetes, hepatitis C serostatus, previous renal transplantation, peak PRA, donor source, donor/recipient human leukocyte antigen mismatch, and initial graft function. Hb, serum bicarbonate, and SCr values were obtained at 1 week and 1, 3, 6, 9, 12, 18, and 24 months posttransplant. Chronic Kidney Disease-Epidemiology Collaboration equations were used to calculate estimated GFR at these time points (29). Data were collected on TAC blood levels, use of ESA, supplemental iron, and ACEi/ARB at 1, 3, 6, 9, and 12, 18, and 24 months. MMF and prednisone doses at 1, 3, 6, and 12 months after transplantation were recorded. All biopsies performed during the study period were reviewed.
Multiple imputation method was used to estimate the missing values for Hb, TAC level, and MMF or prednisone dose. To estimate the missing values, regression models using corresponding values at the previous time point, age, gender, race, and retransplantation, donor source, induction agent, DGF, human leukocyte antigen mismatch, and P-PRA as the predictor variables were used and the average of 20 imputations were used for analyses.
Data are reported as mean±standard deviation. Comparison between the variables was performed using Student’s t test, as appropriate. The Kaplan-Meier method was used to estimate the unadjusted association of degree of anemia with graft and patient survival. Comparison between the groups was made using the log-rank test. Cox proportional hazard models were used to evaluate the independent association of Hb, measured at multiple time points, with the outcome after adjustment for other time-dependent and independent potential confounding factors. Schoenfeld residuals were calculated to assess the violation of proportionality assumption. Because Hb value as a continuous variable did not conform, it was categorized into five sex-specific groups. For graft survival analysis, in addition to standard Cox models with censoring at death, we used competing-risks regression. In this model, death was considered as an outcome competing with graft failure. Generalized estimating equations, using independent correlation with robust standard errors, were used to identify factors that were associated with Hb level.
Intercooled Stata 11.1 software package (Stata Corporation, College Station, TX) was used for statistical analysis.
. Horwich TB, Fonarow GC, Hamilton MA, et al.. Anemia
is associated with worse symptoms, greater impairment in functional capacity and a significant increase in mortality in patients with advanced heart failure. J Am Coll Cardiol 2002; 39: 1780.
. Tang WH, Tong W, Jain A, et al.. Evaluation and long-term prognosis of new-onset, transient, and persistent anemia
in ambulatory patients with chronic heart failure. J Am Coll Cardiol 2008; 51: 569.
. Ibrahim HN, Ishani A, Foley RN, et al.. Temporal trends in red blood transfusion among US dialysis patients, 1992-2005. Am J Kidney Dis 2008; 52: 1115.
. Heinze G, Kainz A, Horl WH, et al.. Mortality in renal transplant recipients given erythropoietins to increase haemoglobin concentration: Cohort study. BMJ 2009; 339: b4018.
. Kolonko A, Pinocy-Mańdok J, Kocierz M, et al.. Anemia
and erythrocytosis after kidney transplantation
: A 5-year graft function and survival analysis. Transplant Proc 2009; 41: 3046.
. Shah N, Al-Khoury S, Afzali B, et al.. Posttransplantation anemia
in adult renal allograft recipients: Prevalence and predictors. Transplantation 2006; 81: 1112.
. Shibagaki Y, Shetty A. Anaemia is common after kidney transplantation
, especially among African Americans. Nephrol Dial Transplant 2004; 19: 2368.
. Vanrenterghem Y, Ponticelli C, Morales JM, et al.. Prevalence and management of anemia
in renal transplant recipients: A European survey. Am J Transplant 2003; 3: 835.
. Petrone H, Arriola M, Re L, et al.. National survey of anemia
prevalence after kidney transplantation
in Argentina. Transplant Proc 2010; 42: 288.
. World Health Organization. Worldwide prevalence of anemia
1993–2005: WHO global database on anemia
, vol. 2011. In: de Benoist B, McLean E, Egli I, Cogswell M, eds. Geneva: World Health Organization; 2008.
. Mix TC, Kazmi W, Khan S, et al.. Anemia
: A continuing problem following kidney transplantation
. Am J Transplant 2003; 3: 1426.
. Yorgin PD, Scandling JD, Belson A, et al.. Late post-transplant anemia
in adult renal transplant recipients. An under-recognized problem? Am J Transplant 2002; 2: 429.
. Turkowski-Duhem A, Kamar N, Cointault O, et al.. Predictive factors of postrenal transplant anemia
. Transplant Proc 2005; 37: 1009.
. Fernandez Fresnedo G, Palomar R, Rodrigo E, et al.. Prevalence of anemia
in renal transplant patients: Results from MOST, an observational trial. Transplant Proc 2005; 37: 3821.
. Molnar MZ, Novak M, Ambrus C, et al.. Anemia
in kidney transplanted patients. Clin Transplant 2005; 19: 825.
. Lorenz M, Kletzmayr J, Perschl A, et al.. Anemia
and iron deficiencies among long-term renal transplant recipients. J Am Soc Nephrol 2002; 13: 794.
. Chadban SJ, Baines L, Polkinghorne K, et al.. Anemia
after kidney transplantation
is not completely explained by reduced kidney function. Am J Kidney Dis 2007; 49: 301.
. Chhabra D, Grafals M, Skaro AI, et al.. Impact of anemia
after renal transplantation on patient and graft survival
and on rate of acute rejection. Clin J Am Soc Nephrol 2008; 3: 1168.
. Winkelmayer WC, Kewalramani R, Rutstein M, et al.. Pharmacoepidemiology of anemia
in kidney transplant recipients. J Am Soc Nephrol 2004; 15: 1347.
. Imoagene-Oyedeji AE, Rosas SE, Doyle AM, et al.. Posttransplantation anemia
at 12 months in kidney recipients treated with mycophenolate mofetil: Risk factors and implications for mortality. J Am Soc Nephrol 2006; 17: 3240.
. Zakai NA, McClure LA, Prineas R, et al.. Correlates of anemia
in American blacks and whites: The REGARDS Renal Ancillary Study. Am J Epidemiol 2009; 169: 355.
. Kamar N, Rostaing L. Negative impact of one-year anemia
on long-term patient and graft survival
in kidney transplant patients receiving calcineurin inhibitors and mycophenolate mofetil. Transplantation 2008; 85: 1120.
. Winkelmayer WC, Lorenz M, Kramar R, et al.. Percentage of hypochromic red blood cells is an independent risk factor for mortality in kidney transplant recipients. Am J Transplant 2004; 4: 2075.
. Ghafari A, Noori-Majelan N. Anemia
among long-term renal transplant recipients. Transplant Proc 2008; 40: 186.
. Sancho A, Pastor MC, Bayés B, et al.. Posttransplant inflammation associated with onset of chronic kidney disease. Transplant Proc 2010; 42: 2896.
. Molnar MZ, Czira ME, Rudas A, et al.. Association between the malnutrition-inflammation score and post-transplant anaemia. Nephrol Dial Transplant 2011; 26: 2000.
. Kasiske L, Vazquez MA, Harmon WE, et al.. Recommendations for the outpatient surveillance of renal transplant recipients. American Society of Transplantation. J Am Soc Nephrol 2000; 11 (suppl 15): S1.
. Racusen LC, Solez K, Colvin RB, et al.. The Banff 97 working classification of renal allograft pathology. Kidney Int 1999; 55: 713.
. Levey AS, Stevens LA, Schmid CH, et al.. A new equation to estimate glomerular filtration rate. Ann Intern Med 2009; 150: 604.