Skip to main content

Relationship between variations in time-dependent response to erythropoiesis-stimulating agents and mortality in hemodialysis patients: a single-center study



Hemodialysis (HD) patients who are resistant to erythropoiesis-stimulating agents (ESAs) have a higher mortality rate. The aim of this study was to assess the relationship between the time-dependent variability of ESA response in HD patients and mortality.


A total of 375 HD patients were enrolled in this study. The ESA resistance index (ERI) was calculated by dividing the weekly weight-adjusted ESA dose by hemoglobin concentration, and the average ERI was calculated from the ERI values every 2 months during a 36-month follow-up period. We divided the patients into six groups based on the patterns of their ERI level fluctuations: low-low (Low group), intermediate-intermediate (Intermediate group), high-high (High group), Low-intermediate group, Intermediate-high group, and Low-high group.


There were 94 (25.1 %) deaths, and they included 51 (13.6 %) deaths from cardiovascular disease (CVD). A multivariate analysis with adjustment for age, serum albumin and C-reactive protein levels, and history of CVD showed that the High group was independently and significantly related to all-cause mortality (odds ratio = 5.53, 95 % CI 2.29–13.99, p = 0.0001) and CVD-related mortality (odds ratio = 3.49, 95 % CI 1.40–8.47, p = 0.0081).


High ERI levels are an independent risk factor for all-cause and CVD mortality in HD patients.


Renal anemia in end-stage renal disease (ESRD) patients is mainly attributable to decreased erythropoietin production by the kidney. Erythropoiesis-stimulating agent (ESA) therapy has resulted in substantial health benefits for ESRD patients, including improved quality of life, a reduced need for blood transfusion, and greater exercise capacity [1]. Unfortunately, a considerable proportion of ESRD patients exhibit a suboptimal hematologic response to ESAs, as evidenced by persistence of the anemia despite adequate dosing or by the need for high-dose ESA therapy to achieve the recommended hemoglobin target [2].

After making adjustments for other variables, previous studies have shown that the need for ESA therapy is an independent predictor of all-cause mortality in HD patients [3]. Lopez-Gomez et al. reported finding a relationship between ESA resistance in HD patients and 1-year mortality [4], and ESA resistance remained a significant predictor of mortality in HD patients [5, 6]. It appears that resistance to ESA therapy in HD patients could be clinically useful as an index of prognostic marker.

However, the time-dependent changes of ESA resistance index (ERI) has not examined, although consistently high ESA resistance was expected to be an adverse effect on mortality of HD patients. The aim of our study was to assess the presence of variability in the time course of ERI and the association between high ERI values and increased mortality of HD patients.


Subjects and protocol

This was a prospective, observational cohort study conducted at a single center in Japan. The subjects were recruited from among patients who had been routinely treated through an arteriovenous fistula in the dialysis unit of the Jyoban Hospital for at least 6 months. A total of 400 HD patients were registered in this study, and the 375 patients for whom sufficient data were available to calculate ERI were the subjects of the analysis in this study. The Institutional Review Board of the Jyoban Hospital approved all study protocols (No. 12), and the protocols were carried out in accordance with the Declaration of Helsinki guidelines regarding ethical principles for medical research involving human subjects. Written informed consent was obtained from every subject.

Baseline data at the time of patient entry into the study were collected on demographics, medical history, laboratory data, anemia therapy, medications, and history of cardiovascular disease (CVD). Data were collected at 2-month intervals on events, including all-cause deaths, and on CVD mortality. A CVD death was defined as a death that was attributed to acute myocardial infarction, ischemic heart disease, cardiac arrest, congestive heart failure, pulmonary edema due to exogenous fluid, cardiac arrhythmia, cerebrovascular accident, including intracranial hemorrhage or ischemic brain damage, and peripheral artery disease. The patients who were enrolled as subjects (n = 375) underwent stable regular HD with a bicarbonate dialysate. Their underlying diseases of ESRD were diabetic nephropathy (n = 162), chronic glomerulonephritis (n = 123), nephrosclerosis (n = 78), and chronic pyelonephritis (n = 4), or unknown origin (n = 8).

Subjects’ anemia was treated according to the guidelines published in 2008 by the Japanese Society of Dialysis Therapy [7]. The target Hb level was 10–11 g/dl, and epoetin beta was administered to achieve the target Hb level. Intravenous iron was administered whenever a subject’s transferrin saturation was found to be <20 % or the serum ferritin level was found to be <100 ng/ml. The conversion ratio adopted in this study was 200 epoetin (EPO):1 darbepoetin-α (DA) [8]. ERI was calculated as the weekly weight-adjusted dose of ESA divided by hemoglobin (Hb) concentration used to evaluate the dose-response effect in ESA treatment as follows [5, 6]: ERI (U/kg/week/g per 100 ml) = weekly ESA dose / (weight × Hb). Ascending quartiles of ERI were assessed based on the ERI levels at enrollment, and the ERI quartile groups were as follows: first, <5.1; second, 5.1 to <8.5; third, 8.5 to <13.7; and fourth, ≥ 13.7 U/kg/week/g per 100 ml. The definition of variability in the time course of ERI is shown in Fig. 1. ERI levels were categorized as low (first quartile of ERI), intermediate (second and third quartiles of ERI), and high (fourth quartile of ERI). The average ERI was calculated from ERI levels every 2 months during the follow-up period for each patient. To assess the size of the ERI fluctuations during the follow-up periods according to ERI quartiles at baseline, we divided the patients into six groups on the basis of their overall fluctuation pattern with reference to a previous report on Hb fluctuations and mortality [9]: low-low (Low group), intermediate-intermediate (Intermediate group), high-high (High group), Low-intermediate group, Intermediate-high group, and Low-high group, and the last three groups were treated as amplitude fluctuation groups.

Fig. 1
figure 1

Definition of variability in the time course of ERI

Statistical analyses

All data are reported as means ± SD unless otherwise specified. The differences between groups were tested for statistical significance by Student’s t test or one-way ANOVA for continuous variables, and by the χ 2 test for categorical variables. First, we evaluated the associations between the six ERI variation groups and cumulative all-cause and CVD-related mortality during the follow-up period by the Kaplan-Meier method. Next, we estimated the odds ratios (ORs) of all-cause and CVD-related mortality according to the ERI variability groups by the logistic regression. Logistic regression was also used to assess associations between patient characteristics and all-cause and CVD-related mortality. The candidates for explanatory variables were age, sex, dialysis vintage, body mass index (BMI), the presence of diabetes, mean serum albumin (≥3.5 vs. <3.5 g/dl), mean C-reactive protein (CRP ≥ 1.0 vs. <1.0 mg/dl), ferritin (≥100 vs. <100 ng/ml), phosphorus (≥6 vs. <6 mg/dl), and history of CVD (yes vs. no or missing). All significant variables in the univariate analysis (p < 0.20) and clinically established factors were entered into an intermediate multivariate logistic model of mortality, and all of the significant variables detected in the intermediate multivariate model and the established factors were entered into a final multivariate logistic mortality model. Statistical significance was defined as p < 0.05. All analyses were performed using the SAS version 9.1 (SAS Institute, Cary, NC, USA).


The data of a total of 375 Japanese HD patients were analyzed, and their mean follow-up period was 31.3 ± 9.4 weeks. Patient characteristics at baseline are summarized in Table 1. The mean ESA doses during the follow-up period calculated in the form of EPO doses were 5444 ± 3538 IU/week in all patients, 5473 ± 3700 IU/week in males, and 5392 ± 3239 IU/week in females. One hundred sixty-three patients were treated with EPO, whereas 212 patients were treated with DA during the follow-up period. The ERI distribution pattern did not appear to differ between the sexes, but the average ERI was significantly higher in females than in males (12.1 ± 8.4 vs. 9.6 ± 7.3 U/kg/week/g per 100 ml, p < 0.001).

Table 1 Baseline characteristics of the study subjects

The patients were divided into the following six ERI variability groups based on their ERI quartiles levels: High group, n = 31 (8.3 %); Intermediate group, n = 68 (18.1 %); Low group, n = 46 (12.3 %); Intermediate-high group, n = 30 (8.0 %); Low-intermediate group, n = 63 (16.8 %); and Low-high group, n = 137 (36.5 %), and the characteristics of the six groups are shown in Table 2. Intravenous iron administration was not performed in the three patients of High group, seven of Intermediate group, six of Intermediate-high group, four of Low-intermediate group, and three of Low-high group. All of the other patients were administered with intravenous iron less than 50 mg/week.

Table 2 Characteristics of the six groups for ERI variability

There were 94 deaths (25.1 %) during the follow-up period. The cause of death was CVD in 51 cases (13.6 %), infection in 23 cases (6.1 %), cancer in 5 cases (1.3 %), and gastrointestinal bleeding in 4 cases (1.1 %). Of the 51 deaths due to CVD, 29 were due to congestive heart failure, 8 were due to cerebrovascular accident including intracranial hemorrhage, and 5 were due to acute myocardial infarction; there were 4 cases of pulmonary edema due to exogenous fluid, 2 cases of death due to cardiac arrhythmia, and 3 cases of sudden death. We plotted the all-cause and CVD-related mortality curves of the six ERI variability groups by the Kaplan-Meier method (Fig. 2a, b), and the all-cause and CVD-related mortality rates were highest in the High ERI group. There was no significant difference in all-cause mortality between EPO group and DA group (Fig. 3, p = 0.7680). Moreover, there was no significant difference between each ERI groups and all-cause mortality in patients treated with EPO (Fig. 4a, p = 0.3189). However, High group had significantly higher mortality risk than those in the other ERI groups in patients treated with DA (Fig. 4b, p < 0.0001).

Fig. 2
figure 2

Kaplan-Meier survival curves of the six ERI variability groups. a All-cause. b CVD-associated

Fig. 3
figure 3

Comparison of mortality risk in patients treated with epoetin (EPO) and darbepoetin (DA)

Fig. 4
figure 4

Relationship between each ERI groups and all-cause mortality in patients treated with epoetin (EPO, a) and darbepoetin (DA, b)

Figure 5 shows the unadjusted ORs for all-cause and CVD-related mortality in each ERI variability group. The risk for all-cause mortality in the High group was significantly higher than those in the other groups (OR = 10.13, 95 % CI 3.56–32.07, p = 0.001). The High group was also significantly associated with CVD-related mortality (OR = 3.67, 95 % CI 1.22–12.03, p = 0.003).

Fig. 5
figure 5

Odds ratio for all-cause and CVD-associated mortality of the six ERI variability groups. a All-cause. b CVD-associated

The results of the multivariate logistic regression model adjusted for the age, sex, dialysis vintage, BMI, the presence of diabetes, ferritin, albumin, CRP, phosphorus, and history of CVD showed that the High group was at significantly higher risk for all-cause mortality (OR = 3.63, 95 % CI 2.03–6.28, p < 0.0001; Table 3). For CVD-related mortality, the High group had significantly higher risk adjusted for the above variables except for phosphorus (OR = 3.81, 95 % CI 1.68–8.04, p = 0.0019). Older age, male, diabetes, and serum albumin levels were independent predictors for all-cause mortality, whereas only older age was an independent predictor for CVD-related mortality.

Table 3 Independent factors associated with all-cause and CVD-related mortality in multivariate logistic regression analysis

Finally, we examined factors related to the High group. The High group was associated with older age and had a lower BMI, Hb and albumin, lower iron dosage, and a higher ferritin and CRP levels (Table 4).

Table 4 Factors associated with high ERI variability


The results of this study showed that the High group, which had consistently high ERI values, was the highest risk group for all-cause and CVD-related mortality in Japanese HD patients compared with three amplitude fluctuation groups. These findings were consistent with the results of a previous study by Fujikawa et al. [10] who showed that factoring out fluctuating ERI increases the ability of consistently high ERI levels as an independent risk factor for all-cause and CVD mortality of 2104 HD patients in the Dialysis Outcomes and Practice Patterns Study (DOPPS). Therefore, consistently high ERI values in HD patients are a strong independent risk factor for all-cause and CVD-related mortality.

The High group in our study was found to be associated with older age, female gender, low BMI, and low serum albumin levels, and these findings are consistent with the results of previous studies [3, 4, 11, 12]. However, CRP was not found to be associated with the High group, although inflammatory conditions can elevate the ERI by inducing the production of cytokines that suppress erythropoiesis [1315]. This was probably because the Low group, which had consistently low ERI values, had an unexpectedly high CRP levels. Examination of high CRP in the Low and High groups will provide an information as to how factors affecting inflammation elevate ERI. Schneider et al. [16] recently reported finding that time-varying ESA resistance was associated with higher rates of sudden death, infectious complications, and all-cause mortality in the German Diabetes and Dialysis study (4D study). The factors predicting ESA resistance may be different from the study cohort.

The ERI values in the present study were lower than in a previous study [4]. This is in part due to the fact that ESA doses in Japan are lower than elsewhere [17], and the low ERI values may predict a favorable outcome in Japanese HD patients [18]. If higher ESA doses were available in the Japanese healthcare system, HD patient conditions causing high resistance to ESA might have higher ERI values. ERI of a less limited range may be expected to predict poor prognosis. Further study of ERI variability should be performed to determine the impact of the dose range. For a clinical understanding of ERI, it would be beneficial to examine how low ERI is affected by a limited range of ESA dosing.

According to the literature, 90 % of renal anemia patients respond to EPO in a dose-dependent manner, whereas the remaining 5–10 % of patients have either a blunted or absent response to ESAs, despite high-dose administration [19]. High ERI variability may be associated with hyporesponsiveness to ESA. Of the various factors that affect patient responsiveness to an ESA, iron deficiency is thought to be the most prevalent cause of ESA hyporesponsiveness [20]. Intravenous iron administration has been shown to improve ESA responsiveness, leading to an increased Hb concentration and permitting the use of a decreased ESA dose [21]. Kuragano et al. have recently shown that iron administration was associated with an increased incidence for adverse events, including CVD, infection, and hospitalization, among HD patients who were treated with high doses of intravenous iron compared with HD patients who were not treated with iron [22]. In the present study, intravenous iron was frequently administered to the high ERI group who had the highest ferritin levels, suggesting the presence of ESA hyporesponsiveness that is associated with an increased mortality risk in high ERI group.

The present study had several limitations that were mainly due to the observational nature of the study. First, since this was a single-center cohort study, the data and results may not be representative of HD patients in other countries, particularly because ESA doses and Hb concentrations are generally lower in Japan than in other countries. However, the mortality rate and other baseline characteristics of our cohort were similar to those of Japanese dialysis patients in general [23]. Second, the longitudinal ERI assessments were based on data collected at 2-month intervals. Since using different measurement frequencies might make it possible to increase the prognostic potential of ERI variability, the most appropriate measurement frequencies and duration should be identified in subsequent investigations. Third, the average dose of EPO in Japan is lower than in other countries because of the healthcare expense reimbursement policy of the national health insurance system in Japan. Until April 2006, the maximum dose of EPO was limited to 9000 U/week of EPO alfa or EPO beta. Fourth, it is unknown why the patients treated with DA in High group had higher mortality risk. ESA selection bias by nephrologists may be associated with the results. Moreover, DA might be administered when response to EPO was insufficient for anemia treatment.


The results of this study demonstrated that consistently high ERI values in HD patients are an independent risk factor for all-cause and CVD mortality. Further study will be necessary to obtain a more detailed understanding of the mechanisms that underlie the causal relationship between ERI and mortality.


  1. Eschbach JW, Egrie JC, Downing MR, Browne JK, Adamson JW. Correction of the anemia of end-stage renal disease with recombinant human erythropoietin. Results of a combined phase I and II clinical trial. N Engl J Med. 1987;316:73–8.

    Article  CAS  PubMed  Google Scholar 

  2. Macdougall IC, Cooper A. The inflammatory response and epoetin sensitivity. Nephrol Dial Transplant. 2002;17 Suppl 1:48–52.

    Article  PubMed  Google Scholar 

  3. Zhang Y, Thamer M, Stefanik K, Kaufman J, Cotter DJ. Epoetin requirements predict mortality in hemodialysis patients. Am J Kidney Dis. 2004;44:866–76.

    Article  CAS  PubMed  Google Scholar 

  4. Lopez-Gomez JM, Portoles JM, Aljama P. Factors that condition the response to erythropoietin in patients on hemodialysis and their relation to mortality. Kidney Int. 2008;74:S75–81.

    Article  Google Scholar 

  5. Okazaki M, Komatsu M, Kawaguchi H, Tsuchiya K, Nitta K. Erythropoietin resistance index and the all-cause mortality of chronic hemodialysis patients. Blood Purif. 2014;37:106–12.

    Article  CAS  PubMed  Google Scholar 

  6. Ogawa T, Shimizu H, Kyono A, eSato M, Yamashita T, Otsuka K, et al. Relationship between responsiveness to erythropoietin-stimulating agent and long-term outcomes in chronic hemodialysis patients: a single center cohort study. Int Urol Nephrol. 2014;46:151–9.

    Article  CAS  PubMed  Google Scholar 

  7. Tsubakihara Y, Nishi S, Akiba T, Hirakata H, Iseki K, Kubota M, et al. 2008 Japanese Society for Dialysis Therapy: guidelines for renal anemia in chronic kidney disease. Ther Apher Dial. 2010;14:240–75.

    Article  CAS  PubMed  Google Scholar 

  8. Locatelli F, Olivares J, Walker R, Wilkie M, Jenkins B, Dewey C, et al. Novel erythropoiesis stimulating protein for treatment of anemia in chronic renal insufficiency. Kidney Int. 2001;60:741–7.

    Article  CAS  PubMed  Google Scholar 

  9. Ebben JP, Gilbertson DT, Foley RN, Collins AJ. Hemoglobin level variability: associations with comorbidity, intercurrent events, and hospitalizations. Clin J Am Soc Nephrol. 2006;1:1205–10.

    Article  PubMed  Google Scholar 

  10. Fujikawa T, Ikeda Y, Fukuhara S, Akiba T, Akizawa T, Kurokawa K, et al. Time-dependent resistance to erythropoiesis-stimulating agent and mortality in hemodialysis patients in the Japan Dialysis Outcomes and Practice Patterns Study. Nephron Clin Pract. 2012;122:24–32.

    Article  CAS  PubMed  Google Scholar 

  11. Panichi V, Rosati A, Bigazzi R, Paoletti S, Mantuano E, Beati S, et al. Anaemia and resistance to erythropoiesis-stimulating agents as prognostic factors in haemodialysis patients: results from the RISCAVID study. Nephrol Dial Transplant. 2011;26:2641–8.

    Article  CAS  PubMed  Google Scholar 

  12. Mallick S, Rafiroiu A, Kanthety R, Iqbal S, Malik R, Rahman M. Factors predicting erythropoietin resistance among maintenance hemodialysis patients. Blood Purif. 2012;33:238–44.

    Article  CAS  PubMed  Google Scholar 

  13. Johnson DW, Pollock CA, Macdougall IC. Erythropoiesis-stimulating agent hyporesponsiveness. Nephrology (Carlton). 2007;12:321–30.

    Article  CAS  Google Scholar 

  14. Kanbay M, Perazella MA, Kasapoglu B, Koroglu M, Covic A. Erythropoiesis stimulatory agent-resistant anemia in dialysis patients: review of causes and management. Blood Purif. 2010;29:1–12.

    Article  PubMed  Google Scholar 

  15. Bamgbola O. Resistance to erythropoietin-stimulating agents: etiology, evaluation, and therapeutic considerations. Pediatr Nephrol. 2012;27:195–205.

    Article  PubMed  Google Scholar 

  16. Schneider A, Gutjahr-Lengsfeld L, Ritz E, Scharnagl H, Gelbrich G, Pilz S, et al. Longitudinal assessment of erythropoietin-stimulating agent responsiveness and the association with specific clinical outcomes in dialysis patients. Nephron Clin Pract. 2014;128:147–52.

    Article  CAS  PubMed  Google Scholar 

  17. Nakai S, Hanafusa N, Masakane I, Taniguchi M, Hamano T, Shoji T, et al. An overview of regular dialysis treatment in Japan (as of 31 December 2008). Ther Apher Dial. 2014;18:535–602.

    Article  PubMed  Google Scholar 

  18. Locatelli F, Pisoni RL, Combe C, Bommer J, Andreucci VE, Piera L, et al. Anaemia in haemodialysis patients of five European countries: association with morbidity and mortality in the Dialysis Outcomes and Practice Patterns Study (DOPPS). Nephrol Dial Transplant. 2004;19:121–32.

    Article  PubMed  Google Scholar 

  19. Priyadarshi A, Shapiro JI. Erythropoietin resistance in the treatment of the anemia of chronic renal failure. Semin Dial. 2006;19:273–8.

    Article  PubMed  Google Scholar 

  20. Ogawa T, Nitta K. Erythropoiesis-stimulating agent hyporesponsiveness in end-stage renal disease patients. Contrib Nephrol. 2015;185:76–86.

    PubMed  Google Scholar 

  21. Kapoian T, O’Mara NB, Singh AK, Moran J, Rizkala AR, Geronemus R, et al. Ferric gluconate reduces epoetin requirements in hemodialysis patients with elevated ferritin. J Am Soc Nephrol. 2008;19:372–9.

    Article  PubMed Central  CAS  PubMed  Google Scholar 

  22. Kuragano T, Matsumura O, Matsuda A, Haga T, Kiyomoto H, Murata T, et al. Association between hemoglobin variability, serum ferritin levels, and adverse events/mortality in maintenance hemodialysis patients. Kidney Int. 2014;86:845–54.

    Article  CAS  PubMed  Google Scholar 

  23. Fukuma S, Yamaguchi T, Hashimoto S, Nakai S, Iseki K, Tsubakihara Y, et al. Erythropoiesis-stimulating agent responsiveness and mortality in hemodialysis patients: results from a cohort study from the dialysis registry in Japan. Am J Kidney Dis. 2012;59:108–16.

    Article  CAS  PubMed  Google Scholar 

Download references


The authors are very grateful to Mr. Toshiaki Naganuma of Jyoban Hospital who understood the clinical importance of this study and provided high-quality data. This study was in part supported by a grant from the Japan Promotion Society for Cardiovascular Disease.

Author information

Authors and Affiliations


Corresponding author

Correspondence to Kosaku Nitta.

Additional information

Competing interests

The authors declare that they have no competing interests.

Authors’ contributions

SS planned the study, searched the literature, assessed studies, extracted data, analyzed data, and prepared article. MK, MO, and TN assessed studies and assisted in the data analysis. HKat and KT searched the literature, assessed studies and assisted in article preparation. HKaw searched the literature and assisted in article preparation. KN planned the study, analyzed the data and assisted in article preparation. All authors read and approved the final manuscript.

Rights and permissions

Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (, which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver ( applies to the data made available in this article, unless otherwise stated.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Shiohira, S., Tsuchiya, K., Kataoaka, H. et al. Relationship between variations in time-dependent response to erythropoiesis-stimulating agents and mortality in hemodialysis patients: a single-center study. Ren Replace Ther 2, 10 (2016).

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: