Abstract
Objective
Sufficient iron substitution leads to a decrease in the required recombinant human erythropoietin (rHuEPO) dose and/or an increased hematocrit in dialysis patients. Intravenous (IV) application of larger doses of iron sucrose may be associated with hyperferritinemia, appearance of catalytically free iron, and impaired phagocyte function. Therefore, we investigated the effectiveness of a low-dose IV iron regimen in peritoneal dialysis (PD) patients.
Patients and Interventions
Forty-five PD patients were followed over a period of 1 year. Serum ferritin, serum transferrin saturation, and hemoglobin were measured monthly. In cases of absolute iron deficiency (serum ferritin < 100 μg/L), 50 mg iron sucrose was given IV every second week. In cases of functional iron deficiency (ferritin ≥ 100 μg/L and transferrin saturation < 20%) and in iron repleted patients (ferritin ≥ 100 μg/L and transferrin saturation ≥ 20%), 50 mg IV iron sucrose was applied monthly. Iron therapy was stopped in cases of acute infection (until complete recovery) and when serum ferritin level was ≥ 600 μg/L.
Results
To analyze the influence of iron substitution on erythropoiesis and rHuEPO requirements, the EPO resistance index (ERI; quotient of rHuEPO dose in units/kilogram/week and hemoglobin in grams per deciliter) was calculated every 3 months. The ERI decreased significantly during the course of the study in the whole patient group (p = 0.009) as well as in the subgroup of 21 patients with absolute iron deficiency (p = 0.01). A nonsignificant decrease in the ERI was observed within the group of 14 iron repleted patients (p = 0.5). There was no significant change in the ERI in 10 patients with functional iron deficiency (p = 0.6).
Conclusion
The low-dose IV iron regimen used in this study substantially decreased rHuEPO requirements in patients with absolute iron deficiency and was effective in maintaining iron stores in iron repleted patients. However, in the absence of significant hyperparathyroidism, aluminum toxicity, or inadequate dialysis, it did not improve the ERI in patients with functional iron deficiency.
Renal anemia is a frequent complication in patients suffering from end-stage renal failure. Therapy of renal anemia with recombinant human erythropoietin (rHuEPO) requires adequate iron stores (1). Iron deficiency is the most important cause of rHuEPO hyporesponsiveness in dialysis patients. Furthermore, sufficient intravenous (IV) iron substitution leads to a decrease in required rHuEPO doses and/or an increased hematocrit in hemodialysis as well as peritoneal dialysis (PD) patients (2-10). Various iron treatment recommendations have been published during the past years (2-7,10,11). High-dose IV single-bolus iron dextran therapy is effective and convenient for PD patients, since no additional visits to the center are necessary (5). However, IV application particularly of larger bolus doses of iron sucrose or iron gluconate may be associated with the risk of hyperferritinemia (6) and the appearance of catalytically active serum iron (12). In this study we investigated the efficacy of a low-dose IV iron regimen in PD patients, using 50 mg iron sucrose once or twice monthly according to the patient's iron status.
Patients and Methods
Forty-five PD patients [median age 51 years, interquartile range (IQR) 43 – 63 years; 19 females, 26 males] participated in this study. Median duration of dialysis before the start of the observation period was 4.2 months (IQR 1.3 – 13.5). Patient characteristics are shown in Table 1. Underlying diseases were diabetic nephropathy (n = 10, 22%), shrunken kidneys of unknown origin (n = 9, 20%), chronic glomerulonephritis (n = 7, 16%), polycystic kidney disease (n = 7, 16%), chronic interstitial nephritis (n = 6, 13%), and vascular nephropathy (n = 6, 13%). Three of the 19 female patients had regular menstruation. None of the patients displayed hematological or malignant diseases, hemolysis, or vitamin B12 or folate deficiency. None of the patients received IV iron or erythrocyte transfusions within 12 months prior to the study. Nine of the 45 patients took oral iron (100 – 200 mg iron sulfate/day) before the observation period started. Thirty-two patients received rHuEPO during the whole study, 9 patients during part of the observation period, 4 patients did not receive rHuEPO.
Patient Characteristics. Median (Interquartile Range 25%–75%)
Chi-square test.
Kruskal Wallis test.
Iron Status
According to the European Best Practice Guidelines (13), we defined absolute iron deficiency as serum ferritin less than 100 μg/L. Functional iron deficiency was defined as serum ferritin greater than or equal to 100 μg/L and transferrin saturation less than 20%. Patients with serum ferritin of 100 μg/L or more and transferrin saturation 20% or more were categorized as iron repleted.
Iron Therapy Regimen
The individual iron dose was adapted monthly, dependent on serum ferritin. In cases of absolute iron deficiency, 50 mg iron sucrose (Venofer; Vifor International AG, St. Gallen, Switzerland) was given IV every second week for as long as serum ferritin remained below 100 μg/L. If ferritin was greater than or equal to 100 μg/L, 50 mg IV iron sucrose was applied monthly (independent of transferrin saturation). Each 50-mg iron dose was injected undiluted in 5 minutes without prior test dose. No single dose higher than 50 mg was given. Intravenous iron therapy was stopped in cases of acute infection (until complete recovery) and when serum ferritin level was greater than or equal to 600 μg/L (until it decreased to < 600 μg/L). The observation period was 1 year.
Laboratory Investigations
Serum ferritin, transferrin saturation, serum iron, and hemoglobin were measured monthly.
To analyze the influence of iron substitution more exactly, the EPO resistance index (ERI; quotient of rHuEPO dose, in units/kilogram/week, and hemoglobin, in grams per deciliter) was calculated every 3 months as described recently (14). Response to IV iron therapy was characterized by an increased hemoglobin and/or decrease in the required rHuEPO dose and, therefore, a decreased ERI.
Serum C-reactive protein (CRP), serum intact parathyroid hormone (iPTH), and weekly Kt/V urea were calculated every 3 months. Serum aluminum levels were measured once during the observation period. Standard laboratory methods were used for all measurements. Since some patients showed increases or decreases in iPTH, CRP, or Kt/V urea during the observation period, the average of 5 measurements (baseline, months 3, 6, 9, and 12) was calculated for these parameters.
Statistical Methods
Continuous variables are presented as the median and interquartile range (IQR; range from the 25th to the 75th percentile). Percentages were calculated for dichotomous variables. The Kruskal Wallis test (which is adequate for analysis of unequal sample sizes) was used to compare continuous variables at baseline between the three patient groups according to iron status. The chi-square test was applied to compare proportions. The nonparametric Friedman test was used for analysis of repeated measurements to compare the course of continuous parameters within the observation period. The Friedman test was chosen instead of ANOVA because of the small sample size in the patient subgroups. All p values are two-sided; a p value less than 0.05 was considered statistically significant. Calculations were performed with SPSS for MS Windows (version 10.0; SPSS Inc., Chicago, Illinois, USA).
Results
Patient Characteristics and Laboratory Results
At the beginning of the study, 21 patients had absolute iron deficiency, 10 patients had functional iron deficiency, and 14 patients were iron repleted. There were no statistically significant differences in patient characteristics (Table 1) or in laboratory results (Table 2) between the three patient subgroups.
Laboratory Results. Median (Interquartile Range 25%–75%)
iPTH = intact parathyroid hormone; CRP = C-reactive protein.
Kruskal Wallis test.
During the observation period, patients with absolute iron deficiency received a median dose of 750 mg (IQR 600 – 950 mg) iron sucrose; patients with functional iron deficiency received a median dose of 600 mg (IQR 563 – 638 mg) iron sucrose; the median iron sucrose dose for iron repleted patients was 525 mg (IQR 463 – 600 mg).
Serum Ferritin and Transferrin Saturation
Table 3 shows a significant increase in serum ferritin during the observation period (1 year) in the whole group of 45 patients as well as in the subgroup of patients with absolute iron deficiency. Serum ferritin did not change significantly in the other two subgroups. In contrast to iron repleted patients, transferrin saturation increased between baseline and month 6 in patients with absolute and those with functional iron deficiency. Thereafter, transferrin saturation declined in both groups, but this decrease was more pronounced in the functional iron deficiency group. However, during the whole course of the study, serum transferrin saturation did not change significantly in either group.
Change in Iron Status (Serum Ferritin and Transferrin Saturation) During the Observation Period. Median (Interquartile Range 25%–75%).
Friedman test.
Hemoglobin Values and Rhuepo Requirements
Median rHuEPO dose decreased significantly in the whole patient group (p = 0.04) (Table 4). This was due mainly to a significant reduction in rHuEPO dose in the group of patients with absolute iron deficiency (p = 0.01). rHuEPO requirements did not change significantly in the other two patient groups (Table 4). Hemoglobin values increased slightly in the whole patient group as well as in each subgroup of patients, without reaching levels of statistical significance (Table 4).
Change of Erythropoietin Dose and Serum Hemoglobin Level During the Observation Period. Median (Interquartile Range 25%–75%).
Friedman test.
The ERI was calculated every 3 months. There was a significant decrease in the ERI during the course of the study in the whole patient group (at baseline 7.2, IQR 3.4 – 12.9; at 6 months 6.0, IQR 3.1 – 11.3; at 12 months 5.3, IQR 2.6 – 10.6; Friedman test p = 0.009). Figure 1 depicts the ERI in the different subgroups of patients, dependent on iron status. Patients with absolute iron deficiency showed a significant decline in ERI during the study period (at baseline 11.5, IQR 6.2 – 13.1; at 6 months 6.3, IQR 4.1 – 11.8; at 12 months 5.4, IQR 2.8 – 10.3; Friedman test p = 0.01). The ERI also decreased in iron repleted patients, but this decrease was slower than in patients with absolute iron deficiency and did not reach levels of significance (at baseline 5.4, IQR 0.0 – 9.8; at 6 months 3.3, IQR 0.6 – 7.4; at 12 months 3.7, IQR 2.2 – 5.9; Friedman test p = 0.5). Although patients with functional iron deficiency showed an increase in the ERI at 6 and 9 months, there was no significant change in the ERI during the whole observation period of 1 year (at baseline 7.8, IQR 4.8 – 12.8; at 6 months 9.8, IQR 4.2 – 12.0; at 12 months 8.9, IQR 5.6 – 11.2; Friedman test p = 0.6) (Figure 1).
Course of erythropoietin resistance index [ERI; quotient of rHuEPO dose (U/kg/week) and hemoglobin (g/dL)] in 45 peritoneal dialysis patients within the observation period (baseline, months 3, 6, 9, and 12) of 1 year, according to iron status grouping. Box plots indicate median, interquartile range, and range. Absolute iron deficiency = serum ferritin < 100 μg/L; functional iron deficiency = serum ferritin ≥ 100 μg/L and transferrin saturation < 20%; iron repleted = serum ferritin ≥ 100 μg/L and transferrin saturation ≥ 20%.
Analysis of data after exclusion of the 9 patients who received oral iron before the start of the study revealed similar results compared to the whole group (data not shown).
Adverse Events
Low-dose IV iron application was well tolerated. Systemic side effects occurred in only 0.5% of 594 applications (hypotension, nausea, and flush after one application, respectively). Local pain at the injection site was reported during two applications (0.3%).
Discussion
Sufficient iron substitution decreases rHuEPO requirements and improves renal anemia in PD patients (2,4-10). The most suitable kind of iron supplementation in this patient group is still a matter of debate. Some authors prefer single high-dose bolus therapy of IV iron dextran (5) or iron sucrose (6). These regimens are effective and can be performed during one or two single visits. However, IV application, especially of higher iron sucrose doses, can result in hyperferritinemia (6), which may be related to impaired phagocyte function (15).
Furthermore, IV bolus iron sucrose therapy is associated with the appearance of catalytically active serum iron. Recently, Parkkinen et al. reported that catalytically active, non transferrin-bound iron, as well as reduced inhibition of bacterial growth, in serum of dialysis patients occurred after IV application of 100 mg iron sucrose (12). Non transferrin-bound iron occurs to a markedly smaller degree after application of iron sucrose doses of 50 mg or less (16). The clinical importance of these findings with respect to a possibly increased risk of infection during IV iron therapy is unclear. In a recent study, Prakash et al. found an increase (nonsignificant) in the number of peritonitis episodes within 6 months after 500 mg bolus iron (dextran or sucrose) infusion (9). During an observation period of 6 months we did not find increased catheter infection or peritonitis rates in PD patients who received 100 – 200 mg IV iron sucrose monthly (7). In a recent study, IV iron dextran supplementation (0.5 – 1 g) in PD patients did not result in higher peritonitis rates within 90 days after application compared to infection rates of patients who did not receive IV iron (17). Nevertheless, it remains essentially important to find the lowest dose of iron that is still sufficient to decrease required rHuEPO doses to an acceptable level. Our results using a low-dose iron regimen (no single dose greater than 50 mg) show a significant decrease in the ERI in a group of 45 PD patients. Patients with ferritin levels greater than or equal to 100 μg/L received IV iron every 4 weeks. Since routine control examinations in most PD centers are performed monthly, this could easily be done during these visits. Patients with ferritin less than 100 μg/L had to make one additional visit per month. Since all included patients lived a rather short distance from the center, these additional visits were well accepted. Furthermore, employed patients were allowed to come in after work in the afternoon or evening.
Low-dose iron substitution was most effective in patients with absolute iron deficiency. During the observation period of 1 year, serum ferritin increased significantly in this subgroup. Furthermore, the ERI decreased 50% in patients with absolute iron deficiency. Median rHuEPO dose after 1 year was 61.9 U/kg/week in this group. This corresponds well to the required rHuEPO doses after sufficient iron substitution in some other studies (6,7), but is still higher than those reached in iron repleted patients in this study.
In iron repleted patients, there was a nonsignificant decrease in the ERI. Domrongkitchaiporn et al. (6) reported increased hemoglobin values without change in rHuEPO dose after IV iron therapy, even in patients with ferritin levels above 200 μg/L and transferrin saturation above 20% — patients who are usually categorized as iron repleted. However, these authors applied 2 doses of 500 mg iron sucrose within an interval of 1 week. It is possible that a low-dose iron regimen as used for iron repleted patients in our study was not sufficient to lead to a response similar to that described in the above-mentioned study. Furthermore, the rHuEPO dose required by our iron repleted patients was much lower even at baseline compared to that used by Domrongkitchaiporn et al. (30.8 U/kg/week in our study vs 71 U/kg/week) (6). Therefore, it is questionable if further optimization of rHuEPO treatment would have been possible in this group.
Patients with functional iron deficiency did not show a decreased ERI during the observation period. There may be several reasons for this: First, a different distribution of factors other than iron status among the subgroups may have resulted in increased rHuEPO requirements in patients with functional iron deficiency. Different studies describe rHuEPO hyporesponsiveness caused by hyperparathyroidism (18), aluminum toxicity (19), inadequate dialysis (20,21), underlying infection or inflammation (22), and malnutrition (23). rHuEPO requirements also increase with age (24). Furthermore, duration of dialysis has been reported as a factor that may aggravate renal anemia (21). In our study, patients with functional iron deficiency were older than those of the other two subgroups. They also had a longer duration of PD before start of the study than patients with absolute iron deficiency. However, differences in age, duration of dialysis before the study began, serum aluminum levels, iPTH, weekly Kt/V, and serum CRP levels were not significant between the three patient groups. The rate of clinically manifest infections during the observation period in patients with functional iron deficiency was comparable with the other two subgroups in our study (data not shown).
A second and more probable explanation for the unsatisfactory response of patients with functional iron deficiency in our study may be that the applied IV iron dose was too low for these patients. Transferrin saturation in patients with functional iron deficiency is typically low (13,25). Since transferrin saturation reflects the iron pool available for erythropoiesis, an iron dose similar to that used in cases of absolute iron deficiency may have resulted in a better response. The decline in transferrin saturation in patients with functional iron deficiency during the second phase of the observation period after its initial increase, as well as the corresponding increase in the ERI at month 6, underscores the fact that, in these patients, the applied IV iron dose should be chosen also in respect to transferrin saturation.
Looking at the whole group of patients, rHuEPO requirements after IV iron supplementation in this study were comparable to those found in our previous study using monthly applications of 100 mg/200 mg iron sucrose (7). Furthermore, the rate of side effects after IV injection of 50 mg iron sucrose in this study was slightly lower than that found in our previous study after application of 100 mg iron sucrose (0.5% vs 0.9%), and markedly lower than the incidence of adverse events noted after the 200-mg IV iron sucrose applications (0.5% vs 5.9%) (7).
Apart from efficacy and adverse events, convenience and cost of applying small IV iron doses more frequently versus large single doses may be an important issue for further investigation.
In summary, during an observation period of 1 year, the low-dose IV iron regimen used in this study substantially decreased rHuEPO requirements in patients with absolute iron deficiency. It also maintained iron stores and rHuEPO requirements in iron repleted patients. Low-dose IV iron therapy did not improve the relatively high ERI in patients with functional iron deficiency. Especially in this latter group, higher iron doses or more frequent iron injections may be necessary to reduce required rHuEPO doses significantly.
Footnotes
Acknowledgments
The authors thank Professor Marcus Müllner for statistical consultation, and the nurses of the Peritoneal Dialysis Unit for their participation in this study.