Transferrin saturation versus reticulocyte hemoglobin content for iron deficiency in Japanese hemodialysis patients

Kidney International, Vol. 63 (2003), pp. 1086 1093 Transferrin saturation versus reticulocyte hemoglobin content for iron deficiency in Japanese hemodialysis patients YOSHIKATSU KANEKO, SHIGERU MIYAZAKI, YOSHIHEI HIRASAWA, FUMITAKE GEJYO, and MASASHI SUZUKI Kidney Center, Shinraku-en Hospital, and Division of Clinical Nephrology and Rheumatology, Niigata University Graduate School of Medical and Dental Sciences, Niigata, Japan Transferrin saturation versus reticulocyte hemoglobin content of recombinant human erythropoietin (rhepo) causes for iron deficiency in Japanese hemodialysis patients. a rapid increase in hematocrit levels, eliminates the need Background. Iron deficiency is a frequent cause of recombifor transfusions, and improves the symptoms accompanied nant human erythropoietin (rhepo)-resistant anemia in hemowith anemia in end-stage renal disease or progressive dialysis patients. Both reticulocyte hemoglobin content (CHr) and transferrin saturation (TSAT) have been proposed as renal failure [1 4]. It is also possible to restore hematomarkers of iron deficiency, but it is unclear which parameter crit to the normal level if sufficient rhepo is adminisis superior. tered [5]. However, iron deficiency is one of the most Methods. To compare the efficacy of CHr and TSAT as an indicator for treatment of iron deficiency, we conducted a single-center, open-label, prospective, randomized, controlled tin therapy [6 11]. Gastrointestinal iron absorption is common causes of a suboptimal response to erythropoie- trial at the Kidney Center in Shinraku-en Hospital of 197 Japa- less than ongoing losses due to loss of blood in the dianese patients on chronic hemodialysis. After 4 weeks of runlyzer [12] or frequent blood tests, so aggressive managein period during which iron supplementation was suspended, 100 patients who were randomized to the CHr group received ment of iron deficiency is necessary to treat anemia in 240 mg iron colloid intravenously over 2 weeks when CHr less hemodialysis patients. Indeed, parenteral iron supplethan 32.5 pg, and 97 patients who were randomized to the mentation in patients treated with rhepo, who have TSAT group received the same doses of iron colloid when insufficient iron stores and even replete iron stores, can TSAT less than 20%. We measured the rhepo dose needed to maintain prestudy hematocrit levels, hematocrit, CHr, TSAT, correct the anemia with decreased mortality and hospi- serum ferritin, percentage of hypochromic red blood cells, and talization, improve the quality of life, and reduce the total iron administered. dose of rhepo [7, 9 11, 13 16]. The reduction of rhepo Results. Sixteen weeks later, 94 patients in the CHr group dose is also economically favorable [10, 11, 14, 17]. and 89 patients in the TSAT group finished the study. The doses of rhepo required decreased by 35.8% (4081 to 2629 The serum ferritin and transferrin saturation (TSAT), U/week, P 0.005) in the TSAT group, but not significantly in both reflecting total body iron stores, are commonly used the CHr group (4121 to 3606 U/week). Although CHr increased to indicate iron status in hemodialysis patients [17 19]. promptly after the iron administration in both groups, TSAT The National Kidney Foundation-Dialysis Outcomes increased only in the TSAT group. Quality Initiative (NKF-DOQI) guidelines recommended Conclusions. Although CHr reflects the iron status more sensitively, TSAT is a better clinical marker for iron supplementation therapy. TSAT at or less than 20% to manage iron deficiency maintaining serum ferritin at or less than 100 ng/ml, or [8]. However, some papers report these cutoff values as insufficient to treat functional iron deficiency, in which Anemia is one of the major complications of chronic sufficient iron is not released from the body iron stores renal failure, especially for patients undergoing chronic to meet an increased request for iron in the red blood hemodialysis for end-stage renal disease. Administration cell production accelerated by rhepo administration, despite a normal or even elevated serum ferritin level Key words: erythropoietin, randomized controlled trial, serum ferritin, anemia, parenteral iron, chronic hemodialysis. Received for publication May 3, 2002 and in revised form July 23, 2002, and October 3, 2002 Accepted for publication October 21, 2002 2003 by the International Society of Nephrology 1086 [9, 10, 13, 18 20]. Other markers, such as zinc protoporphyrin [21] or soluble transferrin receptor [22], have been proposed as alternative indicators of iron status, but they are less widely available and appear to offer no increase in diagnostic sensitivity and specificity over serum ferritin and

Kaneko et al: CHr versus TSAT 1087 TSAT [9], and their usefulness as markers of iron defi- before enrollment; and (2) no prior adverse reactions ciency is still controversial [23 25]. to parenteral iron. Exclusion criteria were hematologic Reticulocyte hemoglobin content (CHr) and percent- malignancies, hemolytic anemia, active gastrointestinal age of hypochromic red blood cells are other indices of bleeding, acute severe infection (C-reactive protein less iron status, which reflect the iron status at the level of the than 10 mg/dl), chronic inflammatory disease, aluminum erythrocyte itself [26 32]. Specifically, CHr is expected toxicity, uncontrolled hypertension, or hyperparathyroid- to provide a snapshot of functional iron status because ism (intact parathyroid hormone less than 1000 pg/ml). of the reticulocytes short lifespan, and, more recently, some papers have reported a correlation between CHr Measurement of hematologic and iron status marker and other indices of iron status in iron deficiency anemia Hematocrit, CHr, percentage of hypochromic red of healthy and chronic hemodialysis patients [26, 29 32]. blood cells, serum iron, unsaturated iron-binding capac- In the United States and in Europe, CHr less than 26.0 ity (UIBC), and serum ferritin were measured at the pg, measured by H*3 hematology analyzer, or CHr less beginning of each hemodialysis therapy twice per month, than 29.0 pg, measured by ADVIA 120 (Bayer, Tarry- 3 days after the previous hemodialysis therapy, to determine town, NY, USA), is used as the indicator of iron deficiency, the dose of rhepo or the implementation of intratown, and several authors reported that CHr increased venous iron therapy. Blood counts and red blood cell promptly within a few days after iron supplementation indices were measured by standard automated counter and was useful as an indicator for iron deficiency [29 32]. methods. CHr (pg) and hypochromic red blood cells In Japan, however, only a few studies were conducted (%), in which the content of hemoglobin was less than about CHr, and one of these studies reported that the cut- 28 g/dl, were measured by a flow cytometric method off value of CHr to define iron deficiency should be set (ADVIA). Serum iron ( g/dl) and UIBC ( g/dl) were by 32.2 pg in Japanese chronic hemodialysis patients (ab- measured by a colorimetric method (7600-020, Hitachi, stract; Okano et al, J Jpn Soc Dial Ther 34:1079, 2001). Tokyo, Japan). Serum ferritin (ng/ml) was measured Little has also been reported on the usefulness of CHr by a chemiluminescence immunoassay method (BN II, as a target of iron supplement therapy compared to Dade Behring, Marburg, Germany). TSAT (%) was cal- TSAT or serum ferritin. The aim of the present study culated as serum iron/serum iron plus UIBC. is to set the cutoff value of CHr and to clarify which parameter, CHr or TSAT, is superior to detect iron deficiency Administration of study medication and to reduce the amount of rhepo required Intravenous iron colloid with chondroitin sulfate (Blu- without decreasing the hematocrit values in Japanese tal, Dainippon Pharmaceutical, Tokyo, Japan), 40 mg, chronic hemodialysis patients. was administered three times per week for 2 weeks at the end of each hemodialysis therapy when CHr was less than 32.5 pg in the CHr group or TSAT was less than METHODS 20% in the TSAT group. Participants in both groups Study design received either epoetin alpha (Espo, Kirin Brewery, This single-center, open, intention to treat, randomized, Tokyo, Japan) or epoetin beta (Epogin, Chugai Phar- controlled clinical trial was performed at the Kidney maceutical, Tokyo, Japan) if needed. The dose of rhepo Center in Shinraku-en Hospital between May 2001 and was adjusted as necessary at the measurement of hemato- September 2001. Patients were randomly assigned in a logic and iron status markers to maintain the hematocrit 1:1 ratio to receive intravenous iron therapy when CHr between 29.5 and 32.5%. Doses of rhepo administered was less than 32.5 pg (CHr group), or TSAT was less were categorized as 0, 750, 1500, 2250, 3000, 4500, 6000, than 20% (TSAT group), at which time all oral iron or 9000 U per week and modified as follows: (1) hemato- supplementation was discontinued. The protocol was re- crit was less than 26.0%, rhepo dose was raised by 200%; viewed and approved by the Institutional Review Board (2) hematocrit was equal to or more than 26.0% and less in Shinraku-en Hospital. Written informed consent was than 29.0%, rhepo dose was raised by 100%; (3) hematocrit obtained from the participants before enrollment. was equal to or more than 29.0% and less than 29.5%, rhepo dose was raised by 50%; (4) hematocrit Patient enrollment was more than 32.5% and less than or equal to 33.0%, Patients 18 years of age or older undergoing chronic rhepo dose was reduced by 33%; (5) hematocrit was hemodialysis treatment were eligible. The range of the more than 33.0% and less than or equal to 36.0%, rhepo time patients had been receiving hemodialysis was from dose was reduced by 50%; and (6) hematocrit was more 2 months to 26 years. Other eligibility criteria were (1) than 36.0%, administration of rhepo was suspended. use of rhepo doses of 750 to 3000 U administered intravenously When the administration of rhepo had been sus- one to three times per week for anemia man- pended and hematocrit was less than 29.5%, 2250 U per agement for at least 4 weeks within the last 3 months week of rhepo was resumed. If the modified rhepo

1088 Kaneko et al: CHr versus TSAT dose in accordance with the rule mentioned above did Table 1. Baseline demographic characteristics of patients not apply to any of the categories, the nearest dose in CHr Group TSAT Group category was adopted. Variable (N 100) (N 97) Outcome measures Men/women number 61/39 60/37 Age year 57.6 13.3 61.2 14.5 Age range year 26 88 27 88 The main outcome of interest was the dose of rhepo Cause of ESRD number required to keep the hematocrit at its baseline level. Chronic glomerulonephritis 62 63 Secondary outcomes included the difference of the total Diabetic nephropathy 12 15 Nephrosclerosis 7 8 dose of iron administered or the difference of hemato- Polycystic kidney 2 3 logic and iron status markers between the two groups. Chronic pyelonephritis 2 3 Congenital renal hypoplasty 2 2 Monitoring of adverse events Purpura nephritis 2 0 Membranous nephropathy 2 0 All patients receiving study medication were evalu- Others a 9 3 ated for safety. Signs and symptoms were evaluated dur- Hematocrit % 31.2 3.3 31.4 2.9 CHr pg 33.2 2.1 33.2 2.5 ing and after each hemodialysis session, and the incidence Hypochromic red blood cells % 3.9 4.9 4.5 5.6 of hospitalization, infections, and deaths was recorded. Transferrin saturation % 25.2 12.0 26.3 16.0 Serum ferritin ng/ml 243.2 323.4 251.6 371.2 Statistical analysis TIBC lg/dl 241.9 47.4 241.1 47.0 Intact PTH pg/ml 278.5 257.5 244.2 234.2 For comparisons between two groups at baseline and Vitamin D 3 dose after the trial, we used the Student t test for all the Alfacalcidol lg/day 0.37 0.13 0.32 0.11 Calcitriol lg/day 0.44 0.67 0.34 0.23 subjects except for doses of rhepo and iron, and the Maxacalcitol lg/day 9.2 5.2 8.3 2.9 Mann-Whitney U test for doses of rhepo and iron. rhepo dose U/week 4263 2982 3969 2872 Comparisons of all the subjects except for doses of rhepo dose range U/week 0 9000 0 9000 Epoetin alpha number 33 29 rhepo in each group were performed using repeated Epoetin alpha dose U/week 4591 286 4603 3224 measure one-way analysis of variance (ANOVA), fol- Epoetin alpha dose range U/week 750 9000 0 9000 lowed by Fisher s protected least significant difference Epoetin beta number 67 68 Epoetin beta dose U/week 4180 3060 3728 2693 as a post-hoc test. Comparisons with respect to rhepo Epoetin beta dose range U/week 0 9000 0 9000 dose in each group were conducted using the Friedman TSAT 20% number 34 37 test, followed by the Wilcoxon signed-rank test with CHr 32.5 pg number (%) 20 (58.8) 22 (59.5) CHr 32.5 pg number (%) 14 (41.2) 15 (40.5) Bonferroni correction as a post-hoc test. Comparison of CHr 32.5 pg number 35 32 the number of patients of which TSAT was less than TSAT 20% number (%) 20 (57.1) 22 (68.8) 20%, and CHr was less than 32.5 pg, or CHr was 32.5 TSAT 20% number (%) 15 (42.9) 10 (31.2) pg or more, was performed using the Fisher s exact probbin Abbreviations are: ESRD, end-stage renal disease; CHr, reticulocyte hemoglo- content; TIBC, total iron binding capacity; PTH, parathyroid hormone; ability test. A P value 0.05 was considered significant. rhepo, recombinant human erythropoietin; TSAT, transferrin saturation. All All statistical calculations were performed using Statbetween results are expressed as mean SD. None of the variables differed significantly View, version 5.0 (SAS Institute Inc., Cary, NC, USA). the groups. a Renal tuberculosis, primary amyloidosis, or unknown causes Role of the funding source The funding source had no role in the collection, analycantly sis, or interpretation of the data or in the decision to between either group (Table 1). We had presis, submit the study for publication. viously verified that there was no difference on the effect to promote erythropoiesis between epoetin alpha and RESULTS epoetin beta (data not published), so we calculated the average dose of rhepo without dividing epoetin alpha Patients and epoetin beta throughout the present study. Of 242 patients who met the study criteria, 197 patients Values of CHr and TSAT of all 197 patients were were enrolled and randomly allocated (100 patients in measured at the enrollment and a weak correlation be- the CHr group and 97 patients in the TSAT group). The tween CHr and TSAT was revealed (r 0.580, r 2 groups did not significantly differ in baseline demo- 0.336, P 0.0001, Fig. 1). We defined TSAT of less than graphic characteristics (Table 1). The average doses of 20% as indicating iron deficiency according to NKF- rhepo, including epoetin alpha and epoetin beta at the DOQI guidelines [8], and when we set the cutoff value time of enrollment, were 4263 U/week in the CHr group of CHr at less than 32.5 pg, it had a sensitivity of 59.2% and 3969 U/week in the TSAT group, a difference of and a specificity of 62.7% for the diagnosis of iron defi- 7%, which was not significant. The dose and proportion ciency. When the cutoff values were set at 29.0 pg, 31.5 of epoetin alpha and epoetin beta did not differ signifi- pg, 32.0 pg, or 33.0 pg, sensitivities were 9.9%, 40.8%,

Kaneko et al: CHr versus TSAT 1089 Hematocrit values in both groups fluctuated during the study, but no difference of the value was observed between the two groups and the hematocrit values at the end of the trial were almost the same as the baseline values (Table 2). The amount of total dosage of iron colloid administered became significantly higher in the TSAT group 13 weeks after the beginning of intervention and this tendency continued to the end of the study (Table 2). Difference of iron status After 16 weeks of the trial, the value of TSAT in- Fig. 1. Correlation between transferrin saturation and reticulocyte hecreased significantly in the TSAT group and was signifimoglobin content. *y 294 0.191x 0.001x 2 ;r 2 0.336, P 0.0001. cantly higher than those in the CHr group (P 0.034). In contrast, no significant increase of TSAT was observed in the CHr group. The serum ferritin value also increased 52.1%, 69.0%, and specificities were 87.5%, 74.4%, dramatically in the TSAT group, and the value was higher 75.6%, 57.0%, respectively. Considering these data, we by 33.3% than the value in the CHr group at the end of decided on CHr less than 32.5 pg as a proper cutoff value the study. Two other markers, CHr and hypochromic red for indication of intravenous iron therapy in the present blood cells, which indicate iron status at the erythrocyte study. level, promptly improved equally from the baseline 4 At the enrollment, 13 patients were taking oral iron weeks and 9 weeks, respectively, after the interventional every day and the others were receiving intravenous iron treatment in both groups. Improvement of CHr was ob- on an as needed basis when TSAT was less than 20%. served 9 weeks earlier than the significant increase of TSAT After 4 weeks of run-in period, during which oral and in the TSAT group was first observed in week 13. Howparenteral iron administration was suspended, 183 pa- ever, no significant differences on the values of CHr and tients (94 patients in the CHr group and 89 patients in hypochromic red blood cells were observed between the the TSAT group) completed the entire 16 weeks of study. two groups. In addition, substantial improvement of the Of the 197 participants, 14 participants (six in the CHr values of CHr and hypochromic red blood cells was not group and eight in the TSAT group) prematurely discon- observed from week 11 to the end of the trial (Table 2). tinued the trial. The reasons for discontinuing the trial Comparison of test variability was measured for CHr, included withdrawn consent (three in the CHr group TSAT, and serum ferritin and was expressed as coefficient and four in the TSAT group), deaths (two in the CHr of variation (% CV). CHr had 6.3% of % CV, group and one in the TSAT group), loss of follow-up which was obviously less than that of TSAT (48.9%) or (one in the CHr group and two in the TSAT group), serum ferritin (130.5%). and massive bleeding (one in the TSAT group). We reevaluated the sensitivity and the specificity of CHr for the diagnosis of iron deficiency at the end of Efficacy the trial. Between the two groups, although specificities Table 2 shows the change of the dose of rhepo administered did not differ (66.7% both in the CHr group and in the to 184 participants who completed the trial. Al- TSAT group), sensitivity was much lower in the CHr though some significant increase of the rhepo dose was group (23.1% in the CHr group and 57.1% in the TSAT observed in the CHr group 4 weeks after the beginning group). Table 3 shows the number of participants who of intervention, the dose of rhepo in the TSAT group were diagnosed as iron deficiency by TSAT less than significantly decreased from the baseline dose 11 weeks 20% among all the participants. Significantly more participants after the beginning of intervention and this tendency in the CHr group had CHr values higher than continued until the end of the trial. The dose of rhepo 32.5 pg despite the fact that TSAT was less than 20% required was significantly lower in the TSAT group compared (P 0.043) and, consequently, the sensitivity of CHr with those in the CHr group (P 0.05) 11 and 13 lowered only in the CHr group. In contrast, same propor- weeks after the start of the trial. Although the difference tion of patients both in the CHr group and the TSAT between the two groups was obvious, but not statistically group had a TSAT value less than 20% among the patients significant at the end of the trial (P 0.10), the dose of who had CHr less than 32.5 pg (Table 4). Consesignificant rhepo required finally decreased from the baseline by quently, when CHr less than 32.5 pg was defined as 35.6% (4081 to 2629 U/week, P 0.005) in the TSAT indicating iron deficiency, sensitivity of TSAT was the group, but did not decrease significantly in the CHr group same (66.7%) in both CHr group and TSAT group at (4121 to 3606 U/week). the end of the trial.

1090 Kaneko et al: CHr versus TSAT Table 2. Average values for recombinant human erythropoietin, hematocrit, transferrin saturation, ferritin, reticulocyte hemoglobin content, hypochromic red blood cells, and total iron Time rhepo Hct TSAT Ferritin CHr2 Hypochromic Total iron period U/week % % ng/ml pg RBC % mg Baseline CHr 4121 2922 31.2 3.2 25.5 12.6 234.5 307.0 33.2 2.2 3.7 4.9 0 TSAT 4081 3123 31.3 2.8 25.7 15.6 257.0 453.4 32.8 2.4 4.1 5.3 0 Week 2 CHr 4574 3261 30.1 a 3.2 26.3 11.8 259.4 296.7 32.7 b 1.8 4.5 b 5.7 68.3 108.0 TSAT 4424 3220 30.4 2.3 25.8 13.3 289.9 405.2 32.7 1.7 5.1 b 6.7 91.7 117.3 Week 4 CHr 5426 a 3481 29.5 c 2.9 28.0 11.5 285.8 c 302.5 33.6 b 1.7 4.1 5.1 164.8 193.6 TSAT 4803 3325 29.9 b 1.9 28.2 13.6 305.5 b 397.2 33.5 c 2.2 4.4 5.4 180.7 199.1 Week 6 CHr 4484 3476 32.1 b 2.8 27.8 12.5 285.6 c 302.0 35.3 c 2.2 3.7 4.3 214.1 242.1 TSAT 3843 3092 32.4 a 2.2 26.5 12.5 304.8 397.5 34.8 c 2.4 3.8 3.9 226.5 254.2 Week 9 CHr 3957 3320 31.5 3.0 28.6 13.2 269.4 b 303.4 34.7 c 1.6 1.9 c 2.0 232.0 268.8 TSAT 3051 2730 32.1 b 2.5 28.5 12.1 323.1 a 423.7 34.4 c 2.2 1.8 c 2.3 283.1 297.4 Week 11 CHr 3638 3276 31.9 2.9 28.9 14.3 279.7 b 298.8 34.6 c 1.8 1.7 c 2.0 247.3 305.5 TSAT 2528 b,d 2736 32.4 a 2.8 28.7 13.6 331.8 a 466.7 34.4 c 1.8 1.7 c 2.2 315.5 312.5 Week 13 CHr 3535 3310 31.4 2.9 28.8 12.4 282.8 b 331.6 34.8 c 1.5 1.5 c 1.7 257.5 334.1 TSAT 2469 b,d 2635 31.7 2.6 33.5 c,d 16.8 364.3 c 473.6 34.8 c 2.0 1.4 c 2.0 358.7 d 341.6 Week 16 CHr 3606 3347 31.3 2.8 28.2 14.3 279.5 b 326.9 34.4 c 1.6 1.7 c 1.9 267.7 353.2 TSAT 2629 b 2640 31.5 2.9 32.7 c,d 14.9 372.6 c 518.1 34.3 c 1.9 1.5 c 2.3 377.5 d 361.6 Abbreviations are: rhepo, recombinant human erythropoietin; Hct, hematocrit; TSAT, transferrin saturation; CHr, reticulocyte hemoglobin content; RBC, red blood cells. All results are expressed as mean SD. a P 0.001 compared to baseline value b P 0.01 compared to baseline value c P 0.0001 compared to baseline value d P 0.05 between the two groups Table 3. Number of patients in whom transferrin saturation was less than 20% and reticulocyte hemoglobin content was under 32.5 pg, or 32.5 pg or more Table 4. Number of patients in whom reticulocyte hemoglobin content was less than 32.5 pg and transferrin saturation was under 20%, or 20% or more TSAT 20% CHr 32.5 pg Group CHr 32.5 pg CHr 32.5 pg Group TSAT 20% TSAT 20% CHr (N 26) 6 (23.1%) a 20 (76.9%) a CHr (N 9) 6 (66.7%) 3 (33.4%) TSAT (N 14) 8 (57.1%) 6 (42.9%) TSAT (N 12) 8 (66.7%) 4 (33.4%) Abbreviations are: TSAT, transferrin saturation; CHr, reticulocyte hemoglobin content. a P 0.043 Abbreviations are: TSAT, transferrin saturation; CHr, reticulocyte hemoglobin content. Safety DISCUSSION Three patients died during this study. Two patients This study demonstrates that TSAT, a marker of iron in the CHr group died; one during week 4 (bacterial status at the total body level, is superior to CHr, a marker pneumonia) and one during week 16 (sudden death by of iron status at the reticulocyte level, for treatment unknown cause) of the trial period. One patient in the of iron deficiency and reducing the amount of rhepo TSAT group died in week 7 because of a liver tumor required in Japanese chronic hemodialysis patients with that was not discovered at patient enrollment and ran- iron deficiency anemia. domization. One patient in the TSAT group was premaan So far, functional iron deficiency has been defined as turely discontinued from the study because of massive impaired iron supply from body iron stores despite bleeding due to a femoral bone fracture and need for a normal or elevated serum ferritin because of the accel- blood transfusion. erated need for iron for erythropoiesis stimulated with No differences in hospitalizations or infection rate administered rhepo [8, 9]. CHr was expected to be a were observed. One patient in the CHr group and one sensitive marker of functional iron deficiency because it patient in the TSAT group were hospitalized for infecerythrocyte directly reflects the iron-deficient erythropoiesis at the tion of renal cysts and internal shunt obstruction, respectively. level regardless of the total body iron stores [29 32].

Kaneko et al: CHr versus TSAT 1091 In this interventional study, increase of CHr was ob- did not increase and iron deficiency was not improved served in week 4, reflecting the change of the iron status significantly. The difference of the iron administered to previous to the significant elevation of TSAT in the participants in the TSAT group may have contributed to TSAT group, which was first observed in week 13. As increasing the amount of iron stores in the bone marrow, reported by several authors [29 32], CHr is an excellent spleen, and liver, as well as in the erythrocytes, as indicated indicator of iron status, especially to detect the change by the increase of TSAT and serum ferritin. These of iron status within a short timespan and to evaluate findings are compatible with several reports that maintaining the efficacy of iron supplement therapy in a short time TSAT at a high level leads to an increase of the period. Considering these results, CHr may reflect the iron stores, enhanced iron delivery to the tissues, and a improvement of iron status more sensitively than TSAT. reduced amount of rhepo required [9, 13, 19, 20]. Although we hypothesized that CHr reflects the func- Iron overload is a cause of the increase in cardiovascular tional iron deficiency more efficiently, and that leads to and infectious morbidity, parenchymal iron infiltrational the reduction of the amount of rhepo required with tion, risk of cancer, and even relative resistance to rhepo effective iron treatment, our study has disclosed that CHr [5, 6, 8 10, 33], so iron supplementation should be imple- was a sensitive, but might be an insufficient, indicator for mented with careful monitoring of hematologic and iron iron deficiency. Although two indices, CHr and hypochromic status markers. If hematocrit does not react to rhepo, red blood cells, promptly improved in both while serum ferritin progressively increases during iron groups after a series of intravenous iron therapy, improvement treatment, iron therapy should be suspended. However, of the values of CHr and hypochromic red these adverse effects are reported to occur when the blood cells stopped 11 weeks after the intervention and serum ferritin level is more than 500 ng/ml [9, 10, 33]. no more improvement was observed until the end of the Although the % CV value of TSAT or serum ferritin trial. On the contrary, a substantial increase of the TSAT was higher than CHr, and it is probable that extreme value and serum ferritin has been observed only in the variability of the test might lead to unnecessary treatment TSAT group until the end of the trial, followed by the of patients without iron deficiency, the serum ferri- reduction of rhepo doses. tin and TSAT values in the TSAT group at the end of In addition, of 26 participants in the CHr group categorized the present study, 372.6 ng/ml and 32.7%, may be levels with iron deficiency by TSAT less than 20%, the with relatively less risk as a target of iron therapy. CHr value of 20 participants was 32.5 pg or more at the The strategy of iron treatment adopted in the present end of the trial. In contrast, the proportion of patients study was significantly different from those adopted in who had TSAT values less than 20% in the TSAT group, the United States and in Europe. We used iron colloid, as well as CHr less than 32.5 pg, was the same as those instead of iron dextran or iron gluconate, which were in the CHr group. This result indicates that the reason widely used in the United States and in Europe [8, 34], why more patients in the CHr group had CHr values but were not available in Japan. The dose of iron in the higher than 32.5 pg despite TSAT less than 20% is not present study was also significantly lower than those used due to the study design, but due to the different charac- in the United States or in Europe [8, 34]. In the present ters of CHr and TSAT. When CHr was less than 32.5 study, to avoid the risk of iron overload, we used 240 pg, two thirds of the patients had TSAT less than 20%, mg of iron over 2 weeks, which was revealed to be a but even if CHr was less than or equal to 32.5 pg, TSAT sufficient dose to improve iron deficiency. was not always more than 20% when iron supplementa- For the present study, we defined iron deficiency as tion was based on CHr value. These results suggest that TSAT less than 20% and set the cutoff value for induc- the increase of CHr does not always lead to the increase tion of intravenous iron therapy as CHr less than 32.5 of TSAT. Although the CHr values in both groups were pg in the CHr group, considering the sensitivity and almost the same, TSAT and serum ferritin were higher specificity for the diagnosis. This cutoff value is significantly in the TSAT group, and this was because the amount of higher than the value used in the United States total iron administered during the trial was 40% higher and in Europe [29 32]. With respect to Japanese chronic in the TSAT group compared with the CHr group. These hemodialysis patients, Okano et al proposed the cutoff results also indicate that the amount of iron administered value of CHr as 32.2 pg, considering the correlation with to participants in the CHr group may be a minimal dose TSAT, and the sensitivity and the specificity of diagnosis, to satisfy iron stores at a reticulocyte level, and intrave- when iron deficiency was defined as TSAT less than 20% nous iron supplementation based on the CHr value increases (Abstract, Okano et al, J Jpn Soc Dial Ther 34:1079, the iron stores only in erythrocytes and that the 2001). Harada et al proposed the cutoff value of CHr in total body iron stores remain unsatisfactory. This may Japanese chronic hemodialysis patients as 32.0 pg, given be because CHr is too sensitive and elevates rapidly to that whether the hematocrit value was improved or not the cutoff value before extra doses of iron to increase by the administration of parenteral iron was clearly de- body iron stores are administered. As a result, TSAT pendent on the value of CHr (i.e., iron deficiency anemia

1092 Kaneko et al: CHr versus TSAT was improved by administration of iron when patient s which rapidly reflects the change of iron status, but may CHr was less than 32.0 pg (personal communication). In be insufficient as a target of iron supplementation therapy. the present study, we set the cutoff value of CHr as 32.5 pg, which was rather higher than the values proposed ACKNOWLEDGMENTS by two studies mentioned above, so as not to overlook The authors thank Drs. Shinji Sakai, Yasuko Yuasa, Tai Sakurabayashi, Ikuo Aoike, Yutaka Osawa, Susumu Haginoshita, Yoshiji Tapatients with possible iron deficiency. Nevertheless, the amount of administered iron and the improvement of kaesu, and the staff at Kidney Center, Shinraku-en Hospital for assis- tance with patient enrollment and follow-up; and Kouichi Yanuki, iron deficiency were greater in the TSAT group. It is Clinical Investigation Center, Shinraku-en Hospital, for measurement unclear what is the reason for the disparity of the CHr of hematologic and iron status marker. We also thank Chugai Pharma- cutoff values between the patients in the United States ceutical for providing us with ADVIA 120 without compensation. Funding for this study was provided in part by Renal Anemia Foundaor in Europe and in Japan. tion, Tokyo, Japan. More recently, a similar randomized trial was reported by Fishbane et al [32]. The results of this study were Reprint requests to Yoshikatsu Kaneko, M.D., Ph.D., Kidney Center, Shinraku-en Hospital, Nishiariake-cho 1-27, Niigata 950-2087, Japan. almost the same as our study (i.e., TSAT and ferritin E-mail: kanekoy@med.niigata-u.ac.jp were elevated due to a greater amount of iron supplementation in a group in which iron management was REFERENCES based on serum ferritin and TSAT compared with a 1. Eschbach JW, Egrie JC, Downing MR, et al: Correction of the group based on CHr). One of major differences between anemia of end-stage renal disease with recombinant human erythropoietin. Results of a combined phase I and II clinical trial. N the two studies was a primary outcome, whether the dose of rhepo required for maintaining baseline hema- Engl J Med 316:73 78, 1987 2. Eschbach JW, Kelly MR, Haley NR, et al: Treatment of the tocrit was significantly reduced or not in the group based anemia of progressive renal failure with recombinant human erythropoietin. N Engl J Med 321:158 163, 1989 on TSAT. The reason for this difference may be due to the differences of the composite, dose, or interval of 3. Eschbach JW, Abdulhadi MH, Browne JK, et al: Recombinant human erythropoietin in anemic patients with end-stage renal disiron therapy, or may be due to the difference of target ease. Results of a phase III multicenter clinical trial. Ann Intern hematocrit value. Target hematocrit value is recom- Med 111:992 1000, 1989 mended to be maintained from 33% to 36% in the 4. Besarab A, Flaharty KK, Erslev AJ, et al: Clinical pharmacology and economics of recombinant human erythropoietin in end-stage United States, following NKF-DOQI guidelines [8], or renal disease: The case for subcutaneous administration. JAmSoc over 33% in Europe, following European Best Practice Nephrol 2:1405 1416, 1992 Guidelines [34]. In the present study, we set the target 5. Besarab A, Bolton WK, Browne JK, et al: The effects of normal as compared with low hematocrit values in patients with cardiac hematocrit value from 29.5% to 32.5%, which is widely disease who are receiving hemodialysis and epoetin. N Engl J Med recommended in Japan, although this value is signifi- 339:584 590, 1998 cantly lower than those in the United States or in Europe. 6. Hörl WH, Cavill I, Macdougall IC, et al: How to diagnose and correct iron deficiency during r-huepo therapy A consensus Given that these differences mentioned above between report. Nephrol Dial Transplant 11:246 250, 1996 in the United States or Europe and in Japan, it is not 7. Macdougall IC: Optimal iron management in patients receiving erythropoietin therapy. Semin Dial 11:10 13, 1998 contradictory that our results are not in line with those 8. National Kidney Foundation Dialysis Outcomes Quality Inireported by Fishbane et al [32]. tiative: NKF-K/DOQI Clinical Practice Guidelines for Anemia of In this study, intravenous administration of iron was Chronic Kidney Disease: Update 2000. Am J Kidney Dis 37(Suppl 1):S182 S238, 2001 designed on an as needed basis, which is rather sub- 9. Besarab A, Frinak S, Yee J: An indistinct balance: The safety optimal, and maintenance parenteral iron administration and efficacy of parenteral iron therapy. J Am Soc Nephrol 10:2029 is recommended to achieve target hematocrit with lower 2043, 1999 10. Tarng D, Huang T, Chen TW, Yang W: Erythropoietin hyporhepo doses [8, 9, 11, 15, 16, 19, 35]. We adopted an responsiveness: From iron deficiency to iron overload. Kidney Int as needed strategy to compare the usefulness of CHr 55(Suppl 69):S107 S118, 1999 and TSAT more clearly, but it may be possible to apply 11. Sunder-Plassmann G, Hörl WH: Importance of iron supply for erythropoietin therapy. Nephrol Dial Transplant 10:2070 2076, the present data to maintenance iron treatment. 1995 12. Lindsay RM, Burton JA, King P, et al: The measurement of dialyzer blood loss. Clin Nephrol 1:24 28, 1973 CONCLUSION 13. Fishbane S, Frei GL, Maesaka J: Reduction in recombinant human erythropoietin doses by the use of chronic intravenous iron In conclusion, we reported the usefulness and limita- supplementation. Am J Kidney Dis 26:41 46, 1995 tions of CHr and TSAT as indicators of iron status for 14. Sepandj F, Jindal K, West M, Hirsch D: Economic appraisal of iron deficiency anemia in Japanese chronic hemodialysis maintenance parenteral iron administration in treatment of anae- mia in chronic haemodialysis patients. Nephrol Dial Transplant patients receiving rhepo therapy. This study demon- 11:319 322, 1996 strates that an increase of TSAT is necessary to improve 15. Taylor JE, Peat N, Porter C, Morgan AG: Regular low-dose functional iron deficiency and concomitantly decreases intravenous iron therapy improves response to erythropoietin in haemodialysis patients. Nephrol Dial Transplant 11:1079 1083, the dose of rhepo required to maintain the hematocrit 1996 between 29.5% and 32.5%. CHr is a sensitive marker, 16. Macdougall IC, Tucker B, Thompson J, et al: A randomized

Kaneko et al: CHr versus TSAT 1093 controlled study of iron supplementation in patients treated with cator of iron deficiency and response to therapy. Blood 83:3100 erythropoietin. Kidney Int 50:1694 1699, 1996 3101, 1994 17. Macdougall IC: Monitoring of iron status and iron supplementaulocytes. 27. Schaefer RM, Schaefer L: Hypochromic red blood cells and retiction Kidney Int 55(Suppl 69):S44 S48, 1999 in patients treated with erythropoietin. Curr Opin Nephrol Hypertens 3:620 625, 1994 28. Tessitore N, Solero GP, Lippi G, et al: The role of iron status markers in predicting response to intravenous iron in haemodial- 18. Fishbane S, Kowalski EA, Imbriano LJ, Maesaka JK: The evaluaysis patients on maintenance erythropoietin. Nephrol Dial Transtion of iron status in hemodialysis patients. J Am Soc Nephrol plant 16:1416 1423, 2001 7:2654 2657, 1996 29. Fishbane S, Galgano C, Langley RC, Jr, et al: Reticulocyte 19. Besarab A, Kaiser JW, Frinak S: A study of parenteral iron hemoglobin content in the evaluation of iron status of hemodialysis regimens in hemodialysis patients. Am J Kidney Dis 34:21 28, 1999 patients. Kidney Int 52:217 222, 1997 20. Besarab A, Amin N, Ahsan M, et al: Optimization of epoetin 30. Mittman N, Sreedhara R, Mushnick R, et al: Reticulocyte hemotherapy with intravenous iron therapy in hemodialysis patients. J globin content predicts functional iron deficiency in hemodialysis Am Soc Nephrol 11:530 538, 2000 patients receiving rhuepo. Am J Kidney Dis 30:912 922, 1997 21. Fishbane S, Lynn RI: The utility of zinc protoporphyrin for prereticulocyte haemoglobin content as markers of iron-deficient eryth- 31. Cullen P, Söffker J, Höpfl M, et al: Hypochromic red cells and dicting the need for intravenous iron therapy in hemodialysis paropoiesis in patients undergoing chronic haemodialysis. Nephrol tients. Am J Kidney Dis 25:426 432, 1995 Dial Transplant 14:659 665, 1999 22. Bovy C, Tsobo C, Crapanzano L, et al: Factors determining the 32. Fishbane S, Shapiro W, Dutka P, et al: A randomized trial of percentage of hypochromic red blood cells in hemodialysis patients. iron deficiency testing strategies in hemodialysis patients. Kidney Kidney Int 56:1113 1119, 1999 Int 60:2406 2411, 2001 23. Braun J, Hammerschmidt M, Schreiber M, et al: Is zinc protoporrial 33. Hoen B, Kessler M, Hestin D, Mayeux D: Risk factors for bactephyrin an indicator of iron-deficient erythropoiesis in maintenance infections in chronic haemodialysis adult patients: A multicen- haemodialysis patients? Nephrol Dial Transplant 11:492 497, 1996 tre prospective survey. Nephrol Dial Transplant 10:377 381, 1995 24. Braun J: Erythrocyte zinc protoporphyrin. Kidney Int 55(Suppl 34. European Best Practice Guidelines for the management of anemia in patients with chronic renal failure: Guideline 8: 69):S57 S60, 1999 Administration of supplemental iron. Nephrol Dial Transplant 25. Ahluwalia N, Skikne BS, Savin V, Chonko A: Markers of masked 14(Suppl 5):17 18, 1999 iron deficiency and effectiveness of EPO therapy in chronic renal 35. Auerbach M, Winchester J, Wahab A, et al: A randomized trial failure. Am J Kidney Dis 30:532 541, 1997 of three iron dextran infusion methods for anemia in EPO-treated 26. Brugnara C: Reticulocyte hemoglobin content (CHr): Early indi- dialysis patients. Am J Kidney Dis 31:81 86, 1998