Optimization of Epoetin Therapy with Intravenous Iron Therapy in Hemodialysis Patients

Transcription

1 J Am Soc Nephrol 11: , 2000 Optimization of Epoetin Therapy with Intravenous Iron Therapy in Hemodialysis Patients ANATOLE BESARAB,* NEETA AMIN, MUHAMMAD AHSAN,* SUSAN E. VOGEL,* GARY ZAZUWA,* STANLEY FRINAK,* JAMES J. ZAZRA, J. V. ANANDAN, and AJAY GUPTA* *Division of Nephrology and Hypertension, Department of Medicine, and Department of Pharmacy, Henry Ford Hospital, Detroit, Michigan; and Life-Chem Labs, Rockleigh, New Jersey. Abstract. Iron deficiency limits the efficacy of recombinant human erythropoietin (rhepo) therapy in end-stage renal disease (ESRD) patients. Functional iron deficiency occurs with serum ferritin 500 ng/ml and/or transferrin saturation (TSAT) of 20 to 30%. This study examines the effects of a maintenance intravenous iron dextran (ivid) protocol that increased TSAT in ESRD hemodialysis patients from conventional levels of 20 to 30% (control group) to those of 30 to 50% (study group) for a period of 6 mo. Forty-two patients receiving chronic hemodialysis completed a 16- to 20-wk run-in period, during which maintenance ivid and rhepo were administered in amounts to achieve average TSAT of 20 to 30% and baseline levels of hemoglobin of 9.5 to 12.0 g/dl. After the run-in period, 19 patients randomized to the control group received ivid doses of 25 to 150 mg/wk for 6 mo. Twentythree patients randomized to the study group received four to six loading doses of ivid, 100 mg each, over a 2-wk period to achieve a TSAT 30% followed by 25 to 150 mg weekly to maintain TSAT between 30 and 50% for 6 mo. Both regimens were effective in maintaining targeted hemoglobin levels. Fifteen patients in the control group and 17 patients in the study group finished the study in which the primary outcome parameter by intention to treat analysis was the rhepo dose needed to maintain prestudy hemoglobin levels. Maintenance ivid requirements in the study group increased from 176 to 501 mg/mo and were associated with a progressive increase in serum ferritin to 658 ng/ml. Epoetin dose requirements for the study group decreased by the third month and remained 40% lower than for the control group, resulting in an overall cost savings in managing the anemia. Secondary indicators of irondeficient erythropoiesis were also assessed. Zinc protoporphyrin did not change in either group. Reticulocyte hemoglobin content increased only in the study group from 28.5 to 30.1 pg. It is concluded that maintenance of TSAT between 30 and 50% reduces rhepo requirements significantly over a 6-mo period. Anemia is a frequent and serious complication of end-stage renal disease (ESRD). Recombinant human erythropoietin (rhepo, Epoetin) therapy produces a dose-dependent increase in hemoglobin levels (1 3) and can restore hemoglobin to normal levels if sufficient rhepo is administered (4). Iron deficiency blunts the response to rhepo (5,6). Iron losses cumulating to 2 g/yr (7) are typical in the hemodialysis (HD) patient and exceed normal total body stores. Given that gastrointestinal iron absorption is less than ongoing iron losses in the majority of HD patients, functional iron deficiency is likely to develop in most patients leading to iron-limited erythropoiesis. In functional iron deficiency, the bone marrow s erythropoietic capacity to respond to Epoetin is limited by iron release from storage depots (8 10) or by a limited capacity of plasma Received May 3, Accepted August 10, Correspondence to Dr. Anatole Besarab, Division of Nephrology, Department of Medicine, University of West Virginia, Room 1264 HSS, P. O. Box 9165, Morgantown, WV. Phone: ; Fax: ; abesarab@pol.net / Journal of the American Society of Nephrology Copyright 2000 by the American Society of Nephrology transport. Iron stores drop precipitously when red blood cell production accelerates following initiation of Epoetin therapy in both non-azotemic as well as dialysis-dependent subjects (11,12). The National Kidney Foundation-Dialysis Outcomes Quality Initiative (NKF-DOQI) guidelines advocate aggressive detection and management of functional iron deficiency (5). Functional iron deficiency is assessed by the response to a course of parenteral iron producing either a decrease in dose of rhepo needed to maintain the target hematocrit or an increase in hemoglobin at the same dose of Epoetin (6,9,10,13,14). Maintenance parenteral iron administration as opposed to asneeded strategies consistently achieve target hematocrit using lower amounts of Epoetin (13 17), presumably by avoiding iron-limited erythropoiesis. Analysis of such iron restorationmaintenance studies reveals that ferritin increases from a pretreatment mean of 209 to 447 ng/ml and mean transferrin saturation (TSAT) from 22 to 35% (18). The optimal strategy for maintenance iron administration, i.e., the frequency of and dose amount, is controversial (15,16,19). Avoidance of iron-limited erythropoiesis anemia depends on its detection. Threshold values for iron repletion therapy in HD patients, TSAT 20% or a serum ferritin 100 ng/ml (5), are frequently inadequate to detect functional iron deficiency (20,21) because marrow iron deficiency can develop in HD

2 J Am Soc Nephrol 11: , 2000 Epoetin Therapy with Intravenous Iron Therapy 531 patients at TSAT values approaching 30% (20,21) or ferritin levels in excess of 500 ng/ml (22,23). The key questions in managing anemia then become: (1) At what levels of TSAT and ferritin are adequate amounts of iron being supplied to support optimal erythropoietic activity? (2) Are there readily available hematopoietic parameters that can detect functional iron deficiency and be measured sequentially to guide iron therapy? We have previously demonstrated that the use of standard intermittent need-based dosing strategy increased the amount of rhepo needed to maintain an average hemoglobin level of 10.5 g/dl when compared with a weekly maintenance intravenous iron dextran (ivid) regimen of 41 to 66 mg (15). Maintenance patients averaged a TSAT of 34.5% compared with only 26.1% for the need-based subjects. We hypothesized that TSAT values of 20 to 30% in ESRD patients could be associated with limited iron delivery to the erythron because of the ESRD-associated decreased plasma iron carrying capacity. The main objective of the present study was to determine whether a maintenance iron protocol that increased TSAT from conventional levels of 20 to 30% to 30 to 50% would increase erythropoiesis and result in reductions in rhepo doses needed to maintain a hemoglobin level of 9.5 to 12.0 g/dl when compared with a maintenance protocol that targeted TSAT levels of 20 to 30%. Secondary goals were to determine whether zinc protoporphyrin (ZPP) and reticulocyte hemoglobin content (CHr), two measures of ongoing erythropoiesis and iron deficiency in hemodialysis patients receiving Epoetin (24 26), could be used to assess adequacy of iron delivery for erythropoiesis. Materials and Methods Patient Selection A prospective, randomized control, intention-to-treat trial was designed, submitted to, and approved by the Institutional Review Board. Informed consent was obtained from all chronic hemodialysis patients recruited for the study. Eligibility criteria included the following: (1) age 18 yr; (2) hematologic parameters of mean cell volume of 80 fl, TSAT between 19 and 30%, serum ferritin between 150 and 600 ng/ml, hemoglobin 9.5 g/dl; (3) stable rhepo (Epogen ; Amgen, Thousand Oaks, CA) dose for anemia management over the previous 3mo( 25%), but this baseline dose had to exceed 700 U intravenously three times per week; and (4) no prior adverse reactions to parenteral iron. Exclusion criteria were hemolytic anemia, known aluminum toxicity, the presence of acute infection/inflammation, hematologic malignancies, active known acute or chronic gastrointestinal bleeding, or moderate hyperparathyroidism (parathyroid hormone 450 pg/dl). Study Design A single-center, open-label, randomized, comparative, prospective outpatient study was conducted in HD patients. Randomization to the two groups was performed after enrollment, at which time all oral iron was discontinued. Patients were then variably supplemented with parenteral iron (INFeD, Schein Pharmaceutical, Inc., Florham Park, NJ) by slow infusion ( 5 mg/min) to maintain TSAT between 20 and 30% during the next 16- to 20-wk run-in phase. Patients were considered to be in steady state if the maximal variation of hemoglobin was 1.2 g/dl (and without a trend by linear regression) and if the maximal variation in Epoetin dosage was 500 U/dose during the last 6 wk of the run-in phase. Once in steady state, patients entered one of the following groups to which they had been previously randomly assigned: Control Group: Intravenous iron dextran administered as weekly maintenance doses of 25 to 150 mg/wk to maintain TSAT at 20 to 30%. Study Group: Intravenous iron dextran administered initially as four to six 100 doses of 100 mg during consecutive hemodialysis sessions sufficient to increase TSAT to 30% and thereafter as weekly maintenance doses of 25 to 150 mg/wk to maintain TSAT at 30 to 50%. The measured monthly TSAT, the group assignment, and the ferritin level determined the amount of iron dextran administered to each patient. The investigator and staff were blinded to the ZPP and CHr measurement results so that these would not influence the ivid dosing decisions. Both groups received Epoetin (Epogen ; Amgen) to maintain the starting hemoglobin (9.5 to 12 g/dl). The rhepo dose was adjusted as necessary (but no more often than monthly) to keep the hemoglobin of each patient at its baseline level (the average of four determinations before randomization for each patient, generally 10 to 12 g/dl). The algorithm for Epoetin dose adjustments was 25% for every variation in hemoglobin of 1 g/dl within any 4-wk period (e.g., if hemoglobin increased by 1 g/dl within 4 wk, the rhepo dose was reduced by 25%). Blood counts and red blood cell indices were measured by standard automated counter methods. Hemoglobin (g/dl) was measured by a colorimetric method and monitored on a weekly basis. Iron indices (serum iron, TSAT, and ferritin) were evaluated monthly. If more than 100 mg of iron dextran had been given in the previous week, iron index measurements were delayed until the following week (19). If ferritin levels exceeded 1500 ng/ml, parenteral iron was withheld for at least 1 mo. Serum iron and total iron binding capacity (TIBC) were measured by a colorimetric method (Hitachi 747; Boehringer Mannheim, Indianapolis, IN). Serum ferritin (ng/ml) was measured monthly by a chemiluminescence immunoassay method (ACCESS; Beckman Instruments, Fullerton, CA) or by a heterogeneous competitive magnetic separation assay using Immuno-1 (Bayer Diagnostics, Tarrytown, NY). TSAT (%) was calculated monthly as serum iron/tibc. Parathyroid hormone and serum aluminum levels were measured every 12 wk unless Al containing phosphate binders were being used, in which case the frequency was increased to every 6 wk. Serum parathyroid hormone levels were measured by immunoradiometric assay at the reference laboratory (LIFE-CHEM, Rockleigh, NJ). ZPP and CHr were measured monthly beginning with month 0 (last month of the run in stabilization period). ZPP was measured fluorometrically at a reference laboratory (SmithKline Beecham, St. Louis, MO) with a normal reference range of 70 g/dl. CHr was measured by a flow cytometric method using a Bayer H3 hematology analyzer (Bayer Diagnostic, Tarrytown, NY) with a reported normal range of 24 to 36 pg (reference range at the time of our measurements, 24.5 to 31 pg). Functional iron deficiency was defined as possibly being present if ZPP was 100 g/dl (27) or a CHr 26 pg (26,28). Signs and symptoms were evaluated during and after each dialysis session, and the incidence of hospitalization, infections, and deaths was recorded. Our computerized database categorized 53 different events that could occur during dialysis.

3 532 Journal of the American Society of Nephrology J Am Soc Nephrol 11: , 2000 Statistical Analyses Sample size was based on a 40% difference in Epoetin dose at a P value of 0.05 and power of 80%; the required number of patients was 14 in each group. Data were collected and maintained in a computerized database. Results are expressed as mean SEM unless otherwise stated. Comparisons at baseline between groups used t test and 2 analysis. Monthly treatment group comparisons with respect to hemoglobin, CHr, serum iron, ferritin, TSAT, and ZPP, as well as the change in Epoetin dose and the change in iron dextran dose, were made using a repeated-measures ANOVA. Because the primary outcome was a change in Epoetin dose that could not be immediately discerned, a modified intention-to-treat analysis was used. For analytical purposes, the last observation was carried forward only for those subjects who finished at least 6 wk of the experimental phase of the study. An analysis was also performed for those who completed the entire 6 mo. Additional analyses included regression analysis. A P value 0.05 was considered significant. Results Sixty-seven patients provided informed consent, 44 completed the run-in period, and 42 met all inclusion criteria to enter randomized treatment. Nineteen and 23 of these 42 patients who achieved steady state had been randomized to the control group and the study group, respectively (Table 1). The Epoetin dosages at the time of enrollment were 4049 U/dose and 3650 U/dose in the control and study groups, respectively, a difference of 10%, which did not achieve statistical significance and the difference narrowed to 4% at entry into the randomized treatment period (Table 1). Thirty-seven patients (control group, 17 patients; study group, 20 patients) completed 6 wk of randomized treatment, and 32 patients (control group, 15 patients; study group, 17 patients) completed the entire 6 mo of study. Data are presented for the 37 patients who completed 6 wk of randomized treatment, using a last-observation-carried-forward approach for any patient who withdrew from the study before completion. The results were not altered by considering only those who completed the full study. Before entry into the randomized treatment phase (last month of the run-in stabilization period or month 0), the two treatment groups were similar in age, male:female ratio, hemoglobin levels, Epoetin dose, iron indices of TIBC, TSAT, and ferritin, and iron requirements (Table 1). Hyperparathyroidism was of mild-to-moderate severity in both groups with mean/median values of 275/186 pg/ml and 262/219 pg/ml for the control and study groups, respectively. Alkaline phosphatase levels were equal at U/L and , respectively. Baseline serum albumin was and in the control and study groups, respectively. Table 2 shows changes in various parameters during the course of the experimental phase. At baseline, mean values of both ZPP and CHr were normal for both groups. Ten of the 37 patients had at baseline either a ZPP 100 g/dl or a CHr 26 pg, indicative of possible iron-limited erythropoiesis. Mean ZPP did not change in either group during the randomized treatment phase of the study (Table 2). At baseline, seven of the 37 subjects had a CHr 26 pg (iron deficiency) despite maintenance iron administration that maintained TSAT 20% Table 1. Demographic characteristics and baseline erythropoietic parameters of the control and study treatment groups at the time of entry into the randomized phase a Variable Control Group Study Group No. of patients male female 9 8 Age (yr) Age range (yr) 30 to to 90 Hemoglobin (g/dl) Serum ferritin (ng/ml) TIBC ( g/dl) Transferrin saturation (%) Maintenance iron (mg/mo) during run-in Total iron during run-in phase (mg) (range) (0 to 2000) (0 to 1800) rhepo dose (U, 3 /wk) rhepo dose range (U) 700 to to 9000 a All results are mean SEM. n 42. None of the variables differed significantly between the groups. The randomized phase followed the 16- to 20-wk run-in phase. The study includes only patients who completed the run-in period and met steady-state requirements. TIBC, total iron binding capacity; rhepo, recombinant human erythropoietin. and ferritin 150 ng/ml. In three of four control group subjects, basal low CHr levels persisted throughout the experimental phase, but a TSAT 20% was seen only transiently. These subjects received the same amount of iron as those control subjects who never developed CHr 26 pg during the treatment phase. A basal CHr 26 pg was corrected in all three study group subjects during the treatment phase of the study. CHr 26 pg developed in three additional control and three study group subjects at some time during the experimental phase even though ferritin 150 ng/ml and TSAT 19% was maintained in five of the six subjects. When all available data were analyzed, there was a weak positive correlation (CHr ZPP, r 0.26, P 0.05) between CHr and ZPP. On multivariate regression, the only significant correlation for CHr was with serum iron (HCr Fe, r ) and previously significant univariate correlations between CHr and TSAT, and CHr and TIBC dropped out of the model. Baseline Epoetin dose was independent of CHr. The hemoglobin level remained unchanged in both groups during the 6-mo randomized treatment period (Figure 1, top panel). Average hemoglobin for all measured values during the 6 mo was in the control group and in the study group. Repeated-measures ANOVA showed no effect of time in either group. As shown in Figure 1 (bottom panel), the doses of Epoetin in the control group remained essentially constant throughout the 6-mo randomized treatment period. The overall mean Epoetin dose in the control group over this

4 J Am Soc Nephrol 11: , 2000 Epoetin Therapy with Intravenous Iron Therapy 533 Table 2. Average monthly values for Hb, CHr, serum iron, ferritin, TSAT, and ZPP a Time Period Hb (g/dl) CHr (pg) Serum Fe ( g/dl) Ferritin (ng/ml) TSAT (%) ZPP b ( g/dl) Baseline control study Month 1 control study Month 2 control study Month 3 control study Month 4 control study Month 5 control study Month 6 control ND study ND a The study includes patients who completed 6 wk of randomized treatment. Hb, hemoglobin; CHr, reticulocyte hemoglobin content; TSAT, transferrin saturation; ZPP, zinc protoporphyrin. b Data available for 12 and 14 patients in the control and study groups, respectively. period was U. Within the study group, the dose of Epoetin progressively decreased over the 6-mo randomized treatment period. The reduction in dose between the study group and the control group reached approximately 40% at each of months 4, 5, and 6. We defined a 20% change in Epoetin dose as being clinically significant, since it approximated our dosing algorithm. The experimental group had more subjects with decreases in Epoetin dose (12 versus 4) during the last 3 mo of the study compared with amount received during their 4 mo stabilization period and fewer patients with increases (2 versus 5) or no change in dose (6 versus 9). The overall effect was highly significant ( , P ). The change in Epoetin dose at the end of the study was independent of the baseline dose in both the Study Group (r , NS) and the control group (r , NS). During the 4-mo run-in period, the amount of iron administered to the two groups was similar: mg/mo in the control group and mg/mo in the study group. The iron dextran administered to patients in the control group to maintain a TSAT of 20 to 30% (mean SEM TSAT % over months 1 to 6) remained unchanged, averaging mg/mo over the 6 mo (Figure 2, middle panel). The monthly doses during months 1 to 6 were not associated with any progressive increases in ferritin (Table 2; Figure 2, bottom panel) or any temporal changes in ZPP or CHr (Figure 3). The mean dose of iron dextran in the study group increased nearly fourfold during the first treatment month to mg (Figure 2, middle panel), in part due to the four to six 100-mg doses administered during consecutive hemodialysis to achieve an initial TSAT 30%. The monthly iron dextran dose required by the study group over the next 5-mo randomized treatment period to maintain the target TSAT level 30% (TSAT % over months 2 to 6) remained high at mg/mo and essentially constant. Administration of these doses was associated with progressive increases in mean ferritin as shown in Figure 2 (from 272 ng/ml at month 0 to 730 ng/ml at month 6). Mean CHr increased from 28.5 pg at month 0 to a peak of 30.1 pg at month 5 (Figure 3). Because of the large difference in ferritin levels between month 6 and month 0 in the control group (24 37 ng/ml) compared to the study group ( ng/ml), we performed a stepwise regression to determine which factors influenced the change in ferritin level between month 6 and month 0 during the randomized treatment phase of the study (all 37 randomized patients). Only the monthly dose of iron (F 20.3) and the presence/absence of an inflammatory state in the last 3 mo of the study (vida infra, Safety) (F 12.3) were statistically significant factors (both P 0.01), whereas the change in hemoglobin, the baseline TSAT value, and the change in TSAT from baseline, iron dextran dose, Epoetin dose, ferritin, base-

5 534 Journal of the American Society of Nephrology J Am Soc Nephrol 11: , 2000 Figure 2. Effect of intravenous iron dextran loading on TSAT and serum ferritin. TSAT separated after the first month in the two groups. Ferritin levels increased in the study group patients (Œ) who received an average of 522 mg/mo of iron dextran. The control group patients (F) received an average of 191 mg/mo during the 6-mo period. Figure 1. Effect of intravenous iron dextran loading on erythropoietin requirements. A 4-mo period was followed by a 6-mo experimental phase. Control subjects (F) were maintained at a transferrin saturation (TSAT) of 20 to 30% and study subjects (Œ) at a TSAT of 30 to 50%. Erythropoietin (EPO) dosage was adjusted every 4 wk to maintain hemoglobin at the same value as at month 0 (top panel). Iron loading with iron dextran (middle panel) decreased EPO dose requirements (bottom panel). line CHr, and the change in CHr were not. When only the patients who finished all 6 mo of study were analyzed, baseline CHr also became a significant predictor. The relationship between change in Epoetin dose and CHr is shown in Figure 4. Epoetin dose was averaged over the last 3 mo of the randomized experimental treatment phase and factored by the dose at month 0 to provide an Epoetin dose ratio. In both groups, there was an inverse relationship between the Epoetin ratio and the change in CHr best fit by an inverse exponential relationship. Comparison of the relationships indicates a significant change in intercept but not slope for the two groups. In the control group, there was no change from baseline in CHr during months 4 to 6 of the study: pg compared with an increase of pg in the study group (P 0.05). The CHr increased in 15 of the 20 study group patients. Alkaline phosphatase and parathyroid hormone were unchanged and similar for both treatment groups (Figure 5) at all time points. The aluminum levels were low (averages 10 g/l with a range of 5 to 31 g/l) and remained unchanged in both groups (data not shown). Safety Three patients died during this study. Two patients in the control group died as a result of underlying disease during the third (cerebrovascular accident [CVA]) and fifth (post-laminectomy surgery) months of the randomized treatment period. One patient in the study group died in the fifth month as a result of catheter-associated sepsis. This elderly patient on hemodialysis for 8 yr had a medical history of hypertension, insulin-dependent diabetes mellitus, CVA, and multiple lacunar infarcts, bilateral above the knee amputation, multiple sacral decubiti, and three permanent accesses that had required five salvage procedures. Because of the latter complications, the patient had had 12 central vein catheters and four previous catheter-related sepsis episodes before entry into the study. She required four catheters during the 5 mo in the study. Late in the third month of the treatment phase, her catheter malfunctioned and was replaced. Ferritin was 450 ng/ml, TSAT 28%, and albumin 3.8 g/dl. Only femoral vein access remained available. Over the course of the next 6 wk, the patient had multiple debridements of her sacral decubiti, developed hypoalbuminemia (3.3 g/dl), and had a preterminal episode of line-related sepsis that did not clear even with line removal and vancomycin. Ferritin was 714 ng/dl, TSAT 37% just before the event. Death followed a massive CVA.

6 J Am Soc Nephrol 11: , 2000 Epoetin Therapy with Intravenous Iron Therapy 535 Figure 3. Effect of intravenous iron dextran loading on two indices of iron utilization. Reticulocyte hemoglobin content (CHr) increased in the iron-loaded patient (Œ) but remained constant in the control group (F). Zinc protoporphyrin (ZPP) did not change in either group. Five patients were prematurely discontinued from study. In the control group, one arose from a massive cerebrovascular event on last day of study and the other from loss to follow-up after 5 mo. Within the study group, one patient was discontinued due to development of acute gastrointestinal bleeding, one patient due to mastectomy complicated by a cerebrovascular event, and one patient due to kidney transplant. No differences in hospitalizations or infection rate were noted. Ten of the 17 control group subjects were hospitalized for non-transplant-related causes on 15 occasions compared with 10 of the 19 study group patients on 15 occasions. Each group had one admission for an infectious etiology (pneumonia in the control group, line-related sepsis in the study group). Infection of inflammation (foot ulcer, inflammatory bowel disease, leukocytosis of unclear cause, chronic bronchitis, arthritis with joint effusion, persistent gastritis) occurred during the last 3 mo of the study in seven of 17 control and six of 19 study patients. Our computerized database categorizes 53 different events occurring during dialysis. None of these differed between the two groups, including the frequency of hypotension, shortness of breath, chest pain, and arthralgia. Cost Effectiveness Comparative treatment costs were determined by comparing monthly and average wholesale price total costs of supplying the two groups with INFeD and rhepo. During weekly maintenance therapy, costs for syringes, saline diluent, and iv Figure 4. Effect of changes in CHr during the experimental phase of the study on Epoetin dose. CHr is the average value for each patient during the last 3 mo of study. EPO dose ratio represents the average dose of Epoetin during the last 3 mo of study divided by the dose at month 0 (entry into study). F, control patients; E, study patients. administration sets are identical in both groups. Similarly, since our unit policy is to measure iron indices monthly, the cost of these is equal in the two groups. The individual monthly costs, as well as the total costs for months 2 through 6, were lower for the study regimen compared with the control regimen. The cost difference reached statistical significance by the third month and remained so for the remainder of the study (P 0.02). Overall cost analysis revealed a potential savings of $109/mo or $1308/yr per patient with maintenance of the TSAT between 30 and 50%. Discussion This study demonstrates the efficacy of maintenance doses of iron dextran in maintaining serum TSAT levels of 20 to 50% over a period of 6 mo. The results indicate that increasing TSAT from 20 to 30% to 30 to 50% with ivid in iron-replete hemodialysis patients results in a decreased need for Epoetin to maintain hemoglobin of 10 to 12 g/dl. Because stable hemoglobin levels were maintained, the rate of erythropoiesis was not altered but the responsiveness to Epoetin was modified since lower doses could be used. Our study is the first to demonstrate that cost effectiveness is possible even in ironreplete patients when iron is aggressively supplemented. Our findings suggest that decreased transferrin levels associated with ESRD limit the delivery of iron to the erythron at TSAT levels in the 20 to 30% range. We believe that the overall effect of reducing Epoetin dose in the experimental group reflects the increase in iron delivery to tissues when higher TSAT were achieved by intensive iron therapy. Classical studies by Cazzola et al. (27) showed that delivery of iron to tissues increases as the amount of iron bound to transferrin also increases, but the number of transferrin molecules that

7 536 Journal of the American Society of Nephrology J Am Soc Nephrol 11: , 2000 Figure 5. Serum parathyroid hormone (PTH) and alkaline phosphatase (Alk Phos) levels over the course of the study. There were no differences between the two groups or a change in parameters in either group over time. leave the plasma per unit time does not. In plasma, transferrin exists as a mixture of unsaturated, partially saturated monoferric, and fully saturated diferric transferrin. Diferric transferrin has a 4.2-fold greater advantage over monoferric transferrin (28) in delivering iron to the developing red blood cell. Erythroid transferrin uptake in turn depends on the activity of the bone marrow, increasing when erythropoiesis is stimulated (29). For the same target hemoglobin, if the limitation to effective erythropoiesis is due to inadequate iron delivery due to decreased transferrin, then less Epoetin is needed if iron delivery can be maximized. Changes in Epoetin dose following deliberate increases in TSAT were not uniform. Changes in Epoetin dose and in CHr in response to parenteral iron were quite variable despite the intent to make the groups as homogeneous as possible. As shown in Figure 5, there was significant dispersion in the change in CHr from baseline in both groups, with CHr serving as an indicator for iron availability to the erythron. These findings reflected the development of intercurrent events such as unsuspected blood losses as well as inflammatory processes. The latter occurred in one-third of the patients but was of equal frequency in the two groups. We used CHr retrospectively to determine the effectiveness of iron incorporation at the level of the bone marrow. Mittman et al. (26) demonstrated that a CHr value of 28 pg is a more sensitive marker of functional iron deficiency in Epoetintreated hemodialysis than transferrin saturation and ferritin. Fishbane et al. (25) found a slightly lower value of 26 pg as the best discriminator for the presence of functional iron deficiency but used a larger 1000-mg dose. Our baseline mean CHr values of 28 to 28.5 pg are comparable to those obtained by others for HD patients (43,45), although higher mean baseline values of 31 to 33 pg have been observed in adult normal volunteers (30,31). In our study, changes in Epoetin doses correlated with the change in HCr for both the control and experimental groups (Figure 5). The curve is shifted downward for patients maintained at a higher TSAT, i.e., loading transferrin with iron alters the response to Epoetin, but the absolute effect achieved depends on the directional change in CHr, which was not under the control of the investigator. The best predictor of CHr is the serum iron and not the TSAT or ferritin. Unlike our previous experience with maintenance iron (15), we had considerably more difficulty in this study in achieving and then maintaining a TSAT of 30 to 50%. In our previous study, once a mini-loading iron dose had been administered we were able to maintain the higher TSAT at a stable ferritin level with much lower ivid doses (232 mg/mo) than in the current study. However, unlike the design of the current study, no predetermined TSAT level determined the iron dosing in that study. Also, the patients in the current study were much more heterogeneous with considerably greater comorbidity. How much these differences in design and population characteristics contributed to the progressive increase in ferritin is unclear. Certainly, the administration of about 500 mg/mo of ivid to the study group far exceeds the average estimated losses of about 150 to 250 mg that we noted in both groups before randomization and in the control group throughout the 6-mo experimental period. Although parenteral iron in this study was well-tolerated over the 6-mo randomized treatment period without any discernible evidence of toxicity as measured by morbidity, hospitalization, and clinical events during dialysis, the sample size is small and the period of observation is short. Because of the potential toxicity of elevated iron burdens, a patient who demonstrates a progressive increase in ferritin during iron therapy without a hematopoietic response should not continue to receive iron. Higher than expected iron dosages in the study group resulted from several unexpected differing processes. First, three of 19 study group patients (and almost 20% of the entire cohort) had baseline CHr of 26 pg, values indicative of functional iron deficiency at study entry, although TSAT was 20% and ferritin 150 ng/dl. These individuals needed more iron than iron-replete patients to correct their baseline deficiency. Second, seven of 19 study patients had baseline CHr 30 and were unlikely to benefit from additional iron loading. Six of these seven had enrollment TSAT 30%, and in four of them a TSAT 30% was difficult to achieve despite large amounts of iron. On retrospective review, all four developed some degree of reticuloendothelial blockade because of a foot ulcer, bronchitis, arthritis with effusion, and a flare of inflammatory bowel disease, respectively. Our protocol did not withhold iron during such events. In these patients, iron loading driven by the TSAT target that could not be attained led to large increases in ferritin without an increase in CHr. A rising ferritin level after iron administration may be a marker for

8 J Am Soc Nephrol 11: , 2000 Epoetin Therapy with Intravenous Iron Therapy 537 patients with reticuloendothelial blockade. The largest increases in ferritin did occur in patients with infection or inflammation. During the experimental phase, one patient s CHr decreased from 27.9 to 24.6 pg despite administration of 4.5 g of iron over the 6-mo period. This patient had increased blood losses. Finally, because of the poor correlation between CHr and TSAT, a CHr 30 pg was achieved in some patients, yet fluctuations in TSAT led to continued high iron dosing. ZPP was not useful in the evaluation or serial evaluation of iron delivery to the erythron in this study. We concur with the observations of Braun et al. (32) that ZPP cannot be used to predict the response to iron supplementation in patients receiving maintenance intravenous iron. Overall, the changes in CHr in this study were modest, which may be related to the fact that the great majority of patients in this study were not in a state of absolute iron deficiency, having received sufficient iron in the 4-mo run-in period to maintain serum ferritin 200 ng/ml and TSAT 20%. Averaged over the last 3 mo of the study, administration of large amounts of ivid increased the mean CHr by pg compared with baseline. This modest increase did lead to a 40% reduction in Epoetin dose, similar to the results of Fishbane et al. (14), who were able to lower the mean rhepo dose by 46%. Conclusion This study demonstrates that ivid (INFeD ) administered initially as a loading regimen and thereafter as intermittent maintenance doses to maintain a TSAT 30% can increase erythropoiesis in ESRD patients receiving chronic hemodialysis and concurrently can decrease the dose of Epoetin required to keep the hemoglobin in the range of 10 to 12 g/dl. The lower Epoetin requirements when maintaining TSAT 30% can potentially result in significant cost savings. Measurements of ZPP levels before and after effective erythropoietic therapy failed to correlate with effective erythropoiesis. Changes in CHr correlated with changes in Epoetin dose, and this parameter should be evaluated further as a primary determinant of iron need in patients on Epoetin. References 1. Eschbach JW, Kelly MR, Haley NR: Treatment of the anemia of progressive renal failure with recombinant human erythropoietin. N Engl J Med 321: , Eschbach JW, Abdulhadi MH, Browne JK: Recombinant human erythropoietin in anemic patients with end-stage renal disease: Results of a phase III multicenter clinical trial. Ann Intern Med 111: , Besarab A, Flaharty KK, Erslev AJ: Clinical pharmacology and economics of recombinant human erythropoietin in end-stage renal disease: The case for subcutaneous administration. JAm Soc Nephrol 2: , Besarab A, Bolton WK, Browne JK, Egrie JC, Nissenson AR, Okamaoto DM, Schwab SC, Goodkin DA: The effects of normal versus anemic hematocrit on hemodialysis patients with cardiac disease. N Engl J Med 339: , National Kidney Foundation-Dialysis Outcomes Quality Initiative: Clinical practice guidelines: Anemia of chronic renal failure. Am J Kidney Dis 30[Suppl 3]: S202 S212, Macdougall IC: Iron management in patients receiving erythropoietin therapy. Semin Dial 11: 10 13, Lindsay RM, Burton JA, King P: The measurement of dialyzer blood loss. Clin Nephrol 1: 24 28, Hörl WH: Consensus statement. How to diagnose and correct iron deficiency during rhuepo therapy: A consensus report. Nephrol Dial Transplant 11: , Macdougall IC: Monitoring of iron status and iron supplementation in patients treated with erythropoietin. Curr Opin Nephrol Hypertens 3: , Macdougall IC, Cavill I, Hulme B: Detection of functional iron deficiency during erythropoietin treatment: A new approach. Br Med J 304: , Major A, Mathez-Loic F, Rohling R, Gautschi K, Brugnara C: The effect of intravenous iron on the reticulocyte response to recombinant human erythropoietin. Br J Haematol 98: , Eschbach JW, Haley NR, Egrie JC, Adamson JW: A comparison of the responses to recombinant human erythropoietin in normal and uremic subjects. Kidney Int 42: , Macdougall IC, Tucker B, Thompson J, Tomson CR, Baker LR, Raine AE: A randomized controlled study of iron supplementation in patients treated with erythropoietin. Kidney Int 50: , Fishbane S, Frei GL, Maesaka J: Reduction in recombinant human erythropoietin doses by the use of chronic intravenous iron supplementation. Am J Kidney Dis 26: 41 46, Besarab A, Kaiser JW, Frinak S: A study of parenteral iron regimens in hemodialysis patients. Am J Kidney Dis 34: 21 27, Taylor JE, Peat N, Porter C, Morgan AG: Regular low-dose intravenous iron therapy improves response to erythropoietin in haemodialysis patients. Nephrol Dial Transplant 11: , Sepandj F, Jindal K, West M, Hirsch D: Economic appraisal of maintenance parenteral iron administration in treatment of anaemia in chronic haemodialysis patients. Nephrol Dial Transplant 11: , Besarab A, Frinak S, Yee J: An indistinct balance: The safety and efficacy of parenteral iron therapy. J Am Soc Nephrol 10: , Auerbach M, Winchester J, Wahab A: A randomized trial of three iron dextran infusion methods for anemia in rhuepotreated dialysis patients. Am J Kidney Dis 31: 81 86, Fishbane S, Kowalski EA, Imbriano LJ, Maesaka JK: The evaluation of iron status in hemodialysis patients. J Am Soc Nephrol 7: , Eschbach JW, Egrie JC, Downing MR, Brown JK, Adamson JW: Correction of anemia of end stage renal disease with rhu-epo. N Engl J Med 316: 73 78, Ali M, Rigolosi R, Fayemi AO, Braun EV, Frascino J, Singer R: Failure of serum ferritin levels to predict bone-marrow iron content after intravenous iron-dextran therapy. Lancet 1: , Kernilde J-J, Folkert V, Mokrzycki M: Functional iron deficiency in hemodialysis patients with high ferritin levels [Abstract]. J Am Soc Nephrol 9: 253A, Fishbane S, Lynn RI: The utility of zinc protoporphyrin for predicting the need for intravenous iron therapy in hemodialysis patients. Am J Kidney Dis 25: , 1995

9 538 Journal of the American Society of Nephrology J Am Soc Nephrol 11: , Fishbane S, Galgano C, Langley RC Jr, Canfield W, Maesaka JK: Reticulocyte hemoglobin content in the evaluation of iron status of hemodialysis patients. Kidney Int 52: , Mittman N, Sreedhara R, Mvshnick R: Reticulocyte hemoglobin content predicts functional iron deficiency in hemodialysis patients. Am J Kidney Dis 30: , Cazzolla M, Huebers HA, Sayers MH, MacPhail PA, Eng M, Finch CA: Transferrin saturation, plasma iron turnover, and transferrin uptake in normal humans. Blood 66: , Huebers HA, Csiba E, Huebers E, Finch CA: Molecular advantage of diferric transferrin in delivering iron to reticulocytes: A comparative study. Proc Soc Exp Biol Med 179: , Cazzola M, Pootrakul P, Huebers HA, Eng M, Eschbach J, Finch CA: Erythroid marrow function in anemic patients. Blood 69: , Major A, Mathez-Loic F, Rohling R, Gautschi K, Brugnara C: The effect of intravenous iron on the reticulocyte response to recombinant human erythropoietin. Br J Haematol 98: , Breymann C, Bauer C, Major A, Zimmermann R, Gautschi K, Huch A, Huch R: Optimal timing of repeated rh-erythropoietin administration improves its effectiveness in stimulating erythropoiesis in healthy volunteers. Br J Haematol 92: , Braun J, Hammerschmidt M, Schreiber M: Is zinc protoporphyrin an indicator of iron-deficiency erythropoiesis on maintenance hemodialysis patients? Nephrol Dial Transplant 11: , 1996 Access to UpToDate on-line is available for additional clinical information at