PEDIATRICS Vol. 107 No. 1 January 2001, pp. 78-85

Effect of Intravenous Iron Supplementation on Erythropoiesis in Erythropoietin-Treated Premature Infants

Arnold Pollak, MD^*, Michael Hayde, MD^*, Marianne Hayn, PhD, Kurt Herkner, PhD^*, Kenneth A. Lombard, MD^§, Gert Lubec, MD^*, Manfred Weninger, MD^*, and John A. Widness, MD

From the ^* Department of Neonatology, University Children's Hospital, Vienna, Austria;  Institute for Biochemistry, University of Graz, Austria; ^§ Department of Pediatrics, Barbara Bush Children's Hospital at Maine Medical Center, Portland, Maine; and the  Department of Pediatrics, University of Iowa, College of Medicine, Iowa City, Iowa.

	ABSTRACT

Top Abstract Methods Results Discussion Conclusion References

Objective.  To test the efficacy and safety of combining intravenous iron in amounts approximating the in utero iron accretion rate and the postnatal iron loss with erythropoietin (EPO) in very low birth weight (VLBW) infants.

Methods.  A prospective, controlled, randomized, unmasked trial lasting 21 days was performed in 29 clinically stable VLBW infants <31 weeks' gestation and <1300 g birth weight not treated with red blood cell transfusions during the study period. Mean (± standard error of the mean) age at study entry was 23 ± 2.9 days. After a 3-day run-in baseline period in which all participants received oral supplements of 9 mg/kg/day of iron polymaltose complex (IPC), participants were randomized to receive 18 days of treatment with: 1) oral IPC alone (oral iron group); 2) 300 U of recombinant human EPO (r-HuEPO) kg/day and daily oral IPC (EPO + oral iron group); 3) 2 mg/kg/day of intravenous iron sucrose, r-HuEPO, and oral iron (intravenous iron + EPO group). To assess efficacy of the 3 treatments, serial blood samples were analyzed for hemoglobin (Hb), hematocrit (Hct), reticulocyte count, red blood cell indices and plasma levels of transferrin, transferrin receptor (TfR), ferritin, and iron. Oxidant injury was assessed before and after treatment by plasma and urine levels of malondialdehyde (MDA) and o-tyrosine.

Results.  At the end of treatment, Hb, Hct, reticulocyte count, and plasma TfR were markedly higher in both of the EPO-treated groups, compared with the oral iron group. At study exit a trend toward increasing Hb and Hct levels and significantly higher reticulocyte counts were observed in the intravenous iron + EPO group, compared with the EPO + oral iron group. During treatment, plasma ferritin levels increased significantly in the intravenous iron + EPO group and decreased significantly in the other 2 groups. By the end of treatment, ferritin levels were significantly higher in the intravenous iron + EPO group compared with the other 2 groups. Although plasma and urine MDA or o-tyrosine did not differ among the 3 groups, plasma MDA was significantly greater in the subgroup of intravenous iron + EPO participants sampled at the end of the 2-hour parenteral iron infusion, compared with values observed immediately before and after parenteral iron-dosing.

Conclusions.  In stable VLBW infants receiving EPO treatment, parenteral supplementation with 2 mg/kg/day of iron sucrose results in a small, but significant, augmentation of erythropoiesis beyond that of r-HuEPO and enteral iron alone. However, to reduce the potential adverse effects of parenteral iron/kg/day on increasing plasma ferritin levels and on causing oxidative injury, we suggest that the parenteral iron dose used should be reduced and/or the time of infusion extended to maintain a serum iron concentration below the total iron-binding capacity.  Key words:  intravenous, iron supplementation, erythropoiesis, erythropoietin-treated infants, very low birth weight infants.

Before hospital discharge, 60% to 80% of very low birth weight (VLBW) infants weighing <1.5 kg at birth receive one or more red blood cell (RBC) transfusions as treatment for anemia resulting from laboratory phlebotomy loss and from physiologic factors.¹ Because clinical trials of erythropoietin (EPO) in this patient group have yielded only moderate success in reducing RBC transfusions,^2,3 the development of promising strategies for improving the efficacy of EPO is desirable. Several reports have indicated that depleted iron stores or limited bioavailability of iron for hemoglobin (Hb) synthesis may attenuate the effectiveness of EPO in premature infants.^4-8

The optimal dose and mode of iron administration in EPO-treated VLBW infants are uncertain. In some neonatal trials, oral iron supplements have failed to enhance erythropoiesis.¹ A theoretic advantage of parenteralcompared with enteraliron is the immediate availability of the former for Hb synthesis.⁹ Although the beneficial effects of parenteral iron on erythropoiesis have been convincingly demonstrated among EPO-treated adults with end-stage renal disease,¹⁰ the combination of parenteral iron and EPO administered to VLBW infants has been inconclusive. This uncertainty is likely a consequence of methodological differences in the few published reports of parenteral iron in EPO-treated premature infants.^6,11-14 The paucity of parenteral iron studies in neonates^11-14 is, in part, the result of previous concerns about the potential of iron for causing oxidant injury¹⁵ and increasing infections among VLBW infants.¹⁶

In the present study we hypothesized that intravenous iron administered in doses approximating the in utero fetal iron accretion rate and postnatal laboratory iron loss would be more effective than EPO alone in providing readily available iron for erythropoiesis in rapidly growing VLBW infantsbut without associated oxidant injury. To test this hypothesis, biochemical indicators of erythropoiesis, iron status, and oxidant injury were compared among 3 infant study groups: those treated with oral iron alone, those treated with EPO and oral iron, and those treated with parenteral iron, EPO, and oral iron. Because a primary objective of this work was to achieve clinically significant increases in erythropoiesis in VLBW infants, we administered what we considered to be highyet safedoses of both recombinant human erythropoietin (r-HuEPO)² and iron.^7,17

	METHODS

Top Abstract Methods Results Discussion Conclusion References

Participants

The ethics committee of the University of Vienna approved the study and written parental consent was obtained.

Eligible infants included those with gestational age <31 weeks and weight <1300 g at birth. Additional inclusion criteria were postnatal age >1 week, hematocrit (Hct) >30%, phlebotomy loss averaging <1.0 mL/day over the previous 3 days, enteral caloric intake >20 kcal/kg/day and stable respiratory status defined as mean oxygen saturation >90% in a fraction of inspired oxygen .40 for >48 hours regardless of ventilation. Excluded were infants who received RBC transfusions 3 days before study entry or during the study, those with an active, culture-proven infection, a major malformation, electroencephalographic-proven seizures, hypertension,¹⁸ intraventricular hemorrhage grade III, immune hemolytic disease, necrotizing enterocolitis (NEC), or surgery.

Study Design

The study was a prospective, controlled, randomized, unmasked trial of 21 days duration. Consecutive infants meeting both birth weight and gestational age criteria were evaluated for enrollment. Those enrolled were stratified based on gestation at birth 28 weeks and >28 weeks. Randomization was performed by drawing sequentially numbered sealed envelopes in blocks of 6 for each stratification group. Infants were removed from study during the treatment for acute illness necessitating the discontinuation of enteral feedings or for protocol violations. Additionally, infants receiving RBC transfusions were removed to permit an objective assessment of the effect of EPO on Hb and Hct levels. Disqualified infants were replaced by the next eligible infant in the same stratification group.

The 21-day study consisted of a 3-day run-in baseline period during which 9 mg/kg/day of iron polymaltose complex (IPC; Ferrum Hausmann Syrup, Vifor International, St Gallen, Switzerland) supplementation was administered to all participants in all groups. This was followed by an 18-day treatment period during which participants received: 1) the same oral iron supplementation dose alone (oral iron group); 2) 300 U/kg/day of r-HuEPO (Erypo, Janssen-Cilag Pharma, Vienna, Austria) as an intravenous bolus infusion administered at 3-day intervals along with the same oral iron supplement as the oral iron group (EPO + oral iron group); or 3) 2 mg of intravenous iron sucrose/kg/day (Venofer, Vifor International) diluted in .9% of sodium chloride to a final concentration of 2 mg/mL and infused daily over 2 hours (intravenous iron + EPO group). To maintain comparability of iron intake among the 3 groups, this last group also received EPO and oral iron in an identical manner as the EPO + oral iron group.

Study infants were gavage fed or fed by nipple every 3 hours using their mother's milk supplemented with powdered milk fortifier containing whey protein, maltodextrin, and minerals. When 5 g of the fortifier is added to 100 mL of human milk, the mixture has a caloric density of ~.85 kcal/mL (FM85, Nestlé, Vevey, Switzerland). All study infants received their daily enteral iron supplement in 4 equally divided doses immediately before feedings. To enhance iron absorption, ascorbic acid was administered simultaneously with the iron in a ratio of 5 mg ascorbic acid to 1 mg of elemental iron. Daily vitamin E and folate supplements were also provided in doses of 25 IU and 50 µg/kg, respectively.

Hematologic, iron status, and oxidation parameters were determined on peripheral venous or on arterial blood drawn on study days 3, 0, 4, 11, and 18. Hematologic parameters were analyzed immediately. Blood samples for iron status and oxidation were centrifuged and aliquoted into plasma and RBC fractions for later analysis. Six infants in the intravenous iron + EPO group had day 0 blood samples drawn immediately before the intravenous iron infusion. The remaining 4 had blood samples drawn immediately after the 2-hour iron infusion. Urine was collected on study days 3, 0, and 4 for determination of malondialdehyde (MDA) and o-tyrosine.

Analytical Procedures

Blood counts, reticulocyte counts, and RBC indices were performed by flow cytometry on 150 µL of whole blood (Technicon H.3 Autoanalyzer Bayer Austria GesmbH, Vienna, Austria). The percentage of hypochromic RBCs was calculated by the instrument based on the frequency distribution analyses of individual cells in the whole blood sample. The flow cytometric analyzer used was gated for neonatal values. Iron was measured in duplicate on 20-µL plasma aliquots using an electrochemical method (Ferrochem II analyzer, ESA, Inc, Bedford, MA). Ferritin was determined in duplicate on 150-µL plasma aliquots by a radioimmunoassay (Bio-Rad Quantimmune Ferritin IRMA, Bio-Rad Diagnostics Group, Hercules, CA). Transferrin was measured on 20-µL duplicate aliquots of plasma by an antigen-antibody turbidimetric method using rabbit human transferrin antiserum (Boehringer Mannheim, Indianapolis, IN).¹⁹ Transferrin receptor (TfR) concentration was determined by an enzyme-linked immunoadsorbent assay procedure (Ramco Laboratories, Inc, Houston, TX) on 10-µL duplicate plasma aliquots and read in a Titertek Multiscan Plus enzyme-linked immunoadsorbent assay Reader (Lab Systems, Helsinki, Finland). Determinations of plasma and urinary o-tyrosine were performed by reversed-phase high-performance liquid chromatography²⁰ and MDA by a thiobarbituric acid high-performance liquid chromatography assay.²¹

Statistical Analyses

Because there have been no erythropoietic data reported in VLBW infants using the high doses of r-HuEPO and iron used in the present study, no power analysis was performed. As such, the present study represents a pilot study. Statistical analyses were performed using microcomputer software (StatView 5.0, Abacus Concepts Inc, Berkeley, CA). Natural logarithm transformed values were used in the analysis of study variables demonstrating a nonparametric distribution. Within-group comparisons were performed by paired t test (2-tailed). Between-group comparisons were performed using analysis of variance and ² testing. For significant F ratios, posthoc comparisons were performed using Fisher's least significant difference procedure. Simple regression analysis was used to test for significant associations among the study variables. Results are expressed as mean ± standard error of the mean (SEM). An  error <.05 was considered significant.

	RESULTS

Top Abstract Methods Results Discussion Conclusion References

From August 1995 to May 1997, 125 infants met birth weight and gestational age entry criteria. Thirty-eight infants were enrolled as study participants, 2 eligible infants were not offered enrollment and the remaining 85 infants were excluded for reasons of death (n = 13), major malformations (n = 10), sepsis (n = 14), intraventricular hemorrhage (n = 6), NEC or suspected NEC (n = 6), surgery of gastrointestinal tract (n = 4), Rh isoimmune disease (n = 1), transfer to another hospital (n = 9), and parental refusal (n = 18). Although no infant was disqualified for seizures or hypertension, 4 additional infants were deemed too sick for enrollment by their attending physician.

Of the 38 study participants who began the study, 9 were disqualified during the treatment period. Those disqualified included 3 infants with sepsis or sepsis-like episodes, 2 with NEC, 2 who received RBC transfusions, and 2 for a protocol violation. Each of the 2 transfused infants met standard NICU transfusion criteria.⁸ The group assignments of the disqualified infants were similar to those who successfully completed the study, ie, 2 in the oral iron group, 3 in the EPO + oral iron group, and 4 in the intravenous iron + EPO group (P = .53).

Demographic and clinical characteristics of the 3 study groups at birth were similar with respect to weight, gestational age, sex, multiple birth, Apgar score, initial pH, and the proportions of small for gestational age and inborn infants (Table 1). At study entry, age, weight, prestudy phlebotomy loss, and the number and volume of RBC transfusions were also not different among the 3 groups. There was no difference in erythropoietic response between infants who did or did not receive RBC transfusions before the study. Demographic and clinical characteristics of the 9 infants disqualified during the study did not differ from those who completed the study (data not shown).

During the study, the 3 groups did not differ in caloric intake, weight gain, and weight at study exit (Table 2). Phlebotomy loss per kg during the treatment period was inexplicably greater in the EPO + oral iron group compared with the other 2 groups. The mean number and volume of RBC transfusions administered from the end of the study until discharge were similar among the groups. No between-group differences were found with respect to mortality, selected morbidities, age at discharge, or weight at discharge (Table 3).

Hematologic indices of erythropoietic activity differed markedly among the groups during the 18-day treatment period. A continuous and significant decline in Hb concentration was observed in the oral iron group with the lowest value recorded on day 18 (P < .0001; Fig 1A). In contrast, Hb levels in the 2 EPO-treated groups remained unchanged during treatment. At the end of treatment, Hb concentrations were significantly higher in the 2 EPO-treated groups compared with the oral iron group (P < .01). On treatment day 18, the intravenous iron + EPO group demonstrated a trend toward higher Hb and Hct (data not shown) values relative to the EPO + oral iron group (P = .12 and .06, respectively).

View larger version (29K): [in this window] [in a new window]   Fig. 1.   Mean (± SEM) parameters of erythropoiesis by treatment received: A) Hb; B) absolute reticulocyte count; C) HFR; and D) plasma human TfR. Posthoc statistical significance compared with the oral iron group is indicated by the asterisk (*P < .05 and **P < .01) and compared with the EPO + oral iron group is indicated by the dagger (P < .05).

Although reticulocyte counts increased significantly from baseline in all 3 groups (P < .05), these values were markedly higher in the 2 EPO-treated groups compared with the oral iron group at all sampling times after the start of EPO treatment (P < .01; Fig 1B). Consistent with trends observed in Hb and Hct, reticulocyte counts on day 18 were significantly higher, ie, better sustained, in the intravenous iron + EPO group compared with the EPO + oral iron group (P = .02). When reticulocyte age was examined by flow cytometric gating, a significant increase in the percentage of reticulocytes classified as young, ie, high-fluorescence reticulocytes (HFRs), was noted for the 2 EPO-treated groups (P = .02) but not for the oral iron group (Fig 1C).²²

The patterns of change observed for plasma TfR levels in each of the 3 groups resembled those of the reticulocytes, ie, TfR levels increased markedly from baseline in both of the EPO-treated groups but remained unchanged among the infants receiving only enteral iron (Fig 1D). Similarly, TfR levels on days 11 and 18 were significantly higher in both EPO groups, compared with the oral iron group (P < .01).

RBC indices varied markedly among the 3 study groups. In the oral iron group, mean corpuscular volume (MCV) decreased slightly during treatment (Fig 2A), whereas mean red cell distribution width (RDW; Fig 2B) and the percentage of hypochromic RBCs remained stable (Fig 2C). In contrast, both EPO-treated groups experienced a slight increase in MCV and the more marked progressive increases in RDW and the percentage of hypochromic RBCs. In the 2 EPO-treated groups, all 3 RBC indices reached their highest levels by the end of treatment. On treatment days 11 and 18, mean MCV, RDW, and the percent hypochromic RBCs were all significantly higher in the 2 EPO groups, compared with the oral iron group (P < .01).

View larger version (16K): [in this window] [in a new window]   Fig. 2.   Mean (± SEM) RBC indices by treatment received during the study for: A) RBC MCV; B) RDW; and C) RBC hypochromia. Posthoc statistical significance compared with the oral iron group is indicated by the asterisk (**P < .01). There was no difference between the 2 EPO-treated groups.

At the start of treatment, iron status parameters were similar among the 3 groups. Among the oral iron group, plasma iron levels demonstrated a tendency to increase gradually during the treatment period and were significantly greater on treatment days 11 and 18, compared with the 2 EPO-treated groups (Fig 3A). Although plasma iron levels in the 2 EPO-treated groups both seemed to decrease, this fall was only significant in the EPO + oral iron group. Plasma iron values were significantly greater in the oral iron group compared with the 2 EPO-treated groups on treatment days 11 and 18. In the subgroup analysis of plasma iron in the intravenous iron + EPO group for day 0, plasma iron was significantly higher in the 4 infants sampled at the end of the 2-hour iron infusion, compared with the 6 sampled just before parenteral iron infusion. With respect to plasma transferrin concentration, the only significant change was the decrease observed in the intravenous iron + EPO group (P < .05; Fig 3B). Plasma ferritin concentration of the oral iron and EPO + oral iron groups declined significantly during treatment (P < .001), whereas ferritin levels increased significantly in the intravenous iron + EPO group (P < .002; Fig 3C). By days 11 and 18 of treatment, ferritin levels in the intravenous iron + EPO group were significantly higher than in the other 2 groups.

View larger version (17K): [in this window] [in a new window]   Fig. 3.   Mean (± SEM) parameters of iron status by treatment received during the study for: A) plasma iron; B) plasma transferrin; and C) plasma ferritin. The open squares indicate mean values for plasma iron samples drawn immediately after the completion of the 2-hour iron sucrose infusion (n = 4). Posthoc statistical significance compared with the oral iron group is indicated by the asterisks (*P < .01 and **P < .01) and compared with the EPO + oral iron group is indicated by the dagger (P < .05).

There were no differences noted among the 3 groups during treatment for any of the oxidant injury parameters. Both plasma and urine o-tyrosine and MDA levels remained unchanged throughout the study (Fig 4). The only exception to this was the higher plasma MDA values noted in the subgroup of 4 infants in the intravenous iron + EPO group sampled immediately after the completion of the iron sucrose infusion, compared with those sampled before starting the iron infusion.

View larger version (34K): [in this window] [in a new window]   Fig. 4.   Plasma and urine parameters of oxidant injury by treatment group: A) plasma MDA; B) urine MDA; C) plasma o-tyrosine; and D) urine o-tyrosine. The open squares indicate mean (± SEM) values for plasma samples drawn immediately after the completion of the 2-hour iron sucrose infusion (n = 4). Posthoc statistical significance compared with the oral iron group is indicated by the asterisk (**P < .01). There was no significant difference among the 3 groups on treatment days 4, 11, or 18.

	DISCUSSION

Top Abstract Methods Results Discussion Conclusion References

Results of the present study indicate that iron supplementation with either intravenous iron sucrose or oral iron polymaltose complex in amounts approximating the combined estimated in utero fetal iron accretion rate²³ and the rate of postnatal laboratory iron blood loss²⁴ is effective in stimulating erythropoiesis in growing VLBW infants being treated with r-HuEPO. Based on the significantly higher reticulocyte count response in the intravenous iron + EPO group compared with the EPO + oral iron group at the end of treatment, it is possible that the trend observed toward higher Hb and Hct levels in the intravenous iron + EPO group might have been more pronounced had the study been extended for a longer period. These data suggest independent effects of EPO and parenteral iron on erythropoiesis, but with EPO exerting the larger effect. Nonetheless, among rapidly growing VLBW infants with plasma ferritin levels indicative of adequate storage iron, intravenous iron offers a slight advantage over oral iron alone in the presence of EPO treatment.

A primary goal of our study was to enhance erythropoiesis in VLBW infants beyond that reported by others.³ To accomplish this we used what we considered to be highyet safedoses of both r-HuEPO and iron. The EPO dose used in the present study (300 IU/kg/day) is, to our knowledge, the largest repeated EPO dose administered to neonates.² Because r-HuEPO has been shown to be safe at lower doses, there has been a tendency to use higher doses of r-HuEPO to increase efficacy. The rationale for the dose of oral and parenteral iron we selected was to provide adequate iron for rapid growth and compensation for blood loss from laboratory testing. Based on our previous work demonstrating 30% absorption in a similar group of VLBW infants, the daily oral supplement selected, ie, 9 mg/kg, was sufficient for accomplishing this.¹⁷ The rationale for administering parenteral iron was to try to overcome functional iron insufficiency in the context of EPO treatment. In maintaining positive iron balance with parenteral iron in non-EPO-treated VLBW infants, it is estimated that an iron supplement of at least 1 mg/kg is required.²⁵ This estimate is consistent with prenatal rates of fetal iron accretion.²³

Together, the favorable erythropoietic responses observed in the present study as assessed by the increases in reticulocyte counts, the percentage of HFRs and in the levels of Hb, Hct, and TfR indicate a rapid and pronounced stimulation of erythropoiesis by EPO treatment. The small benefit of intravenous iron treatment over that of EPO and oral iron supplementation alone did not quite reach statistical significance for Hb and Hct by the end of the 18-day treatment period as it had for reticulocyte count. Although the magnitude of the rise in Hb and Hct levels observed in the present study is consistent with reports of other EPO trials,³ the present study is unique in that a requirement for continuation in the study was not receiving a RBC transfusion. In a study similar to our own, but with a 6-week period of treatment, Carnielli et al¹³ administered EPO in a dose of 1200 U/kg/week in combination with a single weekly bolus infusion of 20 mg/kg of iron sucrose and found that significantly fewer RBC transfusions were administered, compared with an untreated control groupbut not compared with a third group of infants treated with EPO and oral iron.

The increased plasma iron concentrations of the oral iron group could be explained by decreased iron needs for incorporation into the RBCs relative to the EPO-treated infants in response to oral iron supplementation. Conversely, plasma iron concentrations decreased significantly in the EPO + oral iron group and nonsignificantly in the intravenous iron + EPO group in response to the increased iron needs for the stimulated erythropoiesis. The high plasma iron levels found on day 0 in the subgroup of intravenous iron + EPO group infants immediately after stopping the intravenous iron infusion was almost certainly attributable to administering the total dose of iron over 2 hours instead of over 24 hours as occurs during fetal life. If this group had received parenteral iron at a lower infusion rate but over a more prolonged period, it is possible that iron plasma levels and, hence, plasma transferrin saturation would have increased relative to those encountered in fetal life, ie, 70% to 90%²⁶ without exceeding the plasma total iron-binding capacity.

In the present study, a marked difference was noted in the plasma ferritin concentration between the 2 EPO-treated groups, with the intravenous iron + EPO group demonstrating a significant progressive increase and the EPO + oral iron group demonstrating a significant progressive decrease. Similar differences in ferritin levels among EPO-treated VLBW infants have been reported previously.^11,13 The increase in plasma ferritin associated with parenteral iron administration in these previous studies was less marked than in our ownpossibly as a result of intravenous iron doses that were approximately one half of those used in the present study. Unfortunately, in none of the other VLBW infant EPO trials in which intravenous iron supplements were administered was there an oral iron EPO or non-EPO comparison group included for judging differences in plasma ferritin.^6,12-14,27 If one third of oral iron supplement is absorbed,¹⁷ the 3 mg of iron per kg absorbed daily should have resulted in positive iron balance in growing infants whose mean daily iron blood loss was estimated to be only .15 mg/kg. It is thus somewhat surprising in this and other studies to observe that oral iron supplementation does not prevent the normal postnatal decline that occurs in plasma ferritin.¹¹ This finding raises questions regarding significance of plasma ferritin levels as indicators of iron stores in VLBW preterm infants.

Although both MCV and RDW increased significantly in the 2 EPO-treated groups, these values remained within normal range for both term²⁸ and preterm infants treated with EPO and intravenous iron.¹³ In both the present and in these previous studies, EPO treatment has resulted in the formation of RBCs of increased size, resembling the erythrocyte macrocytosis during fetal life.²⁹ Similar changes in RBC indices described in adults with marginal iron stores receiving EPO treatment have been classified as functional iron deficiency caused by inadequate iron availability for erythropoiesis.⁹ Although plasma ferritin levels were well above those used to define classic iron deficiency, the high percentage of hypochromic RBCs in the 2 EPO-treated groups suggest a state of functional iron deficiency, ie, iron being unavailable for incorporation into RBCs.³⁰ Changes in the rate of RBC production, iron incorporation into RBCs, and altered Hb synthesis in response to EPO treatment may explain the production of macrocytic, hypochromic RBCs seen in ours and previous studies.^11,13 Insufficient folate and/or vitamin B₁₂ supplementation are alternative explanations that were not studied.³¹

In addition to iron's essential role in mitochondrial electron transfer processes, iron acts as an electron donor in the formation of free radicals capable of inducing oxidative tissue injury.¹⁵ As such, over the past decade concern has been expressed that iron supplementation or the appearance of free iron as a result of RBC transfusions could be a major contributor to a spectrum of diseases frequently encountered among VLBW infants experiencing intense oxygen exposure and/or having immature endogenous antioxidant defense systems.^32-34 It is, thus, surprising that ours is the first study to report results of biochemical markers of oxidant injury in infants treated with either oral or parenteral iron.

To exclude the possibility that the high-dose oral iron used in this study resulted in evidence of oxidant injury, blood and urine samples were analyzed before and 3 days after starting treatment with enteral iron. None was found. The only evidence of oxidant injury observed in the present study was the difference in plasma MDA levels between the 2 subgroups of the intravenous iron + EPO group. Because slight transitory increases in MDA have been observed under a variety of physiologic conditions such as exercise and dietary manipulations,^35-37 the slightly higher MDA levels observed in the 4 infants sampled immediately after stopping the 2-hour intravenous iron sucrose infusion is of questionable clinical significance. In the present, it does not seem that major increases in lipid peroxidation occurred as a result of the modest increases in plasma iron levels, which exceeded the total plasma iron binding capacity for what was likely only brief periods.

In contrast to our plasma MDA findings, no increase in plasma or urine o-tyrosine was identified. In a previous study of critically ill neonates receiving oxygen for treatment of cardiorespiratory illness, we demonstrated that nonenzymatic hydroxylation of phenylalanine resulted in an increase in plasma o-tyrosine, a specific and sensitive indicator of hydroxyl radical attack.²⁰ Because the 3 study groups in the present study experienced similar durations of oxygen and ventilator treatment, hydroxyl radical attack was most likely not associated with parenteral and/or oral iron administration.

From the standpoint of the potential role of iron as a substantial contributor to neonatal morbidities resulting from free radical injury, it is reassuring that there was no difference detected among the 3 groups in bronchopulmonary dysplasia, intraventricular hemorrhage, and retinopathy of prematurity. It is also reassuring that infection rates were not different. However, because of the small number of study participants, these results must be viewed with great caution as proof of safety.

Because the VLBW infants in the present study were 3 weeks of age and clinically stable at the time of enrollment, other less mature, more critically ill, oxygen-treated infants might have experienced different outcomes when exposed to the iron supplements used in the present study. As such, in future studies it may be prudent to prolong the parenteral iron dosing and/or to reduce the dose below 2 mg/kg/day. Results of follow-up studies of VLBW infants who received relatively large doses of parental iron over brief periods will be of importance in supporting such a recommendation.^6,11,13

	CONCLUSION

Top Abstract Methods Results Discussion Conclusion References

We found that VLBW infants receiving combined high-dose EPO and high-dose iron therapy experience a marked stimulation of erythropoiesis, compared with non-EPO-treated infants. The group of EPO-treated infants concomitantly treated with parenteral iron experienced a significantly better reticulocyte response than did the EPO group receiving oral iron. Whether the increase in ferritin associated with intravenous infusion of iron correlates with increased storage iron is uncertain but, if true, this may be one rationale for recommending parenteralover enteraliron therapy. Under the present study conditions, both enteral and parenteral iron treatment supplementation seem relatively safe in terms of clinical and biochemical indicators of oxidant injury. Nevertheless, before advocating the routine use of large doses of oral and/or parenteral iron supplements to EPO-treated VLBW infants, future studies should focus more rigorously on safety of administering iron supplementation to VLBW infants at even greater risk of oxidant injury, eg, those critically ill infants requiring high-ambient oxygen concentrations.

	ACKNOWLEDGMENTS

This study was supported by the March of Dimes Birth Defects Foundation (Grant FY 95-0220); by Vifor International, St Gallen, Switzerland; by Janssen-Cilag Pharma, Vienna, Austria; by the Verein zur Förderung der wissenschaftlichen Forschung auf dem Gebiet der Neonatologie und Kinderintensivmedizin; and by the Red Bull Company, Salzburg, Austria.

The study was greatly facilitated by the enthusiastic support of the nursing and physician staff of the Department of Neonatology, University Children's Hospital, Vienna, Austria.

	FOOTNOTES

Received for publication Aug 31, 1999; accepted Apr 13, 2000.

Reprint requests to (A.P.) Department of Neonatology, University Children's Hospital, AKH, Waehringer Guertel 18-20, A-1090, Vienna, Austria. E-mail: arnold.pollak{at}akh-wien.ac.at

	ABBREVIATIONS

VLBW, very low birth weight; RBC, red blood cell; EPO, erythropoietin; Hb, hemoglobin; r-HuEPO, recombinant human erythropoietin; Hct, hematocrit; NEC, necrotizing enterocolitis; IPC, iron polymaltose complex; MDA, malondialdehyde; TfR, transferrin receptor; SEM, standard error of the mean; HFR, high-fluorescence reticulocyte; MCV, mean corpuscular volume; RDW, red cell distribution width.

	REFERENCES

Top Abstract Methods Results Discussion Conclusion References

1. Strauss RG Recombinant erythropoietin for the anemia of prematurity: still a promise, not a panacea. J Pediatr 1997; 131:653-655 [Medline]
2. Widness JA, Strauss RG Recombinant erythropoietin in treatment of the premature newborn. Semin Perinatol 1998; 21:20-27
3. Doyle JJ The role of erythropoietin in the anemia of prematurity. Semin Perinatol 1997; 21:20-27 [CrossRef][Medline]
4. Maier RF, Obladen M, Messinger D, Wardrop CAJ Factors related to transfusion in very low birth weight infants treated with erythropoietin. Arch Dis Child 1996; 74:F182-F186
5. Meyer MP, Meyer JH, Commerford A, Recombinant human erythropoieitin in the treatment of the anemia of prematurity: results of a double-blind, placebo-controlled study. Pediatrics 1994; 93:918-923 [Abstract/Free Full Text]
6. Carnielli V, Montini G, DaRiol R, Dall'Amico R, Cantarutti F Effect of high doses of human recombinant erythropoietin on the need for blood transfusions in preterm infants. J Pediatr 1992; 121:98-102 [CrossRef][Medline]
7. Shannon K Recombinant erythropoietin in anemia of prematurity: five years later. Pediatrics 1993; 92:614-617 [Abstract/Free Full Text]
8. Shannon KM, Keith JF, Mentzer WC, Recombinant human erythropoietin stimulates erythropoiesis and reduces erythrocyte transfusions in very low birth weight preterm infants. Pediatrics 1995; 95:1-8 [Abstract/Free Full Text]
9. Macdougall IC, Hutton RD, Cavill I, Coles GA, Williams JD Poor response to treatment of renal anaemia with erythropoietin corrected by iron given intravenously. Br Med J 1989; 299:157-158
10. Sunder-Plassman G, Hörl WH Iron metabolism and iron substitution during erythropoietin therapy. Clin Invest 1994; 72:11-15
11. Meyer M, Haworth C, Meyer J, Commerford A A comparison of oral and intravenous iron supplementation in preterm infants receiving recombinant erythropoietin. J Pediatr 1996; 129:258-263 [CrossRef][Medline]
12. Ohls RK, Harcum J, Schibler KR, Christensen RD The effect of erythropoietin on the transfusion requirements of preterm infants 750 grams: a randomized, double-blind, placebo-controlled study. J Pediatr 1997; 131:661-665 [CrossRef][Medline]
13. Carnielli VP, Riol RD, Montini G Iron supplementation enhances response to high doses of recombinant human erythropoietin in preterm infants. Arch Dis Child 1998; 79:F44-F48 [CrossRef]
14. Ohls RK, Veerman MW, Christensen RD Pharmacokinetics and effectiveness of recombinant erythropoietin administered to preterm infants by continuious infusion in total parenteral nutrition solution. J Pediatr 1996; 128:518-523 [CrossRef][Medline]
15. Halliwell B, Gutteridge JMC. Protection Against Oxidants in Biological Systems. Oxford, United Kingdom: Clarendon Press; 1989:543
16. Hershko C. Iron and infection. In: Fomon SJ, Zlotkin S, eds. Nutritional Anemias. New York, NY: Raven Press; 1992:53-64
17. Widness JA, Lombard KA, Ziegler EE, Erythrocyte incorporation and gastrointestinal absorption of ⁵⁸Fe in erythropoietin treated premature infants. Pediatr Res 1997; 41:416-423 [Medline]
18. Fanaroff AA, Wright E Profiles of mean arterial blood pressure (MAP) for infants weighing 501-1500 grams. Pediatr Res 1990; 27:205A
19. Borum PR, Bennett SG A simple and inexpensive method for plasma transferrin determination. Nutr Res 1985; 5:937-940 [CrossRef]
20. Lubec G, Widness JA, Hayde M, Menzel D, Pollak A Hydroxyl radical generation in oxygen-treated infants. Pediatrics 1997; 101:700-704
21. Nyyssönen K, Poulsen HE, Hayn M, Effect of supplementation of smoking men with plain or slow release ascorbic acid on lipoprotein. Eur J Clin Nutr 1997; 51:154-163 [CrossRef][Medline]
22. Böck A, Herkner KR Reticulocyte maturity pattern analysis as a predictive marker of erythropoiesis in pediatrics: pilot study for clinical application. Clin Lab Haemat 1994; 16:343-348
23. Widdowson EM Trace elements and foetal and early postnatal development. Proc Nutr Soc 1974; 33:275-284 [CrossRef][Medline]
24. Obladen M, Sachsenweger M, Stahnke M Blood sampling in very low birth infants receiving different levels of intensive care. Eur J Pediatr. 1988; 147:399-404 [CrossRef][Medline]
25. Friel JK, Andrews WL, Hall MS, Intravenous iron administration to very-low-birth-weight newborns receiving total and partial parenteral nutrition. J Parent Ent Nutr 1995; 19:114-118
26. Okuyama T, Tawada T, Furuya H, Villee CA The role of transferrin and ferritin in the fetal-maternal placenta unit. Am J Obstet Gynecol 1985; 152:344-350 [Medline]
27. Ohls RK, Ehrenkrantz RA, Lemons JA, A multicenter randomized double-masked placebo-controlled trial of early erythropoietin and iron admistration to preterm infants. Pediatr Res 1999; 45:216A
28. Matoth Y, Zaizov R, Varsano I Postnatal changes in some red cell parameters. Acta Paediatr Scand 1971; 60:317-321 [Medline]
29. Oski FA, Naiman JL. Normal blood values in the newborn period. In: Hematologic Problems in the Newborn. Philadelphia, PA: WB Saunders Company; 1982:1-31
30. Jones CH, Richardson D, Ayers S, Newstead CH, Will EJ, Davison AM Percentage hypochromic red cells and the response to intravenous iron therapy in anemic patients. Nephrol Dial Transplant 1998; 3:229-238
31. Worthington-White DA, Behnke M, Gross S Premature infants require additional folate and vitamin B-12 to reduce the severity of the anemia of prematurity. Am J Nutr 1994; 60:930-935 [Abstract/Free Full Text]
32. Saugstad OD Oxygen radical disease in neonatology. Semin Neonatol 1998; 3:229-238
33. Sullivan JL Iron, plasma antioxidants, and the "oxygen radical disease of prematurity." Am J Dis Child 1988; 142:1341-1344 [Abstract/Free Full Text]
34. Frank L Development of the antioxidant defences in fetal life. Semin Neonatol 1998; 3:173-182
35. Leaf DA, Kleinman MT, Hamilton M, Barstow TJ The effect of exercise intensity on lipid peroxidation. Med Sci Sports Exerc 1997; 29:1036-1039 [Medline]
36. Dixon ZR, Shie FS, Warden BA, Burri BF, Neidlinger TR The effect of low carotenoid diet on malondialdehyde-thiobarbituric acid (MDA-TBA) concentrations in women: a placebo-controlled dobule-blind study. J Am Coll Nutr 1998; 17:54-58 [Abstract/Free Full Text]
37. Sastre J, Asensi M, Gascó E, Exhaustive physical excise causes oxidation of glutathione status in blood: prevention by antioxidant administration. Am J Physiol 1992; 263:R992-R995 [Abstract/Free Full Text]