Anemia of prematurity (AOP)

INTRODUCTION — 

Erythropoiesis decreases markedly after birth. This occurs due to the onset of breathing and closure of the ductus arteriosus, which increase tissue oxygenation and reduce production of erythropoietin (EPO) [1]. In healthy term infants, a mild anemia develops as hemoglobin (Hb) levels decline over the first 8 to 12 weeks after birth (referred to as the "physiologic nadir") (figure 1). (See "Approach to the child with anemia", section on 'Age of patient'.)

Preterm infants have lower Hb values at birth compared with term infants and the postnatal decline in Hb occurs earlier and is more pronounced than the physiologic anemia seen in term infants. This is compounded by other factors that contribute to more severe anemia (blood loss from phlebotomy, reduced red blood cell [RBC] lifespan, depleted iron stores). Together, these processes are referred to as anemia of prematurity (AOP).

The pathogenesis, clinical features, and management of AOP will be reviewed here, including a summary of indications for RBC transfusion in preterm neonates. A more detailed discussion of RBC transfusion in neonates is provided separately. (See "Red blood cell (RBC) transfusions in the neonate".)

DEFINITIONS OF PREMATURITY — 

Definitions of different degrees of prematurity based upon gestational age (which is calculated from the first day of the mother's last period (calculator 1)) or birth weight (BW) are provided in the table (table 1).

CAUSES AND CONTRIBUTING FACTORS — 

The primary cause of AOP is the impaired ability to increase serum erythropoietin (EPO) appropriately in the setting of anemia and decreased tissue availability of oxygen [2,3]. Circulating and bone marrow red cell progenitors respond to EPO, if present, indicating that the impaired erythropoiesis is due to lack of EPO, not a failure to respond to the hormone [4-6]. Other hematopoietic growth factors (eg, granulocyte-macrophage colony-stimulating factor) are not affected.

While AOP is directly due to impaired EPO production, several other factors can contribute to anemia in preterm infants, including blood loss from phlebotomy, reduced red blood cell (RBC) life span, and iron depletion.

Impaired erythropoietin production — EPO is produced by the fetal liver and the cortical interstitial cells of the kidney in response to hypoxia. Its production is regulated by the transcription factor hypoxia inducible factor-1 (HIF-1). Its primary function is to regulate erythrocyte production. EPO does not cross the placenta; fetal production increases with gestational age [7-10].

In preterm infants, EPO levels increase modestly in response to anemia and tissue hypoxia; however, the response is less than what is seen in older children and adults with the same degree of anemia [2,11]. (See "Regulation of erythropoiesis".)

In AOP, the specific mechanisms leading to the discrepancy between serum EPO concentration and the severity of the anemia are uncertain. Proposed pathogenetic pathways involve the site of EPO production and the developmental regulation of transcription factors in the liver versus the kidney.

●The liver is the principal site of EPO production in the fetus [12,13]. The feedback increase in hepatic EPO mRNA in response to anemia or hypoxia may be less than that of the kidney [14]. EPO mRNA expression in the kidney is present in the fetus and increases significantly after 30 weeks gestation, suggesting that the switch to the kidney as the main site of EPO production is developmentally regulated.

●The fetal or neonatal environment may alter the response to hypoxic signals by the liver. Support for this hypothesis comes from the observation that hepatic transplantation from fetal and neonatal lambs into adult sheep increased EPO production by the transplanted liver [15].

●Transcriptional regulatory factors, such as HIF-1, may contribute to low levels of EPO in preterm infants. These factors activate target genes, including those encoding EPO, in response to decreased cellular oxygen concentration [16,17]. They appear to be developmentally regulated in some fetal tissues, which might account for the decreased expression of EPO in response to anemia in preterm infants [1,18].

Blood loss from phlebotomy — Iatrogenic blood loss due to phlebotomy for blood tests is a major contributor to anemia in preterm infants. There is a strong correlation between the amount of blood lost from phlebotomy and the number of transfusions preterm infants require [19]. The volume of blood loss increases with illness severity and decreasing gestational age since critically ill, extremely preterm infants tend to require more frequent laboratory tests [20,21]. Phlebotomy-related blood loss in preterm infants may be greater than is necessary for the care of the infant [22].

In studies involving extremely low birth weight (ELBW; <1000 g) infants, total phlebotomy-related blood loss reached approximately 40 to 50 mL/kg by the end of the second week of the neonatal intensive care unit (NICU) stay and 80 mL/kg after 10 weeks [19,21]. In one study, approximately 50 percent of iatrogenic blood loss was related to sampling for blood gas analysis [19].

These studies emphasize the need for NICU policies to ensure that only the minimal volume required for testing is drawn, and unnecessary tests are avoided. Strategies to avoid iatrogenic blood loss are discussed elsewhere. (See 'Reduction of iatrogenic blood loss' below.)

Reduced red blood cell life span — Red blood cell (RBC) survival in newborn term infants is approximately 60 to 80 days, but decreases with decreasing gestational age to a range of 45 to 50 days in extremely low birth weight infants (ELBW) (BW below 1000 g) [23]. The reduced RBC life span contributes to the severity of anemia. Increased susceptibility to oxidant injury may contribute to shortened RBC survival in the infant [24,25]. (See "Diagnosis of hemolytic anemia in adults", section on 'Conceptual framework'.)

Iron depletion — Although it is not involved in its pathogenesis, iron depletion may impair recovery from AOP. Because of their rapid growth rate, preterm infants have increased utilization and depletion of iron stores and, as noted above, blood loss from phlebotomy. The administration of iron does not inhibit the fall in hemoglobin (Hb) concentration due to AOP. However, in term infants, it reduces the incidence of iron deficiency anemia in the first year of life [2]. (See "Iron deficiency in infants and children <12 years: Screening, prevention, clinical manifestations, and diagnosis".)

Low levels of other nutrients, such as vitamin B12 or folate, do not appear to contribute to neonatal anemia [1]. However, limited clinical trial data suggest that in preterm infants receiving EPO therapy, supplementation with folate and B12 may enhance erythropoiesis [26,27].

PHYSIOLOGIC CONSEQUENCES — 

The main physiologic consequence of anemia is impaired oxygen delivery. Oxygen delivery is the rate at which oxygen is transported from the lungs to the tissues. It depends upon the following (see "Oxygen delivery and consumption"):

●Cardiac output

●Hemoglobin (Hb) concentration

●Oxygen carrying capacity (affinity) of Hb

●Arterial oxygen saturation and oxygen tension

In preterm infants with anemia, compensatory physiologic changes that attempt to maintain adequate oxygen delivery include increases in heart rate and stroke volume, which improve cardiac output [28]. These are similar to compensatory changes that occur in older children and adults with anemia.

However, anemic preterm infants may be less able to maintain oxygen delivery because of the following:

●Higher levels of hemoglobin F (HbF) – HbF-containing red blood cells (RBC) in the preterm infant have a considerably higher oxygen affinity than adult RBC, resulting in reduced release of oxygen to tissues (figure 2). HbF binds poorly to 2,3 diphosphoglycerate, a potent modulator that diminishes the affinity of Hb for oxygen. Decreased binding increases oxygen affinity and shifts the oxyhemoglobin dissociation curve to the left, resulting in decreased peripheral oxygen delivery (figure 2). The proportion of HbF increases with decreasing gestational age, and is regulated developmentally so that HbF levels are similar at the same postmenstrual age [29-31]. The concentration of HbF in an infant born at 28 weeks gestation is approximately 90 percent, and decreases to approximately 60 percent at 10 weeks after birth, a value that is similar to that of an infant newly born at 38 weeks gestation [29]. (See "Structure and function of normal hemoglobins".)

●Concomitant respiratory disease – Many preterm infants have hypoxia due to respiratory disorders, such as respiratory distress syndrome and bronchopulmonary dysplasia. (See "Respiratory distress syndrome (RDS) in preterm neonates: Clinical features and diagnosis".)

●Limitations on target oxygen saturation – For preterm infants requiring respiratory support, standard practice is to avoid hyperoxia due to its harmful effects (eg, risk of bronchopulmonary dysplasia and retinopathy of prematurity). (See "Retinopathy of prematurity (ROP): Risk factors, classification, and screening", section on 'Risk factors' and "Bronchopulmonary dysplasia (BPD): Clinical features and diagnosis", section on 'Postnatal risk factors'.)

CLINICAL FEATURES

Timing of onset — AOP typically occurs at 3 to 12 weeks after birth in infants <32 weeks gestational age. The onset of AOP occurs earlier in more premature infants [1,32,33]. The anemia typically resolves by three to six months of age.

In a study of 40 very low birth weight (VLBW; <1500 g) infants, average hemoglobin (Hb) concentrations fell from 18.2 g/dL at birth to a mean nadir of 9.5 g/dL at six weeks of age [34]. Values of 7 to 8 g/dL were common even in the absence of significant phlebotomy losses. Hematocrit (HCT) values were lowest in the smallest infants, with average nadirs of 21 percent in infants with birth weights (BW) less than 1000 g, and 24 percent in infants with BW between 1000 and 1500 g.

Signs and symptoms — Many infants are asymptomatic despite having Hb values <7 g/dL [35,36]. However, other infants with AOP are symptomatic at similar or even higher Hb levels because of a reduced capacity to compensate for the degree of anemia. (See 'Physiologic consequences' above.)

Symptoms associated with AOP may include:

●Tachycardia that is otherwise unexplained

●Poor weight gain

●Increased oxygen or respiratory support requirement

●Increased episodes of apnea or bradycardia

In a prospective study of preterm infants with BW <1500 g, apneic episodes increased with decreasing HCT values and improved after red blood cell (RBC) transfusion [37].

Laboratory features — The following laboratory findings are characteristic of AOP:

●Normocytic and normochromic anemia.

●Low reticulocyte count.

●While not routinely performed, bone marrow examination shows reduced RBC precursors [2].

●While not routinely measured in clinical practice, serum erythropoietin (EPO) levels are low in preterm infants during the first postnatal month compared with normal values in older children and adults (<10 versus 15 mU/mL) and remain inappropriately low through the second postnatal month [3].

DIAGNOSIS

Monitoring infants at risk — Infants at risk for AOP should be monitored for clinical signs and symptoms, which can include tachycardia, increased oxygen requirement, and increased apneic or bradycardic episodes. (See 'Signs and symptoms' above.)

Laboratory monitoring should be tailored to clinical symptoms and risk factors.

●Hematocrit (HCT) or hemoglobin (Hb) – For extremely low birth weight (ELBW) and very low birth weight (VLBW) infants (table 1), we monitor the HCT or Hb on a weekly basis for the first weeks of life. Subsequently, it is generally not necessary to frequently monitor HCT or Hb in healthy growing preterm infants who are clinically stable; however, it is reasonable to check them every four weeks until discharge.

Infants with other clinical conditions (eg, respiratory distress syndrome, patent ductus arteriosus, sepsis) may warrant more frequent testing depending on the severity of the condition and the volume of blood tests sent for management of these conditions. Infants with persistent illness (eg, bronchopulmonary dysplasia) or surgical issues may require ongoing monitoring.

●Reticulocyte count – Measuring the reticulocyte count is not routinely necessary but may be useful to assess the state of the infant's bone marrow activity in certain situations. Examples include the following:

•If there is concern for another etiology of anemia besides AOP (eg, hemolytic anemia), measuring the reticulocyte count may help with making that distinction. (See "Overview of hemolytic anemias in children", section on 'Reticulocyte count'.)

•In patients with borderline HCT/Hb values (ie, values that are approaching thresholds for transfusion), the reticulocyte count can occasionally help guide decisions about red blood cell (RBC) transfusion, as discussed separately. (See "Red blood cell (RBC) transfusions in the neonate", section on 'Transfusion thresholds'.)

•In infants who are receiving erythropoiesis stimulating agents (ESAs), reticulocyte indices (if available) can help guide the management of iron supplementation. (See 'Dosing and additional supplementation' below.)

Although judicious laboratory monitoring is important to identify AOP in at-risk infants, it is also important to implement strategies to reduce unnecessary blood tests and prevent worsening of AOP. (See 'Blood loss from phlebotomy' above and 'Reduction of iatrogenic blood loss' below.)

Diagnostic criteria — We make the diagnosis of AOP in infants with all of the following features:

●Clinical features:

•ELBW or VLBW infant (table 1)

•Four to eight weeks old postdelivery

•Clinically stable

●Laboratory findings:

•Hb 6.5 to 9 g/dL (HCT 19.5 to 27 percent)

•If a complete blood count and/or reticulocyte count was tested:

-Normocytic/normochromic RBC

-Nonelevated reticulocyte count

-No nucleated RBC

The finding of anemia in an infant who does not meet these criteria should prompt evaluation for other causes (algorithm 1). (See "Approach to the child with anemia", section on 'Evaluation'.)

PREEMPTIVE MANAGEMENT — 

Clinicians who care for preterm infants should anticipate the development of AOP, particularly in very low birth weight (VLBW; <1500 g) and extremely low birth weight (ELBW; <1000 g) infants. Optimal nutrition (including iron supplementation) should be provided, and patients should be monitored for signs of anemia. Blood loss from phlebotomy should be minimized. Use of erythropoiesis stimulating agents (ESAs) is not standardized and varies by institution; however, they may be used in certain clinical settings.

Routine iron supplementation — All preterm infants require iron supplementation. The target for total daily intake is approximately 4 mg/kg per day; the approach depends on the feeding method:

●Breastfed infants – Breastfed infants should receive iron supplementation of 2 to 4 mg/kg per day through the first year of life. (See "Breastfeeding the preterm infant", section on 'Vitamin D and iron supplements'.)

●Formula-fed infants – All formula-fed infants should receive iron-fortified formula. The amount of additional supplemental iron required varies depending on the specific formula used, but it is typically 1 to 3 mg/kg per day.

Iron-fortified formulas have higher iron concentrations compared with breast milk, and therefore formula-fed preterm infants require less additional iron supplementation compared with exclusively breastfed infants. Low iron-containing preterm formulas should not be used, as they do not adequately provide the necessary iron required for these patients. (See "Nutritional composition of human milk and preterm formula for the premature infant".)

●Infants not tolerating enteral feeds – For such infants, we do not routinely provide parenteral iron administration; limited data suggest that it has no benefit in the prevention of AOP and may increase the risk of infection [38] (see "Parenteral nutrition in premature infants", section on 'Trace minerals'). Once they are able to tolerate enteral feeds, we initiate oral iron supplementation, as above.

Infants who are treated with ESAs require additional iron supplementation, as discussed below. (See 'Dosing and additional supplementation' below.)

The iron content at birth is lower in preterm infants than in term infants, and the iron stores of preterm infants often are depleted by two to three months of age. Although iron supplementation does not prevent AOP, it allows for greater iron substrate when erythropoiesis is stimulated [39].

Iron supplementation does not necessarily prevent development of iron-deficiency anemia later in the first year of life. Thus, all preterm and term infants should undergo routine screening for iron-deficiency anemia, as discussed separately. (See "Iron deficiency in infants and children <12 years: Screening, prevention, clinical manifestations, and diagnosis".)

Reduction of iatrogenic blood loss — Strategies to reduce iatrogenic blood loss include:

●Limiting blood sampling to essential testing only [22,40]

●Using microtechniques (sampling and testing with very small volumes of blood) [22,40]

●Using point-of care devices that can perform testing on small blood volumes

●Using noninvasive methods for monitoring when possible (eg, pulse oximetry and transcutaneous carbon dioxide monitors rather than frequent blood gas sampling)

Erythropoiesis stimulating agents (ESAs) — Use of ESAs is not standardized and varies by institution. Erythropoietin (EPO) production is impaired in AOP; therapy with ESAs, including recombinant human erythropoietin (epoetin alfa) and its longer-acting analog darbepoetin, is intended to correct that deficit.

Indications

Uncertain role for high risk neonates — The optimal approach to using ESAs to prevent or treat severe anemia in infants at high risk for AOP (eg, ELBW infants) is uncertain. At the author's institution, ESAs are not routinely used for this purpose; however, there is substantial practice variation and other centers, including those of other UpToDate contributors, use ESAs routinely in all ELBW infants and selectively in more mature infants at high risk for anemia (ie, those requiring frequent blood draws or who are undergoing surgery). (See "Red blood cell (RBC) transfusions in the neonate", section on 'Strategies to reduce RBC transfusion'.)

The main argument in favor of routine use of ESAs in ELBW and selected VLBW infants is that randomized trials show they reduce transfusion rates. Limited data suggest they may be cost-saving [41]. However, trials have failed to show that ESAs reduce blood donor exposures among transfused infants, regardless of whether the ESA is administered early (within the first week of life) or late (at or after eight days of life). Trials also suggest that ESAs do not improve neurodevelopmental outcomes. Thus, the additional benefit of ESA therapy may be marginal when other strategies are used to reduce transfusion and donor exposure, as discussed elsewhere. (See 'Approach to transfusion' below and "Red blood cell (RBC) transfusions in the neonate", section on 'Administration'.)

●Early use – In a meta-analysis of 19 trials (1750 infants), early ESA therapy (predominantly epoetin, started within the first eight days after birth) reduced the likelihood of needing a transfusion compared with no ESA therapy (52 versus 69 percent; relative risk [RR] 0.79, 95% CI 0.74-0.85) [42]. In the two trials that provided information on donor exposures among transfused infants (n = 165 infants), there was little to no difference in the number of donor exposures between infants treated with ESAs compared with placebo (mean difference 0.05, 95% CI -0.33 to 0.42). Rates of necrotizing enterocolitis and intraventricular hemorrhage were lower in ESA-treated infants compared with control. Rates of retinopathy of prematurity were higher in ESA-treated infants compared with control, but the difference was not statistically significant. Mortality rates were similar in both groups. Many of the trials included in the meta-analysis had important methodologic limitations (lack of blinding, incomplete or selective reporting of outcomes). Thus, the certainty of these findings is low.

●Late use – In a meta-analysis of 21 trials (1202 infants), late ESA therapy (predominantly epoetin, started 8 to 28 days after birth) reduced the likelihood of needing a transfusion compared with no ESA therapy (43 versus 60 percent; RR 0.72, 95% CI 0.65-0.79) [43]. However, based upon five trials (n = 197 infants), the total volume of blood transfused was similar in both groups (mean difference 1.6 mL/kg less in the ESA group, 95% CI 5.8 mL/kg less to 2.6 mL/kg more). In the two trials that provided information on donor exposures among transfused infants (n = 190 infants), the number of donor exposures in ESA-treated infants was slightly higher compared with placebo (mean difference 0.45, 95% CI 0.20-0.69). Mortality rates were similar in both groups, as were rates of necrotizing enterocolitis, intraventricular hemorrhage, sepsis, and bronchopulmonary dysplasia. Rates of retinopathy of prematurity were higher in ESA-treated infants compared with control, but the finding did not achieve statistical significance (RR 1.27, 95% CI 0.99-1.64).

●Impact on neurodevelopment – In the previously described meta-analysis of trials evaluating early ESA therapy, rates of neurodevelopmental impairment (NDI) at 18 to 22 months corrected age were lower in ESA-treated infants compared with control (13 versus 21 percent; RR 0.62, 95% CI 0.48-0.80; based on four trials [1130 infants]) [42].

However, in two subsequent large multicenter trials that were not included in the meta-analysis, there was no apparent benefit of early ESA therapy on neurodevelopmental outcomes [44-46]. In the PENUT trial, which included 941 extremely preterm infants who were randomized to epoetin or placebo starting within 24 hours after delivery, both groups had similar rates of moderate to severe NDI at age 22 to 26 months (38 percent in both arms; RR 0.97; 95% CI 0.79 to 1.18) [44]. In a separate report from the PENUT trial, both groups had similar brain magnetic resonance imaging (MRI) findings at term equivalent, which correlated with neurodevelopmental outcome as assessed by the Bayley Scales [47]. In the Swiss EPO trial, which included 448 preterm infants (gestational age between 26 weeks and 31 weeks and 6 days) randomized to epoetin or placebo starting at three hours after birth, both groups had similar mean scores on standardized developmental testing at age two years (mean Bayley MDI score 93.5 versus 94.5; difference -1.0, 95% CI -4.5 to 2.5) [45]. Neurodevelopmental outcomes at five years of age were also similar [46].

Although the use of ESAs in preterm infants appears to be safe, monitoring is warranted as adverse events have been reported in other patients receiving ESAs. (See 'Duration of therapy and monitoring' below.)

Selected patients with comorbidities — ESAs may be used in infants with the following comorbidities:

●Infants with alloimmune hemolytic disease of the newborn. This is discussed separately. (See "Alloimmune hemolytic disease of the newborn: Postnatal diagnosis and management", section on 'Erythropoiesis stimulating agents'.)

●Infants with chronic kidney disease. This is discussed separately. (See "Chronic kidney disease in children: Complications", section on 'Erythropoiesis-stimulating agents'.)

ESAs may also be useful in VLBW and ELBW infants with severe anemia whose parents/caregivers refuse blood transfusions for religious reasons. (See "Approach to the patient who declines blood transfusion", section on 'Erythropoiesis-stimulating agents (ESAs/EPO)'.)

Dosing and additional supplementation

●Dosing – When the decision is made to give an ESA, one of several agents and dosing regimens can be used (table 2) [48-51]:

•Darbepoetin alfa – We prefer darbepoetin alfa, 10 mcg/kg per dose given subcutaneously or intravenously (IV) once weekly. Darbepoetin has a longer half-life and may result in fewer injections if multiple doses per vial are obtained [51]. If darbepoetin is not available, epoetin alfa is an acceptable alternative.

•Subcutaneous epoetin alfa – Subcutaneous epoetin alfa 400 units/kg per dose can be given three times per week. For infants <1000 g, this is given as 0.2 mL/kg of the 2000 units/mL solution. As the infant grows, preparations with higher concentrations (up to 10,000 units/mL) can be used. An alternative regimen is 1200 units/kg per dose given subcutaneously once weekly [52]. The three times-a-week dose is thought to be more effective for preterm infants due to their increased volume of distribution and clearance; however, a small study found that the once-weekly dose results in similar levels of erythropoiesis [52].

•IV epoetin alfa – Epoetin alfa can also be given intravenously mixed in a protein-containing solution (eg, 5 percent albumin), although the optimal dose is uncertain. Some UpToDate experts use 200 units/kg per dose once daily as a continuous infusion with total parenteral nutrition (TPN) over 24 hours [53].

●Additional supplementation

•Iron supplementation – Infants treated with ESAs require additional iron supplementation. ELBW infants (birth weight [BW] <1000 g) receiving ESA therapy are at risk for iron deficiency even if they are receiving iron supplementation [54]. When determining the total dose of iron supplementation, all sources of iron should be included.

Our suggested approach is as follows (table 2) [50]:

-For infants tolerating enteral feeds – Iron is given enterally. We typically administer ferrous sulfate 6 mg/kg of elemental iron once daily. Some UpToDate experts start ferrous sulfate at 8 mg/kg/day divided into one to two daily doses for ELBW infants tolerating 60 to 80 mL/kg/day of enteral feeds. For iron insufficiency (based on reticulocyte hemoglobin or ferritin monitoring), the dose is advanced, as discussed elsewhere (table 2). (See 'Duration of therapy and monitoring' below.)

-For infants not yet tolerating enteral feeds (ie, those receiving TPN) – Supplementation can be given via IV with either iron sucrose or iron dextran (for the latter, a test dose is recommended prior to starting maintenance dosing). Various dosing regimens have been described, including 0.5 mg/kg per day, 1 mg/kg per day, 1.5 mg/kg twice weekly, or 3 to 6 mg/kg once weekly (table 2) [44,50,55-57]. At some centers, IV iron is added to the TPN [58].

•Other supplements – A daily multivitamin is also typically given to infants receiving ESA therapy (table 2). Some UpToDate experts administer folate (50 micrograms/day) and either an MVI or vitamin E (15 to 25 IU/day) when infants are no longer receiving TPN [50,52]. Folate is not included in MVI supplements.

Duration of therapy and monitoring — The central hematocrit (HCT) or hemoglobin (Hb) concentration is measured prior to starting ESA therapy. We remeasure these as well as the reticulocyte count (eg, ret-He) or ferritin level and absolute neutrophil count (ANC) every two to four weeks while the infant is receiving ESA therapy.

These laboratory values are used to determine the duration of ESA therapy, assess the effect of iron supplementation, and monitor for adverse effects:

●Duration of therapy – We suggest discontinuing ESA therapy once the HCT is ≥50 percent or once the infant's corrected gestational age (CGA) is ≥35 weeks, whichever is earlier. ESA therapy should be discontinued sooner if the infant experiences adverse effects related to therapy (eg, neutropenia, hypertension, thrombosis), as discussed below. For ELBW infants, ESAs can be discontinued if the Hct is >50 percent after two weeks of age; ESAs should be restarted if the Hct decreases to <30 percent and the infant is <34 to 36 weeks gestation.

ESA and iron therapy should generally not be held or discontinued following red blood cell (RBC) transfusion, except in the case of double exchange transfusion for alloimmune hemolytic anemia. (See "Alloimmune hemolytic disease of the newborn: Postnatal diagnosis and management", section on 'Exchange transfusion'.)

●Assessing effect of iron supplementation – The infant's response to supplemental iron therapy can be monitored with either of the following tests:

•Reticulocyte hemoglobin content (ret-He; also called CHr) – If available at the center, the ret-He can be used to monitor the adequacy of iron therapy during ESA treatment. Low ret-He values reflect early iron-deficient erythropoiesis [59-62]. The ret-He is checked at 14 to 28 days of age and then every two to four weeks while on treatment. The goal is to maintain the ret-He in the range of 29 to 35 pg [51,63]. Iron supplementation should be increased or decreased if the ret-He is below or above this range, respectively, as outlined in the table (table 2). Many automated hematology instruments routinely report ret-He and other reticulocyte parameters. However, the ret-He may not be available in all laboratories. (See "Automated complete blood count (CBC)", section on 'Reticulocytes'.)

•Serum ferritin – Ferritin levels can also be used to assess the adequacy of iron supplementation. If this strategy is used, ferritin levels should be monitored every two to four weeks. The target ferritin level is 50 to 300 ng/mL. Iron supplementation should be increased or decreased if the ferritin level is below or above this range, respectively, as outlined in the table (table 2). If the level is >400 ng/mL, iron therapy should be discontinued [44].

●Monitoring for adverse effects – Use of ESAs in preterm infants appears to be safe. In the available studies involving preterm infants, the most common adverse effects were:

•Neutropenia, which is generally transient and resolves with stopping the agent [64,65]. ESA should be held if the ANC falls below 1000/microL and resumed once the ANC is greater than 1000/microL.

•Iron deficiency, particularly if supplementation is inadequate [65-67].

Adverse effects that have been observed in other populations receiving ESAs (eg, patients with advanced kidney disease) include hypertension, seizures, rash, bone pain, and development of anti-EPO antibodies. These have not been reported in preterm infants. Nevertheless, ESA therapy should be discontinued if the infant experiences thrombotic complications, clinically significant hypertension, or seizures. (See "Treatment of anemia in patients on dialysis", section on 'Adverse effects of ESAs' and "Introduction to recombinant hematopoietic growth factors", section on 'Toxicity of colony-stimulating factors'.)

APPROACH TO TRANSFUSION — 

Red blood cell (RBC) transfusion is the most rapidly effective treatment for AOP and is generally warranted when the degree of anemia causes symptoms or compromises oxygen delivery. Guidelines for RBC transfusion are based upon the degree of anemia (ie, the hematocrit [HCT]/hemoglobin [Hb]) and the patient's clinical status (ie, hemodynamic stability, respiratory support requirement, degree of symptoms) (see 'Signs and symptoms' above). However, transfusion is a temporary measure, has well-established risks (eg, transfusion-transmitted infections, immune-mediated transfusion reactions, graft-versus-host disease, toxic effects of anticoagulants or preservatives), and has the disadvantage of further inhibiting erythropoiesis. (See "Red blood cell (RBC) transfusions in the neonate", section on 'Risks'.)

Restrictive transfusion strategy — For most preterm infants with AOP, we recommend using a restrictive transfusion strategy (ie, transfusing at a lower Hb level) rather than a liberal strategy (transfusing at a higher Hb level). The thresholds we use to trigger transfusion are based chiefly upon HCT/Hb levels, postnatal age, and clinical status (table 3). Additional details regarding RBC transfusion, including the selection of RBC products and guidance on administration of transfusions, are provided separately. (See "Red blood cell (RBC) transfusions in the neonate".)

The practice of using a restrictive transfusion approach in preterm infants is supported by randomized clinical trials and meta-analyses demonstrating that using restrictive transfusion thresholds reduces exposures to transfusion without increasing mortality or serious morbidity [68-74]. Most of these trials involved extremely low birth weight (ELBW) infants.

In a meta-analysis of five trials (3325 infants), patients assigned to restrictive versus liberal transfusion protocols had similar mortality rates (14 percent in both groups; relative risk [RR] 0.99, 95% CI 0.84-1.17) [68]. Other neonatal morbidities (bronchopulmonary dysplasia, sepsis, necrotizing enterocolitis, retinopathy of prematurity, intraventricular hemorrhage) were also similar in both groups, as was hospital length of stay. In the two trials (1739 patients) that assessed neurodevelopmental outcomes at 18 to 24 months, rates of neurodevelopmental impairment were similar in both groups (39 versus 36 percent; RR 1.08, 95% CI 0.88-1.33) [68,70,74].

In one of the largest trials (the ETTNO trial, which included 1013 ELBW infants), the restrictive threshold group had a lower incidence of any transfusion (60 versus 79 percent) and lower cumulative volume of transfused blood (median 19 versus 40 mL) [70]. Similarly, in another large multicenter trial (the TOP trial, which included 1824 ELBW infants), patients in the restrictive threshold group, on average, received approximately two fewer transfusions compared with patients in the liberal threshold group (mean transfusions 4.4 versus 6.2) [74].

The transfusion thresholds used in the restrictive and liberal groups varied somewhat between the different trials. Most protocols were based upon postnatal age and respiratory support requirement. For example, in the ETTNO trial, the restrictive transfusion threshold for stable neonates ≤7 days old was HCT <28 versus <35 percent in the liberal transfusion protocol. For critically ill neonates ≤7 days old, transfusion thresholds were HCT <34 and <41 percent, respectively [70].

Other strategies to reduce donor exposure — The use of satellite packs (RBC units from a single donor that are divided into multiple smaller aliquots, allowing for repeated transfusions from the same donor to the individual infant) can decrease donor exposure [75,76]. This approach is discussed in more detail elsewhere. (See "Red blood cell (RBC) transfusions in the neonate", section on 'Administration'.)

Preemptive strategies to reduce the severity of anemia, such as optimizing nutrition and minimizing iatrogenic blood loss, are intended to reduce the need for transfusion, thereby reducing donor exposure. (See 'Routine iron supplementation' above and 'Reduction of iatrogenic blood loss' above and "Approach to enteral nutrition in the premature infant".)

Erythropoesis stimulating agents (ESAs) lower transfusion requirements in high-risk infants but may not reduce the number of blood donor exposures in transfused patients. (See 'Uncertain role for high risk neonates' above.)

Options for patients who refuse transfusion — ESAs may be used in very low birth weight (VLBW) and ELBW infants whose parents/caregivers refuse blood transfusions for religious reasons. This is discussed in more detail elsewhere. (See 'Erythropoiesis stimulating agents (ESAs)' above and "Approach to the patient who declines blood transfusion", section on 'Erythropoiesis-stimulating agents (ESAs/EPO)'.)

SUMMARY AND RECOMMENDATIONS

●Etiology, pathogenesis, and pathophysiology – Newborn infants have a fall in hematocrit (HCT) soon after birth due primarily to impaired production of erythropoietin (EPO). Preterm infants have lower hemoglobin (Hb) values at birth compared with term infants, and the postnatal decline in Hb occurs earlier and is more pronounced than the physiologic anemia seen in term infants. This is compounded by other factors that contribute to more severe anemia (blood loss from phlebotomy, reduced red blood cell [RBC] lifespan, depleted iron stores). Together, these processes are referred to as anemia of prematurity (AOP). (See 'Causes and contributing factors' above.)

Preterm infants are at particular risk for impaired oxygen delivery in the setting of anemia because they often have concomitant respiratory disease and have high levels of hemoglobin F (HbF), which is less efficient in releasing oxygen to tissues. (See 'Physiologic consequences' above.)

●Clinical features

•Timing of onset – AOP typically occurs at 3 to 12 weeks after birth in infants, and it typically resolves by three to six months of age. (See 'Timing of onset' above.)

•Signs and symptoms – Many infants remain asymptomatic despite having Hb levels <7 g/dL. However, other infants may be symptomatic at similar or even higher Hb levels. Symptoms may include tachycardia, poor weight gain, increased requirement of supplemental oxygen, and/or increased frequency of apnea or bradycardia. (See 'Signs and symptoms' above.)

•Laboratory features – The laboratory findings characteristic of AOP include normocytic and normochromic RBC, low reticulocyte count, and low EPO levels. (See 'Laboratory features' above.)

●Monitoring – Infants at risk for AOP should be monitored for clinical signs of anemia. The HCT or (Hb) concentration should be monitored on a weekly basis in extremely low birth weight (ELBW) infants in the first weeks of life. It is not necessary to routinely monitor HCT/Hb in healthy growing preterm infants; however, more frequent monitoring may be needed in infants with persistent illness. (See 'Monitoring infants at risk' above.)

●Diagnosis – We make the diagnosis in clinically stable ELBW or very low birth weight (VLBW) infants (table 1) who are four to eight weeks old postdelivery and have Hb 6.5 to 9 g/dL (HCT 19.5 to 27 percent). If checked, tests will show that RBC are normocytic/normochromic, reticulocyte count is nonelevated, and there are no nucleated RBC. Anemia in an infant who does not meet these criteria should prompt evaluation for other causes (algorithm 1). (See 'Diagnostic criteria' above.)

●Preemptive management – Components of preempting clinically significant AOP include:

•Optimal nutrition – This includes universal iron supplementation. (See 'Routine iron supplementation' above and "Approach to enteral nutrition in the premature infant".)

•Limiting blood loss from phlebotomy – This includes performing only essential tests and using microtechniques. (See 'Reduction of iatrogenic blood loss' above.)

•Uncertain role of erythropoiesis stimulating agents (ESAs) for routine prophylaxis – There is considerable practice variation regarding use of ESAs for routine prophylaxis in all ELBW and select VLBW infants, and the optimal approach is uncertain. The advantages and disadvantages are discussed above. (See 'Uncertain role for high risk neonates' above.)

ESAs may be warranted for other comorbidities, such as anemia of chronic kidney disease and hemolysis. They may also be useful for those with severe anemia who decline transfusion. (See 'Selected patients with comorbidities' above.)

Details regarding dosing, supplementation, and monitoring while on ESA therapy are discussed elsewhere (table 2). (See 'Dosing and additional supplementation' above and 'Duration of therapy and monitoring' above.)

●Restrictive transfusion strategy – RBC transfusion is generally warranted when the degree of anemia causes symptoms or compromises oxygen delivery. For most preterm infants with AOP, we recommend using a restrictive transfusion strategy (ie, transfusing at a lower Hb level) rather than a liberal strategy (transfusing at a higher Hb level) (Grade 1B). The thresholds we use to trigger transfusion are based chiefly upon HCT/Hb levels, postnatal age, and clinical status (table 3). (See 'Restrictive transfusion strategy' above and "Red blood cell (RBC) transfusions in the neonate", section on 'Indications for transfusion'.)

Using satellite packs is another transfusion strategy to reduce donor exposure. (See "Red blood cell (RBC) transfusions in the neonate", section on 'Administration'.)

ACKNOWLEDGMENT — 

The editorial staff at UpToDate acknowledge Joyce M Koenig, MD, who contributed to an earlier version of this topic review.