Neonatal Erythrocyte Disorders

Fetal Erythropoiesis

Embryonic and fetal hematopoiesis is a complex system, simultaneously adapting to and supporting the ever-changing anatomy and milieu within which it develops. Fetal erythropoiesis occurs sequentially during embryonic development in three different sites: yolk sac (extra-embryonic hematopoiesis), liver, and bone marrow. The initial waves of cells originate in blood islands within the yolk sac and serve to meet the immediate needs of the growing embryo. The first types of blood cells produced are morphologically distinct, large nucleated erythrocytes which have been termed “primitive erythrocytes.” The yolk sac is also the source of the first wave of “definitive erythrocytes” or erythro-myeloid progenitors (EMPs) which ultimately seed the liver where they differentiate. Yolk sac formation of red blood cells (RBCs) is maximal between 2 and 10 weeks of gestation as the developing fetal liver does not become the predominant source of erythrocytes and other hematopoietic cells until after the first trimester. Bone marrow production of RBCs begins around week 18. Slowly, the liver’s function transitions from hematopoiesis to metabolic and by the 30th week of fetal life, bone marrow is the major erythropoietic organ. At full term gestation, almost all RBCs are produced in the bone marrow, although a low level of hepatic erythropoiesis persists through the first few days of life. Sites of fetal erythropoiesis occasionally are reactivated in older patients with hematologic disorders such as myelofibrosis, aplastic anemia, and severe hemolytic anemia.

The growth factors and cytokines that regulate embryonic hematopoiesis remain areas of controversy, and animal work suggests that they differ from those that regulate proliferation and differentiation of stem cells in later life. RBC production in extrauterine life is controlled in part by erythropoietin (EPO), a humoral erythropoietic-stimulating factor produced primarily by the kidney. The role of erythropoietin in the developing fetus has not been completely defined, but EPO is not thought to influence the earliest stages of yolk sac erythropoiesis. Within the past decade, neural crest cells have been determined to be the first site of EPO production. As production within the neural crest and neuroepithelial cells begins prior to circulation, delivery to the yolk sac does not occur until roughly day 20, with the first heartbeats. Drive of primitive erythropoiesis (prior to day 20) is therefore EPO-independent. As the sites of early hematopoiesis transition, so too do the sites of EPO production from the neuronal EPO-producing cells to the fetal hepatocytes in late embryonic stages where EPO acts in a paracrine manner, and finally a gradual transition to the fibroblasts of the kidneys (renal erythropoietin-producing cells) beginning at week 30 of gestation.

EPO is detected in fetal blood and amniotic fluid during the third trimester of pregnancy. The concentration of this hormone increases directly with the period of gestation, and thus, EPO levels in term newborns are significantly higher than in premature infants. This difference may reflect some degree of fetal hypoxia during late intrauterine life. Increased EPO titers also are seen in placental dysfunction, fetal anemia, and maternal hypoxia. Fetal RBC formation is not influenced by maternal EPO. Animal studies have demonstrated that EPO does not cross the placenta. In humans, transfusion-induced maternal polycythemia which decreases maternal EPO levels has no effect on fetal erythropoiesis. Additionally, maternal nutritional status is not a significant factor in the regulation of fetal erythropoiesis, because iron, folate, and vitamin B ₁₂ are trapped by the fetus irrespective of maternal stores, to a point. Studies have demonstrated that women with severe iron deficiency bear children with normal total body hemoglobin content.

The placenta has been the focus of much of the research in maternal-fetal iron homeostasis and indeed has revealed a tightly regulated system of placental iron transport mechanisms. In the setting of maternal iron deficiency, the expression of various placental proteins is increased and include but are not limited to divalent metal transporter 1 (DMT1), transferrin receptor 1 (TFR1), and ferroportin (FPN) in more severe iron deficiency. Despite this intricate dance, recent studies suggest that the placental acquisition of iron is prioritized over fetal endowment. This is not surprising given the well described correlation between maternal iron deficiency and adverse neonatal outcomes such as preterm birth, low birth weight, impaired immune function, and numerous neurobehavioral effects. Additionally, parameters such as cord blood hemoglobin and cord blood ferritin may not accurately reflect the neonate’s total body stores. Neonatal iron has been shown to preferentially be used for heme synthesis which maintains normal total body hemoglobin content while sacrificing brain iron endowment.

Hemoglobin, hematocrit, and RBC count increase throughout fetal life ( Table 69.1 ). Extremely large RBCs (mean corpuscular volume [MCV] of 180 fL) with an increased hemoglobin content (mean corpuscular hemoglobin [MCH] of 60 pg/cell) are produced early in fetal life. The size and hemoglobin content of these cells decrease throughout gestation, but the mean corpuscular hemoglobin concentration (MCHC) does not change significantly. Even at birth, the MCV and MCH are greater than those in older children and adults. Many nucleated RBCs and reticulocytes are present early in gestation, and the percentage of these cells also decreases as the fetus ages.

Hemoglobin production increases markedly during the last trimester of pregnancy. The actual hemoglobin concentration increases, but, more important, body weight, blood volume, and total body hemoglobin triple during this period. Fetal iron accumulation parallels the increase in total body hemoglobin content. The neonatal iron endowment at birth, therefore, is directly related to total body hemoglobin content and length of gestation. Term infants have more iron than premature infants.

Red Blood Cell Physiology at Birth

In utero, the PO ₂ in blood delivered to the tissues is only one third to one fourth the value in adults. This relative hypoxia may be responsible for the increased content of erythropoietin and signs of active erythropoiesis (nucleated RBCs, increased reticulocytes) seen in newborns at birth. When lungs become the source of oxygen, hemoglobin-oxygen saturation increases to 95% and erythropoiesis decreases. Within 72 hours after birth, erythropoietin is undetectable, nucleated RBCs disappear, and by 7 days, reticulocytes decrease to less than 1%.

The concentration of hemoglobin during the first few hours of life increases to values greater than those in cord blood. This is both a relative increase caused by a reduction in plasma volume and an absolute increase caused by placental blood transfusion. The umbilical vein remains patent long after umbilical arteries have constricted, and thus transfusion of placental blood occurs when newborns are placed at a level below the placenta. The placenta contains approximately 100 mL of fetal blood (30% of the infant’s blood volume). Approximately 25% of placental blood enters the newborn within 15 seconds of birth, and by one minute, 50% is transfused. The time of cord clamping is thus a direct determinant of neonatal blood volume. The blood volume in term infants (mean of 85 mL/kg) varies considerably (50 to 100 mL/kg) because of different degrees of placental transfusion. These differences are readily apparent when the effects of early versus delayed cord clamping are compared at 72 hours of age: 82.3 mL/kg (early clamping) versus 92.6 mL/kg (delayed clamping). These changes are largely the result of differences in RBC mass, 31 mL/kg (early clamping) versus 49 mL/kg (delayed clamping). The blood volume in premature infants (89 to 105 mL/kg) is slightly greater than that in term infants, but this difference is due in large part to an increased plasma volume. The RBC mass in premature infants, expressed in milliliters per kilogram, is the same as in term newborns.

Fetal and Neonatal Hemoglobin Function

A variety of hemoglobins are present during fetal and neonatal life. Fetal hemoglobin (hemoglobin F) is the major hemoglobin in utero, whereas hemoglobin A is the normal hemoglobin of extrauterine life. A single RBC may contain both hemoglobin F and hemoglobin A in varying proportions, depending on gestational and postnatal age. One major difference between hemoglobins A and F is related to oxygen transport.

The transport of oxygen to peripheral tissues is regulated by several factors, including blood oxygen capacity, cardiac output, and hemoglobin-oxygen affinity. Oxygen capacity is a direct function of hemoglobin concentration (1 g hemoglobin combines with 1.34 mL oxygen). Compensatory changes in cardiac output can maintain normal oxygen delivery under conditions in which oxygen capacity is significantly reduced. The oxygen affinity of hemoglobin also influences oxygen delivery to tissues. Hemoglobin A is 95% saturated at an arterial PO ₂ of 100 mm Hg, but this decreases to 70% to 75% saturation at a venous PO ₂ of 4 mm Hg. The difference in O ₂ content at arterial and venous oxygen tensions reflects the amount of oxygen that can be released. Changes in hemoglobin affinity for oxygen can influence oxygen delivery ( Fig. 69.1 ). At any given PO ₂ , more oxygen is bound to hemoglobin when oxygen affinity is increased. Stated in physiologic terms, increased hemoglobin-oxygen affinity reduces oxygen delivery, whereas decreased hemoglobin-oxygen affinity increases oxygen release to peripheral tissues.

The oxygen affinity of hemoglobin A in solution is greater than that of hemoglobin F. Paradoxically, however, whole blood from normal children (hemoglobin A) has a lower oxygen affinity than that of neonatal blood (hemoglobin F). This difference is related to an intermediate of RBC metabolism, 2,3-diphosphoglycerate (2,3-DPG). This organic phosphate compound interacts with hemoglobin A to decrease its affinity for oxygen, thereby enhancing O ₂ release. Fetal hemoglobin does not interact with 2,3-DPG to any significant extent ; consequently, cells containing hemoglobin F have a higher oxygen affinity than those containing hemoglobin A. The increased oxygen affinity of fetal RBCs is advantageous for extracting oxygen from maternal blood within the placenta.

A few months after birth, infant blood acquires the same oxygen affinity as that of older children ( Fig. 69.2 ). The postnatal decrease in oxygen affinity is due to a reduction in hemoglobin F and an increase in hemoglobin A (which interacts with 2,3-DPG). Oxygen delivery (the difference in arterial and venous O ₂ content) increases while oxygen capacity (hemoglobin concentration) decreases during the first week of life ( Fig. 69.3 ). This enhanced delivery is largely a reflection of the decreased oxygen affinity of infant blood. The oxygen affinity of blood from premature infants is higher than that of term infants, and the normal postnatal changes (decrease in oxygen affinity, increase in oxygen delivery) occur much more gradually in premature infants (see Fig. 69.3 ).

Another role of the fetal erythrocyte that has only relatively recently been appreciated is maternal immune tolerance. In a mouse model, the immunosuppressive role of CD71+ erythroid cells both on the maternal side as well as the fetal side has been demonstrated. In pregnant mice as well as pregnant women, a higher proportion of peripheral CD71+ erythroids was found compared to their non-pregnant counterparts. This upregulation was also present locally, at the fetomaternal interface (placental tissue). Through regulation of L-arginine levels and site-specific expression of PDL-1, the work of Delyea et al. suggests that fetal CD71+ erythroid cells play an integral role in the immune tolerance which is essential for a successful allogeneic pregnancy.

Evaluation of Anemia

A simple classification of neonatal anemia based on physical examination and basic laboratory tests is presented in Table 69.2 . More extensive RBC testing is discussed elsewhere.

Red Blood Cell Count, Hemoglobin, Hematocrit, and Red Blood Cell Indices

RBC values during the neonatal period are more variable than at any other time of life. The diagnosis of anemia must therefore be made in terms of “normal” values for gestational and postnatal ages. The mean cord blood hemoglobin of healthy term infants ranges between 14 and 20 g/dL ( Table 69.3 ). Shortly after birth, however, hemoglobin concentration increases. This increase is both relative (owing to a reduction of plasma volume) and absolute (owing to placental RBC transfusion). Failure of hemoglobin to increase during the first few hours of life may be the initial sign of hemorrhagic anemia. RBC values at the end of the first week are virtually identical to those seen at birth. Anemia during the first week of life is thus defined as any hemoglobin value less than 14 g/dL. A significant hemoglobin decrease during this time is suggestive of hemorrhage or hemolysis. For example, a hemoglobin of 14.5 g/dL at 7 days of age is abnormal for a term infant whose hemoglobin was 18.5 g/dL at birth. A slight hemoglobin reduction normally occurs in premature infants during the first week of life. Beyond the first week, however, the hemoglobin concentration decreases in both term and premature infants (see Physiologic Anemia of Infancy and Anemia Prematurity, later).

Table 69.3

Red Blood Cell Values in Term and Premature Infants During the First Week of Life

	Hb (g/100 mL)	Hct (%)	Reticulocytes (%)	Nucleated RBCs (Cells/1000 RBCs)
Term
Cord blood	17.0 (14–20)	53.0 (45–61)	<7	<1.00
Day 1	18.4	58.0	<7	<0.40
Day 3	17.8	55.0	<3	<0.01
Day 7	17.0	54.0	<1	0
Premature (Birthweight <1500 g)
Cord blood	16.0 (13.0–18.5)	49	<10	<3.00
Day 7	14.8	45	<3	<0.01

Hb , Hemoglobin; Hct , hematocrit; RBC , red blood cell.

The electronic equipment used for blood counts also gives statistical information regarding erythrocyte size (MCV) and hemoglobin content (MCH). The normal MCV in older children ranges from 75 to 90 fL. MCV values of less than 75 fL are considered microcytic, whereas those over 100 fL indicate macrocytosis. Normal infant RBCs are large (MCV 105 to 125 fL), and not until 8 to 10 weeks of age does cell size approach that in older children. Neonatal microcytosis is defined as an MCV of less than 95 fL at birth. The RBC hemoglobin content of neonatal cells (MCH 35 to 38 pg/cell) is greater than that seen in older children (MCH 30 to 33 pg/cell). Neonatal hypochromia is defined as an MCH of less than 34 pg/cell. Hypochromia and microcytosis generally occur together and are due to hemoglobin production defects. Neonatal hypochromic microcytosis is seen with iron deficiency and thalassemia disorders (alpha and gamma thalassemias). Both the MCV and MCH are higher in preterm infants as shown in Fig. 69.4 .

The site from which blood is obtained is important, because hemoglobin and hematocrit are higher in capillary blood than in simultaneously obtained central venous samples (up to 20%). This difference can be minimized by warming an extremity to obtain “arterialized capillary blood.” In the face of acute hemorrhage, however, central venous samples must be obtained because of marked peripheral vasoconstriction.

Management of Anemia

A hemoglobin of 14 g/dL corresponds to an RBC mass of 31 mL/kg. Thus, an RBC transfusion of 2 mL/kg will increase the hemoglobin concentration by approximately 1 g/dL. Packed RBCs (hematocrit approximately 67%) contain 2 mL of RBCs/3 mL of packed RBCs. Thus, the transfusion of 3 mL of packed RBCs/kg increases hemoglobin concentration by approximately 1 g/100 mL.

Packed RBCs are the product of choice when transfusion is necessary for simple anemia, as occurs in hemolysis. If anemia is accompanied by hypovolemia from acute blood loss, volume expansion must be achieved promptly, using packed RBCs and normal saline or a colloid such as 5% serum albumin (infused separately). The previously common practice of reconstituting RBCs with fresh frozen plasma to make “whole blood” is no longer acceptable because the increased donor exposure increases the risk of transmitting infectious disease. When packed RBCs need to be diluted to facilitate nonurgent transfusion, isotonic saline is the preferred diluent. Although fresh blood less than 2 days old is ideal because there is a reduced risk of hyperkalemia, this is not usually available. An acceptable substitute is packed RBCs less than 4 to 5 days old. These packed RBCs provide adequate oxygen delivery; hyperkalemia can be prevented by washing the RBCs once in saline and then reconstituting with normal saline. Washing is not required for the usually small, simple transfusions of packed RBCs, because the small volume of plasma minimizes any toxic effect of increased concentration of potassium in the plasma.

Blood currently available in most blood banks is anticoagulated with citrate-phosphate-dextrose (CPD), CPD-adenine (CPDA-1), or adenine-saline (AS-3), with a shelf life of 21, 35, or 42 days, respectively. Hematocrit usually ranges between 65% and 80% for packed RBCs. Near-normal 2,3-DPG levels are maintained for up to 12 to 14 days, which is advantageous in transfusing infants with acute hypoxia or those receiving large volumes of blood. Hematocrits range from 55% to 65%, thus facilitating flow during infusion. The newest of these preparations, AS-3, is well tolerated by newborns even after up to 42 days of storage, so long as only small-volume transfusions are given at any one time. However, when larger-volume (≥20 mL/kg) transfusions are required, the theoretical concern is for fetal/neonatal renal and hepatic exposure to potentially toxic levels of the additives within these products such as adenine, mannitol, sodium chloride, dextrose, citrate, and phosphate. To date, there have been no trials assessing the safety of these additive solutions when transfusion needs increase, and many institutions utilize them in certain situations. A survey conducted in 2015 of 21 facilities across the US found that for large-volume transfusions in neonatal patients, 43% of responding centers used AS-3 RBC units, 29% used AS-1 RBCs, and 28% used CPD or CPDA RBC units. Risks were mitigated using fresh units or re-washing previously radiated units so as to decrease the chance of hyperkalemia.

It is important to adopt practices that limit donor exposure in order to reduce the risk of transfusion-associated infections. Splitting and aliquoting a single red cell donation for multiple use by one neonate, using a unit through its outdate, the use of restrictive transfusion thresholds, and larger transfusion volumes are all effective ways to limit exposures.

Preterm infants born weighing less than 1250 g are uniquely susceptible to potentially serious cytomegalovirus (CMV) infection from transfused blood, particularly if they lack immunity because their mothers are seronegative. Practices to prevent CMV infection have involved utilizing blood products from seronegative donors but because approximately 40% to 60% of adults are seropositive, there is limited availability of seronegative donors. Alternatively, because CMV resides mainly in leukocytes, removal of such cells also can prevent transmission of the virus, and the use of high-efficiency leukocyte depletion filters has proven effective. A potential disadvantage of using CMV-seronegative blood in CMV-positive infants receiving large amounts of blood is dilution of infant’s antibody level, resulting in increased susceptibility to nursery-acquired CMV infection. Currently, a majority of neonatal services utilize leukocyte-reduced red cell products rather than relying on CMV-negative products to prevent CMV infection. Although the efficacy of using leukoreduced and CMV-seronegative products was demonstrated in a recent prospective cohort study of 539 infants, there remains quite a bit of controversy in terms of which products should be utilized for which patient populations. This ambiguity was highlighted when in 2016, the American Association of Blood Banks Clinical Transfusion Medicine Committee decided not to issue guidelines regarding the use of such products to reduce the risk of transfusion-transmitted CMV. This controversy plays out in practice as was depicted in a US practice survey which was conducted in 2015 and published in 2018.

Graft-versus-host (GVH) reaction rarely follows transfusion and occurs mainly in certain newborns at risk. For this to occur, viable lymphocytes in cellular blood products must be able to engraft and react against foreign antigens on tissues of the recipient. Infants at risk include those with congenital or acquired defects of cellular immunity, those who as fetuses received intrauterine transfusion of RBCs or platelets, newborns receiving exchange transfusion following intrauterine transfusion, and infants receiving directed blood donations from first-degree relatives (whose genetic similarity may increase the likelihood of engraftment). Irradiation of RBCs and platelets with a minimum of 1500 rads has proved effective in preventing GVH reaction. Reports of GVH reaction after RBC transfusion in very premature infants without known risk factors have prompted most neonatal services to irradiate all RBC blood products.

Causes of Fetal and Neonatal Anemia