Robert D. Christensen, MD

Normal Erythrocyte Values of Neonates

Expected values, also called “reference ranges,” for Hgb and hematocrit on the day of birth are a function of gestational age, increasing gradually through the second and third trimesters. Studies with very large sample sizes of neonates on the day of birth reveal no differences in Hgb or hematocrit associated with the infants’ sex. Reference ranges for blood Hgb concentrations are shown as Figure 12-1. The fifth percentile value (the lowest expected limit) at term is 14 g/dL. Thus the value of 11.8 g/dL in this patient is low. ¹

The 95th percentile reference range at term is 22.5 g/dL (see Figure 12-1). The value of 24 g/dL in this patient is therefore high.

Yes. Values from capillary beds (heel stick) are generally about 15% higher and also more variable than venous or arterial values. This observation is due in part to the changing peripheral perfusion in the hours after birth. Poorly perfused heels seem to have capillary pooling or sludging of red blood cells (RBCs) that result in higher values. This phenomenon of higher Hgb level in capillary as opposed to central blood sources is not so significant in older children and adults. If the Hgb value of 24 g/dL in the patient in Question 2 was drawn from a heel stick, you would want to repeat it using a central vascular determination before you decide whether the neonate is truly polycythemic.

The MCV and the MCH both diminish gradually through the second and third trimesters, as shown in Figure 12-2. The MCHC, however, does not change during this period but remains in the range of 31 to 34 g/dL. MCHC values greater than 36 g/dL should alert you to the possibility of hereditary spherocytosis or pyropoikilocytosis, two conditions that generally present with a low MCV and a high MCHC. They commonly also demonstrate hyperbilirubinemia and sometimes a diminishing Hgb concentration. If the MCV is consistently greater than 36 g/dL you should assess the morphology of the RBCs, and you will need to ask whether other family members have abnormally shaped RBCs and have had anemia, neonatal jaundice, or early cholelethiasis (bilirubin stones).

NRBCs can be reported as the number of NRBCs per 100 WBCs or as NRBCs per μL. The latter provides a more accurate accounting because of changing WBC counts over the first several days. The former is more commonly used in clinical practice. Reference ranges for NRBCs per 100 WBCs are shown in Figure 12-3. For term infants on the day of birth, the 95th percentile (highest expected limit) is 15 per 100 WBCs. Therefore the value of 100 in this patient is abnormally high.

A value of zero (0) NRBCs per 100 WBCs is always within the reference range, regardless of gestational age. Thus the value of 0 in this patient is within the expected (normal) range.

Hgb is a tetramer of globin chains, usually of two distinct types, bound to a heme moiety. Adult hemoglobin (Hgb A) consists of two alpha chains and two beta chains, whereas fetal hemoglobin (Hgb F) consists of two alpha chains and two gamma chains ( Figure 12-4).

Embryonic Hgbs are present in the first 8 weeks after conception and consist of Hgb Gower 1 (zeta 2, epsilon 2), Hgb Gower 2 (alpha 2, epsilon 2), and Hgb Portland (zeta 2, gamma 2). Hgb Barts consists of four gamma chains and occurs in the absence or deficiency of alpha chains. Thus 10% of the Hgb observed as Barts is abnormal and suggests a deficiency of at least one, and probably two, of the four alpha chain genes. Deletion of one alpha gene results in a phenotypically normal individual, and deletion of two can result in mild microcytic anemia with the presence of Barts Hgb during the fetal and early newborn period. Deletion of three genes gives rise to Hgb H disease; deletion of all four gives rise to a lethal syndrome of hydrops fetalis with all or most of the Hgb being Barts, and no Hgb A or F or alpha 2 (because there are no alpha chains). ²

Hgb F binds oxygen more avidly. The oxyhemoglobin dissociation curve for Hgb F is shifted to the left of the adult curve. The higher affinity for oxygen facilitates transfer of oxygen from maternal Hgb A but also results in decreased oxygen release to the fetal tissues. The latter situation is not a disadvantage, however, because fetal tissues use oxygen primarily for growth; metabolic functions are mostly handled by the mother.

Factors that shift the oxyhemoglobin dissociation curve to the right (increasing oxygen delivery to tissues) include increased temperature, increased partial pressure of carbon dioxide in the blood (PaCO₂), increased RBC 2,3-diphosphoglycerate content, and decreased pH.

Anemia in the Fetus and Newborn Infant

None of the four is normal. The bilirubin and the MCHC are high, and the Hgb and the MCV are low.

All the preceding steps may be helpful in the diagnosis and management of this case. Consider the following:

The initially stable Hgb level does not exclude the diagnosis of a subgaleal hemorrhage. Hemorrhage itself does not lower the Hgb and hematocrit. The fall occurs only when extravascular fluid moves into the vascular space as a physiologic response to hypovolemia.

Although much less common than a caput, a subgaleal hemorrhage can be life-threatening and therefore demands aggressive monitoring and support. Head wrapping has been attempted in the past as a potential method for tamponade, but in general this approach has not been successful because it tends to increase the intracranial pressure.

With signs of hypovolemia, transfusion support may be warranted. Coagulopathy can occur, generally secondary to shock and disseminated intravascular coagulopathy (DIC), and can worsen the hemorrhage. Portable cranial ultrasound will generally confirm the presence of a subgaleal hemorrhage. Computed tomography or magnetic resonance imaging will provide more accurate and detailed information, but these are usually not needed to make the diagnosis of a subgaleal hemorrhage.

False. Most neonatal subgaleal hemorrhages do indeed follow vacuum extraction, but some follow forceps delivery and some occur with nonoperative delivery. However, in a series from Intermountain Healthcare, we found that every neonate with a subgaleal hemorrhage that required one or more RBC transfusions was delivered by either vacuum or forceps extraction.

All neonates who had a “spontaneous” subgaleal hemorrhage (not delivered by vacuum or forceps extraction) lacked signs of shock, had no transfusions, and generally had a good outcome. Thus vacuum delivery is the most significant risk factor for developing a neonatal subgaleal hemorrhage. ⁴

True. In a recent report from Taiwan, one in 218 vacuum deliveries developed a subgaleal hemorrhage. In a study from Intermountain Healthcare, a subgaleal hemorrhage was diagnosed in one in 598 vacuum deliveries. A subgaleal hemorrhage is therefore rare, even after a vacuum delivery, but because of the vigilance needed for proper diagnosis and management, the possibility of a subgaleal hemorrhage should be considered after any operative delivery in which scalp fluctuance is observed.

False. Some publications describing cases from the 1980s and earlier did indeed report a mortality rate this high, but more recent series suggest the mortality rate is 5% to 10%. Regardless, this injury is devastating. Vigilance and aggressive management are likely responsible for the observed improvement in outcome.

Neonatal hemolytic jaundice is typical, particularly manifesting in the following ways:

HDFN resulting from maternal anti-Kell antibody presents with fetal or neonatal anemia but no reticulocytosis, no evidence of hemolysis, and no hyperbilirubinemia. These cases present with hyporegenerative anemia. The explanation is that the Kell antigen is expressed on erythroid progenitor cells, whereas most other blood group antigens are not expressed until the cells clonally mature. As a consequence of maternal anti-Kell antibody binding to fetal erythroid progenitors, fetal RBC production is reduced and hyporegenerative anemia results. The KEL gene (7q33) encodes a 93 kilodalton transmembrane zinc-dependent endopeptidase that is responsible for cleaving endothelin-3. The Kell protein has recently been designated CD238.

True. Women lacking the A and the B erythrocyte antigens often have anti-A and anti-B antibodies even before pregnancy. In the case of women with blood type O, their anti-A and anti-B antibodies are sometimes of the immunoglobulin G type and therefore can cross the placenta and bind to fetal antigens. Unlike the situation observed with maternal anti-D, in neonates with ABO hemolytic disease the principal problem is usually jaundice. Anemia, erythroblastosis, and hydrops are all very rare. The ABO locus is on chromosome 9 and has three main allelic forms: A, B, and O. The A and B alleles encode glycosyltransferases. The O allele differs from the A allele by deletion of only one nucleotide—guanine at position 261. This deletion causes a frame shift and results in premature termination of translation of the mRNA. ⁶

The H antigen is the precursor to the ABO blood group antigens. The H locus is on chromosome 19 and encodes the H antigen, which is expressed on the RBC surface. The H antigen is then modified by the A or the B antigen to produce the final A, B, or O antigen. Very rarely an individual lacks the H antigen because of a mutation in the H gene. This results in type O blood, but because the precursor molecule (the H antigen) is also missing, even type O blood cannot be transfused because the individual recognizes the H antigen in the type O blood as foreign. This unusual O blood type is called Bombay blood group and occurs in approximately four per million people, except in parts of India where it may be as common as 1 in 10,000. Neonates who are type O on the basis of Bombay can hemolyze if transfused with type O blood.

The donor (anemic, smaller) twin is more likely to have a hyporegenerative neutropenia, similar to that seen in neonates born after pregnancy-induced hypertension. This situation is likely to present with no left shift (a normal immature-to-total neutrophil ratio) and a duration of only about 2 or 3 days. Similarly, the donor twin is more likely to have a moderately low platelet count with a normal mean platelet volume (MPV). The pathogenesis of these findings is not known with certainty but likely relates to the accelerated erythropoietic effort in the anemic twin, with a concomitant temporary reduction in platelet and neutrophil production. Also, the anemic twin will usually have a higher NRBC count, generally above the reference range for age (see Figure 12-3). ⁸

Typical transfusion reactions (i.e., those commonly reported for adult recipients) are only rarely observed in transfused neonates. These include febrile nonhemolytic reactions, urticarial (allergic) reactions, hypothermia, circulatory overload, hypotensive reactions, citrate toxicity (e.g., peripheral paresthesia, tingling, buzzing, cramps, nausea, vomiting), and acute hemolysis resulting from undetected incompatibility. Transfusion-transmitted diseases include bacterial contamination, which is considerably more common than the hepatitis and other viruses transmitted in past decades, before development and implementation of modern hemovigilance techniques and procedures.

Adverse associations with transfusions that are unique to neonates include transfusion-associated necrotizing enterocolitis (generally very-low-birth-weight neonates 3 to 4 weeks old receiving a “late” transfusion) and severe intraventricular hemorrhage in extremely-low-birth-weight neonates after an “early” transfusion. Transfusion-related acute lung injury (TRALI) reactions involve acute onset of (or acute worsening of) respiratory distress after transfusion. All plasma-containing blood products have been implicated in TRALI reactions. TRALI is now among the three leading causes of transfusion-related fatalities, along with ABO incompatibility and bacterial contamination, but it is rarely reported (perhaps because it is rarely recognized) in neonatal transfusion recipients. ⁹¹⁰

Several differences can be anticipated. Awareness of these differences prevents confusion when you compare the CBCs before and after the exchange.

Polycythemia: Diagnosis and Management

Technically, these two words describe different aspects of illness, but in neonates the two tend to occur together. In older children and adults marked increases in blood concentrations of leukocytes and serum proteins can result in hyperviscous blood, even with a normal hematocrit level, and they can therefore have hyperviscosity without polycythemia. Neonates with leukocyte concentrations up to and over 100,000/μL, however, have been reported to have normal blood viscosity measurements. In neonates hyperviscosity is nearly always secondary to polycythemia, and polycythemia (particularly a “central” hematocrit exceeding 70%) essentially always indicates hyperviscosity. ¹¹

A hematocrit (or blood Hgb concentration) exceeding the 95th percentile limit (see Fig. 12-1) on the first day of life is, by definition, abnormal. However, not all neonates with a hematocrit above the 95th percentile need a reduction transfusion. If they did, 5% of all neonates would be subjected to a reduction transfusion. A general recommendation is that if the central (noncapillary) value exceeds 70% and the neonate has physiologic disturbances consistent with hyperviscosity, a reduction transfusion is warranted. Those disturbances include tachypnea, tachycardia, plethora, hypoglycemia, and tremulousness. An additional general recommendation is that if the central hematocrit exceeds 75%, a reduction transfusion may be warranted even if the neonate is asymptomatic. Ideally, avoiding the signs associated with hyperviscosity by reducing the hematocrit before intravascular problems result is preferable.

An isovolemic reduction is better tolerated than simply withdrawing blood. In hyperviscocity syndrome withdrawal of blood alone may produce increased intravascular sludging and increase the risk of symptoms. To accomplish this process, the clinician can simultaneously administer sterile saline while removing an equal volume of blood. This exchange can be done using two separate sites (pushing through one intravenous line and pulling through the other) or through one site, such as an umbilical venous catheter, pushing and then pulling in increments not exceeding 5 mL/kg body weight in each cycle. The procedure should be set up in a sterile manner, using continuous heart rate, respiratory rate, and pulse oximetry monitoring. In general, the total volume of saline infused will equal exactly the total volume of blood removed. This volume can be calculated as follows.

Multiply the estimated blood volume (about 80 to 90 mL/kg body weight) by the observed hematocrit minus the desired hematocrit (aim for 60%) divided by the observed hematocrit. In general, you will be performing a reduction on a term neonate (3 kg) with a hematocrit of 75%, so the equation will be calculated as follows:

This type of reduction transfusion has been documented to reverse the clinical symptoms of neonatal hyperviscosity, but it is not clear whether any long-term improvements occur as a result.

Normal Platelet Values of Mothers and Neonates

Reference ranges for platelet counts during pregnancy are shown in Figure 12-9. The 5th percentile value (the lowest expected limit) at term is 100,000/μL; thus a value of 135,000/μL is within the normal range. As the figure shows, the reference range diminishes steadily throughout gestation, generally falling by an increment of about 50,000/μL from early pregnancy to term. This fall is at least partly caused by the normal hemodilution of pregnancy. In contrast, no change in MPV normally occurs during pregnancy. Neonates born to women who have platelet counts in the range of 100,000 to 150,000/μL do not have an increased risk of neonatal thrombocytopenia. Therefore if the neonate appears healthy, it is not necessary to obtain a platelet count on the basis of the mother’s count of 135,000/μL. ¹²

Reference ranges for platelet counts on the day of birth, according to gestational age, are shown in Figure 12-10. The 5th percentile (lower) expected range at 30 weeks’ gestation is about 110,000/μL, and thus the observed count of 129,000/μL is normal.

Reference ranges for platelet counts during the first 90 days after birth are shown in Figure 12-11. The 95th percentile (upper limit) value at 3 weeks is about 650,000/μL, and the value of 595,000/μL is therefore within the expected reference range. The reference range for platelets increases after birth, reaching a first peak at 14 to 20 days. This change is most likely due to the physiologic thrombopoietin surge that occurs at birth. This surge—and the subsequent increase in platelet count—occurs after either vaginal or cesarean delivery and in preterm as well as term neonates. A comparable increase in MPV also occurs during the first 2 to 3 weeks, consistent with increased platelet production. The cause of the second peak in platelet count at 40 to 50 days (see Figure 12-11) is not known.

Thrombocytopenia and Platelet Transfusion

All the preceding steps can be appropriate depending on the circumstances.

Repeating an abnormal platelet count is generally wise because artifactually low platelet counts can occur with platelet clumping, as sometimes happens in a difficult and slowly oozing capillary draw, where platelets can aggregate at the wound edge and render the platelet count artifactually low. However, the presence of petechiae in this case strongly suggests that the platelet count is indeed pathologically low; do not let repeating the count delay your other orders.

In a neonate with severe congenital thrombocytopenia, an estimate of the platelet size can be an important diagnostic aid. Very small platelets in a male can suggest Wiscott–Aldrich syndrome (OMIM #301000, Xp11.23) or X-linked thrombocytopenia (OMIM #313900, Xp11.22). Very large platelets can suggest accelerated platelet destruction (as with neonatal alloimmune thrombocytopenia), a familial macrothrombocytopenia such as Bernard–Soulier syndrome (OMIM # 231200), or the MHY9-related disorders.

Ordering a platelet transfusion for this patient would be in keeping with usual practice in the United States. In a recent survey more than 90% of neonatologists in the United States and Canada would order a platelet transfusion for a neonate with a platelet count below 10,000/μL on the day of birth, even if no clinical bleeding manifestations (other than petechiae) were identified. Measuring the platelet count after (within 30 minutes) completing the platelet transfusion can help you determine whether the thrombocytopenia is the result of reduced platelet production (adequate rise with transfusion) or accelerated destruction (poor rise with transfusion).

A low platelet count in the mother could be an important diagnostic finding, suggesting an immunologic basis for the low platelets in both mother and neonate. Idiopathic thrombocytopenic purpura, systemic lupus erythematosis, or any maternal autoimmune disorder could be responsible, but the platelet count of 9000/μL in this patient is very low for maternal autoimmune thrombocytopenia.

Most neonates whose mothers have autoimmune thrombocytopenia do not have platelet counts below 30,000 or 40,000/μL. Also, a variety of familial thrombocytopenias might be found in the mother, and the platelet size in these cases can be either normal or large. Familial thrombocytopenia with normal-size platelets include Paris–Trousseau syndrome (11q23 deletion), RUNX1 mutation, and the ANKRD26 mutation.

Thrombocytopenia is part of many syndromes. For instance, neonates with trisomy 13 and 18 are almost always thrombocytopenic. However, the platelet counts in most syndromes with dysmorphia rarely have platelet counts this low, generally not below 40,000/μL. Severe thrombocytopenia, as observed in this case, can be seen with TAR syndrome (thrombocytopenia and absent radii), but the patient must have short forearms with normal-looking thumbs for TAR to be a consideration. Another syndrome associated with severe thrombocytopenia is ATRUS (amegakaryocytic thrombocytopenia with radioulnar synostosis) and should be expected if the forearms look normal but radioulnar rotation is severely limited (you are unable to supinate and pronate the forearm). Another is CAMT (congenital amegakaryocytic thrombocytopenia), which generally occurs with no other malformations but rarely can accompany other conditions, such as Noonan syndrome.

Congenital infections (e.g., TORCH) can be associated with thrombocytopenia, and some cases result in values as low as seen in this patient. The kinetic mechanism for this variety of neonatal thrombocytopenia is often a mixture of reduced production and accelerated destruction.

The prinicipal considerations for this appropriately grown, 37-week-gestation, otherwise healthy-appearing neonate with a platelelet count of 9000/μL (if the MPV is normal or slightly increased, perhaps approximately 11 or 12 fL) include neonatal alloimmune thrombocytopenia, congenital infection, CAMT, and ATRUS.

In this case, an x-ray ( Figure 12-12) of the forearms showed radioulnar synostosis, making ATRUS the likely diagnosis (a mutation in HOXA11 mapping to 7p15). ATRUS is an autosomal dominant condition, but neonates can be much more severely affected than their affected parent. Generally, the affected parent has radioulnar synostosis with limited forearm pronation and supination, but his or her platelet counts tend to be only moderately low (50,000/μL) with minimal clinical bleeding problems. Neonates with ATRUS are likely to be dependent on platelet transfusion, and marrow or cord blood transplantation is the only known curative procedure. ¹³

A transfusion of 15 mL donor platelets per kg will generally increase the recipient’s platelet count by 70,000 to 100,000/μL. The life span of normal platelets in vivo is approximately 10 days. Therefore, when platelets are harvested from a donor, about 10% of these are newly formed and might survive in the recipient for up to 10 days. However, about 10% are effete and will not survive at all in the recipient. On average the donor platelets will decay after transfusion with a half-life of between 1 and 2 days. Thus in 3 or 4 days most (>75%) of the transfused platelets will be gone, and by 5 or 6 days essentially all the transfused platelet will be gone.

In cases of ATRUS, the rise in platelet count following transfusion and the disappearance of the transfused platelets should conform to these principles, because the kinetic mechanism responsible for the thrombocytopenia is reduced platelet production. However, in cases of neonatal thrombocytopenia caused by a platelet consumptive process, such as DIC, a propagating thrombus, or immune-mediated thrombocytopenia, the platelet count will not increase by the expected amount after transfusion and will fall much more rapidly than previously described. ¹⁴

According to a recent survey of neonatologists in North America, about half of the respondents routinely order a platelet transfusion volume of 10 mL/kg and about half order 10 to 15 mL/kg. Because every platelet transfusion given results in another donor exposure, one larger transfusion might offer an advantage over two smaller ones (each from a different donor). On that basis the author recommends a volume of 15 mL/kg. About half of respondents use single-donor platelets, about 10% use platelets pooled from 2 to 3 donors, and about 25% use apheresis-prepared platelets. Pooling donors results in more donor exposures, and the volume reduction needed after pooling reduces platelet viability. The Intermountain Healthcare NICUs (where the author works) use apheresis-prepared platelets exclusively. Between 60% and 70% of respondents use only irradiated platelets for all neonatal platelet transfusions.

Reference Ranges for Neutrophil Counts and Neutropenia

Newborn nurseries in Colorado, New Mexico, and Utah, 4000 to 5500 feet above sea level, have reported much higher reference ranges for blood neutrophil concentrations during the first 3 days after birth than nurseries at or near sea level. It is curious that hematocrit/Hgb levels and NRBC/μL are not higher at the high-altitude hospitals; the mechanism explaining the difference in neutrophil counts is not known. Figure 12-13 shows the reference ranges for high-altitude and sea-level centers superimposed. One value in recognizing that these differences exist is that NICUs at high altitude can falsely label neonates as neutrophilic if the sea-level range is used. Perhaps the opposite is also true; neonates at sea level could have the diagnosis of neutrophilia missed if the reference range for high-altitude centers is used.

Neutropenia is indeed common in neonates born after PIH, particularly if the neonate is preterm and small for gestational age. However, the marked “left shift” is not typical of that variety of neutopenia. It is important to consider that the hypotension, metabolic acidosis, and respiratory distress accompanying the neutropenia and left shift are manifestations of sepsis. Therefore it would be unwise to withhold antibiotic treatment for this patient under the possibly erroneous assumption that the baby has neutropenia associated with PIH and being small for gestational age.

All the preceding actions are appropriate.

Table 12-1 categorizes neonatal neutropenia into subtypes. Some of these varieties are very common, and others are exceedingly rare. Some are the result of reduced neutrophil production, and others the result of accelerated neutrophil utilization (sepsis) or destruction (immune mediated). Although this could be one of the subtypes of severe congenital neutropenia, those are extremely rare. Given that the patient is not ill and has no left shift, this is not the neutropenia of overwhelming sepsis. Because this infant is not small for gestational age and the mother did not have PIH, it is not that variety. Most likely this is a case of alloimmune neonatal neutropenia (ANN), wherein the mother has immunoglobulin G antibody against a neutrophil antigen she lacks but that is expressed by father and fetus. ¹⁵

ANN is a common variety of severe congenital neutropenia, with a prevalence estimated to be as high as one to two per thousand births.

ANN is a potentially critical disorder, with a mortality rate estimated to be as high as 5%, on the basis of acquiring a significant infection during a period of severe neutropenia.

The pathogenesis of ANN involves the passive transfer of neutrophil-specific maternal immunoglobulin G antibodies across the placenta. The antibodies bind to fetal neutrophils, which express an antigen inherited from the father that is absent in the mother. These maternal immunoglobulin G antibodies can result in severe neutropenia before and after birth. Neutrophil-specific antibodies directed against the human neutrophil antigens (HNAs) HNA-1a, HNA-1b and HNA-2a, are detected in more than 50% of ANN cases. Antibodies to HNA-1c, HNA-3a, and HNA-4a are occasionally identified.

A reference laboratory skilled in the diagnosis of ANN is required to ensure an accurate diagnosis. One such outstanding laboratory is the American Red Cross North Central Blood Services in St. Paul, Minnesota. Simply looking for maternal antineutrophil antibodies is not always sufficient. Generally, the evaluation begins using the mother’s serum and granulocyte agglutination (GA) or granulocyte immunofluorescence (GIF) assays. If maternal antineutrophil antibodies are identified, neutrophil genotyping can be performed on whole blood collected from the father and the mother to identify the HNA involved.

Work with your pediatric hematology consultant to decide whether to treat the neonate with ANN with recombinant granulocyte colony-stimulating factor. We generally do so if the neutropenia is severe (<500/μL) for several days, if it is in the range of 500 to 999/μL for approximately 1 week, or if the patient has a bacterial infection. A dose of 10 μg/kg given subcutaneously once daily for about 3 days will usually result in an absolute neutrophil count greater than 1000/μL. Subsequent doses may be needed to keep the absolute neutrophil count above 1000/μL. The duration of the condition roughly corresponds to the disappearance of maternal antineutrophil antibody from the neonate, which sometimes takes up to 2 months or so.

ELANE (ELAstase, Neutrophil Expressed) is the gene encoding neutrophil elastase and is located at 19p13.3. Most cases of severe congential neutropenia (excluding cases of severe alloimmune neutropenia) are the result of ELANE mutations. Severe congenital neutropenia type 1 (OMIM #202700) is an autosomal dominant disorder. More that 70 different ELANE mutations have been identified that result in type 1 severe congenital neutropenia. These mutations each produce a gene product that folds into an incorrect three-dimensional shape. The abnormal neutrophil elastase protein accumulates in neutrophils and damages or kills these cells before they are fully mature.

Other ELANE mutations result in cyclic neutropenia, which can also be categorized as a type of severe congenital neutropenia, wherein cyclic drops in the absolute neutrophil count occur on an every-3-to-4-week cycle. The ELANE mutations causing cyclic neutropenia are generally single nucleotide substitutions that produce a defective neutrophil elastase protein retaining some level of function. ¹⁶¹⁷

No. The Swedish kindred described by Dr. Kostmann in 1956 was an autosomal recessive disorder. The phenotype of Kostmann syndrome is similar to that of severe congenital neutropenia type 1 but is more clinically heterogenous. The OMIM number for Kostmann syndrome is 610738, and it is currently termed severe congenital neutropenia type 3. The condition results from mutations in the HAX1 gene, located at 1q21. HAX1 encodes a mitochondrial-associated protein involved in neutrophil signal transduction and cytoskeletal regulation. Patients with severe congenital neutropenia type 3 can be either homozygous for the mutation or compound heterozygotes, meaning that they inherited one mutation in HAX1 from one parent and a different HAX1 mutation from the other parent.

Eosinophilia

Expected values, also called reference ranges, for eosinophil counts on the day of birth are a function of gestational age, increasing gradually through the second and third trimesters. Reference ranges for eosniophil counts at birth are shown in Figure 12-14. The 95th percentile value (the highest expected limit) at 34 weeks is about 1100/μL. Thus a value of 1000/μL is normal. ¹⁸

Although the eosinophil count has increased significantly from that measured on the day of birth, the value is within the expected range. The reference range for blood eosinophil concentration during the first 28 days after birth is shown in Figure 12-15.

Eosinophils are effector cells involved in allergic and nonallergic inflammatory conditions. Circulating eosinophils are derived from myelocytic progenitors within the marrow and within extramedullary sites as well. After exiting the site of production and entering the blood, eosinophils circulate for approximately 1 day (T½ 18 hours), after which they transmigrate to tissues, primarily in the gastrointestinal tract, where they produce cytokines and chemokines. Several pathologic conditions in neonates are associated with eosinophilic tissue infiltration, and these conditions are often accompanied by blood eosinophilia. The conditions include erythema toxicum, neonatal eosinophilic pustulosis, and bronchopulmonary dysplasia. Other inflammatory conditions associated with eosinophilia are neonatal eosinophilic esophagitis, eosinophilic colitis, subcutaneous fat necrosis with eosinophilic granules, a variety of infectious diseases, and necrotizing enterocolitis after erythrocyte transfusion. A slight increase can occasionally be expected in preterm infants corresponding to the time weight gain is established.

Coagulation

First, you might inquire if the mother knows the type of von Willebrand disease that she has. You should also find out if she had any bleeding problems as a baby and about her own bleeding history. Perhaps the parents already know that von Willebrand disease is the most common inherited disorder of coagulation, with a prevalence as high as 1% of the general population. They might also already know that von Willebrand disease is not a single disorder and that the most common variety, type 1 (the most likely type the baby may have inherited from mother), is hardly ever problematic during the newborn period. Only type 3, a very rare autosomal recessive variety completely lacking von Willebrand factor antigen and activity, is likely to result in excessive bleeding in infants. Table 12-2 reviews the features of von Willebrand subtypes.¹⁹

The factor V Leiden mutation and the prothrombin 20210 mutation are so commonly found in women (men, too, for that matter) that this specific question of neonatal risk with an affected parent comes up regularly in NICU practice.

Up to 10% of people of Northern European ancestry carry one copy of the factor V Leiden mutation (OMIM #188055), which is a single nucleotide substitution (R506Q) in the F5 gene on 1q24.2. This variation results in activated factor V that is resistant to inactivation by its physiologic regulator, activated protein C. In the very rare instance when both parents carry the factor V Leiden mutation, one in four of their offspring can inherit both defective genes in a homozygous fashion, and the neonate can indeed be at high risk for neonatal thrombosis. However, neonates with a single copy of the factor V Leiden mutation (heterozygote) appear to be at no increased risk for neonatal thrombosis. Thrombotic risks similar to that of the affected parent will be present during adolescence and adult life.

Up to 2% of Caucasians carry the prothrombin 20210 mutation, which increases the amount of prothrombin in the circulation by 15% to 30%. The human prothrombin (F2) gene is located at 11p11-q12, and the 20210 mutation is a single nucleotide substitution. Neonates who inherit this mutation from their father or mother are not at increased risk, as neonates, for thrombotic disorders. However, in the very rare instance when both father and mother have the prothrombin 20210 mutation, the fetus can be homozygous for this mutation. Such neonates have been reported to have a significant risk of neonatal thrombosis.

Hemophilia A (OMIM # 306700) is an X-linked recessive disorder caused by a deficiency in the activity of coagulation factor VIII.

The gene F8 encoding factor VIII is located at Xq28. Bleeding severity in patients with hemophilia A correlates with the plasma levels of factor VIII. Patients are expected to have “mild” bleeding problems with levels in the range of 6% to 30% of normal, “moderate” with levels 2% to 5% of normal, and “severe” with levels less than 1% of normal.

Hemophilia B (OMIM # 306900) is also an X-linked recessive disorder but is caused by a deficiency in the activity of coagulation factor IX. The gene F9 encoding factor IX is located at Xq27.1.

The parents of this boy are wise to determine whether their son has hemophilia before a circumcision is done. In fact, the CDC recommends that a baby who might have hemophilia (born to a known carrier, or to a mother with brothers known to have hemophilia) should avoid circumcision. However, if parents of such a neonate insist that a circumcision is performed, the Centers for Disease Control and Prevention recommend that a pediatric hematologist be consulted before the procedure to ensure that the child receives proper treatment to prevent excessive bleeding. Many neonates with mild or moderate hemophilia have been circumcised with little or no abnormal bleeding, but it is unwise to proceed with circumcision given this family history without first making a diagnosis and without involving a hematologist.

It is helpful for parents who are planning to deliver a son with a risk of hemophilia (based on the family history) to inform you before the delivery. This knowledge will allow you to obtain the necessary tests from fetal blood drawn from the placental end of the umbilical vein immediately after placental delivery. This method does not require blood to be drawn from the neonate for diagnostic testing.

Using umbilical cord (fetal) plasma, you can quantify factor VIII and factor IX and obtain a prothrombin time (PT) and activated partial thromboplastin time (aPTT). You can also order a CBC from the umbilical cord blood so that you are aware of the initial platelet count and Hgb level. Because factor IX is a vitamin K–dependent factor, normal levels are relatively low at birth. Therefore mild hemophilia B is difficult to confirm at times, based on low factor IX levels, for the first month or so. However, factor VIII plasma levels should be at adult levels even in the cord blood. Thus a diagnosis of hemophilia A can be made using umbilical cord blood.

DIC in a neonate is actually more of a clinical than a laboratory diagnosis—bleeding at multiple sites with hypotension. Laboratory confirmation includes thrombocytopenia, or at least a falling platelet count, along with a low fibrinogen or at least a falling fibrinogen level, and elevated or rising D-dimers. Prolongation in the PT and aPTT also support the diagnosis. No single test is sufficient to unequivocally diagnose DIC in a neonate.

Acute bleeding as a result of liver failure is rare in neonates but can present much like DIC. This condition, largely the result of poor production of procoagulant factors by damaged or abnormal hepatocytes, can be distinguished from DIC by the factor VIII level. Factor VIII is not made by hepatocytes but by endothelial cells, and plasma levels increase in liver failure but fall in DIC.

The only treatment for DIC in neonates that works reliably is to reverse the DIC triggers: sepsis, hypoxia, acidosis, hypotension. The nonintuitive treatment of anticoagulation has not been adequately tested in hemorrhaging neonates. It sometimes seems to work, but it is best left to the rare cases of DIC that present with multiple thrombotic sites, such as purpura fulminans.

Transfusing fresh frozen plasma and platelets has been said to exacerbate DIC, but when the bleeding is severe and difficult to control and all else is failing, these transfusions probably have a place.

In case reports recombinant factor VIIa has been used successfully to treat life-threatening bleeding in neonates with DIC. Also, activated protein C is under investigation as an adjunctive treatment for neonates with DIC, but the risks of either of these potential treatments are great, and risk-to-benefit evaluations are needed.