CURRENT Diagnosis and Treatment Pediatrics, (Current Pediatric Diagnosis & Treatment) 22nd Edition

30. Hematologic Disorders

Daniel R. Ambruso, MD

Rachelle Nuss, MD

Michael Wang, MD


The normal ranges for peripheral blood counts vary significantly with age. Normal neonates show a relative polycythemia with a hematocrit concentration of 45%–65%. The reticulocyte count at birth is relatively high at 2%–8%. Within the first few days of life, erythrocyte production decreases, and the values for hemoglobin and hematocrit fall to a nadir at about 6–8 weeks. During this period, known as physiologic anemia of infancy, normal infants have hemoglobin values as low as 10 g/dL and hematocrits as low as 30%. Thereafter, the normal values for hemoglobin and hematocrit gradually increase until adult values are reached after puberty. Premature infants can reach a nadir hemoglobin level of 7–8 g/dL at 8–10 weeks. Anemia is defined as a hemoglobin concentration two standard deviations below the mean for a normal population of the same gender and age.

Newborns have larger red cells than children and adults, with a mean corpuscular volume (MCV) at birth of more than 94 fL. The MCV subsequently falls to a nadir of 70–84 fL at about age 6 months. Thereafter, the normal MCV increases gradually until it reaches adult values after puberty.

The normal number of white blood cells (WBCs) is higher in infancy and early childhood than later in life. Neutrophils predominate in the differential white count at birth and in the older child. Lymphocytes predominate (up to 80%) between about ages 1 month and 6 years.

Normal values for the platelet count are 150,000–400,000/μL and vary little with age.


Failure of the marrow to produce adequate numbers of circulating blood cells may be congenital or acquired and may cause pancytopenia (aplastic anemia) or involve only one cell line (single cytopenia). Constitutional and acquired aplastic anemias are discussed in this section and the more common single cytopenias in later sections. Bone marrow failure caused by malignancy or other infiltrative disease is discussed in Chapter 31. It is important to remember that many drugs and toxins may affect the marrow and cause single or multiple cytopenias.

Suspicion of bone marrow failure is warranted in children with pancytopenia and in children with single cytopenias who lack evidence of peripheral red cell, white cell, or platelet destruction. Macrocytosis often accompanies bone marrow failure. Many of the constitutional bone marrow disorders are associated with a variety of congenital anomalies.



image Progressive pancytopenia.

image Macrocytosis.

image Multiple congenital anomalies.

image Increased chromosome breakage in peripheral blood lymphocytes.

image General Considerations

Fanconi anemia, which is the number 1 inherited bone marrow failure syndrome, is characterized by defective DNA repair that is caused by a variety of genetic mutations. Inheritance is autosomal recessive, and this disease occurs in all ethnic groups; 75%–90% of affected individuals develop bone marrow failure in the first 10 years of life.

image Clinical Findings

A. Symptoms and Signs

Symptoms are determined principally by the degree of hematologic abnormality. Thrombocytopenia may cause purpura, petechiae, and bleeding; neutropenia may cause severe or recurrent infections; and anemia may cause weakness, fatigue, and pallor. Congenital anomalies are present in at least 50% of patients. The most common include abnormal pigmentation of the skin (generalized hyperpigmentation, café au lait or hypopigmented spots), short stature with delicate features, and skeletal malformations (hypoplasia, anomalies, or absence of the thumb and radius). More subtle anomalies are hypoplasia of the thenar eminence or a weak or absent radial pulse. Associated renal anomalies include aplasia, horseshoe kidney, and duplication of the collecting system. Other anomalies are microcephaly, microphthalmia, strabismus, ear anomalies, and hypogenitalism.

B. Laboratory Findings

Thrombocytopenia or leukopenia typically occurs first, followed over the course of months to years by anemia and progression to severe aplastic anemia. Macrocytosis is virtually always present; is usually associated with anisocytosis and an elevation in fetal hemoglobin levels; and is an important diagnostic clue. The bone marrow reveals hypoplasia or aplasia. The diagnosis is confirmed by the demonstration of an increased number of chromosome breaks and rearrangements in peripheral blood lymphocytes. The use of diepoxybutane to stimulate these breaks and rearrangements provides a sensitive assay that is virtually always positive in children with Fanconi anemia, even before the onset of hematologic abnormalities.

Specific Fanconi genes (FANCA, B, C, and others) have been identified and transmission is generally autosomal, although FANCB is on the X chromosome.

image Differential Diagnosis

Because patients with Fanconi anemia frequently present with thrombocytopenia, the disorder must be differentiated from idiopathic thrombocytopenic purpura (ITP) and other more common causes of thrombocytopenia. In contrast to patients with ITP, those with Fanconi anemia usually exhibit a gradual fall in the platelet count. Counts less than 20,000/μL are often accompanied by neutropenia or anemia. Fanconi anemia may also be manifested initially by pancytopenia, and must be differentiated from acquired aplastic anemia and other disorders, such as acute leukemia. Examination of the bone marrow and chromosome studies of peripheral blood lymphocytes (chromosomal breakage) will usually distinguish between these disorders.

image Complications

Complications are those related to thrombocytopenia and neutropenia. Endocrine dysfunction may include growth hormone deficiency, hypothyroidism, or impaired glucose metabolism. Persons with Fanconi anemia have a significantly increased risk of developing malignancies, especially acute nonlymphocytic leukemia (800-fold), head and neck cancers, genital cancers, and myelodysplastic syndromes.

image Treatment

Attentive supportive care is critical. Patients with neutropenia who develop fever require prompt evaluation and parenteral broad-spectrum antibiotics. Transfusions are important, but should be used judiciously, especially in the management of thrombocytopenia, which frequently becomes refractory to platelet transfusions as a consequence of alloimmunization. Transfusions from family members should be discouraged because of the negative effect on the outcome of bone marrow transplant. At least 50% of patients with Fanconi anemia respond, albeit incompletely, to oxymetholone, and many recommend institution of androgen therapy before transfusions are needed. However, oxymetholone is associated with hepatotoxicity, hepatic adenomas, and masculinization, and is particularly troublesome for female patients.

The definitive treatment is a reduced intensity hematopoietic stem cell transplant, ideally from a human leukocyte antigen (HLA)–identical sibling donor, although matched unrelated and cord transplant may be considered. Before transplant, any prospective sibling donor must be screened for Fanconi anemia.

image Prognosis

Many patients succumb to bleeding, infection, or malignancy in adolescence or early adulthood. Stem cell transplant does not reduce the increased susceptibility for malignancy.

Mehta P, Locatelli F, Stary J, Smith FO: Bone marrow transplantation for inherited bone marrow failure syndromes. Pediatr Clin North Am 2010;57:147–170 [PMID: 20307716].

Younghoon K, D’Andrea AD: Molecular pathogenesis and clinical management of Fanconi anemia. J Clin Invest 2012;122: 3799–3806 [PMID: 23114602].



image Weakness and pallor.

image Petechiae, purpura, and bleeding.

image Frequent or severe infections.

image Pancytopenia with hypocellular bone marrow.

image General Considerations

Acquired aplastic anemia is characterized by peripheral pancytopenia with a hypocellular bone marrow. Approximately 50% of cases in childhood are idiopathic. Other cases are secondary to idiosyncratic reactions to drugs such as phenylbutazone, sulfonamides, nonsteroidal anti-inflammatory drugs (NSAIDs), and anticonvulsants. Toxic causes include exposure to benzene, insecticides, and heavy metals. Infectious causes include viral hepatitis, infectious mononucleosis (Epstein-Barr virus [EBV]), and human immunodeficiency virus (HIV). In immunocompromised children, aplastic anemia has been associated with human parvovirus B19. Immune mechanisms of marrow suppression are suspected in most cases.

image Clinical Findings

A. Symptoms and Signs

Weakness, fatigue, and pallor result from anemia; petechiae, purpura, and bleeding occur due to thrombocytopenia; and fevers due to generalized or localized infections are associated with neutropenia. Hepatosplenomegaly and significant lymphadenopathy are unusual.

B. Laboratory Findings

Anemia is usually normocytic, with a low reticulocyte count. The WBC count is low, with a marked neutropenia. The platelet count is typically below 50,000/μL and is frequently below 20,000/μL. Bone marrow biopsy shows a marked decrease in cellularity typically less than 20% of normal in severe aplastic anemia.

image Differential Diagnosis

Examination of the bone marrow usually excludes pancytopenia caused by peripheral destruction of blood cells or by infiltrative processes such as acute leukemia, storage diseases, and myelofibrosis. Many of these other conditions are associated with hepatosplenomegaly. Preleukemic conditions also may present with pancytopenia and hypocellular bone marrows. Cytogenetic analysis of the marrow is helpful, because a clonal abnormality may predict the subsequent development of leukemia. Since congenital anomalies may not be apparent in some children with Fanconi anemia, patients with newly diagnosed aplastic anemia should be studied for chromosome breaks and rearrangements in peripheral blood lymphocytes.

image Complications

Acquired aplastic anemia is characteristically complicated by infection and hemorrhage, which are the leading causes of death. Other complications are those associated with therapy.

image Treatment

Comprehensive supportive care is essential. Febrile illnesses require prompt evaluation and usually parenteral antibiotics. Red blood cell (RBC) transfusions alleviate symptoms of anemia. Platelet transfusions may be lifesaving, but they should be used sparingly because many patients eventually develop platelet alloantibodies and become refractory to platelet transfusions.

Immunomodulation, usually with antithymocyte globulin and cyclosporine or tacrolimus, is associated with a high response rate and improved overall survival. However, incomplete response, relapse, and progression to myelodysplasia/leukemia may occur. Hematopoietic stem cell transplant is the treatment of choice for severe aplastic anemia when an HLA-identical sibling donor is available. Because the likelihood of success with transplant is influenced adversely by receipt of transfusions, HLA typing of family members should be undertaken at the time of diagnosis. Increasingly, patients who lack HLA-identical siblings are able to find matched donors through cord blood banks or the National Marrow Donor Program.

image Prognosis

Children receiving early bone marrow transplant from an HLA-identical sibling have a long-term survival rate of greater than 80%. Sustained, complete remissions may be seen in 65%–80% of patients receiving immunosuppressive therapy. However, both therapies are associated with an increased risk of myelodysplastic syndromes, acute leukemia, and other malignancies in long-term survivors.

Korthof ET, Kekassy AN, Hussein AA: Management of acquired aplastic anemia in children. Bone Marrow Transplant 2013;48:191–195 [PMID: 23292240].

Samarasinghe S et al: Excellent outcome of matched unrelated donor transplantation in paediatric aplastic anaemia following failure with immunosuppressive therapy: a United Kingdom multicentre retrospective experience. Br J Haematol 2012;157:339–346 [PMID: 22372373].



Anemia is a relatively common finding, and identifying the cause is important. Even though anemia in childhood has many causes, the correct diagnosis can usually be established with relatively little laboratory cost. Frequently the cause is identified with a careful history. Nutritional causes should be sought by inquiry about dietary intake; growth and development; and symptoms of chronic disease, malabsorption, or blood loss. Hemolytic disease may be associated with a history of jaundice (including neonatal jaundice) or by a family history of anemia, jaundice, gallbladder disease, splenomegaly, or splenectomy. The child’s ethnicity may suggest the possibility of certain hemoglobinopathies or deficiencies of red cell enzymes, such as glucose-6-phosphate dehydrogenase (G6PD). The review of systems may reveal clues to a previously unsuspected systemic disease associated with anemia. The patient’s age is important because some causes of anemia are age related. For example, patients with iron-deficiency anemia (IDA) and β-globin disorders present more commonly at ages 6–36 months than at other times in life.

The physical examination may also reveal clues to the cause of anemia. Poor growth may suggest chronic disease or hypothyroidism. Congenital anomalies may be associated with constitutional aplastic anemia (Fanconi anemia) or with congenital hypoplastic anemia (Diamond-Blackfan anemia). Other disorders may be suggested by the findings of petechiae or purpura (leukemia, aplastic anemia, hemolytic uremic syndrome), jaundice (hemolysis or liver disease), generalized lymphadenopathy (leukemia, juvenile rheumatoid arthritis, HIV infection), splenomegaly (leukemia, sickle hemoglobinopathy syndromes, hereditary spherocytosis, liver disease, hypersplenism), or evidence of chronic or recurrent infections.

The initial laboratory evaluation of the anemic child consists of a complete blood count (CBC) with differential and platelet count, review of the peripheral blood smear, and a reticulocyte count. The algorithm in Figure 30–1 uses limited laboratory information, together with the history and physical examination, to reach a specific diagnosis or to focus additional laboratory investigations on a limited diagnostic category (eg, microcytic anemia, bone marrow failure, pure red cell aplasia, or hemolytic disease). This diagnostic scheme depends principally on the MCV to determine whether the anemia is microcytic, normocytic, or macrocytic, according to the percentile curves of Dallman and Siimes (Figure 30–2).


image Figure 30–1. Investigation of anemia.


image Figure 30–2. Hemoglobin and red cell volume in infancy and childhood. (Reproduced, with permission, from Dallman PR, Siimes MA: Percentile curves for hemoglobin and red cell volume in infancy and childhood. J Pediatr 1979;94:26.)

Although the incidence of iron deficiency (ID) in the United States has decreased significantly with improvements in infant nutrition, it remains an important cause of microcytic anemia, especially at ages 6–24 months. A trial of therapeutic iron is appropriate in such children, provided the dietary history is compatible with ID and the physical examination or CBC does not suggest an alternative cause for the anemia. If a trial of therapeutic iron fails to correct the anemia and microcytosis, further evaluation is warranted.

Another key element of Figure 30–1 is the use of both the reticulocyte count and the peripheral blood smear to determine whether a normocytic or macrocytic anemia is due to hemolysis. Typically hemolytic disease is associated with an elevated reticulocyte count, but some children with chronic hemolysis initially present during a period of a virus-induced aplasia when the reticulocyte count is not elevated. Thus, review of the peripheral blood smear for evidence of hemolysis (eg, spherocytes, red cell fragmentation, sickle forms) is important in the evaluation of children with normocytic anemias and low reticulocyte counts. When hemolysis is suggested, the correct diagnosis may be suspected by specific abnormalities of red cell morphology or by clues from the history or physical examination. Autoimmune hemolysis is usually excluded by a negative direct antiglobulin test (DAT). Review of blood counts and the peripheral blood smears of the mother and father may suggest genetic disorders such as hereditary spherocytosis. Children with normocytic or macrocytic anemias, with relatively low reticulocyte counts and no evidence of hemolysis on the blood smear, usually have anemias caused by inadequate erythropoiesis in the bone marrow. The presence of neutropenia or thrombocytopenia in such children suggests the possibility of aplastic anemia, malignancy, or severe folate or vitamin B12 deficiency, and usually dictates examination of the bone marrow.

Pure red cell aplasia may be congenital (Diamond-Blackfan anemia), acquired, and transient (transient erythroblastopenia of childhood); a manifestation of a systemic disease such as renal disease or hypothyroidism; or associated with malnutrition or mild deficiencies of folate or vitamin B12.

Janus J, Moerschel S: Evaluation of anemia in children. Am Fam Physician 2010;15:1462–1471 [PMID: 20540485].


Infants and children with normocytic or macrocytic anemia, a low reticulocyte count, and normal or elevated numbers of neutrophils and platelets should be suspected of having pure red cell aplasia. Examination of the peripheral blood smear in such cases is important because signs of hemolytic disease suggest chronic hemolysis complicated by an aplastic crisis due to parvovirus infection. Appreciation of this phenomenon is important because chronic hemolytic disease may not be diagnosed until the anemia is exacerbated by an episode of red cell aplasia and subsequent rapidly falling hemoglobin level. In such cases, cardiovascular compromise and congestive heart failure may develop quickly.

1. Congenital Hypoplastic Anemia (Diamond-Blackfan Anemia)


image Age: birth to 1 year.

image Macrocytic anemia with reticulocytopenia.

image Bone marrow with erythroid hypoplasia.

image Short stature or congenital anomalies in one-third of patients.

image General Considerations

Diamond-Blackfan anemia is a relatively rare cause of anemia that usually presents at 2–3 months of age. To date, mutations of genes encoding ribosomal proteins occurring autosomal dominance has been recognized. Early diagnosis is important because treatment with corticosteroids results in increased erythropoiesis in 80% of patients, thus avoiding the difficulties and complications of long-term chronic transfusion therapy.

image Clinical Findings

A. Symptoms and Signs

Signs and symptoms are generally those of chronic anemia, such as pallor and congestive heart failure. Jaundice, splenomegaly, or other evidence of hemolysis are usually absent. Short stature or other congenital anomalies are present in 50% of patients. A wide variety of anomalies have been described; craniofacial and triphalangeal thumbs are the most common.

B. Laboratory Findings

Diamond-Blackfan anemia is characterized by severe macrocytic anemia and marked reticulocytopenia. The neutrophil count is usually normal or slightly decreased, and the platelet count is normal, elevated, or decreased. The bone marrow usually shows a marked decrease in erythroid precursors but is otherwise normal. In older children, fetal hemoglobin levels are usually increased and there is evidence of persistent fetal erythropoiesis, such as the presence of the i antigen on erythrocytes. In addition, the level of adenosine deaminase in erythrocytes is elevated.

image Differential Diagnosis

The principal differential diagnosis is transient erythroblastopenia of childhood. Children with Diamond-Blackfan anemia generally present at an earlier age, often have macrocytosis, and have evidence of fetal erythropoiesis and an elevated level of red cell adenosine deaminase. In addition, short stature and congenital anomalies are not associated with transient erythroblastopenia. Lastly, transient erythroblastopenia of childhood usually resolves within 6–8 weeks of diagnosis, whereas Diamond-Blackfan anemia is a lifelong affliction. Other disorders associated with decreased red cell production such as renal failure, hypothyroidism, and the anemia of chronic disease need to be considered.

image Treatment

Oral corticosteroids should be initiated at the time of diagnosis. Eight percent of patients respond to prednisone, 2 mg/kg/d, and many who respond subsequently tolerate significant tapering of the dose. Patients who are unresponsive to prednisone require chronic transfusion therapy, which inevitably causes transfusion-induced hemosiderosis and the need for chelation. Bone marrow transplant is an alternative definitive therapy that should be considered for transfusion-dependent patients who have HLA-identical siblings. Unpredictable spontaneous remissions occur in up to 20% of patients.

image Prognosis

The prognosis for patients responsive to corticosteroids is generally good, particularly if remission is maintained with low doses of alternate-day prednisone. Patients dependent on transfusion are at risk for the complications of hemosiderosis. There is an increased risk for the development of solid tumors.

Horos R, vonLindern: Molecular mechanisms of pathology and treatment in Diamond Blackfan Anaemia. Bri J Haematol 2012;159:514–527 [PMID: 23016900].

2. Transient Erythroblastopenia of Childhood


image Age: 6 months to 4 years.

image Normocytic anemia with reticulocytopenia.

image Absence of hepatosplenomegaly or lymphadenopathy.

image Erythroid precursors initially absent from bone marrow.

image General Considerations

Transient erythroblastopenia of childhood is a relatively common cause of acquired anemia in early childhood. The disorder is suspected when a normocytic anemia is discovered during evaluation of pallor or when a CBC is obtained for another reason. Because the anemia is due to decreased red cell production, and thus develops slowly, the cardiovascular system has time to compensate. Therefore, children with hemoglobin levels as low as 4–5 g/dL may look remarkably well. The disorder is thought to be autoimmune in most cases, because IgG from some patients has been shown to suppress erythropoiesis in vitro.

image Clinical Findings

Pallor is the most common sign, and hepatosplenomegaly and lymphadenopathy are absent. The anemia is normocytic, and the peripheral blood smear shows no evidence of hemolysis. The platelet count is normal or elevated, and the neutrophil count is normal or, in some cases, decreased. Early in the course, no reticulocytes are identified. The Coombs test is negative, and there is no evidence of chronic renal disease, hypothyroidism, or other systemic disorder. Bone marrow examination shows severe erythroid hypoplasia initially; subsequently, erythroid hyperplasia develops along with reticulocytosis, and the anemia resolves.

image Differential Diagnosis

Transient erythroblastopenia of childhood must be differentiated from Diamond-Blackfan anemia, particularly in infants younger than age 1 year. In contrast to Diamond-Blackfan anemia, transient erythroblastopenia is not associated with macrocytosis, short stature, or congenital anomalies, or with evidence of fetal erythropoiesis prior to the phase of recovery. Also in contrast to Diamond-Blackfan anemia, transient erythroblastopenia is associated with normal levels of red cell adenosine deaminase. Transient erythroblastopenia of childhood must also be differentiated from chronic disorders associated with decreased red cell production, such as renal failure, hypothyroidism, and other chronic states of infection or inflammation. As with other single cytopenias, the possibility of malignancy (ie, leukemia) should always be considered, particularly if fever, bone pain, hepatosplenomegaly, or lymphadenopathy is present. In such cases, examination of the bone marrow is generally diagnostic. Confusion may sometimes arise when the anemia of transient erythroblastopenia is first identified during the early phase of recovery when the reticulocyte count is high. In such cases, the disorder may be confused with the anemia of acute blood loss or with hemolytic disease. In contrast to hemolytic disorders, transient erythroblastopenia of childhood is not associated with jaundice or peripheral destruction of red cells.

image Treatment & Prognosis

By definition, this is a transient disorder. Some children require red cell transfusions if cardiovascular compromise is present. Resolution of the anemia is heralded by an increase in the reticulocyte count, which generally occurs within 4–8 weeks of diagnosis. Transient erythroblastopenia of childhood is not treated with corticosteroids because of its short course.


1. Iron-Deficiency Anemia


image Pallor and fatigue.

image Poor dietary intake of iron (ages 6–24 months).

image Chronic blood loss (age > 2 years).

image Microcytic hypochromic anemia.

image General Considerations

Iron deficiency (ID) and iron-deficiency anemia (IDA) are a worldwide concern. ID is defined as a state in which there is insufficient iron to maintain normal physiologic functions such that iron stores (serum ferritin or bone marrow iron content) are reduced. IDA is defined as a hemoglobin more than 2 standard deviations below normal for age and gender, which has developed as a consequence of ID.

Normal-term infants are born with sufficient iron stores to prevent ID for the first 4 months of life, whereas premature infants have reduced iron stores since iron is predominantly acquired in the last trimester. Thus premature infants, as well as those with low birth weight, neonatal anemia, perinatal blood loss, or subsequent hemorrhage may have reduced iron stores. Breast milk is low in iron relative to cow’s milk and fortified formulas, and without iron supplementation, ID may develop in exclusively breast-fed children.

image Clinical Findings

A. Symptoms and Signs

Symptoms and signs vary with the severity of the deficiency. ID is usually asymptomatic. IDA may be associated with, pallor, fatigue, and irritability. A history of pica is common. It is controversial whether or not ID/IDA adversely affects long-term neurodevelopment and behavior. IDA is associated with increased lead absorption and subsequent neurotoxicity.

B. Laboratory Findings

According to the American Academy of Pediatrics (AAP) 2010 guidelines, screening for anemia should be performed at about 12 months of age with determination of hemoglobin concentration and an assessment of risk factors for ID/IDA. Risks include low socioeconomic status, prematurity or low birth weight, lead exposure, exclusive breast-feeding beyond 4 months of age without iron supplementation, weaning to whole milk or complementary foods that do not include iron, feeding problems, poor growth, and inadequate nutrition. If the hemoglobin is less than 11 mg/dL or there is a high risk for ID, an iron evaluation should be performed. There is no single measurement that will document the iron status; recommended tests include serum ferritin and C-reactive protein or reticulocyte hemoglobin concentration.

image Differential Diagnosis

The differential diagnosis is that of microcytic, hypochromic anemia. The possibility of thalassemia (α-thalassemia, β-thalassemia, and hemoglobin E disorders) should be considered, especially in infants of African, Mediterranean, or Asian ethnic background. In contrast to infants with ID, those with thalassemia generally have an elevated red cell number and are less likely, in mild cases, to have an elevated RBC distribution width (the index of the MCV divided by the red cell number is usually < 13). Thalassemias are associated with normal or increased levels of serum iron and ferritin and with normal iron-binding capacity. The hemoglobin electrophoresis in β-thalassemia minor typically shows an elevation of hemoglobin A2 levels, but coexistent ID may lower the percentage of hemoglobin A2 into the normal range. Hemoglobin electrophoresis will also identify children with hemoglobin E, a cause of microcytosis common in Southeast Asians. In contrast, the hemoglobin electrophoresis in α-thalassemia trait is normal. Lead poisoning has also been associated with microcytic anemia, but anemia with lead levels less than 40 mg/dL is often due to coexistent ID.

The anemia of chronic inflammation or infection is normocytic but in late stages may be microcytic. This anemia is usually suspected because of the presence of a chronic systemic disorder and an elevated CRP. Relatively mild infections, particularly during infancy, may cause transient anemia. Thus, screening tests for anemia should not be obtained within 3–4 weeks of such infections.

image Treatment

The AAP has published guidelines for routine iron intake for children. If a child has a hemoglobin of 10–11 mg/dL at the 12-month screening visit, the child can be closely monitored or empirically treated with iron supplementation with a recheck of hemoglobin in one month.

If a child is found to have ID/IDA, the recommended oral dose of elemental iron is 6 mg/kg/d in three divided daily doses. Parenteral administration of iron is rarely necessary. Iron therapy results in an increased reticulocyte count within 3–5 days, which is maximal between 5 and 7 days. The rate of hemoglobin rise is inversely related to the hemoglobin level at diagnosis. Resolution of the anemia is within 4–6 weeks. Treatment is generally continued for a few additional months to replenish iron stores.

Baker RD, Greer FR; the Committee on Nutrition: Diagnosis and prevention of iron deficiency and iron-deficiency anemia in infants and young children (0–3 years of age). Pediatrics 2010;126:1040–1050 [PMID: 2093825].

Eden AN, Sandoval C: Iron deficiency in infants and toddlers in the United States. Pediatr Hematol Oncol 2012;29:704–709 [PMID: 2303474].

2. Megaloblastic Anemias


image Pallor and fatigue.

image Nutritional deficiency or intestinal malabsorption.

image Macrocytic anemia.

image Megaloblastic bone marrow changes.

image General Considerations

Megaloblastic anemia is a macrocytic anemia caused by deficiency of cobalamin (vitamin B12), folic acid, or both. Cobalamin deficiency due to dietary insufficiency may occur in infants who are breast fed by mothers who are strict vegetarians or who have pernicious anemia. Intestinal malabsorption is the usual cause of cobalamin deficiency in children and occurs with Crohn disease, chronic pancreatitis, bacterial overgrowth of the small bowel, infection with the fish tapeworm (Diphyllobothrium latum), or after surgical resection of the terminal ileum. Deficiencies due to inborn errors of metabolism (transcobalamin II deficiency and methylmalonic aciduria) also have been described. Malabsorption of cobalamin due to deficiency of intrinsic factor (pernicious anemia) is rare in childhood.

Folic acid deficiency may be caused by inadequate dietary intake, malabsorption, increased folate requirements, or some combination of the three. Folate deficiency due to dietary deficiency alone is rare but occurs in severely malnourished infants and has been reported in infants fed with goat’s milk not fortified with folic acid. Folic acid is absorbed in the jejunum, and deficiencies are encountered in malabsorptive syndromes such as celiac disease. Anticonvulsion medications (eg, phenytoin and phenobarbital) and cytotoxic drugs (eg, methotrexate) also have been associated with folate deficiency, caused by interference with folate absorption or metabolism. Finally, folic acid deficiency is more likely to develop in infants and children with increased requirements. This occurs during infancy because of rapid growth and also in children with chronic hemolytic anemia. Premature infants are particularly susceptible to the development of the deficiency because of low body stores of folate.

image Clinical Findings

A. Symptoms and Signs

Infants with megaloblastic anemia may show pallor and mild jaundice as a result of ineffective erythropoiesis. Classically, the tongue is smooth and beefy red. Infants with cobalamin deficiency may be irritable and may be poor feeders. Older children with cobalamin deficiency may complain of paresthesias, weakness, or an unsteady gait and may show decreased vibratory sensation and proprioception on neurologic examination.

B. Laboratory Findings

The laboratory findings of megaloblastic anemia include an elevated MCV and mean corpuscular hemoglobin (MCH). The peripheral blood smear shows numerous macro-ovalocytes with anisocytosis and poikilocytosis. Neutrophils are large and have hypersegmented nuclei. The white cell and platelet counts are normal with mild deficiencies but may be decreased in more severe cases. Examination of the bone marrow is not indicated, but it typically shows erythroid hyperplasia with large erythroid and myeloid precursors. Nuclear maturation is delayed compared with cytoplasmic maturation, and erythropoiesis is ineffective. The serum indirect bilirubin concentration may be slightly elevated.

Children with cobalamin deficiency have a low serum vitamin B12 level, but decreased levels of serum vitamin B12 may also be found in about 30% of patients with folic acid deficiency. Negative results should not negate treatment if clinically compatible symptoms are present. The level of red cell folate is a better reflection of folate stores than is the serum folic acid level. Elevated serum levels of metabolic intermediates (methylmalonic acid and homocysteine) may help establish the correct diagnosis. Elevated methylmalonic acid levels are consistent with cobalamin deficiency and generally decrease with treatment, whereas elevated levels of homocysteine occur with both cobalamin and folate deficiency.

image Differential Diagnosis

Most macrocytic anemias in pediatrics are not megaloblastic. Other causes of an increased MCV include drug therapy (eg, anticonvulsants, anti-HIV nucleoside analogues), Down syndrome, an elevated reticulocyte count (hemolytic anemias), bone marrow failure syndromes (Fanconi anemia, Diamond-Blackfan anemia), liver disease, and hypothyroidism.

image Treatment

Treatment of cobalamin deficiency due to inadequate dietary intake is readily accomplished with high-dose oral supplementation that is as effective as parenteral treatment if absorption is normal. Folic acid deficiency is treated effectively with oral folic acid in most cases. Children at risk for the development of folic acid deficiencies, such as premature infants and those with chronic hemolysis, are often given folic acid prophylactically.

Stabler SP: Vitamin B12 deficiency. New Engl J Med 2013;368:149–160 [PMID: 23301732].


Anemia is a common manifestation of many chronic illnesses in children. In some instances, causes may be mixed. For example, children with chronic disorders involving intestinal malabsorption or blood loss may have anemia of chronic inflammation in combination with nutritional deficiencies of iron, folate, or cobalamin. In other settings, the anemia is due to dysfunction of a single organ (eg, renal failure, hypothyroidism), and correction of the underlying abnormality resolves the anemia.

1. Anemia of Chronic Inflammation

Anemia is frequently associated with chronic infections or inflammatory diseases. The anemia is usually mild to moderate in severity, with a hemoglobin level of 8–12 g/dL. In general, the severity of the anemia corresponds to the severity of the underlying disorder, and there may be microcytosis, but not hypochromia. The reticulocyte count is low. The anemia is thought to be due to inflammatory cytokines that inhibit erythropoiesis, and shunting of iron into, and impaired iron release from, reticuloendothelial cells. High levels of hepcidin, a peptide produced in the liver during infection or inflammation, reduce iron absorption by the duodenum and release from macrophages. Levels of erythropoietin are relatively low for the severity of the anemia. The serum iron concentration is low, but in contrast to ID, anemia of chronic inflammation is not associated with elevated iron-binding capacity and is associated with an elevated serum ferritin level. Treatment consists of correction of the underlying disorder, which, if controlled, generally results in improvement in hemoglobin level.

Ganz T: Hepcidin and iron regulation, 10 years later. Blood 2011;117:4425 [PMID: 21346250].

2. Anemia of Chronic Renal Failure

Severe normocytic anemia occurs in most forms of renal disease that have progressed to renal insufficiency. Although white cell and platelet production remain normal, the bone marrow shows significant hypoplasia of the erythroid series and the reticulocyte count is low. The principal mechanism is deficiency of erythropoietin, a hormone produced in the kidney, but other factors may contribute to the anemia. In the presence of significant uremia, a component of hemolysis may also be present. Recombinant human erythropoietin (epoetin alfa) and iron correct the anemia, largely eliminating the need for transfusions.

3. Anemia of Hypothyroidism

Some patients with hypothyroidism develop significant anemia. Occasionally, anemia is detected before the diagnosis of the underlying disorder. A decreased growth velocity in an anemic child suggests hypothyroidism. The anemia is usually normocytic or macrocytic, but it is not megaloblastic and, hence not due to deficiencies of cobalamin or folate. Replacement therapy with thyroid hormone is usually effective in correcting the anemia.


The congenital hemolytic anemias are divided into three categories: defects of the red cell membrane; hemoglobinopathies; and disorders of red cell metabolism. Hereditary spherocytosis and elliptocytosis are the most common red cell membrane disorders. The diagnosis is suggested by the peripheral blood smear, which shows characteristic red cell morphology (eg, spherocytes, elliptocytes). These disorders usually have an autosomal dominant inheritance, and the diagnosis may be suggested by family history. The hemolysis is due to the deleterious effect of the membrane abnormality on red cell deformability. Decreased cell deformability leads to entrapment of the abnormally shaped red cells in the spleen. Many patients have splenomegaly, and splenectomy usually alleviates the hemolysis.

1. Hereditary Spherocytosis


image Anemia and jaundice.

image Splenomegaly.

image Positive family history of anemia, jaundice, or gallstones.

image Spherocytosis with increased reticulocytes.

image Increased osmotic fragility.

image Negative DAT.

image General Considerations

Hereditary spherocytosis is a relatively common inherited hemolytic anemia that occurs in all ethnic groups, but is most common in persons of northern European ancestry, in whom the incidence is about 1:5000. The disorder is marked by variable degrees of anemia, jaundice, and splenomegaly. In some persons, the disorder is mild and there is no anemia because erythroid hyperplasia fully compensates for hemolysis. In others, symptoms and transfusion dependence may be present prior to splenectomy. The hallmark of hereditary spherocytosis is the presence of microspherocytes in the peripheral blood. The disease is inherited in an autosomal dominant fashion in about 75% of cases; the remaining cases are thought to be autosomal recessive or due to de novo mutations.

Hereditary spherocytosis is secondary to alteration of genes encoding for spectrin, band 3, ankyrin, or protein 4.2 of the red cell membrane; spectrin abnormalities are more often diagnosed in childhood and band 3 in adulthood. The vertical linkages in the membrane are impaired so that spherocytes form. These are poorly deformable resulting in a shortened life span because they are trapped in the microcirculation of the spleen and engulfed by splenic macrophages. The specific membrane defect is of no major clinical implication.

image Clinical Findings

A. Symptoms and Signs

Hemolysis causes significant neonatal hyperbilirubinemia in 50% of affected children. Splenomegaly subsequently develops in the majority and is often present by age 5 years. Jaundice is variably present and in many patients may be noted only during infection. Patients with significant chronic anemia may complain of pallor, fatigue, or malaise. Intermittent exacerbations of the anemia are caused by increased hemolysis, splenic sequestration or by aplastic crises, and may be associated with severe weakness, fatigue, fever, abdominal pain, or even heart failure.

B. Laboratory Findings

Most patients have mild chronic hemolysis with hemoglobin levels of 9–12 g/dL. In some cases, the hemolysis is fully compensated and the hemoglobin level is in the normal range. Rare cases of severe disease require frequent transfusions. The anemia is usually normocytic and hyperchromic, and many patients have an elevated MCH concentration. The peripheral blood smear shows numerous microspherocytes and polychromasia. The reticulocyte count is elevated, often higher than might be expected for the degree of anemia. WBC and platelet counts are usually normal. The osmotic fragility is increased, particularly after incubation at 37°C for 24 hours. Serum bilirubin usually shows an elevation in the unconjugated fraction. The DAT is negative.

image Differential Diagnosis

Spherocytes are frequently present in persons with immune hemolysis. Thus, in the newborn, hereditary spherocytosis must be distinguished from hemolytic disease caused by ABO or other blood type incompatibilities. Older patients with autoimmune hemolytic anemia (AIHA) frequently present with jaundice, splenomegaly, and spherocytes on the peripheral blood smear. The DAT is positive in most cases of immune hemolysis and negative in hereditary spherocytosis. Occasionally, the diagnosis is confused in patients with splenomegaly from other causes, especially when hypersplenism increases red cell destruction and when some spherocytes are noted on the blood smear. In such cases, the true cause of the splenomegaly may be suggested by signs or symptoms of portal hypertension or by laboratory evidence of chronic liver disease. In contrast to children with hereditary spherocytosis, those with hypersplenism typically have some degree of thrombocytopenia or neutropenia.

image Complications

Severe jaundice may occur in the neonatal period and, if not controlled by phototherapy, may occasionally require exchange transfusion. Intermittent or persistent splenomegaly occurs in 10%–25% of patients and may require removal. Splenectomy is associated with an increased risk of overwhelming bacterial infections, particularly with pneumococci. Gallstones occur in 60%–70% of adults who have not undergone splenectomy and may form as early as age 5–10 years.

image Treatment

Supportive measures include the administration of folic acid to prevent the development of red cell hypoplasia due to folate deficiency. Acute exacerbations of anemia, due to increased rates of hemolysis or to aplastic crises caused by infection with human parvovirus, may be severe enough to require red cell transfusions. Splenectomy may be indicated depending on clinical severity and always results in significant improvement. This increases survival of the spherocytic red cells and leads to complete correction of the anemia in most cases. Except in unusually severe cases, the procedure should be postponed until the child is at least age 5 years because of the greater risk of postsplenectomy sepsis prior to this age. All patients scheduled for splenectomy should be immunized with pneumococcal, Haemophilus influenzae type b (Hib), and meningococcal vaccines prior to the procedure, and some clinicians recommend daily penicillin prophylaxis afterward. Asplenic patients with fever should be promptly evaluated for severe infection. Splenectomy prevents the subsequent development of cholelithiasis and eliminates the need for the activity restrictions. However, these benefits must be weighed against the risks of the surgical procedure and the subsequent lifelong risk of postsplenectomy sepsis.

image Prognosis

Splenectomy eliminates signs and symptoms in all but the most severe cases and reduces the risk of cholelithiasis. The abnormal red cell morphology and increased osmotic fragility persist without clinical consequence.

An X: Disorders of red cell membrane. Br J Haematol 2008;141:367 [PMID: 18341630].

2. Hereditary Elliptocytosis

Hereditary elliptocytosis is a heterogeneous disorder that ranges in severity from an asymptomatic state with almost normal red cell morphology to about 10% having moderate to severe hemolytic anemia. Most affected persons have numerous elliptocytes on the peripheral blood smear, but mild or no hemolysis. Those with hemolysis have an elevated reticulocyte count and may have jaundice and splenomegaly. This disorder is caused by weakened horizontal linkages in the red cell membrane skeleton due either to a defective spectrin dimer-dimer interaction or a defective spectrin-actin protein 4.1R junctional complex. Inheritance is autosomal dominant. Because most patients are asymptomatic, no treatment is indicated. Patients with significant degrees of hemolytic anemia may benefit from folate supplementation or splenectomy.

Some infants with hereditary elliptocytosis present in the neonatal period with moderate to marked hemolysis and significant hyperbilirubinemia. This disorder has been termed transient infantile pyknocytosis because such infants exhibit bizarre erythrocyte morphology with elliptocytes, budding red cells, and small misshapen cells that defy description. The MCV is low, and the anemia may be severe enough to require red cell transfusions. Typically, one parent has hereditary elliptocytosis, usually mild or asymptomatic. The infant’s hemolysis gradually abates during the first year of life, and the erythrocyte morphology subsequently becomes more typical of hereditary elliptocytosis.


The hemoglobinopathies are an extremely heterogeneous group of congenital disorders that may occur in all ethnic groups. The relatively high frequency of these genetic variants is related to the malaria protection afforded to heterozygous individuals. The hemoglobinopathies are generally classified into two major groups. The first, the thalassemias, are caused by quantitative deficiencies in the production of globin chains. These quantitative defects in globin synthesis cause microcytic and hypochromic anemias. The second group of hemoglobinopathies consists of those caused by structural abnormalities of globin chains. The most important of these, hemoglobins S, C, and E, are all the result of point mutations and single amino acid substitutions in β-globin. Many, but not all, infants with hemoglobinopathies are identified by routine neonatal screening.

Figure 30–3 shows the normal developmental changes that occur in globin-chain production during gestation and the first year of life. At birth, the predominant hemoglobin is fetal hemoglobin (hemoglobin F), which is composed of two α-globin chains and two γ-globin chains. Subsequently, the production of γ-globin decreases and the production of β-globin increases so that adult hemoglobin (two α chains and two β chains) predominates after 2–4 months. Because α-globin chains are present in both fetal and adult hemoglobin, disorders of α-globin synthesis (α-thalassemia) are clinically manifest in the newborn as well as later in life. In contrast, patients with β-globin disorders such as β-thalassemia and sickle cell disease are generally asymptomatic during the first 3–4 months of age and present clinically after γ-chain production and therefore fetal hemoglobin levels have decreased substantially.


image Figure 30–3. Changes in hemoglobin polypeptide chains during human development. (Reproduced, with permission, from Miller DR, Baehner RL: Blood Diseases of Infancy and Childhood, 6th ed. Mosby; 1989.)

1. α-Thalassemia


image Predominately African, Mediterranean, Middle Eastern, Chinese, or Southeast Asian ancestry.

image Microcytic, hypochromic anemia of variable severity.

image Generally Bart’s hemoglobin detected by neonatal screening.

image General Considerations

Most of the α-thalassemia syndromes are the result of deletions of one or more of the four α-globin genes on chromosome 16, although nondeletional α+ mutations also occur. Excess non–α-globin chains damage the red cell membrane, causing extravascular hemolysis. The variable severity of the α-thalassemia syndromes is related to the number of gene deletions (Table 30–1). The severity of the α-thalassemia syndromes varies among affected ethnic groups, depending on the genetic abnormalities prevalent in the population. In persons of African ancestry, α-thalassemia is usually caused by the deletion of only one of the two α-globin genes on each chromosome. Thus, in the African population, heterozygous individuals are silent carriers and homozygous individuals have α-thalassemia trait. In Asians, deletions of one or of both α-globin genes on the same chromosome are common; heterozygous individuals are either silent carriers or have α-thalassemia trait, and homozygous individuals or compound heterozygous individuals have α-thalassemia trait, hemoglobin H disease, or hydrops fetalis. Thus, the presence of α-thalassemia in a child of Asian ancestry may have important implications for genetic counseling, whereas this is not usually the case in families of African ancestry.

Table 30–1 The α-thalassemias.


image Clinical Findings

The clinical findings depend on the number of α-globin genes deleted. Table 30–1 summarizes the α-thalassemia syndromes.

Persons with three α-globin genes (one-gene deletion) are asymptomatic and have no hematologic abnormalities. Hemoglobin electrophoresis in the neonatal period shows 0%–3% Bart’s hemoglobin, which is a variant hemoglobin composed of four γ-globin chains. Hemoglobin electrophoresis after the first few months of life is normal. Thus, this condition is usually suspected only in the context of family studies or when a small amount of Bart’s hemoglobin is detected by neonatal screening for hemoglobinopathies.

Persons with two α-globin genes (two-gene deletion) are typically asymptomatic. The MCV is usually less than 100 fL at birth. Hematologic studies in older infants and children show a normal or slightly decreased hemoglobin level with a low MCV and a slightly hypochromic blood smear with some target cells.

Persons with one α-globin gene (three-gene deletion) have a mild to moderately severe microcytic hemolytic anemia (hemoglobin level of 7–10 g/dL), which may be accompanied by hepatosplenomegaly and some bony abnormalities caused by the expanded medullary space. The reticulocyte count is elevated, and the red cells show marked hypochromia and microcytosis with significant poikilocytosis and some basophilic stippling. Incubation of red cells with brilliant cresyl blue (hemoglobin H preparation) shows inclusion bodies formed by denatured hemoglobin H.

The deletion of all four α-globin genes causes severe intrauterine anemia and results in hydrops fetalis and fetal demise or neonatal death shortly after delivery. Extreme pallor and massive hepatosplenomegaly are present. Hemoglobin electrophoresis reveals a predominance of Bart’s hemoglobin with a complete absence of normal fetal or adult hemoglobin.

image Differential Diagnosis

α-Thalassemia trait (two-gene deletion) must be differentiated from other mild microcytic anemias, including ID and β-thalassemia minor (see the next section). In contrast to children with ID, children with α-thalassemia trait show normal or increased levels of ferritin and serum iron. In contrast to children with β-thalassemia minor, children with α-thalassemia trait have a normal hemoglobin electrophoresis after age 4–6 months. Finally, the history of a low MCV (96 fL) at birth or the presence of Bart’s hemoglobin on the neonatal hemoglobinopathy screening test suggests α-thalassemia.

Children with hemoglobin H disease may have jaundice and splenomegaly, and the disorder must be differentiated from other hemolytic anemias. The key to the diagnosis is the decreased MCV and the marked hypochromia on the blood smear. With the exception of β-thalassemia, most other significant hemolytic disorders have a normal or elevated MCV and the RBCs are not hypochromic. Infants with hydrops fetalis due to severe α-thalassemia must be distinguished from those with hydrops due to other causes of anemia, such as alloimmunization.

image Complications

The principal complication of α-thalassemia trait is the needless administration of iron, given in the belief that a mild microcytic anemia is due to ID. Persons with hemoglobin H disease may have intermittent exacerbations of their anemia in response to oxidant stress or infection, which occasionally require blood transfusions. Splenomegaly may exacerbate the anemia and may require splenectomy. Women pregnant with hydropic α-thalassemia fetuses are subject to increased complications of pregnancy, particularly toxemia and postpartum hemorrhage.

image Treatment

Persons with α-thalassemia trait require no treatment. Those with hemoglobin H disease should receive supplemental folic acid and avoid the same oxidant drugs that cause hemolysis in persons with G6PD deficiency, because exposure to these drugs may exacerbate their anemia. The anemia may also be exacerbated during periods of infection, and transfusions may be required. Hypersplenism may develop later in childhood and require surgical splenectomy. Genetic counseling and prenatal diagnosis should be offered to families at risk for hydropic fetuses.

Cunningham MJ: Update on thalassemia: clinical care and complications. Hematol Oncol Clin North Am 2010;24:215 [PMID: 20113904].

Harteveld CL: α-thalassemia. Orphanet J Rare Dis 2010;5:13 [PMID: 20507641].

Lal A: Heterogeneity of hemoglobin H disease in childhood. N Engl J Med 2011;364:710 [PMID: 21345100].

2. β-Thalassemia


β-Thalassemia minor:

image Normal neonatal screening test.

image Predominantly African, Mediterranean, Middle Eastern, or Asian ancestry.

image Mild microcytic, hypochromic anemia.

image No response to iron therapy.

image Elevated level of hemoglobin A2.

β-Thalassemia intermedia:

image Ancestry as above.

image Microcytic, hypochromic anemia that usually becomes symptomatic after the first few years of life with hepatosplenomegaly.

β-Thalassemia major:

image Neonatal screening shows hemoglobin F only.

image Mediterranean, Middle Eastern, or Asian ancestry.

image Severe microcytic, hypochromic anemia with marked hepatosplenomegaly.

image General Considerations

In contrast to the four α-globin genes, only two β-globin genes are present in diploid cells, one on each chromosome 11. Excess non–β-globin chains damage the red cells, causing extravascular hemolysis. Individuals heterozygous for one β-thalassemia gene generally have β-thalassemia minor. Homozygous individuals generally have β-thalassemia major (Cooley anemia), a severe transfusion-dependent anemia. Individuals with β-thalassemia intermedia, which is more severe than thalassemia minor but is not generally transfusion-dependent, are usually heterozygous for two β+ mutations (produce reduced quantities of normal A hemoglobin) or have both β0 (production of no normal A hemoglobin) and a β+ mutation, or have an alternative mutation such as E in conjunction with a severe β+ or β0 mutation. Thalassemia major is the most common worldwide cause of transfusion-dependent anemia in childhood. In addition, β-thalassemia genes interact with genes for structural β-globin variants, such as hemoglobin S and hemoglobin E to cause serious disease in compound heterozygous individuals. These disorders are discussed further in the sections dealing with sickle cell disease and with hemoglobin E disorders.

image Clinical Findings

A. Symptoms and Signs

Persons with β-thalassemia minor are usually asymptomatic with a normal physical examination. The time to presentation for those with β-thalassemia intermedia is variable. Those with β-thalassemia major are normal at birth but develop significant anemia during the first year of life. If the disorder is not identified and treated with blood transfusions, such children grow poorly and develop massive hepatosplenomegaly and enlargement of the medullary space with thinning of the bony cortex. The skeletal changes (due to ineffective erythropoiesis) cause characteristic facial deformities (prominent forehead and maxilla) and predispose the child to pathologic fractures.

B. Laboratory Findings

Children with β-thalassemia minor have normal neonatal screening results but subsequently develop a decreased MCV with or without mild anemia. The peripheral blood smear typically shows hypochromia, target cells, and sometimes basophilic stippling. Hemoglobin electrophoresis performed after 6–12 months of age is usually diagnostic when levels of hemoglobin A2, hemoglobin F, or both are elevated. β-Thalassemia major is often initially suspected when hemoglobin A is absent on neonatal screening. Such infants are hematologically normal at birth, but develop severe anemia after the first few months of life. The peripheral blood smear typically shows a severe hypochromic, microcytic anemia with marked anisocytosis and poikilocytosis. Target cells are prominent, and nucleated RBCs often exceed the number of circulating WBCs. The hemoglobin level usually falls to 5–6 g/dL or less, and the reticulocyte count is elevated, but the reticulocyte index is normal to decreased. Platelet and WBC counts may be increased, and the serum indirect bilirubin level is elevated. The bone marrow shows marked erythroid hyperplasia, but this finding is not needed for diagnosis. Hemoglobin electrophoresis shows only fetal hemoglobin and hemoglobin A2 in children with homozygous β0-thalassemia.

Those with β+-thalassemia genes make a variable amount of hemoglobin A, depending on the mutation, but have a marked increase in fetal hemoglobin and hemoglobin A2 levels. The diagnosis of homozygous β-thalassemia, intermedia or minor, may also be supported by DNA testing.

image Differential Diagnosis

β-Thalassemia minor must be differentiated from other causes of mild microcytic, hypochromic anemias, principally ID and α-thalassemia. In contrast to patients with IDA, those with β-thalassemia minor typically have an elevated number of RBCs, and the index of the MCV divided by the red cell count is under 13. Generally, the finding of an elevated hemoglobin A2 level is diagnostic; however, the A2 level is lowered by coexistent ID. Thus, in children thought to be iron-deficient, hemoglobin electrophoresis with quantitation of hemoglobin A2 is sometimes deferred until after a course of iron therapy.

β-Thalassemia major is rarely confused with other disorders. Hemoglobin electrophoresis and family studies readily distinguish it from hemoglobin E/β-thalassemia, which is the other increasingly important cause of transfusion-dependent thalassemia.

image Complications

The principal complication of β-thalassemia minor is the unnecessary use of iron therapy in a futile attempt to correct the microcytic anemia. Children with β-thalassemia major who are inadequately transfused experience poor growth and recurrent infections and may have hepatosplenomegaly, thinning of the cortical bone, and pathologic fractures. Without treatment, most children die within the first decade of life. The principal complications of β-thalassemia major in transfused children are hemosiderosis, splenomegaly, and hypersplenism. Transfusion-related hemosiderosis requires chelation therapy to prevent cardiac, hepatic, and endocrine dysfunction. Noncompliance with chelation in adolescents and young adults may lead to death from congestive heart failure, cardiac arrhythmias, or hepatic failure. Even with adequate transfusions, many patients develop splenomegaly and some degree of hypersplenism. This may require surgical splenectomy because of the increasing transfusion requirements, but the procedure increases the risk of thrombosis, pulmonary hypertension, and overwhelming septicemia.

image Treatment

β-Thalassemia minor requires no specific therapy, but the diagnosis may have important genetic implications for the family. The time to presentation for those with β-thalassemia intermedia is variable. For β-thalassemia major, two treatments are available: chronic transfusion with iron chelation and stem cell transplant. Programs of blood transfusion are generally targeted to maintain a nadir hemoglobin level of 9–10 g/dL. This approach gives increased vigor and well-being, improved growth, and fewer overall complications. However, maintenance of good health requires iron chelation. Small doses of supplemental ascorbic acid may enhance the efficacy of iron chelation. Patients who undergo splenectomy to reduce transfusion requirements, and hence iron loading, should receive pneumococcal vaccine prior to the procedure and prophylactic penicillin and urgent treatment of all febrile illness after splenectomy. Chronic transfusion therapy is infrequently indicated for individuals with β-thalassemia intermedia.

Bone marrow or umbilical cord blood transplant is an important therapeutic option for children with β-thalassemia major who have an HLA-identical sibling donor. The probability of hematologic cure is greater than 90% when transplant is performed prior to the development of hepatomegaly or portal fibrosis.

Angelucci E: Hematopoietic stem cell transplantation in thalassemia. Hematology Am Soc Hematol Educ Program 2010;2010:456 [PMID: 21239835].

Taher AT: Optimal management of β thalassemia intermedia. Br J Haematol 2011;152:512 [PMID: 21250971].

3. Sickle Cell Disease


image Neonatal screening test usually with hemoglobins FS, FSC, or FSA (S > A).

image Predominantly African, Mediterranean, Middle Eastern, Indian, or Caribbean ancestry.

image Anemia, elevated reticulocyte count, usually jaundice.

image Recurrent episodes of musculoskeletal or abdominal pain.

image Often hepatomegaly and splenomegaly that resolves.

image Increased risk of bacterial sepsis.

image General Considerations

Sickle cell disease encompasses a family of disorders with manifestations secondary to the propensity of deoxygenated sickle hemoglobin (S) to polymerize. Polymerization of sickle hemoglobin distorts erythrocyte morphology; decreases red cell deformability; causes a marked reduction in red cell life span; increases blood viscosity; and predisposes to inflammation, coagulation activation, and episodes of vaso-occlusion. Sickle cell anemia, the most severe sickling disorder, is caused by homozygosity for the sickle gene and is the most common form of sickle cell disease. Other clinically important sickling disorders are compound heterozygous conditions in which the sickle gene interacts with genes for hemoglobins C, DPunjab, OArab, CHarlem, or β-thalassemia.

Overall, sickle cell disease occurs in about 1 of every 400 African-American infants. Eight percent of African Americans are heterozygous carriers of the sickle gene and are said to have sickle cell trait.

image Clinical Findings

A. Symptoms and Signs

These are related to the hemolytic anemia, tissue ischemia, and organ dysfunction caused by vaso-occlusion. They are most severe in children with sickle cell anemia or sickle β0-thalassemia. Physical findings are normal at birth, and symptoms are unusual before age 3–4 months because high levels of fetal hemoglobin inhibit sickling. A moderately severe hemolytic anemia may be present by age 1 year. This causes pallor, fatigue, and jaundice, and predisposes to the development of gallstones during childhood and adolescence. Intense congestion of the spleen with sickled cells may cause splenomegaly in early childhood and results in functional asplenia as early as age 3 months in sickle cell anemia. This places children at great risk for overwhelming infection with encapsulated bacteria, particularly pneumococci. Up to 30% of patients experience one or more episodes of acute splenic sequestration, characterized by sudden enlargement of the spleen with pooling of red cells, acute exacerbation of anemia, and, in severe cases, shock and death. Acute exacerbation of anemia also occurs with aplastic crises, usually caused by infection with human parvovirus and other viruses.

Recurrent episodes of vaso-occlusion and tissue ischemia cause acute and chronic problems. Dactylitis, or hand-and-foot syndrome, is the most common initial symptom of the disease and occurs in up to 50% of children with sickle cell anemia before age 3 years. Recurrent episodes of abdominal and musculoskeletal pain may occur throughout life. Overt strokes occur in about 11% of children with sickle cell anemia and tend to be recurrent. Recurrence is significantly reduced with chronic red cell transfusions. The so-called acute chest syndrome, characterized by fever, pleuritic chest pain, and acute pulmonary infiltrates with hypoxemia, is caused by pulmonary infection, infarction, or fat embolism from ischemic bone marrow. All tissues are susceptible to damage from vaso-occlusion, and multiple organ dysfunction is common by adulthood in those with sickle cell anemia or sickle β0-thalassemia. The common manifestations of sickle cell disease are listed in Table 30–2. Manifestations generally develop less frequently in those with SC and S β+-thalassemia.

Table 30–2. Common clinical manifestations of sickle cell disease.


B. Laboratory Findings

Children with sickle cell anemia generally show a baseline hemoglobin level of 7–10 g/dL. This value may fall to life-threatening levels at the time of a splenic sequestration or aplastic crisis; often this occurs in association with parvovirus B19 infection. The baseline reticulocyte count is elevated markedly. The anemia is usually normocytic or macrocytic, and the peripheral blood smear typically shows the characteristic sickle cells as well as numerous target cells. Patients with sickle β-thalassemia also generally have a low MCV and hypochromia. Those with sickle β+-thalassemia tend to have less hemolysis and anemia. Persons with sickle hemoglobin C disease have fewer sickle forms and more target cells, and the hemoglobin level may be normal or only slightly decreased because the rate of hemolysis is much less than in sickle cell anemia.

Most infants with sickle hemoglobinopathies born in the United States are identified by neonatal screening. Results indicative of possible sickle cell disease require prompt confirmation with hemoglobin electrophoresis. Children with sickle cell anemia and with sickle β0-thalassemia have only hemoglobins S, F, and A2. Persons with sickle β+-thalassemia have a preponderance of hemoglobin S with a lesser amount of hemoglobin A and elevated A2. Persons with sickle hemoglobin C disease have equal amounts of hemoglobins S and C. The use of solubility tests to screen for the presence of sickle hemoglobin should be avoided because a negative result is frequently encountered in infants with sickle cell disease, and because a positive result in an older child does not differentiate sickle cell trait from sickle cell disease. Thus, hemoglobin electrophoresis is always necessary to accurately identify a sickle disorder. Solubility tests also will not identify hemoglobin variants other than S.

image Differential Diagnosis

Hemoglobin electrophoresis and sometimes hematologic studies of the parents are usually sufficient to confirm sickle cell disease, although DNA testing is available. It is critical to determine whether the child with only F and S hemoglobins on newborn screening has sickle cell anemia, sickle β0 thalassemia, or is a compound heterozygote for sickle hemoglobin and pancellular hereditary persistence of fetal hemoglobin. Such children, when older, typically have 30% fetal hemoglobin and 70% hemoglobin S, and are well.

image Complications

Repeated tissue ischemia and infarction causes damage to virtually every organ system. Table 30–2 lists the most important complications. Patients who require multiple transfusions are at risk of developing transfusion-related hemosiderosis and infections as well as red cell alloantibodies.

image Treatment

The cornerstone of treatment is enrollment in a program involving patient and family education, comprehensive outpatient care, and appropriate treatment of acute complications. Important to the success of such a program are psychosocial services, blood bank services, and the ready availability of baseline patient information in the setting in which acute illnesses are evaluated and treated.

Management of sickle cell anemia and sickle β0-thalassemia includes daily prophylactic penicillin, which should be initiated by age 2 months and continued at least until age 5 years. The routine use of penicillin prophylaxis in sickle hemoglobin C disease and sickle β+-thalassemia is controversial. Pneumococcal conjugate and polysaccharide vaccines should be administered to all children who have sickle cell disease. Other routine immunizations, including yearly vaccination against influenza, should be provided. All illnesses associated with fever greater than 38.5°C should be evaluated promptly, bacterial cultures performed, parenteral broad-spectrum antibiotics administered, and careful inpatient or outpatient observation conducted.

Treatment of painful vaso-occlusive episodes includes the maintenance of adequate hydration (with avoidance of overhydration), correction of acidosis if present, administration of adequate analgesia, maintenance of normal oxygen saturation, and the treatment of any associated infections.

Red cell transfusions play an important role in management. Transfusions are indicated to improve oxygen-carrying capacity during acute severe exacerbations of anemia, as occurs during episodes of splenic sequestration or aplastic crisis. Red cell transfusions are not indicated for the treatment of chronic steady-state anemia, which is usually well tolerated, or for uncomplicated episodes of vaso-occlusive pain. Simple or partial exchange transfusion to reduce the percentage of circulating sickle cells is indicated for some severe acute vaso-occlusive events and may be lifesaving. These events include stroke, moderate to severe acute chest syndrome, and acute life-threatening failure of other organs. Transfusions may be administered prior to high-risk procedures such as surgery with general anesthesia and arteriograms with high ionic contrast materials. Some patients with severe complications may benefit from chronic transfusion therapy. The most common indications for transfusions are stroke or an abnormal transcranial Doppler assessment indicating an increased risk for stroke. Extended matching for red cell antigens reduces the incidence of alloimmunization.

Successful stem cell transplant cures sickle cell disease, but to date its use has been limited because of the risks associated with the procedure, the inability to predict in young children the severity of future complications, and the paucity of HLA-identical sibling donors. Daily administration of oral hydroxyurea increases levels of fetal hemoglobin, decreases hemolysis, and reduces episodes of pain and dactylitis in young children with sickle cell anemia, as well as some evidence for a reduction in acute chest syndrome, hospitalization rates, and transfusions. The hematologic effects and short-term toxicity of hydroxyurea in children are similar to those in adults. Thus, hydroxyurea is being increasingly prescribed for children and adolescents with sickle cell anemia and sickle β0-thalassemia; efficacy in SC and β+-thalassemia has not been formally studied.

image Prognosis

Early identification by neonatal screening of infants with sickle cell disease, combined with comprehensive care that includes prescription of prophylactic penicillin, instruction on splenic palpation, and education on the need to urgently seek care when fever occurs, has markedly reduced mortality in childhood. Most patients now live well into adulthood, but eventually succumb to complications.

Lasalle-Williams M: Extended red blood cell antigen matching for transfusions in sickle cell disease: a review of a 14-year experience from a single center. Transfusion 2011 Feb 18 [Epub ahead of print] [PMID: 21332724].

Strouse JJ, Heeney MM: Hydroxyurea for the treatment of sickle cell disease: efficacy, barriers, toxicity, and management in children. Pediatr Blood Cancer 2012;59:365–371 [PMID: 22517797].

4. Sickle Cell Trait

Individuals who are heterozygous for the sickle gene have sickle cell trait. Neonates are identified by neonatal screening that shows hemoglobin FAS (A > S). Accurate identification of older persons with sickle cell trait depends on hemoglobin electrophoresis, which typically shows about 60% hemoglobin A and 40% hemoglobin S. No anemia or hemolysis is present, and the physical examination is normal. Persons with sickle cell trait are generally healthy with normal life expectancy.

However, sickle trait erythrocytes are capable of sickling, with acidemia and hypoxemia. Thus, the kidney may be affected with the most common manifestation of sickle trait being hyposthenuria. Painless hematuria, usually microscopic, affects about 4% of those with sickle trait and does not progress to significant renal dysfunction. Fewer than 40 individuals have been reported with an exceedingly rare malignancy, renal medullary carcinoma, and all but one have had sickle trait. The incidence of bacteriuria and pyelonephritis may be increased during pregnancy, but overall rates of maternal and infant morbidity and mortality are not affected by sickle cell trait.

Exertion at moderate altitudes rarely precipitates splenic infarction. Whether or not the risk of sudden unexplained death during strenuous exercise, as occurs during military basic training, is increased in men with sickle cell trait is controversial. In general, exercise tolerance seems to be normal; the incidence of sickle cell trait in black professional football players is similar to that of the general African-American population.

There is no reason to restrict strenuous activity for individuals with sickle cell trait. As is true for all individuals performing strenuous activity, it is important to be conditioned, dress appropriately, have access to fluids, rest periodically, and perform moderate activity in extreme heat and humidity. Sickle cell trait is most significant for its genetic implications.

5. Hemoglobin C Disorders

Hemoglobin C is detected by neonatal screening. Two percent of African Americans are heterozygous for hemoglobin C and are said to have hemoglobin C trait. Such individuals have no symptoms, anemia, or hemolysis, but the peripheral blood smear may show some target cells. Identification of persons with hemoglobin C trait is important for genetic counseling, particularly with regard to the possibility of sickle hemoglobin C disease in offspring.

Persons with homozygous hemoglobin C have a mild microcytic hemolytic anemia and may develop splenomegaly. The peripheral blood smear shows prominent target cells. As with other hemolytic anemias, complications of homozygous hemoglobin C include gallstones and aplastic crises.

6. Hemoglobin E Disorders

Hemoglobin E is the second most common hemoglobin variant worldwide, with a gene frequency up to 60% in northeast Thailand and Cambodia. Persons heterozygous for hemoglobin E show hemoglobin FAE by neonatal screening and are asymptomatic and usually not anemic, but they may develop mild microcytosis. Persons homozygous for hemoglobin E are also asymptomatic but may have mild anemia; the peripheral blood smear shows microcytosis and some target cells.

Compound heterozygotes for hemoglobin E and β0-thalassemia are normal at birth and, like infants with homozygous E, show hemoglobin FE on neonatal screening. Unlike homozygotes, they subsequently develop mild to severe microcytic hypochromic anemia. Such children may exhibit jaundice, hepatosplenomegaly, and poor growth if the disorder is not recognized and treated appropriately. In some cases, the anemia becomes severe enough to require lifelong transfusion therapy. Even without regular transfusions, hemosiderosis may occur. In certain areas of the United States, hemoglobin E/β0-thalassemia has become a more common cause of transfusion-dependent anemia than homozygous β-thalassemia.

7. Other Hemoglobinopathies

At least 500 human globin-chain variants have been described. Some, such as hemoglobins D and G, are relatively common. Heterozygous individuals, who are frequently identified during the course of neonatal screening programs for hemoglobinopathies, are generally asymptomatic and usually have no anemia or hemolysis. The principal significance of most hemoglobin variants is the potential for disease in compound heterozygous individuals who also inherit a gene for β-thalassemia or sickle hemoglobin. For example, children who are compound heterozygous for hemoglobins S and DPunjab (DLos Angeles) have sickle cell disease.


Erythrocytes depend on the anaerobic metabolism of glucose for the maintenance of adenosine triphosphate levels sufficient for homeostasis. Glycolysis also produces the 2,3-diphosphoglycerate (2,3-DPG) levels needed to modulate the oxygen affinity of hemoglobin. Glucose metabolism via the hexose monophosphate shunt is necessary to generate sufficient reduced nicotinamide adenine dinucleotide phosphate (NADPH) and reduced glutathione to protect red cells against oxidant damage. Congenital deficiencies of many glycolytic pathway enzymes have been associated with hemolytic anemias. In general, the morphologic abnormalities present on the peripheral blood smear are nonspecific, and the inheritance of these disorders is autosomal recessive or X-linked. Thus, the possibility of a red cell enzyme defect should be considered during the evaluation of a patient with a congenital hemolytic anemia in the following instances: when the peripheral blood smear does not show red cell morphology typical of membrane or hemoglobin defects (eg, spherocytes, sickle forms, target cells); when hemoglobin disorders are excluded by hemoglobin electrophoresis and by isopropanol precipitation tests; and when family studies do not suggest an autosomal dominant disorder. The diagnosis is confirmed by finding a low level of the deficient enzyme.

The two most common disorders of erythrocyte metabolism are G6PD deficiency and pyruvate kinase deficiency.

1. Glucose-6-Phosphate Dehydrogenase Deficiency


image African, Mediterranean, or Asian ancestry.

image Neonatal hyperbilirubinemia.

image Sporadic hemolysis associated with infection or with ingestion of oxidant drugs or fava beans.

image X-linked inheritance.

image General Considerations

Deficiency of glucose-6-phosphate dehydrogenase (G6PD) is the most common red cell enzyme defect that causes hemolytic anemia. The disorder has an X-linked recessive inheritance and occurs with high frequency among persons of African, Mediterranean, and Asian ancestry. Hundreds of different G6PD variants have been characterized. In most instances, the deficiency is due to enzyme instability; thus, older red cells are more deficient than younger ones and are unable to generate sufficient nicotinamide adenine dinucleotide (NADH) to maintain the levels of reduced glutathione necessary to protect the red cells against oxidant stress. Thus, most persons with G6PD deficiency do not have a chronic hemolytic anemia; instead, they have episodic hemolysis at times of exposure to the oxidant stress of infection or of certain drugs or food substances. The severity of the disorder varies among ethnic groups; G6PD deficiency in persons of African ancestry usually is less severe than in other ethnic groups.

image Clinical Findings

A. Symptoms and Signs

Infants with G6PD deficiency may have significant hyperbilirubinemia and require phototherapy or exchange transfusion to prevent kernicterus. The deficiency is an important cause of neonatal hyperbilirubinemia in infants of Mediterranean or Asian ancestry, but less so in infants of African ancestry. Older children with G6PD deficiency are asymptomatic and appear normal between episodes of hemolysis. Hemolytic episodes are often triggered by infection or by the ingestion of oxidant drugs such as antimalarial compounds and sulfonamide antibiotics (Table 30–3). Ingestion of fava beans may trigger hemolysis in children of Mediterranean or Asian ancestry but usually not in children of African ancestry. Episodes of hemolysis are associated with pallor, jaundice, hemoglobinuria, and sometimes cardiovascular compromise.

Table 30–3. Some common drugs and chemicals that can induce hemolytic anemia in persons with G6PD deficiency.


B. Laboratory Findings

The hemoglobin concentration, reticulocyte count, and peripheral blood smear are usually normal in the absence of oxidant stress. Episodes of hemolysis are associated with a variable fall in hemoglobin. “Bite” cells or blister cells may be seen, along with a few spherocytes. Hemoglobinuria is common, and the reticulocyte count increases within a few days. Heinz bodies may be demonstrated with appropriate stains. The diagnosis is confirmed by the finding of reduced levels of G6PD in erythrocytes. Because this enzyme is present in increased quantities in reticulocytes, the test is best performed at a time when the reticulocyte count is normal or near normal.

image Complications

Kernicterus is a risk for infants with significant neonatal hyperbilirubinemia. Episodes of acute hemolysis in older children may be life-threatening. Rare G6PD variants are associated with chronic hemolytic anemia; the clinical course of patients with such variants may be complicated by splenomegaly and by the formation of gallstones.

image Treatment

The most important treatment issue is avoidance of drugs known to be associated with hemolysis (see Table 30–3). For some patients of Mediterranean, Middle Eastern, or Asian ancestry, the consumption of fava beans must also be avoided. Infections should be treated promptly and antibiotics given when appropriate. Most episodes of hemolysis are self-limiting, but red cell transfusions may be lifesaving when signs and symptoms indicate cardiovascular compromise.

2. Pyruvate Kinase Deficiency

Pyruvate kinase deficiency is an autosomal recessive disorder observed in all ethnic groups but is most common in northern Europeans. The deficiency is associated with a chronic hemolytic anemia of varying severity. Approximately one-third of those affected present in the neonatal period with jaundice and hemolysis that require phototherapy or exchange transfusion. Occasionally, the disorder causes hydrops fetalis and neonatal death. In older children, the hemolysis may require red cell transfusions or be mild enough to go unnoticed for many years. Jaundice and splenomegaly frequently occur in the more severe cases. The diagnosis of pyruvate kinase deficiency is occasionally suggested by the presence of echinocytes on the peripheral blood smear, but these findings may be absent prior to splenectomy. The diagnosis depends on the demonstration of low levels of pyruvate kinase activity in red cells.

Treatment of pyruvate kinase depends on the severity of the hemolysis. Blood transfusions may be required for significant anemia, and splenectomy may be beneficial. Although the procedure does not cure the disorder, it ameliorates the anemia and its symptoms. Characteristically, the reticulocyte count increases and echinocytes become more prevalent after splenectomy, despite the decreased hemolysis and increased hemoglobin level.


1. Autoimmune Hemolytic Anemia


image Pallor, fatigue, jaundice, and dark urine.

image Splenomegaly.

image Positive DAT.

image Reticulocytosis and spherocytosis.

image General Considerations

Acquired autoimmune hemolytic anemia (AIHA) is rare during the first 4 months of life but is one of the more common causes of acute anemia after the first year. It may arise as an isolated problem or may complicate an infection (hepatitis, upper respiratory tract infections, EBV mononucleosis, or cytomegalovirus [CMV] infection); systemic lupus erythematosus and other autoimmune syndromes; immunodeficiency states, including autoimmune lymphoproliferative syndrome (ALPS); or, very rarely, malignancies. Drugs may induce antibody-associated hemolytic anemia, and recently third-generation cephalosporins have become a more common cause for this adverse event of antibiotic therapy.

image Clinical Findings

A. Symptoms and Signs

The disease usually has an acute onset manifested by weakness, pallor, dark urine, and fatigue. Jaundice is a prominent finding, and splenomegaly is often present. Some cases have a more chronic, insidious onset. Clinical evidence of an underlying disease may be present.

B. Laboratory Findings

The anemia is normochromic and normocytic and may vary from mild to severe (hemoglobin concentration < 5 g/dL). The reticulocyte count and index are usually increased but occasionally are normal or low. Spherocytes and nucleated red cells may be seen on the peripheral blood smear. Although leukocytosis and elevated platelet counts are a common finding, thrombocytopenia occasionally occurs. Other laboratory data consistent with hemolysis are present, such as increased indirect and total bilirubin, lactic dehydrogenase, aspartate aminotransferase, and urinary urobilinogen. Intravascular hemolysis is indicated by hemoglobinemia or hemoglobinuria. Examination of bone marrow shows marked erythroid hyperplasia and hemophagocytosis, but is seldom required for the diagnosis.

Serologic studies are helpful in defining pathophysiology, planning therapeutic strategies, and assessing prognosis (Table 30–4). In almost all cases, the direct and indirect antiglobulin (DAT and IAT) tests are positive. Further evaluation allows distinction into one of three syndromes. The presence of IgG and no or low level of complement activation on the patient’s RBCs, maximal in vitro antibody activity at 37°C, and either no antigen specificity or an Rh-like specificity constitute warm AIHA with extravascular destruction by the reticuloendothelial system. In contrast, the detection of complement alone on RBCs, optimal reactivity in vitro at 4°C, and I or i antigen specificity are diagnostic of cold AIHA with intravascular hemolysis. Although cold agglutinins are relatively common (∼10%) in normal individuals, clinically significant cold antibodies exhibit in vitro reactivity at 30°C or above. Warm reactive IgM antibodies are rare.

Table 30–4. Classification of autoimmune hemolytic anemia (AIHA) in children.


Paroxysmal cold hemoglobinuria presents a different category of disease. The laboratory evaluation is identical to cold AIHA except for antigen specificity (P) and the exhibition of an in vitro hemolysin. Paroxysmal cold hemoglobinuria is almost always associated with significant infections, such as Mycoplasma, EBV, and CMV.

image Differential Diagnosis

AIHA must be differentiated from other forms of congenital or acquired hemolytic anemias. The DAT discriminates antibody-mediated hemolysis from other causes, such as hereditary spherocytosis. The presence of other cytopenias and antibodies to platelets or neutrophils suggests an autoimmune (eg, lupus) syndrome, immunodeficiency (eg, ALPS, congenital immunodeficiency), or Evans syndrome (AIHA and other cytopenias associated with autoantibodies). Up to one-third of patients diagnosed as Evans syndrome may have ALPS.

image Complications

The anemia may be very severe and result in cardiovascular collapse, requiring emergency management. The complications of an underlying disease, such as disseminated lupus erythematosus or an immunodeficiency state, may be present.

image Treatment

Medical management of the underlying disease is important in symptomatic cases. Defining the clinical syndrome provides a useful guide to treatment. Most patients with warm AIHA (in which hemolysis is mostly extravascular) respond to prednisone (2 mg/kg/d). After the initial treatment, the dose of corticosteroids may be decreased slowly. Patients may respond to 1 g of intravenous immune globulin (IVIG) per kilogram per day for 2 days, but fewer patients respond to IVIG than to prednisone. Although the rate of remission with splenectomy may be as high as 50%, particularly in warm AIHA, this should be carefully considered in younger patients and withheld until other treatments have failed. In severe cases, unresponsive to more conventional therapy, immunosuppressive agents such as mycophenolate, sirolimus, cyclosporine, cyclophosphamide, azathioprine, or busulfan may be tried alone or in combination with corticosteroids. The initial three may induce less myelosuppression and be helpful when hemolysis is associated with Evans syndrome or ALPS. In severe cases, rituximab may be a successful alternative; however, this drug should be avoided in AIHA associated with ALPS. Transplantation has been used in a small number of cases.

Patients with cold AIHA and paroxysmal cold hemoglobinuria are less likely to respond to corticosteroids or IVIG. Because these syndromes are most apt to be associated with infections and have an acute, self-limited course, supportive care alone may be sufficient. Plasma exchange may be effective in severe cold autoimmune (IgM) hemolytic anemia because the offending antibody has an intravascular distribution.

Supportive therapy is crucial. Patients with cold-reacting antibodies, particularly paroxysmal cold hemoglobinuria, should be kept in a warm environment. Transfusion may be necessary because of the complications of severe anemia but should be used only when there is no alternative. In most patients, cross-match compatible blood will not be found, and the least incompatible unit may be identified. Transfusion must be conducted carefully, beginning with a test dose (see Transfusion Medicine section, later in this chapter). Identification of the patient’s phenotype for minor red cell alloantigens may be helpful in avoiding alloimmunization or in providing appropriate transfusions if alloantibodies arise after initial transfusions. Patients with severe intravascular hemolysis may have associated disseminated intravascular coagulation (DIC), and heparin therapy should be considered in such cases.

image Prognosis

The outlook for AIHA in childhood usually is good unless associated diseases are present (eg, congenital immunodeficiency, ALPS, AIDS, lupus erythematosus), in which case hemolysis is likely to have a chronic course. In general, children with warm (IgG) AIHA are at greater risk for more severe and chronic disease with higher morbidity and mortality rates. Hemolysis and positive antiglobulin tests may continue for months or years. Patients with cold AIHA or paroxysmal cold hemoglobinuria are more likely to have acute, self-limited disease (< 3 months). Paroxysmal cold hemoglobinuria is almost always associated with infection (eg, Mycoplasma infection, CMV, and EBV).

Aladjidi N: New insights into childhood autoimmune hemolytic anemia: a French national observational study of 265 children. Haematologica 2011;96:655 [PMID: 21228033].

Arndt PA: Serologic characteristics of ceftriaxone antibodies in 25 patients with drug-induced anemia. Transfusion 2012;52:602 [PMID: 21880048].

Petz LD: Unusual problems regarding autoimmune hemolytic anemias. In: Petz LD, Garratty G (eds): Acquired Immune Hemolytic Anemias, 2nd ed. Churchill Livingstone; 2004:341–344.

Rao VK: Use of rituximab for refractory cytopenias associated with autoimmune lymphoproliferative syndrome (ALPS). Pediatr Blood Cancer 2009;52:847 [PMID: 19214977].

2. Nonimmune Acquired Hemolytic Anemia

Hepatic disease may alter the lipid composition of the red cell membrane. This usually results in the formation of target cells and is not associated with significant hemolysis.

Occasionally, hepatocellular damage is associated with the formation of spur cells and brisk hemolytic anemia. Renal disease may also be associated with significant hemolysis; hemolytic-uremic syndrome is one example. In this disorder, hemolysis is associated with the presence, on the peripheral blood smear, of echinocytes, helmet cells, fragmented red cells, and spherocytes.

A microangiopathic hemolytic anemia with fragmented red cells and some spherocytes may be observed in several conditions associated with intravascular coagulation and fibrin deposition within vessels. This occurs with DIC complicating severe infection, but may also occur when the intravascular coagulation is localized, as with giant cavernous hemangiomas (Kasabach-Merritt syndrome). Fragmented red cells may also be seen with mechanical damage (eg, associated with artificial heart valves).



In pediatrics, polycythemia is usually secondary to chronic hypoxemia. The disorder differs from polycythemia vera in that only RBCs are affected; the WBC and platelet counts are normal. It occurs as an autosomal dominant or recessive disorder. There are usually no physical findings except for plethora and splenomegaly. The hemoglobin level may be as high as 27 g/dL, and symptoms are generally limited to headache and lethargy. Treatment is not indicated unless symptoms are marked. Phlebotomy is the treatment of choice.


Secondary polycythemia occurs in response to hypoxemia. The most common cause of secondary polycythemia in children is cyanotic congenital heart disease, but it also occurs in chronic pulmonary disease such as cystic fibrosis. Persons living at extremely high altitudes, as well as some with methemoglobinemia, develop polycythemia. Polycythemia may occur in the neonatal period; it is particularly exaggerated in infants who are preterm or small for gestational age. It may occur in infants of diabetic mothers, in Down syndrome, and as a complication of congenital adrenal hyperplasia.

Iron deficiency may complicate polycythemia and aggravate the associated hyperviscosity. This complication should always be suspected when the MCV falls below the normal range. Coagulation and bleeding abnormalities, including thrombocytopenia, mild consumption coagulopathy, and elevated fibrinolytic activity, have been described in severely polycythemic cardiac patients. Bleeding at surgery may be severe.

The ideal treatment of secondary polycythemia is correction of the underlying disorder. When this cannot be done, phlebotomy may be necessary to control symptoms. Iron sufficiency should be maintained. These measures help prevent the complications of thrombosis and hemorrhage.


When heme iron is oxidized, it changes from the ferrous to the ferric state and methemoglobin is produced. Normally, methemoglobin is enzymatically reduced back to hemoglobin. Methemoglobin is unable to transport oxygen and causes a shift in the oxygen dissociation curve. Cyanosis is seen with methemoglobin levels greater than 15%.

1. Hemoglobin M

This designation is given to several abnormal hemoglobins associated with methemoglobinemia due to either amino acid substitutions in α- or β-globin chains. Hemoglobin M is transmitted as an autosomal dominant disorder. Hemoglobin electrophoresis at the usual pH will not always demonstrate the abnormal hemoglobin, and isoelectric focusing may be needed. Affected individuals are cyanotic, but they have normal exercise tolerance and life expectancy. No treatment is indicated.

2. Congenital Methemoglobinemia Due to Enzyme Deficiencies

Congenital methemoglobinemia is caused most frequently by congenital deficiency of the reducing enzyme diaphorase I (coenzyme factor I) and is transmitted as an autosomal recessive trait. Affected individuals may have as much as 40% methemoglobin but usually have no symptoms, although a mild compensatory polycythemia may be present. Patients with diaphorase I deficiency respond to treatment with ascorbic acid and methylene blue (see the next section), but treatment is not usually indicated.

3. Acquired Methemoglobinemia

Nitrites and nitrates, chlorates, and quinines such as aniline dyes, sulfonamides, acetanilid, phenacetin, bismuth subnitrate, and potassium chlorate generate methemoglobin. Recreational use of volatile nitrites (“poppers”) and cocaine may precipitate methemoglobinemia. Poisoning with a drug or chemical containing one of these substances should be suspected with sudden onset cyanosis. Methemoglobin levels in such cases may be extremely high and can produce anoxia, dyspnea, unconsciousness, circulatory failure, and death. Because of transiently deficient NADH methemoglobin reductase, newborns are more susceptible to drug- or chemical-induced methemoglobinemia. Infants with metabolic acidosis may also develop methemoglobinemia.

Patients with acquired methemoglobinemia respond dramatically to intravenous methylene blue. Ascorbic acid administered orally or intravenously also reduces methemoglobin, but the response is slower.




image Increased frequency of infections.

image Ulceration of oral mucosa and gingivitis.

image Decreased absolute neutrophil count; normal numbers of red cells and platelets.

image General Considerations

Neutropenia is an absolute neutrophil (granulocyte) count of less than 1500/μL in childhood, or less than 1100/μL between ages 1 month and 2 years. During the first few days of life, an absolute neutrophil count of less than 3500/μL may be considered neutropenia in term infants. Neutropenia results from absent or defective myeloid stem cells; ineffective or suppressed myeloid maturation; decreased production of hematopoietic cytokines (eg, granulocyte colony-stimulating factor [G-CSF] or granulocyte-macrophage colony-stimulating factor [GM-CSF]); decreased marrow release; increased neutrophil apoptosis; destruction or consumption; or, in pseudoneutropenia, from an increased neutrophil marginating pool (Table 30–5). A decrease in neutrophil mass diminishes delivery of these cells to areas where the balance favors bacterial proliferation and invasion.

Table 30–5. Classification of neutropenia of childhood.


The most severe types of congenital neutropenia include reticular dysgenesis (congenital aleukocytosis), Kostmann syndrome (severe neutropenia with maturation defect in the marrow progenitor cells), Shwachman syndrome (neutropenia with pancreatic insufficiency), neutropenia with immune deficiency states, cyclic neutropenia, and myelokathexis or dysgranulopoiesis. Genetic mutations for Chédiak-Higashi syndrome, Kostmann syndrome, Shwachman syndrome, and cyclic neutropenia and the newly described glucose-6-phosphatase catalytic subunit 3 (G6PC3) have been identified. Neutropenia may also be associated with storage and metabolic diseases and immunodeficiency states. The most common causes of acute neutropenia are viral infection or drugs, resulting in decreased neutrophil production in the marrow, increased peripheral turnover, or both. Severe bacterial infections may be associated with neutropenia. Although not commonly seen, neonatal alloimmune neutropenia can be severe and associated with infection. Autoimmune neutropenia occurs with chronic benign neutropenia of childhood, immunodeficiency syndromes, autoimmune disorders, or, in the newborn, as a result of passive transfer of antibody from the mother to the fetus. Malignancies, osteopetrosis, marrow failure syndromes, and hypersplenism usually are not associated with isolated neutropenia.

image Clinical Findings

A. Symptoms and Signs

Acute severe bacterial or fungal infection is the most significant complication of neutropenia. Although the risk is increased when the absolute neutrophil count is less than 500/μL, the actual susceptibility is variable and depends on the cause of neutropenia, marrow reserves, and other factors. The most common types of infection include septicemia, cellulitis, skin abscesses, pneumonia, and perirectal abscesses. Sinusitis, aphthous ulcers, gingivitis, and periodontal disease also cause significant problems. In addition to local signs and symptoms, patients may have chills, fever, and malaise. In most cases, the spleen and liver are not enlarged. Staphylococcus aureus and gram-negative bacteria are the most common pathogens.

B. Laboratory Findings

Neutrophils are absent or markedly reduced in the peripheral blood smear. In most forms of neutropenia or agranulocytosis, the monocytes and lymphocytes are normal and the red cells and platelets are not affected. The bone marrow usually shows a normal erythroid series, with adequate megakaryocytes, but a marked reduction in the myeloid cells or a significant delay in maturation of this series may be noted. Total cellularity may be decreased.

In the evaluation of neutropenia (eg, persistent, intermittent, cyclic), attention should be paid to the duration and pattern of neutropenia, the types of infections and their frequency, and phenotypic abnormalities on physical examination. A careful family history and blood counts from the parents are useful. If an acquired cause, such as viral infection or drug, is not obvious as an acute cause and no other primary disease is present, WBC counts, white cell differential, and platelet and reticulocyte counts should be completed twice weekly for 6 weeks to determine the possibility of cyclic neutropenia. Bone marrow aspiration and biopsy are most important to characterize the morphologic features of myelopoiesis. Measuring the neutrophil counts in response to corticosteroid infusion may document the marrow reserves. Other tests that aid in the diagnosis include measurement of neutrophil antibodies, immunoglobulin levels, antinuclear antibodies, and lymphocyte phenotyping to detect immunodeficiency states. Tissue culture of bone marrow is important for defining the numbers of stem cells and progenitors committed to the myeloid series or the presence of inhibitory factors. Cytokine levels in plasma or mononuclear cells can be measured directly. Some neutropenia disorders have abnormal neutrophil function, but severe neutropenia may preclude collection of sufficient cells to complete assays. Recent studies have documented abnormalities in an antiapoptotic gene, HAX1, and the elastase gene, ELA2, in Kostmann syndrome and ELA2 mutations in cyclic neutropenia. A mutation for Shwachman syndrome has been described. Increased apoptosis in marrow precursors or circulating neutrophils has been described in several congenital or genetic disorders.

image Treatment

Underlying disorders should be identified and treated or associated agents should be eliminated. Infections should be aggressively assessed and treated. Prophylactic antimicrobial therapy is not indicated for afebrile, asymptomatic patients, but may be considered in rare cases with recurrent infections. Recombinant G-CSF will increase neutrophil counts in most patients; GM-CSF may be considered, but is less extensively used. Patients may be started on 3–5 mcg/kg/d of G-CSF (filgrastim) given subcutaneously or intravenously once a day. Depending on the counts, the dose may be adjusted to keep the absolute neutrophil count below 10,000/μL. Some patients maintain adequate counts with G-CSF given every other day or three times a week. Treatment will decrease infectious complications but may have little effect on periodontal disease. However, not all patients with neutropenia syndromes require G-CSF (eg, chronic benign neutropenia of childhood). Patients with cyclic neutropenia may have a milder clinical course as they grow older. Immunizations should be given if the adaptive immune system is normal. Hematopoietic stem cell transplant may be considered for patients with severe complications, especially those with severe congenital neutropenia.

image Prognosis

The prognosis varies greatly with the cause and severity of the neutropenia. In severe cases with persistent agranulocytosis, the prognosis is poor in spite of antibiotic therapy; in mild or cyclic forms of neutropenia, symptoms may be minimal and the prognosis for normal life expectancy excellent. G-CSF has the potential to prolong life expectancy. Up to 50% of patients with Shwachman syndrome may develop aplastic anemia, myelodysplasia, or leukemia during their lifetime. Patients with Kostmann syndrome also have a potential for leukemia, as do patients with neutropenia associated with some immune disorders. Hematopoietic stem cell transplant may be the only curative therapy for some disorders.

Audrain M: Autoimmune neutropenia in children: analysis of 116 cases. Pediatr Alergy Immunol 2011:22;494 [PMID: 21771084].

Boztug K: Extended spectrum of human glucose-6-phosphatase catalytic subunit deficiency: novel genotypes and phenotype variability in congenital neutropenia. J Pediatr 2012;160:674 [PMID: 22050868].

Boztug K: Genetic etiologies of severe congenital neutropenia. Curr Opin Pediatr 2011;23:21 [PMID: 21206270].

Carlsson G: Hematopoietic stem cell transplantation in severe congenital neutropenia. Pediatr Blood Cancer 2011;56:444 [PMID: 21072829].

Fioredda F: Infectious complications in children with severe congenital, autoimmune or idiopathic neutropenia: a retrospective study form the Italian neutropenia registry. Pediatr Infect Dis J 2012 [Epub ahead of print] [PMID: 23249920].


Neutrophilia is an increase in the absolute neutrophil count in the peripheral blood to greater than 7500–8500/μL for infants, children, and adults. To support the increased peripheral count, neutrophils may be mobilized from bone marrow storage or peripheral marginating pools. Neutrophilia occurs acutely in association with bacterial or viral infections, inflammatory diseases (eg, juvenile rheumatoid arthritis, inflammatory bowel disease, Kawasaki disease), surgical or functional asplenia, liver failure, diabetic ketoacidosis, azotemia, congenital disorders of neutrophil function (eg, chronic granulomatous disease, leukocyte adherence deficiency), and hemolysis. Drugs such as corticosteroids, lithium, and epinephrine increase the blood neutrophil count. Corticosteroids cause release of neutrophils from the marrow pool, inhibit egress from capillary beds, and postpone apoptotic cell death. Epinephrine causes release of the marginating pool. Acute neutrophilia has been reported after stress, such as from electric shock, trauma, burns, surgery, and emotional upset. Tumors involving the bone marrow, such as lymphomas, neuroblastomas, and rhabdomyosarcoma, may be associated with leukocytosis and the presence of immature myeloid cells in the peripheral blood. Infants with Down syndrome have defective regulation of proliferation and maturation of the myeloid series and may develop neutrophilia. At times this process may affect other cell lines and mimic myeloproliferative disorders or acute leukemia.

The neutrophilias must be distinguished from myeloproliferative disorders such as chronic myelogenous leukemia and juvenile chronic myelogenous leukemia. In general, abnormalities involving other cell lines, the appearance of immature cells on the blood smear, and the presence of hepatosplenomegaly are important differentiating characteristics.


Neutrophils play a key role in host defenses. Circulating in capillary beds, they adhere to the vascular endothelium adjacent to sites of infection and inflammation. Moving between endothelial cells, the neutrophil migrates toward the offending agent. Contact with a microbe that is properly opsonized with complement or antibodies triggers ingestion, a process in which cytoplasmic streaming results in the formation of pseudopods that fuse around the invader, encasing it in a phagosome. During the ingestion phase, the oxidase enzyme system assembles and is activated, taking oxygen from the surrounding medium and reducing it to form toxic oxygen metabolites critical to microbicidal activity. Concurrently, granules from the two main classes (azurophil and specific) fuse and release their contents into the phagolysosome. The concentration of toxic oxygen metabolites (eg, hydrogen peroxide, hypochlorous acid, hydroxyl radical) and other compounds (eg, proteases, cationic proteins, cathepsins, defensins) increases dramatically, resulting in the death and dissolution of the microbe. Complex physiologic and biochemical processes support and control these functions. Defects in any of these processes may lead to inadequate cell function and an increased risk of infection.

image Classification

Table 30–6 summarizes congenital neutrophil function defects. Recently reported is variant CGD with p40phox deficiency manifested by inflammatory bowel disease. Also described is a syndrome of severe neutrophil dysfunction and severe infections associated with a mutation in a GTPase signaling molecule, Rac2. New syndromes of innate immune dysfunction include defects in interferon and interleukin (IL)-12 receptor and signaling pathways, leading to monocyte and macrophage dysfunction and defective toll-like receptor signaling pathways (IL-1 receptor-associated [IRAK] deficiency) associated with recurrent bacterial infections. Leukocyte adhesion deficiency (LAD) III is a disorder characterized by severe bleeding, impaired leukocyte adhesion, and endothelial inflammation, and is associated with mutations of FERMT3 gene, which encodes for a protein, Kindlin-3, critical for intracellular function of β integrins. Other congenital or acquired causes of mild to moderate neutrophil dysfunction include metabolic defects (eg, glycogen storage disease 1b, G6PC3 deficiency, diabetes mellitus, renal disease, and hypophosphatemia), viral infections, and certain drugs. Neutrophils from newborn infants have abnormal adherence, chemotaxis, and bactericidal activity. Cells from patients with thermal injury, trauma, and overwhelming infection have defects in cell motility and bactericidal activity similar to those seen in neonates.

Table 30–6. Classification of congenital neutrophil function deficits.


image Clinical Findings

Recurrent bacterial or fungal infections are the hallmark of neutrophil dysfunction. Although patients will have infection-free periods, episodes of pneumonia, sinusitis, cellulitis, cutaneous and mucosal infections (including perianal or peritonsillar abscesses), and lymphadenitis are frequent. As with neutropenia, aphthous ulcers of mucous membranes, severe gingivitis, and periodontal disease are also major complications. In general, S aureus or gram-negative organisms are commonly isolated from infected sites; other organisms may be specifically associated with a defined neutrophil function defect. In some disorders, fungi account for an increasing number of infections. Deep or generalized infections, such as osteomyelitis, liver abscesses, sepsis, meningitis, and necrotic or gangrenous soft-tissue lesions, occur in specific syndromes (eg, leukocyte adherence deficiency or chronic granulomatous disease). Patients with severe neutrophil dysfunction may die in childhood from severe infections or associated complications. Table 30–6 summarizes pertinent laboratory findings.

image Treatment

The mainstays of management of these disorders are anticipation of infections and aggressive attempts to identify the foci and the causative agents. Surgical procedures to achieve these goals may be both diagnostic and therapeutic. Broad-spectrum antibiotics covering the range of possible organisms should be initiated without delay, switching to specific antimicrobial agents when the microbiologic diagnosis is made. When infections are unresponsive or they recur, granulocyte transfusions may be helpful.

Chronic management may include prophylactic antibiotics. Trimethoprim-sulfamethoxazole and some other antibiotics (eg, rifampin) enhance the bactericidal activity of neutrophils from patients with chronic granulomatous disease. Some patients with Chédiak-Higashi syndrome improve clinically when given ascorbic acid. Recombinant γ-interferon decreases the number and severity of infections in patients with chronic granulomatous disease. Demonstration of this activity with one patient group raises the possibility that cytokines, growth factors, and other biologic response modifiers may be helpful in other conditions in preventing recurrent infections. Bone marrow transplant has been attempted in most major congenital neutrophil dysfunction syndromes, and reconstitution with normal cells and cell function has been documented. Combining genetic engineering with autologous bone marrow transplant may provide a future strategy for curing these disorders.

image Prognosis

For mild to moderate defects, anticipation and conservative medical management ensure a reasonable outlook. For severe defects, excessive morbidity and significant mortality still exist. In some diseases, the development of noninfectious complications, such as the lymphoproliferative phase of Chédiak-Higashi syndrome or inflammatory syndromes in chronic granulomatous disease, may influence prognosis.

Ambruso DR, Johnston RB Jr: Chronic granulomatous disease of childhood and common variable immunodeficiency. In: Chernick V, Boat TF (eds): Kendig’s Disorders of the Respiratory Tract in Children. Elsevier Saunders; 2012.

Dinauer MC: Disorders of neutrophil function: an overview. Methods Mol Biol 2007;412:489 [PMID: 18453130].

Kuhns DB: Residual NADPH oxidase and survival in chronic granulomatous disease. N Engl J Med 2010;363:2600 [PMID: 21190454].

Robert P: A novel leukocyte deficiency III variant: kindlin-3 deficiency results in integrin- and nonintegrin-related defects in different steps of leukocyte adhesion. J Immunol 2011;186:5273 [PMID: 21441448].

Walach B: Lessons learned from phagocytic function studies in a large cohort of patients with recurring infections. J Clin Immunol 2012;32:454 [PMID: 222072252].


From the first week up to the fifth year of life, lymphocytes are the most numerous leukocytes in human blood. The ratio then reverses gradually to reach the adult pattern of neutrophil predominance. An absolute lymphocytosis in childhood is associated with acute or chronic viral infections, pertussis, syphilis, tuberculosis, and hyperthyroidism. Other noninfectious conditions, drugs, and hypersensitivity and serum sickness–like reactions cause lymphocytosis.

Fever, upper respiratory symptoms, gastrointestinal complaints, and rashes are clues in distinguishing infectious from noninfectious causes. The presence of enlarged liver, spleen, or lymph nodes is crucial to the differential diagnosis, which includes acute leukemia and lymphoma. Most cases of infectious mononucleosis are associated with hepatosplenomegaly or adenopathy. The absence of anemia and thrombocytopenia helps to differentiate these disorders. Evaluation of the morphology of lymphocytes on peripheral blood smear is crucial. Infectious causes, particularly infectious mononucleosis, are associated with atypical features in the lymphocytes, such as basophilic cytoplasm, vacuoles, finer and less-dense chromatin, and an indented nucleus. These features are distinct from the characteristic morphology associated with lymphoblastic leukemia. Lymphocytosis in childhood is most commonly associated with infections and resolves with recovery from the primary disease.


Eosinophilia in infants and children is an absolute eosinophil count greater than 300/μL. Marrow eosinophil production is stimulated by the cytokine IL-5. Allergies, particularly those associated with asthma and eczema, are the most common primary causes of eosinophilia in children. Eosinophilia also occurs in drug reactions, with tumors (Hodgkin and non-Hodgkin lymphomas and brain tumors), and with immunodeficiency and histiocytosis syndromes. Increased eosinophil counts are a prominent feature of many invasive parasitic infections. Gastrointestinal disorders such as chronic hepatitis, ulcerative colitis, Crohn disease, and milk precipitin disease may be associated with eosinophilia. Increased blood eosinophil counts have been identified in several families without association with any specific illness. Rare causes of eosinophilia include the hypereosinophilic syndrome, characterized by counts greater than 1500/μL and organ involvement and damage (hepatosplenomegaly, cardiomyopathy, pulmonary fibrosis, and central nervous system injury). This is a disorder of middle-aged adults and is rare in children. Eosinophilic leukemia has been described, but its existence as a distinct entity is very rare.

Eosinophils are sometimes the last type of mature myeloid cell to disappear after marrow ablative chemotherapy. Increased eosinophil counts are associated with graft-versus-host disease after bone marrow transplant, and elevations are sometimes documented during rejection episodes in patients who have solid organ grafts.


Bleeding disorders may occur as a result of (1) quantitative or qualitative abnormalities of platelets, (2) quantitative or qualitative abnormalities in plasma procoagulant factors, (3) vascular abnormalities, or (4) accelerated fibrinolysis. The coagulation cascade and fibrinolytic system are shown in Figures 30–4 and 30–5.


image Figure 30–4. The procoagulant system and formation of a fibrin clot. Vascular injury initiates the coagulation process by exposure of tissue factor (TF); the dashed lines indicate thrombin actions in addition to clotting of fibrinogen. The dotted lines associated with VIIa indicate the feedback activation of the VII-TF complex by Xa and IXa. Ca++, calcium; HK, high-molecular-weight kininogen; PL, phospholipid; PK, prekallikrein. (Reproduced, with permission, from Good-night SH, Hathaway WE [eds]: Disorders of Hemostasis & Thrombosis: A Clinical Guide, 2nd ed. McGraw-Hill; 2001.)


image Figure 30–5. The fibrinolytic system. Solid arrows indicate activation; dashed line arrows indicate inhibition. ECM, extracellular matrix; FDP, fibrinogen-fibrin degradation products; MMP, matrix metalloproteinases; PAI, plasminogen activator inhibitor; TAFI, thrombin activatable fibrinolysis inhibitor; tPA, tissue plasminogen activator; uPA, urokinase; uPAR, cellular urokinase receptor. (Reproduced, with permission, from Goodnight SH, Hathaway WE [eds]: Disorders of Hemostasis & Thrombosis: A Clinical Guide, 2nd ed. McGraw-Hill; 2001.)

The most critical aspect in evaluating the bleeding patient is obtaining detailed personal and family bleeding histories, including bleeding complications associated with dental interventions, surgeries, suture placement and removal, and trauma. Excessive mucosal bleeding is suggestive of a platelet disorder, von Willebrand disease (vWD), dysfibrinogenemia, or vasculitis. Bleeding into muscles and joints may be associated with a plasma procoagulant factor abnormality. In either scenario, the abnormality may be congenital or acquired. A thorough physical examination should be performed with special attention to the skin, oro- and nasopharynx, liver, spleen, and joints. Screening and diagnostic evaluation in patients with suspected bleeding disorders should initially include the following laboratory testing:

1. Prothrombin time (PT) to assess clotting function of factors VII, X, V, II, and fibrinogen.

2. Activated partial thromboplastin time (aPTT) to assess clotting function of high-molecular-weight kininogen, prekallikrein, XII, XI, IX, VIII, X, V, II, and fibrinogen.

3. Platelet count and size determined as part of a CBC.

4. Platelet functional assessment by platelet function analyzer-100 (PFA-100), template bleeding time, or whole blood platelet aggregometry.

5. Fibrinogen functional level by clotting assay.

The following laboratory tests may also be useful:

1. Thrombin time to measure the generation of fibrin from fibrinogen following conversion of prothrombin to thrombin, as well as the antithrombin effects of fibrin-split products and heparin. The thrombin time may be prolonged in the setting of a normal fibrinogen concentration if the fibrinogen is dysfunctional (ie, dysfibrinogenemia).

2. Euglobulin lysis time (ELT), if available, to evaluate for accelerated fibrinolysis if the preceding workup is nonrevealing despite documented history of pathologic bleeding. If the ELT is shortened, assessment of plasminogen activator inhibitor-1 and α2-antiplasmin is warranted, as congenital deficiency in these fibrinolytic inhibitors may cause hyperfibrinolysis. In ill patients, measurement of fibrin degradation products may assist in the diagnosis of DIC.

Goodnight SH, Hathaway WE (eds): Disorders of Hemostasis & Thrombosis: A Clinical Guide, 2nd ed. McGraw-Hill, 2001:41–51.


Thrombocytopenia in the pediatric age range is often immune-mediated (eg, ITP, neonatal auto- or alloimmune thrombocytopenia, heparin-induced thrombocytopenia), but is also caused by consumptive coagulopathy (eg, DIC, Kasabach-Merritt syndrome), acute leukemias, rare disorders such as Wiskott-Aldrich syndrome and type 2b vWD, and artifactually in automated cytometers (eg, Bernard-Soulier syndrome), where giant forms may not be enumerated as platelets by automated cell counters.

1. Idiopathic Thrombocytopenic Purpura


image Otherwise healthy child.

image Decreased platelet count.

image Petechiae, ecchymoses.

image General Considerations

Acute idiopathic thrombocytopenic purpura (ITP) is the most common bleeding disorder of childhood. It occurs most frequently in children aged 2–5 years and often follows infection with viruses, such as rubella, varicella, measles, parvovirus, influenza, EBV, or acute and chronic HIV. Most patients recover spontaneously within a few months. Chronic ITP (> 12 months’ duration) occurs in 10%–20% of affected patients. The thrombocytopenia results from clearance of circulating IgM- or IgG-coated platelets by the reticuloendothelial system. The spleen plays a predominant role in the disease by forming the platelet cross-reactive antibodies and sequestering the antibody-bound platelets.

image Clinical Findings

A. Symptoms and Signs

Onset of ITP is usually acute, with the appearance of multiple petechiae and ecchymoses. Epistaxis is also common at presentation. No other physical findings are usually present. Rarely, concurrent infection with EBV or CMV may cause hepatosplenomegaly or lymphadenopathy, simulating acute leukemia.

B. Laboratory Findings

1. Blood—The platelet count is markedly reduced (usually < 50,000/μL and often < 10,000/μL), and platelets frequently are of larger size on peripheral blood smear, suggesting accelerated production of new platelets. The white blood count and differential are normal, and the hemoglobin concentration is preserved unless hemorrhage has been significant.

2. Bone marrow—The number of megakaryocytes is increased. Erythroid and myeloid cellularity is normal.

3. Other laboratory tests—Platelet-associated IgG or IgM, or both, may be demonstrated on the patient’s platelets or in the serum. PT and aPTT are normal.

image Differential Diagnosis

Table 30–7 lists common causes of thrombocytopenia. ITP is a diagnosis of exclusion. Family history or the finding of predominantly giant platelets on the peripheral blood smear is helpful in separating from thrombocytopenia that is hereditary. Bone marrow examination should be performed if the history is atypical (ie, the child is not otherwise healthy, or if there is a family history of bleeding), if abnormalities other than purpura and petechiae are present on physical examination, or if other cell lines are affected on the CBC. The importance of performing a bone marrow examination prior to using corticosteroids in the treatment for ITP remains controversial.

Table 30–7. Common causes of thrombocytopenia.


image Complications

Severe hemorrhage and bleeding into vital organs are the feared complications of ITP. Intracranial hemorrhage is the most serious (but rarely seen) complication, occurring in less than 1% of affected children. The most important risk factors for hemorrhage are a platelet count less than 10,000/μL and mean platelet volume less than 8 fL.

image Treatment

A. General Measures

Treatment is optional in most children in the absence of bleeding. Aspirin and other medications (eg, NSAIDs such as Advil, Naproxin, etc) that compromise platelet function should be avoided. Bleeding precautions (eg, restriction from physical contact activities and use of helmets) should be observed. Platelet transfusion should be avoided except in circumstances of life-threatening bleeding, in which case emergent splenectomy may be considered. In this setting, administration of corticosteroids and IVIG is also advisable.

B. Corticosteroids

Patients with clinically significant but non–life-threatening bleeding (ie, epistaxis, hematuria, and hematochezia) and those with a platelet count of less than 10,000/μL may benefit from prednisone at 1–2 mg/kg/d for 2–3 weeks with a maximum dose of 60–80 mg/d. A higher dose initially (3–5 mg/kg/d) for 3–7 days may lead to faster count recovery. The dosage is then tapered and stopped. No further prednisone is given regardless of the platelet count unless significant bleeding recurs, at which time prednisone is administered in the smallest dose that achieves resolution of bleeding episodes (usually 2.5–5 mg twice daily). Follow-up continues until the steroid can again be discontinued, spontaneous remission occurs, or other therapeutic measures are instituted. Toxicity (Cushingoid facies, weight gain, change in behavior, hyperglycemia, and hypertension) is usually mild for short treatment courses.

C. Intravenous Immunoglobulin

Intravenous immunoglobulin (IVIG) is the treatment of choice for severe, acute bleeding, and may also be used as an alternative or adjunct to corticosteroid treatment in both acute and chronic ITP of childhood. IVIG may be effective even when the patient is resistant to corticosteroids; responses are prompt and may last for several weeks. Most patients receive 0.8–1 g/kg/d for 1–2 days. Infusion time is typically 4–6 hours. Platelets may be given simultaneously during life-threatening hemorrhage but are rapidly destroyed. Adverse effects of IVIG are common, including transient neurologic complications in one-third of patients (eg, headache, nausea, and aseptic meningitis). These symptoms may mimic those of intracranial hemorrhage and necessitate radiologic evaluation of the brain. A transient decrease in neutrophil number may also be seen.

D. Anti-Rh(D) Immunoglobulin

This polyclonal immunoglobulin binds to the D antigen on RBCs. The splenic clearance of anti-D–coated red cells interferes with removal of antibody-coated platelets, resulting in improvement in thrombocytopenia. This approach is effective only in Rh(+) patients with a functional spleen. The time required for platelet increase is slightly longer than with IVIG. However, approximately 80% of Rh(+) children with acute or chronic ITP respond well. Significant hemolysis may occur transiently with an average hemoglobin concentration decrease of 0.8 g/dL. However, severe hemolysis occurs in 5% of treated children, and clinical and laboratory evaluation following administration is warranted in all patients. Rh(D) immunoglobulin is less expensive and infused more rapidly than IVIG but is more expensive than corticosteroids.

E. Splenectomy

Many children with chronic ITP have platelet counts greater than 30,000/μL. Up to 70% of such children spontaneously recover with a platelet count greater than 100,000/μL within 1 year. For the remainder, corticosteroids, IVIG, and anti-D immunoglobulin are typically effective treatment for acute bleeding. Splenectomy produces a response in 70%–80%, but it should be considered only after persistence of significant thrombocytopenia for at least 1 year. Preoperative treatment with corticosteroids, IVIG, or anti-D immunoglobulin is usually indicated. Postoperatively, the platelet count may rise to 1 million/μL. This reactive thrombocytosis is not associated with thrombotic complications in children. The risk of overwhelming infection (predominantly with encapsulated organisms) is increased after splenectomy, particularly in the young child. Therefore, the procedure should be postponed, if possible, until age 5 years. Administration of pneumococcal, H influenzae type b and meningococcal vaccines at least 2 weeks prior to splenectomy is recommended. Daily penicillin prophylaxis should be started postoperatively and continued at least until 5 years of age.

F. Rituximab (Anti-CD20 Monoclonal Antibody)

There have been no randomized trials for rituximab in children. The efficacy of treating childhood chronic ITP in several series and case studies has demonstrated a response rate of 60%. Because of significant adverse events, this therapy may be reserved for refractory cases with significant bleeding or as an alternative to splenectomy.

G. New Agents

One randomized clinical trial in children has been conducted with romiplostim, a thrombopoietin receptor agonist, with an 88% response rate and improved quality of life. Further studies in larger numbers of pediatric patients are needed to address the possibility that some patients with chronic ITP have a response in platelet production that is not maximally increased.

image Prognosis

Ninety percent of children with ITP will have a spontaneous remission. Features associated with the development of chronic ITP include female gender, age greater than 10 years at presentation, insidious onset of bruising, and the presence of other autoantibodies. Older child- and adolescent-onset ITP is associated with an increased risk of chronic autoimmune diseases or immunodeficiency states. Appropriate screening by history and laboratory studies (eg, antinuclear antibody) is warranted.

Blanchette V: Childhood immune thrombocytopenic purpura: diagnosis and management. Pediatr Clin North Am 2008;55:393 [PMID: 18381093].

Journeycake J: Childhood immune thrombocytopenia: role of rituximab, recombinant thrombopoietin, and other new therapeutics. Hematology Am Soc Hematol Educ Program 2012;2012:444 [PMID: 23233617].

Neunert C: The American Society of Hematology 2011 evidence-based practice guideline for immune thrombocytopenia. Blood 2011;117:4190 [PMID: 21325604].

2. Thrombocytopenia in the Newborn

Thrombocytopenia is one of the most common causes of neonatal hemorrhage and should be considered in any newborn with petechiae, purpura, or other significant bleeding. Defined as a platelet count less than 150,000/μL, thrombocytopenia occurs in approximately 0.9% of unselected neonates. Several specific entities may be responsible (see Table 30–7); however, half of such neonates have alloimmune thrombocytopenia. Infection and DIC are the most common causes of thrombocytopenia in ill full-term newborns and in preterm newborns. In the healthy neonate, antibody-mediated thrombocytopenia (alloimmune or maternal autoimmune), viral syndromes, hyperviscosity, and major-vessel thrombosis are frequent causes of thrombocytopenia. Management is directed toward the underlying etiology.

A. Thrombocytopenia Associated With Platelet Alloantibodies (Neonatal Alloimmune Thrombocytopenia)

Platelet alloimmunization occurs in 1 out of approximately 350 pregnancies. Unlike in Rh incompatibility, 30%–40% of affected neonates are first-born. Thrombocytopenia is progressive over the course of gestation and worse with each subsequent pregnancy. Alloimmunization occurs when a platelet antigen of the infant differs from that of the mother, and the mother is sensitized by fetal platelets that cross the placenta into the maternal circulation. In Caucasians, alloimmune thrombocytopenia is most often due to human platelet antigen (HPA)-1a incompatibility. Sensitization of a mother homozygous for HPA-1b to paternally acquired fetal HPA-1a antigen results in severe fetal thrombocytopenia in 1 in 1200 fetuses. Only 1 in 20 HPA-1a–positive fetuses of HPA-1a–negative mothers develop alloimmunization. Other platelet-specific alloantigens may be etiologic. The presence of antenatal maternal platelet antibodies on more than one occasion and their persistence into the third trimester is predictive of severe neonatal thrombocytopenia; a weak or undetectable antibody does not exclude thrombocytopenia. Severe intracranial hemorrhage occurs in 10%–30% of affected neonates as early as 20 weeks’ gestation. Petechiae or other bleeding manifestations are usually present shortly after birth. The disease is self-limited, and the platelet count normalizes within 4 weeks.

If alloimmunization is associated with clinically significant bleeding, transfusion of platelet concentrates harvested from the mother is more effective than random donor platelets in increasing the platelet count. Transfusion with HPA-matched platelets from unrelated donors or treatment with IVIG or methylprednisolone to acutely block macrophage uptake of sensitized cells has also been successful in raising the platelet count and achieving hemostasis. If thrombocytopenia is not severe and bleeding is absent, observation alone is often appropriate.

Intracranial hemorrhage in a previous child secondary to alloimmune thrombocytopenia is the strongest risk factor for severe fetal thrombocytopenia and hemorrhage in a subsequent pregnancy. Amniocentesis or chorionic villus sampling to obtain fetal DNA for platelet antigen typing is sometimes performed if the father is heterozygous for HPA-1a. If alloimmunization has occurred with a previous pregnancy, irrespective of history of intracranial hemorrhage, screening cranial ultrasound for hemorrhage should begin at 20 weeks’ gestation and be repeated regularly. In addition, cordocentesis should be performed at approximately 20 weeks’ gestation, with prophylactic transfusion of irradiated, leukoreduced, maternal platelet concentrates. If the fetal platelet count is less than 100,000/μL, the mother should be treated with weekly IVIG. Delivery by elective cesarean section is recommended if the fetal platelet count is less than 50,000/μL, to minimize the risk of intracranial hemorrhage associated with birth trauma.

B. Thrombocytopenia Associated With ITP in the Mother (Neonatal Autoimmune Thrombocytopenia)

Infants born to mothers with idiopathic thrombocytopenic purpura (ITP) or other autoimmune diseases (eg, antiphospholipid antibody syndrome or systemic lupus erythematosus) may develop thrombocytopenia as a result of transfer of antiplatelet IgG from the mother to the infant. Unfortunately, maternal and fetal platelet counts and maternal antiplatelet antibody levels are unreliable predictors of bleeding risk. Antenatal corticosteroid administration to the mother is generally instituted once maternal platelet count falls below 50,000/μL, with or without a concomitant course of IVIG.

Most neonates with autoimmune thrombocytopenia do not develop clinically significant bleeding, and thus treatment for thrombocytopenia is not often required. The risk of intracranial hemorrhage is 0.2%–2%. If diffuse petechiae or minor bleeding are evident, a 1- to 2-week course of oral prednisone, 2 mg/kg/d, may be helpful. If the platelet count remains consistently less than 20,000/μL or if severe hemorrhage develops, IVIG should be given (0.8–1 g/kg daily for 1–2 days). Platelet transfusions are only indicated for life-threatening bleeding, and may only be effective after removal of antibody by exchange transfusion. The platelet nadir is typically between the fourth and sixth day of life and improves significantly by 1 month; full recovery may take 2–4 months. Platelet recovery may be delayed in breast-fed infants because of transfer of IgG to the milk.

C. Neonatal Thrombocytopenia Associated With Infections

Thrombocytopenia is commonly associated with severe generalized infections during the newborn period. Between 50% and 75% of neonates with bacterial sepsis are thrombocytopenic. Intrauterine infections such as rubella, syphilis, toxoplasmosis, CMV, herpes simplex (acquired intra- or postpartum), enteroviruses, and parvovirus are often associated with thrombocytopenia. In addition to specific treatment for the underlying disease, platelet transfusions may be indicated in severe cases.

D. Thrombocytopenia Associated With Kaposiform Hemangioendotheliomas (Kasabach-Merritt Syndrome)

A rare but important cause of thrombocytopenia in the newborn is kaposiform hemangioendotheliomas, a benign neoplasm with histopathology distinct from that of classic infantile hemangiomas. Intense platelet sequestration in the lesion results in peripheral thrombocytopenia and may rarely be associated with a DIC-like picture and hemolytic anemia. The bone marrow typically shows megakaryocytic hyperplasia in response to the thrombocytopenia. Corticosteroids, α-interferon, and vincristine are all useful for reducing the size of the lesion and are indicated if significant coagulopathy is present, the lesion compresses a vital structure, or the lesion is cosmetically unacceptable. If consumptive coagulopathy is present, heparin or aminocaproic acid may be useful. Depending on the site, embolization may be an option. Surgery is often avoided because of the high risk of hemorrhage.

Hall GW: Kasabach-Merritt syndrome: pathogenesis and management. Br J Haematol 2001;112:851 [PMID: 11298580].

Kaplan RN: Differential diagnosis and management of thrombocytopenia in childhood. Pediatr Clin North Am 2004;51:1109 [PMID: 15275991].

3. Disorders of Platelet Function

Individuals with platelet function defects typically develop abnormal bruising and mucosal bleeding similar to that occurring in persons with thrombocytopenia. Historically, platelet function has been screened by measuring the bleeding time. If this is prolonged, in vitro platelet aggregation is studied using agonists, such as adenosine diphosphate, collagen, arachidonic acid, and ristocetin. While labor-intensive, platelet aggregometry remains important in selected clinical situations, the PFA-100 has become available to evaluate platelet dysfunction and vWD and has replaced the template bleeding time in many clinical laboratories. Unfortunately, none of these screening tests of platelet function is uniformly predictive of clinical bleeding severity.

Platelet dysfunction may be inherited or acquired, with the latter being more common. Acquired disorders of platelet function may occur secondary to uremia, cirrhosis, sepsis, myeloproliferative disorders, congenital heart disease, and viral infections. Many pharmacologic agents decrease platelet function. The most common offending agents in the pediatric population are aspirin and other NSAIDs, synthetic penicillins, and valproic acid. In acquired platelet dysfunction, the PFA-100 closure time is typically prolonged with collagen-epinephrine, while normal with collagen-ADP.

The inherited disorders are due to defects in platelet-vessel interaction, platelet-platelet interaction, platelet granule content or release (including defects of signal transduction), thromboxane and arachidonic acid pathway, and platelet-procoagulant protein interaction. Individuals with hereditary platelet dysfunction generally have a prolonged bleeding time with normal platelet number and morphology by light microscopy. PFA-100 closure time, in contrast to that in acquired dysfunction, is typically prolonged with both collagen-ADP and collagen-epinephrine.

Congenital causes of defective platelet–vessel wall interaction include Bernard-Soulier syndrome, which is characterized by increased platelet size and decreased platelet number. The molecular defect in this autosomal recessive disorder is a deficiency or dysfunction of glycoprotein Ib-V-IX complex on the platelet surface resulting in impaired von Willebrand factor (vWF) binding, and hence, impaired platelet adhesion to the vascular endothelium.

Glanzmann thrombasthenia is an example of platelet-platelet dysfunction. In this autosomal recessive disorder, glycoprotein IIb-IIIa is deficient or dysfunctional. Platelets do not bind fibrinogen effectively and exhibit impaired aggregation. As in Bernard-Soulier syndrome, acute bleeding is treated by platelet transfusion.

Disorders involving platelet granule content include storage pool disease and Quebec platelet disorder. In individuals with storage pool disease, platelet-dense granules lack adenosine dinucleotide phosphate and adenosine trinucleotide phosphate and are often found to be low in number by electron microscopy. These granules are also deficient in Hermansky-Pudlak, Chédiak-Higashi, and Wiskott-Aldrich syndromes. Whereas deficiency of a second granule class, α-granules, results in the gray platelet syndrome, Quebec platelet disorder is characterized by a normal platelet α-granule number, but with abnormal proteolysis of α-granule proteins and deficiency of platelet α-granule multimerin. α-Granule abnormality in this disorder also results in increased serum levels of urokinase-type plasminogen activator. Epinephrine-induced platelet aggregation is markedly impaired.

Platelet dysfunction has also been observed in other congenital syndromes, such as Down and Noonan syndromes, without a clear understanding of the molecular defect.

image Treatment

Acute bleeding in many individuals with acquired or selected congenital platelet function defects responds to therapy with desmopressin acetate, likely due to an induced release of vWF from endothelial stores and/or upregulated expression of glycoprotein Ib-V-IX on the platelet surface. If this therapy is ineffective, or if the patient has Bernard-Soulier syndrome or Glanzmann syndrome, the mainstay of treatment for bleeding episodes is platelet transfusion, possibly with HLA type-specific platelets. Recombinant VIIa has variable efficacy and may be helpful in platelet transfusion-refractory patients.

Israels S: Platelet disorders in children: a diagnostic approach. Pediatr Blood Cancer 2011;56:975 [PMID: 21294245].


Table 30–8 lists normal values for coagulation factors. The more common factor deficiencies are discussed in this section. Individuals with bleeding disorders should avoid exposure to medications that inhibit platelet function. Participation in contact sports should be considered in the context of the severity of the bleeding disorder.

Table 30–8. Physiologic alterations in measurements of the hemostatic system.



1. Factor VIII Deficiency (Hemophilia A, Classic Hemophilia)


image Bruising, soft-tissue bleeding, hemarthrosis.

image Prolonged aPTT.

image Reduced factor VIII activity.

image General Considerations

Factor VIII activity is reported in units per milliliter, with 1 U/mL equal to 100% of the factor activity found in 1 mL of normal plasma. The normal range for factor VIII activity is 0.50–1.50 U/mL (50%–150%). Hemophilia A occurs predominantly in males as an X-linked disorder. One-third of cases are due to a new mutation. The incidence of factor VIII deficiency is 1:5000 male births.

image Clinical Findings

A. Symptoms and Signs

Patients with severe hemophilia A (< 1% plasma factor VIII activity) have frequent spontaneous bleeding episodes involving skin, mucous membranes, joints, muscles, and viscera. In contrast, patients with mild hemophilia A (5%–40% factor VIII activity) bleed only at times of trauma or surgery. Those with moderate hemophilia A (1% to < 5% factor VIII activity) typically have intermediate bleeding manifestations. The most crippling aspect of factor VIII deficiency is the tendency to develop recurrent hemarthroses that incite joint destruction.

B. Laboratory Findings

Individuals with hemophilia A have a prolonged aPTT, except in some cases of mild deficiency. The PT is normal. The diagnosis is confirmed by finding decreased factor VIII activity with normal vWF activity. In two-thirds of families of hemophilic patients, the females are carriers and some are mildly symptomatic. Carriers of hemophilia can be suspected by determination of the ratio of factor VIII activity to vWF antigen and definitely diagnosed by molecular genetic techniques. In a male fetus or newborn with a family history of hemophilia A, cord blood sampling for factor VIII activity is accurate and important in subsequent care.

image Complications

Intracranial hemorrhage is the leading disease-related cause of death among patients with hemophilia. Most intracranial hemorrhages in moderate to severe deficiency are spontaneous (ie, not associated with trauma). Hemarthroses begin early in childhood and, particularly when recurrent, can result in joint destruction (ie, hemophilic arthropathy). Large intramuscular hematomas can lead to a compartment syndrome with resultant neurologic compromise. Although these complications are most common in severe hemophilia A, they may be experienced by individuals with moderate or mild disease. A serious complication of hemophilia is the development of an acquired circulating antibody to factor VIII after treatment with factor VIII concentrate. Such factor VIII inhibitors develop in up to 30% of patients with severe hemophilia A, and most commonly among patients with large deletions in the factor VIII gene. Inhibitors may be amenable to desensitization with regular factor VIII infusion (immune tolerance therapy) with or without immunosuppressive therapy. The “bypassing agent,” recombinant factor VIIa, has become a therapy of choice for treatment of acute hemorrhage in patients with hemophilia A and a high-titer inhibitor.

In prior decades, therapy-related complications in hemophilia A have included infection with the HIV, hepatitis B virus, and hepatitis C virus. Through stringent donor selection, the implementation of sensitive screening assays, the use of heat or chemical methods for viral inactivation, and the development of recombinant products, the risk of these infections is minimal. Inactivation methods do not eradicate viruses lacking a lipid envelope, however, so that transmission of parvovirus and hepatitis A remains a concern with the use of plasma-derived products. Immunization with hepatitis A and hepatitis B vaccines is recommended for all hemophilia patients.

image Treatment

The general aim of management is to raise the factor VIII activity to prevent or stop bleeding. Some patients with mild factor VIII deficiency may respond to desmopressin via release of endothelial stores of factor VIII and vWF into plasma; however, most patients require administration of exogenous factor VIII to achieve hemostasis. The in vivo half-life of factor VIII is generally 8–12 hours but may exhibit considerable variation among individuals depending on comorbid conditions. Non–life-threatening, non–limb-threatening hemorrhage is treated initially with 20–30 U/kg of factor VIII, to achieve a rise in plasma factor VIII activity to 40%–60%. Large joint hemarthrosis and life- or limb-threatening hemorrhage is treated initially with approximately 50 U/kg of factor VIII, targeting a rise to 100% factor VIII activity. Subsequent doses are determined according to the site and extent of bleeding and the clinical response. In circumstances of suboptimal clinical response, recent change in bleeding frequency, or comorbid illness, monitoring the plasma factor VIII activity response may be warranted. For most instances of non–life-threatening hemorrhage in experienced patients with moderate or severe hemophilia A, treatment can be administered at home, provided adequate intravenous access and close contact with the hemophilia clinician team can be achieved.

Prophylactic factor VIII infusions (eg, two or three times weekly) may prevent the development of arthropathy in severe hemophiliacs, and this approach is the standard of care in pediatric hemophilia.

image Prognosis

The development of safe and effective therapies for hemophilia A has resulted in improved long-term survival in recent decades. In addition, more aggressive management and the coordination of comprehensive care through hemophilia centers have greatly improved quality of life and level of function.

Blanchette VS: Meeting unmet needs in inhibitor patients. Haemophilia 2010;16(Suppl 3):46–51 [PMID: 20586802].

Gringeri A; ESPRIT Study Group: A randomized clinical trial of prophylaxis in children with hemophilia A (the ESPRIT Study). J Thromb Haemost 2011;9:700–710 [PMID: 21255253].

2. Factor IX Deficiency (Hemophilia B, Christmas Disease)

The mode of inheritance and clinical manifestations of factor IX deficiency are the same as those of factor VIII deficiency. Hemophilia B is 15%–20% as prevalent as hemophilia A. As in factor VIII deficiency, factor IX deficiency is associated with a prolonged aPTT, but the PT and thrombin time are normal. However, the aPTT is slightly less sensitive to factor IX deficiency than factor VIII deficiency. Diagnosis of hemophilia B is made by assaying factor IX activity, and severity is determined similarly to factor VIII deficiency. In general, clinical bleeding severity correlates less well with factor activity in hemophilia B than in hemophilia A.

The mainstay of treatment in hemophilia B is exogenous factor IX. Unlike factor VIII, about 50% of the administered dose of factor IX diffuses into the extravascular space. Therefore, 1 U/kg of plasma-derived factor IX concentrate or recombinant factor IX is expected to increase plasma factor IX activity by approximately 1%. Factor IX typically has a half-life of 20–22 hours in vivo, but because of variability, therapeutic monitoring may be warranted. As for factor VIII products, viral-inactivation techniques for plasma-derived factor IX concentrates appear effective in eradicating enveloped viruses. Only 1%–3% of persons with factor IX deficiency develop an inhibitor to factor IX, but patients may be at risk for anaphylaxis when receiving exogenous factor IX. The prognosis for persons with factor IX deficiency is comparable to that of patients with factor VIII deficiency. Gene therapy research efforts are ongoing for both hemophilias.

Aledort LM: Optimizing the treatment of haemophilia B: laboratory and clinical perspectives. Haemophilia 2010;16(Suppl 6): 1–2 [PMID: 20561350].

Santagostino E: Prophylaxis in haemophilia B patients: unresolved issues and pharmacoeconomic implications. Haemophilia 2010; 16(Suppl 60):13–17 [PMID: 20561353].

3. Factor XI Deficiency (Hemophilia C)

Factor XI deficiency is a genetic, autosomally transmitted coagulopathy, typically of mild to moderate clinical severity. Cases of factor XI deficiency account for less than 5% of all hemophilia patients. Homozygotes generally bleed at surgery or following severe trauma, but do not commonly have spontaneous hemarthroses. In contrast to factor VIII and IX deficiencies, factor XI activity is least predictive of bleeding risk. Although typically mild, pathologic bleeding may be seen in heterozygous individuals with factor XI activity as high as 60%. The aPTT is often considerably prolonged. In individuals with deficiency of both plasma and platelet-associated factor XI, the PFA-100 may also be prolonged. Management typically consists of perioperative prophylaxis and episodic therapy for acute hemorrhage. Treatment includes infusion of fresh frozen plasma (FFP); platelet transfusion may also be useful for acute hemorrhage in patients with deficiency of platelet-associated factor XI. Desmopressin has been used in some cases. The prognosis for an average life span in patients with factor XI deficiency is excellent.

Gomez K: Factor XI deficiency. Haemophilia 2008;14(6):1183–1189 [PMID: 18312365].

4. Other Inherited Bleeding Disorders

Other hereditary single clotting factor deficiencies are rare. Transmission is generally autosomal. Homozygous individuals with a deficiency or structural abnormality of prothrombin, factor V, factor VII, or factor X may have excessive bleeding.

Persons with dysfibrinogenemia (ie, structurally or functionally abnormal fibrinogen) may develop recurrent venous thromboembolic episodes or bleeding. Immunologic assay of fibrinogen is normal, but clotting assay may be low and the thrombin time prolonged. The PT and aPTT may be prolonged. Cryoprecipitate, which is rich in fibrinogen, has been the treatment of choice. Fibrinogen concentrates are available in the United States, but the clinical experience is limited

Afibrinogenemia resembles hemophilia clinically but has an autosomal recessive inheritance. Affected patients can experience a variety of bleeding manifestations, including mucosal bleeding, ecchymoses, hematomas, hemarthroses, and intracranial hemorrhage, especially following trauma. Fatal umbilical cord hemorrhage has been reported in neonates. The PT, aPTT, and thrombin time are all prolonged. A severely reduced fibrinogen concentration in an otherwise well child is confirmatory of the diagnosis. As in dysfibrinogenemia, fibrinogen concentrate or cryoprecipitate infusion is used for surgical prophylaxis and for acute hemorrhage.

Todd T: A review of long-term prophylaxis in the rare inherited coagulation factor deficiencies. Haemophilia 2010;16(4):569–583 [PMID: 19906159].



image Easy bruising and epistaxis from early childhood.

image Menorrhagia.

image Prolonged PFA-100 (or bleeding time), normal platelet count, absence of acquired platelet dysfunction.

image Reduced activity or abnormal structure of vWF.

image General Considerations

von Willebrand disease (vWD) is the most common inherited bleeding disorder among Caucasians, with a prevalence of 1%. vWF is a protein present as a multimeric complex in plasma, which binds factor VIII and is a cofactor for platelet adhesion to the endothelium. An estimated 70%–80% of all patients with vWD have classic vWD (type 1), which is caused by a partial quantitative deficiency of vWF. vWD type 2 involves a qualitative deficiency of (ie, dysfunctional) vWF, and vWD type 3 is characterized by a nearly complete deficiency of vWF. The majority (> 80%) of individuals with type 1 disease are asymptomatic. vWD is most often transmitted as an autosomal dominant trait, but it can be autosomal recessive. The disease can also be acquired, developing in association with hypothyroidism, Wilms tumor, cardiac disease, renal disease, or systemic lupus erythematosus, and in individuals receiving valproic acid. Acquired vWD is most often caused by the development of an antibody to vWF or increased turnover of vWF.

image Clinical Findings

A. Symptoms and Signs

A history of increased bruising and excessive epistaxis is often present. Prolonged bleeding also occurs with trauma or at surgery. Menorrhagia is often a presenting finding in females.

B. Laboratory Findings

PT is normal, and aPTT is sometimes prolonged. Prolongation of the PFA-100 or bleeding time is usually present since vWF plays a role in platelet adherence to endothelium. Platelet number may be decreased in type 2b vWD. Factor VIII and vWF antigen are decreased in types 1 and 3, but may be normal in type 2 vWD. vWF activity (eg, ristocetin cofactor or collagen binding) is decreased in all types. Since normal vWF antigen levels vary by blood type (type O normally has lower levels), blood type must be determined. Complete laboratory classification requires vWF multimer assay. The diagnosis requires confirmation of laboratory testing and bleeding history is often helpful when present.

image Treatment

The treatment to prevent or halt bleeding for most patients with vWD types 1 and 2 is desmopressin acetate, which causes release of vWF from endothelial stores. Desmopressin may be administered intravenously at a dose of 0.3 mcg/kg diluted in at least 20–30 mL of normal saline and given over 20–30 minutes. This dose typically elicits a three- to fivefold rise in plasma vWF. A high-concentration desmopressin nasal spray (150 mcg/spray), different than the preparation used for enuresis, may alternatively be used. Because response to vWF is variable among patients, factor VIII and vWF activities are typically measured before 60 minutes and 4 hours after infusion, to document the adequacy of the response. Desmopressin may cause fluid shifts, hyponatremia, and seizures; therefore, fluid restriction should be discussed, especially in children younger than 2 years. Because release of stored vWF is limited, tachyphylaxis often occurs with desmopressin.

If further therapy is indicated, vWF-replacement therapy (eg, plasma-derived concentrate) is recommended; such therapy is also used in patients with type 1 or 2a vWD who exhibit suboptimal laboratory response to desmopressin, and for all individuals with type 2b or 3 vWD. Antifibrinolytic agents (eg, ε-aminocaproic acid) may be useful for control of mucosal bleeding. Topical thrombin and fibrin glue may also be of benefit, although antibodies that inhibit these clotting proteins and therefore the affect of the glue have been described. Estrogen-containing contraceptive therapy may be helpful for menorrhagia.

image Prognosis

With the availability of effective treatment and prophylaxis for bleeding, life expectancy in vWD is normal.

Branchford BR, Di Paola J: Making a diagnosis of VWD. Hematology Am Soc Hematol Educ Program 2012;2012:161 [PMID: 23233576].

Halimeh S: Long-term secondary prophylaxis in children, adolescents and young adults with von Willebrand disease. Results of a cohort study. Thromb Haemost 2011 Apr;105(4):597–604 [PMID: 21301780].

Mikhail S: von Willebrand disease in the pediatric and adolescent population. J Pediatr Adolesc Gynecol 2010;23(6 Suppl):S3–S10 [PMID: 20934894].


1. Disseminated Intravascular Coagulation


image Presence of disorder known to trigger DIC.

image Evidence for consumptive coagulopathy (prolonged aPTT, PT, or thrombin time; increase in FSPs [fibrin-fibrinogen split products]; decreased fibrinogen or platelets).

image General Considerations

Disseminated intravascular coagulation (DIC) is an acquired pathologic process characterized by tissue factor–mediated coagulation activation in the host. DIC involves dysregulated, excessive thrombin generation, with consequent intravascular fibrin deposition and consumption of platelets and procoagulant factors. Microthrombi, composed of fibrin and platelets, may produce tissue ischemia and end-organ damage. The fibrinolytic system is frequently activated in DIC, leading to plasmin-mediated destruction of fibrin and fibrinogen; this results in fibrin-fibrinogen degradation products (FDPs) which exhibit anticoagulant and platelet-inhibitory functions. DIC commonly accompanies severe infection and other critical illnesses in infants and children. Conditions known to trigger DIC include endothelial damage (eg, endotoxin, virus), tissue necrosis (eg, burns), diffuse ischemic injury (eg, shock, hypoxia acidosis), and systemic release of tissue procoagulants (eg, certain cancers, placental disorders).

image Clinical Findings

A. Symptoms and Signs

Signs of DIC may include (1) signs of shock, often including end-organ dysfunction, (2) diffuse bleeding tendency (eg, hematuria, melena, purpura, petechiae, persistent oozing from needle punctures or other invasive procedures), and (3) evidence of thrombotic lesions (eg, major vessel thrombosis, purpura fulminans).

B. Laboratory Findings

Tests that are most sensitive, easiest to perform, most useful for monitoring, and best reflect the hemostatic capacity of the patient are the PT, aPTT, platelet count, fibrinogen, and FSPs. The PT and aPTT are typically prolonged and the platelet count and fibrinogen concentration may be decreased. However, in children, the fibrinogen level may be normal until late in the course. Levels of FSPs are increased, and elevated levels of D-dimer, a cross-linked fibrin degradation byproduct, may be helpful in monitoring the degree of activation of both coagulation and fibrinolysis. However, D-dimer is nonspecific and may be elevated in the context of a triggering event (eg, severe infection) without concomitant DIC. Often, physiologic inhibitors of coagulation, especially antithrombin III and protein C, are consumed, predisposing to thrombosis. The specific laboratory abnormalities in DIC may vary with the triggering event and the course of illness.

image Differential Diagnosis

DIC can be difficult to distinguish from coagulopathy of liver disease (ie, hepatic synthetic dysfunction), especially when the latter is associated with thrombocytopenia secondary to portal hypertension and hypersplenism. Generally, factor VII activity is decreased markedly in liver disease (due to deficient synthesis of this protein, which has the shortest half-life among the procoagulant factors), but only mildly to moderately decreased in DIC (due to consumption). Factor VIII activity is often normal or even increased in liver disease, but decreased in DIC.

image Treatment

A. Therapy for Underlying Disorder

The most important aspect of therapy in DIC is the identification and treatment of the triggering event. If the pathogenic process underlying DIC is reversed, often no other therapy is needed for the coagulopathy.

B. Replacement Therapy for Consumptive Coagulopathy

Replacement of consumed procoagulant factors with FFP and of platelets via platelet transfusion is warranted in the setting of DIC with hemorrhagic complications, or as periprocedural bleeding prophylaxis. Infusion of 10–15 mL/kg FFP typically raises procoagulant factor activities by approximately 10%–15%. Cryoprecipitate can also be given as a rich source of fibrinogen; one bag of cryoprecipitate per 3 kg in infants or one bag of cryoprecipitate per 6 kg in older children typically raises plasma fibrinogen concentration by 75–100 mg/dL.

C. Anticoagulant Therapy for Coagulation Activation

Continuous intravenous infusion of unfractionated heparin is sometimes given in order to attenuate coagulation activation and consequent consumptive coagulopathy. The rationale for heparin therapy is to maximize the efficacy of, and minimize the need for, replacement of procoagulants and platelets; however, clinical evidence demonstrating benefit of heparin in DIC is lacking. Prophylactic doses of unfractionated heparin or low-molecular-weight heparin (LMWH) in critically ill and nonbleeding patients with DIC may be considered for prevention of venous thromboembolism. Unfractionated heparin dosing and monitoring is listed on page 979.

D. Specific Factor Concentrates

A nonrandomized pilot study of antithrombin concentrate in children with DIC and associated acquired antithrombin deficiency demonstrated favorable outcomes, suggesting that replacement of this consumed procoagulant may be of benefit. Protein C concentrate has also shown promise in two small pilot studies of meningococci-associated DIC with purpura fulminans. Activated protein C has reduced mortality in septic adults in a large randomized multicenter trial; in additional studies, efficacy in adults and pediatrics is increasingly controversial.

2. Liver Disease

The liver is the major synthetic site of prothrombin, fibrinogen, high-molecular-weight kininogen, and factors V, VII, IX, X, XI, XII, and XIII. In addition, plasminogen and the physiologic anticoagulants (antithrombin III, protein C, and protein S) are hepatically synthesized. α2-Antiplasmin, a regulator of fibrinolysis, is also produced in the liver. Deficiency of factor V and the vitamin K–dependent factors (II, VII, IX, and X) is most often a result of decreased hepatic synthesis and is manifested by a prolonged PT and often a prolonged aPTT. Extravascular loss and increased consumption of clotting factors may contribute to PT and aPTT prolongation. Fibrinogen production is often decreased, or an abnormal fibrinogen (dysfibrinogen) containing excess sialic acid residues may be synthesized, or both. Hypofibrinogenemia or dysfibrinogenemia is associated with prolongation of thrombin time and reptilase time. FDPs and D-dimers may be present because of increased fibrinolysis, particularly in the setting of chronic hepatitis or cirrhosis. Thrombocytopenia secondary to hypersplenism may occur. DIC and coagulopathy of liver disease also mimic vitamin K deficiency; however, vitamin K deficiency has normal factor V activity. Treatment of acute bleeding in the setting of coagulopathy of liver disease consists of replacement with FFP and platelets. Desmopressin may shorten the bleeding time and aPTT in patients with chronic liver disease, but its safety has not been well established. Recombinant VIIa also is efficacious for life-threatening hemorrhage.

3. Vitamin K Deficiency

The newborn period is characterized by physiologically depressed activity of the vitamin K–dependent factors (II, VII, IX, and X). If vitamin K is not administered at birth, a bleeding diathesis previously called hemorrhagic disease of the newborn, now termed vitamin K deficiency bleeding (VKDB), may develop. Outside of the newborn period, vitamin K deficiency may occur as a consequence of inadequate intake, excess loss, inadequate formation of active metabolites, or competitive antagonism.

One of three patterns is seen in the neonatal period:

1. Early VKDB of the newborn occurs within 24 hours of birth and is most often manifested by cephalohematoma, intracranial hemorrhage, or intra-abdominal bleeding. Although occasionally idiopathic, it is most often associated with maternal ingestion of drugs that interfere with vitamin K metabolism (eg, warfarin, phenytoin, isoniazid, and rifampin). Early VKDB occurs in 6%–12% of neonates born to mothers who take these medications without receiving vitamin K supplementation. The disorder is often life threatening.

2. Classic VKDB occurs at 24 hours to 7 days of age and usually is manifested as gastrointestinal, skin, or mucosal bleeding. Bleeding after circumcision may occur. Although occasionally associated with maternal drug usage, it most often occurs in well infants who do not receive vitamin K at birth and are solely breast fed.

3. Late neonatal VKDB occurs on or after day 8. Manifestations include intracranial, gastrointestinal, or skin bleeding. This disorder is often associated with fat malabsorption (eg, in chronic diarrhea) or alterations in intestinal flora (eg, with prolonged antibiotic therapy). Like classic VKDB, late VKDB occurs almost exclusively in breast-fed infants.

The diagnosis of vitamin K deficiency is suspected based on the history, physical examination, and laboratory results. The PT is prolonged out of proportion to the aPTT (also prolonged). The thrombin time becomes prolonged late in the course. The platelet count is normal. This laboratory profile is similar to the coagulopathy of acute liver disease, but with normal fibrinogen level and absence of hepatic transaminase elevation. The diagnosis of vitamin K deficiency is confirmed by a demonstration of noncarboxylated proteins in the absence of vitamin K in the plasma and by clinical and laboratory responses to vitamin K. Intravenous or subcutaneous treatment with vitamin K should be given immediately and not withheld while awaiting test results. In the setting of severe bleeding, additional acute treatment with FFP or recombinant factor VIIa may be indicated.

4. Uremia

Uremia is frequently associated with acquired platelet dysfunction. Bleeding occurs in approximately 50% of patients with chronic renal failure. The bleeding risk conferred by platelet dysfunction associated with metabolic imbalance may be compounded by decreased vWF activity and procoagulant deficiencies (eg, factor II, XII, XI, IX) due to increased urinary losses of these proteins in some settings of renal insufficiency. In accordance with platelet dysfunction, uremic bleeding is typically characterized by purpura, epistaxis, menorrhagia, or gastrointestinal hemorrhage. Acute bleeding may be managed with infusion of desmopressin acetate, factor VIII concentrates containing vWF, or cryoprecipitate with or without coadministration of FFP. Red cell transfusion may be required. Prophylactic administration of erythropoietin before the development of severe anemia appears to decrease the potential for bleeding. Recombinant VIIa may be useful in refractory bleeding.

Levi M: Guidelines for the diagnosis and management of disseminated intravascular coagulation. British Committee for Standards in Haematology. Br J Haematol 2009;145:24–33 [PMID: 19222477].

Shearer MJ: Vitamin K deficiency bleeding (VKDB) in early infancy. Blood Rev 2009;23:49–59 [PMID: 18804903].

Witmer CM: Off-label recombinant factor VIIa use and thrombosis in children: a multi-center cohort study. J Pediatr 2011;158:820–825 [PMID: 21146180].


1. Henoch-Schönlein Purpura (Anaphylactoid Purpura)


image Purpuric cutaneous rash.

image Migratory polyarthritis or polyarthralgias.

image Intermittent abdominal pain.

image Nephritis.

image General Considerations

Henoch-Schönlein purpura (HSP), the most common type of small vessel vasculitis in children, primarily affects boys 2–7 years of age. Occurrence is highest in the spring and fall, and upper respiratory infection precedes the diagnosis in two-thirds of children.

Leukocytoclastic vasculitis in HSP principally involves the small vessels of the skin, gastrointestinal tract, and kidneys, with deposition of IgA immune complexes. The most common and earliest symptom is palpable purpura, which results from extravasation of erythrocytes into the tissue surrounding the involved venules. Antigens from group A β-hemolytic streptococci and other bacteria, viruses, drugs, foods, and insect bites have been proposed as inciting agents.

image Clinical Findings

A. Symptoms and Signs

Skin involvement may be urticarial initially, progresses to a maculopapules, and coalesces to a symmetrical, palpable purpuric rash distributed on the legs, buttocks, and elbows. New lesions may continue to appear for 2–4 weeks, and may extend to involve the entire body. Two-thirds of patients develop migratory polyarthralgias or polyarthritis, primarily of the ankles and knees. Intermittent, sharp abdominal pain occurs in approximately 50% of patients, and hemorrhage and edema of the small intestine can often be demonstrated. Intussusception may develop. From 25% to 50% of those affected develop renal involvement in the second or third week of illness with either a nephritic or, less commonly, nephrotic picture. Hypertension may accompany the renal involvement. In males, testicular torsion may also occur, and neurologic symptoms are possible due to small vessel vasculitis.

B. Laboratory Findings

The platelet count is normal or elevated, and other screening tests of hemostasis and platelet function are typically normal. Urinalysis frequently reveals hematuria, and sometimes proteinuria. Stool may be positive for occult blood. The antistreptolysin O (ASO) titer is often elevated and the throat culture positive for group A β-hemolytic streptococci. Serum IgA may be elevated.

image Differential Diagnosis

The rash of septicemia (especially meningococcemia) may be similar to skin involvement in HSP, although the distribution tends to be more generalized. The possibility of trauma should be considered in any child presenting with purpura. Other vasculitides should also be considered. The lesions of thrombotic thrombocytopenic purpura (TTP) are not palpable.

image Treatment

Generally, treatment is supportive. NSAIDs may be useful for the arthritis. Corticosteroid therapy may provide symptomatic relief for severe gastrointestinal or joint manifestations but does not alter skin or renal manifestations. If culture for group A β-hemolytic streptococci is positive or if the ASO titer is elevated, a therapeutic course of penicillin is warranted.

image Prognosis

The prognosis for recovery is generally good, although symptoms frequently (25%–50%) recur over a period of several months. In patients who develop renal manifestations, microscopic hematuria may persist for years. Progressive renal failure occurs in fewer than 5% of patients with HSP, with an overall fatality rate of 3%.

McCarthy HJ: Clinical practice: diagnosis and management of Henoch-Schönlein purpura. Eur J Pediatr 2010;169:643–650 [PMID: 20012647].

2. Collagen Disorders

Mild to life-threatening bleeding occurs with some types of Ehlers-Danlos syndrome, the most common inherited collagen disorder. Ehlers-Danlos syndrome is characterized by joint hypermobility, skin extensibility, and easy bruising. Coagulation abnormalities may sometimes be present, including platelet dysfunction and deficiencies of coagulation factors VIII, IX, XI, and XIII. However, bleeding and easy bruising, in most instances, relates to fragility of capillaries and compromised vascular integrity. Individuals with Ehlers-Danlos syndrome types 4 and 6 are at risk for aortic dissection and spontaneous rupture of aortic aneurysms. Surgery should be avoided for patients with Ehlers-Danlos syndrome, as should medications that induce platelet dysfunction.

Key NS, DE Paepe A, Malfait F, Shovlin CL: Vascular hemostasis. Haemophilia 2010 Jul;16(Suppl 5):146 [PMID: 20590874].


image General Considerations

Although uncommon in children, thrombotic disorders are being recognized with increasing frequency, particularly with heightened physician awareness and improved survival in pediatric intensive care settings. Several clinical conditions have a potential association with thrombotic events in childhood (see the next section Clinical Risk Factors).

image Clinical Findings

Initial evaluation of the child who has thrombosis includes an assessment for potential triggering factors, as well as a family history of thrombosis and early cardiovascular or cerebrovascular disease.

A. Clinical Risk Factors

Clinical risk factors are present in more than 90% of children with acute venous thromboembolism (VTE). These conditions include the presence of an indwelling vascular catheter, cardiac disease, infection, trauma, surgery, immobilization, collagen-vascular or chronic inflammatory disease, renal disease, sickle cell anemia, and malignancy. Prospective findings employing serial radiologic evaluation as screening indicate that the risk of VTE is nearly 30% for short-term central venous catheters placed in the internal jugular veins.

Retrospective data suggest that approximately 8% of children with cancer develop symptomatic VTE.

1. Inherited Thrombophilia (Hypercoagulable) States

A. PROTEIN C DEFICIENCYProtein C is a vitamin K–dependent protein that is normally activated by thrombin bound to thrombomodulin and inactivates activated factors V and VIII. In addition, activated protein C promotes fibrinolysis. Two phenotypes of hereditary protein C deficiency exist. Heterozygous individuals with autosomal dominant protein C deficiency often present with VTE as young adults, but the disorder may manifest during childhood or in later adulthood. In mild protein C deficiency, anticoagulant prophylaxis is typically limited to periods of increased prothrombotic risk. Homozygous or compound heterozygous protein C deficiency is rare but phenotypically severe. Affected children generally present within the first 12 hours of life with purpura fulminans (Figure 30–6) and/or VTE. Prompt protein C replacement by infusion of protein C concentrate or (if unavailable) FFP every 6–12 hours, along with therapeutic heparin administration, is recommended. Subsequent management requires chronic anticoagulation with warfarin, sometimes along with routine protein C concentrate infusion in order to permit lower warfarin dosing. Recurrent VTE is common, especially during periods of subtherapeutic anticoagulation or in the presence of conditions associated with increased prothrombotic risk.


image Figure 30–6. Purpura fulminans in an infant with severe protein C deficiency.

B. PROTEIN S DEFICIENCYProtein S is a cofactor for protein C. Neonates with homozygous protein S deficiency have a course similar to those with homozygous or compound heterozygous protein C deficiency. Lifelong warfarin therapy is indicated in homozygous/severe deficiency, or in heterozygous individuals who have experienced recurrent VTE. Efforts must be made to distinguish these conditions from acquired deficiency, which can be antibody-mediated or secondary to an increase in C4b-binding protein induced by inflammation.

C. ANTITHROMBIN DEFICIENCYAntithrombin, which is the most important physiologic inhibitor of thrombin, inhibits activated factors IX, X, XI, and XII. Antithrombin deficiency is transmitted in an autosomal dominant pattern and is associated with VTE, typically with onset in adolescence or young adulthood. Therapy for acute VTE involves replacement with antithrombin concentrate and therapeutic anticoagulation. The efficiency of heparin may be significantly diminished in the setting of severe antithrombin deficiency and it often requires supplementation of antithrombin via concentrate. Patients with homozygous/severe deficiency or recurrent VTE are maintained on lifelong warfarin.

D. FACTOR V LEIDEN MUTATIONAn amino acid substitution in the gene coding for factor V results in factor V Leiden, a factor V polymorphism that is resistant to inactivation by activated protein C. The most common cause of activated protein C resistance in Caucasians, factor V Leiden is present in approximately 5% of the Caucasian population, 20% of Caucasian adults with deep vein thrombosis, and 40%–60% of those with a family history of VTE. VTE occurs in both heterozygous and homozygous individuals. In the former case, thrombosis is typically triggered by a clinical risk factor (or else develops in association with additional thrombophilia traits), whereas in the latter case, it is often spontaneous. Population studies suggest that the risk of incident VTE is increased two- to sevenfold in the setting of heterozygous factor V Leiden, 35-fold among heterozygous individuals taking the oral contraceptives, and 80-fold in those homozygous for factor V Leiden.

E. PROTHROMBIN MUTATIONThe 20210 glutamine to alanine mutation in the prothrombin gene is a relatively common polymorphism in Caucasians that enhances its activation to thrombin. In heterozygous form, this mutation is associated with a two- to threefold increased risk for incident VTE. This mutation also appears to modestly increase the risk for recurrent VTE.

F. OTHER INHERITED DISORDERSQualitative abnormalities of fibrinogen (dysfibrinogenemias) are usually inherited in an autosomal dominant manner. Most individuals with dysfibrinogenemia are asymptomatic. Some patients experience bleeding, while others develop venous or arterial thrombosis. The diagnosis is suggested by a prolonged thrombin time with a normal fibrinogen concentration. Hyperhomocysteinemia can be an inherited or an acquired condition and is associated with an increased risk for both arterial and venous thromboses. In children, it may also serve as a risk factor for ischemic arterial stroke. Hyperhomocysteinemia is quite uncommon in the setting of dietary folate supplementation (as in the United States) and is observed almost uniquely in cases of renal insufficiency or metabolic disease (eg, homocystinuria). Methylene tetrahydrofolate reductase receptor mutations do not appear to constitute a risk factor for thrombosis in US children if homocysteine is not elevated.

Lipoprotein(a) is a lipoprotein with homology to plasminogen. In vitro studies suggest that lipoprotein(a) may both promote atherothrombosis and inhibit fibrinolysis. Some evidence to date suggests that elevated plasma concentrations of lipoprotein(a) are associated with an increased risk of VTEs and ischemic arterial stroke in children.

Increased factor VIII activity is a risk factor for incident VTE and is common among children with acute VTE. Elevation in factor VIII may persist in the long-term follow-up of these patients, and can be inherited.

2. Acquired Disorders

A. ANTIPHOSPHOLIPID ANTIBODIESThe development of antiphospholipid antibodies is the most common form of acquired thrombophilia in children. Antiphospholipid antibodies, which include the lupus anticoagulant, anticardiolipin antibodies, and β2-glycoprotein-1 antibodies (among others) are common in acute childhood VTE. The lupus anticoagulant is demonstrated in vitro by its inhibition of phospholipid-dependent coagulation assays (eg, aPTT and dilute Russell viper venom time), whereas immunologic techniques (eg, enzyme-linked immunosorbent assays) are often used to detect anticardiolipin and β2-glycoprotein-1 antibodies. Although common in persons with autoimmune diseases such as systemic lupus erythematosus, antiphospholipid antibodies may also develop following certain drug exposures, infection, acute inflammation, and lymphoproliferative diseases. Sometimes VTE and antiphospholipid antibodies may predate other signs of lupus for long periods of time. Viral illness is a common precipitant in children, and in many cases, the inciting infection may be asymptomatic.

If an antiphospholipid antibody persists for 12 weeks following the acute thrombotic event, the diagnosis of this syndrome is confirmed. Optimal duration of anticoagulation in this setting is unclear, such that current pediatric treatment guidelines recommend a 3-month to lifelong course.

B. DEFICIENCIES OF INTRINSIC ANTICOAGULANTSAcquired deficiencies of proteins C and S and antithrombin may occur in the clinical context of antibodies (eg, protein S antibodies in acute varicella) or in excessive consumption, including sepsis, DIC, major-vessel or extensive VTE, and post–bone marrow transplant sinusoidal obstruction syndrome (formerly termed hepatic veno-occlusive disease). Pilot studies in children have suggested a possible therapeutic role for antithrombin or protein C concentrates in sepsis-associated DIC (eg, meningococcemia) and severe posttransplant sinusoidal obstruction syndrome.

C. ACUTE PHASE REACTANTSAs part of the acute phase response, elevations in plasma fibrinogen concentration, plasma factor VIII, and platelet count may occur, all of which may contribute to an acquired prothrombotic state.

Reactive thrombocytosis is rarely associated with VTEs in children when the platelet count is less than 1 million/μL.

B. Symptoms and Signs

Presenting features of thrombosis vary with the anatomic site, extent of vascular involvement, degree of vaso-occlusion, and presence of end-organ dysfunction. The classic presentation of deep venous thrombosis of an upper or lower extremity is painful acute or subacute extremity swelling, while that for pulmonary embolism commonly involves dyspnea and pleuritic chest pain, and in cerebral sinovenous thrombosis (CSVT) often includes severe or persistent headache, with or without neurologic deficit in otherwise well children. It is not infrequently preceded by sign/symptoms of otitis media progressing to mastoiditis. Arterial thrombosis of the lower extremity (eg, neonatal umbilical artery catheter–associated) as well as vasospasm without identified thrombosis, often manifests with diminished distal pulses and dusky discoloration of the limb.

C. Laboratory Findings

A comprehensive laboratory investigation for thrombophilia (ie, hypercoagulability) is recommended by the International Society on Thrombosis and Haemostasis in order to disclose possible underlying congenital or acquired abnormalities that may affect acute or long-term management. Testing evaluates for intrinsic anticoagulant deficiency (proteins C and S and antithrombin), procoagulant factor excess (eg, factor VIII), proteins and genetic mutations mediating enhanced procoagulant activity or reduced sensitivity to inactivation (antiphospholipid antibodies; factor V Leiden and prothrombin 20210 polymorphisms), biochemical mediators of endothelial damage (homocysteine), and markers or regulators of fibrinolysis (eg, D-dimer, plasminogen activator inhibitor-1, and lipoprotein[a]). Interpretation of procoagulant factor and intrinsic anticoagulant levels should take into account the age dependence of normal values for these proteins. Among these VTE risk factors, antiphospholipid antibodies and elevated levels of homocysteine and lipoprotein(a) have also been demonstrated as risk factors for arterial thrombotic and ischemic events.

D. Imaging

Appropriate radiologic imaging is essential to objectively document the thrombus and to delineate the type (venous vs arterial), occlusiveness, and extent (proximal and distal termini) of thrombosis. Depending on site, typical imaging modalities include compression ultrasound with Doppler, computed tomographic (CT) venography, magnetic resonance venography, and conventional angiography.

image Treatment

Current guidelines for the treatment of first-episode VTE in children have been largely based on adult experience and include therapeutic anticoagulation for at least 3 months. During the period of anticoagulation, bleeding precautions should be followed, as previously described (see Treatment under Idiopathic Thrombocytopenic Purpura, earlier); in CVST, presentation with postthrombotic hemorrhage often does not preclude anticoagulant therapy. Initial therapy for acute VTE employs continuous intravenous unfractionated heparin or subcutaneous injections of LMWH) for at least 7 days, monitored by anti-Xa activity level to maintain anticoagulant levels of 0.3–0.7 or 0.5–1.0 IU/mL, respectively. Subsequent extended anticoagulant therapy is given with LMWH or daily oral warfarin, the latter agent monitored by the PT to maintain an international normalized ratio (INR) of 2.0–3.0 (2.5–3.5 in the presence of an antiphospholipid antibody). During warfarin treatment, the INR optimally should be within the therapeutic range for 2 consecutive days before discontinuation of heparin. Warfarin pharmacokinetics are affected by acute illness, numerous medications, and changes in diet, and can necessitate frequent monitoring. In children, warfarin dose is determined by age. LMWH offers the advantage of infrequent need for monitoring but is far more expensive than warfarin. Anatomic contributions to venous stasis (eg, mastoiditis or depressed skull fracture as risk factors for CVST; congenital left iliac vein stenosis in DVT of proximal left lower extremity with May-Thurner anomaly) should be addressed to optimize response to anticoagulation. In cases of limb- or life-threatening VTEs, including major proximal pulmonary embolus, and in cases of progressive VTE despite therapeutic anticoagulation, thrombolytic therapy (eg, tissue-type plasminogen activator) may be considered. A recent cohort study has indicated that initial thrombolytic therapy may also reduce the risk of the postthrombotic syndrome (PTS) in children with veno-occlusive deep venous thrombosis of the proximal limbs in whom adverse prognostic biomarkers (ie, elevated factor VIII and D-dimer levels) are present at diagnosis; however, the safety and efficacy of this approach must be further evaluated in larger studies. In adolescent females, estrogen-containing contraceptives are relatively contraindicated in those with prior VTE, particularly if an additional genetic cause for impairment of protein C pathway is disclosed.

image Prognosis

Registries and cohort studies have suggested that recurrent VTE occurs in approximately 10% of children within 2 years. Persistent thrombosis is evident following completion of a standard therapeutic course of anticoagulation in up to 30% of children, with unclear clinical importance. Approximately one in four children with deep venous thrombosis involving the extremities develop post thrombotic syndrome, a condition of venous insufficiency of varying severity characterized by chronic skin changes, edema, and dilated collateral superficial venous formation, and often accompanied by functional limitation (pain with activities or at rest). The presence of homozygous anticoagulant deficiencies, multiple thrombophilia traits, or persistent antiphospholipid antibodies following VTE diagnosis has been associated with increased risk of recurrent VTE, leading to consideration of extended anticoagulation in these instances. Complete veno-occlusion and elevated levels of factor VIII and D-dimer at VTE diagnosis have been identified as prognostic factors for PTS among children with deep venous thrombosis affecting the limbs. In CSVT failure to provide antithrombotic therapy has been associated with adverse neurologic outcome.

Goldenberg NA: Definition of post-thrombotic syndrome following lower extremity deep venous thrombosis and standardization of outcome measurement in pediatric clinical investigations. J Thromb Haemost 2012;10:477 [PMID: 22482118].

Goldenberg NA: The “age” of understanding VKA dose. Blood 2010;116:5789–5790 [PMID: 21183695].

Monagle P: Antithrombotic therapy in neonates and children: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines (8th Edition). Chest 2012 Feb;141 (2 Suppl):e737S [PMID: 22315277]

Raffini L: Dramatic increase in venous thromboembolism in children’s hospitals in the United States from 2001 to 2007. Pediatrics 2009;124:1001–1008 [PMID: 19736261].



The differential diagnosis of splenomegaly includes the general categories of congestive splenomegaly, chronic infections, leukemia and lymphomas, hemolytic anemias, reticuloendothelioses, and storage diseases (Table 30–9).

Table 30–9. Causes of chronic splenomegaly in children.


Splenomegaly due to any cause may be associated with hypersplenism and the excessive destruction of circulating red cells, white cells, and platelets. The degree of cytopenia is variable and, when mild, requires no specific therapy. In other cases, the thrombocytopenia may cause life-threatening bleeding, particularly when the splenomegaly is secondary to portal hypertension and associated with esophageal varices or the consequence of a storage disease. In such cases, treatment with surgical splenectomy or with splenic embolization may be warranted. Although more commonly associated with acute enlargement, rupture of an enlarged spleen can be seen in more chronic conditions such as Gaucher disease.

Stone DL: Life threatening splenic hemorrhage in two patients with Gaucher’s disease. Am J Hematol 2000;64:140 [PMID: 10814997].


Children who lack normal splenic function are at risk for sepsis, meningitis, and pneumonia due to encapsulated bacteria such as pneumococci and H influenzae. Such infections are often fulminant and fatal because of inadequate antibody production and impaired phagocytosis of circulating bacteria.

Congenital asplenia is usually suspected when an infant is born with abnormalities of abdominal viscera and complex cyanotic congenital heart disease. Howell-Jolly bodies are usually present on the peripheral blood smear, and the absence of splenic tissue is confirmed by technetium radionuclide scanning. The prognosis depends on the underlying cardiac lesions, and many children die during the first few months. Prophylactic antibiotics, usually penicillin, and pneumococcal conjugate, Hib, and meningococcal vaccines, are recommended.

The risk of overwhelming sepsis following surgical splenectomy is related to the child’s age and to the underlying disorder. Because the risk is highest when the procedure is performed earlier in life, splenectomy is usually postponed until after age 5 years. The risk of postsplenectomy sepsis is also greater in children with malignancies, thalassemias, and reticuloendothelioses than in children whose splenectomy is performed for ITP, hereditary spherocytosis, or trauma. Prior to splenectomy, children should be immunized against Streptococcus pneumoniaeH influenzae, and Neisseria meningitidis. Additional management should include penicillin prophylaxis and prompt evaluation for fever 38.8°C or above or signs of severe infection.

Children with sickle cell anemia develop functional asplenia during the first year of life, and overwhelming sepsis is the leading cause of early deaths in this disease. Prophylactic penicillin reduces the incidence of sepsis by 84%.

Pickering L: American Academy of Pediatrics: immunization in special circumstances. Red Book. Am Acad Pediat 2012; 74–90.



Minimizing the risks of transfusion begins by screening volunteer donors with a questionnaire that will protect the recipient from transmission of infectious agents as well as other risks of transfusions. In addition, information defining high-risk groups whose behavior increases the possible transmission of HIV, hepatitis, and other diseases is provided, with the request that persons in these groups not donate blood. Positive responses may result in temporary or permanent deferral from donation.

Before blood components can be released for transfusion, donor blood is screened for hepatitis B surface antigen; antibodies to hepatitis B core antigen, hepatitis C, HIV-1 and 2, and human T-cell lymphotropic virus (HTLV) I and II; and a serologic test for syphilis (Table 30–10). Screening donor blood for viral genome (nucleic acid amplification [NAT] testing) was mandated for HIV, HCV, and West Nile virus. NAT testing for other viruses may be added in the future, and screening for Chagas disease was accepted in 2007. Positive tests are repeated.

Table 30–10. Transmission risks of infectious agents for which screening of blood products is routinely performed.


Upon confirmation of a screening test, the unit in question is destroyed and the donor is notified and deferred from future donations. Many of the screening tests used are very sensitive and have a high rate of false-positive results. As a result, confirmatory tests have been developed to check the initial screening results and separate the false-positives from the true-positives, allowing for donors with repeat reactive screening tests on a specific donation to be reentered into the donor pool in the future if they meet other requirements. Recently, bacterial culture of platelet concentrates was added to the testing paradigm.

With these approaches, the risk of an infectious complication from blood components has been minimized (see Table 30–10). Autologous donation is recognized by some centers as a safe alternative to homologous blood. Issues of donor size make the techniques of autologous donation difficult to apply to the pediatric population.

Primary CMV infections are significant complications of blood transfusion in transplant recipients, neonates, and immunodeficient individuals. Transmission of CMV can be avoided by using seronegative donors, apheresis platelet concentrates collected by techniques ensuring low numbers of residual white cells, or red cell or platelet products leukocyte-depleted by filtering (< 5 million WBCs per packed red cell unit or apheresis platelet concentrate equivalent).


Whole blood is routinely fractionated into packed red cells, platelets, and FFP or cryoprecipitate for most efficient use of all blood components. The storage conditions and biologic characteristics of the fractions are summarized in Table 30–11. The conditions provide the optimal environment to maintain appropriate recovery, survival, and function, and are different for each blood component. For example, red cells undergo dramatic metabolic changes during their 35- to 42-day storage, with a virtual disappearance of 2,3-DPG by day 14 of storage, a decrease in adenosine triphosphate, a gradual loss of intracellular potassium, but in vivo recovery of greater than or equal to 80% transfused cells in red cells by day of storage outdate. Fortunately, these changes are reversed readily in vivo within hours to days after the red cells are transfused. However, in certain clinical conditions, these effects may define the type of components used. For example, blood less than 7–10 days old would be preferred for exchange transfusion in neonates, red cell exchanges in older patients, or replacement of red cells in persons with severe cardiopulmonary disease to ensure adequate oxygen-carrying capacity. Storage time is not an issue when administering transfusions to those with chronic anemia.

Table 30–11. Characteristics of blood and blood components.




If extracellular potassium in older packed red cells may present a problem, one may use blood less than 10 days old, making packed cells out of an older unit of whole blood or washing blood stored as packed cells. Regardless of the blood’s age, greater than 80% of the red cells will circulate after transfusion and approximate normal survival in the circulation.

Platelets are stored at 22°C for a maximum of 5 days; criteria for 7-day storage are being developed. At the extremes of storage, there should be at least a 60% recovery, a survival time that approximates turnover of fresh autologous platelets, and normalization of the bleeding time or PFA-100 in proportion to the peak platelet count. Frozen components, red cells, FFP, and cryoprecipitate are outdated at 10 years, 1 year, and 1 year, respectively. Frozen red cells retain the same biochemical and functional characteristics as the day they were frozen. FFP contains 80% or more of all of the clotting factors of fresh plasma. Factors VIII and XIII, vWF, and fibrinogen are concentrated in cryoprecipitate.


Both the donated blood and the recipient are tested for ABO and Rh(D) antigens and screened for auto- or alloantibodies in the plasma. The cross-match is required on any component that contains red cells. In the major cross-match, washed donor red cells are incubated with the serum from the patient, and agglutination is detected and graded. The antiglobulin phase of the test may then be performed; Coombs reagent, which will detect the presence of IgG or complement on the surface of the red cells, is added to the mixture, and agglutination is evaluated. In the presence of a negative antibody screen in the recipient, a negative immediate spin cross-match test confirms the compatibility of the blood and antiglobulin phase is not required. Further testing is required if the antibody screen or the cross-match is positive, and blood should not be given until the nature of the reactivity is delineated. An incompatible cross-match is evaluated first with a DAT or Coombs test to detect IgG or complement on the surfaces of the recipient’s red cells. The indirect antiglobulin test is also used to determine the presence of antibodies that will coat red cells or activate complement, and additional studies are completed to define the antibody.


General Rules

Several rules should be observed in administering any blood component:

1. In final preparation of the component, no solutions should be added to the bag or tubing set except for normal saline (0.9% sodium chloride for injection, USP), ABO-compatible plasma, or other specifically approved expanders. Hypotonic solutions cause hemolysis of red cells, and, if these are transfused, a severe reaction will occur. Any reconstitution should be completed by the blood bank.

2. Transfusion products should be protected from contact with any calcium-containing solution (eg, lactated Ringer); recalcification and reversal of the citrate effect will cause clotting of the blood component.

3. Blood components should not be warmed to a temperature greater than 37°C. If a component is incubated in a water bath, it should be enclosed in a watertight bag to prevent bacterial contamination of entry ports.

4. Whenever a blood bag is entered, the sterile integrity of the system is violated, and that unit should be discarded within 4 hours if left at room temperature or within 24 hours if the temperature is 4–6°C.

5. Transfusions of products containing red cells should not exceed 4 hours. Blood components in excess of what can be infused during this time period should be stored in the blood bank until needed.

6. Before transfusion, the blood component should be inspected visually for any unusual characteristics, such as the presence of flocculent material, hemolysis, or clumping of cells, and mixed thoroughly.

7. The unit and the recipient should be identified properly.

8. The administration set includes a standard 170- to 260-μm filter. Under certain clinical circumstances, an additional microaggregate filter may be used to eliminate small aggregates of fibrin, white cells, and platelets that will not be removed by the standard filter.

9. The patient should be observed during the entire transfusion and especially during the first 15 minutes. With the onset of any adverse symptoms or signs the transfusion should be stopped, an evaluation initiated immediately and the reaction reported promptly to the transfusion service.

10. When cross-match–incompatible red cells or whole blood unit(s) must be given to the patient (as with AIHA), a test dose of 10% of the total volume (not to exceed 50 mL) should be administered over 15–20 minutes; the transfusion is then stopped and the patient observed. If no changes in vital signs or the patient’s condition are noted, the remainder of the volume can be infused carefully.

11. Blood for exchange transfusion in the newborn period may be cross-matched with either the infant’s or the mother’s serum. If the exchange is for hemolysis, 500 mL of whole blood stored for less than 7 days will be adequate. If replacement of clotting factors is a key issue, packed red cells (7 days old) reconstituted with ABO type-specific FFP may be considered. Based on posttransfusion platelet counts, platelet transfusion may be considered. Other problems to be anticipated are acid-base derangements, hyponatremia, hyperkalemia, hypocalcemia, hypoglycemia, hypothermia, and hypervolemia or hypovolemia.

Choice of Blood Component

Several principles should be considered when deciding on the need for blood transfusion. Indications for blood or blood components must be well defined, and the patient’s medical condition, not just the laboratory results, should be the basis for the decision. Specific deficiencies exhibited by the patient (eg, oxygen-carrying capacity, thrombocytopenia) should be treated with appropriate blood components and the use of whole blood minimized. Information about specific blood components is summarized in Table 30–11. In general, very little is known about specific indications for blood component transfusion and outcomes. A recent review evaluates what is known and presents fertile areas for investigation (see Josephson, et al).

A. Whole Blood

Whole blood may be used in patients who require replacement of oxygen-carrying capacity and volume. More specifically, it may be considered during massive blood loss, after initial response to replace volume with crystalloid and oxygen carrying capacity. Doses vary depending on volume considerations (see Table 30–11). In acute situations, the transfusion may be completed rapidly to support blood volume.

B. Packed Red Cells

Packed red cells (which include leukocyte-poor, filtered, or frozen deglycerolized products) prepared from whole blood by centrifugal techniques are the appropriate choice for almost all patients with deficient oxygen-carrying capacity. Exact indications will be defined by the clinical setting, the severity of the anemia, the acuity of the condition, and any other factors affecting oxygen transport.

C. Platelets

The decision to transfuse platelets depends on the patient’s clinical condition, the status of plasma phase coagulation, the platelet count, the cause of the thrombocytopenia, and the functional capacity of the patient’s own platelets. In the face of decreased production, clinical bleeding, and platelet counts less than 10,000/μL, the risk of severe, spontaneous bleeding is increased markedly. In the presence of these factors and in the absence heparin-induced thrombocytopenia, TTP, or antibody-mediated thrombocytopenia, transfusion may be considered. Under certain circumstances, especially with platelet dysfunction or treatment that inhibits the procoagulant system, transfusions at higher platelet counts may be necessary.

Transfused platelets are sequestered temporarily in the lungs and spleen before reaching their peak concentrations, 45–60 minutes after transfusion. A significant proportion of the transfused platelets never circulate but remain sequestered in the spleen. This phenomenon results in reduced recovery; under best conditions, only 60%–70% of the transfused platelets are accounted for when peripheral platelet count increments are used as a measure of response.

In addition to cessation of bleeding, two variables indicate the effectiveness of platelet transfusions. The first is platelet recovery, as measured by the maximum number of platelets circulating in response to transfusion. The practical measure is the platelet count at 1 hour after transfusion. In the absence of immune or drastic nonimmune factors that markedly decrease platelet recovery, one would expect a 7000/μL increment for each random donor unit and a 40,000–70,000/μL increment for each single-donor apheresis unit in a large child or adolescent. For infants and small children, 10 mL/kg of platelets will increase the platelet count by at least 50,000/μL. The second variable is the survival of transfused platelets. If the recovery is great enough, transfused platelets will approach a normal half-life in the circulation. In the presence of increased platelet destruction, the life span may be shortened to a few days or a few hours. Frequent platelet transfusions may be required to maintain adequate hemostasis.

A particularly troublesome outcome in patients receiving long-term platelet transfusions is the development of a refractory state characterized by poor (≤ 20%) recovery or no response to platelet transfusion (as measured at 1 hour). Most (70%–90%) of these refractory states result from the development of alloantibodies directed against HLA antigens on the platelet. Platelets have class I HLA antigens, and the antibodies are primarily against HLA A or B determinants. A smaller proportion of these alloantibodies (< 10%) may be directed against platelet-specific alloantigens. The most effective approach to prevent HLA sensitization is to use leukocyte-depleted components (< 5 million leukocytes per unit of packed red cells or per apheresis or 6–10 random donor unit concentrates). For the alloimmunized, refractory patient, the best approach is to provide HLA-matched platelets for transfusion. Reports have suggested that platelet cross-matching procedures using HLA-matched or unmatched donors may be helpful in identifying platelet concentrates most likely to provide an adequate response.

D. Fresh Frozen Plasma

The indication for fresh frozen plasma (FFP) is replacement of plasma coagulation factors in clinical situations in which a deficiency of one or more clotting factors exists and associated bleeding manifestations are present. In some hereditary factor deficiencies, such as factor VIII deficiency or vWD, commercially prepared concentrates contain these factors in higher concentrations and, because of viral inactivation, impose less infectious risk and are more appropriate than plasma. Treatment with FFP is indicated for decreased liver production of clotting factors or generalized consumption (DIC) when INR is greater than 1.5.

E. Cryoprecipitate

This component may be used for acquired or congenital disorders of hypofibrinogenemia or afibrinogenemia. Although cryoprecipitate is a rich source for factor VIII or vWF, commercial concentrates that contain these factors are more appropriate (see preceding section). The dose given depends on the protein to be replaced. Cryoprecipitate can be given in a rapid transfusion over 30–60 minutes and is currently under investigation.

F. Granulocytes

With better supportive care over the past 20 years, the need for granulocytes in neutropenic patients with severe bacterial infections has decreased. Indications still remain for severe bacterial or fungal infections unresponsive to vigorous medical therapy in either newborns or older children with bone marrow failure, or patients with neutrophil dysfunction. Newer mobilization schemes using G-CSF and steroids in donors result in granulocyte collections with at least 50 billion neutrophils. This may provide a better product for patients requiring granulocyte support.

G. Apheresis Products and Procedures

Apheresis equipment allows one or more blood components to be collected while the rest are returned to the donor. Apheresis platelet concentrates, which have as many platelets as 6–10 units of platelet concentrates from whole blood donations, are one example; granulocytes are another. Apheresis techniques can also be used to collect hematopoietic stem cells that have been mobilized into the blood by cytokines (eg, G-CSF) given alone or after chemotherapy or mononuclear cells for immunotherapy. These stem cells are used for allogeneic or autologous bone marrow transplantation. Blood cell separators can be used for the collection of single-source plasma or removal of a blood component that is causing disease. Examples include red cell exchange in sickle cell disease and plasmapheresis in Goodpasture syndrome or in Guillain-Barré syndrome.

Adverse Effects

The noninfectious complications of blood transfusions are outlined in Table 30–12. Most complications present a significant risk to the recipient.

Table 30–12. Adverse events following transfusions.


Allain JP: Transfusion-transmitted infectious diseases. Biologicals 2009;37:71 [PMID: 19231236].

Dodd R: Transfusion medicine and safety. Biologicals 2009;37:62 [PMID: 19230706].

Fasano R: Blood component therapy. Pediatr Clin North Am 2008;55:421 [PMID: 18381094].


Goodnight SH, Hathaway WE (ed): Disorders of Hemostasis and Thrombosis: A Clinical Guide, 2nd ed. McGraw-Hill; 2001.

Hillyer CD et al (eds): Blood Banking and Transfusion Medicine, Basic Principles and Practice. Churchill Livingstone; 2007.

Petz LD, Garratty G (eds): Acquired Immune Hemolytic Anemias, 2nd ed. Churchill Livingstone; 2004.

Roback JD et al: Technical Manual, 17th ed. AABB; 2010.