Pierre Noel and Elizabeth A. Jaben
HEMATOLOGIC COMPLICATIONS OF PREGNANCY
Anemia in Pregnancy
During normal pregnancies, plasma volume increases by 40% to 60% and red cell mass by 20% to 40%. The hematocrit typically decreases 30% to 32%, and the lower limit of normal for hemoglobin declines to 11 g/dL in the first trimester and 10 g/dL in the second and third trimesters. The most common forms of anemia of pregnancy in North America are due to iron and folate deficiencies.
One thousand milligrams of additional iron are required during pregnancy. The normal 500 mg iron storage pool is insufficient, and iron deficiency anemia develops in the absence of iron supplementation throughout pregnancy. The recommended daily allowance for iron during pregnancy is 27 mg of elemental iron. The United States Centers for Disease Control recommends routine low-dose iron supplementation (30 mg of elemental iron daily) for all pregnant women, beginning at the first prenatal visit. Calculations of dosage for iron preparations should be based on the amount of iron in each preparation: Ferrous sulfate contains 20% elemental iron, 12% ferrous gluconate, and 33% ferrous fumarate. Low values of serum iron and ferritin are reliable indicators of iron deficiency in pregnancy. The consequences of maternal iron deficiency on the neonate are controversial. Mild-to-moderate maternal iron deficiency anemia is not associated with significant anemia in the fetus.
Folate needs are increased during pregnancy. Folate deficiency is associated with anemia, neural tube defects, and cleft palate. Neural tube closure occurs during the fourth week of pregnancy; therefore, folate supplementation is necessary prior to conception to prevent neural tube defects. Most prenatal vitamins contain both folate and iron.
Sickle Cell Disease in Pregnancy
Women with sickle cell anemia are in a high-risk pregnancy group. With modern obstetric and perinatal care, maternal mortality is less than 1% and perinatal mortality is less than 15%.
Prophylactic red cell transfusions are associated with fewer maternal painful episodes, but they have no impact on maternal morbidity, infant birth weight, gestational age, fetal distress, or perinatal mortality.
Maintenance transfusions should be administered to women who are symptomatic of vasoocclusive or anemia-related problems or when signs of fetal distress are present.
Thrombocytopenia in Pregnancy
Platelet count decreases by approximately 10% during pregnancy, mostly in the third trimester.
The most common cause of thrombocytopenia is incidental thrombocytopenia of pregnancy (75%), followed by thrombocytopenia complicating hypertensive disorders of pregnancy (20%) and finally immunological disorders of pregnancy (5%).
Thrombocytopenia of less than 100,000/μL in the first trimester of pregnancy is most typical for immune thrombocytopenic purpura. Thrombocytopenia of over 70,000/μL occurring late during the second trimester or during the third trimester, in the absence of hypertension or proteinuria, usually represents incidental thrombocytopenia of pregnancy. Platelet-associated IgG is elevated in both incidental thrombocytopenia of pregnancy and immune thrombocytopenic purpura.
It is important in any patient with thrombocytopenia to consider human immunodeficiency virus infection, systemic lupus erythematosus, and thrombocytopenia associated with antiphospholipid antibodies in the differential diagnosis.
Incidental Thrombocytopenia of Pregnancy
The platelet count in incidental thrombocytopenia generally remains above 100,000/μL. Incidental thrombocytopenia usually develops in the third trimester and is not associated with neonatal thrombocytopenia. The likelihood of a more serious cause of thrombocytopenia increases when the platelet count drops below 70,000/μL. The pathogenesis of incidental thrombocytopenia is not clearly defined but may involve a combination of hemodilution and decreased platelet half-life.
Incidental thrombocytopenia remains a diagnosis of exclusion. The diagnosis is made by the lack of other physical or laboratory abnormalities in patients who do not have an antecedent history of immune thrombocytopenia. Women with incidental thrombocytopenia should receive standard obstetrical care. A platelet count greater than 80,000/μL is felt to be sufficient for epidural anesthesia.
Immune Thrombocytopenic Purpura
Immune thrombocytopenia (ITP) is the most common cause of severe thrombocytopenia in the first trimester of pregnancy. An antecedent history of ITP or autoimmune disorder makes the diagnosis more likely. The nadir platelet count in ITP usually occurs in the third trimester.
Patients with platelet counts greater than 20,000 to 30,000/μL and no evidence of bruising or mucosal bleeding generally do not require treatment in the first two trimesters of pregnancy. A platelet count of greater than 50,000/μL is considered safe for normal vaginal delivery or cesarean section. Although there is no consensus, a platelet count greater than 80,000/μL is sufficient for epidural anesthesia. The bleeding time is not an accurate predictor of risk of bleeding in these situations.
Optimal first-line therapy for ITP in pregnant patients is controversial. Corticosteroids are the least expensive option, but they have been associated with pregnancy-induced hypertension, gestational diabetes, osteoporosis, excessive weight gain, and premature rupture of fetal membranes. The placenta metabolizes 90% of the administered dose of prednisone, and thus serious fetal side effects are unlikely. Prednisone is initiated at a dose of 1 mg/kg/day (based on the pre-pregnancy weight) and subsequently tapered to the minimum hemostatically effective dose. Intravenous immunoglobulin (IVIg) should be considered if the maintenance dose of prednisone is in excess of 10 mg/day. IVIg given at a dose of 1 g/kg (based on pre-pregnancy weight) is associated with a response in over 60% of patients, and response duration averages a month.
In patients refractory to corticosteroids and IVIg, splenectomy should be considered. Splenectomy is best performed in the second trimester of pregnancy. Splenectomy in the first trimester may induce labor, and splenectomy in the third trimester may be technically difficult. Splenectomy has been successfully performed laparoscopically during pregnancy. High-dose methylprednisolone and intravenous anti-D immunoglobulin have been used in small series of refractory patients.
Very little data are available regarding the safety and efficacy during pregnancy of thrombopoietin receptor mimetic agonists (romiplostim, eltrombopag). Experience with immunosuppressive and cytotoxic agents during pregnancy also is limited. Danazol and vinca alkaloids are best avoided. Interventions that raise maternal platelet count are not effective in augmenting the platelet count of the fetus.
The use of non-steroidal anti-inflammatory drugs should be avoided postpartum in patients with platelet counts less than 100,000/μL. Thromboprophylaxis should be considered in all women with a platelet count greater than 50,000/μL; if they have undergone surgical delivery, are immobilized for a prolonged amount of time, or have acquired or congenital thrombophilia.
Neonatal mortality is less than 1% in ITP. Five percent of neonates will have a platelet count of less than 20,000/μL; most hemorrhagic events in neonates occur 24 to 48 hours after delivery, at the nadir of the platelet count. There is no evidence that cesarean section is safer for the neonate than is vaginal delivery. The mode of delivery should be decided on the basis of routine obstetric indications.
Maternal platelet count, maternal platelet antibody levels, or a history of maternal splenectomy for ITP are not accurate predictors of neonatal platelet counts. The most accurate predictor of fetal thrombocytopenia is a history of thrombocytopenia at delivery in a prior sibling. Fetal scalp blood sampling and cordocentesis have been generally abandoned.
A cord platelet count should be determined following delivery in every neonate. Thrombocytopenic neonates should be followed closely following delivery, the platelet count nadir may not occur before 2 to 5 days. Neonates presenting with clinical bleeding or a platelet count less than 20,000 μL should be managed using IVIg at a dose of 1 g/kg. Life threatening bleeding can be treated with a combination of IVIg and platelet transfusions. Neonates with platelet counts less than 50,000/μL should undergo transcranial ultrasound to exclude intracranial hemorrhage.
Preeclampsia and HELLP Syndrome
Preeclampsia is defined as hypertension (systolic pressure greater than 140 mm of mercury or diastolic pressure greater than 90 mm of mercury) and proteinuria (greater than 300 mg of protein/24 hours) occurring after 20 weeks of gestation. Preeclampsia occurs in 5% of all pregnancies, and it accounts for 18% of maternal deaths in the United States. Predisposing factors include age below 20 or over 30, increased body mass index, chronic hypertension, and insulin resistance. Thrombocytopenia develops in 50% of patients with preeclampsia. Endothelial damage and activation of the coagulation system with thrombin generation may explain the thrombocytopenia. D-dimers and thrombin–anti-thrombin complexes are increased in patients with thrombocytopenia.
The criteria for HELLP syndrome (hemolysis, elevated liver enzymes, and low platelets) are as follows:
microangiopathic hemolytic anemia
increased transaminases
lactic dehydrogenase greater than 600 units per milliliter
thrombocytopenia (less than 100,000/μL).
HELLP occurs in up to 10% of women with severe preeclampsia. Proteinuria is present in 75% of patients with HELLP syndrome, but only 50% to 60% have hypertension. The syndrome usually occurs in white, multiparous women above the age of 25 years. Maternal mortality is 1% and fetal mortality is 10% to 20%. Fetal mortality is attributed to placental ischemia, abruption of the placenta, immaturity, and intrauterine asphyxia. Neonatal thrombocytopenia can occur in both preeclampsia and HELLP. The mechanism of neonatal thrombocytopenia remains unclear. There is a 3% risk of recurrence of HELLP in subsequent pregnancies.
The definitive treatment for eclampsia and HELLP is delivery of the fetus. Management focuses on stabilization of the patient and maturation of the fetal lung. The presence of multiorgan dysfunction, fetal distress, or a gestational age greater than 34 weeks warrants immediate delivery. Coagulopathy resulting from preeclampsia-associated diffuse intravascular coagulation (DIC) occurs in 20% of patients. The clinical manifestations of preeclampsia and HELLP resolve within a few days of delivery. Rarely, HELLP syndrome can present postpartum. If the manifestations worsen or persist after 1 or 2 days, plasma exchange is indicated.
Acute fatty liver of pregnancy occurs in the third trimester and is associated with hypertension and proteinuria in 50% of patients. Microangiopathic hemolytic anemia and thrombocytopenia are not prominent in this syndrome. Patients usually have a prolonged prothrombin time, a low fibrinogen, and low antithrombin levels.
Thrombotic Thrombocytopenic Purpura and Hemolytic Uremic Syndrome
Thrombotic thrombocytopenic purpura (TTP) and hemolytic uremic syndrome (HUS) occur in only 0.004% of pregnancies. The following are the classic pentad of symptoms of TTP:
Microangiopathic hemolytic anemia
Thrombocytopenia
Neurologic abnormalities
Fever
Renal dysfunction
However, the classic pentad is present in only 40% of patients. Pregnancy is a precipitating factor for TTP. The mean time of onset of TTP is at 23.5 weeks of pregnancy. Plasma exchange therapy is recommended for the management of the pregnant TTP patient, and delivery is indicated only for patients who do not respond to plasma exchange. Pregnancy termination is not considered therapeutic in TTP or HUS.
Ultralarge von Willebrand factor (VWF) multimers are found in TTP, thought to be secondary to the deficiency of a specific VWF-cleaving protease, identified as ADAMTS-13. ADAMTS-13 levels in TTP are typically less than 10%. ADAMTS-13 deficiency can be congenital or acquired. Acquired deficiency is associated with the presence of autoantibodies directed against ADAMTS-13. Reduced ADAMTS-13 levels are not specific for TTP; reduced levels are seen in the third trimester of pregnancy, in uremia, acute inflammation, malignancy, and in DIC. The majority of patients with TTP respond to plasma exchange. The role of corticosteroids remains controversial, mainly because of associated side effects. Patients who develop pregnancy-associated TTP are at high risk of recurrence with subsequent pregnancies.
The mean time of onset of HUS is 26 days following delivery. Patients with HUS present with microangiopathic hemolytic anemia and acute renal failure. VWF levels are usually elevated while multimer analysis may or may not show ultra-large multimers. Deficiency of VWF—cleaving protease is usually not associated with this syndrome.
Several women with a familial history of pregnancy-associated HUS have developed their first episode of HUS during pregnancy, and HUS has occurred in such patients with the use of oral contraceptives. Postpartum HUS is associated with a poor prognosis. Plasma exchange is less effective in reversing renal failure in pregnancy-associated HUS. Nevertheless, a trial of plasma exchange is indicated. Dialysis and other supportive-care measures may also need to be initiated (Table 26.1).
Diffuse Intravascular Coagulation
Placental abruption is the most common cause of DIC (Table 26.2). There is an increased incidence of placental abruption in cocaine addicts. The incidence of DIC complicating placental abruption and dead fetus syndrome has decreased with advances in ultrasonography and prenatal care.
Table 26.2 Causes of Obstetrical Disseminated Intravascular Coagulation
Placental abruption
Fetal death syndrome
Amniotic fluid embolism
HELLP syndrome
Clostridial sepsis
Sepsis
Major obstetrical hemorrhage
HELLP, hemolysis, elevated liver enzymes, and low platelets.
Fetal death syndrome is recognized by ultrasonography. Delivery of the dead fetus removes the source of tissue thromboplastin release. Blood component support and the use of antithrombin-3 have been useful in the management of the coagulopathy.
Placental abruption is managed with blood component support followed by delivery. Antithrombin-3 and activated protein C have been used with success in this disorder.
Transient DIC occurs in patients undergoing hypertonic saline abortions, but the DIC usually resolves once the fetus is delivered. Clostridial sepsis following abortions is associated with DIC and poor clinical outcome.
Venous Thromboembolism in Pregnancy
The per-day risk of venous thromboembolism (VTE) is increased 7- to 10-fold for antepartum VTE and 15- to 35-fold for postpartum VTE. The risk of VTE diminishes rapidly after delivery, returning to the antepartum risk level by 3 weeks postpartum and to the non-pregnant level after 6 weeks. Venous thrombi occur predominantly in the left leg, partly because of the compression of the left iliac vein by the right iliac artery as they cross.
Hemodynamic changes causing venous stasis and hypercoagulability play a role in the increased risk of VTE during pregnancy. Hypercoagulability is thought to be secondary to an increase in fibrinogen, factor VIII, and von Willebrand factor. Further, a decrease in protein S, the development of acquired protein C resistance, and reduced fibrinolytic activity from increased plasminogen activator inhibitor type 1 and 2 activity and decreased tissue plasminogen activator activity may contribute.
History of prior VTE, body mass index greater than 25, prolonged immobilization, inherited thrombophilias, antiphospholipid antibodies, and a family history of thrombosis all increase the risk of VTE during pregnancy.
Diagnosis of Venous Thromboembolism in Pregnancy
The diagnosis of VTE during pregnancy is complicated by the potential for fetal oncogenicity and teratogenicity due to ionizing radiation in diagnostic purposes.
Compression ultrasonography (CU) of the entire proximal venous system to the trifurcation should be performed as the initial test for suspected deep vein thrombosis (DVT) in pregnancy. A normal CU does not exclude a calf DVT. The CU needs to be repeated at day 2 and day 7 to exclude an extending calf-vein thrombosis. A limited venogram with fetal shielding can be used in equivocal cases. When iliac DVT is suspected, pulsed Doppler ultrasound should be used; if the results are negative or equivocal, magnetic resonance venography (MRV) or venography should be considered.
In patients with suspected pulmonary emboli during pregnancy, bilateral compression lower extremity ultrasounds should be performed. If the ultrasound is negative a ventilation/perfusion lung scan (V/Q) should be the next procedure. If the results of the V/Q scan are equivocal, computed tomography pulmonary angiography (CTPA) should be performed. However, should an isolated subsegmental defect be suggested by CTPA, additional testing is suggested because of the high false-positive rate.
D-dimer levels increase throughout pregnancy. The D-dimer test has a high sensitivity, relatively low specificity and very high negative predictive value.
Treatment of Venous Thromboembolic Disease in Pregnancy
Unfractionated heparin (UFH) and low molecular weight heparin (LMWH) do not cross the placenta; therefore, there is no risk of fetal bleeding or teratogenicity. Heparin-induced thrombocytopenia, bleeding, and heparin-induced osteoporosis are more common with UFH than with LMWH.
There are case reports describing the use of fondaparinux, argatroban, and lepirudin in pregnant women with heparin-induced thrombocytopenia. The three drugs are classified under United States Food and Drug Administration (FDA) Class B, indicating that animal studies have not shown harm in pregnancy, but there are no data from humans. Fondaparinux appears to cross the placenta in very low concentrations, but lepirudin and argatroban do not appear to cross the placenta. Coumarin derivatives do transit the placenta and have been associated with fetal bleeding and teratogenicity. Central nervous system abnormalities have occurred after the use of coumarin derivatives in every trimester of pregnancy: nasal hypoplasia and/or stippled epiphyses have been associated with these drugs employed between the sixth and twelfth week of pregnancy.
The activated partial thromboplastin time response to UFH is blunted in pregnancy because of increased factor VIII levels and increased heparin-binding proteins. This blunted response may lead to heparin overdosing. Measuring anti-activated factor (FXa) levels may obviate the problem. LMWHs have less non-specific binding to heparin-binding proteins, hence they have a more predictable doseresponse than UFH.
UFH and LMWH are not secreted in breast milk, nor are argatroban or bivalirudin. Fondaparinux, however, is present in breast milk. Clinical evidence suggests that warfarin sodium is not excreted in breast milk and that breast-feeding is safe when mothers are under warfarin sodium therapy.
The initial dose of LMWH is based on patient weight. Because of variation of weight and glomerular filtration rate during pregnancy, it is recommended to monitor anticoagulation by monthly anti-FXa levels. Dose reduction to ¾ dose after 3 to 4 weeks of full treatment appears safe and may obviate the need for continued anti-FXa monitoring. LMWH should be discontinued 24 hours prior to elective induction of labor and neuroaxial anesthesia. Intravenous UFH can be initiated in patients at high risk for thrombosis and discontinued 4 to 6 hours prior to the time of expected delivery. LMWH can usually be restarted within 12 hours of delivery.
UFH or LMWH should be continued for at least 4 days after initiation of warfarin, until the international normalized ratio has been therapeutic, 2.0 or more, for two consecutive days.
Graduated compression stockings providing a pressure of 30 to 40 mm Hg should be prescribed to decrease the incidence of post-phlebitic syndrome.
The placement of an inferior vena cava filter should be considered, if the VTE is diagnosed after 37 weeks of pregnancy. Discontinuation of anticoagulation for delivery without a filter in place is associated with high rates of morbidity and mortality.
Prophylactic Anticoagulation in Patients with a Previous History of Venous Thromboembolism
Prophylactic antepartum anticoagulation is indicated in patients with a history of unprovoked VTE. Patients with a previous provoked VTE secondary to a temporary risk factor, who do not have an identifiable thrombophilia, are at low risk of recurrence at the time of a subsequent pregnancy and do not require antepartum prophylactic anticoagulation. All women with prior VTE should receive prophylactic postpartum anticoagulation for 6 weeks.
Patients with no prior VTE who are heterozygous for Factor V Leiden or prothrombin gene mutation have a low risk of antepartum VTE without prophylaxis. Antepartum anticoagulation is warranted in patients with antithrombin III deficiency and in patients who are double heterozygotes for Factor V Leiden and prothrombin gene mutation (Table 26.3).
Patients with idiopathic VTE who are pregnant or plan to become pregnant should undergo screening for thrombophilias. Patients with history of fetal loss, abruption, preeclampsia, and intrauterine fetal growth retardation should also be screened for thrombophilias.
Cesarean section is not a risk factor for VTE. Pharmacologic or mechanical thromboprophylaxis is recommended in patients with one VTE risk factor. Combined pharmacologic and mechanical thromboprophylaxis are recommended in patients with multiple VTE risk factors. In high-risk patients, 6 weeks of thromboprophylaxis is recommended.
Table 26.3 Most Common Thrombophilias
Inherited
Factor V Leiden
Prothrombin G20210A mutation
4G/4G mutation of the plasminogen activator inhibitor gene (PAI-I)
Thermolabile variant of methylenetetrahydrofolate reductase, the most common cause of homocystinemia
Antithrombin III deficiency
Protein C deficiency
Protein S deficiency
Acquired
Antiphospholipid antibody
Thrombophilias and Recurrent Miscarriage
Recurrent miscarriage is defined as three consecutive, spontaneous abortions of an intrauterine pregnancy, each occurring at less than 20 weeks gestation. Anticardiolipin antibodies have been linked with recurrent miscarriage. There are insufficient data to include inherited thrombophilias in the evaluation of women with recurrent miscarriage.
Prednisone, low dose aspirin, UFH, LMWH, and IVIg have been used in the management of this problem. Prednisone was found to be equally effective as low-dose subcutaneous UFH in preventing pregnancy loss, but aspirin was associated with increased toxicities. UFH and aspirin have been shown to be superior to aspirin alone in preventing pregnancy loss. LMWH can be employed instead of UFH. The optimal dosages of UFH and LMWH remain to be defined.
HEMATOLOGICAL MANIFESTATIONS OF TROPICAL DISEASE
Malaria
Anemia is a serious complication of malaria, especially Plasmodium falciparum infection. The prevalence and degree of anemia depends on the nutritional and immune status of the patient. The degree of anemia cannot be explained entirely by intravascular rupture of parasitized red cells. Several mechanisms are involved in the anemia of malaria (Table 26.4). P. vivax, and P. ovale invade only reticulocytes, P. malariae invades only mature red cells, and P. falciparum invades red cells of all ages. The proportion of cells parasitized in P. vivax malaria rarely exceeds 1%, whereas as many as 50% of red cells may be parasitized in P. falciparum infections.
There are two major clinical patterns in malaria: (1) acute malaria in the nonimmune and (2) recurrent malaria. Acute malaria is associated with a rapid drop in hemoglobin. Recurrent malaria causes splenomegaly and less severe anemia, and there are only scanty asexual forms and some gametocytes on the peripheral blood smear (Table 26.6). In tropical areas, anemia tends to be more prevalent and most severe in children from 1 to 5 years of age and during pregnancy. Pregnant women who are not immune to P. falciparum develop severe malaria during pregnancy, and they have high rates of abortion, premature delivery, and perinatal and maternal mortality. In women who are immune, extravascular hemolysis and secondary folic acid deficiency play a major role in the pathogenesis of anemia. The extravascular hemolysis in immune women peaks during the second trimester and is accompanied by progressive splenomegaly.
Table 26.4 Causes of Anemia in Malaria
Intravascular rupture of parasitized red cells Hypersplenism
Autoimmune hemolysis (50% of patients have a positive direct Coombs)
Reticulocytopenia (Anemia of chronic disease) Dyserythropoiesis (Cytokine mediated)
Secondary bacterial, fungal, or viral infections
Nutritional anemias
Table 26.5 Protective Genetic Alterations
Southeast Asian ovalocytosis (Autosomal dominant, 27 base pair deletion in the band 3 gene).
Heterozygotes for beta-thalassemias (Protection against P. falciparum)
HbE, HbS
Hereditary persistence of fetal hemoglobin
Glucose-6-phosphate dehydrogenase deficiency
Duffy-null phenotype (The Duffy antigen receptor for chemokines serves as a receptor for red cell invasion by P. vivax. Individuals who are Duffy-null are resistant to vivax malaria.)
Glycophorin A deficient phenotypes [En(a-), Mk] (Glycophorins are important ligands for the attachment and invasion of P. falciparum merozoites).
Glycophorin B deficient phenotypes [S-s-U-]
CD35 (Knops antigen) variants (CD35 is involved in the resetting of P. falciparum-infected red cells with uninfected cells)
Hyperreactive malarial splenomegaly (HMS) is characterized by splenomegaly, hypersplenism, a polyclonal B-lymphocyte proliferation, high IgM levels, and raised titers of antibodies against the predominant species of malaria. Sickle cell trait is protective against HMS. Patients with HMS have a persistence of malaria-induced IgM lymphocytotoxic antibodies, which reduce the numbers of T-suppressor lymphocytes and permit the proliferation of B-lymphocytes. HMS has been associated with the development of splenic lymphoma with villous lymphocytes. Significant lymphocytosis will develop in 15% of patients with HMS, and may be mistaken for chronic lymphocytic leukemia.
Visceral Leishmaniasis
Visceral leishmaniasis (VL, Kala-Azar) is caused by one of three species of Leishmania donovani complex. L. donovani is transmitted by phlebotomine sand flies. VL can also be transmitted through sexual contact, blood transfusions, and vertically.
Chronic VL infection can cause marrow hypoplasia, gelatinous transformation, dyserythropoiesis, and myelofibrosis.
African Trypanosomiasis (Sleeping Sickness)
African trypanosomiasis (AT or sleeping sickness) is endemic in sub-Saharan Africa. Trypanosoma brucei gambiense and Trypanosoma brucei rhodesiense are the etiologic agents. The tsetse fly is the vector. The infection is associated by the proliferation of macrophages and lymphocytes. Patients typically develop splenomegaly, pancytopenia secondary to hypersplenism, polyclonal hypergammaglobulinemia, monocytosis, and lymphocytosis.
Helminth Infections
Eosinophilia is present during the invasive migrating phase of hookworms, Strongyloides, and Ascaris. Hookworm is second in frequency only to malaria as an infectious cause of anemia. The daily loss of blood in the gut is 0.03 to 0.05 mL for each Necator americanus worm and 0.15 to 0.23 mL for each Ancylostoma duodenale worm. The development of iron deficiency is related to the dietary intake of iron, the size of the iron stores, and the hookworm load. Iron depletion is more common in women, during pregnancy, and in children. Less frequent causes of iron deficiency are outlined in Table 26.7.
CLONAL EOSINOPHILIC DISORDERS
Blood eosinophilia is defined as an eosinophil count superior to 450/μL. Eosinophils are much more abundant in tissues than in the peripheral blood. Sustained eosinophilia is associated with end-organ damage in a minority of patients (Table 26.8).
IL-5, IL-3, and GM-CSF all stimulate eosinophil production and inhibit eosinophil apoptosis. Eotaxin-1, eotaxin-2, and RANTES (regulated on activation T cell expressed and secreted) are chemotactic cytokines, causing eosinophils to migrate into tissues. Eosinophils are the source of multiple cytokines (IL-2, IL-3, IL-4, IL-5, IL-7, IL-13, IL-16, TNF-alpha, TGF-beta, and RANTES). They are also the source of cationic proteins such as eosinophil cationic protein, eosinophil peroxidase, major basic protein, eosinophil derived neurotoxin, and Charcot-Leyden crystal lysophospholipase.
Helminthic infections are the most common cause of eosinophilia worldwide, and atopic disorders are the most common cause in industrialized countries. Clonal eosinophilic disorders account for only a small proportion of all eosinophilias (Table 26.9).
Sustained hypereosinophilia, whether reactive or clonal, can damage tissue. The risk factors for end-organ damage are undefined.
The evaluation of a patient with eosinophilia is influenced by the patient’s geographical origin and travel history. Serial stool examinations for ova and parasites may need to be supplemented by endemically relevant serologies and occasionally tissue biopsies.
A clonal eosinophilic disorder needs to be investigated in patients without evidence of infectious or reactive causes of eosinophilia. Clonal eosinophilic disorders can be subdivided into (1) clonal T-cell disorders, (2) clonal myeloid disorders, (3) suspected clonality that cannot be proven [idiopathic hypereosinophilic syndrome (IHES)]. The number of patients classified as having IHES is decreasing as our diagnostic tools are improving (Table 26.10).
Table 26.9 Diseases Commonly Associated with Eosinophilia
Infectious (helminth, protozoa, fungi, HIV, HTLV-1)
Allergic diseases (asthma, atopic dermatitis, allergic rhinitis, urticarias, allergic drug reactions)
Respiratory tract disorders (hypersensitivity pneumonitis, Loeffler’s syndrome, allergic bronchopulmonary aspergillosis, tropical pulmonary eosinophilia)
Endocrinologic disorders (Addison’s disease)
Gastrointestinal disorders (inflammatory bowel disease, eosinophilic gastroenteritis)
Cutaneous and subcutaneous disorders (atopic dermatitis, eosinophilic cellulitis, scabies, episodic angioedema with eosinophilia, chronic idiopathic urticaria, recurrent granulomatous dermatitis, eosinophilic fasciitis)
Immunodeficiency syndromes
Connective tissue disease (Churg-Strauss and cutaneous necrotizing eosinophilic vasculitis)
Neoplastic (Lymphomas,T-ALL,T-cell lymphoproliferative disorders, solid tumors)
Myeloid leukemias and myeloproliferative disorders (acute eosinophilic leukemia, myelomonocytic leukemia with eosinophilia, chronic myelomonocytic leukemia with eosinophilia, chronic myeloid leukemia)
Idiopathic hypereosinophilic syndrome
Cytokines (IL-2, GM-CSF)
L-tryptophan and toxic oil syndrome
GM-CSF, granulocyte macrophage colony-stimulating factor; HIV, human immunodeficiency virus; HTLV-1, human T-cell leukemia virus; IL-2, interleukin 2;T-ALL,T-cell acute lymphoblastic leukemia.
Table 26.10 Clonal Hypereosinophilic Disorders
Clonal T-cell Disorders
Clonal Myeloid Disorders
FIP1L1-PDGFRα hypereosinophilic disorders
Clonal Hypereosinophilic Disorders
The evaluation of patients with suspected clonal hypereosinophilic disorders should include:
AML, acute myeloid leukemia; CBC, complete blood count; CT, computed tomography; HIV, human immunodeficiency virus; IgE, immunoglobulin E; PCR, polymerase chain reaction;T-ALL,T-cell acute lymphoblastic leukemia.
The clonality of eosinophils can be demonstrated by the expression of a single alloenzyme of glucose-6-phosphate dehydrogenase in purified eosinophils from female heterozygotes. Polymerase chain reaction amplification of the human androgen receptor gene locus (HUMARA) can also document clonality of eosinophils in female patients.
T-Cell Clonal Disorders
IL-5 overproduction by TH2 lymphocytes has been demonstrated in both clonal and reactive hypereosinophilic disorders. Aberrant clones of T–lymphocytes are present in 25% of patients with clonal hypereosinophilic disorders. The aberrant phenotypes are heterogeneous ([CD3+, CD4+, CD8−], [CD3+, CD4−, CD8+], [CD3+, CD4−, CD8−], [CD3−, CD4+]). In most cases, an activated T cell phenotype is present with expression of CD25 and HLA-DR. In 50% of patients, a clonal rearrangement of the T-cell receptor gene (beta) or (gamma) is found. T-cell lymphomas develop in a proportion of these patients.
Patients with aberrant CD4+, CD3− T cells producing high levels of IL-5, IL-4, and IL-13 typically present with skin manifestations, lack of severe end-organ involvement, and they have elevated IgE levels and polyclonal hypergammaglobulinemia.
The optimal treatment of patients with aberrant T-cell clones remains unclear. Corticosteroids have been associated with some responses. Interferon-alpha has in-vitro antiapoptotic effects on the clonal CD4+CD3− population and may increase the risk of lymphomatous transformation. The role of alemtuzumab in the management of CD52 positive clonal T-cell disorders is under evaluation.
Acute Leukemias
Acute eosinophilic leukemia is rare. Cyanide-resistant peroxidase can be used to identify eosinophilic blasts. Myelomonocytic leukemia (M4-Eo) with eosinophilia is associated with inv (16) (p13;q22) and t(16;16) (p13;q22). The core binding factor-beta is a transcription factor located at 16q22, and the smooth muscle myosin heavy chain is located at 16p13. The eosinophils in M4Eo frequently have a dysplastic appearance.
Chronic Myelomonocytic Leukemia with Eosinophilia
The two predominant subtypes of chronic myelomonocytic leukemia with eosinophilia involve, respectively, platelet-derived growth factor receptor beta (PDGFR-β) and fibroblast growth factor receptor 1 (FGFR1). In both subtypes, fusion oncoproteins are constitutively activated and are able to activate downstream stimulatory and antiapoptotic pathways.
Chronic Myelomonocytic Leukemia with Eosinophilia Subtypes
Platelet-derived growth factor receptor-β subtype:
Age 50 to 60
Male predominance (>90%)
Monocytosis, eosinophilia, splenomegaly
Most common chromosome abnormality is t(5;12)(p12;q31-32) [ETV6-PDGFR]
Imatinib responsive
Fibroblast growth factor receptor 1 subtype (FGFR1):
Median age: 32
Male: female ratio (1.5:1)
Associated with lymphoblastic lymphoma transformation (B and T)
Associate with rearrangement of FGFR1 at the 8p11-12 locus
Not responsive to imatinib
FIP1L1-PDGFR-α Hypereosinophilic Disorders
FIP1L1-PDGFR-α is a constitutively activated tyrosine kinase that was first described in a patient with hypereosinophilic syndrome, who genetically had an occult 800-kb interstitial deletion of chromosome 4q12. The FIP1L1-PDGFRα cannot be detected by conventional karyotyping; fluorescence in situ hybridization (CHIC-2) or reverse transcription polymerase chain reaction is required. The dose of imatinib (100 mg/day) required to inhibit FIP1L1-PDGFRα tyrosine kinase is less than the dose (400 mg/day) needed to block activity of the BCR-ABL tyrosine kinase in chronic myeloid leukemia. The large majority of patients with the FIP1L1-PDGFR-α treated with imatinib obtain a molecular remission within 3 months of initiating therapy. The optimal dosage and duration of treatment remain to be defined, but discontinuation of imatinib leads to relapse. Molecular remissions can be reestablished by reinitiating imatinib therapy. Imatinib suppresses but does not eliminate the FIP1L1-PDGFR-α positive clone. Resistance to imatinib is associated with a T6741 mutation in PDGFR-α; this mutation occurs in the adenosine triphosphate (ATP)-binding region of PDGFR-α at the same position as the T3151 mutation in BCR-ABL. There is limited experience with the use of other tyrosine-kinase inhibitors (desatinib, nilotinib) in hypereosinophilia harboring the FIP1L1-PDGFR-α mutation.
Serum tryptase and vitamin B12 levels are increased in the majority of patients. Splenomegaly is reported in over 60% of cases.
A serum troponin and cardiac echocardiogram should be obtained prior to initiating imatinib. An increased level of serum cardiac troponin correlates with the presence of cardiomyopathy. Prophylactic use of corticosteroids during the first 7 to 10 days of treatment with imatinib is recommended
in patients with evidence of eosinophil-mediated cardiomyopathy and in patients with other cardiac comorbidities.
Idiopathic Hypereosinophilic Syndrome
Idiopathic hypereosinophilic syndrome is arbitrarily defined as eosinophilia in excess of 1,500/μL for more than 6 months with evidence of end-organ damage, without an evident primary or secondary cause of eosinophilia. Idiopathic hypereosinophilia is the term favored when end-organ damage is absent. Corticosteroids represent the first line therapy for idiopathic hypereosinophilic syndrome. Hydroxyurea, interferon-alpha, and monoclonal antibodies directed against interleukin-5 have been used in steroid-resistant patients. The role of imatinib as front-line therapy in patients without FIP1L1-PDGFR-α fusion proteins remains to be established.
NEUTROPENIA
Neutropenia is defined as a decrease in neutrophils below 1,500/μL. Severe neutropenia is defined as a decrease in neutrophils below 500/μL. In patients of African origin, the neutrophil count may normally be as low as 1,000/μL.
Neutropenias can be divided into intrinsic disorders of the hematopoietic system and secondary forms. The secondary forms are caused by extrinsic factors such as: immune causes, hypersplenism, infections, and drugs (Table 26.11).
Intrinsic Disorders
Congenital Neutropenias
Congenital neutropenias include Kostmann syndrome, cyclic neutropenia, congenital immunodeficiency syndromes, as well as several other rare syndromes that will not be discussed in this chapter.
Kostmann syndrome is an autosomal-dominant disorder presenting in the newborn. Characteristic findings include: neutrophils below 200/μL, monocytosis, anemia, thrombocytosis, splenomegaly, and maturation arrest in the marrow at the promyelocyte level. The accelerated apoptosis of neutrophilic precursors is secondary to a mutation of neutrophil elastase. Ninety percent of children with Kostmann syndrome respond to granulocyte colony stimulating factor (G-CSF). Evolution to myelodysplasia and acute leukemia occurs in some patients. It is unclear if G-CSF increases this risk.
Cyclic neutropenias can be congenital (autosomal-dominant congenital disorder) or acquired with clonal large granular lymphocyte syndrome. Congenital cyclic neutropenia is due to mutations at the enzyme active site of the neutrophil elastase gene, which leads to accelerated apoptosis of neutrophils. Clinically, patients present with cycles of neutropenia every 21 to 56 days. The neutropenia can be severe and last 3 to 6 days. Fever, mucosal ulcers, and lymphadenopathy can occur during the nadir of the cycles. G-CSF is useful in the management of cyclic neutropenia.
Table 26.11 Neutropenia Classification
Intrinsic Disorders
Congenital
Acquired
Extrinsic Disorders
Immune neutropenias
Neutropenia associated with autoimmune disorders
Neutropenia associated with large granular lymphocytes
Hypersplenism
Neutropenia associated with infectious diseases
Drug-related neutropenias
Nutritional deficiencies (B12, folate, copper)
Congenital immunodeficiency syndromes frequently associated with neutropenia, include: X-linked agammaglobulinemia, X-linked hyperimmunoglobulin M syndrome, and reticular dysgenesis.
Acquired Neutropenias
Acquired intrinsic disorders include leukemias, myelodysplastic syndromes, lymphoproliferative disorders, aplastic anemia, neutropenia of prematurity, and chronic idiopathic neutropenia.
Chronic idiopathic neutropenia occurs in both children and adults. The neutropenia in some patients can be severe. Patients have negative antineutrophil antibodies, normal marrow cytogenetics, and either normocellular marrows or marrows showing decreased postmitotic cells. The prognosis is excellent; patients do not progress to myelodysplasia or leukemia. A proportion of these patients may have autoimmune neutropenia, with undetectable antineutrophil antibodies. G-CSF is effective in increasing the neutrophil count.
Extrinsic Disorders
Immune Neutropenias
Five neutrophil-specific antigens carried on two different glycoproteins have been described. The NA antigens (NA1, NA2, SH) are expressed on FcμRIIIb (CD16), which is a low-affinity receptor for IgG1 and IgG3. The NB antigen is expressed on glycoprotein CD177. There are data suggesting that ANCA can be implicated in the pathogenesis of secondary AIN. The granulocyte immunofluorescence test (GIFT), the granulocyte agglutination test (GAT) and the monoclonal antibody immobilization of granulocyte antigens assay (MAIGA) can be used to detect antineutrophil antibodies. A combination of GIFT and GAT is recommended as the best approach.
Alloimmune neonatal neutropenia occurs when maternal antibodies cross the placenta and react with the infant’s neutrophils. In isoimmune neutropenia, the mother produces an antibody to the paternal CD16 isotype that is different from her own.
Primary autoimmune neutropenia (AIN) is diagnosed in patients with isolated neutropenia who have detectable antineutrophil antibodies. NA1 antibodies are detected in 35% to 40% of patients. The clinical course is usually benign, and spontaneous remissions are common.
Neutropenia Associated with Autoimmune Disorders
AIN is associated with common variable immunodeficiency. The condition should be excluded in patients with recurrent immune cytopenias and granulomatous disease. A high incidence of antineutrophil antibodies is found in patients with X-linked autoimmune lymphoproliferative syndrome (ALPS).
In systemic lupus erythematosus, Fas-mediated apoptosis of mature neutrophils and CD34 positive hematopoietic progenitor cells play an important role in the pathogenesis of neutropenia. Sjogren’s syndrome, systemic sclerosis, and primary biliary cirrhosis and Grave’s disease have all been associated with AIN.
Felty’s syndrome patients typically have deforming rheumatoid arthritis, splenomegaly, and elevated rheumatoid factor titers. The neutropenia in Felty’s is thought to be antibody mediated. In a proportion of patients with Felty’s syndrome, the neutropenia is secondary to the presence of clonal large granular lymphocytes.
Neutropenia Associated with Large Granular Lymphocyte Syndrome
Large granular lymphocyte (LGL) syndrome is caused by an expansion of either T-lymphocytes or natural killer cells (NK) cells. The NK-cell subtype is more aggressive and accounts for 15% of cases. Forty percent of LGL is associated with other diseases such as rheumatoid arthritis.
The T-cells in clonal LGL express the CD3-TCR complex and have rearranged T cell receptor genes. These cells are thought to represent in-vivo activated cytotoxic T cells. Clonal LGLs express high levels of Fas ligand. Normal neutrophil survival is regulated by the Fas-Fas ligand apoptotic system. The neutropenia in clonal LGL syndrome appears to be mediated by increased peripheral destruction of neutrophils secondary to immune complexes and bone marrow suppression of granulopoiesis by Fas ligand secretion.
Table 26.12 Mechanisms of Drug-Induced Isolated Neutropenia
Dose-dependent inhibition of granulopoiesis
Immune-mediated destruction of neutrophils and neutrophil precursors
Direct toxic effect on marrow granulocytic precursors
Neutropenia Associated with Infectious Diseases
The most common cause of acquired neutropenia is infection. Gram-negative septicemia, Staphylococcus aureus, typhoid fever, paratyphoid fever, tularemia, and brucellosis can cause neutropenia. Infectious hepatitis, influenza, measles, Colorado tick fever, mononucleosis, cytomegalovirus, Kawasaki disease, HIV, parvovirus B12 are included in the differential diagnosis of neutropenia associated with infectious diseases.
Parvovirus B19 is frequently associated with transient neutropenia and may cause protracted leucopenia in immunosuppressed patients. Neutropenia is seen in over 70% of patients with acquired immunodeficiency syndrome and can be associated with hypersplenism and the presence of antineutrophil antibodies.
Drug-Induced Neutropenia
The second most common cause of neutropenia is medication exposure: approximately 70% of agranulocytosis cases in the United States are attributed to medications. Procainamide, antithyroid drugs, and sulfasalazine are most commonly implicated. An exhaustive list of drugs causing neutropenia is beyond the scope of this chapter (see instead Kaufman DW et al. The Drug Etiology of Agranulocytosis and Aplastic Anemia, Oxford University Press, 1991).
Three pathogenetic mechanisms for isolated neutropenia include: dose dependent inhibition of granulopoiesis, immune-mediated destruction of neutrophils and their precursors, and direct toxic effect on marrow granulocytic precursors (Table 26.12).
The onset of neutropenia is rapid (1–2 days) in immune-mediated destruction of neutrophils and more variable with agents causing either direct toxic effect or dose-dependent inhibition. Immune-mediated destruction of neutrophils and their precursors occurs by two mechanisms. With hapten mediation, the agent acts as a hapten to induce antibody formation and needs to be present for neutropenia to occur. In the immune complex mechanism, once the complex is formed, continued drug presence is not required for neutrophil destruction.
Ipilimumab, fludarabine, and rituximab have been associated with AIN. Rituximab-associated neutropenia is delayed. Rituximab late onset neutropenia appears after a median of 38 to 175 days from the last rituximab dose, and its median duration is 5 to 77 days. The pathogenesis is not completely understood. The role of G-CSF remains controversial, and the decision as to its use should be made on an individual basis.
Suggested Reading
Abdalla SH, Pasvol G. Approach to the patient in the tropics with anemia. In: Guerrant RL, Walker DH, Weller PF, eds. Tropical Infectious Diseases: Principles, Pathogens and Practice. 3rd ed. Philadelphia, PA: Elsevier; 2011.
Akhtari M, Curtis B, Waller EK. Autoimmune neutropenia in adults. Autoimmun Rev. 2009;9:62-66.
Boztug K, Klein C. Genetic etiologies of severe congenital neutropenia. Curr Opin Pediatr. 2011;23:21-26.
Checkley AM, Chiodini PL, Dockrell DH, et al. Eosinophilia in returning travellers and migrants from the tropics: UK recommendations for investigation and initial management. J Infect. 2010;60:1-10.
Gotlib J. World Health Organization-defined eosinophilic disorders: 2011 update on diagnosis, risk stratification, and management. Am J Hematol. 2011;86:677-688.
Kaufman DW, Kelly JP, Levy M, Shapiro S. The Drug Etiology of Agranulocytosis and Aplastic Anemia. Oxford: Oxford University Press; 1991.
McCrae KR. Thrombocytopenia in pregnancy. Hematology Am Soc Hematol Educ Program. 2010;2010:397-402.
Noel P. Eosinophilic myeloid disorders. Semin Hematol. 2012;49:120-127.
Pels SG, Paidas MJ. Microangiopathic disorders in pregnancy. Hematol Oncol Clin North Am. 2011;25:311-322.
Pierangeli SS, Leader B, Barilaro G, et al. Acquired and inherited thrombophilia disorders in pregnancy. Obstet Gynecol Clin North Am. 2011;38:271-295.
Rodger M. Evidence base for the management of venous thromboembolism in pregnancy. Hematology Am Soc Hematol Educ Program. 2010;2010:173-180.
Szczepiorkowski ZM, Winters JL, Bandarenko N, et al. Guidelines on the use of therapeutic apheresis in clinical practice: evidence-based approach from the apheresis. Applications Committee of the American Society for Apheresis. J Clin Apher. 2010;25:83-177.
Wilson ME, Weller PF. Eosinophilia. In: Guerrant RL, Walker DH, Weller PF, eds. Tropical Infectious Diseases: Principles, Pathogens & Practice. 3rd ed. Philadelphia, PA: Elsevier; 2011.
Wolach O, Bairey O, Lahav M. Late-onset neutropenia after rituximab treatment. Medicine. 2010;89:308-318.
Young NS. Agranulocytosis. In: Neal S Young eds. Bone Marrow Failure Syndromes. Philadelphia, PA: Saunders; 2000.