Key Abbreviations
Chorionic villus sampling CVS
Hemolysis, elevated liver enzymes, low platelets HELLP
Hemolytic uremic syndrome HUS
Human immunodeficiency virus HIV
Immune thrombocytopenic purpura ITP
Immunoglobulin G IgG
Intravenous immunoglobulin IVIG
Mean corpuscular volume MCV
Polymerase chain reaction PCR
Red blood cell RBC
Thrombotic thrombocytopenic purpura TTP
Unusually large multimers of von Willebrand factor ULVWf
Urinary tract infection UTI
von Willebrand cleaving enzyme ADAMTS13
von Willebrand disease vWD
von Willebrand factor vWF
Pregnancy-Associated Thrombocytopenia
Affecting approximately 4% of pregnancies, thrombocytopenia is the most frequent hematologic complication of pregnancy that results in consultation. As gestation progresses, platelet counts generally fall slightly owing to hemodilution and increased destruction. Commonly, platelet counts will reach a nadir of 120,000/mm 3 during pregnancy; however, they should not fall below the normal range. In pregnancy, the vast majority of cases of mild to moderate thrombocytopenia are caused by gestational thrombocytopenia. This form of thrombocytopenia has little likelihood of causing maternal or neonatal complications. The obstetrician, however, should rule out other etiologies of thrombocytopenia that are associated with severe maternal or perinatal morbidity. The common and rare causes of thrombocytopenia in the gravida at term are shown in the Box 44-1 .
Common Causes
Gestational thrombocytopenia
Severe preeclampsia
Hemolysis, elevated liver enzymes, low platelets (HELLP) syndrome
Disseminated intravascular coagulation
Uncommon Causes
Immune thrombocytopenic purpura
Antiphospholipid antibody syndrome
Systemic lupus erythematosus
Human immunodeficiency virus (HIV) infection
Rare Causes
Thrombotic thrombocytopenic purpura
Hemolytic uremic syndrome
Type 2b von Willebrand syndrome
Hemoglobin SC crisis with splenic sequestration
Folic acid deficiency
Hematologic malignancies
May-Hegglin anomaly (congenital thrombocytopenia)
Wiskott-Aldrich syndrome
Gestational Thrombocytopenia
Most patients with gestational thrombocytopenia generally have a platelet count of 120,000 to 149,000/mm 3 . However, about 1% of patients with gestational thrombocytopenia will have a platelet count of 50,000 to 99,000/mm 3 . These patients require no therapy, and the fetus appears to be at negligible risk of being born with clinically significant thrombocytopenia or a bleeding diathesis. This distinct entity was first suggested but not specifically defined in a study published in 1986 by Hart and colleagues. In this report, 28 of 116 pregnant women (24%) who were evaluated prospectively during an 8-month period had platelet counts less than 150,000/mm 3 at least once during pregnancy. In all 17 patients who were followed after delivery, platelet counts returned to normal. These researchers were actually describing gestational thrombocytopenia before the condition had been recognized as a distinct entity. Samuels and colleagues also investigated 74 mothers with gestational thrombocytopenia. Regardless of platelet antibody status, none of the infants born to these mothers demonstrated thrombocytopenia. Burrows and Kelton have further shown that there is little risk to the mother or neonate in cases of gestational thrombocytopenia. In their study of 1357 healthy, pregnant women, 112 (8.3%) had platelet counts less than 150,000/mm 3 . The lowest platelet count was 97,000/mm 3 . The incidence of thrombocytopenia (platelet count < 150,000/mm 3 ) in the infants of these 112 women was 4.3%, not statistically different from infants born to healthy pregnant women without thrombocytopenia (1.5%). None of these infants had platelet counts less than 100,000/mm 3 . Indeed, the reports by Samuels and colleagues and Burrows and Kelton have convincingly demonstrated that gestational thrombocytopenia is a distinct and common entity that requires no treatment. However, the obstetrician must use judgment in giving this diagnosis because no test exists for this disorder. If platelet counts continue to fall to levels below 50,000/mm 3 , other diagnoses should be entertained.
The decrease in platelet count that occurs in gestational thrombocytopenia is not merely the result of dilution of platelets with increasing blood volume; it also appears to be due to an acceleration of the normal increase in platelet destruction that occurs during pregnancy. This is demonstrated by the fact that the mean platelet volume (MPV) is increased in patients with gestational thrombocytopenia. However, if the platelet counts fall below 20,000/mm 3 or if clinical bleeding is present, further investigation and intervention are warranted. This scenario, however, is rare, and it is difficult to determine whether these patients with profound thrombocytopenia have gestational thrombocytopenia or thrombocytopenia from another cause. Platelet antibody testing should only be utilized if suspicion is high for immune thrombocytopenic purpura (ITP). This would include a platelet count less than 50,000/mm 3.
Immune Thrombocytopenic Purpura
Immune thrombocytopenic purpura affects 1 to 3 per 1000 pregnancies and rarely causes neonatal complications. Although rare cases of neonatal thrombocytopenia have been reported, fetal complications are almost nonexistent. Therefore the focus should be on maternal disease and well-being.
In general, pregnancy has not been determined to cause ITP or to change its severity , but rare exceptions do exist. Harrington and associates were the first to demonstrate that ITP was humorally mediated, and Shulman and colleagues showed that the mediator of this disorder was immunoglobulin G (IgG). These findings were confirmed when Cines and Schreiber developed the first platelet antiglobulin test, a radioimmunoassay, in 1979. Today, this test is usually performed using an enzyme-linked immunosorbent assay (ELISA) or flow cytometry. Newer assays have shown that these autoantibodies may be directed against specific platelet surface glycoproteins, including the IIb/IIIa and Ib/IX complexes. In vivo, after the platelets are coated with antibody, they are removed from circulation by binding to the Fc receptors of macrophages in the reticuloendothelial system, especially the spleen. Approximately 90% of women with ITP have platelet-associated IgG. Unfortunately, this is not specific for ITP, because studies have shown that these tests are also positive in women with gestational thrombocytopenia and preeclampsia.
To make the issue more confusing, the pathogenesis of ITP in children and adults usually differs. Childhood ITP most often follows a viral infection and clinically presents with petechiae and bleeding. This form of ITP is generally self-limited and disappears over time. Conversely, adults have milder bleeding and easy bruisability and are often diagnosed after a prolonged period of subtle symptoms. Adult ITP usually runs a chronic course, and long-term therapy is often eventually needed. Many pregnancies occur in women in their late teens and early twenties. In these women with a history of ITP, it may be difficult to ascertain whether the patient has childhood ITP or adult ITP. The distinction is important for counseling concerning long-term prognosis.
ITP has a predisposition for women aged 18 to 40 years, with an overall female-male ratio of 1.7. It is a diagnosis of exclusion. The patient must have isolated thrombocytopenia with an unremarkable peripheral smear. She must have only bleeding clinically consistent with a depressed platelet count, such as petechiae. She must not be taking any medication, herbal compound, or illicit drug that may cause thrombocytopenia. Finally, the patient must have no other disease process than can cause thrombocytopenia, such as those listed in the box earlier in this chapter. The American Society of Hematology has published a review of ITP that details the diagnostic and therapeutic guidelines.
Thrombotic Thrombocytopenic Purpura and Hemolytic Uremic Syndrome
These two conditions are characterized by microangiopathic hemolytic anemia and severe thrombocytopenia. Pregnancy does not predispose a patient to these conditions, but they should be considered when evaluating the gravida with severe thrombocytopenia. Thrombotic thrombocytopenic purpura (TTP) is characterized by a pentad of findings, which are shown in Box 44-2 . The complete pentad occurs only in approximately 40% of patients, but approximately 75% have a triad of microangiopathic hemolytic anemia, thrombocytopenia, and neurologic changes. Pathologically, these patients have thrombotic occlusions of arterioles and capillaries. These occur in multiple organs, and no specific clinical manifestation for the disease is recognized. The clinical picture reflects the organs that are involved.
* The classic pentad is found in only 40% of patients.
TTP/hemolytic uremic syndrome (HUS) may mimic preeclampsia. Because preeclampsia is much more common than this disorder, it should be considered first. However, delay in diagnosing TTP/HUS can have fatal consequences .
To diagnose the hemolytic anemia associated with TTP, the indirect antiglobulin (Coombs) test must be negative. This rules out an immune-mediated cause for the hemolytic anemia. Lactate dehydrogenase (LDH) should be elevated, the indirect bilirubin should be increased, and haptoglobin should be decreased, indicating ongoing hemolysis. Schistocytes are usually seen on the peripheral smear, if it is carefully reviewed. These tests all signify hemolysis, but specificities and sensitivities differ. For instance, LDH can be elevated in liver disease. Schistocytes are very specific but manifest themselves once the hemolysis is severe. The clinician should use the clinical picture, as well as some of these tests, to make the diagnosis of hemolysis. To be classified as TTP, the platelet count should be less than 100,000/mm 3 . In renal insufficiency associated with TTP, the urine sediment is usually normal with an occasional red blood cell (RBC). This finding helps distinguish this disorder from a lupus flare, which more often has associated hematuria and casts. The serum creatinine is usually greater than 2 mg/dL. This degree of renal dysfunction is unusual, but not rare, in preeclampsia. Proteinuria, more than a trace amount, is usually seen on urine dipstick.
The neurologic findings in TTP are usually nonspecific. They include headache, confusion, and lethargy . Infrequently, generalized tonic-clonic seizures occur. Terrell and coworkers examined the epidemiology of TTP/HUS occurring in Oklahoma between 1996 and 2004. In 206 reported cases, they found that 37% were idiopathic. However, 13% were associated with an autoimmune disease, and 7% occurred in pregnancy and postpartum. These researchers were able to project that the annual incidence of suspected TTP/HUS is 11 cases per million population, whereas the annual incidence of proven cases is 4.5 cases per million. If this disease is so rare, why include it in a text on obstetrics? Because if untreated, TTP carries a 90% mortality rate, whereas treatment with plasma exchange decreases the mortality rate to 20%. Therefore obstetricians must be aware of this disease process so it can be quickly and aggressively treated.
Tsai and colleagues have found that a decrease of ADAMTS13 (the von Willebrand cleaving enzyme) activity is strongly associated with TTP. This metalloprotease, also known as von Willebrand cleaving enzyme, cleaves unusually large multimers of von Willebrand factor (ULVWf). Activity can be decreased from a decrease in the metalloprotease or antibodies against it. If a deficiency in the activity and/or concentration of ADAMTS13 is apparent, ULVWf circulates in increased amounts, leading to increased platelet aggregation and the initiation of TTP. ADAMTS13 can be readily assayed in clinical laboratories. Ferrari and colleagues have shown that all four immunoglobulin subclasses of anti-ADAMTS13 antibodies are associated with TTP, but the IgG4 subclass is most common. Congenital TTP is usually associated with a mutation of ADAMTS13 that leads to a profound decrease in its activity. Moatti-Cohen and colleagues queried the French registry of thrombotic microangiopathies and found that 24% of women who developed TTP during pregnancy had the congenital type (Upshaw-Schulman syndrome) compared with less than 5% of total adult cases. Weiner has published the most extensive literature review concerning TTP. In this series of 45 patients, 40 developed the disease antepartum, and 50% occurred before 24 weeks’ gestation. The mean gestational age at onset of symptoms was 23.4 weeks. This finding may be helpful when trying to distinguish TTP from other causes of thrombocytopenia and microangiopathic hemolytic anemia that occur during gestation. In Weiner’s review, the fetal and maternal mortality rates were 84% and 44%, respectively. These mortality rates are overly pessimistic, because this series included many patients who contracted the disease before plasma infusion/exchange therapy was utilized to treat TTP.
However, TTP may be confused with rarely occurring early-onset severe preeclampsia. In preeclampsia, antithrombin III levels are frequently low, and this is not the case with TTP. This test, therefore, may be a useful discriminator between these two disorders.
Although HUS has many features in common with TTP, it usually has its onset in the postpartum period. Patients with HUS display a triad of microangiopathic hemolytic anemia, acute nephropathy, and thrombocytopenia. HUS is rare in adults, and the thrombocytopenia is usually milder than that seen in TTP, with only 50% of patients having a platelet count less than 100,000/mm 3 at the time of diagnosis. The thrombocytopenia worsens as the disease progresses. A major difference between TTP and HUS is that 15% to 25% of patients with HUS develop chronic renal disease. HUS often follows infections with verotoxin-producing enteric bacteria. Cyclosporine therapy, cytotoxic drugs, and oral contraceptives may predispose adults to develop HUS. The majority of cases of HUS that occur in pregnancy develop at least 2 days after delivery. In fact, in one series, only 9 of 62 cases (14.5%) of pregnancy-associated HUS occurred antepartum. Four of these nine patients developed symptoms on the day of delivery. The mean time from delivery to development of HUS in patients in this series was 26.6 days. The maternal mortality rate may exceed 50% in postpartum HUS; however, this mortality rate is based on historic data. With plasmapheresis and dialysis, the likelihood of maternal death is probably much less. It is not important to make the distinction between TTP and HUS, because the initial therapy for both disorders is plasmapheresis.
Evaluation of Thrombocytopenia during Pregancy and the Puerperium
Before deciding on a course to follow in treating the patient with thrombocytopenia, the obstetrician must evaluate the patient and attempt to ascertain the etiology of her low platelet count, realizing that gestational thrombocytopenia will be the most likely diagnosis. Important management decisions are dependent on arriving at an accurate diagnosis; therefore a complete medical history is critically important. It is essential to learn whether the patient has previously had a depressed platelet count or bleeding diathesis. It is also important to know whether these clinical conditions occur coincidentally with pregnancy. A complete medication history should be elicited, because certain medications—such as heparin, many antibiotics, and histamine-2 blockers—can result in profound maternal thrombocytopenia. The obstetric history should focus on whether any maternal or neonatal bleeding problems occurred in the past. Excessive bleeding from an episiotomy site or cesarean delivery incision site or bleeding from intravenous (IV) sites during labor should alert the physician to the possibility of thrombocytopenia in the previous pregnancy. The obstetrician should also question whether the infant had any bleeding diathesis or any problem occurred following a circumcision. The obstetrician should also ask pertinent questions to determine whether severe preeclampsia or hemolysis, elevated liver enzymes, and low platelets (HELLP) syndrome is the cause of her thrombocytopenia. The treatment of preeclampsia and HELLP are discussed in Chapter 31 . All thrombocytopenic pregnant women should be carefully evaluated for the presence of risk factors for human immunodeficiency virus (HIV) infection, because this infection can cause an ITP-like syndrome. Also, a family history should be elicited because familial forms of thrombocytopenia exist.
An accurate assessment of gestational age should also be carried out. This is important not only in helping to determine the etiology of the thrombocytopenia but also in the timing of delivery. A thorough physical examination of the patient should be performed, and the physician should look for the presence of ecchymoses or petechiae. The conjunctivae and nail beds often reveal petechiae when they are not readily apparent elsewhere on the body. Blood pressure should be determined to ascertain whether the patient has impending preeclampsia. If the patient is developing HELLP syndrome, scleral icterus may be present, and an eye exam should be performed to look for evidence of arteriolar spasm or hemorrhage.
It is imperative that a peripheral blood smear be examined by an experienced physician or technologist whenever a case of pregnancy-associated thrombocytopenia is diagnosed. The presence or absence of evidence of microangiopathic hemolysis on the smear will help establish a diagnosis. This specialist can also rule out platelet clumping, which will result in a factitious thrombocytopenia. Platelet clumping in ethylenediaminetetraacetic acid (EDTA, a lavender-top tube) occurs in about 3 per 1000 individuals and may lead to a spurious diagnosis of thrombocytopenia. If platelet clumping is suspected, the physician should ask the laboratory to perform a platelet count on citrate-collected blood (a blue-top tube). If the count is normal, platelet clumping is likely, and the patient is not thrombocytopenic. Other laboratory evaluations should be performed as necessary to rule out preeclampsia and HELLP syndrome as well as disseminated intravascular coagulation (DIC). If a diagnosis of ITP is entertained, appropriate platelet antibody testing may aid in the diagnosis but is of limited utility during pregnancy.
After determining the etiology of thrombocytopenia, the physician can better determine whether imminent delivery is necessary, if the thrombocytopenia should be treated before initiating delivery, or if the low platelet count should be monitored during an ongoing pregnancy.
Therapy of Thrombocytopenia during Pregnancy
Gestational Thrombocytopenia
Gestational thrombocytopenia, the most common form of thrombocytopenia encountered in the third trimester, requires no special intervention or therapy. The most important therapeutic issue is to refrain from treatment and testing that may lead to unnecessary intervention or iatrogenic preterm delivery. In patients with mild to moderate thrombocytopenia and no antenatal or antecedent history of thrombocytopenia, the patient should be treated as a normal pregnant patient. If the maternal platelet count drops below 50,000/mm 3 , the patient may still have gestational thrombocytopenia, but not enough data are available on mothers with counts this low to determine whether any maternal or fetal risks exist. These patients, therefore, should be treated as if they have de novo ITP. Although approximately 4% of patients have gestational thrombocytopenia, less than 1% of alternate uncomplicated pregnant women will have gestational thrombocytopenia with platelet counts less than 100,000/mm 3 .
Immune Thrombocytopenic Purpura
Treatment of the gravida with ITP during pregnancy and the puerperium requires special attention to the mother, because platelet counts can drop to very low numbers during gestation. As in other cases of thrombocytopenia, maternal therapy needs to be instituted only if a bleeding diathesis is evident or to prevent a bleeding complication if surgery is anticipated. Usually no spontaneous bleeding is present unless the platelet count falls below 20,000/mm 3 . In a meta-analysis of 17 studies, the risk of fatal hemorrhage in an individual younger than 40 years of age with a platelet count less than 30,000/mm 3 was 0.4%. The predicted 5-year mortality rate in this setting was 2.2%. Surgical bleeding does not usually occur until the platelet count is less than 50,000/mm 3 . The American Society of Hematology presently recommends that hospital admission is not necessary unless the platelet count falls to below 20,000/mm 3 or if clinical bleeding is present.
The conventional forms of raising the platelet count in the patient with ITP include glucocorticoid therapy, IV gamma globulins, platelet transfusions, and splenectomy. If the patient has clinical bleeding or if the platelet count is below 20,000/mm 3 , there is usually a need to raise the platelet count in a relatively short period of time. Although oral glucocorticoids can be used, IV glucocorticoids may work more rapidly. Any steroid with a glucocorticoid effect can be used. However, hematologists have had the most experience with methylprednisolone. This medication can be given intravenously and has very little mineralocorticoid effect. It is important to avoid steroids with strong mineralocorticoid effects because these agents can disturb electrolyte balance, cause fluid retention, and result in hypertension. The usual dose of methylprednisolone is 1.0 to 1.5 mg/kg of total body weight intravenously daily in divided doses. It usually takes approximately 2 days for a response, but it may take up to 10 days for a maximum response. Even though methylprednisolone has very little mineralocorticoid effect, some may be observed because of the large dose being administered. Therefore it is important to follow the patient’s electrolytes. The likelihood is low that methylprednisolone will cause neonatal adrenal suppression because little crosses the placenta. It is metabolized by placental 11β-dehydrogenase type 1 to an inactive 11-keto metabolite. Park-Wyllie and colleagues performed a meta-analysis, which confirmed the general safety of glucocorticoids during pregnancy. They did, however, find a 3.4-fold increased risk of cleft lip and palate with first-trimester exposure. The risk/benefit ratio should be discussed with the patient before initiation of therapy (see Chapter 8 ).
After the platelet count has risen satisfactorily using IV methylprednisolone, the patient can be switched to oral prednisone. The usual dose is 60 to 100 mg/day. Prednisone can be given in a single dose, but less gastrointestinal upset occurs with divided doses. The physician can rapidly taper the dose to 30 to 40 mg/day and decrease it slowly thereafter. The dose should be titrated to keep the platelet count at approximately 100,000/mm 3 . If therapy is initiated with oral prednisone, the usual daily dose is 1 mg/kg total body weight.
The likelihood of a favorable response to glucocorticoids is about 70%. It is important to realize that if the patient has been taking glucocorticoids for a period of at least 2 to 3 weeks, she may have adrenal suppression and should undergo increased doses of steroids during labor and delivery to avoid an adrenal crisis. Tapering should be done slowly thereafter. Also, if the patient has been taking glucocorticoids for some time, she may experience significant side effects, including fluid retention, hirsutism, acne, striae, poor wound healing, and monilia vaginitis. In rare circumstances, patients on long-term steroids during gestation can develop osteopenia or cataract formation. The chance of any fetal or neonatal side effects from the glucocorticoids, however, is remote.
Although glucocorticoids are the mainstays of treating maternal thrombocytopenia, up to 30% of patients do not respond to these medications. In such cases, IV immunoglobulin (IVIG) is used. This agent probably works by binding to the IgG Fc receptors on reticuloendothelial cells and preventing destruction of platelets. It may also adhere to receptors on platelets and prevent antiplatelet antibodies from binding to these sites. The usual dose is 0.4 g/kg/day for 3 to 5 days. However, it may be necessary to use as much as 1 g/kg/day. The response usually begins in 2 to 3 days and peaks in 5 days. An alternative regimen is to give 1 g/kg once and observe the patient. Often this single dose will result in an adequate increase in platelets. The length of this response is variable, and the timing of the dose is extremely important. If the obstetrician wants a peak platelet count for delivery, therapy should be instituted about 5 to 8 days before the planned delivery. The most frequent adverse reaction is postinfusion headache, which may be lessened by slowing the infusion rate.
IVIG is a blood product from many pooled donors. Early in its use, concerns were raised about hepatitis C transmission. No recent cases of viral infection from IVIG use have been reported. This is due to careful donor screening as well as an intensive purification process. IVIG should be used before seriously contemplating splenectomy, because some patients experience long-term remission with IVIG, and others have a spontaneous increase in platelet counts postpartum. In severe life-threatening hemorrhage, recombinant factor VIIa can be used in conjunction with other therapies. This is a very expensive and complicated therapy that should only be undertaken with the assistance of a physician familiar with its use.
IV anti-D has been used in emergent settings in Rh-positive, direct antiglobulin–negative patients. In life-threatening situations, when other methods fail, this could be considered an option. The usual dose is 50 to 75 µg/kg. Anti-D binds to IgG Fc receptors different than those bound by IVIG.
The American Society of Hematology (ASH) has made specific recommendations for treating ITP in pregnancy. They state that virtually no indication exists for cordocentesis to determine the fetal platelet count. This group recommends no pharmacologic treatment in the first or second trimesters unless the platelet count is less than 30,000/mm 3 or if clinically significant bleeding is evident. If the count is between 10,000/mm 3 and 30,000/mm 3 in the second or third trimester, IVIG is recommended. The ASH does not recommend platelet transfusion unless the platelet count falls below 10,000/mm 3 .
In the midtrimester, splenectomy can also be used to raise the maternal platelet count. This procedure is reserved for those who do not respond to medical management, with the platelet count remaining below 20,000/mm 3 and with clinical bleeding. It can also be performed postpartum if the patient does not respond to medical management. In extremely emergent cases of life-threatening bleeding or unresponsiveness to other therapies, splenectomy can be performed at the time of cesarean delivery after extending a midline incision cephalad.
In an emergent situation, platelets can be transfused during cesarean delivery if significant clinical bleeding is evident. Platelets can be transfused before a vaginal delivery if the mother’s platelet count is less than 10,000/mm 3 or at any count if clinically significant bleeding is present. Each “pack” of platelets increases the platelet count by approximately 10,000/mm 3 . The half-life of these platelets is extremely short because the same antibodies and reticuloendothelial cell clearance rates that affect the mother’s endogenous platelets also affect the transfused platelets. However, if platelets are transfused at the beginning of surgery, hemostasis adequate to carry out the surgical procedure should be provided.
If the patient with profound thrombocytopenia undergoes cesarean delivery, certain surgical precautions should be taken. Needless to say, the key is adequate surgical hemostasis. The bladder flap may be left open to avoid hematoma formation. If the parietal peritoneum is closed, subfascial drains are helpful if hemostasis is imperfect. If the peritoneum is not closed, the peritoneal edges should be carefully inspected to make certain no bleeding vessels are present. Small “bleeders” may not be apparent if the patient is hypotensive; therefore the operator must watch for bleeding as the blood pressure rises toward the end of the surgical case. If severe, life-threatening hemorrhage occurs, recombinant factor VIIa and platelet transfusion can be used.
In summary, the treatment of thrombocytopenia during gestation is dependent on its etiology. The obstetrician need not act on the mother’s platelet count unless it is below 30,000/mm 3 , if it is below 50,000/mm 3 with evidence of clinical bleeding, or if surgery is anticipated. In these cases, the treatment depends on the diagnosis. Furthermore, whether delivery needs to be expedited or can be delayed is also dependent on the etiology of thrombocytopenia, the patient’s physical health, fetal well-being, and gestational age.
Management of Thrombotic Thrombocytopenic Purpura and Hemolytic Uremic Syndrome
Before the use of plasma exchange, maternal and fetal outcomes in pregnancies complicated by TTP were uniformly poor. The first cases treated with plasma exchange for TTP during pregnancy were reported in 1984, and no large series of patients with TTP in pregnancy have been undertaken. A review of 11 patients described in case reports reveals that the prognosis has improved greatly with plasma infusion and plasma exchange. These researchers also demonstrated that cyclosporin may increase the duration of remission. In one case report, TTP relapses were prevented by using prophylactic monthly plasma exchange throughout gestation. If TTP is suspected, plasma exchange should be initiated immediately.
HUS has been more difficult to treat, and only a few case reports have appeared. Supportive therapy remains the mainstay in cases of HUS, although dialysis is often necessary with close attention to fluid management. Platelet function inhibitors were used in two cases during pregnancy. Plasma infusion and plasma exchange can be attempted, but the results have not been as good as those observed in cases of TTP. Vincristine has been administered with some success in nonpregnant patients but has not been tried in pregnancy, and prostacyclin infusion has been effective in children but has not been used during pregnancy.
Fetal/Neonatal Alloimmune Thrombocytopenia
In neonatal alloimmune thrombocytopenia, a rare disorder, the mother lacks a specific platelet antigen and develops antibodies to this antigen. The disease is somewhat analogous to Rh isoimmunization but involves platelets. If the fetus inherits an antigen from its father and the mother lacks the antigen, maternal antibody can develop and can cross the placenta. This results in severe neonatal thrombocytopenia and possibly fetal intracranial hemorrhage. The mother, however, will have a normal platelet count. The most common antibodies noted in these patients is anti–HPA-1a antibodies, although several other antibodies have been identified. If this disorder is suspected, the mother’s blood should be sent to a reference laboratory with experience in diagnosing neonatal alloimmune thrombocytopenia. These patients should be managed in a tertiary care center with experience caring for mothers and infants with this rare disorder. Transfusion of maternal platelets into the neonate has improved outcome in these cases. After birth or in utero, the child can be transfused with the mother’s platelets (because she lacks the antigen) or with donor platelets known to lack the antigen. Bussel and colleagues demonstrated that neonates who had an older sibling that was affected, especially with an antenatal intracranial hemorrhage, had lower platelet counts than the index pregnancy. Pacheco and colleagues have described an excellent algorithm based on risk stratification for evaluating mothers at risk of having a neonate with alloimmune thrombocytopenia. It describes all of the testing that should be performed. McQuilten and colleagues reviewed the experience in Australia and showed that cordocentesis with transfusion, IVIG administration, and corticosteroids have all been used with good results. Kamphuis and Oepkes reviewed the experience in the Netherlands and demonstrated that weekly IVIG alone can virtually prevent fetal/neonatal intracranial hemorrhage in neonatal alloimmune thrombocytopenia; therefore they believe that fetal blood sampling should be abandoned because of its risks. Rayment and coworkers searched the Cochrane database and the Childbirth Group’s trial register to ascertain whether they could discern the optimal management of alloimmune thrombocytopenia. They reviewed four trials that involved 206 patients. Because of incomplete data and differences in interventions, they could not conclude the best treatment plan for these patients. They convincingly show that more randomized studies to look at medication doses and timing need to be performed.
Iron Deficiency Anemia
During a singleton pregnancy, maternal plasma volume gradually expands by approximately 50% (1000 mL). The total RBC mass also increases but only by approximately 300 mg (25%), and this starts later in pregnancy. It is not surprising, therefore, that hemoglobin and hematocrit levels usually fall during gestation. These changes are not necessarily pathologic but usually represent a physiologic alteration of pregnancy. By 6 weeks postpartum, in the absence of excessive blood loss during the puerperium, hemoglobin and hematocrit levels have returned to normal if the mother has adequate iron stores.
Most clinicians diagnose anemia when the hemoglobin concentration is less than 11 g/dL or the hematocrit is less than 32%. Using these criteria, 50% of pregnant women are anemic. Many women have hemoglobin concentrations as low as 10 gm/dL and recover. The incidence of anemia changes depending on the population studied. It is unfortunate that this problem is often ignored; in developing nations, iron deficiency is an overwhelming problem, and worldwide, many maternal deaths occur because of excessive blood loss in those who were already anemic. Causes of anemia are shown in Box 44-3 .
Common Causes: 85% of Anemia
Physiologic anemia
Iron deficiency
Uncommon Causes
Folic acid deficiency
Vitamin B 12 deficiency
Hemoglobinopathies
- •
Sickle cell disease
- •
Hemoglobin SC
- •
β-Thalassemia minor
- •
Bariatric surgery
Gastrointestinal bleeding
Rare Causes
Hemoglobinopathies
- •
β-Thalassemia major
- •
α-Thalassemia
- •
Syndromes of chronic hemolysis
- •
Hereditary spherocytosis
- •
Paroxysmal nocturnal hemoglobinuria
- •
Hematologic malignancy
Approximately 75% of anemia that occurs during pregnancy is secondary to iron deficiency. Ho and colleagues performed elaborate hematologic evaluations of 221 gravidas at term in Taiwan. None of the studied patients received an added iron preparation during gestation. Of the previously nonanemic patients, 10.4% developed clinical anemia after a full-term delivery. Of these 23 patients, 11 (47.8%) developed florid iron deficiency anemia, and another 11 demonstrated moderate iron depletion. The other anemic patient in the group had folate deficiency. Of the 198 nonanemic gravidas at term, 46.5% showed evidence of iron depletion stores even though they had a normal hematocrit.
To distinguish the normal physiologic changes of pregnancy from those of pathologic iron deficiency, the normal iron requirements of pregnancy ( Table 44-1 ) and the proper use of hematologic laboratory parameters must be understood. In adult women, iron stores are located in the bone marrow, liver, and spleen in the form of ferritin, which constitutes approximately 25% (500 mg) of the 2 g of iron stores found in the normal woman. Approximately 65% of stored iron is located in the circulating RBCs. If the dietary iron intake is poor, the interval between pregnancies is short, or the delivery is complicated by hemorrhage, iron deficiency anemia readily and rapidly develops.
| FUNCTION | REQUIREMENT |
|---|---|
| Increased red blood cell mass | 450 mg |
| Fetus and placenta | 360 mg |
| Vaginal delivery | 190 mg |
| Lactation | 1 mg/day |
The first pathologic change to occur in iron deficiency anemia is the depletion of bone marrow, liver, and spleen iron stores. This may take a few weeks to a few months depending on the level of the woman’s iron stores. Over a period of a few weeks, the serum iron level falls, as does the percentage saturation of transferrin. The total iron-binding capacity rises simultaneously with the fall of iron, because this is a reflection of unbound transferrin. A falling hemoglobin and hematocrit follow within 2 weeks. Microcytic hypochromic RBCs are released into the circulation. If this is a pure iron deficiency, a reticulocytosis will occur within 3 days of initiating therapy, and the hemoglobin concentration will increase within a week. However, it may take more than a month to completely replete iron stores. A patient who has a very low hemoglobin as a result of iron deficiency immediately postpartum should return to normal by her 6-week postpartum visit. If iron deficiency is combined with folate or vitamin B12 deficiency, normocytic and normochromic RBCs are observed on the peripheral blood smear.
Care must be taken when using laboratory parameters to establish the diagnosis of iron deficiency anemia during gestation. A serum iron concentration less than 60 mg/dL with a less than 16% saturation of transferrin is suggestive of iron deficiency. Conversely, a single normal serum iron concentration does not rule out iron deficiency. For example, a patient may take iron for several days, and this may result in a transiently normal serum iron concentration while iron stores are still negligible. An increase in iron-binding capacity is not reliable, however, because 15% of pregnant women without iron deficiency show an increase in this parameter. If a patient has been iron deficient for an extended period of time, her serum iron level can rise before she has depleted her iron stores. The ferritin level indicates the total status of her iron stores. Serum ferritin levels normally decrease minimally during pregnancy. However, a significantly reduced ferritin concentration is indicative of iron deficiency anemia and is the best parameter to judge the degree of iron deficiency. However, ferritin levels are variable and can change 25% from one day to the next. Tran and colleagues have demonstrated that iron deficiency is the only possible diagnosis for a low ferritin. If a ferritin of 41 ng/mL is used as a cutoff, serum ferritin has 98% sensitivity and 98% specificity in diagnosing iron deficiency. This is true if there is no concomitant infectious or inflammatory process.
Ahluwalia compared iron status in normal and obese pregnant women and found by comparing ferritin that obese women had decreased iron stores. It was also found that obese pregnant women had higher concentrations of the inflammatory marker hepcidin, the concentration of which correlated directly with iron status; it was therefore surmised that chronic inflammation in obese pregnant women may impede their ability to absorb iron.
As part of a large study that included 1171 pregnant women between 1999 and 2006, Mei and colleagues, working through the National Centers for Chronic Disease Prevention and Health Promotion, assessed total body iron using ferritin and soluble transferrin receptor concentrations. They found that iron deficiency increased during pregnancy from 6.9% ± 2.2% in the first trimester to 29.5% ± 2.7% in the third trimester. The prevalence of iron deficiency was highest in women with a parity of at least two. Iron deficiency was significantly higher in Mexican-American and non-Hispanic black women. Statistical analysis showed that this difference was not due to educational level or family income.
Bone marrow aspiration is rarely necessary for the diagnosis of iron deficiency. It is reserved for persistent anemia with confusing hematologic parameters and can be safely performed during pregnancy.
Whether all women should receive prophylactic iron in addition to that contained in prenatal vitamins during pregnancy remains controversial. In reviewing the Cochrane database, Milman and colleagues found that 20% of fertile women have iron stores greater than 500 mg, which is the required minimum for pregnancy. They also noted that 40% of women have iron stores between 100 and 500 mg, and 40% have virtually no iron stores. Based on these data, most women do need some iron supplementation. No consensus was reached, however, on how much iron supplementation may be needed in patients with iron deficiency.
In pregnancy, iron absorption from the duodenum increases, providing 1.3 to 2.6 mg of elemental iron daily. An acid environment in the duodenum helps this absorption; therefore the frequent ingestion of antacid medications commonly used by many patients decreases the absorption of iron. Chronic use of H 2 blockers and proton pump inhibitors also diminishes iron absorption. Vitamin C, in addition to the iron, may increase the acidic environment of the stomach and increase absorption. In patients who do not show clear signs of iron deficiency, it is uncertain whether prophylactic iron, in addition to what is in prenatal vitamins, leads to an increased hemoglobin concentration at term. Iron prophylaxis, however, is safe because only amounts that can be used are absorbed. With the exception of dyspepsia and constipation, side effects are few. One 325 mg tablet of ferrous sulfate daily provides adequate prophylaxis. It contains 60 mg of elemental iron, 10% of which is absorbed. If the iron is not needed, it will not be absorbed and will be excreted in the feces. The standard generic iron tablets and the amount of elemental iron they provide are listed in Table 44-2 .
| PREPARATION | ELEMENTAL IRON (mg) |
|---|---|
| Ferrous gluconate 325 mg | 37-39 |
| Ferrous sulfate 325 mg | 60-65 |
| Ferrous fumarate 325 mg | 107 |
In iron-deficient patients, one iron tablet three times daily has been recommended, although the evidence-based source of this recommendation is difficult to ascertain. Most individuals can absorb as much iron as they need taking iron twice daily. Iron should be taken 30 minutes before meals to allow maximum absorption. However, when taken in this manner, dyspepsia and nausea are more common. Therapy, therefore, must be individualized to maximize patient compliance. Reveiz and colleagues examined the Cochrane database to see whether an optimal treatment for iron deficiency during pregnancy could be discerned. They identified 23 trials that comprised 3198 women. Many of the trials were from low-income countries, were generally small, and had poor methodology. Although oral iron supplementation led to a reduction in the incidence of anemia, it was not possible to assess the effects of treatment by severity of anemia. The authors concluded that despite a high incidence and significant ramifications of this disease, a paucity of good-quality trials exists. Whereas these trials could be relatively easy to design and would be relatively inexpensive compared with so many other investigations being carried out, a lack of interest among researchers and funding institutions in the United States is apparent.
Young and colleagues studied the effectiveness of weekly iron supplementation and found it to be almost as effective as daily supplementation in raising the hemoglobin concentration in iron-deficient patients. This approach can be used in patients with less than optimal compliance. Yakoob and Bhutta systematically reviewed 31 studies in the Cochrane database to determine whether routine iron supplementation affects the incidence of anemia in pregnancy. They included studies that used iron alone and iron with folic acid and found a 73% reduction in anemia with routine supplementation. However, no difference was found in rates of anemia at term with intermittent iron-folate when compared with daily supplementation (relative risk [RR], 1.61; 95% confidence interval [CI], 0.82 to 3.14).
For those patients who are noncompliant or unable to take oral iron and are severely anemic, IV iron can be given. Singh and colleagues found that parenteral iron can be safely given and significantly raises the hematocrit in patients. It also raises the serum ferritin. Hallak and coworkers examined the safety and efficacy of parenteral iron administration. Of 26 patients receiving parenteral iron, only one developed signs of mild allergy during the test dose and was excluded from the study. The remaining 21 pregnant patients completed the course of therapy and received a mean of 1000 mg of elemental iron. Their hemoglobin increased an average of 1.6 g/dL from the beginning to the end of therapy and rose another 0.8 g/dL during the following 2 weeks. Ferritin levels increased from 2.9 ng/mL at the beginning of therapy to 122.8 ng/dL by the end of treatment. Ferritin levels decreased to a mean of 109.4 ng/mL 2 weeks later, demonstrating that the iron was being utilized. Only mild transient side effects were noted; therefore the authors concluded that parenteral iron therapy can be used safely during pregnancy.
Parenteral iron is indicated in those who cannot or will not take oral iron therapy and are not anemic enough to require transfusion. In fact, by building iron stores in the patients before delivery, we may be able to prevent a need for transfusion postpartum in the severely anemic patient. Iron dextran comes in a concentration of 50 mg/mL. It can be given intramuscularly or intravenously, although intramuscular injection is very painful. Iron dextran can result in anaphylaxis caused by dissociation of the iron and carbohydrate components. The reaction may be immediate or delayed; therefore a 0.5 mL test dose should be given, and epinephrine should be readily available. Anaphylaxis usually occurs within several minutes but may take 2 days to develop. In the past 3 years, our group has given iron dextran to 14 patients. Two developed a severe reaction within minutes of the test dose. Although neither patient developed shortness of breath, both exhibited severe bone pain and myalgias. The dosage for iron dextran therapy is shown in Table 44-3 . Although iron dextran is rarely used in the United States and Canada, in developing nations, this compound is the only parenteral iron readily available, and therefore its discussion is included in this chapter.
