ANTICOAGULANTS IN PREGNANCY

Indications for anticoagulant therapy during pregnancy include prevention of venous thromboembolism in patients with familial thrombophilia or previous venous thrombosis. treatment of acute venous thromboembolism. and prevention of systemic embolism in patients with prosthetic heart valves or native valvular heart disease. Heparin and/or acetylsalicylic acid are also used in selected patients to prevent fetal growth retardation and pregnancy loss in individuals with the antiphospholipid antibody syndrome or pregnancy-induced hypertension. Any anticoagulant therapy has the potential to cause maternal bleeding. Heparin can also cause maternal osteoporosis and thrombocytopenia (Levine. 1986), whereas warfarin has teratogenic potential and can cause fetal bleeding. In this chapter. we will review the effects on the fetus and infant of anticoagulant therapy during pregnancy and lactation. respectively. Effects of heparin on the fetus: Heparin is a heterogenous sulphated. long-chain acidic glycosaminoglycan that is extracted from beef or porcine lung or gut. It functions as a catalytic cofactor for an endogenous inhibitor of serine proteases called anti-thrombin. Heparin exerts its effect by complexing with and changing the molecular configuration of antithrombin. thereby increasing its ability to neutralize thrombin. Factor Xa. Factor IXa. Factor XIa. and Factor XIIa (Salzman et al., 1994). The molecular weight of heparin ranges from 4000 to 30 000 daltons. with an average of 12 000 to 15 000 daltons (Salzman et al., 1994). Heparin does not cross the placenta (Flessa et al., 1963) and. therefore. might not be expected to produce fetal complications. However. a review of anticoagulant therapy during pregnancy did conclude that the use of anticoagulants was associated with adverse fetal outcomes in approximately one-third of cases. regardless of whether heparin or coumarin derivatives were used. In 135 published cases in which

heparin was the sole anticoagulant used during pregnancy. one-eighth of the infants were stillborn and one-fifth were premature. Of the latter. one-third died (Hall et al., 1980). However. approximately one-third of the pregnancies in which heparin was used were associated with significant maternal co-morbid conditions. including severe toxaemia. glomerulonephritis. and recurrent abortions. and almost two-thirds of the adverse fetal outcomes occurred in this sub-group. Moreover. many of the adverse outcomes in the heparin-treated patients were prematurity (gestational age less than 37 weeks) where the infants ultimately suffered no permanent disability or deformity. Two subsequent publications suggest that maternal heparin therapy is safe for the fetus. In the first. Ginsberg et al reviewed 186 studies published between 1966 and 1986 which described fetal or infant outcomes in 1325 pregnancies associated with anticoagulant therapy (Table 15). The cases were classified into three groups according to anticoagulant used: heparin alone. oral anticoagulants alone. or heparin and oral anticoagulants. Four analyses were performed. The first included the reported adverse fetal outcomes of all the pregnancies. In this analysis. the rates were similar in all three groups. with adverse events being identified in 21.7% of the heparin-treated patients. 27.9% of the patients treated with oral anticoagulants alone. and 19.4% of those treated with both heparin and oral anticoagulants. The second analysis excluded all pregnancies. regardless of fetal outcome. that had co-morbid conditions independently associated with adverse fetal outcomes. These co-morbid conditions were determined prior to data accumulation and analysis by an obstetrician experienced in high risk pregnancies. Because a large number of pregnancies in the heparintreated patients were associated with maternal co-morbid conditions. the frequency of adverse events in that group fell dramatically from 21.7 to 10.4% with the exclusion of such pregnancies. The frequency of adverse outcomes fell only slightly in the other two groups--to 26.5% in those treated with oral anticoagulants alone and 18.3% in those receiving both heparin and oral anticoagulants. In the third analysis. premature births (less than 37 weeks' gestation) in which the

infant suffered no permanent deformity or disability were excluded from the second analysis. Again. a significant decrease in the frequency of adverse outcomes was noted in the group of patients treated with heparin alone; the rate fell from 10.4 to 3.6% (Bates & Ginsberg . 1997). deaths. and congenital malformations were similar to those of a comparative normal population. Adverse neonatal or fetal outcomes were noted in 17.0% of all pregnancies. Prematurity occurred in 11.6%. spontaneous abortion in 8.8%. stillbirth in 3.1%. neonatal death in 1.1% and congenital malformation in 1.1% of all pregnancies. If pregnancies with co-morbid conditions identified as independent risk factors for adverse neonatal or fetal outcomes were excluded. then prematurity occurred in 8.8%. spontaneous abortion in 3.3%. stillbirth in 3.2%. neonatal death in 1.1%. and congenital malformation in none of the remaining pregnancies (Ginsberg et al., 1989b).

Little change was noted in the other two groups. with adverse outcomes occurring in 26.1% of the patients treated solely with oral anticoagulants and 18.3% of those treated with both heparin and oral anticoagulants. The prematurity rate of 6.8% in the heparin-treated group was not dissimilar to that of a normal population with which it was compared. The fourth analysis included only abortions. stillbirths and neonatal deaths as adverse outcomes in those pregnancies without comorbid conditions. In that analysis. the rate of adverse outcomes was 2.5% in the patients treated with heparin alone. 16.8% in the patients treated with oral anticoagulants alone. and 11.5% in the patients treated with both heparin and oral anticoagulants. The death rate in the heparin treated patients was again similar to that of a normal population. This study. therefore. concluded that the previously reported high rate of adverse fetal and infant outcomes associated with heparin. but not coumarin. therapy could be explained by the frequent presence of co-morbid conditions and the higher rate of uncomplicated prematurity in that population (Ginsberg et al., 1989a). The second report was a retrospective cohort study of 100 consecutive pregnancies in 77 women associated with the use of heparin therapy (Table 2). In 98 pregnancies. heparin therapy was given for prevention or treatment of venous thromboembolic disease. whereas in the other two it was give deaths. and congenital malformations were similar to those of a comparative normal population. Adverse neonatal or fetal outcomes were noted in 17.0% of all pregnancies. Prematurity occurred in 11.6%. spontaneous abortion in 8.8%. stillbirth in 3.1%. neonatal death in 1.1% and congenital malformation in 1.1% of all pregnancies. If pregnancies with co-morbid conditions identified as independent risk factors for adverse neonatal or fetal outcomes were excluded. then prematurity occurred in 8.8%. spontaneous abortion in 3.3%. stillbirth in 3.2%. neonatal death in 1.1%. and congenital malformation in none of the remaining pregnancies (Ginsberg et al., 1989b).

n because of the presence of prosthetic heart valves. The rates of prematurity. abortions. stillbirth. Neonatal The true incidence of warfarin embryopathy is difficult to estimate from studies because of the likelihood of case-reporting bias and a lack of uniform reporting of the trimester of exposure to anticoagulants. In one study. of the patients taking warfarin between the sixth and ninth weeks of gestation. approximately 8% of infants were reported to have features of this embryopathy (Hall et al., 1980). However. in a prospective study of 72 pregnancies in 63 women with prosthetic heart valves. Iturbe-Alessio et al reported that warfarin embryopathy occurred in 28.6% of all infants exposed to warfarin between the sixth and twelfth week of gestation (Iturbe-Alessio et al., 1986). It is likely that the incidence of warfarin embryopathy can be markedly reduced by avoiding coumarin exposure between the sixth and twelfth weeks of gestation. In the study by Hall et al., (1980). all 24 patients in whom warfarin embryopathy was described were exposed to warfarin between the sixth and ninth week of gestation. In another study. no cases of warfarin embryopathy occurred in the 19 patients in whom heparin was substituted for warfarin between the sixth and twelfth week of pregnancy (Iturbe-Alessio et al., 1986). In the review by Hall et al (1980). the severity of warfarin embryopathy was variable. One-half had no severe disability. In severe cases. however. the nose appeared sunke into the face. In some. true choanal stenosis and variable upper airway obstruction was reported. The nasal hypoplasia may not fully resolve during growth in childhood. The calcific stippling is incorporated into epiphyses and has not been reported to result in asymmetric growth (Hall et al., 1980). The central nervous system abnormalities associated with maternal oral anticoagulant use include dorsal midline dysplasia with agenesis of the corpus callosum. Dandy-Walker malformations. and midline cerebellar atrophy; ventral midline dysplasia with optic atrophy and blindness. and haemorrhage (Hall et al. 1980). Unlike

warfarin embryopathy. which has been reported with only firsttrimester warfarin exposure. central nervous system abnormalities appear to occur after warfarin exposure during any trimester (Hall et al., 1980). In the literature review by Hall et al., sequelae from the central nervous system abnormalities appeared more significant and debilitating than those of the warfarin embryopathy. None of the affected infants were normal on follow-up. Mental retardation was found in all. blindness in one-half. spasticity in some one-third. and seizures in approximately one-quarter of all liveborn infants (Hall et al. 1980). Like warfarin embryopathy. the true incidence of central nervous system abnormalities is difficult to determine. In the study by Hall et al., 3% of live-born infants exposed to coumarin derivatives during pregnancy demonstrated central nervous system abnormalities (Hall et al., 1980). None of the 54 infants in the study by Iturbe-Alessio et al (1986) had any central nervous system anomalies. Similar results have been reported in multiple other studies (Ibarra-Perez et al. 1976; Chong et al. 1984; Salazar et al. 1984; Vitali et al., 1986). Suggesting that the incidence of such events is low although not negligible. Early studies by Quick (1946) and by Kraus and his colleagues (Kraus et al., 1949) demonstrated a high incidence of fetal haemorrhage and death when coumarin was given to pregnant dogs and rabbits. In both studies. because large doses of anticoagulants were given until term. the fetuses were delivered when their coagulation systems were severely impaired. In an early literature review. Villasanta (1963) suggested that similar outcomes could be expected in humans. Hirsh. Cade and Gallus re-examined the effects of giving oral anticoagulants to pregnant rabbits with the goal of determining the causal factors responsible for fetal and neonatal haemorrhage. When coumarin was given from early pregnancy until term to five rabbits. all 26 fetuses were stillbom with evidence of widespread haemorrhage. However. when coumarin was stopped 4-5 days prior to delivery. allowing the level of coagulation factors to return to near-normal levels. and when delivery was performed by caesarean section at a time when the fetal

coagulation defect was severe. all fetuses were born alive and without haemorrhage (Hirsh et al., 1970a). Their results. therefore. suggested that bleeding in the fetus with abnormally depressed vitamin K-de prospective investigation by Hirsh. Cade and O'Sullivan of 14 patients treated with anticoagulants during 15 pregnancies. reported no fetal or neonatal complications when oral anticoagulants were withdrawn after 37 weeks' gestation and replaced with heparin (Hirsh et al., 1970b). It can. therefore be concluded that coumarin derivatives should. if possible. be avoided during pregnancy. A woman who conceives while taking oral anticoagulants should be advised to discontinue their use prior to 6 weeks of gestation. if this is feasible. If warfarin is necessary in situations with a high risk of thrombosis. its use should be restricted to the second and third trimesters and avoided between 6 and 12 weeks of gestation and near term to avoid delivery of an anticoagulated fetus (Hirsh and Hoak. 1996).

Effects of low-molecular-weight heparins on the fetus: Low-molecular-weight heparins are heparin fragments produced by chemical or enzymatic depolymerization of unfractionated heparin (Hirsh and Levine. 1992). They differ in preparation method. molecular weight distribution. specific activity. clearance routes and rates. and optimal dosage and. therefore. one cannot assume that specific clinical properties are common to different low-molecular-weight heparin preparations (Hirsh and Levine. 1992). Low-molecular-weight heparins may have some advantages during pregnancy because of: their long half-pendent coagulation factors was related to the trauma of delivery. A performed on the plasma of each mother on multiple occasions revealed prothrombin times of 20-30% of control. and Factor II. Factor VII and Factor X levels of approximately 20%. No warfarin was detected in the breast milk. The prothrombin times and Factor II. Factor VII and Factor assays

performed on the plasma of their infants showed 100% of control activity.