Maternal Antiplatelet Antibodies Research Articles

Thrombocytopenia in pregnant women.

Thrombocytopenia is one of the two most common hematological problems in pregnant women. It is defined as the platelet (PLT) count below 150 × 103/μL. Gestational incidental thrombocytopenia (GIT) represents about 75% of thrombocytopenia cases in pregnancy and it is believed that GIT is secondary to accelerated platelet destruction and increased plasma volume associated with pregnancy. The pregnancy complications such as preeclampsia and its most severe form - HELLP syndrome account for 20% cases of thrombocytopenia in pregnancy and primary immune thrombocytopenic purpura (ITP) - for 3-4 percent. During ITP, maternal antiplatelet antibodies can pass through the placenta and bind to fetal thrombocytes leading to the development of fetal thrombocytopenia which occurs in about 50% cases. Even if the maternal platelet count stabilizes, the estimated fetal and neonatal risk of thrombocytopenia in ITP is approximately 30%. Other types of thrombocytopenia in pregnant women constitute 1-2% of cases (disseminated intravascular coagulation, autoimmunological diseases, congenital, infection and drug-related, concomitant with blood neoplastic diseases). Although thrombocytopenia in pregnant women usually has a mild course, in case of a significant decrease in PLT count may lead to dangerous bleeding, especially when the platelet count falls below 20 × 103/μL. Since it is important to identify the cause of thrombocytopenia and to determine the risk for both the mother and the child, this paper presents the influence of maternal thrombocytopenia on the pregnancy course as well as its etiology and diagnostics. The treatment principles are discussed.

Circulating antiplatelet antibodies in pregnant women with immune thrombocytopenic purpura as predictors of thrombocytopenia in the newborns

Newborns from mothers with immune thrombocytopenic purpura (ITP) have a risk of thrombocytopenia due to passage of maternal antiplatelet antibodies into fetal/neonatal circulation. We looked for predictors of neonatal thrombocytopenia (nTP) in pregnant women with ITP. One hundred pregnant women with platelet count <100 × 109/l, no non-immune causes of thrombocytopenia and increased platelet associated IgG (PA-IgG) were included in the study. Thirty seven and 63 of them gave birth to babies with and without nTP, respectively (nTP+ and nTP− groups). Platelet count, mean platelet volume, PA-IgG, antiplatelet circulating antibodies (cAB), time of ITP onset (before or during pregnancy), and frequency of corticosteroid treatment were compared in these groups. There were no differences in all test parameters between nTP+ and nTP− groups except cAB. These antibodies were detected in 33 out of 37 in nTP+ group and in 2 out of 63 mothers in nTP− group (p < 0.001). The sensitivity of this test was 89% and its specificity was 97%. A strong reverse correlation (r = −0.749, p < 0.001) was established between maternal cAB titer and neonatal platelet count. Antibodies against glycoproteins IIb–IIIa and/or Ib were identified in antigen specific MAIPA (Monoclonal Antibody Immobilization of Platelet Antigen) assay only in 10 out of 19 (53%) test sera with cAB. Antiplatelet cAB in pregnant women with ITP could serve as reliable predictors of nTP in their babies.

From thrombasthenia to next generation thrombocytopenia: Neonatal alloimmune thrombocytopenia induced by maternal Glanzmann thrombasthenia.

Glanzmann thrombasthenia (GT) is a rare autosomal recessive disorder of platelet function caused by mutations in the genes coding for integrin αIIbβ3. The aim of this study was to examine the outcome of newborns of GT mothers, with emphasis on thrombocytopenia and bleeding manifestations and their relation to maternal antiplatelet antibodies. Medical files of all female patients with GT treated in a single tertiary center from 1999 to 2017 were searched for details on pregnancy and birth. The medical files of their newborns were retrieved, and data on the postnatal course were collected. Nine babies were born to five patients with GT at our center during the study period. Three of the nine newborns had severe thrombocytopenia, and all three were offspring of GT mothers who were positive for antiplatelet antibodies. Pregnant GT patients should be examined for platelet antibodies. Assessment and management protocols (including treatment with intravenous immunoglobulins) for fetal and neonatal alloimmune thrombocytopenia should be considered.

Fetal and Neonatal Alloimmune Thrombocytopenia: Lessons Learned from Animal Models

Fetal and neonatal alloimmune thrombocytopenia (FNAIT) is a severe alloimmune disorder that results from fetal/neonatal platelet opsonization by maternal antibodies, which cross the placenta and result in fetal/neonatal platelet destruction. The frequency of FNAIT has been estimated at 0.5-1.5/1,000 liveborn neonates. However, this number does not include fetuses that die from this disease, since the incidence of FNAIT miscarriage has not been adequately studied. Analogous to autoimmune thrombocytopenia (ITP), the major target antigens in FNAIT are the platelet GPIIIa (β3 integrin) and GPIbα. However, severe bleeding is much more frequent in FNAIT, particularly the occurrence of intracranial hemorrhage (ICH). It is not known why the reported incidence of FNAIT mediated by anti-GPIbα antibodies is at a much lower frequency when compared to ITP, and whether the severe bleeding tendency in FNAIT is due to β3 integrin expression on angiogenic vessels in the developing fetus, which are targeted by cross-reacting maternal anti-β3 integrin antibodies. To study the pathogenesis and to develop new strategies for prevention and treatment, we have established murine models of FNAIT using β3 integrin and GPIbα deficient (-/-) mice. We first transfused these deficient mice with wild-type (WT) platelets to induce anti-β3 or anti-GPIbα antibody responses; we then bred these immunized female mice with WT males, causing FNAIT in the offspring. We found that maternal antiplatelet antibody titer correlated with the severity of FNAIT. These two murine models have revealed fundamental differences between the pathogenesis of anti-β3 and anti-GPIbα-mediated FNAIT. In anti-β3-mediated FNAIT, we found severe thrombocytopenia, ICH, and miscarriage. We also found the impairment of angiogenesis, which may contribute to ICH and intrauterine growth retardation. In contrast, the anti-GPIbα-mediated model revealed a nonclassical form of FNAIT (e.g., miscarriage but not bleeding disorders in neonates). We found that anti-GPIbα antibodies caused thrombosis in the placentas and miscarriage in most pregnant mice, which may partially explain the rarity of anti-GPIbα-mediated FNAIT reported in humans. Despite these substantial differences, there are also similarities between anti-β3 and anti-GPIbα-mediated FNAIT, such as the role of the neonatal Fc receptor (FcRn). FcRn is important for serum IgG homeostasis and for IgG transportation across the placenta. Using FcRn-/- mice, we demonstrated that fetal (but not maternal) FcRn is required to transport maternal antibodies to the fetal circulation and is indispensable for FNAIT. Blocking FcRn with an anti-FcRn antibody markedly reduced the severity of both anti-β3 and anti-GPIbα-mediated FNAIT. This important finding may lead to the development of new therapies (e.g., anti-FcRn antibody) against FNAIT and other maternal pathogenic antibody-mediated fetal/neonatal diseases. These models of FNAIT have allowed us to investigate the efficacy and mechanism of action of several therapies, including the aforementioned anti-FcRn antibody as well as antibody-mediated immune suppression (AMIS) and intravenous IgG (IVIG). In β3-/- mice, prophylactic administration of anti-HPA-1a antibody or murine β3 antisera induced AMIS against human HPA-1a-positive or murine WT platelets, respectively. Importantly, AMIS induced by β3 antisera suppressed the antibody response, thrombocytopenia, and miscarriage in FNAIT mice. These findings support the hypothesis that anti-HPA-1a antibody administration to HPA-1a-negative women after delivery of an HPA-1a-positive child may prevent FNAIT in subsequent pregnancies. The efficacy of IVIG is inconsistent, sources are limited, and the mechanism of action is not fully understood. We demonstrated that IVIG ameliorated both anti-β3 and anti-GPIbα-mediated FNAIT. IVIG decreased antiplatelet antibodies in both maternal and neonatal circulations, fetal platelet clearance, bleeding, and fetal mortality. In summary, we have uncovered fundamental differences in the pathogenesis of anti-β3 and anti-GPIbα-mediated FNAIT and have greatly enhanced our understanding of emerging and existing therapies. We will continue to investigate the pathogenesis of FNAIT so that we may develop more tailored and accessible therapeutic strategies for patients. Disclosures:No relevant conflicts of interest to declare.

Are maternal antiplatelet antibodies a prothrombotic condition leading to miscarriage?

Fetal and neonatal alloimmune thrombocytopenia (FNAIT) is a condition characterized by thrombocytopenia in the newborn. If severe, the thrombocytopenia can lead to intracranial hemorrhage. FNAIT arises when maternal antibodies specific for platelet antigens, most commonly β3 integrin, cross the placenta and destroy fetal platelets. Surprisingly, few cases of FNAIT are associated with antibodies specific for the platelet antigen GPIbα, which is a common target in patients with immune thrombocytopenia. In this issue of the JCI, Li et al. have identified a potential reason for this - they find that in the majority of pregnant mice, anti-GPIbα antibodies enhance platelet activation and accelerate thrombus formation in the placenta and that this leads to miscarriage.

Murine Model of Fetal and Neonatal Alloimmune Thrombocytopenia Mediated by Anti-GPIb α Versus Anti-β3 Integrin Antibodies.

Fetal and neonatal alloimmune thrombocytopenia (FNAITP) is a life-threatening bleeding disorder which results from maternal anti-platelet antibodies that cross the placenta and destroy fetal platelets. In most cases FNAITP is mediated by anti-β3 integrin antibodies, whereas reported cases of anti-GPIbα mediated FNAITP are rare. This difference can not be solely explained by the frequency of their respective polymorphisms. It is unclear whether this is due to:the GPIbα antigen being less immunogenic, resulting in less maternal antibody generation during pregnancy, oranti-GPIbα antibodies mediating a less severe pathology, resulting in a reduced number of reported cases.To study the immunogenecity and antibody response to the GPIbα antigen, GPIbα deficient mice (GPIbα−/−, Black Swiss background) were transfused with wild-type (WT) platelets. No detectable anti-GPIbα IgG antibody was induced even after 8 transfusions (108 platelets/transfusion). This clearly differs from the immune response generated in β3 integrin deficient (β3−/−, BALB/c background) mice, in which anti-β3 integrin antibody can be easily detected after 2–4 WT platelet transfusions. This suggested that the MHC complex in Black Swiss mice might not be able to present the GPIbα antigen. We thus introduced the BALB/c background to the GPIbα−/− mice via backcrossing these mice with BALB/c mice for 9 generations (F9). To further minimize genetic background differences, the F9 BALB/c GPIbα−/− mice were bred with F9 BALB/c β3−/− mice to generate heterozygous BALB/c GPIbα+/− and β3+/− mice. These mice were subsequently used to generate littermate GPIbα−/− or β3−/− mice for the following studies. We found that BALB/c GPIbα−/− mice are immunoresponsive to the GPIbα antigen, but antibody titers after 2 and 4 platelet transfusions were significantly lower than those seen with the anti-β3 integrin model (1:50 & 1:200 vs 1:400 & 1:3200 respectively, P<0.05). We then established a FNAITP model with these GPIbα−/− mice and compared it with syngeneic background β3−/− mice. Naïve (Group I) or 2-time-platelet-immunized (Group II) female GPIbα−/− or β3−/− mice were bred with WT male mice. In Group I, neonatal platelet counts after the first delivery were normal and no maternal antibody was detected for both groups. After the third delivery, GPIbα+/− pups did not have any bleeding disorders and the maternal antibody was also negative. In contrast, 14.3% of β3+/− pups were stillborn or had subcutaneous bleeding and maternal antibody was detected. In Group II, 8 pups were delivered from a GPIbα−/− female and 2 of them have minor intestinal and subcutaneous bleeding. 10 pups were delivered from 2 β3−/− female mice and 6 of them were stillborn. Of these dead pups, there was 1 case of intracranial hemorrhage and 2 immature pups. When the anti-GPIbα antibody level was increased via 4 platelet transfusions, a GPIbα−/− female had a miscarriage and severe FNAITP was observed. Our preliminary data suggested that fewer FNAITP cases reported with alloantibodies against GPIbα might not be due to the severity of FNAITP caused by this antibody, but is more likely because GPIbα is less immunogenic with a limited repertoire of MHC in pregnant women.

Neonatal alloimmune thrombocytopenia in the Irish population: a discrepancy between observed and expected cases.

To estimate the rate of detection of neonatal alloimmune thrombocytopenia (NAITP) in the Irish population, to investigate clinical presentation and outcome in affected infants, and to determine the extent, if any, to which this condition is underdiagnosed at present. Cases were collected in a retrospective fashion from a review of platelet serology laboratory records from January 1992 to December 2000. Clinical data were obtained from hospital records. Testing for maternal antiplatelet antibody was by one or more of the following: the platelet suspension immunofluorescence test, a commercial antigen capture enzyme linked immunosorbent assay (GTI-PakPlus), and the monoclonal antibody immobilisation of platelet antigens assay. Platelet antigen typing was by the polymerase chain reaction technique with sequence specific primers. Twenty seven serologically verified cases of NAITP were identified in 18 families. Maternal antibody to human platelet antigen 1a accounted for 25 of the 27 confirmed cases. Twenty one of 26 infants were born with severe thrombocytopenia. Nineteen of 27 infants had bleeding manifestations at birth. Petechiae and bruising were most commonly observed (n = 17). There were no documented cases of intracranial haemorrhage in this group but systematic cranial ultrasound was not performed. Screening studies in predominantly white populations have estimated the incidence of NAITP to be between 1 in 1000 and 1 in 2000 live births. With 50 000 births each year in Ireland, these results give a clinical detection rate for NAITP of just 1 case in 16 500 live births, strongly suggesting that NAITP is currently underdiagnosed. Antenatal screening to detect women at risk of having babies with NAITP is now scientifically feasible and should be considered.

Neonatal alloimmune thrombocytopenia: Antenatal management

Objective: The optimal management of pregnancies at risk for neonatal alloimmune thrombocytopenia is debated. Proposed management includes the administration of intravenous immunoglobulin and serial determination of the fetal platelet count. The aims of our study were to determine the effectiveness and likely mechanism of action of intravenous immunoglobulin and to evaluate the safety of cordocentesis in cases of neonatal alloimmune thrombocytopenia. Study Design: Eighteen mother-infant pairs were studied. All were at risk for neonatal alloimmune thrombocytopenia on the basis of delivery of a previously affected infant and confirmation of specific maternal antiplatelet antibodies. The pertinent antigen was HPA-1a in 13 cases, HPA-3a in 2 cases, and undetermined in 3 cases. Serial cordocenteses were used to determine fetal platelet counts. If the platelet count was <50,000/μL before 37 weeks’ gestation, treatment was initiated with intravenous immunoglobulin administered to either the fetus (n = 2) or the mother (n = 8). In 3 cases fetal and maternal immunoglobulin G levels were determined before and after treatment. Results: Seven (39%) fetuses had adequate platelet counts, were not treated, and were delivered of infants with normal platelet counts. Eleven (61%) fetuses were thrombocytopenic. Eight thrombocytopenic infants were treated with maternally administered intravenous immunoglobulin. In 6 (75%) of 8 cases the fetal platelet count increased after administration of intravenous immunoglobulin, but 2 fetuses remained severely thrombocytopenic. Two thrombocytopenic fetuses were treated with intravenous immunoglobulin infusion directly into the umbilical vein; both remained thrombocytopenic. Moreover, fetal immunoglobulin G levels did not correlate well with the response to intravenous immunoglobulin. Two (5.3%) of 38 cordocenteses were complicated by hemorrhagic complications, necessitating immediate cesarean delivery despite the use of prophylactic platelet transfusion in one case. Conclusion: Severe fetal alloimmune thrombocytopenia does not always occur in subsequent fetuses. Thus either fetal antigen status or platelet counts or both of these are necessary to determine whether treatment is needed. The effect of intravenous immunoglobulin on raising the fetal platelet count is inconsistent and appears to be caused by maternal or placental factors rather than a direct inhibition of fetal platelet destruction by immunoglobulin. The risk of hemorrhagic complications from cordocentesis in pregnancies complicated by neonatal alloimmune thrombocytopenia is higher than generally appreciated and is not always avoided by platelet transfusion at the time of the procedure. (Am J Obstet Gynecol 2000;182:1233-8.)

Maternal antiplatelet antibodies in predicting risk of neonatal thrombocytopenia

Objective: To determine the incidence of maternal antiplatelet antibodies in cases of thrombocytopenia during pregnancy, using the monoclonal antibody–specific immobilization of platelet antigens assay; and to assess the usefulness of this assay for predicting risk of neonatal thrombocytopenia. Methods: A total of 6770 pregnant women were included in the study, and the monoclonal antibody–specific immobilization of platelet antigens assay was done when platelet counts were less than 150 × 10 9/L. Platelet counts were determined in 6103 newborns. Results: The incidence of maternal thrombocytopenia was 11.6% (95% confidence interval [CI] 10.8, 12.4). Among newborns, 1.3% (95% CI 0.5, 2.7) born to thrombocytopenic mothers were thrombocytopenic, compared with 0.4% (95% CI 0.2, 0.6) born to nonthrombocytopenic women. Antiplatelet antibodies were detected in 37 (8.6%) of 430 thrombocytopenic women; autoantibodies were detected in 28 cases (circulating or bound to platelets), alloantibodies in eight cases, and an association of alloantibodies and autoantibodies in one case. The positive and negative likelihood ratios for predicting neonatal thrombocytopenia were 4.6 and 0.7, respectively. Conclusion: The monoclonal antibody–specific immobilization of platelet antigens assay did not predict the risk of neonatal thrombocytopenia in an unselected population of thrombocytopenic pregnant women.

Maternal Antiplatelet Antibodies in Predicting Risk of Neonatal Thrombocytopenia

Maternal Antiplatelet Antibodies in Predicting Risk of Neonatal Thrombocytopenia

Neonatal thrombocytopenia: incidence and characterization of maternal antiplatelet antibodies by MAIPA assay.

Neonatal thrombocytopenia (NNT) which is frequent in distressed newborns was uncommon in a non-selected population of neonates. The aim of this prospective study was to determine the frequency of NNT and, in confirmed NNT, to search for maternal antiplatelet antibodies with a monoclonal antibody-specific immobilization of platelet antigens (MAIPA) assay. Among the 8388 newborns studied, 40 (0.5%, 95% CI 0.3-0.6) had confirmed NNT, which was severe (platelet count < 50 x 10[9]/l) in 10 cases (0.12%, 95% CI 0.05-0.20). Antiplatelet antibodies were detected in 10/31 studied mothers of thrombocytopenic newborns (32.3%): they were alloantibodies in five cases and autoantibodies in five other cases. Among these 10 newborns, seven had severe thrombocytopenia and four had bleeding complications. As controls, antiplatelet antibodies were also searched for in mothers of non-thrombocytopenic newborns: antiplatelet antibodies were present in 8.5% (95% CI 5.9-11.7) of thrombocytopenic mothers (n = 400) and 3.2% (95% CI 0.7-9.0) of non-thrombocytopenic mothers (n = 95). The difference was significant between the control groups and the group of mothers of thrombocytopenic newborns. In conclusion, our data indicate that an immune origin is frequent in NNT and should be looked for, particularly when the platelet count is < 50 x 10(9)/l.

In vitro culture of colony forming unit-megakaryocyte (CFU-MK) in fetal alloimmune thrombocytopenia.

Perinatal alloimmune thrombocytopenia (PAITP) causes intracranial haemorrhage in the fetus and neonate. However, the severity of the thrombocytopenia correlates poorly with maternal anti-platelet antibody titres. To test the hypothesis that reduced platelet production contributes to fetal thrombocytopenia in PAITP, maternal sera from three HPA-1a-negative mothers whose pregnancies were complicated by anti-HPA-1a (two severe cases, one mild case) were added to colony forming unit-megakaryocyte (CFU-MK) cultures from HPA-1a positive and negative individuals. Sera from the two severely affected pregnancies containing anti-HPA-1a caused 66-100% inhibition of HPA-1a-positive fetal and neonatal CFU-MK, whereas CFU-MK from two HPA-1a-negative mothers were not inhibited by the anti-HPA-1a-containing sera. Maternal serum from the case of mild PAITP caused only mild inhibition of HPA-1a-positive cord and adult CFU-MK and did not inhibit HPA-1a-positive fetal CFU-MK. Taken together, these findings suggest that reduced megakaryocyte production contributes to fetal thrombocytopenia due to maternal anti-HPA-1a antibodies and also that the degree of CFU-MK inhibition correlates with severity of fetal thrombocytopenia.

Autoimmune Thrombocytopenic Purpura in Pregnancy: Maternal Risk Factors Predictive of Neonatal Thrombocytopenia

Pregnancy in ATP women is not unusual. The problem of this association concerns the possibility of disease transmission to the fetus due to the crossing of maternal antiplatelet antibodies through the placenta. Maternal risk factors predictive of neonatal thrombocytopenia, can be identified as follows: severe thrombocytopenia, previous splenectomy, high titre of PA-IgG and/or SPB-IgG. In 63 pregnancies in ATP patients, we have evaluated whether the above maternal risk factors, considered in the third trimester, can provide useful criteria for the prediction of neonatal thrombocytopenia. In the third trimester, the distribution of maternal risk factors was as follows: 0 in 7 cases, 1 in 27 cases, 2 in 15 cases, 3 in 12 cases, 4 in 2 cases. From a statistical evaluation, the neonatal platelet values and the maternal risk factors seem inversely correlated (r -0.437; p = 0.0005). In particular, neonatal and maternal platelet count correlated positively (r = 0.249; p = 0.025); moreover, neonatal platelet count correlated negatively with Splenectomy (r = -0.209; p = 0.05), PA-IgG (r = -0.401; p < 0.0005) and SPB-IgG (r = -0.338; p < 0.005). We tried to apply a multiple regression model for all the above parameters which appears statistically significant (p = 0.001); the variability was about 30%. This regression model could be validated if applied to a larger number of cases, and it could represent an alternative to the invasive methods used for the diagnosis of neonatal thrombocytopenia.

Favorable Neurological Outcome in 7 Cases of Perinatal Intracranial Hemorrhage Due to Immune Thrombocytopenia

Neonatal intracranial hemorrhage secondary to immune thrombocytopenia has been uniformly associated with neurological sequelae in survivors. These sequelae are seizures, hydrocephalus, mental retardation, and developmental delay. We report 7 survivors of intracranial hemorrhage who were prospectively evaluated regarding their long-term outcome at a mean of approximately 5 years of age. Five children were completely normal. One was delayed in speech, and one had a ventriculoperitoneal (VP) shunt in place and a residual hemiparesis. Four children had had seizures including the two with sequelae (speech delay and hemiparesis); only the patient with the VP shunt was still taking anticonvulsant medication. This latter patient was also the only one who required special education classes in which she was maintaining her grade level. In summary, a good long-term outcome can be expected in at least some patients with neonatal intracranial hemorrhage in cases of severe neonatal thrombocytopenia caused by maternal antiplatelet antibodies. This good outcome may be a result of, and should encourage, early diagnosis and vigorous supportive care in the neonatal intensive care unit.

Neonatal alloimmune thrombocytopenic purpura and congenital porencephaly associated with a new maternal antiplatelet antibody [letter

Neonatal alloimmune thrombocytopenic purpura and congenital porencephaly associated with a new maternal antiplatelet antibody [letter

NEONATAL ALLOIMMUNE THROMBOCYTOPENIC PURPURA AND CONGENITAL PORENCEPHALY ASSOCIATED WITH A NEW MATERNAL ANTIPLATELET ANTIBODY

NEONATAL ALLOIMMUNE THROMBOCYTOPENIC PURPURA AND CONGENITAL PORENCEPHALY ASSOCIATED WITH A NEW MATERNAL ANTIPLATELET ANTIBODY

Neonatal Alloimmune Thrombocytopenic Purpura and Congenital Porencephaly in Two Siblings Associated With a “New” Maternal Antiplatelet Antibody

We report a brother and sister, both of whom have porencephaly, hydrocephalus, optic atrophy, severe mental retardation, and spastic quadriplegia. In the younger child, abnormal intracranial structure was demonstrated by sonography at 32 weeks' gestation and was suspected earlier. Both children had transient severe thrombocytopenia as newborns. The mother is healthy and has never had purpura or other bleeding symptoms. However, her serum was found to react strongly with platelets from the father and from both children. The antibody in the mother's serum is platelet-specific but does not appear to be directed against any of the known antigens associated with neonatal alloimmune thrombocytopenic purpura (NATP) in other families, including PlA1, PlE2, or Baka. Although the mother's serum reacts with platelets from all 47 unrelated normal donors tested and from both the mothers and the fathers of 17 other children with suspected NATP, it does not react with her own platelets or with platelets from a patient with Glanzmann's thrombasthenia. These observations suggest that the serum from this woman identifies a previously undescribed high-frequency platelet-specific alloantigen and that sensitization to this determinant caused severe immune thrombocytopenia in both her children. It is likely that this led to intracranial hemorrhage in utero in these cases.