Diagnosis and Management of Fetal and Neonatal Alloimmune Thrombocytopenia: An Update 2025

Authors: Ulrich J. Sachs, Ivonne Bedei, Sandra Wienzek-Lischka, Nina Cooper, Harald Ehrhardt, Roland Axt-Fliedner, Gregor Bein

Abstract

Background

Antibodies of the mother, which are directed against paternal antigens on platelets of the child, can lead to the destruction of the fetal blood cells in the circulation after diaplacental passage. The clinical picture of fetal-neonatal alloimmune thrombocytopenia (FNAIT) is characterized by bleeding, of which intracranial bleeding is particularly feared. Our understanding of the pathophysiology of FNAIT and its targeted prophylaxis and therapy has improved significantly in recent years.

Summary

FNAIT by anti-HPA-1a is the best studied. How exactly the mother is immunized is not known for certain, but, in clinically apparent cases, immunization usually occurs in the first pregnancy of an HLA-DRB3*01:01-positive, HPA-1a-negative woman. There is no convincing basis for assigning immunization against HPA-5b and against HLA class I any significance in the development of fetal thrombocytopenia. Newborns of mothers with anti-HPA-1a present a broad clinical picture ranging from isolated, clinically unremarkable thrombocytopenia to intracranial hemorrhage (ICH; in approx. 1–10% of cases). ICH usually occurs intrauterine (before week 28). There are indications that, in addition to the fetal platelets, the placenta can also be affected by anti-HPA-1a. As there are no screening programmes, the index diagnosis of FNAIT is random. It is made by serological and genetic laboratory tests. Predicting outcome in a subsequent pregnancy is problematic if the child is antigen-positive.

Key Messages

With a first-born child with severe thrombocytopenia, the probability of a recurrence of severe thrombocytopenia is around 70%. Without ICH, the probability of ICH in the subsequent pregnancy is low but not zero, and with ICH the recurrence rate is high. There is no established laboratory diagnostic method to predict the severity of thrombocytopenia or the occurrence of ICH. Prophylaxis with immunoglobulins (IVIgs) is considered effective. Pharmaceutics that block placental transport are currently undergoing clinical trials and may replace IVIgs in the future. Intrauterine platelet transfusions should no longer be performed. For the mature, thrombocytopenic newborn without internal hemorrhage, a platelet transfusion is advisable for platelet counts <25 g/L.

Introduction

Fetal and neonatal alloimmune thrombocytopenia (FNAIT) is caused by maternal antibodies to a human platelet antigen (HPA) on fetal platelets that is inherited from the father. The pathogenesis of FNAIT is similar to that of hemolytic disease of the fetus and newborn (HDFN). In contrast to HDFN, immunization against platelet antigens with subsequent fetal thrombocytopenia usually occurs during the first pregnancy in clinically apparent cases [1]. In women of European ancestry, antibodies to HPA-1a are most commonly detected. Upon transfer into the fetal circulation, the antibodies bind to their target antigen and can lead to subsequent destruction of fetal platelets, suppression of platelet production, and risk of hemorrhage. Fetal intracranial hemorrhage (ICH) is feared because it can lead to intrauterine fetal death or lifelong disabilities. ICH occurs in around 10–25% of cases with severe thrombocytopenia (platelet count <50 g/L) depending on study type [2]. Early detection and treatment of FNAIT can prevent serious complications in the affected fetus or newborn, as well as in a subsequent pregnancy.

Due to the lack of screening programs, FNAIT is usually diagnosed in an otherwise healthy newborn with clinical signs of bleeding, such as petechiae and hematomas, or with evidence of ICH on prenatal ultrasound. Neonatal thrombocytopenia may also be discovered incidentally when blood counts are obtained for other reasons.

Incidence of Maternal Alloimmunization and Incidence of Severe Disease Courses

In a meta-analysis of prospective screening and intervention studies in pregnant women, the incidence of alloimmunization to the HPA-1a antigen was estimated to be 200 per 100,000 (1:500) pregnancies [3]. A meta-analysis of prospective studies on the incidence of thrombocytopenia in unselected newborns (n = 59,425) showed an incidence of severe thrombocytopenia (platelet count <50 g/L) of 150 per 100,000 newborns (1:666) [2]. FNAIT was detected in 27% of these newborns (approximately 40 per 100,000 newborns, 2,500), making it the most common cause of severe neonatal thrombocytopenia. The incidence of FNAIT with ICH was estimated in this meta-analysis to be 10 per 100,000 (1:10,000) newborns [2]. In a prospective, blinded observational study with retrospective unblinding, 81 HPA-1a incompatible fetuses/newborns of anti-HPA-1a-immunized pregnant women were identified at a screening appointment at 27 weeks of gestational age. One fetus (1.2%; 95% CI: 0–6.7%) was diagnosed with ICH. The authors of this study estimate the incidence of ICH due to FNAIT to be 11 (95% CI: 0–32) per 10,000 HPA-1a-negative pregnancies (<1 per 100,000 unselected pregnancies) [4]. This means that approximately 1% of anti-HPA-1a-immunized pregnant women, diagnosed in a screening program, will experience a serious fetal/neonatal bleeding complication.

Without systematic screening as part of antenatal care or a regular blood count for the newborn, many cases of FNAIT will go undetected [5]. For reasons that are as yet unexplained, clinically apparent cases of FNAIT are more frequently diagnosed in male fetuses/newborns [6, 7].

Human Platelet Antigens

Most blood group antigens on platelets are characterized by a point mutation resulting in a single amino acid exchange on a platelet glycoprotein (biallelic polymorphism). In the typical case of immunization against the HPA-1a antigen, the pregnant woman is homozygous for the rare allele (HPA-1bb, phenotype frequency approximately 2%). She is immunized against the common antigen, which the fetus inherits from its father. The fetus is always heterozygous (HPA-1ab); the father may be homozygous or heterozygous for the immunogenic allele (HPA-1aa or HPA-1ab). Currently, more than 36 HPAs are known [8]. In the HPA-1, -2, -3, -4, -5, and -15 systems, the common antigen is designated with the letter a, while the rare antigen is designated with the letter b (e.g., HPA-1a, HPA-1b). The other systems (HPA-6 to HPA-14 and HPA-16 to HPA-35) are characterized by rare antigens, which are designated by the letter b (e.g., HPA-9b). These systems are often referred to as low-frequency HPA. Antibodies against the common protein chain (e.g., HPA-9a) have not yet been described for these systems. The epitopes of the platelet antigens HPA-1 to HPA-35 are localized on the glycoprotein complexes GPIIb/IIIa (fibrinogen receptor), Ib/IX (von Willebrand factor receptor), Ia/IIa (collagen receptor) and on CD109 (TGF-β receptor). The glycoprotein complexes IIb/IIIa, Ib/IX and Ia/IIa consist of 2 or more proteins that form a complex in a fixed ratio. The fibrinogen receptor GPIIb/IIIa consists of the GPIIb chain, also known as αIIb, and the GPIIIa chain, also known as β3. Most described HPAs, including HPA-1, are localized to the β3 integrin chain. A specific antigen is the CD36 protein, which some individuals do not express. Immunization against this protein can occur as isoimmunization, as is seen in pregnant women immunized against Rhesus D (Table 1).

In people of European ancestry, FNAIT is associated with anti-HPA-1a antibodies in approximately 75–80% of cases [9–11]. In suspected cases of FNAIT, anti-HPA-5b antibodies are found in approximately 10–15% of all cases in which HPA antibodies have been diagnosed [9–11]. However, if the denominator is not all cases with HPA antibodies, but rather all suspected cases, then the prevalence of anti-HPA-5b antibodies is comparable to the prevalence of anti-HPA-5b antibodies in unselected pregnant women (review in [12]). It is therefore controversial whether HPA-5b antibodies are the cause of clinically apparent FNAIT [12]. In people of Japanese ancestry, clinically evident FNAIT is most commonly caused by antibodies to HPA-4b [13]. Isoantibodies to CD36 (GP IV, previously Nak^a^) are predominantly found in people of Oriental, Asian, and African ancestry with type I CD36 deficiency. Type I CD36 deficiency is characterized by absence of CD36 on both platelets and monocytes, while type II CD36 deficiency involves the absence of CD36 on platelets but its presence on monocytes [14, 15]. CD36 is expressed on red blood cells, and maternal antibodies to CD36 can also cause fetal anemia in affected fetuses [16, 17].

Immunization of Pregnant Women

Immunization of pregnant women occurs in 40–60% of cases during the first pregnancy when evaluating case series in which the newborns have clinical signs of FNAIT [18, 19]. However, data from the largest prospective screening study indicated that in the majority of primigravida, anti-HPA-1a antibodies were detected only postpartum and not during pregnancy [20, 21]. This finding is significant as it shows that in only a few immunized women, an early and strong immune response leads to fetal bleeding complications already in their first pregnancy. The majority (60–75%) of fetal ICHs also occurs during the first pregnancy, beginning at about 18 weeks of gestational age [7, 19]. The mechanism of immunization is only partially understood. The receptors on which the platelet alloantigens are located are not expressed exclusively on platelets. Although expression of the fibrinogen receptor GPIIb/IIIa (αIIbβ3) is restricted to the megakaryocytic lineage [22], another member of this receptor family, the vitronectin receptor αvβ3, uses the same β3 chain. It is expressed on other cell lines, e.g., on syncytiotrophoblast cells as early as in the first trimester. It has been hypothesized that debris from syncytiotrophoblast cells could lead to immunization of pregnant women [23].

Genetic Risk Factor HLA-DRB3*01:01

Immunization against the HPA-1a antigen is genetically associated with the HLA class II allele HLA-DRB301:01 (review in [24]). The two antigens of the HPA-1 system are characterized by an amino acid exchange in the β3 integrin HPA-1a (Leu59) and HPA-1b (Pro59) [25]. A peptide of the HPA-1a alloform of the β3 integrin chain (PMCAWCSDEALPL^59^GSPR) is efficiently presented by the HLA class II molecule DRB301:01 to T helper cells, which in turn control the humoral immune response against HPA-1a [26, 27]. Pooled data from three prospective screening studies show that 27% of pregnant women with the HPA-1bb phenotype who carry the HLA-DRB301:01 allele produce antibodies to HPA-1a (positive predictive 0.269). In contrast, only 2% of HLA-DRB301:01-negative pregnant women are found to have such an antibody (negative predictive 0.983) [24]. In a retrospective study of 101 women with clinically apparent FNAIT due to anti-HPA-1a immunization, 99 women were HLA-DRB301:01-positive. No serious complications (ICH) occurred in the two HLA-DRB301:01-negative cases [28]. In a review of prospective studies evaluating an association between HLA-DRB301:01 and the neonatal outcome including 198 immunized women, all 64 newborns with severe thrombocytopenia (<50 g/L) and all three cases with ICH were observed in the HLA-DRB301:01-positive group (n = 180) and none in the negative group (n = 18) [29]. In conclusion, the absence of the HLA class II allele HLA-DRB301:01 is a significant protective factor. For prospective screening programs as part of antenatal care, it is therefore recommended to limit follow-up examinations during pregnancy to women who carry both the HPA-1bb phenotype and the HLA-DRB301:01 risk allele [24, 30] (Fig. 1).

IgG Transport

IgG antibodies are transported by transcytosis from the maternal circulation into the fetal capillaries of the placental villi. The maternal IgG antibodies represent the passive humoral immunity of the newborn. The neonatal Fc receptor (FcRn), which is expressed in syncytiotrophoblast cells and develops a high affinity for the Fc portion of IgG antibodies under acidic endosomal pH conditions, is essential in the active transport process (review in [31]). IgG transport into the fetal circulation begins as early as 13 weeks or earlier. After a steady increase of fetal IgG levels to 50% of maternal IgG concentrations at 28–32 weeks of gestation, there is a steep increase after 36 weeks of gestation, so that fetal IgG levels at the time of birth are approximately 20–30% higher than maternal IgG levels [32] (Fig. 2).

Association between Alloantibody Specificity and Clinical Presentation of FNAIT

Anti-HPA-1a Antibodies

In people of European ancestry, more than 80% of all antibodies detected in suspected cases of FNAIT are anti-HPA-1a antibodies. Larger epidemiologic studies of the incidence and severity of FNAIT therefore refer exclusively to cases with this antibody specificity. The occurrence of ICH has been associated with anti-HPA-1a antibodies that preferentially bind to the vitronectin receptor avβ3 on endothelial cells [33]. So far, this finding has not been replicated in an independent study.

Anti-HPA-5b Antibodies

Due to the high prevalence of anti-HPA-5b antibodies in unselected pregnant women, it is currently controversial whether maternal anti-HPA-5b antibodies are causally associated with the occurrence of fetal/neonatal hemorrhage (review in [12]). In this meta-analysis, the pooled prevalence of anti-HPA-5b antibodies in unselected pregnant women of European ancestry was 1.92% (n = 3,147), compared with 3.32% (n = 5,003) in women with suspected FNAIT. It is unclear whether this marginal difference actually exists or is just due to bias. A study of 500 female blood donors with a history of pregnancy found that 3.2% had anti-HPA-5b antibodies, a rate comparable to that in suspected FNAIT cases [34]. A single Japanese study in unselected pregnant women demonstrated weak evidence that a small proportion of women presenting with anti-HPA-5b antibodies will give birth to a newborn with mild thrombocytopenia [13]. A retrospective analysis of 817 families with suspected FNAIT found no association between the presence or absence of maternal anti-HPA-5b antibodies and neonatal platelet count [12]. Maternal anti-HPA-5b antibodies have also been reported in association with fetal ICH in several cases [11, 35]. Note, in these cases, the neonatal platelet count was normal or moderately decreased in both studies. Thus, there could be many other concurrent conditions that predisposed these newborns to ICH. Results from a retrospective cohort study of 105 cases of fetal/neonatal ICH from a Swedish registry suggest coincidence with anti-HPA-5b antibodies, which were detected in 2 women (1.9%) [36]. This is the rate that would be expected to occur by chance. In conclusion, there is no convincing evidence, suggesting that maternal anti-HPA-5b antibodies cause fetal or neonatal hemorrhage. A recent consensus panel concluded that a potential screening program for FNAIT should not include anti-HPA-5b screening [30].

Antibodies to Other HPA Antigens

Due to the rarity of antibodies against other HPA antigens, it is difficult to make reliable statements about the prognosis in the event of a subsequent pregnancy. Clinical management is therefore usually based on the recommendations for the treatment of FNAIT following immunization against HPA-1a.

Antibodies to HLA Class I Antigens

Maternal antibodies to HLA class I antigens have been associated with FNAIT in many case reports. However, postpartum, anti-HLA antibodies are detected in up to 50% of women, depending on the sensitivity of the detection method (review in [6]). In a retrospective cohort study of 817 suspected cases of FNAIT, maternal HLA class I antibody detection was not associated with any of the endpoints in the overall cohort (immunization against HPA-1a, neonatal platelet count) and none of the endpoints in the cohort with FNAIT due to immunization against HPA-1a (neonatal platelet count, birth weight, ICH) [6]. Colvin et al. [37] found no association between HLA class I antibody strength and nadir neonatal platelet count in a small study of 31 suspected FNAIT cases. In this study, FNAIT cases were found to have both higher HLA antibody strength and broader HLA antibody reactivity, compared to controls. There may have been a bias between cases and maternal blood samples in cases were taken after delivery and in controls between 24 and 28 weeks of gestation. Regan et al. [38] showed that anti-paternal cytotoxic HLA antibodies are related to the gestational age of the pregnancy and are rarely demonstrable before 28 weeks of gestation. In our opinion, there is currently no scientific evidence that HLA class I antibodies cause clinically apparent FNAIT [6]. An appropriate animal model could be used to test this hypothesis.

Antibodies to ABO Antigens

Very rarely, FNAIT may occur if the pregnant woman has very high anti-ABO antibody titers and the child also belongs to the 1–2% of all individuals with extremely high ABO expression density on platelet glycoproteins (review in [39]).

Clinical Signs

The clinical spectrum of the disease is broad, ranging from clinically asymptomatic (and therefore possibly unrecognized) thrombocytopenia to severe ICH, which can lead to significant perinatal mortality and severe neurological sequelae [7]. Petechiae and hematomas are the most common clinical signs of FNAIT observed in the majority of neonates. Other bleeding manifestations such as melena (30%), hemoptysis (8%), hematuria (3%), or retinal hemorrhage (7%) are less common [18]. ICH may occur in approximately 10% of children (see below). Fetal/neonatal ICH due to maternal anti-HPA-1a antibodies is usually associated with severe thrombocytopenia [11, 35]. However, thrombocytopenia per se does not cause bleeding – it is only when the endothelial lining is damaged that bleeding may occur. In a recent retrospective study, median platelet counts were lower in the cases with bleeding symptoms; severely thrombocytopenic infants were observed in the group with severe and minor bleeding but also in children without bleeding symptoms [35]. Very rare cases of newborns with ICH but without severe thrombocytopenia [11, 35] should be interpreted cautiously. There could be many other concurrent clinical conditions that predispose these newborns to ICH. It has been observed that the birth weight of male newborns with FNAIT is approximately 200–500 g less than that of healthy male newborns, a phenomenon not observed in girls [40].

In a retrospective study of FNAIT-affected mothers, 54% reported psychological problems >6 weeks, including anxiety, stress, depression, and sleep disturbances. The prevalence of autoimmune diseases associated with the HLA ancestral haplotype A1-B8-DR3 (including HLA-DRB3*01:01) was slightly increased, indicating the need for clinical care of affected mothers [41].

ICH

ICH usually occurs early in pregnancy, often before 28 weeks of gestation, while peripartum hemorrhage is rare [7]. Birth mode reportedly has no association with the occurrence of ICH [42]. In about 60 percent of the cases, the first-born child was affected by an ICH [7]. The perinatal mortality rate for children with FNAIT and ICH is as high as 30–50% [7]. The hemorrhage may be parenchymal, intra- or periventricular hemorrhage, or mixed or unclassified. Bleeding before 28 weeks of gestation is almost exclusively intra- or periventricular hemorrhage [7]. In the retrospective multicenter analysis by Tiller et al. [7], 23 of 43 (54%) of ICHs occurred before 28 weeks of gestation. In the absence of intrauterine fetal death, parenchymal hemorrhage typically results in large hemispheric porencephalic cysts, and ventriculomegaly is also common. In one retrospective observational study, 60% of surviving children had severe neurological developmental delay [43], including severe mental retardation, cortical blindness, seizures, and cerebral palsies such as paraplegia, diplegia, and spastic quadriplegia. For unknown reasons, ICH is more common in boys. Bleedings are also more often fatal in boys than in girls [7].

Children newly diagnosed with FNAIT are at an increased risk of developing long-term neurodevelopmental problems by a median age of 12 years, ranging from mild to severe impairment. This is true even for children without clinically diagnosed ICH but with anti-HPA-1a-positive mothers [44].

Placental Pathology

Several small studies of placental histology in FNAIT provide evidence of inflammatory processes. It is reasonable to assume that anti-HPA-1a antibodies also find their target structures on the syncytiotrophoblast, leading to inflammation through complement activation [45]. Dubruc et al. [46] found signs of inflammatory activity in the examination of 21 placentas compared to age-matched controls, especially signs of chronic chorioamnionitis (OR: 14, 95% CI: 1.7–113.8) and chronic basal villitis (OR: 17, 95% CI: 2–145.6). The study also revealed evidence of a possible link between placental inflammation and fetal complications, such as ICH. Recently, we have shown that high levels of the antiangiogenic placental factor soluble fms-like tyrosine kinase-1 in the serum of newborns with FNAIT are associated with low platelet counts [47]. A similar observation was made in HPA-1a-alloimmunized mothers. Increasing levels of soluble fms-like tyrosine kinase-1 protein were associated with lower neonatal platelet count [48]. Presumably based on placental pathology, maternal immunization against HPA-1a appears to be associated with hypertensive pregnancy complications, reduced birth weight and prematurity [4].

Diagnosis

The goal of FNAIT diagnostics is to rapidly and reliably identify maternal antibodies against platelets antigens inherited by the fetus from its father. Antibodies against rare paternal platelet antigens should be excluded so that fetal/neonatal hemorrhage prophylaxis can be administered in a future pregnancy, if necessary. Complete work up of a suspected case of FNAIT is performed in specialized laboratories, where the presence of alloantibodies to low-frequency or private antigens, otherwise undetectable, can also be excluded by a crossmatch procedure between maternal serum and paternal platelets [10, 49]. For this purpose, reference laboratories use glycoprotein-specific in-house assays based on the monoclonal antibody-specific immobilization of platelet antigens (MAIPA) test developed by Kiefel et al. [50]. This assay remains the gold standard in platelet serology, since it is robust against HLA class I antibodies, which are common in parous women. Screening and antibody identification in indirect MAIPA is performed with HPA-typed donor platelets or paternal platelets in which the glycoprotein complexes GPIIb/IIIa, GPIb/IX, GPIa/IIa, CD109 (optionally CD36) with a known antigen pattern are immobilized and incubated with maternal serum. The use of in vitro-generated megakaryocytes has also been suggested for the detection of platelet antigen-specific antibodies [51]. The common antibody specificities anti-HPA-1a and anti-HPA-5b can also be excluded with good sensitivity and specificity in an initial screening using commercial test systems such as the bead-based PakLx assay [52]. The advantage is that results are available quickly (within 3 h; MAIPA takes up to 8 h). However, the PakLx assay, based on Luminex technology, has insufficient sensitivity for anti-HPA-3 antibodies; anti-HPA-15 antibodies cannot be detected at all due to the assay design [53]. The detection of anti-HPA-15 antibodies by MAIPA should be based on platelets that express high levels of the HPA-15 antigen to avoid false negative findings [54]. For details on the detection of platelet antibodies, we refer to a recent review by Porcelijn et al. [55].

HPA antigens are determined by molecular genetic methods. Genotyping of mother, father and child for HPA-1, -2, -3, -4, -5, -6, -9 and -15 is recommended for diagnostic work up in cases of suspected FNAIT [56]. In parents of Asian descent, HPA-21 should also be considered due to the frequency distribution. Genotyping is primarily used to confirm the serologic results of antibody detection, but can guide initial neonatal treatment, identifying an HPA mismatch before serological testing is completed. In particular, some of the low-frequency antigens could disappear from the platelet surface due to long transport times of blood samples and escape detection in the crossmatch technique. Therefore, rare antigens should also be included in genotyping in order not to miss a possible immunization. The father is genotyped to determine whether he is a homozygous or heterozygous carrier of the implicated antigen. This information is used to counsel the family and to manage a subsequent pregnancy. Genotyping does not replace serologic detection of an alloantibody.

In clinical cases of suspected FNAIT, in approximately 5–10% of mothers with the HPA-1bb phenotype and a newborn with the HPA-1ab phenotype at birth, no antibody to HPA-1a can be detected by standard methods. In this constellation, a low-affinity antibody could be detected in 30–50% of cases using a highly sensitive method (surface plasmon resonance) that does not require washing steps [57, 58]. Therefore, if FNAIT is clinically suspected and serology is negative, it is recommended that a reference laboratory work up be performed 2–8 weeks after delivery, as HPA alloantibodies may be detected with a delay, presumably due to boosting during delivery and/or affinity maturation.

Risk Stratification

The detection of an anti-HPA antibody in a woman during or after pregnancy has very limited predictive value for the current or planned future pregnancy. This is true even if the child is or (in the case of a homozygous father) will be a confirmed carrier of the corresponding antigen.

In the Norwegian nationwide screening study, 45 anti-body-positive women were followed in at least one subsequent pregnancy [59]. However, there was only one FNAIT with ICH. In this study, neonatal thrombocytopenia in subsequent pregnancies was less severe (18%), unchanged (52%), or more severe (30%). If thrombocytopenia was severe in the first pregnancy, it was also severe in the subsequent pregnancy in 71% (10 of 14 cases); and although in 15 cases the child did not have FNAIT despite antibody detection in the first pregnancy, 5 of 15 children developed FNAIT in the subsequent pregnancy; three of them with severe thrombocytopenia. If the neonate had severe thrombocytopenia in the first pregnancy, severe thrombocytopenia would be expected in the majority of cases in the subsequent pregnancy.

In Norway, IVIg has traditionally not been used for pregnant platelet-immunized women [60]. In a recent retrospective study, the risk of ICH was investigated in patients without IVIg treatment in a subsequent pregnancy [61]. If no ICH occurred in the first pregnancy with FNAIT, there is a low risk of ICH in an untreated subsequent pregnancy. In 64 subsequent pregnancies of mothers with FNAIT due to immunization against HPA-1a (previous sibling without ICH), no case of ICH occurred (0.0%, 95% CI: 0.0–5.7) [61]. This risk is low, but not zero, similar to the ICH risk (0.6% [2/313]; 95% CI: 0.2–2.3) reported in IVIg-treated patients (previous sibling without ICH) in the systematic review by Winkelhorst et al. [62]. Knightly et al. [63] conducted a retrospective survey of members of the patient organization NAITbabies.org on the same question. The authors estimated the risk of ICH in a subsequent untreated pregnancy in mothers with FNAIT without ICH in the previous pregnancy to be 3.2% [63]. This study may a priori include more severe cases compared to the Norwegian screening study, as it is assumed that mainly mothers with severe FNAIT will join a patient organization.

Children with FNAIT and ICH are a subgroup of all children with FNAIT. A summary of anecdotal reports by Bussel and colleagues [64] on the outcome of subsequent pregnancies after ICH in the first pregnancy with FNAIT showed a rate of 11 out of 13 ICH cases (85%; 95% CI: 79–91%). A recent retrospective comparative study found that the ICH recurrence rate in untreated pregnancies was considerably lower (29% [2/7]; 95% CI: 8.2–64.1). This means that if the child had an ICH due to FNAIT in the previous pregnancy, there is a high ICH risk in untreated subsequent pregnancies.

The limited predictability of the consequences for a subsequent pregnancy demonstrates that the presence of an HPA alloantibody is not sufficient to predict the severity of thrombocytopenia. In addition, severe thrombocytopenia often does not lead to ICH, so better prediction of the degree of thrombocytopenia does not necessarily contribute to better prediction of the risk of ICH (review in [65]).

Three modulating factors may become important in the future for assessing fetal antibody glycosylation [66, 67], the ability of antibodies to react with endothelial cells and induce ICH [33], and antibody concentration [68]. For details, we refer to a recent review [65].

Unfortunately, the only diagnostic test currently available is the determination of antibody concentration. Mechanistically, it is conceivable that a high concentration of antibodies in maternal blood may result in more anti-HPA-1a antibodies entering the infant’s circulation and contributing to greater platelet degradation. Studies have provided conflicting results on this issue. Killie et al. [21] analyzed maternal sera collected during the Norwegian screening study and concluded that the positive predictive value of antibody concentration under specific analytical conditions was 54% (95% CI: 43–63%). This means that “severe FNAIT” cannot be predicted by determining the antibody concentration. Anti-HPA-1a antibody levels have been used in Norway as a stratification tool [60]. However, there were too few cases of ICH in all the studies to draw conclusions. Therefore, antibody concentration cannot be used to predict the risk of ICH. Another obstacle to the potential adoption of titers for clinical risk stratification is the lack of standardization and reproducibility of titers. Despite efforts to standardize anti-HPA-1a titration [69, 70], the coefficient of variation of titers between reference laboratories is very high (up to 144%) [71]. As a result, a recent consensus panel could not agree on an anti-HPA-1a antibody cutoff value for risk stratification [30].

For possible risk stratification, only the history of older siblings regarding platelet count at birth, clinical signs and severity of bleeding, and occurrence of ICH, and in this case, the time of onset of bleeding (before or after 28 weeks of gestation) can currently be used for clinical decisions in the management of subsequent pregnancies.

Management of Pregnancy with Known Risk

Because FNAIT is a rare and serious condition, early management of high-risk pregnancies should be provided by a multidisciplinary team in centers with appropriate expertise. First, a detailed history of previous pregnancies, the clinical manifestations of FNAIT and their timing should be taken. It is important to note if and when ICH occurred in the previous pregnancy. The previous immunohematologic findings should also be recorded. When a new pregnancy is diagnosed, the presence of the HPA alloantibody should be confirmed. If not done postpartum after the index pregnancy, maternal and paternal HPA genotyping is required. If the partner is a homozygous carrier of the implicated HPA allele, 100% of the offspring will also be allele carriers. If the partner is heterozygous for the implicated allele, there is a 50% chance that the offspring will not be a carrier, and in the latter case, there is no indication for monitoring the pregnancy (if there is no doubt about paternity). The homozygous or heterozygous status of the father may be determined before the subsequent pregnancy and can already be included in a preconception consultation. However, when the presumed father is not the biological father, there are adverse medical and ethical implications. Therefore, it was recommended that noninvasive prenatal diagnosis be used exclusively to determine the fetal HPA allele in question [72] (Fig. 3).

Noninvasive Prenatal Diagnosis of Fetal HPA Alleles

If the father of the child is heterozygous, unknown, or unavailable for genotyping, fetal HPA genotyping is performed to determine the fetal risk in the current pregnancy, which can be performed noninvasively from 12 weeks of gestation using maternal cell-free plasma DNA [73–76]. With the discovery of cell-free fetal DNA in maternal plasma, the invasive procedures of chorionic villus sampling or amniocentesis to obtain DNA for fetal HPA typing have been largely replaced, as amniocentesis and especially chorionic villus sampling increase the risk of maternal alloimmunization [77]. Next-generation sequencing is used not only for aneuploidy screening and noninvasive expanded prenatal genetic diagnosis but also for the determination of fetal HPA alleles [74]. Next-generation sequencing offers the possibility to determine the fetal DNA fraction in a maternal plasma sample by sequencing many gene fragments in parallel. A valid noninvasive fetal blood group diagnosis is possible if the fetal fraction is above a predetermined threshold (e.g., 4% fetal fraction) [78].

Prenatal Care and Ultrasound Monitoring of Pregnancy

There are currently no recommendations for ultrasound monitoring of a pregnancy at risk for FNAIT. While Doppler ultrasound can be used for HDFN, as there is a direct correlation between fetal hematocrit and blood flow velocity in fetal vessels, there is no noninvasive method to estimate fetal platelet count. Percutaneous umbilical cord puncture for fetal blood platelet count is now only used in exceptional cases due to the risks of the procedure involved [79]. A possible algorithm could be that after an early malformation ultrasound in the first trimester and a detailed exclusion of malformations in week 20–22 of pregnancy, further ultrasound examinations are performed every 4 weeks and in high-risk pregnancies every 2 weeks. Ultimately, however, only a hemorrhage can be detected and a decision must be made as to whether it is justifiable and desirable to continue the current pregnancy.

According to data from a Swedish registry study [36], severe thrombocytopenia (platelet count <50 g/L) is rarely the cause of incidental ICH (8.5% of cases), and antiplatelet antibodies were found in only 2.9% of cases. This means that FNAIT is a rare cause of ICH in fetuses of unselected pregnant women. The main causes are genetic and congenital infections, including cytomegalovirus infection, ischemic infarcts, and trauma. In a French registry study of 194 consecutive ICH cases, pathogenic variants in the COL4A1/COL4A2 genes were detected in 19% of fetuses (de novo in 70% of cases). Pathogenic variants in two megakaryopoiesis genes (MPL and MECOM) were detected in two families [80]. Whole exome screening in cases with negative COL4A1/COL4A2 testing revealed additional pathogenic or likely pathogenic variants in other genes in 8% of fetuses [81]. It can be concluded that the main cause of fetal ICH is of genetic origin. Genetic testing and counseling should be provided to the family.

In addition to ultrasound, fetal magnetic resonance imaging (MRI) has gained considerable clinical importance in recent years. In prenatal diagnosis, fetal MRI can be used to complement ultrasound findings [82].

Fetal and Neonatal Bleeding Prophylaxis

Preventing ICH and other serious fetal bleeding is the goal of prenatal treatment. Secondary prophylaxis of fetal and neonatal hemorrhage in a patient with a history of FNAIT and ICH was first performed in 1984 by intrauterine transfusion of maternal platelets a few hours before the planned delivery [83]. Subsequently, serial intrauterine platelet transfusion (IUPT) after diagnostic fetal blood sampling (FBS) was introduced into the management of patients with FNAIT. FBS is associated with a high risk of bleeding when fetal platelet counts are low [84]. The cumulative risk of miscarriage with the IUPT treatment strategy has been reported to be 6% per pregnancy [85]. Noninvasive prophylaxis of fetal and neonatal bleeding complications with intravenous immunoglobulin (IVIg), with or without dexamethasone, was introduced by Bussel et al. [86].

A meta-analysis comparing both strategies – IUPT or IVIg – analyzed a total of 26 trials, including 4 with randomization to different treatment arms [62]. In the IVIg prophylaxis trials, the most commonly used dose was 1 g/kg body weight (bw) per week (range 0.4–2 g/kg bw). Treatment outcomes in most trials were similar with respect to the incidence of ICH. In contrast, invasive strategies with FBS and/or IUPT were associated with an overall complication rate of 11%; the most common complication was emergency cesarean section. In one-third of cases, complications of invasive procedures were associated with intrauterine fetal death. In patients who received IVIg alone, treatment was successful in 98.7% of ICH occurred in only 4 of 315 pregnancies treated with IVIg.

In several studies, corticosteroids were administered to pregnant women in addition to IVIg (review in [87]). An additional benefit cannot be inferred with certainty from the published data. There is international consensus not to use antenatal steroids in standard-risk pregnancies (previous pregnancy without ICH); there is no consensus on the use of steroids in high-risk pregnancies (previous pregnancy with ICH) [30]. For high-risk pregnancies (older sibling with ICH), the available data support initiation of treatment at 12–16 weeks of gestation, and for standard-risk pregnancies (no sibling with ICH), initiation of treatment at 20–24 weeks of gestation with a weekly IVIg dose of 1 g/kg bw [87]. There are currently insufficient data to recommend other dosages (e.g., 0.5 g/kg bw for standard-risk pregnancies or 2 g/kg bw for very high-risk pregnancies) [62].

Because the risk of fetal ICH in a subsequent pregnancy in a standard-risk pregnancy (no sibling with ICH) is very low (see above), the benefit of IVIg therapy in a standard-risk pregnancy has been questioned [88]: the risk of fetal ICH in a subsequent pregnancy is comparable in patients with and without IVIg treatment (see above). Therefore, the benefits and risks of IVIg therapy should be carefully weighed up with the patient before a subsequent pregnancy. The use of IVIg for fetal and neonatal bleeding prophylaxis in pregnant women with FNAIT is off-label. Patients must be informed about this off-label use.

The above statements on fetal and neonatal bleeding prophylaxis concern FNAIT cases caused by antibodies against HPA-1a. The question of whether suspected FNAIT cases with detection of anti-HPA-5b antibodies should be treated with IVIg in a subsequent pregnancy is controversial. A recent consensus panel did not vote for treatment in a standard-risk pregnancy (previous pregnancy without ICH). However, the panel voted for IVIg treatment in an anti-HPA-5b-positive high-risk pregnancy (previous pregnancy with ICH) [30]. In our view, the outcome of the vote reflects the ethical dilemma faced by physicians when counseling women with a previous history of fetal ICH and is not necessarily supported by epidemiological evidence (see above).

The most common adverse effects of IVIg infusions in pregnant women are headache, nausea/vomiting, and pruritus/dermatitis [89]; clinically relevant hemolysis and pancytopenia have been rarely reported [90]. Monitoring of hemoglobin levels in pregnant women with blood groups A, B, or AB is recommended to exclude relevant hemolysis [89]. The risk of neurodevelopmental disorders in children whose mothers were treated prenatally with IVIg for FNAIT is comparable to that of the general population [91]. In conclusion, the administration of IVIg during pregnancy is state of the art for the prophylaxis of ICH in the fetus and newborn in a high-risk pregnancy.

Birth Mode

Modes of delivery for pregnancies with FNAIT vary from center to center. In a study of 32 pregnant women with known standard risk (previous pregnancy without ICH) after prepartum IVIg treatment, in whom vaginal delivery was preferred (n = 23) and cesarean section was performed only for obstetric reasons (n = 10), there was no evidence of relevant differences in neonatal outcome between delivery modes; therefore, there is no objection to vaginal delivery in the standard-risk group [92]. In the high-risk group (previous pregnancy with ICH), elective cesarean section is preferred also for organizational reasons (provision of antigen-negative apheresis platelet units, optimal neonatal care, possible need for advanced neuroradiological imaging). An international consensus panel could not reach a consensus on the mode of delivery in either low-risk or high-risk pregnancy [30].

In addition to the patient’s risk constellation, the obstetric history must be considered in planning. There is consensus that placement of a cephalic electrode for monitoring during labor, fetal blood testing during labor, and forceps- or vacuum-assisted vaginal delivery should always be avoided. Routine cordocentesis and FBS to check fetal platelet count are also not recommended because of the increased complication rate [87].

The optimal time for delivery is also not known with certainty. In high-risk pregnancies (previous pregnancy with ICH), delivery could take place between 36 and 37 weeks of gestation by primary cesarean section [1, 60]. For standard-risk pregnancy (previous pregnancy without ICH), the planned delivery could take place between weeks 37 and 38 of pregnancy [1]. Delivery should always take place in a specialized center with a high level of obstetric and neonatal expertise. Immediate postnatal determination of the infant’s platelet count should be performed; HPA-matched platelets should be available at delivery [30].

Therapeutic Approach for a Newborn with FNAIT

Petechiae, hematomas, and spontaneous bleeding are typical clinical manifestations of neonatal hemorrhage due to thrombocytopenia. Common differential diagnoses such as early onset infection, perinatal asphyxia, or maternal causes are validated based on history and available clinical findings. In early-onset thrombocytopenia, FNAIT is the most important differential diagnosis, accounting for approximately 10% of isolated thrombocytopenias (incidence 1,000) and approximately 30% of severe thrombocytopenias (<50 g/L; incidence 2,500) [2]. Even in mild forms of thrombocytopenias (>50 g/L), FNAIT must be considered in the differential diagnosis, as maternal immunization in a subsequent pregnancy can lead to severe FNAIT and corresponding risks for the fetus or newborn. We refer to a recently published review on the neonatal differential diagnosis of thrombocytopenia [93]. FNAIT is the main differential diagnosis in otherwise healthy neonates with isolated thrombocytopenia.

In addition to cutaneous and mucocutaneous bleeding, ICH is the most common and severe bleeding complication of FNAIT. Extracranial bleeding is much less common and can involve various organ systems (gastrointestinal tract, lungs, others) [94]. Due to the high risk of lifelong neurological impairment, every newborn with suspected FNAIT should be examined for ICH as soon as possible, regardless of the current platelet count [87]. Although ultrasound with high-frequency transducers and sound windows in addition to the anterior fontanel can now detect hemorrhages in the typical locations with a high degree of certainty, MRI imaging of the CNS is superior to ultrasound, especially in recent (<6 h), extracerebral, and infratentorial hemorrhages, and should be considered at least in questionable cases [95]. If a diagnosis of ICH is clinically suspected in a thrombocytopenic newborn, platelets should be transfused before the results of the imaging tests are available [87].

Recommendations for transfusion of platelets in neonates are largely based on expert opinion in the absence of prospective studies. Although low platelet counts in preterm infants are associated with an increased risk of ICH, neither the severity of thrombocytopenia nor transfusion of platelets is associated with the incidence of ICH [96]. Recently, a prospective randomized study in preterm infants <34 weeks of gestation without FNAIT showed that higher thresholds (>50 g/L) are detrimental with respect to the occurrence of a bleeding event or death, regardless of the predicted bleeding risk of the infant [97, 98].

In a systematic review of postnatal treatment of FNAIT, 24 of 29 newborns with ICH were found to have a platelet count <30 g/L [99]. Current recommendations suggest neonatal platelet transfusion if the platelet count is <25 g/L in the case of a previous pregnancy without ICH and <50 g/L in the case of a previous pregnancy with ICH [30]. For acute life-threatening bleeding (e.g., ICH or gastrointestinal bleeding), the platelet count should be at least 100 g/L initially and at least 50 g/L for approximately 1 week [87].

Platelet units from HPA-1a-/HPA-5b-negative donors are compatible with maternal antibodies in over 90% of people of European ancestry. HPA-selected platelet units should be used when readily available. Data from a retrospective, multicenter cohort study suggest that HPA-compatible transfusions result in a higher median platelet count increase than HPA-incompatible transfusions. Whether the use of HPA-compatible platelet units is also associated with a lower risk of bleeding remains unclear [100].

If HPA-compatible units are not available, a sufficient increase in the platelet count out of the critical range can usually be achieved by transfusion of nonselected platelet units [99]. However, a rapid decrease in platelet count can be expected during follow-up. Because of the risk of ICH, the search for a suitable donor must not delay the initial transfusion; rather, platelet transfusion must be initiated as soon as there is a reasonable clinical suspicion of FNAIT.

The optimal transfusion volume (10 versus 20 mL/kg bw) is not known. Risk of circulatory overload should be considered [93]. It should be noted that the platelet count in untreated neonates with FNAIT often continues to fall in the days after birth, requiring close monitoring at 4–24 h intervals until the platelet count has stabilized above the threshold without (re)intervention [87]. Occasionally, thrombocytopenia may persist in newborns for 8–12 weeks [101]. The addition of IVIg or the administration of IVIg alone does not provide any therapeutic benefit, nor does the administration of corticosteroids [99]. In cases of persistent thrombocytopenia, alternative diagnoses such as inherited thrombocytopenia should be ruled out. If FNAIT is known prenatally, an HPA-compatible platelet unit should be available for delivery [30]. In this case, cord blood platelet count is recommended immediately after birth.

Screening

The screening of pregnant women for FNAIT and fetal bleeding prophylaxis as part of regular antenatal care has been discussed for over 30 years. Recently, de Vos et al. [102] proposed a screening program and cost-utility analysis. The authors suggested screening at 20 and 27 weeks of gestation and weekly prophylaxis with IVIg until delivery if the anti-HPA-1a antibody titer is >3 IU/mL. The titer threshold used is based on data from the Norwegian screening In this study, 55% of pregnant women with an anti-HPA-1a titer >3 IU/mL at 22 and/or 34 weeks of gestation delivered a newborn with a platelet count <50 g/L [21]. Fetal ICH occurred in only 1.2% (95% CI: 0–6.7%) of immunized pregnant women in the Dutch observational screening study [4]. Treatment of 55% of all anti-HPA-1a-immunized women to avoid ICH in 1.2% implies a high burden for many women who do not benefit from the treatment. In our view, international standardization of anti-HPA-1a antibody quantification would be a prerequisite for a future multicenter study on the association of antibody titer with fetal/neonatal ICH (see above). Introducing FNAIT screening into antenatal care is likely to require significantly better risk assessment than is currently possible to avoid overtreatment.

Future Perspectives

A phase II trial with the FcRn receptor blocker nipocalimab was successfully conducted for the prevention of early-onset severe HDFN [103]. We anticipate that upon completion of the ongoing phase III trials, FcRn blockade may replace IVIg prophylaxis in patients at risk for FNAIT in the near future. Although preclinical data indicated that immunoprophylaxis with anti-HPA-1a (comparable to prophylaxis with anti-D) could be feasible [104, 105], the phase 2 clinical trial was recently discontinued. It is unclear whether this approach can be pursued further.

Conclusion

Experienced multidisciplinary centers have a very high success rate in diagnosing and treating fetal and neonatal thrombocytopenia in the year 2025. There is an urgent need for better risk assessment in anti-HPA-1a-immunized pregnant women to implement FNAIT screening in regular antenatal care.

Conflict of Interest Statement

G.B. served on the Scientific Advisory Board of Janssen Pharmaceuticals; U.J.S. has received research funding from Johnson & Johnson and served on the Data Safety Monitoring Board for Rallybio. G.B. was a member of the journal’s Editorial Board at the time of submission.

Funding Sources

This study was not supported by any sponsor or funder.

Author Contributions

U.J.S. and G.B. were responsible for conceptualizing this review and harmonized and edited the final manuscript. All authors wrote one or more chapters and approved the final version of the review.