Vaginal delivery for fetuses at risk of alloimmune thrombocytopenia?


Dr D Oepkes, Department of Obstetrics, Leiden University Medical Centre, K6-31, PO Box 9600, 2300 RC Leiden, the Netherlands. Email


Objectives  To evaluate the safety of vaginal delivery in pregnancies with fetal and neonatal alloimmune thrombocytopenia (FNAIT).

Design  Prospective data collection.

Setting  Leiden University Medical Centre, the national centre for management of severe red cell and platelet alloimmunisation.

Population  Thirty-two pregnancies with FNAIT, with a sibling with thrombocytopenia but without an intracranial haemorrhage (ICH).

Methods  The mode of delivery, platelet count in cord blood and neonatal outcome were analysed. All women received weekly intravenous immunoglobulin from 32 to 38 weeks of gestation. Head ultrasound scan was performed in all neonates.

Main outcome measures  Signs of ICH or other bleeding in the neonates.

Results  Twenty-three women delivered vaginally. Nine caesarean sections were performed, all for obstetric reasons. Median platelet count at birth was 142 × 109/l (range, 4–252 × 109/l), with severe thrombocytopenia (<50 ×109/l) in four neonates, of which three were born vaginally. None of the neonates showed signs of ICH or other bleeding.

Conclusions  In pregnancies with FNAIT and a thrombocytopenic sibling without ICH, vaginal delivery was not associated with neonatal intracranial bleeding. These initial results support our noninvasive management of these pregnancies with FNAIT.


Fetal and neonatal alloimmune thrombocytopenia (FNAIT) is a rare but potentially devastating disease. It is the most common cause of thrombocytopenia in term neonates. In contrast to haemolytic disease due to rhesus D antibodies, the index pregnancy is often affected. Consequently, treatment is only possible in the next pregnancy. Many controversies still exist about the management, including mode of delivery.

The recurrence rate of severe thrombocytopenia and the risk to develop an intracranial haemorrhage (ICH) in the absence of ICH in a previous child is not known. Radder et al.1 estimated the ICH risk to be 7% in a subsequent untreated pregnancy after a previous child with thrombocytopenia but without ICH.

Moreover, in 80% of neonates with ICH, this bleeding would occur before labour,2 implying that, during labour or postpartum, the chance to develop ICH is approximately 1.4% in this group. In these milder FNAIT cases, the estimated 2.8% risk of complications associated with fetal blood sampling (FBS) and the 1.6% risk of fetal loss per procedure may not be justified when balanced against the risk of ICH.1 Caesarean section is often routinely employed for delivery in these cases. Practice guidelines advise vaginal delivery as an option in case of a platelet count > 50 × 109/l established by FBS, with or without an intrauterine platelet transfusion.3–5

However, there is no evidence that a vaginal delivery poses the fetus with a platelet count < 50 × 109/l at higher risk for ICH than caesarean section. We report our experience with the safety of vaginal delivery in FNAIT pregnancies without ICH in a previous child.


The Department of Obstetrics at the Leiden University Medical Centre is the national referral centre for pregnancies complicated by FNAIT in the Netherlands.

Our less invasive treatment strategy in women who are at risk for FNAIT and who have an index child with severe thrombocytopenia (<50 × 109/l) but without an ICH has been described previously.6 We prospectively collected all data from pregnancies complicated by FNAIT referred to us between March 1989 and August 2004. For this study, we selected all women with an index child with thrombocytopenia due to FNAIT but without an ICH. All women received weekly 1 g/kg bodyweight of intravenous immunoglobulins (IVIG) from 32 to 38 weeks followed by induction of labour. Head ultrasound scan was performed in all fetuses before starting the IVIG treatment and after birth. No diagnostic FBSs were performed. Assisted vaginal delivery was considered contraindicated. Caesarean section was only performed for obstetric reasons. The mode of delivery, the platelet count in cord blood and neonatal outcome were analysed.


Between March 1989 and August 2004, 29 women with 32 pregnancies met the inclusion criteria. The characteristics of the study group are given in Table 1. Twenty-three neonates from the 29 untreated siblings had a platelet count < 50 × 109/l. Four neonates were delivered by assisted vaginal delivery, having a platelet count of 11, 14, 21 and 30 × 109/l, respectively.

Table 1.  Characteristics of the study group
  • HPA-1a, human platelet antigen-1a.

  • *

    Delivered at 36 weeks.

Mothers (n= 29)
Median age, years (range)33 (29–40)
Children involved, n32
Previous pregnancy (n= 29)
Median platelet count (range)30 × 109/l (4–134)
Platelet count < 50 × 109/l, n (%)23/29 (79)
Delivery mode; caesarean section, n (%)4/29 (14)
Delivery mode; instrumental delivery, n (%)4/29 (14)
Current pregnancy (n= 32)
HPA-1a, n28
Mean IVIG (weeks), n (range)5 (3–7)
Preterm, n1*

In Table 2, the characteristics of the neonates are given. Twenty-three deliveries were by vaginal route. Nine caesarean sections were performed, four because of breech presentation, one for transverse presentation and two because of an earlier caesarean section. Two secondary caesarean sections were performed for failure to progress, one breech and one vertex presentation.

Table 2.  Data on delivery, neonatal outcome and treatment
 Spontaneous vaginal delivery (n= 23)Caesarean section (n= 9)
  1. There were no significant differences between the two groups for all parameters.

ICH, n00
Median platelet count (range)146 × 109/l (12–252)144 × 109/l (4–231)
Platelet count < 50 × 109/l, n31
Neonatal treatment63
 IVIG only, n30
 Platelet transfusion only, n03
 Combination of IVIG and donor platelet transfusion, n30

Median platelet count in cord blood at birth was 145 × 109/l (range, 4–252 × 109/l). Median platelet count in the caesarean section group was 144 × 109/l (range, 4–231 × 109/l), with severe thrombocytopenia (4 × 109/l) in one neonate. Median platelet count in the vaginal group was 146 × 109/l (range, 12–252 × 109/l), with three neonates being severely thrombocytopenic (12, 40 and 41 × 109/l, respectively). Nine neonates needed treatment for thrombocytopenia; three received platelet transfusions combined with IVIG. Three others needed only platelet transfusions and three neonates received only IVIG after birth. None of the neonates had signs of ICH at ultrasound examination. In Figure 1, the relationship between the platelet count at birth in the index group and the platelet count at birth in the treated group is shown.

Figure 1.

Relationship between the platelet count at birth in the index group and the platelet count in the treated group.

Discussion and conclusion

In our study group of pregnancies complicated by FNAIT, we achieved a 72% vaginal delivery rate. None of the 32 neonates developed an ICH, although in four neonates, the platelet count was less than 50 × 109/l. Three of these four children were safely born vaginally. The relatively high number of malpresentations could be explained by contraindication to external cephalic version in this group.

For the mother, the benefits of a vaginal delivery against a caesarean section are obvious. Although a caesarean section is safer then ever before, risks are not negligible. Maternal mortality, even with a planned section, is three times higher than with a vaginal birth7 and maternal morbidity—thrombosis, hysterectomy, infections, extended hospital stay and chance of rehospitalisation—is also higher.8,9 Uterine scar is also associated with increased risks in future pregnancies.10,11

Predelivery FBS is not without risk, especially for fetuses with thrombocytopenia. Mortality associated with FBS performed in a low-risk population is estimated to be 1.5% per procedure.12,13 Based on a review in the literature, the complication rate of FBS in FNAIT pregnancies was calculated as 1.6% fetal loss and 2.8% other complications.1

In the group of FNAIT pregnancies with a sibling without ICH, not treated with IVIG, the risk to develop ICH is estimated to be 7%.1 IVIG strongly reduces ICH in pregnancies with a sibling with ICH, but not completely, and at least four cases of recurrent ICH have been reported despite maternal IVIG treatment.14–17 There is only one case report of an ICH after maternal IVIG treatment in a pregnancy in which the previous child had a platelet count of 8 × 109/l but no ICH.18 We can therefore assume that the risk to develop an ICH after optimal IVIG treatment in the group without a previous ICH is low, even in fetuses or neonates without an adequate response to IVIG.

Although our study group of 32 pregnancies is relatively large for such a rare disease, the number of patients is still too low to sustain our results statistically. In the vaginal delivery group of 23 children, only three children had severe thrombocytopenia (<50 × 109/l). The fact that none of these three children had an ICH may be due to chance. Nevertheless, our experience supports that the mode of delivery in FNAIT without a history of ICH should be further investigated.

To achieve this, an international registry ( has been developed, in which all FNAIT cases from participating centres will be collected both retrospectively and prospectively. We expect to find more conclusive evidence through this data collection to optimise management of pregnancies with FNAIT.