Delayed diagnosis of fetal and neonatal alloimmune thrombocytopenia: a cause of perinatal mortality and morbidity


Dr D Oepkes, Leiden University Medical Centre, Department of Obstetrics, K6-35, PO Box 9600, 2300 RC Leiden, the Netherlands. Email


Please cite this paper as: Madani K, Kamphuis M, Lopriore E, Porcelijn L, Oepkes D. Delayed diagnosis of fetal and neonatal alloimmune thrombocytopenia: a cause of perinatal mortality and morbidity. BJOG 2012;119:1612–1616.

Objective  To evaluate the rate and consequences of a late or missed diagnosis of fetal and neonatal alloimmune thrombocytopenia (FNAIT).

Design  Retrospective analysis of prospectively collected data of a national cohort.

Setting  National referral centre for fetal therapy in the Netherlands.

Population  Twenty-six women with pregnancies complicated by FNAIT and at least one previous pregnancy with a thrombocytopenic child.

Methods  Retrospective analysis of data from our electronic FNAIT database. In a consecutive cohort managed between July 2008 and July 2010, timing of first diagnosis of FNAIT was correlated to severity and outcome in the subsequent pregnancies.

Main outcome measures  Occurrence of delayed diagnosis of FNAIT, and possibly associated intracranial haemorrhage (ICH).

Results  In four of 26 pregnancies, timely diagnostic testing for FNAIT was not performed despite fetal or neonatal thrombocytopenia or ICH. Down syndrome, dysmaturity and birth trauma were perceived to be the cause of the thrombocytopenia/ICH. In two of these four subsequent, untreated pregnancies, severe fetal ICH occurred. The other 22 women were treated for FNAIT using intravenous immunoglobulin, all children are alive and well.

Conclusions  All neonates with thrombocytopenia at birth should be evaluated for FNAIT. Missing this diagnosis can have severe consequences for subsequent pregnancies.


Thrombocytopenia is a common clinical problem in neonates: 1–5% of newborns present with this problem.1 Fetal and neonatal alloimmune thrombocytopenia (FNAIT) is the most common cause of thrombocytopenia in otherwise healthy term infants.2 It is the result of maternal alloimmunisation to antigens on fetal platelets, inherited from the father. The maternal immunoglobulin G (IgG) antibodies cross the placenta and cause destruction of fetal/neonatal platelets, with resultant thrombocytopenia and associated risk of bleeding.

A potentially devastating condition, FNAIT may lead to intracranial haemorrhage (ICH) in the fetus or neonate, often with death or major neurological damage as consequence. The reported incidence of FNAIT ranges from 1:5003 to 1:1100.4 With an annual birth rate of 185 000, the expected number of affected pregnancies in the Netherlands would range between 168 and 370. FNAIT is most often diagnosed after the birth of a clinically affected child with signs of bleeding. Coincidental detection of thrombocytopenia is not uncommon with laboratory tests for other reasons. Further testing then reveals the presence of alloantibodies in the maternal serum.

Currently, preventive measures are only possible in a subsequent pregnancy of women with a previously affected child and known FNAIT. Experience shows that FNAIT is not routinely taken into consideration by paediatricians as a possible cause of neonatal thrombocytopenia.5

This study was undertaken to evaluate the rate and consequences of a late or missed diagnosis of FNAIT by assessing the clinical presentation of first affected children, the timing of diagnosis and the outcomes of subsequent children.



The Leiden University Medical Centre is the national referral centre for pregnancies complicated by FNAIT in the Netherlands. We retrospectively evaluated prospectively collected data using our FNAIT database and neonatal records from all women and infants treated at our centre between July 2008 and July 2010.

Medical records were evaluated to obtain obstetric history, clinical presentation of the first affected child and timing of diagnosis of FNAIT.

Case definition

Early diagnosis was defined as a diagnosis of FNAIT made following clinical signs or suspicion after birth of a first affected child. Diagnosis was considered delayed if diagnostic testing for FNAIT was not performed after a first clinically affected child, with FNAIT being unknown at the outset of the subsequent pregnancy.

Study outcomes

Antenatal management and outcome in subsequent pregnancies of women with undiagnosed FNAIT in previous pregnancies were compared with those with early diagnosed FNAIT. Primary outcome was the occurrence of fetal or neonatal ICH. Secondary outcomes included other bleeding signs, the cord blood platelet count at birth and type of neonatal treatment.

Statistical analysis

Data are expressed as mean (SD) values for continuous variables or as median (range) for categorical variables. The Fisher’s exact test and the Mann–Whitney test were used for comparison as appropriate. A P-value <0.05 was considered significant. Analyses were performed using spss 16.0 for Windows statistical package (SPSS Inc., Chicago, IL, USA).


Between July 2008 and July 2010, 26 women were referred to the Leiden University Medical Centre for management of pregnancies complicated by FNAIT. Table 1 outlines the characteristics of the study group.

Table 1. Characteristics of pregnancies with FNAIT following a previous affected child (n = 26)
 Early diagnosis of FNAIT (n = 22)Delayed diagnosis of FNAIT* (n = 4)
  1. *Defined as diagnosis of FNAIT made only during or after the subsequent pregnancy.

Parity(before index pregnancy)
Pregnancy loss
One miscarriage61
Recurrent miscarriage20
Immature delivery11
Type of HPA antibodies
HPA-1a and HPA-5b1 
HPA-5a 1

Clinical presentation of older siblings

In the older siblings, 16 of the 26 children presented at birth with bleeding manifestations: skin bleeding only in 13, and ICH in three. Nine infants (35%) presented with asymptomatic thrombocytopenia, detected through blood drawing for other purposes. Seven had severe thrombocytopenia (platelet count <50 × 109/l) and two infants had moderate thrombocytopenia (50 × 109–100 × 109/l).

Early diagnosis of FNAIT and primary outcome of index cases

In 22/26 pregnancies, the diagnosis was made directly after the birth of the first affected child. Four were multiparous women. Of these four women, two lost their child in the neonatal period as a result of unrelated congenital anomalies. Thrombocytopenia was first detected in their second child. The other two multiparous women first had asymptomatic children, with symptomatic thrombocytopenic children in the next pregnancy.

Intracranial haemorrhage occurred in one index case (1/22, 5%) of the early diagnosis group. The previous child of this mother had a low platelet count (5 × 109/l), without ICH. Ultrasound at 20 weeks of gestation showed no bleeding. Just before initiation of intravenous immunoglobulin (IVIG) therapy at 28 weeks of gestation, small intracranial bleedings were identified on ultrasound. The IVIG treatment (1.0 g/kg/week) was given as planned, and an asymptomatic child was born at 36 weeks of gestation. Magnetic resonance imaging confirmed several minor bleeding sites. The child is developing normally at 1 year of age.

Late diagnosis of FNAIT and primary outcome of index cases

In four of 26 women, there was a long delay in diagnosing FNAIT. Despite the birth of a child with a low platelet count or ICH, the possible diagnosis of FNAIT was not considered and no tests for FNAIT were performed. In the delayed diagnosis group, two of the four (50%) subsequent children suffered from ICH. One neonate died 2 hours postpartum from a massive subarachnoidal haemorrhage. The other child had no clinical sequelae at 1 year of age. The four pregnancies are summarised in Table 2.

Table 2. Characteristics of pregnancies with delayed diagnosis of FNAIT
  1. IUGR, intrauterine growth restriction; CS, Caesarean section.

  2. *IVIG planned for 10 weeks but preterm birth.

Parity (at index pregnancy)G 3 P 2G2 P1G2 P1G2 P1
Clinical presentation of first affected childICHAsymptomaticAsymptomaticAsymptomatic
Platelet count of first affected childUnknown30 × 109/l67 × 109/l11 × 109/l
Presumed cause of ICH/thrombocytopeniaBirth traumaDown syndromeIUGRIUGR
Gestational age at diagnosis in index pregnancy20-week scan (diagnosis of ICH)After birth at 32 weeks (ICH)35 weeks13 weeks
Antenatal management of index pregnancyIVIG 1.0 g/kg/week for 15 weeksIVIG 1.0 g/kg/week for 3 weeksIVIG 0.5 g/kg/week for 4 weeks*
Clinical outcome of index pregnancyCS, 38 weeks
ICH, asymptomatic
CS for fetal distress at 32 weeks
ICH, neonatal death
Vaginal birth at 38 weeks. HealthyCS for fetal distress at 32 weeks. Healthy
Platelet count at birth of index pregnancy158 × 109/l11 × 109/l16 × 109/l184 × 109/l

Secondary outcomes of index cases

In Table 3, the platelet count at birth, bleeding signs other than ICH, and antenatal and neonatal treatment are given.

Table 3. Data on neonatal outcome, antenatal and neonatal treatment and type of delivery in index pregnancies with FNAIT
 Early diagnosis of FNAIT (n = 22)Delayed diagnosis of FNAIT (n = 4) P-value
  1. Values given as median (range).

  2. IVIG, intravenous immunoglobulin: 0.5 g/kg/week from 28 weeks when affected sibling had no ICH, or 1.0 g/kg/week from 16 weeks if sibling had ICH.

  3. *In one early diagnosis, a predelivery fetal blood sample was taken followed by one intrauterine platelet transfusion and a vaginal birth.

Pregnancy outcome
ICH, n (%)1 (5)2 (50)0.051
Perinatal death, n (%)01 (25)0.154
Platelet count at birth, × 109/l66 (6–278)87 (11–184)0.561
Severe thrombocytopenia (<50 × 109/l), n (%)9 (41)2 (50)0.386
Haemorrhagic symptoms excluding ICH, n (%)2 (9)1 (25)0.355
Antenatal treatment
IVIG, n (%)*21 (95)2 complete (50), 1 incomplete, 3 weeks0.271
Neonatal treatment
Platelet transfusion only, n (%)4 (18)1 (25)0.445
Platelet transfusion and IVIG, n (%)3 (14)00.592
None, n (%)15 (68)3 (75)0.437
Mode of delivery
Vaginal, n (%)13 (59)1 (25)0.206
Caesarean section, n (%)9 (41)3 (75)


In this cohort study of 26 women with FNAIT, delay of diagnosis was identified in four pregnancies (15%). Two of these four fetuses suffered from severe ICH. Several factors were presumed to have caused the low platelets in the previous pregnancies—Down syndrome, intrauterine growth restriction and birth trauma—and kept the clinicians from requesting the appropriate investigations.

The most feared complication of fetal or neonatal thrombocytopenia is ICH. Untreated newborns with FNAIT are reported to be affected by ICH in 7–26% of pregnancies.6–8 After the birth of a child with ICH, the recurrence rate of ICH is as high as 79% (95% CI 61–79%).9–11 Surviving children suffer from severe neurological sequelae including mental retardation, cerebral palsy, cortical blindness and seizures in 14–26%.5,8 As safe and nearly 100% effective antenatal treatment is available, using IVIG,12 it is of clinical importance to detect neonates with FNAIT.

In our study, one ICH occurred in the group with known FNAIT, just before starting IVIG. This child is now developing normally at 1 year of age. The mother had received IVIG for 8 weeks, which could be the reason for the good long-term outcome. However, as this is the only failure in over 10 years and more than 120 women treated with noninvasive management (IVIG only), we have not changed our policy of starting IVIG at 28 weeks.

The limited number of pregnancies means that our analysis has limited power to detect statistically significant differences. However, these illustrative examples of delayed diagnosis show that missing the diagnosis of FNAIT can have devastating consequences for subsequent children, including perinatal death.

More than one-third of the neonates in our study with FNAIT presented with asymptomatic thrombocytopenia. A Norwegian study showed that in the absence of a routine screening programme, only 14% of affected neonates would be identified.13 Our data confirm their conclusion that for effective prevention of ICH due to FNAIT, routine screening of all pregnant women for human platelet antigen 1a (HPA-1a) antibodies needs to be implemented.

The same recommendation was given by Knight et al.5 following a national study in the UK which showed significantly more children diagnosed with ICH when FNAIT was unknown or unrecognised at the onset of pregnancy compared with pregnancies with known diagnosis of FNAIT (at 1 year of age 10 versus 0% severe disability or death).

The debate on implementation of routine antenatal screening for FNAIT depends mostly on cost-effectiveness. Several studies provided calculations that all reach the conclusion that screening is likely to be cost-effective.4,8,14–16 However, until then, given the serious consequences of missing the diagnosis of FNAIT, we recommend platelet antibody detection for all pregnancies involving (suspected) antenatal fetal haemorrhage and neonatal thrombocytopenia. We propose a limited serological investigation for antenatally suspected cases, by maternal HPA-1a typing and screening for the clinically important HPA antibodies. This can be extended for newborns with proven thrombocytopenia with a cross-match between maternal serum and paternal platelets, maternal and neonatal (or paternal) HPA typing and platelet autoantibody investigation depending on the differential diagnosis.


Delay in diagnosing FNAIT, resulting in withholding appropriate preventive measures in the subsequent pregnancy, may lead to perinatal death or severe brain damage in newborns. We recommend that in all neonates with thrombocytopenia at birth, the diagnosis FNAIT should be considered followed by appropriate testing. Screening all pregnant women for HPA-type would be even more effective.

Disclosure of interests

None of the authors has any relevant interest to declare.

Contribution to authorship

All authors have significantly contributed to this scientific work and approved the final version of the manuscript. All authors contributed in gathering of the data. DO was involved in the design of the study and supervised the data analyses and writing process. KM performed the data analyses and wrote the first draft of the manuscript. All authors subsequently revised the manuscript critically and assisted in drafting of the final version of the manuscript.

Details of ethics approval

Ethical approval was not required for this study.


There was no associated funding.