Introduction

Neonatal alloimmune thrombocytopenia (NAIT) is among the most common causes of thrombocytopenia, manifesting around birth by marked thrombocytopenia [1, 2]. The incidence of NAIT is estimated to be 1 in 2000–3000 births [3, 4]. NAIT is caused by fetomaternal platelet incompatibility that results in transplacental passage of maternal antibodies against fetal human platelet antigens (HPA) and/or human leukocyte antigens (HLA). Anti-HLA and anti-HPA have been reported in 24.4% and 0.91% of pregnant women, respectively [5, 6]. However, much fewer reports with very small numbers of cases have described NAIT resulting from anti-HLA antibodies rather than from anti-HPA antibodies.

Down syndrome (DS) was the first described chromosomal disorder and is the most common autosomal trisomy. The incidence of thrombocytopenia in neonates with DS is higher than that in the general population [7]. Moreover, DS may present with hematological disorders such as transient abnormal myelopoiesis (TAM), occurring in approximately 10% of DS neonates [8, 9]. Thus, thrombocytopenia is among the most common hematological abnormalities associated with DS. However, cases of NAIT with DS caused by anti-HLA antibodies have barely been reported [10], although resolving the differential diagnosis of thrombocytopenia is very important to improve outcome of DS.

Here, we report a case of DS with NAIT, attributed to HLA antibodies with A31 and B61 specificity.

Case report

The mother (gravida 9 para 5) was 38 years old with no history of blood transfusion or autoimmune disease; she was not on any medications. Her pregnancy was uneventful, with negative screening results for toxoplasmosis, rubella, cytomegalovirus, herpes simplex virus, and vaginal group B streptococcus. Complete blood cell counts were normal throughout the pregnancy and her blood type was B Rh (D) positive. Her fetus was suspected of having DS due to brachycephaly by ultrasonography at 20 weeks’ gestation. Her platelet (PLT) counts were 265 × 109/L before and 223 × 109/L after delivery.

A boy was delivered by vacuum extraction at 36 weeks’ gestation due to late decelerations in fetal heart rate. He weighed 1,936 g (− 2.3 SD), and his height and head circumstance were 46.0 cm (− 0.5 SD) and 31.0 cm (− 1.2 SD), respectively. His Apgar scores were 6 and 7 at one and five minutes. He was admitted to our neonatal intensive care unit shortly after birth due to decreased oxygen saturation and clinical features of DS. Laboratory data demonstrated metabolic acidosis, polycythemia, and thrombocytopenia; chromosomal analysis confirmed 47, XY, + 21 (Table 1). There was no evidence of hemorrhage or other abnormal findings by ultrasound of the head and abdomen. As shown in Fig. 1, the initial PLT count on the peripheral venous blood sample was 109 × 109/L. On day 2, the patient had no signs of bleeding, such as petechiae; however, the PLT count declined to 13.0 × 109/L. Therefore, random-donor PLT concentrates were transfused on days 2 and 3. On day 4, the PLT count increased to 107 × 109/L after PLT transfusion. On day 3, immunoserological tests for NAIT were conducted, with positive results on day 4. Intravenous immunoglobulin (IVIG) with 500 mg/kg/day was administered for 3 days. The PLT count increased to 76.0 × 109/L by day 6 and to 225 × 109/L by day 14. Thereafter, the thrombocytopenia did not relapse without clinical complications, and he was discharged on day 26.

Table 1 Laboratory values on admission (day 0)
Full size table
Fig. 1
figure 1

Platelet counts during the patient’s clinical course. The vertical axis shows platelet count and the horizontal axis marks the days after birth. Inverted filled arrow random-donor platelet concentrates (10 mL/kg) were transfused on days 2 and 3. Inverted filled arrow intravenous immunoglobulin was administrated at a dose of 500 mg/kg/day on days 4, 5 and 6. IVIG intravenous immunoglobulin, PC platelet concentrate, PLT platelet(s)

Full size image

Materials and methods

Preserved blood samples from the mother at 36 weeks’ gestation and the cord blood at birth were further investigated on day 3. The blood samples from the patient on day 3 and those from the parents on day 3 and 8 months after childbirth were collected, after obtaining written informed consent from the parents for immunoserological testing. Additional informed consent was obtained from both parents to report this case.

HLA genotyping

The blood samples collected from the parents on day 3 and the cord blood at birth were screened for HLA genotyping using the polymerase chain reaction reverse sequence-specific oligonucleotide-Luminex method (Genosearch HLA Kit, MBL, Nagoya, Japan) with a Luminex 100 IS fluoroanalyzer (Luminex Inc., Austin, TX, USA), in accordance with the manufacturer’s instructions.

HPA and Naka (CD36) antigen genotyping

The blood samples collected from the parents on day 3 and the cord blood at birth were screened for HPA and Naka antigen genotyping using a commercially available kit (WAKFlow HPA typing kit, Wakunaga, Hiroshima, Japan). Platelet antigens of the panel were genotyped for HPA-1 through HPA-7, HPA-15, HPA-21, and Naka.

HLA antibody screening, identification, titration, and cross-matching

Serum samples of the mother collected during pregnancy and those of patient collected on day 3 were screened for anti-HLA class I antibodies using the Luminex® method with fluorescent beads coated with HLA-A, -B, and -Cw antigens (One Lambda LABScreen Single Antigen Class I & Supplement, Veritas Corporation, Tokyo, Japan). Maternal serum samples during pregnancy were screened for the antibodies against a panel of platelets using a commercial mixed passive hemagglutination assay (MPHA) kit (ANTI-HPA·MPHA Panel, Beckman Coulter, USA). Maternal serum samples collected during pregnancy were tested against platelets from the paternal blood sample collected on day 3 using MPHA. To distinguish between anti-platelet and anti-HLA class I antibodies, the reactivity of serum against chloroquine-treated platelets was determined. The titer of each serum sample was based on serial dilutions in PBS until a negative test result was obtained. To detect anti-HLA antibody and analyze its specificity, maternal serum samples collected during pregnancy were tested against lymphocytes from the paternal blood sample collected on day 3 with an anti-human immunoglobulin lymphocytotoxicity test using a panel of lymphocytes. Neonatal platelets were not available for testing.

HPA antibody screening and cross-matching

MPHA, enzyme-linked immunosorbent assay (ELISA), and a modified rapid monoclonal antibody (mAb)-specific immobilization of platelet antigens (MAIPA) assay were performed to detect the HPA antibodies for HPA-1 through HPA-7, HPA-15, HPA-21, and Naka. Maternal serum samples collected 8 months after childbirth and the preserved cord blood were screened for HPA antibodies using the MPHA, ELISA, and modified rapid MAIPA. Maternal serum collected 8 months after childbirth and the cord blood preserved from the time of childbirth were tested against paternal platelets collected 8 months after childbirth using the modified rapid MAIPA. The preserved blood samples of the mother that were collected during pregnancy were also tested for HPA-15. MPHA was performed using a commercially available kit (ANTI-HPA·MPHA Panel, Beckman Coulter, USA) to detect HPA-1 through -6 and Naka. ELISA was performed using a commercially available kit (PAKPLUS, Immucor, Norcross, GA) to detect HPA-1 through -5 and Naka. The modified rapid MAIPA protocols were based on the original protocol [11, 12] with four capture mAbs: the antibodies against GP-IIb/IIIa, Ib/IX, Ia/IIa, and CD109.

Results

From HPA and HLA genotyping (Table 2a), maternal-paternal or neonatal mismatch of HLA-A31, B61, Cw15 and HPA-15a were observed. It was confirmed that the paternal platelets or white blood cells strongly reacted to the maternal serum in HLA cross-matching (Table 2b). No anti-HPA antibody was detected in the maternal or neonatal serum and no reactions were observed in HPA cross-matching between paternal platelets and maternal or neonatal serum (Table 2b). These results did not suggest HPA antibody involvement in NAIT. Multispecific anti-HLA class I antibodies were detected in maternal and neonatal serum samples, and the antibodies were specific to HLA-A31 and B61 among the father-derived HLA antigens in maternal serum sample (Table 2c). Hence, along with the results of cross-matching, this patient was diagnosed as NAIT owing to the presence of anti-HLA A31 and B61 antibodies.

Table 2 Immunoserological investigation
Full size table

Discussion

NAIT is diagnosed based on the presence of IgG alloantibodies in the maternal serum that react with neonatal platelet antigens. In Japan, the most frequently occurring cases of NAIT are caused by HPA-4b [6]. The ratio of positive anti-HLA antibodies increases per pregnancy with raising their titers [5]. However, the incidence of NAIT is about 20 times higher with anti-HPA antibodies than with anti-HLA antibodies [6]. This may be in part explained by the fact that maternal anti-HLA antibodies tend to be attenuated and are difficult to transfer to the baby because they are neutralized by the HLA antigens of the baby that are expressed in the placenta [13].

In the present case, there were HLA-A31, HLA-B61, HLA-Cw15, and HPA-15a mismatches between the mother and the patient. However, the HPA antibody screening and HPA crossmatching did not show the suggestive result of HPA antibody involvement in NAIT. In contrast, anti-HLA class I antibodies were detected both in maternal and neonatal serum, although specificities to HLA A31 and B61 antibodies were determined only in the maternal serum. Together with the result of cross-matching, this patient was diagnosed as NAIT due to the anti-HLA A31 and B61 antibodies. Little sample could be taken from the baby, and the antibody was assumed to be consumed by neonatal tissue [14], resulting in the limitation to determine specificities of anti-HLA antibodies in this small patient.

Growth retardation in trisomy 21 fetuses is partly attributable to placental structural abnormalities [15]. The chorionic villi in trisomy 21 were reported to be bigger and more asymmetric and irregular in shape than were normal villi [15]. Dysmorphic villi of trisomy 21 show villous vascular growth. Therefore, maternal anti-HLA antibodies may not be attenuated and may be transferred to the baby because of no neutralization by the baby’s HLA antigens that are expressed in the placenta owing to the placental structural abnormalities in trisomy 21. This could be explained by the fact that maternal HLA antibodies result in the development of NAIT. In the present case, the mother had prior pregnancies, and the anti-HLA antibody titer was likely to increase due to HLA antigen stimulation from them [5]. But that alone did not lead to onset. It is thought that placental dysfunction may have existed, and the damaged placenta may have facilitated the leakage of maternal blood.

The incidence of thrombocytopenia in neonates with DS is higher than that in the general population. DS is associated with several hematological disorders, such as TAM. Pediatricians who are involved in the care of patients with DS need to take care of the bleeding and other clinical findings due to hematological abnormalities. NAIT is often asymptomatic or presents with only petechial hemorrhages in the absence of complications, such as intracranial hemorrhage, making it difficult to establish a diagnosis based on the symptoms. As mentioned earlier, DS may have placental structural abnormalities that may predispose the patient to NAIT. If DS is complicated by thrombocytopenia, NAIT should be considered as a differential diagnosis from the beginning, and medical history-taking and a careful clinical examination are crucial for establishing a correct diagnosis. Very few DS case with NAIT due to anti-HLA antibodies has been reported, but NAIT should figure prominently in the differential diagnosis of thrombocytopenia in this patient population.