David L. Allen1, Geoffrey F. Lucas2 & Michael F. Murphy3
1NHS Blood and Transplant, John Radcliffe Hospital, Oxford, UK
2NHS Blood and Transplant, Bristol, UK
3University of Oxford and NHS Blood and Transplant and Department of Haematology, John Radcliffe Hospital, Oxford, UK
Antigens on platelets and granulocytes
Antigens on human platelets and granulocytes can be categorized according to their biochemical nature into:
· carbohydrate antigens on glycolipids and glycoproteins:
a. A, B and O,
b. P and Le on platelets, I on granulocytes;
· protein antigens:
a. human leucocyte antigen (HLA) class I (A, B and C),
b. glycoprotein (GP)IIb/IIIa, GPIa/IIa, GPIb/IX/V, etc., on platelets,
c. FcγRIIIb (CD16), CD177, etc., on granulocytes;
· hapten-induced antigens:
a. quinine, quinidine,
b. some antibiotics, e.g. penicillins and cephalo-sporins,
These antigens can be targeted by some or all of the following types of antibodies:
· isoantibodies and
· drug-dependent antibodies.
Many platelet and granulocyte antigens, e.g. ABO and HLA class I, are shared with other cells (Table 5.1); others, however, are restricted to single lineages. This chapter is divided into two sections: the first focuses on proteins expressed predominantly on platelets, and in particular the human platelet antigens (HPAs), whilst the second section focuses on the equivalent proteins and alloantigens expressed predominantly on neutrophils (human neutrophil antigens, HNAs).
Table 5.1 Antigen expression on peripheral blood cells.
Human platelet antigens
Twenty-one polymorphisms have been described (Table 5.2); most were first discovered during investigation of cases of neonatal alloimmune thrombocytopenia (NAIT). The majority of these antigens are located on the GPIIIa subunit of the GPIIb/IIIa integrin (CD41/CD61), which is present at high density on the platelet membrane and seems to be particularly polymorphic and immunogenic. Others are located on the GPIIb subunit, on GPIa/IIa (CD49b), GPIb/IX/V and CD109.
Table 5.2 HPA systems.
These receptor complexes are critical to platelet function and are responsible for the stepwise process of platelet attachment to the damaged vessel wall. GPIb/IX/V is the receptor for the von Willebrand factor (vWF) and is implicated in the initial tethering of platelets to damaged endothelium. The GPIbα-bound vWF interacts with collagen, facilitating the interaction of collagen with its signalling (GPVI) and attachment receptors (GPIa/IIa). Outside-in signalling via GPVI leads to conformational changes in integrins GPIIb/IIIa and GPIa/IIa from ‘locked’ to ‘open’ configurations, exposing the high affinity binding sites for collagen and fibrinogen, respectively. GPIIb/IIIa is the main platelet fibrinogen receptor and is critical to the final phase of platelet aggregation, but it also binds fibronectin, vitronectin and vWF. The function of CD109 has not been fully elucidated although recent studies suggest a role in regulation of transforming growth factor beta (TGF-β)-mediated signalling. Glanzmanns thrombasthenia and Bernard–Soulier syndrome are rare and severe, autosomal recessive, platelet bleeding disorders caused by deletions or mutations in the genes encoding GPIIb and GPIIIa, or GPIbα, GPIbβ and GPIX, respectively.
Inheritance and nomenclature
Most HPAs have been shown to be bi-allelic, with each allele being codominant, although recently the HPA-1, -5 and -7 systems have been shown to have third alleles. Historically, platelet-specific antigens were named using an abbreviation of the name of the propositus in which the alloantibody was first detected. Some systems were described simultaneously by different investigators and, confusingly, several names were assigned to the same polymorphism, e.g. Zw and PlA and Zav, Br and Hc. In 1990, an International Society of Blood Transfusion working group agreed a new nomenclature for platelet-specific alloantigens, the HPA nomenclature, and subsequently guidelines for naming of newly discovered platelet-specific alloantigens . Each system is now numbered consecutively (HPA-1, -2, -3 and so on) (Table 5.2) according to its date of discovery, with the major allele in each system designated ‘a’ and the minor allele ‘b’. Antigens are only included in a system if antibodies against the alloantigen encoded by both the major and minor alleles have been reported; if an antibody against only one system antigen has been reported, a ‘w’ (for workshop) is added after the antigen name, e.g. HPA-10bw.
With the advent of techniques such as immunoprecipitation of radioactive-labelled platelet membrane proteins, the monoclonal antibody-specific immobilization of platelet antigen (MAIPA) assay (Figure 5.1) and the polymerase chain reaction (PCR), the genetic and molecular basis of all HPAs has been elucidated (Figure 5.2 and Table 5.2). For all but one of the 21 HPAs, the difference between the two alleles is a single nucleotide polymorphism (SNP), which changes the amino acid in the corresponding protein (Figure 5.2). Twelve of the HPAs are grouped into six HPA systems (HPA-1 to -5 and HPA-15) and for all of these, except HPA-3 and HPA-15, the minor allele frequency is ≤0.2. Homozygosity for the minor allele is therefore relatively rare so providing compatible blood components for patients with antibodies against high frequency HPA antigens can be difficult. Some SNPs are population-specific, e.g. SNP rs5918 (HPA-1 system) is absent in populations of the Far East; conversely, SNP rs5917 (HPA-4) is not present in Caucasians but is present in Far Eastern populations. It is therefore important to take ethnicity into account when investigating clinical cases of suspected HPA alloimmunization.
Fig 5.1 MAIPA assay. (1) Human serum and murine monoclonal antibody (MoMab) directed against glycoprotein being studied; e.g. GPIIb/IIIa are sequentially incubated with target platelets: in (a) the test serum contains anti-HPA-1a and in (b) no platelet antibodies are present. (2) After incubation a trimeric (a) or dimeric (b) complex is formed. Excess serum antibody and MoMab is removed by washing. (3) The platelet membrane is solubilized in nonionic detergent, releasing the complexes into the fluid phase; particulate matter is removed by centrifugation. (4) The lysates containing the glycoprotein/ antibody complexes are added to wells of a microtitre plate previously coated with goat antimouse antibody. (5) Unbound lysate is removed by washing and enzyme- conjugated goat antihuman antibody is added. (6) Excess conjugate is removed by washing and substrate solution is added. Cleavage of substrate, i.e. a colour reaction, indicates binding of human antibody to target platelets.
Fig 5.2 Representation of the platelet membrane and the glycoproteins (GP) on which the human platelet antigens (HPA) are localized. From left to right are depicted GPIa/IIa, GPIIb/IIIa, CD109 and GPIb/IX/V. The molecular basis of the HPAs are indicated by black dots, with the amino acid change in single-letter code and by residue number in the mature protein.
Typing for HPAs
Until the early 1990s, HPA typing was performed by serological assays (‘phenotyping’). This required the use of monospecific antisera, which were relatively uncommon as the majority of immunized individuals produced HLA class I antibodies in addition to HPA antibodies. The development of more advanced assays, e.g. the MAIPA assay (Figure 5.1) that were able to elucidate complex mixtures of antibodies against different GPs, permitted more extensive phenotyping, but some antisera were simply not available.
Many DNA-based typing techniques have been developed; these have largely replaced phenotyping in the majority of platelet immunology laboratories. One such assay is the polymerase chain reaction using sequence-specific primers (PCR-SSP). This is a fast and reliable technique and has become one of the cornerstone techniques in platelet immunology laboratories (see Figure 5.3). High throughput HPA SNP typing techniques with automated readout, such as Taqman assays, are now also in routine use and allow typing at reduced cost. Novel techniques for the simultaneous detection of numerous SNPs are emerging and these may become the routine method for donor typing in blood centres.
Fig 5.3 PCR-SSP determination of HPA-1, -2, -3, -4, -5 and -15 genotypes. The upper band present in all lanes is the 429-bp product of human growth hormone; lower bands are the products of sequence-specific primers. The results are read from left to right, i.e. lane 1 HPA-1a, lane 2 HPA-1b, etc. The HPA genotype in this case is 1b/1b, 2a/2a, 3a/3a, 4a/4a, 5a/5a, 15b/15b. Courtesy of Dr Paul Metcalfe (NIBSC).
Genotyping of fetal DNA from amniocytes or from chorionic villus biopsy samples is of clinical value in cases of HPA alloimmunization in pregnancies where there is a history of severe NAIT and the father is heterozygous for the implicated HPA SNP. Noninvasive HPA genotyping assays based on the presence of trace amounts of fetal DNA in maternal plasma have been described and reduce the risk to the fetus from invasive sampling procedures.
Platelet isoantigens, autoantigens and hapten-induced antigens
GPIV is absent from the platelet membrane in 4% of African Blacks and 3–10% of Japanese. If these individuals are exposed to GPIV-positive blood through pregnancy or transfusion, they may produce GPIV isoantibodies. These antibodies may cause NAIT or platelet refractoriness and may be responsible for febrile nonhaemolytic transfusion reactions (FNHTRs). Similarly, formation of isoantibodies can complicate both the pregnancies and transfusion support of patients with Glanzmanns' thrombasthenia or Bernard–Soulier syndrome.
The GPs carrying the HPAs are the target of autoantibodies in autoimmune thrombocytopenia (AITP); these autoantibodies bind to the platelets of all individuals, regardless of their HPA type. Platelet autoimmunity is frequently associated with B-cell malignancies and during immune cell re-engraftment following haemopoietic stem cell transplantation. In both situations, the presence of autoantibodies may contribute to the refractoriness to donor platelets.
Some drugs too small to elicit an immune response in their own right may bind to platelet GPs in vivo and act as a hapten . In some patients, the haptenized platelet GP can trigger the formation of antibodies that only bind to the GP in the presence of hapten; a classic example is quinine. Typically, quinine-dependent antibodies bind to GPIIb/IIIa and/or GPIb/IX/V although other GPs are sometimes the target. Many other drugs, including several antibiotics, have been associated with hapten-mediated thrombocytopenia. In haemato-oncology patients, who often receive a spectrum of drugs, unravelling the causes of persistent thrombocytopenia or poor responses to platelet transfusions can be complex because of the many possible causes of thrombocytopenia. If the thrombocytopenia is hapten-mediated, withdrawal of the drug will result in rapid recovery of the platelet count.
Another form of drug-dependent thrombocytopenia may be observed in coronary artery disease patients treated with ReoPro (Abciximab). This function-blocking chimeric human–mouse F(Ab) fragment against GPIIb/IIIa causes precipitous thrombocytopenia in approximately 1% of patients due to the presence of pre-existing antibodies against ReoPro-induced neoepitopes.
The interaction of heparin with platelet factor 4 induces epitope formation that can cause antibody production and lead to heparin-induced thrombocytopenia (HIT) (see Chapter 30), but the reduction in platelet count is less profound than in classic examples of hapten-mediated immune thrombocytopenia. The risk of thrombotic complications is the main concern in patients with HIT who show a mild but significant reduction in their platelet count after heparin administration.
Detection of HPA alloantibodies
Over the last five decades, techniques for the detection of HPA antibodies have evolved significantly. Early assays, e.g. the platelet agglutination test, were both insensitive and nonspecific. The platelet immunofluorescence test (PIFT) improved sensitivity but is still classified as a nonspecific assay as it is unable to distinguish between HPA and HLA class I antibodies. Despite this limitation, the PIFT with a flow cytometric endpoint remains widely used since it is a whole cell assay capable of detecting a wide range of antibody specificities and is especially useful in detecting both autoantibodies and alloantibodies against HPA-2 and -3. The principles of the PIFT are shown in Plate 5.1 (see the plate section). Later assays that use purified or captured GPs, e.g. the MAIPA assay and solid-phase ELISA assays, were developed; these are both sensitive and specific and have become cornerstone techniques for the detection and identification of HPA antibodies. The widely used MAIPA assay captures specific GPs using monoclonal antibodies and can be used to analyse complex mixtures of platelet antibodies in patient sera . The principle of this assay is shown in Figure 5.1. The MAIPA assay requires considerable operator expertise in order to ensure maximum sensitivity and specificity and selection of appropriate screening cells is critical. The use of platelets heterozygous for the relevant HPA or from donors who have a low expression of particular antigens, e.g. HPA-15, may result in the failure to detect clinically significant alloantibodies. Furthermore, recent evidence suggests that integrin conformation is an important factor in assay sensitivity. A disadvantage of assays that use purified GPs, rather than captured GPs as in the MAIPA assay, is that not all clinically important GPs are available, e.g. CD109. It can thus be seen that the detection and characterization of platelet antibodies can be problematic. More recently the use of biotinylated recombinant GPIIIa proteins coupled to microbeads suitable for use in high throughput assays has been described, with the potential to multiplex and streamline laboratory investigations of platelet alloimmunization.
Fig 5.4 Seventh pregnancy of a patient who has had five miscarriages. The last of these was shown to have hydrops and hydrocephalus and a platelet count of only 17 × 109/L, and the serological findings supported a diagnosis of anti-HPA-1a mediated NAIT. The fetal platelet count was <10 × 109/L at 25 weeks' gestation in the sixth pregnancy, and a cord haematoma developed during FBS resulting in fetal death. In the seventh pregnancy, prednisolone 20 mg/day and IvIgG 1 g/kg/week were administered to the mother from 16 weeks until delivery. The figure shows pre- and posttransfusion platelet counts following serial FBS and platelet transfusions. The fetal platelet count was <10 × 109/L at 26 weeks. The aim was to maintain the fetal platelet count above 30 × 109/L by raising the immediate posttransfusion platelet count to above 300 × 109/L after each transfusion. The fetal platelet count fell below 10 × 109/L on one occasion when there were problems in preparing the fetal platelet concentrate and the dose of platelets was inadequate. CS, caesarean section.
Fig 5.5 Representation of the amino acid substitutions resulting in the HNA-1a, -1b, -1c forms of FcγRIIIb. The positions of the amino acid substitutions arising from the allelic variation of the FcγRIIIb gene are depicted by black dots. Amino acids are given in three-letter acronyms. The intrachain disulfide bonds create two domains, which are closely related to the C-terminal heavy chain domains of IgG.
Clinical significance of HPA alloantibodies
HPA alloantibodies are responsible for the following clinical conditions:
· NAIT: this condition is described in detail below (but also see Chapter 32);
· refractoriness to platelet transfusions (described in detail in Chapter 28); and
· posttransfusion purpura (PTP) (described in detail in Chapter 12).
Neonatal alloimmune thrombocytopenia
The first case of NAIT was described by van Loghem in 1959. The existence of the platelet equivalent of haemolytic disease of the fetus and newborn (HDFN) had long been suspected, but laboratory confirmation was delayed because the detection of platelet antibodies was extremely technically demanding.
NAIT is now a well-recognized clinical entity with an estimated incidence of severe thrombocytopenia due to maternal HPA antibodies of 1 per 1000–1200 live births. Unlike HDFN, about 30% of cases of NAIT occur in first pregnancies.
Definition and pathophysiology
NAIT is due to maternal HPA alloimmunization caused by feto-maternal incompatibility for a fetal HPA inherited from the father and which is absent in the mother. Maternal IgG alloantibodies against the fetal HPA cross the placenta and bind to fetal platelets and, depending on a number of factors, may reduce platelet survival. Severe thrombocytopenia in the term neonate, accompanied by haemorrhage, is generally caused by HPA-1a antibodies if the mother is Caucasian or Black African. Antibodies against HPA-2 and HPA-4 antigens are generally implicated in cases of Far Eastern ethnicity. In the latter and in Black Africans, GPIV deficiency should also be considered. Anti-HPA-5b tends to cause much less severe NAIT than anti-HPA-1a.
NAIT due to alloantibodies against other HPAs is infrequent and HLA class I antibodies, present in 15–25% of multiparous women, are not thought to cause NAIT. Destruction of IgG-coated fetal platelets takes place predominantly in the spleen through interaction with mononuclear cells bearing Fcγ receptors for the constant domain of IgG.
HPA-1a is known to be expressed on fetal platelets from 16 weeks' gestation and placental transfer of IgG antibodies can occur from 14 weeks, so thrombocytopenia can occur early in pregnancy and ICH has been reported as early as 16 weeks' gestation.
Prospective screening of pregnant Caucasian women has shown that about 1 in 1200 neonates has severe thrombocytopenia (<50 × 109/L) because of alloimmunization against HPA-1a. However, the authors' experience and other studies, where prospective screening was not carried out, indicate that the number of samples referred for investigation of suspected NAIT is considerably less, which suggests that many cases are undiagnosed . HPA-5b antibodies are often found in pregnant women, but they cause clinically significant platelet destruction much less frequently than anti-HPA-1a, possibly due to the low copy number of the GPIa/IIa complex (<2000/platelet compared to 50 000/platelet for GPIIb/IIIa).
A typical case of NAIT presents with skin bleeding (purpura, petechiae and/or ecchymoses) or more serious haemorrhage, such as intracranial haemorrhage (ICH), in a full-term and otherwise healthy newborn with a normal coagulation screen and isolated thrombocytopenia. There are less common presentations in utero, including ventriculomegaly, cerebral cysts and hydrocephalus, which may be discovered by routine ultrasound. Although rare, hydrops fetalis has been reported in association with NAIT and this diagnosis should be considered if there are no other obvious reasons for the hydrops.
The precise incidence of ICH due to NAIT is unknown, but conservative estimates suggest that it is as low as 1 in 20 000 live births, which equates to approximately 35 cases per annum in the UK. Nearly 50% of severe ICHs occur in utero, usually between 30 and 35 weeks' gestation, but sometimes before 20 weeks. At the other end of the clinical spectrum, NAIT can be discovered incidentally when a blood count is performed for other reasons.
Severe NAIT in a neonate is a serious condition and requires correction of the platelet count. Appropriate management (see below) is essential to prevent ICH and the possibility of a lifelong disability.
Other causes of neonatal thrombocytopenia are infection, prematurity, intrauterine growth retardation, inherited chromosomal abnormalities (particularly trisomy 21), maternal AITP and, very rarely, inherited forms of inadequate megakaryopoiesis. Precise figures on the incidence of neonatal thrombocytopenia caused by viral infection are not available. Maternal platelet autoimmunity is rarely associated with severe thrombocytopenia in the neonate, but should be considered in women with a history of AITP.
Platelet-type von Willebrand's disease, in which mutations in the GPIbα gene are associated with a propensity for in vitro platelet aggregation, can lead to falsely low platelet counts.
Only antibodies against HPAs or isoantibodies against GPIIb/IIIa, GPIb/IX, CD109 and GPIV are thought to cause alloimmune thrombocytopenia in the fetus and neonate, although there are reports of platelet autoantibodies from patients with AITP crossing the placenta and causing neonatal thrombocytopenia.
For appropriate clinical management, the cause of severe thrombocytopenia in an otherwise healthy neonate should be urgently investigated. Screening for maternal HPA antibodies must be carried out, using techniques with appropriate sensitivity and specificity. The combination of two techniques such as the indirect PIFT and MAIPA assays, using a panel of group O, HPA-typed platelets, remains the preferred option in many reference laboratories and is an approach supported by results of quality assessment exercises.
HPA antibodies are detected in approximately 15% of referrals of suspected NAIT referred to the National Platelet Immunology Reference Laboratory in England. The most frequently detected antibody specificities are HPA-1a and HPA-5b, which are implicated in about 85% and 10% of clinically diagnosed cases of NAIT, respectively. The ability of an HPA-1a-negative mother to form anti-HPA-1a is partly controlled by the HLA DRB3*01:01 allele. This allele is present in approximately 30% of Caucasoid women and the chance of HPA-1a antibody formation is greatly enhanced in HPA-1a-negative women who are HLA DRB3*01:01-positive compared to DRB3*01:01-negative women (odds ratio of 140). This high level association between HLA class II type and the formation of alloantibodies has not been observed for any other blood group antigen. The absence of HLA DRB3*01:01 has a negative predictive value of >90% for HPA-1a alloimmunization but its positive predictive value is only 35%, limiting its potential usefulness as part of an antenatal screening programme. However, it remains of clinical use when counselling female siblings from index cases who have formed HPA-1a antibodies in pregnancy. About 15% of HPA-1a-negative pregnant women develop anti-HPA-1a in pregnancy, and of these about 30% will deliver a neonate with a platelet count <50 × 109/L. The HPA-15 system was described a decade ago, but its clinical relevance has only recently been demonstrated, with a number of studies having shown that it is the third most commonly encountered HPA antibody specificity and that its effects may be as severe as in anti-HPA-1a-mediated NAIT.
Molecular typing of the parents and neonate for HPA-1, -2, -3, -5 and HPA-15 should be performed because the results will be informative when interpreting antibody investigation results. For patients from the Far East, HPA-4 must also be included and the platelets should be investigated for GPIV expression status.
Alloimmunization against low frequency HPAs, e.g. HPA-9bw, explain some NAIT referrals that have a negative antibody screen for the common HPA antibody specificities . A practical approach to detecting antibodies against low frequency antigens that are absent from cells used in antibody screening panels is to perform a cross-match against paternal platelets, although it is necessary to exclude positive findings due to ABO or HLA class I antibodies.
Genotyping of the maternal, paternal and affected infants' DNA samples for HPA-bw SNPs is also a cost-effective approach in this clinical setting.
A cord platelet count of <100 × 109/L should be repeated using a venous sample and a blood film examined. The neonate should be examined for skin or mucosal bleeding if a low platelet count is confirmed. If the platelet count is <30 × 109/L or if there are signs of bleeding with a low count, it is strongly recommended that the neonate is transfused with donor platelets that are HPA-1a and -5b-negative, as these will be compatible with the maternal HPA alloantibody in ≥95% of NAIT cases. The authors have shown that the transfusion of such platelets to infants affected by HPA-1a or -5b mediated NAIT results in a higher increment and more prolonged platelet survival than transfusion of random donor (HPA-1a-positive) platelets. However, if HPA-1a- and -5b-negative platelets are not immediately available and there is an urgent clinical need for transfusion then random, ABO and RhD compatible, donor platelets should be used in the first instance. A platelet count should be performed approximately 1 hour after completion of the platelet transfusion, and subsequently at least daily until the platelet count has been demonstrated not to be falling.
The results of laboratory investigations should not delay immediate platelet transfusion, as full investigation may be time-consuming and the risk of cerebral bleeds is highest in the first 48 hours post-delivery. In a typical case, the platelet count should recover to normal within a week, although a more protracted recovery can occur. Intravenous immunoglobulin (IvIgG) is not recommended as a first-line treatment as it is only effective in about 75% of cases and there is a delay of 24–48 hours before a satisfactory count is achieved; this is in contrast to the immediate effect of transfusion of HPA-compatible donor platelets. A cerebral ultrasound scan of the baby within the first week of life should be considered if the platelet count is <50 × 109/L, and is recommended when the platelet count is 30 × 109/L.
In a subsequent pregnancy of a mother with a known history of NAIT, the clinical management needs to be planned by a team experienced in the management of the risks of this condition. Treatment during the subsequent pregnancy is based on the history of haemorrhage and fetal/neonatal thrombocytopenia in previous pregnancies.
The mother should be advised to avoid nonsteroidal anti-inflammatory drugs as well as aspirin. There are two main treatment options: high dose IvIgG to the mother, or in utero platelet transfusion. Over the last decade, it has become increasingly clear that the former is the safest and most effective intervention to reduce the risk of ICH in the fetus . The dose is 1 g/kg body weight at weekly intervals, usually from 20 weeks' gestation onwards; some fetal medicine specialists use a lower dose (0.5 g/kg/week) and may start between 12 and 20 weeks' gestation, depending on the history of NAIT in previous pregnancies. Early commencement of treatment is indicated where there is a history of antenatal ICH in previous pregnancies, because the earliest reports of ICH are at 16 weeks. A beneficial effect of IvIgG on the fetal platelet count occurs in approximately 70% of cases. There is a debate about the need to perform fetal blood sampling (FBS) to ascertain the platelet count and many centres do this at 28 weeks (usually after 8 weeks' treatment with IvIgG). If FBS reveals IvIgG to be ineffective in achieving a safe fetal platelet count, doubling the dose of IvIgG and/or adding corticosteroids (prednisolone, 0.5 mg/kg body weight) should be considered. If the increased intensity of treatment is ineffective, it may be necessary to switch to weekly fetal platelet transfusions. In uteroplatelet transfusions, carried out with FBS, carry a significant risk of fetal morbidity and mortality and should not be chosen as the first-line treatment but as a rescue therapy or in the management of pregnancies with a history of treatment failure on IvIgG. The transfusion of platelets has more complications compared to red cell transfusions for HDFN, e.g. bradycardia, post-needle withdrawal cord bleeds. Once commenced, the technically demanding procedure of in utero platelet transfusions is generally repeated at weekly intervals (Figure 5.4) or as indicated by the fetal platelet count.
The delivery needs careful planning between obstetric, paediatric and haematology teams to ensure an appropriate mode of delivery, and close liaison with blood transfusion services for timely provision of HPA-compatible platelets for the neonate. For neonates that have been transfused in utero, irradiation of cellular blood components is recommended.
Counselling of couples with an index case about the risks of severe fetal/neonatal thrombocytopenia in a subsequent pregnancy needs to be based on disease severity in the infant(s) and outcome of immunological investigations. The following should be taken into account:
· thrombocytopenia in subsequent cases is as severe or, generally, more severe;
· the best predictors of severe fetal thrombocytopenia in a future pregnancy are antenatal ICH and severe thrombocytopenia (platelet count <30 × 109/L) in a previous pregnancy;
· antibody specificity;
· antibody titre and bioactivity have been investigated to determine if these parameters have a predictive role in determining the severity of NAIT and contradictory data have been obtained [7–9], and currently are probably of no value in informing clinical management;
· HPA zygosity of the partner.
HPA-typed donor panels
Establishing donor panels for fetal and neonatal platelet transfusion requires a major commitment from blood services and identification of suitable donors requires the use of high throughput typing techniques. Although the frequency of HPA-1a-negative individuals amongst Caucasians is 2.5%, potential donors for fetal/neonatal transfusions must also be negative for the mandatory microbiological tests, negative for antibodies against red cells, platelets and leucocytes, and be cytomegalovirus (CMV) seronegative. The donors should also ideally be HPA-5b-negative to ensure that the panel is of potential benefit to the maximum number of cases of suspected NAIT. In order to recruit a single HPA-1a-negative donor satisfying the above criteria, approximately 1500–2000 donors have to be typed for HPA-1a. In addition, therapeutic platelets should be RhD matched, as small amounts of red cells present in platelet concentrates may immunize RhD-negative recipients and be negative for high titre anti-A and anti-B antibodies. To recruit a single RhD-negative HPA-1a-negative donor whose platelets will be suitable for a first fetal or neonatal platelet transfusion, where the fetal/neonatal blood group is unknown, approximately 6000–7000 donors need to be typed.
Human neutrophil antigens
Antigens on the membrane of human neutrophils can, as with platelets, be divided into different categories. There are common antigens that have a wider distribution on other blood cells and tissues, e.g. I and P blood group systems and HLA class I. There are ‘shared’ antigens that have a limited distribution amongst other cell types, e.g. HNA-4a and HNA-5a polymorphisms associated with CD11/18. There are also a limited number of truly neutrophil-specific antigens, e.g. HNA-1a, HNA-1b, HNA-1c polymorphisms on FcγRIII or CD16. The current nomenclature for the HNA systems includes polymorphisms that are both cell-specific and ‘shared’ (see Table 5.3) .
Table 5.3 HNA systems.
The three antigens that comprise the HNA-1 system are localized on neutrophil FcγRIIIb (CD16), one of two low affinity receptors (R) for the constant domain (Fc) of human IgG(γ) that are found on neutrophils. There are normally 100 000–200 000 copies of FcγRIIIb per neutrophil. Four amino acid differences with arginine/serine, asparagine/serine, aspartic acid/asparagine and valine/isoleucine substitutions at positions 36, 65, 82 and 106, respectively, define the difference between HNA-1a and -1b, while a single amino acid substitution (alanine 78 > asparagine) defines the HNA-1c polymorphism (see Figure 5.5). The expression of HNA-1c is frequently associated with the presence of an additional FcγRIIIb gene and increased expression of FcγRIIIb. The expression of HNA-1 antigens varies with ethnicity, with HNA-1a being more common in Chinese and Japanese populations than in Caucasians.
Two other FcγRIIIb-associated high frequency alloantigens, LAN and SAR, have been reported. The FcγRIIIb ‘Null’ phenotype is rare and is based on a double deletion or mutation of the FcγRIIIb gene and is, in some cases, associated with a deletion of the FcγRIIc gene. A maternal deficiency of FcγRIIIb can cause immune neutropenia in the newborn due to maternal FcγRIIIb isoantibodies. The FcγRIIIb molecule (but not necessarily the HNA-1 sites) can also be the antigenic target in autoimmune neutropenias.
HNA-2, formerly known as HNA-2a or NB1, is localized on a 58–64 kDa glycoprotein (CD177), expressed as a glycosylphosphatidylinositol-anchored membrane GP found both on the neutrophil surface membrane and on secondary granules. The term HNA-2a should no longer be used since it is now known that there is no antithetical antigen.
The percentage of neutrophils expressing HNA-2 varies between individuals and HNA-2 alloantibodies typically give a bimodal fluorescence profile with granulocytes from HNA-2-positive donors in immunofluorescence tests with a flow cytometric endpoint. HNA-2 antigen status can be determined by phenotyping with polyclonal or monoclonal antibodies. The HNA-2-negative phenotype is associated with particular sequence haplotypes from which nonproductive HNA-2 transcripts are generated, thereby causing a failure to express HNA-2.
HNA-3a (previously known as 5b) and HNA-3b (previously known as 5a) were originally described using antisera obtained from women immunized during pregnancy and were reported as being present on granulocytes, platelets, lymphocytes, kidney, spleen and lymph node tissue. Biochemical studies localized the HNA-3a antigen to a granulocyte glycoprotein with a molecular weight of 70–95 kDa. The polymorphism has been further localized to choline transporter-like protein 2 (CTL2) and a single point mutation at amino acid 154 (HNA-3a is encoded by Arginine and HNA-3b by Glutamine). The gene frequency of HNA-3a and HNA-3b has been reported as 0.792 and 0.207, respectively, in the German population. HNA-3a antibodies have been implicated in transfusion-related acute lung injury (TRALI) and neonatal alloimmune neutropenia (NAIN). HNA-3b antibodies have also been described but the clinical significance of these antibodies is not known.
Alloantigens on CD11a and CD11b
The genes encoding the αM and αL subunits of the β2 integrin or CD11b and CD11a are polymorphic and are associated with HNA-4a and HNA-5a, respectively. Alloantibody formation against these two polymorphisms has been observed in transfusion recipients, and recently cases of NAIN due to HNA-4a and HNA-5a antibodies have been described. The low incidence of neonatal neutropenia associated with these antibodies is probably explained by the wide distribution of these proteins on granulocytes, monocytes and lymphocytes.
Detection of neutrophil antibodies
The reliable detection and identification of neutrophil antibodies can be technically difficult. The main problems are the abundant expression of the two low affinity receptors (FcγRII and FcγRIIIb) for the constant domain of human IgG, which results in increased and variable binding of circulating immunoglobulins in normal sera and the requirement for fresh and typed donor neutrophils as panel cells, since neutrophils cannot be stored. The incidence of antibody-mediated neutropenias is comparatively rare and, therefore, the best strategy for investigation of clinical cases is a national reference laboratory where adequate technical expertise and reagents are available.
Many techniques for neutrophil/granulocyte antibody detection have been evaluated over the years. Early assays such as the granulocyte cytotoxicity and agglutination tests had a very low specificity. The granulocyte immunofluorescence and granulocyte chemiluminescence tests have the advantage of good sensitivity but are not specific, i.e. they cannot readily distinguish between granulocyte-specific and HLA class I antibodies without further investigations. For some HNA systems, e.g. antigens expressed on CD16, CD177 and CD11/18, the monoclonal antibody immobilization of granulocyte antigens (MAIGA) assay can be applied but, otherwise, immunoprecipitation of membrane-labelled neutrophil proteins remains the only reliable technique to determine the nature of the antigen. The principles of the granulocyte immunofluorescence test and the MAIGA assays are analogous to the equivalent platelet tests (Plate 5.1 and Figure 5.1, respectively). Increased understanding of the molecular nature of HNAs has opened up the potential to develop recombinant HNAs and both cell lines expressing recombinant proteins (rHNA) and soluble rHNA coupled to a solid phase have been described. These new assays have shown promise but, currently, generally lack the sensitivity and specificity of established techniques. The introduction of such techniques does, however, have the potential to transform the serological investigation of granulocyte alloantibodies and the high throughput capability of assays using beads coupled to rHNA proteins would be particularly beneficial for blood donor screening in TRALI-reduction programmes .
HNA typing has historically been based on the use of monospecific alloantisera derived from immunized patients or blood donors, but monoclonal antibodies against HNA-1a, -1b and -2 are now available. Recent advances in the understanding of the molecular basis of the HNA systems means that it is now possible to type for HNA-1a, -1b, -1c, -3a, -3b, -4a, -4b, -5a and -5b using PCR-SSP or sequence-based typing techniques.
Clinical significance of HNA antibodies
Neutrophil-specific antibodies are implicated in:
· neonatal alloimmune neutropenia (NAIN);
· febrile nonhaemolytic transfusion reactions (see Chapter 8);
· transfusion-related acute lung injury (TRALI) (see Chapter 9);
· transfusion-related alloimmune neutropenia (TRAIN);
· autoimmune neutropenia; and
· persistent post-bone marrow transplant neutropenia.
Neonatal alloimmune neutropenia
Maternal alloimmunization against neutrophil-specific alloantigens on fetal/neonatal neutrophils is a condition analogous to NAIT in terms of pathophysiology but, with an estimated incidence of 0.1–0.2% of live births, is comparatively rare as a clinically significant entity although there are no reliable figures. Clinical presentation is mainly one of bacterial infections with isolated neutropenia being the only haematological abnormality. The neutropenia may be severe but is reversible and newborn infants may require treatment with antibiotics and/or GCSF to control bacterial infections and hasten recovery to a normal neutrophil count. The neutropenia in some cases caused by HNA-1a and -1b antibodies has been reported to extend for up to 28 weeks. HNA-1a and -2 are the most commonly implicated antibody specificities.
FNHTR and TRALI (see Chapters 8 and 9)
FNHTRs have a number of different causes. They can occasionally be associated with the presence of leucocyte (HLA and HNA) alloantibodies in the recipient. In the UK, where there is universal leucocyte reduction of blood components, investigations for other causes of high fever associated with transfusion are carried out, e.g. tests for bacterial contamination and IgA-deficiency. Serological investigations for platelet, HLA and granulocyte antibodies are of limited clinical value as the diagnostic specificity of these tests for FNHTRs is low. Nonetheless, testing for HNA antibodies may be required in rare cases in which a severe FNHTR cannot be otherwise explained and washed components have proved ineffective. The management of FNHTRs is described in detail in Chapter 8.
TRALI is a severe and sometimes life-threatening transfusion reaction. The majority of cases are caused by donor leucocyte alloantibodies against alloantigens present on the patient's leucocytes, although patient alloantibodies may be involved in some cases. In most TRALI cases, HLA class I- and II-specific antibodies are implicated but HNA antibodies have also been implicated as causal agents with HNA-1a and HNA-3a antibody specificities being found most commonly . TRALI investigations are logistically complex because of the need to contact all of the implicated donors to obtain fresh blood samples. Investigations are usually stratified to include only female donors initially. The donor samples are screened for both HLA and HNA alloantibodies. If antibodies are found it is necessary to type the patient to determine whether they have the cognate antigen and to type the donor to establish that they lack the antigen. In some cases, it may be necessary to screen a recipient's serum for antibodies or to perform a cross-match between donor sera and the patient's granulocytes and lymphocytes.
Many blood transfusion services have taken steps to reduce the incidence of TRALI, e.g. by reducing the proportion of female donors for plasma and platelet components, and more recently by screening female donors for HLA and HNA antibodies. The success of these strategies has been demonstrated by the reduced incidence of TRALI in haemovigilance schemes.
Transfusion-related alloimmune neutropenia (TRAIN)
The first documented case of TRAIN occurred following the infusion of 80 mL of plasma-reduced blood after cardiac surgery on a 4-week-old infant . The plasma from the blood donor was found to contain HNA-1b alloantibodies that resulted in an absolute neutropenia in the infant, who typed as HNA-1a(+), -1b(+). The neutropenia resolved after 7 days after treating the infant with GCSF. The condition is of interest since it demonstrates that, in some circumstances, passively infused HNA antibodies can trigger neutropenia rather than TRALI. This clinical entity has recently been confirmed by an additional report.
Autoimmune neutropenia is a rare condition that can occur as a transient, self-limiting autoimmunity in young children  or a chronic form in adults . The autoantibodies tend to target the FcγRIIIb (CD16), CD177 or CD11/18 molecules but can also be HNA-specific, e.g. HNA-1a antibodies are found in autoimmune neutropenia of childhood.
The most sensitive method for the detection of autoantibodies is to test the patient's neutrophils using the direct immunofluorescence test. However, the combination of severe neutropenia, high blood sample volume requirements to recover sufficient granulocytes and the need for a fresh sample limits the applicability of this test, especially in children. Screening of a patient's serum with a panel of typed neutrophils in the indirect granulocyte immunofluorescence and granulocyte chemiluminescence or granulocyte agglutination tests provides a suitable alternative and, in some studies, this approach has been found to be only slightly less sensitive than performing a direct test.
Persistent post-bone marrow transplant neutropenia
Antibody-mediated neutropenia may be a serious complication of bone marrow transplantation. In this context, as the neutrophil antibodies may be autoimmune and/or alloimmune in nature, laboratory investigation requires serological and typing studies to elucidate the nature of the antibodies involved.
1. Allo-, auto-, iso- and drug-induced antigens may be found on platelets and neutrophils and are implicated in a range of immune cytopenias.
2. Alloantigens on platelets are known as HPAs; alloantigens on neutrophils are known as HNAs.
3. Reliable detection and identification of HPA- and HNA-specific antibodies requires the use of both whole-cell type assays such as the PIFT/GIFT and antigen-capture type assays such as the MAIPA/MAIGA assays.
4. HPA and HNA types can mostly be determined using PCR-based methodologies.
5. NAIT is a common disorder and HPA-1a or HPA-5b antibodies are responsible for approximately 95% of cases.
6. Optimal postnatal treatment of NAIT is the transfusion of HPA-1a/5b-negative donor platelets.
7. Optimal antenatal treatment of NAIT is yet to be determined but maternal treatment with IvIgG is the recommended initial treatment.
8. HNA antibodies can be associated both with alloimmune and autoimmune neutropenia.
9. TRALI can be a life-threatening condition, especially if HNA-3a antibodies are involved.
10. The incidence of antibody-mediated TRALI has been reduced by implementation of a number of different strategies.
1. Metcalfe P, Watkins NA, Ouwehand WH, Kaplan C, Newman P, Kekomaki R et al. Nomenclature of human platelet antigens. Vox Sanguinis 2003 October; 85(3): 240–245.
2. Aster RH, Curtis BR, McFarland JG & Bougie DW. Drug-induced immune thrombocytopenia: pathogenesis, diagnosis, and management. J Thromb Haemost 2009 June; 7(6): 911–918.
3. Kiefel V, Santoso S, Weisheit M & Mueller-Eckhardt C. Monoclonal antibody-specific immobilization of platelet antigens (MAIPA): a new tool for the identification of platelet-reactive antibodies. Blood 1987 December; 70(6): 1722–1726.
4. Tiller H, Killie M, Skogen B, Oian P & Husebekk A. Neonatal alloimmune thrombocytopenia in Norway: poor detection rate with nonscreening versus a general screening programme. BJOG 2009 March; 116(4): 594–598.
5. Ghevaert C, Rankin A, Huiskes E, Porcelijn L, Javela K, Kekomaki R et al. Alloantibodies against low-frequency human platelet antigens do not account for a significant proportion of cases of fetomaternal alloimmune thrombocytopenia: evidence from 1054 cases. Transfusion 2009 October; 49(10): 2084–2089.
6. Kamphuis MM & Oepkes D. Fetal and neonatal alloimmune thrombocytopenia: prenatal interventions. Prenat Diagn 2011 July; 31(7): 712–719.
7. Bertrand G, Drame M, Martageix C & Kaplan C. Prediction of the fetal status in noninvasive management of alloimmune thrombocytopenia. Blood 2011 March 17; 117(11): 3209–3213.
8. Killie MK, Husebekk A, Kjeldsen-Kragh J & Skogen B. A prospective study of maternal anti-HPA 1a antibody level as a potential predictor of alloimmune thrombocytopenia in the newborn. Haematologica 2008 April 28; 93(6): 870–877.
9. Ghevaert C, Campbell K, Stafford P, Metcalfe P, Casbard A, Smith GA et al. HPA-1a antibody potency and bioactivity do not predict severity of fetomaternal alloimmune thrombocytopenia. Transfusion 2007 July; 47(7): 1296–1305.
10. Bux J. Nomenclature of granulocyte alloantigens. ISBT Working Party on Platelet and Granulocyte Serology, Granulocyte Antigen Working Party. International Society of Blood Transfusion. Transfusion 1999 June; 39(6): 662–663.
11. Lucas G, Win N, Calvert A, Green A, Griffin E, Bendukidze N et al. Reducing the incidence of TRALI in the UK: the results of screening for donor leucocyte antibodies and the development of national guidelines. Vox Sanguinis 2012 July; 103(1): 10--17.
12. Bux J & Sachs UJ. The pathogenesis of transfusion-related acute lung injury (TRALI). Br J Haematol 2007 March; 136(6): 788–799.
13. Wallis JP, Haynes S, Stark G, Green FA, Lucas GF & Chapman CE. Transfusion-related alloimmune neutropenia: an undescribed complication of blood transfusion. The Lancet 2002 October 5; 360(9339): 1073–1074.
14. Bruin M, Dassen A, Pajkrt D, Buddelmeyer L, Kuijpers T & de Haas M. Primary autoimmune neutropenia in children: a study of neutrophil antibodies and clinical course. Vox Sanguinis 2005 January; 88(1): 52–59.
15. Shastri KA & Logue GL. Autoimmune neutropenia. Blood 1993 April 15; 81(8): 1984–1995.
Bassler D, Greinacher A, Okascharoen C et al. A systematic review and survey of the management of unexpected neonatal alloimmune thrombocytopenia. Transfusion 2008; 48: 92–98.
Bux J. Human neutrophil alloantigens. Vox Sanguinis 2008; 94: 277–285.
Capsoni F, Sarzi-Puttini P & Zanella A. Primary and secondary autoimmune neutropenia. Arthritis Res Therapy 2005; 7: 208–214.
Fung YL, Minchinton RM & Fraser JF. Neutrophil antibody diagnostics and screening: review of the classical versus the emerging. Vox Sanguinis 2011; 101: 282–290.
Lucas GF & Metcalfe P. Platelet and granulocyte glycoprotein polymorphisms. Transfus Med 2000; 10: 157–174.
Murphy MF & Bussel JB. Advances in the management of alloimmune thrombocytopenia. Br J Haematol 2007; 136: 366–378.
Ouwehand WH, Stafford P, Ghevaert C et al. Platelet immunology, present and future. ISBT Sci Ser 2006; 1: 96–102.
Warkentin TE & Smith JW. The alloimmune thrombocytopenic syndromes. Transfus Med Rev 1997; 11: 296–307.