- Top of page
- Patients and methods
Summary. Fcγ receptor-mediated destruction of autoantibody-sensitized platelets is central to the immune pathophysiology of childhood immune thrombocytopenic purpura (ITP). Allelic variants exist among the random population for some Fcγ receptors. The variants represent single nucleotide polymorphisms, leading to functional differences in the ability to bind immunoglobulin (Ig)G or IgG subclasses. The genotypic frequencies for two Fcγ receptor single nucleotide polymorphisms, FcγRIIa-131 arginine (R) versus histidine (H) and FcγRIIIa-158 valine (V) versus phenylalanine (F) were examined in 98 children diagnosed with childhood ITP. The genotype frequencies were compared with those of 130 healthy control subjects. Chi-square analysis was used to determine whether the allelic frequencies of the high-affinity receptor variants were associated with childhood ITP. Both the FcγRIIa-131H and the FcγRIIIa-158V were significantly over-represented in children with ITP versus the control subjects (P-values 0·03). The same statistical difference was noted with the combined FcγRIIa-131H and FcγRIIIa-158V allelic gene frequencies. There was no statistical difference between children who later developed chronic ITP compared with children with acute ITP, suggesting that additional factors are responsible for the development of the chronic form of the disease. These observations underscore the importance of Fcγ receptor-mediated cell clearance in childhood ITP.
Immune thrombocytopenic purpura (ITP) in children is a relatively common autoimmune haematolgical disorder presenting at a peak age of 4–6 years (Schulman, 1964). The thrombocytopenic episode is usually acute and primary in origin, i.e. not secondary to other immunological disorders such as systemic lupus erythematosus. Childhood ITP is often preceded by an infectious illness, most often a viral upper respiratory tract infection, a few weeks prior to the onset of clinical symptoms. Thrombocytopenia (platelet count < 150 × 109/l) resolves within 6 months of onset in over 80% of children with primary acute ITP (Buchanan, 1987); persistence of thrombocytopenia for longer than 6 months defines the chronic form of the disorder.
The immune pathophysiology of ITP has the following characteristic: antiplatelet autoantibodies, usually of the immunoglobulin (Ig)G class, can be demonstrated on the platelet surface and in the circulation. The direct platelet antibody test is reported to be positive in approximately 70–80% of adults with ITP and similarly in children with chronic ITP (Berchtold et al, 1989). Although not diagnostic of ITP, these autoantibodies may target specific glycoproteins (GPs) on the platelet membrane, e.g. GPIIb/IIIa, GPIa/IIa and GPIb/IX (Szatkowski et al, 1986; Hagenstrom et al, 2000). Reticuloendothelial cells, mainly macrophages found principally in the spleen, bearing receptors for the Fc portion of IgG (FcγRs), clear sensitized platelets from the circulation. Thus, the expression of platelet autoantibodies and Fcγ receptor function are central to the disease process.
The human FcγR family consists of three major classes of receptors encompassing 12 transcripts derived from eight genes: FcγRIa, Ib and Ic; FcγRIIa, IIb and IIc; FcγRIIIa and FcγRIIIb (Van de Winkel & Capel, 1993), all of which are present on chromosome 1. These receptors utilize an intracellular tyrosine-based activation motif (ITAM) for signal transduction with the exception of FcγRIIb, which expresses an intracellular tyrosine-based inhibition motif (ITIM). FcγRI is a high-affinity receptor for monomeric IgG. FcγRIIa and FcγRIIIa are low-affinity receptors that interact only with complexed or multimeric IgG. Both FcγRIIa and FcγRIIIa can bind IgG1 and IgG3 while FcγRIIa is the sole FcγR that can bind IgG2 with high affinity (Parren et al, 1992). Heterogeneity is further increased among FcγRIIa, IIIa and IIIb isoforms by inherited functional single nucleotide polymorphisms. These allelic variants are distributed randomly among populations and their distribution varies between different ethnic groups (Joutsi et al, 1998). They are co-dominantly inherited according to classical Mendelian genetics.
The FcγRIIa genetic polymorphism results from a single nucleotide substitution (A or G) which codes for the amino acids histidine (H) or arginine (R), respectively, at position 131. This amino acid substitution is associated with a markedly altered ability of the receptor to bind human IgG2 (Warmerdam et al, 1991). The FcγRIIa-131H allele has a higher binding affinity for IgG2 than FcγRIIa-131R. Small but significant differences in the interaction of these variants with IgG1 and IgG3 also exist (Parren et al, 1992; Denomme et al, 1997). Similarly for FcγRIIIa, valine (V) or phenylalanine (F) at codon 158 alters the affinity of the receptor for IgG1 and IgG3. The FcγRIIIa-158V form, expressed on natural killer (NK) cells and monocytes, binds IgG1 and IgG3 much more efficiently when compared with FcγRIIIa-158F (Koene et al, 1997; Wu et al, 1997). Inheritance of higher or lower affinity Fcγ receptor alleles may predispose individuals to Fcγ-dependent immune disorders (Bredius et al, 1994a, 1994b; Sanders et al, 1994; Salmon et al, 1996, 1999).
Identification and recruitment of patients. Patients were identified by a retrospective review of health records (coded by discharge diagnosis), from laboratory bone marrow aspiration reports and by identifying patients followed for ITP by haematologists at The Hospital for Sick Children, Toronto, Canada. All patients were eligible for study regardless of the type of ITP, i.e. acute or chronic. Patients diagnosed with ITP at less than 6 months of age were considered ineligible for this study, as the diagnosis of ITP in this age group is often questionable. Patients were contacted by telephone by one of the investigators and invited to participate in the study. Informed consent was obtained during the subsequent hospital visit and prior to venepuncture. Approximately half of all potential ITP patients were recruited into the study.
Patient information obtained via chart review included patient age at diagnosis, sex, type of ITP (acute or chronic) and history of splenectomy. Patients were categorized as having acute or chronic ITP based on the duration of thrombocytopenia. Chronic ITP was defined as thrombocytopenia (platelet count < 150 × 109/l) persisting for longer than 6 months.
- Top of page
- Patients and methods
A total of 98 children were evaluated. Mean and median ages at presentation of ITP were 7·1 years and 5·6 years respectively. Twenty-seven patients had acute ITP (ITP that had resolved within 6 months of presentation) and 71 had chronic ITP. Age at presentation was significantly different (P = 0·003) between patients who had acute ITP (mean age at presentation: 4·8 years) and those with chronic ITP (mean age at presentation: 8·0 years). Eighteen (25%) of the chronic ITP patients had undergone splenectomy.
The FcγRIIa-131H and -131R allelic gene frequencies, calculated from the numbers for each genotype among the control subjects, were 0·496 and 0·504 respectively. The gene frequencies for the control group approximated those previously reported (Denomme et al, 1997; Joutsi et al, 1998; Foster et al, 2001). The FcγRIIa-131H and -131R allelic gene frequencies for ITP patients were 0·600 [95% confidence interval (CI) = 0·532–0·667] and 0·400 (95% CI = 0·330–0·469). The frequency of the three genotypic classes for ITP patients differed significantly from the control group (P-value = 0·03), with an over-representation of the homozygous FcγRIIa-131H/H genotype among ITP patients compared with the control subjects (Table II).
Table II. FcγRIIa phenotypes based on the single nucleotide polymorphism. (A) The number of control subjects and ITP patients for each of the FcγRIIa phenotypes and their frequencies (%). : (B) The FcγRIIa allele frequencies for the con- trol subjects and ITP patients.
| ||131H/H (n = 67)||131H/R (n = 110)||131R/R (n = 49)||Total (n = 226)|
|Control subjects||n = 30||n = 69||n = 31||n = 130|
|ITP patients||n = 37||n = 41||n = 18||n = 96*|
For FcγRIIIa-158F and -158V, the allelic gene frequencies for the control group were 0·690 vs 0·310 respectively. For ITP patients, the FcγRIIIa-158F and -158V allelic gene frequencies were 0·577 (95% CI = 0·509–0·644) and 0·423 (95% CI = 0·353–0·492) respectively. The genotypic classes for ITP patients differed significantly from the control group (P-value = 0·03), with the largest difference noted in the frequency of the homozygous FcγRIIIa-158F/F group. The FcγRIIIa-158F versus -158V genotypic frequencies for ITP patients showed a significant trend towards an increased prevalence of the FcγRIIIa-158V allele (Table III).
Table III. FcγRIIIa phenotypes based on the single nucleotide polymorphism. (A) The number of control subjects and ITP patients for each of the FcγRIIIa phenotypes and their frequencies (%). : (B) The FcγRIIIa allele frequencies for the con- trol subjects and ITP patients.
| ||158V/V (n = 28)||158V/F (n = 106)||158F/F (n = 92)||Total (n = 226)|
|Control subjects||n = 13||n = 54||n = 62||n = 129|
|ITP patients||n = 15||n = 52||n = 30||n = 97*|
Recognizing that the combination of different polymorphic variants may have a synergistic effect, we compared patients that had the combination of at least one FcγRIIa-131H allele and one FcγRIIIa-158V allele (the alleles with the higher affinities for certain immunoglobulin subclasses) (Table IV). The combination of the FcγRIIa-131H and FcγRIIIa-158V alleles was found to be more common in ITP patients than in control subjects (61%vs 46%; P-value = 0·02). The corollary to this was that the combination of the lower affinity alleles (FcγRIIa-131R and FcγRIIIa-158F) was less common in ITP patients than in control subjects (54%vs 70%; P-value = 0·02).
Table IV. Combined FcγRIIa and FcγRIIIa phenotypes. (A) FcγRIIa-131H and FcγRIIIa-158V (high IgG binding) allele frequencies. : (B) FcγRIIa-131R and FcγRIIIa-158F (low IgG binding) allele frequencies.
| ||131H & 158V* (n = 118)||No 131H & 158V (n = 107)||Total (n = 225)|
|Control subjects||n = 59||n = 70||n = 129|
|ITP patients||n = 59||n = 37||n = 96|
| ||131R & 158F* (n = 142)||No 131R & 158F (n = 83)||Total (n = 225)|
|Control subjects||n = 90||n = 39||n = 129|
|ITP patients||n = 52||n = 44||n = 96|
Lastly, patients were subdivided into those with acute versus chronic ITP. We compared the distribution of polymorphisms in these two groups and found no statistical difference (data not shown). The distribution of polymorphisms (either FcγRIIa or FcγRIIIa) did not correlate with age at presentation or sex.
- Top of page
- Patients and methods
Primary immune system alterations or dysfunction are characteristic of ITP (Semple & Freedman, 1991; Semple et al, 1996). In this disorder, thrombocytopenia results from the accelerated clearance of antibody-coated platelets by FcγR-bearing cells of the RES. For the most part, FcγR polymorphic variants have different affinities for immunoglobulins, resulting in variable clearance of immune complexes which may contribute to an increased or decreased susceptibility to certain disorders (Van der Pol & Van de Winkel, 1998). The association of FcγR polymorphisms and functional variation was first observed for the binding of murine IgG1 antibodies to the FcγRIIa-131H variant (Tax et al, 1983). Following this report, a number of associations between the FcγRIIa functional polymorphisms and diseases have been found (Bredius et al, 1994a, 1994b; Sanders et al, 1994; Denomme et al, 1997; Carlsson et al, 1998; Sugita et al, 1999). Taken together, FcγR polymorphisms may contribute not only to disease occurrence, but to severity and clinical presentation as well.
The mechanism underlying the association of a particular FcγR variant to the pathophysiology of an autoimmune disease is more difficult to ascertain. Patients with systemic lupus erythematosus have an increased frequency of the lower affinity FcγRIIa-131R variant. One explanation for this observation is that the presence of the FcγRIIa-131R variant results in less efficient immune complex clearance, thereby predisposing patients to a greater risk of immune complex complications (Sanders et al, 1994; Koene et al, 1997; Wu et al, 1997; Dijstelbloem et al, 1999).
In this study, we found that the genotypic frequencies for FcγRIIa-131H and -131R, and FcγRIIIa-158F and -158V in children with ITP were significantly different to control populations, suggesting that the inheritance of these receptor variants may convey a small but significant risk to the development of childhood ITP. Moreover, the combination of the higher affinity alleles (FcγRIIa-131H and FcγRIIIa-158V) was over-represented in children with ITP as compared with control subjects. It is important to note that our results reveal only a trend, with ITP patients having slightly, and statistically significant, higher genotypic frequencies for the FcγR variants with the higher affinities for IgG. Also, allelic frequencies were determined among patients who presented with thrombocytopenia. It is uncertain whether individuals with low-affinity genotypes develop a subclinical form of the disease.
Our findings corroborate those of Foster et al (2001), which involved a much smaller cohort of children with ITP. The genotype distribution among the control subjects in their study (FcγRIIa-131R/R:H/R:H/H, 25%:44%:31% and FcγRIIIa-158F/F:V/F:V/V, 50%:39%:11%) was similar to our findings. Also, the genotypic frequencies in their 37 children with early chronic ITP (FcγRIIa-131R/R:H/R:H/H, 19%:36%:44% and FcγRIIIa-158F/F:V/F:V/V, 35%:62%:3%) were comparable to the results obtained in our study. Our study results, involving a much larger number of children with ITP, lend further strength to the argument that FcγRIIa and IIIa polymorphisms contribute to the pathogenesis of childhood ITP both individually and in combination.
Our results are in contrast to those reported by Williams et al (1998) and Joutsi et al (1998). Williams et al (1998) found that the FcγRIIa-131R allele was predominant in a group of 27 adults with refractory idiopathic (immune) thrombocytopenic purpura. They suggested that the decreased affinity of FcγRIIa-131R for IgG2 results in a reduced or incomplete clearance of antibody–infectious agent complexes, leading to a predisposition to develop ITP. One explanation for the difference in findings between our results and those obtained by Williams et al (1998) is that the two populations studied are different. We studied children with ITP whereas their study involved adults. The two conditions may represent different disorders. Furthermore, the study by Williams et al (1998) involved adults with severe refractory ITP whereas our study involved children with all forms of ITP (acute and chronic). Nevertheless, no significant differences in genotype distribution were detected when we subdivided our patients between those with acute and chronic ITP.
Joutsi et al (1998) genotyped for the FcγRIIa-131H and -131R variants among adult patients with thrombocytopenia and found no difference in their genotype distributions compared with the control population. Differences in genotype distribution were apparent in subgroups of patients with ITP, for example younger women with ITP tended to have a predominance of the FcγRIIa-131R allele.
Two possible mechanisms may underlie the reason for the over-representation of FcγRIIa-131H and FcγRIIIa-158V in childhood ITP as seen in our study. An over-representation of the high IgG binding alleles might result in increased clearance of IgG-sensitized platelets. It is interesting to note that viral infection often precedes childhood ITP and autoantibody production has been demonstrated in the sera of some individuals during the antiviral immune response (Winiarski, 1990; Nieminen et al, 1993). Moreover, some antiviral antibodies have been shown to cross-react with platelets (Wright et al, 1996). Thus, more efficient clearance of cross-reactive antibody-sensitized platelets by the higher binding FcγRs may lead to enhanced platelet antigen presentation by major histocompatibility complex (MHC) class II cells. FcγR frequencies determined from a large cohort of children with ITP who have confirmed evidence of recent viral infection or vaccination may provide evidence for this hypothesis. Alternatively, platelet consumption could increase as a result of viral immune complex binding to the higher affinity FcγRs on platelets, leading to increased self-antigen presentation again by phagocytic cells bearing MHC class II. Both processes are not mutually exclusive. Childhood ITP probably results from a spectrum of polymorphic susceptibility factors, including FcγR functional variants.
Our study suffered some limitations. Given the case–control nature of this study, there was an over-representation of chronic cases as these children are more likely to be followed and be available for study. In fact, the immunological dysfunction leading to chronic ITP in children may be similar to the adult form of the disease. If this is the case, an over-representation of chronic ITP could potentially bias our statistical analysis. Whether the distribution of FcγR polymorphisms are different between acute versus chronic ITP suffers from the small number of children recruited with acute ITP (n = 27). A large prospective study of polymorphic immune response elements in children with ITP is needed to determine whether the outcome or management of the disorder can benefit from the knowledge of these functional variants.