Summary
Immune thrombocytopenic purpura (ITP) is an autoimmune thrombocytopenia with shortened platelet survival and relative bone marrow failure. The pathogenesis involves antibody production, cytokine release, T cell impairment, complement activation and clearance of platelets. We measured plasma levels of C3, C4, C1q and sC5b‐9 in 80 ITP patients in acute phase, 50 ITP patients in complete (CR) or partial (PR) remission and 50 age‐ and sex‐matched healthy volunteers. Statistical analyses showed that acute ITP patients had higher plasma levels of sC5b‐9 and C1q than CR or PR patients (median = sC5b‐9: 200 versus 98 mg/dl, P‐value < 0·001) (median C1q = 2·11 versus 1·00 mg/dl, P‐value < 0·001). CR and PR ITP patients had sC5b‐9 and C1q plasma levels comparable to those observed in healthy volunteers. There was a significant correlation between sC5b‐9 and C1q plasma levels (Spearman’s rho correlation index on 130 ITP patients equal to 0·58, P‐value < 0·001). We also found that sC5b‐9 plasma level is inversely correlated with the number of platelets. Furthermore, we divided acute ITP patients into subjects with detectable (24 of 80, 30%) or undetectable (56 of 80, 70%) anti‐platelet antibodies; patients with detectable anti‐platelet antibodies have significantly higher plasma levels of C1q and sC5b‐9. This research will potentially offer novel therapeutic strategies in light of new drugs affecting complement activation for monitoring therapy response.
Keywords: Anti‐platelet antibody, C1q, complement system, immunoglobulin, ITP, sC5b‐9, thrombocytopenia
Immune Thrombocytopenic Purpura (ITP) is an autoimmune thrombocytopenia with a shortened platelet survival and relative bone marrow failure. We measured plasma levels of C3, C4, C1q, sC5b‐9 in 80 ITP patients in acute phase, 50 ITP patients in complete (CR) or partial (PR) remission, and 50 age‐ and sex‐matched healthy volunteers. Statistical analyses showed that acute ITP patients had higher plasma levels of sC5b‐9 and C1q than CR or PR patients We also found that sC5b‐9 plasma level is inversely correlated with the number of platelets.
Introduction
Immune thrombocytopenic purpura (ITP) is an acquired autoimmune bleeding disorder in which a variable peripheral platelet destruction is combined with impaired platelet production [1, 2].
Bleeding complications in patients with ITP range from skin bruises to life‐threatening complications such as intracranial hemorrhages (ICH) [3]. Severe bleeding is uncommon when the platelet count is > 30 × 109/l, and usually occurs only when the platelet count drops to < 10 × 109/l.
ITP is classified as primary when there is no evidence of any systemic disease (80% of all cases) and as secondary when associated with an infection such as human immunodeficiency virus, cytomegalovirus, hepatitis C virus or Helicobacter pylori, or with an autoimmune or lymphoproliferative disease [1, 2]. The etiology of ITP remains unknown and the diagnosis is based on exclusion of other causes of thrombocytopenia [2]. Thrombocytopenia lasting longer than 12 months is considered chronic ITP. Among these patients, the 1‐year risks of cardiovascular complications range from 1 to 2%, while nearly 8% of ITP patients experience a bleeding episode within 1 year [4].
The immune mechanisms underlying the shortened platelet survival and the reduced platelet production are complex, involving not only an antibody‐mediated clearance of platelets in the reticulum–endothelial system, but also impaired T cell immunity [5], complement activation and the clearance of platelets opsonized by immunoglobulins [5, 6]. Although the initial event inducing the production of anti‐platelet antibodies remains a matter of debate, platelet autoantibodies are often detected at the time of diagnosis. Antibodies induce opsonization of antibody–platelet complexes by antigen‐presenting cells (APCs), facilitating intracellular processing of platelets and presenting to T cells via major the histocompatibility complex (MHC)‐II as an array of ‘foreign’ platelet peptides. The presentation of platelet peptides by MHC‐II activates T cells. These events lead to enhancement of the anti‐platelet immune response and the possibility of epitope spread to additional platelet antigens, macrophages and dendritic cells of the reticuloendothelial system phagocyte antibody‐targeted platelets [5].
In patients with ITP, autoantibodies are frequently directed against GpIb/IX and GPIIb/IIIa, but specificity for other platelet or megakaryocyte antigens has been described [2, 4]. Although anti‐platelet autoantibodies appear to play a pivotal role in the pathogenesis of ITP, they may be undetectable in some patients at the time of diagnosis; many factors, including limitations of laboratory methods and biology of ITP, may be the cause [5]. Factors include: (1) the detection limit of anti‐platelet antibody titers may be reduced below its limit by the clearance of some types of antibody–platelet complexes; (2) tightly bound anti‐platelet antibodies may be difficult to dissociate; (3) antibodies directed against minor or cryptic platelet antigen or against megakaryocytes antigens may be missed; and (4) finally, there is a subset of patients in which anti‐platelet antibodies are lacking.
The role of complement in ITP is poorly understood. The complement system is involved in various diseases associated with platelet dysregulation and micro thrombosis, e.g. hemolytic uremic syndrome (HUS) [6, 7, 8]. Interestingly, in HUS renal dysfunction is secondary to microthrombosis in the renal vascular system, and mutations in C3 were shown to predispose to development of atypical hemolytic uremic syndrome (aHUS) [8, 9, 10, 11]. Complement receptors for C3a and C5a are shown to be expressed on platelets [12]. Recently, it has been observed that platelets express binding sites for classical complement components, especially C1q; inducing GPIIB‐IIIa‐mediated platelet aggregation and plasma from ITP patients was shown to be capable of fixing complement to platelets in vitro [10, 11, 12, 13].
The aim of this study is to evaluate whether or not complement activation is involved in the pathogenesis of ITP. Secondary end‐points are to evaluate whether complement activation is major during the acute phase of ITP compared to complete (CR) or partial relapse (PR) phases of the disease course.
Materials and methods
Study protocol and design
The institutional review board of our University Hospital (Luigi Sacco Hospital, University of Milan, Italy) approved the study protocol. All enrolled patients expressed and signed an informed consent. From March 2015 to August 2019, patients with a diagnosis of ITP were screened. Eighteen newly diagnosed acute‐phase ITP patients and 50 ITP patients in CR or PR fulfilled the inclusion criteria, were enrolled into the study and underwent blood sampling for the following tests: anti‐nuclear antibody, anti‐cardiolipin antibody, thyrotropin, quantitative immunoglobulins, serum electrophoresis, hepatitis B and C screening and serological testing of adults for Helicobacter pylori. The diagnosis of ITP was based on: (1) clinical evaluation of history of isolated bleeding symptoms consistent with thrombocytopenia without constitutional symptoms (e.g. weight loss, bone pain, night sweats), bleeding symptoms and the absence of hepato‐splenomegaly or lymphadenopathy; and (2) a complete blood count characterized by isolated thrombocytopenia with normal red blood cells and a white blood cell count. Bone marrow examination was carried out only in patients with abnormalities in blood count and differential count or in patients with unexplained concomitant anemia or leukopenia.
As defined by an international working group reporting on the standardization of terminology in ITP, all patients included in the analysis were considered to have chronic ITP if thrombocytopenia lasted for 12 months or more [5]. ITP patients were divided into subgroups as follows: patients with acute‐phase ITP and patients with CR or PR ITP, according to the response criteria of Rodeghiero et al. [14].
Exclusion criteria were severe cognitive impairment, previous history of autoimmune diseases, hepatitis C virus, cytomegalovirus immunoglobulin (Ig)M antibody positivity or human immunodeficiency virus infection.
In all ITP patients we evaluated C1q, C4, C3 and sC5b‐9 plasma levels and the presence of anti‐platelet antibodies. Fifty age‐ and sex‐matched healthy volunteers served as the control group.
Among patients with acute‐phase ITP, we identified patients in whom anti‐platelet antibodies were detectable and compared complement parameters between this subgroup of patients and the subgroup with undetectable autoantibodies.
Bleeding complications were defined by the presence of bleeding manifestations at presentation requiring treatment or by the occurrence of new bleeding symptoms needing additional therapeutic interventions with a different platelet‐enhancing agent or an increased dose [15].
Laboratory methods
Blood samples were obtained by clean venipuncture and minimal stasis. Plasma for the measurement of antibody‐platelets, C1q, C4, C3 and sC5b‐9 was collected in ethylenediamine tetraacetic acid (EDTA), with a final concentration of 20 mM. Blood samples were centrifuged within 1 h at 2000 g and 20°C for 20 min. The plasma was collected, divided into aliquots and stored at −80°C until tested.
C1q, C4 and C3 plasma levels were measured by radial immunodiffusion (RID) (NOR‐Partigen; Siemens Healthcare Diagnostics, Munich, Germany) and sC5b‐9 by EIA (Quidel Corporation, San Diego, CA, USA); the required sample dilution for the latter test is 1 : 10 [normal range = 75–219 ± 2 standard deviations (s.d.)]. A dot‐plot with distribution of results of sC5b9 is provided in Fig. 3.
Fig. 3.
Scatter‐plot of sC5b‐9 plasma levels versus platelet counts, respectively, in acute‐phase immune thrombocytopenic purpura (ITP) (circles labeled ‘1’) and in complete remission/partial remission (CR/PR) ITP patients (circles labeled ‘2’).
Anti‐platelet autoantibodies were evaluated on EDTA‐anti‐coagulated plasma of 130 ITP patients using monoclonal antibody‐specific immobilization of platelet antigens (MAIPA) assay, according to Kiefel et al [16]. Briefly, antibodies were searched using platelets from a pool of group O donors selected according to platelet genotype (Advanced Practical Diagnostics, BVBA, Turnhout, Belgium). Platelets were incubated with serum and mouse monoclonal antibodies specific for Ia/IIa, Ib/IX and IIb/IIIa. Lysates were transferred to microplate wells precoated with goat anti‐mouse IgG. The bound complex was detected using goat peroxidase‐coupled anti‐human IgG and revealed by peroxidase substrate O‐phenylenediamine. All positive antibody screens were identified using a standard six‐cell genotyped panel. A subject was considered to have anti‐platelet antibodies if one or more platelet anti‐glycoproteins were identified.
Statistical analysis
Ordinal data were compared using the t‐test. The number of subjects for each group was 50 or more, so it was plausible using the random variable T with (n–1) degree of freedom as an approximation to the normal distribution (Central Limit Theorem). The bi‐directional t‐test for independent samples of data was performed to verify if the previously noted differences between the groups of patients effectively subsisted. Nominal data (positive/negative) were analyzed using the two‐tailed Fisher’s exact test.
Spearman’s rho coefficient was used as a non‐parametric statistical correlation measure, which quantifies the degree of relationship between two variables (a number between 0 and 1) for which there is no other hypothesis of the ordinal measure, but is possibly continuous. Unlike the linear Pearson’s correlation coefficient, Spearman’s coefficient does not measure a linear relationship even when interval measurements are used.
Multivariate analysis was used to evaluate the relationship between complement activation and platelet count. P‐values < 0·05 were considered statistically significant.
Results
We enrolled 130 patients with a diagnosis of ITP. Eighty patients had acute‐phase ITP and 50 patients had ITP in CR or PR, according to Rodeghiero et al. [14]. The clinical and demographic characteristics of ITP patients and controls are shown in Table 1. Patients with ITP ranged in age from 18 to 53 years; the mean patient age was 51·2 ± 2 years. The gender distribution in the ITP group was 60% female and 40% male. Platelet count ranged from 9000 to 50 000 and from 100 000 to 120 000 in acute‐phase ITP patients and in CR or PR ITP patients, respectively.
Table 1.
Study population and treatment.
| ITP patients in acute phase (n = 80) | ITP patients in remission (n = 50) | |
|---|---|---|
| Demographic and clinical information | ||
| Age in year ± s.d. (range) | 51·2 ± 2 (18–53) | 50 ± 3 (65–83) |
| Female, n (%), male, n (%) | 48 (60%), 32 (40%) | 21 (40%), 29 (60%) |
| Anti‐platelet antibodies, n (%) | 33 (42%) | 15 (30%) |
| Platelet count × 103/mm3 (median) | 9–92 (24) | 100–120 (26) |
| Therapeutics | ||
| Steroids | 80 (100%) | 50 (100%) |
| Intravenous immunoglobulins (IVIG) | 20 (25%) | 25 (5%) |
| Splenectomy | 5 (6·25%) | 1 (2%) |
| Rituximab | 10 (12·5%) | 29 (40%) |
| Thrombopoietin analogues | 10 (12·5%) | 2 (4%) |
| Anti‐D Ig | 1 (1·25%) | 1 (2%) |
| Danazol | 2 (3·5%) | 1 (2%) |
| Vincristine | 0 | 0 |
| Azathioprine | 0 | 0 |
| Comorbidities | ||
| Hypertension | 15 (19%) | 29 (40%) |
| Diabetes mellitus | 10 (6·25%) | 10 (20%) |
| Ischemic heart disease | 5 (3·1%) | 11 (22%) |
ITP = immune thrombocytopenic purpura; s.d. = standard deviation.
ITP patients were treated with steroids, intravenous immunoglobulins, anti‐D immunoglobulins, rituximab, splenectomy, thrombopoietin analogues, danazol and vincristine.
Tables 2 and 3 present complement parameters in the study population.
Table 2.
Levels of C3, C4, sC5b‐9 and C1q (mean value ± s.d.) in, respectively, ITP patients in acute phase, ITP patients in remission (CR/PR) and healthy volunteers
| ITP patients in acute phase (n = 80)(a) | ITP patients in CR/PR (n = 50)(b) | Healthy volunteers (n = 50)(c) | Statistical significance | |||
|---|---|---|---|---|---|---|
| (a) versus (b) | (a) versus (c) | (b) versus (c) | ||||
| C3 % of pooled plasma | 99·03 ± 6·1 | 96·42 ± 17·9 | 102·94 ± 4·6 | > 0·05 | > 0·05 | > 0·05 |
| C4 % of pooled plasma | 92·15 ± 3·1 | 89·90 ± 1·9 | 89·96 ± 3·7 | > 0·05 | > 0·05 | > 0·05 |
| sC5b‐9 level (ng/ml) | 206·45 ± 20·9 | 101·06 ± 7·1 | 103·04 ± 6·1 | < 0·001 | < 0·001 | > 0·05 |
| C1q level (g/l) | 2·11 ± 1·9 | 1·0 ± 0·1 | 1·0 ± 0·1 | < 0·001 | < 0·001 | > 0·05 |
ITP = immune thrombocytopenic purpura; CR/PR = complete remission/partial remission.
Table 3.
Plasma levels of C3, C4, sC5b‐9 and C1q (mean value ± s.d.) in ITP patients in the acute phase and in ITP patients in complete or partial remission (CR/PR), respectively, with or without detectable anti‐platelet (anti‐plt) antibodies
| Acute‐phase ITP patients with anti‐plt antibody (n = 24/80, 30%)(a) | Acute‐phase ITP patients without anti‐plt antibody (n = 56/80, 70%)(b) | CR/PR ITP patients with anti‐plt antibody (n = 10/50, 20%)(c) | CR/PR ITP patients without anti‐plt antibody (n = 40, 80%)(d) | Statistical significance | ||
|---|---|---|---|---|---|---|
| (a) versus (b) | (c) versus (d) | |||||
| C3 % of pooled plasma | 94·08 ± 7·1 | 101·14 ± 4·1 | 94·20 ± 12·0 | 96·97 ± 19·4 | > 0·05 | > 0·05 |
| C4 % of pooled plasma | 92·04 ± 3·6 | 92·20 ± 2·9 | 90·20 ± 1·1 | 89·82 ± 2·1 | > 0·05 | > 0·05 |
| sC5b‐9 level (ng/ml) | 233·63 ± 14·7 | 194·80± 8·8 | 110·00 ± 0·1 | 98·82 ± 6·1 | < 0·001 | < 0·001 |
| C1q level (g/l) | 4·91 ± 2·1 | 1·16 ± 0·5 | 1·0 ± 0·1 | 1·0 ± 0·1 | < 0·001 | > 0·05 |
ITP = immune thrombocytopenic purpura; s.d. = standard deviation.
We found a significant increase of C1q and sC5b‐9 plasma levels in patients with acute‐phase ITP compared to both CR or PR ITP. The median value of sC5b‐9 was 200 versus 98 ng/dl (P‐value < 0·001) and the median value of C1q was 2·11 versus 1·00 mg/dl (P‐value < 0·001). ITP patients in CR or PR had plasma levels of sC5b‐9 and C1q comparable to those observed in healthy volunteers. Outliers did not affect the obtained results: the 95% confidence interval (CI) for the mean level of sC5b‐9 ranges from 201·79 to 211·11 mg/dl for acute‐phase ITP patients, from 99·04 to 103·08 mg/dl for CR or PR ITP patients and from 101·29 to 104·79 mg/dl for healthy volunteers. Similarly, the CI for the mean level of C1q ranges from 1·68 to 2·54 mg/dl for acute‐phase ITP patients, while it is constantly equal to 1 mg/dl for other ITP patients and healthy volunteers.
The box‐plot distribution of both the mentioned factors is shown in Fig. 1. Moreover, a significant correlation was found between sC5b‐9 and C1q plasma levels (Spearman’s rho correlation index on 130 ITP patients = 0·58, P‐value < 0·001).
Fig. 1.
Distribution of sC5b‐9 and C1q plasma levels, respectively, in acute‐phase immune thrombocytopenic purpura (ITP), complete remission/partial remission (CR/PR) ITP patients and healthy volunteers.
In addition, statistical analysis clearly showed that high values of platelets were associated with low values of sC5b‐9 and, inversely, that low values of platelets were associated with high values of sC5b‐9. These two factors were highly inversely correlated (Pearson’s correlation index equal to −0·54, P‐value < 0·001).
An additional analysis was conducted in 80 active ITP patients. The patients were divided into those with detectable (24 of 80, 30%) and undetectable (56 of 80, 70%) anti‐platelet antibodies. The bi‐directional t‐test indicates that acute ITP patients with detectable anti‐platelet antibodies had sC5b‐9 levels almost 20% higher than in patients with undetectable autoantibodies (233·63 versus 194·80 mg/dl, P‐value < 0·001). Similarly, the mean value of C1q was almost four times higher in patients with detectable anti‐platelet antibodies than with undetectable antibodies (4·32 versus 1·16 mg/dl, P‐value < 0·001). The box‐plot distribution of both aforementioned factors is shown in Fig. 2. No significant difference in C3 or C4 plasma levels was found between the two groups of patients. Statistical analysis showed no significant correlation between levels of complement and response to therapies (data not shown).
Fig. 2.
Distribution of sC5b‐9 and C1q plasma levels in acute‐phase immune thrombocytopenic purpura (ITP) and complete remission/partial remission (CR/PR) ITP patients with and without detectable anti‐platelet antibodies and in healthy volunteers.
Bleeding complications
Thirteen patients experienced bleeding complications. We observed three mucosal bleeding complications (grade 3) among active ITP patients and 10 mucosal bleeding complications among CR or PR patients. None of the patients had intracranial hemorrhages or fatal bleeding. All patients with bleeding complications had significantly lower platelet counts. There was no difference in sC5b‐9 level between patients with and without bleeding complications (Table 4 and Fig. 3).
Table 4.
Plasma level of sC5b‐9 and platelet count (mean value ± s.d.) in ITP patients, both in the acute phase or complete/partial remission, with or without bleeding complications. Data suggest that ITP patients with bleeding events had moderately lower values of sC5b‐9 (146·73 versus 168·18, P‐value > 0·05) and significantly lower platelet count (9363·64 versus 30 236·84, P‐value < 0·001) than ITP patients without bleeding events
| All ITP patients with bleeding complications (n = 13/130, 10%)(a) | All ITP patients without bleeding complications (n = 117/130, 90%)(b) | Statistical significance | |
|---|---|---|---|
| (a) versus (b) | |||
| sC5b‐9 level (ng/ml) | 146·73 ± 65·8 | 168·18 ± 52·9 | > 0·05 |
| Platelet count (n/mm3) | 9363·64 ± 4201·7 | 30 236·84 ± 24 146·8 | < 0·001 |
ITP = immune thrombocytopenic purpura; s.d. = standard deviation.
Discussion
In this study, we aimed to evaluate if complement is activated during ITP, if complement activation is related to the presence or absence of detectable anti‐platelets and if complement activation correlates with the complete or partial remission of chronic of ITP.
In ITP patients, the drop in platelet mass caused by accelerated platelet clearance was mainly attributed to autoantibody‐mediated destruction by macrophages in the spleen.
In ITP, platelet mass falls as a result of accelerated platelet clearance due mainly, but not exclusively, to autoantibody‐mediated destruction by macrophages in the spleen. In addition, platelet production tends to be moderately impaired due to the cytotoxic T cell‐mediated megakaryocytic damage [1, 2, 3, 17]. The importance of anti‐platelet antibodies in its pathogenesis is supported by the following observations: (1) the presence of platelet‐associated autoantibodies in plasma of patients with immune thrombocytopenia and their absence in control subjects and patients affected by non‐immune thrombocytopenia; (2) infusion of plasma or IgG‐rich plasma fractions from ITP patients into healthy subjects induces thrombocytopenia; and (3) procedures affecting antibody production, such as splenectomy, cyclophosphamide and combination chemotherapy, decreases or induces disappearance of autoantibody titers. The complement system has long been suspected to participate in platelet elimination in patients with ITP [18].
It has been observed that, in a large number of patients with chronic ITP, anti‐platelet autoantibodies mainly directed to GP IIb⁄IIIa and GP Ib⁄IX are capable of activating the classical complement pathway. Complement fixation is even present in ITP sera without detectable autoantibodies. Anti‐platelet antibodies in patient plasma can bind to platelets in the test system and increase classical complement activation above baseline. These findings support the hypothesis that current techniques for autoantibody detection may fail to detect them [19]. Although there is no evidence of direct cytolysis mediated by C5b9 on platelets [20], complement activation may contribute together with anti‐platelet antibodies to platelet destruction either via clearance of platelets opsonized with immunoglobulin by the reticuloendothelial system or via complement receptors on macrophages in the spleen [21].
In this study we investigated the role of complement in ITP and in the different phases of the disease course. This study provides the changing of complement activation products in the serum of ITP patients in the acute phase and in CR or PR. The remarkable increments of C1q and sC5b9 indicate a strong activation of complement during the acute phase of ITP. Moreover, among acute‐phase patients, those with detectable anti‐platelet antibodies have higher levels of C1q and sC5b‐9, suggesting a higher rate of complement activation. In this study, patients in CR or PR have complement parameters similar to those of control subjects, although we cannot rule out the possible effects of ITP drugs in reducing the title of platelet antibodies. Furthermore, plasma levels of C1q and sC5b‐9 may discriminate between active and inactive ITP, and between major flares and no active disease. These complement‐related mechanisms may provide potential therapeutic targets in patients with ITP; for example, drugs to reduce the activation of C5 (Eculizumab) and C3 (Compstatin). In order to investigate the role of anti‐platelet antibodies in complement activation, we evaluated the level of complement activation in ITP patients with and without detectable anti‐platelet antibodies. Statistical analysis showed significant complement activation in the presence of anti‐platelet antibodies compared to undetectable anti‐platelet antibodies. In addition, we found that the degree of thrombocytopenia positively correlates with complement activation.
Although the results of this study must be confirmed in a larger number of ITP patients, they provide some novel observations: (1) ITP patients in the acute phase have higher levels of sC5b‐9; (2) patients with anti‐platelet antibodies share higher levels of C1q and sC5b‐9 compared with patients without anti‐platelet antibodies; and (3) increased levels of sC5b‐9 inversely correlate to the severity of thrombocytopenia.
Taken together, these findings provide evidence that in ITP complement is activated according to disease stage and detectability of anti‐platelet antibodies. Complement activation strongly correlates with severity of disease.
Platelet autoantibodies can induce the activation of the classical pathway of complement, leading to elevated deposition of complement proteins which stimulate either enhanced platelet phagocytosis or complement‐mediated generation of platelet microparticles. The latter can trigger platelets inducing the expression of platelet membrane procoagulant activity through the expression of tissue factor‐inducing clot activation and thrombosis. This phenomenon might potentially explain the increased risk of thrombosis that patients with chronic ITP have despite platelet reduction [22, 23].
Taken together, these observations provide evidence that in ITP complement is activated according to disease stage and the detectability of anti‐platelet antibodies. Complement activation strongly correlates with severity of disease.
Disclosures
None.
Acknowledgements
We thank Dr Luca Cilumbriello for his advice in statistical processing and Dr Kari Bohlke for her review of the manuscript.
References
- 1. George JN, Woolf SH, Raskob GE. Idiopathic thrombocytopenic purpura: a practice guideline developed by explicit methods for the American Society of Hematology. Blood 1996; 88:43–46. [PubMed] [Google Scholar]
- 2. Cines DB, Blanchette V. Idiopathic thrombocytopenic purpura. N Engl J Med 2002; 364:995–1008. [DOI] [PubMed] [Google Scholar]
- 3. Cohen YC, Djulbegovic B, Shamai‐Lubovitz O, Mozes B. The bleeding risk and natural history of idiopathic thrombocytopenic purpura in patients with persistent low platelet counts. Arch Intern Med 2000; 160:1630–8. [DOI] [PubMed] [Google Scholar]
- 4. Adelborg K, Kristensen NR, Norgard M. Cardiovascular and bleeding outcomes in a population based cohort of patients with chronic immune thrombocytopenia. J Thromb Haemost 2019; 17:912–24. [DOI] [PubMed] [Google Scholar]
- 5. Beardsley DS, Ertem M. Platelet autoantibodies in immune thrombocytopenic purpura. Transfus Sci 1998; 19:237–44. [DOI] [PubMed] [Google Scholar]
- 6. Li J, Sullivan JA. Pathophysiology of autoimmune thrombocytopenia. Curr Opin Hematol 2018; 25:373–81. [DOI] [PubMed] [Google Scholar]
- 7. Kayser W, Mueller‐Eckhardt C, Bhakdi S. Platelet‐associated complement C3 in thrombocytopenic states. Br J Haematol 1983; 54:353–63. [DOI] [PubMed] [Google Scholar]
- 8. Mannucci PM, Cugno M. The complex differential diagnosis between thrombotic thrombocytopenic purpura and the atypical hemolytic uremic syndrome: laboratory weapons and their impact on treatment choice and monitoring. Thromb Res 2015; 136:851–4. [DOI] [PubMed] [Google Scholar]
- 9. Ardissino G, Tel F, Sgarbanti M. Complement functional tests for monitoring eculizumab treatment in patients with atypical hemolytic uremic syndrome: an update. Pediatr Nephrol 2018; 33:457–61. [DOI] [PubMed] [Google Scholar]
- 10. Le Clech A, Simon‐Tillaux N, Provôt F, Delmas Y. Atypical and secondary hemolytic uremic syndromes have a distinct presentation and no common genetic risk factors. Kidney Int 2019; 95:1443–52. [DOI] [PubMed] [Google Scholar]
- 11. Frémeaux‐Bacchi V, Miller EC, Liszewski MK. Mutations in complement C3 predispose to development of atypical hemolytic uremic syndrome. Blood 2008; 112:4948–53. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Polley MJ, Nachmann RL. Human platelet activation by C3a and C3a Des‐Arg. J Exp Med 1983; 158:603–15. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Fletcher MP, Sthal GL, Longhurst JC. C5a‐ induced myocardial ischemia: role for CD 18 dependent PMN localization and PMN‐platelet interactions. Am J Physiol 1993; 265:1750–61. [DOI] [PubMed] [Google Scholar]
- 14. Rodeghiero F, Stasi R, Gernsheimer T. Standardization of terminology, definitions and outcome criteria in immune thrombocytopenic purpura of adults and children: report from an international working group. Blood 2009; 113:2386–93. [DOI] [PubMed] [Google Scholar]
- 15. Rodeghiero F, Michel M, Gernsheimer T. Standardization of bleeding assessment in immune thrombocytopenia: report from the International Working Group. Blood 2013; 14:2596–606. [DOI] [PubMed] [Google Scholar]
- 16. Kiefel V, Santoso S, Weisheit M. Monoclonal antibody‐specific immobilization of platelet antigens (MAIPA): a new tool for the identification of platelet‐reactive antibodies. Blood 1987; 70:1722–6. [PubMed] [Google Scholar]
- 17. Audia S, Mahévas M, Samson M et al Pathogenesis of immune thrombocytopenia. Autoimmun Rev 2017; 16:620–32. [DOI] [PubMed] [Google Scholar]
- 18. Peerschke EI, Panicker S, Bussel J. Classical complement pathway activation in immune thrombocytopenia purpura: inhibition by a novel C1s inhibitor. Br J Haematol 2016; 173:942–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Najaoui A, Bakchoul T, Stoy J et al Autoantibody‐mediated complement activation on platelets is a common finding in patients with immune thrombocytopenic purpura (ITP). Eur J Haematol 2011;88:167–74. [DOI] [PubMed] [Google Scholar]
- 20. Martel C, Cointe S, Maurice P, et al Requirements for membrane attack complex formation and anaphylatoxins binding to collagen‐activated platelets. PLOS ONE 2011; 6:e18812. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Peerschke EIB, Andemariam B, Yin WQ, Bussel JB. Complement activation on platelets correlates with a decrease in circulating immature platelets in patients with immune thrombocytopenic purpura. Br J Haematol 2009; 148:638–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Peerschke EI, Yin W, Ghebrehiwet B. Complement activation on platelets: implications for vascular inflammation and thrombosis. Mol Immunol 2010; 47:2170–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Yin W, Ghebreiwet B, Peerschke EIB. Expression of complement components and inhibitors on platelet microparticles. Platelets 2007; 19:225–33. [DOI] [PMC free article] [PubMed] [Google Scholar]