If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Clinical Division of Hematology and Hemostaseology, Department of Medicine I, Medical University of Vienna, Vienna, AustriaInstitute of Vascular Biology and Thrombosis Research, Centre of Physiology and Pharmacology, Medical University of Vienna, Vienna, Austria
The bleeding phenotype in immune thrombocytopenia (ITP) is heterogeneous, but usually mild and only partly dependent on the severity of thrombocytopenia. Platelet reactivity has previously been suggested to underly the mild phenotype.
Platelet function was assessed as basal and agonist-induced surface expression of P-selectin and activation of GPIIb/IIIa via flow cytometry, and soluble (s)P-selectin levels were assessed in plasma of 77 patients with primary ITP, 19 hemato-oncologic thrombocytopenic controls (TC) and 20 healthy controls (HC). The association of platelet function with laboratory and clinical parameters such as bleeding manifestations at inclusion and previous thrombosis was analyzed.
ITP patients showed tendency towards increased surface P-selectin and elevated levels of activated GPIIb/IIIa. Platelet activation after stimulation with all agonists including TRAP-6, ADP, arachidonic acid and CRP was decreased compared to HC. Compared to TC, only GPIIb/IIIa activation but not surface P-selectin was higher in ITP. Levels of soluble (s)P-selectin were significantly higher in ITP patients compared to TC, but similar to HC. Higher sP-selectin levels were associated with blood group O and current therapy, with highest levels in TPO-RA treated patients. Platelet reactivity was not associated with platelet count or size, platelet antibodies, treatment regime, or blood group. No correlation between platelet activation with the bleeding phenotype or previous thrombotic events could be observed.
ITP patients did not have hyper-reactive platelets compared to HC, but partly higher reactivity compared to TC. Further studies are needed to understand the underlying mechanism behind the bleeding and pro-thrombotic phenotype in ITP.
]. These clinical observations have been attributed to a higher amount of young and large platelets in ITP patients, as these are highly functional and possibly more activatable and thus might compensate for decreased platelet counts and prevent severe bleeding events [
Platelet activation status can be assessed via activation of the fibrinogen receptor GPIIb/IIIa, which is recognized by a specific antibody (PAC-1), and surface expression of P-selectin (CD62P), which derives from platelet α-granules and is only exposed upon degranulation [
]. Platelet activation can also be accessed indirectly via shedding of soluble P-selectin (sP-selectin) in plasma. As a biomarker of platelet activation sP-selectin has been associated to arterial and venous thromboembolic events [
]. Furthermore, the association of platelet function with clinical aspects such as bleeding symptoms or history of thrombosis has never been systematically investigated.
In this study, we aimed to investigate whether platelets in ITP patients have altered functions or activatability by measuring the platelet activation status and reactivity, and sP-selectin in a large cohort of well-characterized ITP patients in comparison to a group of non-immunological thrombocytopenia patients and healthy controls (HC). Furthermore, we evaluated the association of platelet activation to clinical outcomes, namely the bleeding phenotype and previous thrombosis, in primary ITP patients, and analyzed the effect of clinical factors and ITP treatment regimens on the platelet activation status.
2.1 Study design and patients
The ITP biobank is a multicenter prospective cohort study, which is conducted at the Department of Hematology and Haemostaseology of the Medical University of Vienna and the Third Medical Department of the Hanusch Hospital Vienna. The study was approved by the Ethics Committee of the Medical University of Vienna (EC1843/2016).
Adult patients, who are newly diagnosed with primary ITP, under treatment or observation at a participating hematological center are included in the study. All patients included in this study met the standard criteria for a diagnosis of primary ITP and had a platelet count of ≤100 × 109/L at the time of diagnosis. Patients without ITP specific treatment, regardless the duration, had to have a platelet count of ≤100 × 109/L at study inclusion, whereas in patients receiving ITP-specific treatment platelet counts >100 × 109/L were accepted. The minimum age for inclusion was 18 years. Patients with secondary (immune) thrombocytopenia or an active neoplastic disease were not included in the ITP biobank.
Furthermore, a group of patients with non-immunological thrombocytopenia (thrombocytopenic controls, TC) was recruited at the ward of the Department of Hematology and Haemostaseology at the Medical University of Vienna (EC1843/2016). Inclusion criteria for the TC group were age ≥ 18 years and a non-immunological thrombocytopenia with platelet counts ≤100 × 109/L. Of the 19 recruited TC, 7 patients had acute myeloid leukemia (36.8 %), 3 patients each had acute lymphatic leukemia (15.8 %) or chronic lymphatic leukemia (15.8 %) or myelodysplastic syndrome (15.8 %), and the latter 3 had diffuse large B-cell lymphoma (5.2 %), Burkitt-lymphoma (5.2 %) or T-lymphoblastic lymphoma (5.2 %). All TC received chemotherapy before inclusion in the study.
Sex- and age-matched HC aged ≥18 years were recruited from patients' relatives and hospital personnel. All patients and participants gave written informed consent before study inclusion.
ITP patients underwent a structured interview with trained personnel at study inclusion, to record general medical and ITP specific clinical data, including onset of ITP, previous treatments, co-medications, and the previous thrombosis history using a standardized questionnaire. Bleeding manifestations at time of inclusion were recorded using a standardized, ITP-specific bleeding assessment tool as reported below. Furthermore, the life-long bleeding history was recorded by the Vicenza bleeding score [
], and information on bleeding at ITP diagnosis, and previous severe bleedings (WHO grade III and IV) were recorded. In line, also in TC and HC general medical data, previous thrombosis and the bleeding score were recorded by medical personnel using the same questionnaires.
Blood samples of all study participants were taken and timely processed either to the assessment of platelet function, routine laboratory measurements, or storage at the biobank, as described in supplementary paragraph 1 [
]. Bleeding symptoms were assessed by physical examination at the time of study inclusion by trained personnel. Bleeding symptoms were recorded within three categories, Skin (S), Mucosa (M) and Organ (O), and graded according to the severity (G) of each manifestation from 0 to 3 or 4, with grade 5 presenting a fatal bleeding. Furthermore, significant severe bleeding at study inclusion was assessed.
2.3 Platelet function assessment
Determination of platelet activation and reactivity by glycoprotein expression naïve and after addition of agonists was determined by flow cytometry on a FACSCanto™ II Cell Analyzer from Becton Dickinson (BD, San Jose, CA, USA) at study inclusion [
Before assessment, within 1 h of blood sampling, platelet counts were determined using a DxH Hematology analyzer (Beckman coulter). In samples with <100 × 109/L platelets, platelet rich plasma (PRP) was diluted with phosphate-buffered-saline (PBS) to result in 20 × 109/L platelets [
]. For PRP preparation, whole blood was centrifuged at 150 xg at room temperature for 7 min, without break. In patients and controls with platelets count ≥100 × 109/L, whole blood was used.
To stain the platelet population, 5 μL of anti-CD42b (clone HIP1, allophycocyanin labelled, Becton Dickinson) were added to 20 μL PRP and incubated at room temperature (RT) for 10 min in the dark. As agonists thrombin receptor activating peptide-6 (TRAP-6, 5 μM; Bachem, Bubendorf, Switzerland), ADP (1 μM, Roche ADPtest), arachidonic acid (AA, 80 μM, Roche ARAtest), CRP-XL (0,04 μg/mL, generous gift from Dr. R. W. Farndale, Department of Biochemistry, University of Cambridge, Cambridge, UK), or PBS (10 μL) were added and incubated for 10 min at RT in the dark. For staining, the following monoclonal antibodies (mAB) were used: fluorescein isothiocyanate (FITC)-conjugated monoclonal antibody PAC-1 (BD), which only binds to the activated conformation of GPIIb/IIIa, phycoerythrin (PE)-conjugated anti-P-selectin monoclonal antibody (CD62P, clone CLB-Thromb/6, Immunotech) or the corresponding isotype-matched controls from Immunotech (IgM control FITC) and BD (IgG1 Platelet Control PE) to determine non-specific binding. Optimal concentrations of antibodies and agonists were determined by titration in preliminary experiments.
After a 15-min incubation at RT in the dark, the reaction was stopped with 500 μL PBS at 4 °C and after additional 10 min in the dark, measurements on a FACSCanto™ II Cell Analyzer from BD were started. To ensure instrument performance over time, the flow cytometer is calibrated daily with FACSDiva™ CS&T IVD Beads by BD. Platelets were gated in a side scatter vs FL4 dot plot to acquire 10.000 events, which were analyzed in histograms for FL-1 for the binding of fibrinogen (PAC-1) and FL-2 for the activation of alpha-granules (CD62P), respectively. Analysis includes the evaluation of median fluorescence intensity (MFI), as well as the percentage of positive events facilitated by setting a marker around 1 (max. 1.2) on a logarithmic scale histogram [
]. Platelet activation data are presented as percentage of P-selectin and GPIIb/IIIa activation compared to an isotype-matched control before and after platelet activation with distinct agonists.
2.4 Quantitative determination of sP-selectin
To determine levels of sP-selectin, the Quantikine® Human P-Selectin/CD62P Immunoassay (Cat. No. DPSE00, R&D Systems®, Minneapolis, USA) was performed according to manufacturer's manual in baseline samples of partcipants stored in the biobank [
]. In short, monoclonal antibodies specific for sP-selectin conjugated to horseradish peroxidase are applied in combination with the kit-provided standards, controls and patient samples. After regular washing-processes, a substrate solution of the mixed color reagents is added to start a color reaction, where the color intensity is proportional to the sP-selectin concentrations and optical density is measured on a Multiskan™ FC Microplate Photometer (Thermo Scientific™, Waltham, USA) [
Continuous variables are described by median values (interquartile ranges, IQR), categorical variables by counts (percentages). Normally distributed continuous clinical and laboratory characteristics were compared between groups using analysis of variance (ANOVA) models, whereas in case of non-normal distributions the nonparametric Kruskal-Wallis test (3 groups) or Wilcoxon rank sum test (2 groups) were used. The chi-square test and Fisher's exact test were performed to evaluate differences in categorical characteristics. Basal platelet activation markers were compared between the three groups (ITP, TC and HC) using ANOVA models. Levels of activation markers after stimulation with TRAP-5, ADP, AA, and collagen, respectively, were compared between the three groups using analysis of covariance (ANCOVA) models, including the basal activation levels as covariates. In case of a statistically significant main (group) effect, pairwise post-hoc comparisons were done using the Tukey-Kramer multiplicity correction. The same procedures were performed to test for differences in basal and stimulated activation markers between ITP patients with platelet count ≤ 50 G/L and ITP patients with platelet count ≥ 50 G/L. To evaluate the potential association of clinical factors with platelet activation markers in ITP patients, multivariable linear regression models were calculated, considering platelet count, immature platelet fraction (IPF), MPV, blood group, anti-platelet antibodies, and current therapy as independent variables. Separate multivariable linear regression models were performed on every single activation marker (dependent variable). Furthermore, differences between ITP patients with or without bleeding symptoms, and between ITP patients with or without history of thromboembolic events, respectively, were tested using the two-sample t-test. Differences of sP-selectin levels between ITP treatment groups were tested using an analysis of variance (ANOVA) model with pairwise post-hoc comparisons applying the Tukey-Kramer multiplicity correction. The base 2 logarithmic transformation was applied on variables with skew distributions to achieve normally distributed variables for statistical calculations. The Spearman correlation coefficient was calculated to quantify the correlation between platelet activation markers and the bleeding phenotype. Two-sided p-values < 0.05 were considered as statistically significant. Due to the exploratory and hypotheses generating character of this study, no further multiplicity corrections were applied. The software SAS (SAS Institute Inc., 2020, Cary, NC, USA) was used for statistical calculations.
3.1 Patients' clinical and laboratory characteristics
Seventy-seven ITP patients, 19 TC and 20 HC, both age- and sex-matched, were included in this study. Clinical and laboratory characteristics of patients and both control groups are shown in Table 1.
Table 1Clinical and laboratory characteristics of 77 ITP patients, 19 thrombocytopenic controls, and 20 healthy controls.
The median (IQR) platelet count in ITP patients was 69 (34.5–147.5) x109/L with a platelet volume (MPV) of 11.5 (10.4–12.6) fL, which was slightly higher than in TC (median MPV: 9.9 (9.7–10.6)) and HC (median MPV: 10.6 (10.2–11.3)). Of the 77 ITP patients, 34 (44.2 %) were under current ITP treatment at the time of study inclusion. Most of them (n = 15, 19.5 %) received corticosteroids, 14 (18.2 %) received a TPO-RA, and 5 (6.5 %) were treated with a combined treatment (4 corticosteroids and TPO-RA, 1 TPO-RA and Rituximab). Fourteen ITP patients (18.2 %) were splenectomized, while none of the TC or HC had a history of splenectomy.
Anti-platelet antibodies were found in 30 ITP patients (39.5 %) on the platelet-surface and in 27 ITP patients (35.5 %) in the serum. In two TC (10.5 %) autoantibodies against GPIIb/IIIa were found, while no HC had anti-platelet antibodies.
There was no difference in the SMOG BS, total or according to the 3 subcategories between ITP patients and TC, while HC had a lower SMOG BS (Table 1). Eleven ITP patients (14.3 %) had a history of a severe bleeding event since diagnosis of ITP including peri−/postpartum bleeding (n = 3), gastrointestinal bleeding (n = 2), posttraumatic/surgery bleeding (n = 3), heavy menstrual bleeding (n = 2) or intracerebral bleeding (n = 1) at inclusion.
Eight ITP patients (10.4 %) and 3 TC (15.8 %) had a history of thrombosis. Of the ITP patients, 5 had a deep vein thrombosis (DVT) and 2 an arterial thrombosis, in one the exact thrombotic location was not reported. One patient with DVT only had TPO-RA as potential risk factor, one patient with DVT had combined treatment with TPO-RA and steroids and hyperlipoproteinemia as potential risk factors, in one patient with DVT a protein C deficiency was diagnosed besides obesity as a potential risk factor, and in the remaining 2 patients with DVT no risk factors were reported. Of the one patient with ischemic stroke, atrial fibrillation and hypertension were the two main risk factors, one patient with an acute coronary syndrome had hypertension, diabetes and hyperlipoproteinemia.
Three TC had an arterial thrombotic event (2 strokes and 1 acute coronary syndrome), of whom one had diabetes, one hypertension and one both diabetes and hypertension as known risk factors for thromboembolic events, besides their malignant hematologic disease.
3.2 Platelet activation and reactivity in ITP patients and control groups
Results of the analysis of glycoprotein expression, naïve and upon addition of agonists are reported in Table 2 and depicted in Fig. 1. The main group effect was significant in all measured platelet activation markers in the performed AN(C)OVA models. In the pairwise comparisons, the naive expression of P-selectin was highest in ITP patients compared to HC and TC, but lacking statistical difference. P-selectin levels were lower in ITP patients compared to HC upon addition of TRAP-6, ADP, AA, or CRP, whereas there was no relevant difference between ITP patients and TC.
Table 2Comparison of platelet activation markers between ITP patients, TC and HC.
There was no relevant difference in basal GPIIb/IIIa activation between ITP patients and TC or HC, but again a trend towards higher basal platelet activation in ITP compared to HC. Again, upon activation with all tested agonists, ITP patients had lower levels of activated GPIIb/IIIa in comparison to HC (Table 2). In comparison to TC, GPIIb/IIIa activation after stimulation with TRAP-6, ADP, AA was higher in ITP patients, whereas there for CRP no significant difference could be reached.
Levels of sP-selectin were significantly higher in ITP patients compared to TC, whereas there was no difference in comparison to HC (Table 2, Fig. 2).
Lastly, there was a higher platelet reactivity in HC compared to TC, shown by surface P-selectin expression and activated GPIIb/IIIa with all agonists. There was no difference in baseline P-selectin expression or GPIIb/IIIa activation, while sP-selectin levels where higher in HC than in TC (Table 2).
Next, we separately analyzed patients with platelet counts ≤50 × 109/L (Supplementary table 1). In this subgroup of 28 patients, the median (IQR) platelet count, MPV and IPF were 29.5 (13.3–37.8), 12.1 (11.5–13.7), and 15.0 (10.4–23.9), respectively. Again, the main group effect was significant in all measured platelet activation markers in the performed AN(C)OVA models. Also, in the separate analysis of patients with platelet counts ≤50 × 109/L, naïve P-selectin expression was highest in ITP patients. ITP patients with platelet counts ≤50 × 109/L again had lower levels of P-selectin expression upon addition of TRAP-6, ADP, AA, and CRP, while there were differences in the expression of P-selectin upon addition of TRAP-6 and AA between ITP patients and TC. In comparison to HC, ITP patients with platelet counts ≤50 × 109/L had higher basal GPIIb/IIIa activation than TC and HC. As in the whole group of ITP patients, the reactivity was lower with all agonist in ITP compared to HC, while there was no difference compared to HC.
Also in patients with platelet counts ≤50 × 109/L, sP-selectin levels were higher than in TC and similar to HC.
Further, ITP patients with higher platelet counts (> 50 G/L, n = 28) had higher platelet reactivity but no difference in basal platelet activation of P-selectin levels and GPIIbIIIa expression than patients with lower platelet counts (≤ 50 G/L, n = 49, Supplementary table 2).
3.3 Clinical parameters influencing platelet activation in ITP patients
To identify clinical factors associated with platelet activation markers in ITP patients, we performed a multivariable linear regression analysis including platelet count, IPF, MPV, blood group (O vs. non O), antiplatelet antibodies and current ITP treatment (Table 3). Only P-selectin expression after addition of TRAP-6 and AA, and GPIIb/IIIa activation after addition of AA were dependent by the platelet count. P-selectin expression after AA addition showed a association with the MPV, while IPF showed no association to any tested agonist-induced reactivity. We found no impact of current ITP treatment on the expression of P-selectin or GPIIb/IIIa activation, neither unstimulated nor in response to agonists. Also, anti-platelet antibodies and blood group O were not associated with expression of P-selectin or GPIIb/IIIa activation.
Table 3Multivariable linear regression of parameters influencing platelet activations markers in ITP.
Levels of sP-selectin were associated with blood group O and current treatment regime (Table 4). The separate analysis of the influence of ITP treatments on sP-selectin levels showed an overall difference between different treatment regimens (no therapy, corticosteroids, TPO-RA or combination of corticosteroids with TPO-RA/rituximab, ANOVA p = 0.035). In post-hoc Tukey-Kramer multiplicity corrected analysis sP-Selectin levels were significantly higher in TPO-RA treated compared to untreated ITP patients (Table 5).
Table 4Multivariable linear regression of parameters influencing sP-selectin (log2 transformed) in ITP patients.
Due to the low number of ITP patients with platelet counts ≤50 × 109/L, no multivariable regression analysis was performed in this subgroup.
3.4 Correlation of platelet activation markers and bleeding or previous thrombosis
To investigate a potential correlation of platelet function with the bleeding phenotype in ITP patients, we compared platelet function parameters between patients with and without bleeding manifestations at time of the investigation of platelet function, which showed no difference (Table 6). Of note, there was no difference in the platelet count between these two groups (mean (SD) 109.2 (91.0) vs. 98.2 (91.4), p = 0.616). We further calculated correlations between the bleeding severity measured by the SMOG BS in total and according to the subcategories with platelet activation markers in ITP patients (Supplementary table 3). We found no correlation between P-selectin expression, activated GPIIb/IIIa levels or sP-selectin levels and the bleeding phenotype at study inclusion.
Table 6Influence of presence of bleeding symptoms on platelet activation markers within ITP patients.
As mentioned before, severe bleeding manifestations (WHO grade III/IV bleeding) since ITP diagnosis occurred in 11 patients. We next compared platelet function at study inclusion of patients who experienced a severe bleeding manifestation to those without. There was no difference in any of the analyzed platelet activation markers and sP-selectin between these groups (Supplementary table 4).
Also, between patients with or without a history of thromboembolic events, there was no significant difference in glycoprotein expression or levels of sP-selectin (Supplementary table 5).
In light of the conflicting data on platelet function in primary ITP patients from previous studies, we investigated platelet function and reactivity in a large and well-characterized cohort of adult ITP patients in comparison to TC after chemotherapy and HC to assess if platelet activation is associated with bleeding and/or previous thromboembolic events. We could show that there is a tendency towards an increased basal platelet activation state in ITP patients compared to healthy controls, whereas platelet reactivity upon stimulation with distinct agonists was lower in ITP patients compared to HC. This was seen in the general ITP cohort, as well as in the subgroup of patients with low platelet counts (≤50 × 109/L). In comparison to TC, platelet reactivity was either similar or increased in ITP patients, as seen by GPIIb/IIIa activation upon agonist stimulation. Platelet function was not associated with clinical outcome parameters, such as the bleeding phenotype or previous thrombosis in ITP patients.
The increased basal P-selectin expression and GPIIb/IIIa activation compared to healthy individuals was in line with previous clinical investigations on smaller cohorts of chronic ITP patients [
]. Previous studies that suspected an increase of thromboembolic risk in patients under TPO-RA treatment due to increased platelet reactivity suffer from small sample size and only compared treated patients to standard of care treated patients, or before and during the treatment [
]. However, we and others previously showed no impact of eltrombopag on platelet reactivity. Further, no effect of baseline expression of P-selectin or GPIIb/IIIa activation on the bleeding phenotype of ITP patients was observed in this study, which is in line with previous results [
] and challenges the clinical relevance of this slightly elevated basal platelet activation.
In contrary to basal platelet activation, platelet reactivity after stimulation with various agonists was diminished in ITP patients. We and others have previously shown lower expression of P-selectin and GPIIb/IIIa activation upon stimulation with TRAP-6, but not ADP in ITP patients compared to HC [
]. However, previous studies differed mainly on the lower sample size but also with regards to disease severity, lower percentage of patient on ITP treatment and higher splenectomy rate. While, in this study, we reported on reduced in vitro platelet reactivity in ITP with various agonists including TRAP-6, AA, CRP and ADP [
]. In our study of the so far biggest analyzed cohort however, platelet reactivity was not associated to IPF or MPV. Also the presence of antiplatelet antibodies did not affect platelet function or levels of sP-selectin in our cohort, in contrast to a previous smaller study on chronic ITP [
], no clear difference in platelet expression of P-selectin between the two thrombocytopenic groups could be observed, while GPIIb/IIIa activation was higher in ITP. Impaired platelet production might result in defects in the cytoskeleton, impaired receptor-mediated intracellular signaling and/or defective granule secretion. For example, impaired platelet function has been described in thrombocytopenic patients with newly diagnosed AML [
]. Further, also in patients with hematological malignancies undergoing chemotherapy, impairment of platelet function, more specifically Ca2+ flux, integrin a β activation, P-selectin expression and platelet spreading on fibrinogen, has been described [
]. On the other hand, ongoing platelet activation and platelet exhaustion and consequently abrogated platelet reactivity is associated to increased mortality and thromboembolic events in solid cancer patients [
]. Nevertheless, the hypothesis of ongoing platelet activation seems unlikely in our TC, as levels of sP-selectin, a platelet activation marker, were lower in TC than in ITP patients in our study. Circulating levels of sP-selectin were increased in TPO-RA treated patients, whereas we did not observe an overall difference in sP-selectin levels compared to HC, which is in line with previous reports [
]. While data on the prospective thrombotic risk of our study cohort are yet not available, we found no differences in sP-selectin levels between ITP patients with or without previous thrombosis. In line with previous data, higher sP-selectin levels were associated with blood group O also in our cohort [
], however the exact underlying effect of blood group O influencing platelet function remains unclear.
It is known that bleeding risk in ITP is only partly predictable by platelet count and not directly related to a distinct threshold platelet count, even though bleeding risk seems to increase when platelet count is below 20 × 109/L [
]. Therefore, we analyzed platelet function in a subgroup of ITP patients with a platelet count ≤50 × 109/L, which showed an even more pronounced impairment of platelet reactivity in comparison to the whole ITP cohort. As a higher platelet activation state has been postulated to underly this relatively moderate bleeding phenotype in ITP by some studies [
], we investigated the association between platelet function and bleeding in our larger cohort. Our results show no association of platelet reactivity and bleeding manifestations.
Our study was performed in a cohort of well-characterized patients with ITP, yet, has some limitations. First, our cohort of ITP patients and control groups are relatively small and sample size was not equal in all groups, though we provide data on the largest ITP patient cohort, in which platelet function of this orphan disease were investigated. Within the Vienna ITP biobank, we include ITP patients who were diagnosed and/or treated at our outpatient department. Thus, our patient cohort is heterogenous in terms of disease state and treatment. Nevertheless, this approach also provides an advantage as it represents a broad approach in a clinically representative cohort of ITP patients.
The current gold standard for testing platelet function is light transmission aggregometry, which is biased by platelet count in thrombocytopenic samples. Flow cytometry to detect the activation-dependent expression of glycoproteins represents a widely used alternative, that is barely not affected by the platelet count [
]. We determined platelet function via two independent markers, expression of P-selectin and GPIIb/IIIa activation, as they have most widely been used in previous studies and reflect platelet degranulation as well as the aggregatory potential [
], therefore a potential influence on platelet reactivity cannot be ruled out. Recently, Veninga et al. found, that among large platelets a subpopulation has elevated GPVI expression, which correlates well with HLA-I expression, RNA content, and increased platelet reactivity [
]. Analyzing other surface markers like GPVI expression would have been also interesting in our cohort. P-selectin and GPIIb/IIIa activation in ITP patients was measured at a predefined platelet count (20 × 109/L) to rule out that observed changes solely depend on platelet density.
In conclusion, our data of the so far biggest number of analyzed ITP patients does not support the hypothesis that hyper-reactive platelets in ITP patients lower the bleeding tendency. The platelet activation state was not associated to the bleeding phenotype or previous thromboembolic events, and further undiscovered factors might influence the bleeding and thrombotic risk in ITP. Additionally, we found that patients with non-immunological thrombocytopenia showed patterns of impaired platelet function shown by impaired GPIIb/IIIa activation, which could underly the generally higher risk for bleeding in these patients. An increase of sP-selectin, but not P-selectin surface expression or GPIIb/IIIa activation, was associated with TPO-RA treatment, however, the significance of this finding remains to be elucidated.
CRediT authorship contribution statement
J. Gebhart, I. Pabinger, C. Ay, D. Mehic and J. Machacek designed the study; J. Gebhart, I. Pabinger, C. Ay, D. Mehic, J. Machacek, L. Buresch, T. Schramm, M. Fillitz, B. Dixer, T. Flasch, T. Anderle and A. Rath recruited patients; H. Haslacher processed and stored the samples; A. Kaider performed statistical analyses; B. Eicherberger and A., J. Gebhart, I. Pabinger, A. Assinger, D. Mehic and J. Machacek interpreted the data; D. Mehic, J. Gebhart and J. Machacek wrote the manuscript, which was reviewed, edited and finally approved by all authors.
IP has occasionally received honoraria from CSL Behring, Novartis, Amgen and Sobi for lectures and advisory board meetings. JG received honoraria for lectures and advisory board meetings and research funding for the Medical University of Vienna from CSL Behring, Novartis, Amgen and Sobi. CA received honoraria from Bayer, CSL Behring, NovoNordisk, Pfizer, Roche, Sobi and Takeda for lectures and participation in advisory board meetings. DM received honoraria from CSL Behring.
The ITP biobank was sponsored by a research collaboration with Novartis, and research grants of Amgen and Sobi.