Search for a command to run...
Immune thrombocytopenia (ITP) is an autoimmune disease caused by auto-antibodies against platelets. Platelet function can be impaired in ITP, which can contribute to bleeding (van Bladel et al, 2014; Frelinger et al, 2015; Suntsova et al, 2017; Frelinger et al, 2018). However, not all studies support this (Connor et al, 2013), and the effects of new thrombopoietin mimetics romiplostim and eltrombopag on platelet function in ITP are unclear (Psaila et al, 2012; Haselboeck et al, 2013; Gerrits et al, 2015; Ignatova et al, 2019). We therefore investigated evolution of platelet function and bleeding in 31 adult chronic ITP patients (Table S1) treated with romiplostim over the course of 6 months. Prior to treatment, 9 patients had bleeding complications and 22 patients were asymptomatic. The platelet counts were not significantly different between these two groups (Fig 1A). Thromboelastography lag times (parameters R and K) were prolonged, while maximum amplitude and angle α were decreased in the bleeding patients compared with the non-bleeding ones (Fig 1B, C; Figure S1P–Q, S2). Flow cytometry parameters (forward scatter-height [FSC-H], sideways scatter-height [SSC-H]) and total glycoprotein levels (CD42b, CD61) of non-stimulated platelets were significantly increased in the ITP platelets compared with healthy controls (Fig 1D; Figure S1A–D, S3). Among these parameters, FSC and CD42b elevation was significantly higher in the bleeding group. All platelet activation markers (PAC1 for integrins, CD62 for α-granules, procoagulant platelets) were significantly increased in the resting platelets of the patients (Fig 1E, F; Figure S1E). CD62P and procoagulant platelets in the non-stimulated samples were higher in the bleeding group. In contrast, platelet activation responses to potent stimulation with thrombin plus collagen receptor agonists were not essentially affected in ITP, and did not differ between the bleeding and non-bleeding groups (Fig 2I; Figure S1F–I). To summarize, these data indicate that platelets in untreated chronic ITP are enlarged and pre-activated, and both these changes are associated with bleeding. Although changes in thromboelastography parameters are potentially interesting, additional research is needed to identify the responsible mechanisms, in particular, because the sensitivity of thromboelastography to platelet function is known to be limited. Over the 6-month treatment with romiplostim, platelet function underwent significant changes (Fig 2; Figure S4). Platelet size parameters (FSC) and pre-activation (PAC1 and procoagulant platelets) were gradually decreased towards the normal values. Electron microscopy confirmed disappearance of pre-activation on therapy (Fig 2C; Figure S5A–D). For some of the patients, changes in the platelet function parameters and cessation of bleeding were observed even without platelet count increase (Figures S6, S7, S8). However, the data are not sufficient to study potential association of the platelet size and pre-activation changes on romiplostim with bleeding risks independently of the platelet count. The mechanisms of these changes are intriguing and require additional exploration. Immunosuppressive therapy was previously reported to decrease platelet size in ITP (Winkler et al, 2012), which could be explained by decreased clearance and therefore relative increase of older platelets. However, this explanation should not be applicable to romiplostim, which is supposed to act by affecting platelet production and not circulation. In order to determine independent platelet function parameters among those sensitive to bleeding risks, we performed correlation analysis (Figure S9), which identified three groups: "thromboelastography" group (either R, K, or alpha), "platelet size" group (FSC-H, SSC-H, CD42b), "platelet pre-activation" group (CD62P). Based on this, we designed a simple bleeding risk calculator (one point added to the risk score for exceeding K, CD62P and FSC-H threshold values of 13·8, 7·5 and 170, respectively), yielding a gradual bleeding risk stratification with an area under the curve of 0·92 ± 0·06, P < 0·001 (Figure S10F). Additional prospective studies are required to evaluate the usefulness of this bleeding risk evaluation tool. The picture of the untreated chronic ITP platelets suggested by our results agrees well with the view that they are enlarged and pre-activated (Frelinger et al, 2015; Frelinger et al, 2018), and that these two features correlate with bleeding independently of the platelet count. However, in contrast to these reports, and in agreement with others (Connor et al, 2013), we did not observe impairment in the ITP platelet functional responses to stimulation. In order to better probe platelet function in ITP, we recruited a second cohort of 26 patients treated with and without romiplostim (Table S2). For this cohort, we additionally studied platelet responses to weak ADP stimulation (Fig 2E, F; Figure S11L, M), thrombus formation under arterial conditions (Fig 2D; Figure S5E) and the levels of platelet-associated immunoglobulins (Figure S5F, G). However, neither of these experiments revealed any functional impairment in the platelets of the untreated ITP patients compared with either romiplostim-treated or normal ones. The main findings of our study could be summarized in the following manner: (i) platelets in chronic adult ITP are enlarged and pre-activated; (ii) increased flow cytometry size parameters and pre-activation markers are associated with bleeding risks independently of the platelet count; (iii) these parameters gradually become decreased on treatment with romiplostim; (iv) main functional responses of healthy, untreated ITP and treated ITP platelets to stimulation are similar; (v) thromboelastography parameters in ITP are changed and could be used for bleeding risk stratification. The present study was partially supported by a grant from Amgen. Development of the flow cytometry and flow perfusion chamber assays for evaluation of platelet function in ITP was supported by grants from the endowment foundation "Doctors, Innovations, Science for Children" and by the Russian Foundation for Basic Research [Grants 17-00-00141 (17-00-00140),18-34-20026 and 17-04-01309]. Electron microscopy was supported by the Russian Science Foundation grant 17-15-01290. Equipment for this study was provided through the Lomonosov Moscow State University Development Program (PNR 5.13). AAI, IAD, VVP, MMP, FIA, AAM, AVM, GAN and MAP planned research and analysed data. AAI, IAD, DMP carried out flow cytometry experiments. EAS, AVP performed thromboelastography measurements. SGK, ONS performed PA-IgG and glycocalicin measurements. Transmission electron microscopy experiments were done by SIO, OSS and IIK. AAR carried out flow perfusion chambers experiments. MAP, AAI wrote the manuscript in consultation with SGK, MMP, IIK, FIA, AVM and with contributions of all authors. All authors read and approved the final manuscript. Figure S1. Status of platelets in ITP before romiplostim treatment. All notation and conditions are as in Fig 1. Figure S2. Typical TEG curves of the ITP patients before treatment. Figure S3. Platelet flow cytometry in ITP. Figure S4. Dynamics of platelet function in ITP. Figure S5. Platelet function in ITP: potential physiologicial significance. Figure S6. Romiplostim effect on platelet function is independent of platelet count. Figure S7. Relationship between laboratory parameters and platelet count in ITP. Figure S8. Individual dynamics. Figure S9. Correlation of the laboratory parameters. Figure S10. Possibility of bleeding risk stratification. Figure S11. Platelet function in ITP: weak ADP stimulation. Table S1. Patients of the first cohort (the 6-month romiplostim protocol)a. Table S2. Patients of the second cohort (the ITP mechanisms study)a. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.