Increased Thromboinflammation Is Associated with High Thrombotic Risk in Patients with Myeloproliferative Neoplasms: A Multicentric Study of 394 Cases By the French Intergroup of Myeloproliferative Neoplasms (FIM)

Introduction

Thrombosis is a major complication of Philadelphia-chromosome-negative myeloproliferative neoplasms (MPN), responsible for a high morbidity and mortality. Assessing thrombotic risk is particularly important to drive cytoreductive drugs initiation in aim to reduce thrombosis occurrence and subsequent mortality. Recent findings highlight a major role of thromboinflammation in the physiopathology of MPN-associated thrombosis, including activation of neutrophils, monocytes, platelets, and endothelial cells. We here hypothesized that thromboinflammation markers can be used as biomarkers of thrombosis risk during MPN.

Objectives

The main objective of our study was to evaluate the association between thromboinflammation markers and thrombotic risk scores in MPN patients. Secondary objectives were to assess thromboinflammation markers in MPN subgroups (PV, ET) and to evaluate JAK2V617F impact on thromboinflammation markers’ concentrations.

Material and methods

We conducted a multicentric, prospective study across 21 hospital centers in France between 2018 and 2022 and recruited newly-diagnosed cytoreductive therapy-free PV and ET patients. All individuals participating were concurrently registered in the national clinico-biological cohort FIM-BANK (Angers University Hospital). We measured plasmatic concentrations of (i) NETosis markers with citrullinated H3 (H3cit), deoxyribonucleic acid (DNA)-H3cit, myeloperoxidase (MPO)-DNA, (ii) neutrophil activation marker with S100A8/A9, (iii) monocyte activation marker with soluble tissue factor (TF), (iv) platelet activation marker with soluble CD62-P, and (v) endothelial cells activation marker with thrombomodulin (TM) and von Willebrand factor (vWF). We analyzed the correlation between these markers and currently available thrombotic risk scores. Conventional risk score and Venous Thrombosis Score (VETS) were evaluated in all patients, whereas IPSET-Thrombosis and revised IPSET-Thrombosis score were evaluated in ET patients only. Statistical analysis included Student’s t-test or Mann-Whitney test for comparison between two groups and Spearman’s correlation for analysis of correlation. Adjustment of p-values for multiple comparisons was performed using Benjamini-Hochberg method.

Results

394 MPN patients were included: 270 patients (69%) with ET and 124 (31%) with PV. 318 patients (81.5%) were JAK2V617F+. 49 ET patients (18.2%) were positive for CALR mutation and 12 (4.5%) were positive for MPL mutation. 43 patients (10.9%) had an history of venous thrombosis and 71 (18.0%) had a previous arterial thrombosis.

We first looked for an association between thromboinflammation markers and thrombotic risk using the conventional thrombotic risk score in the whole cohort. We observed higher concentrations of H3cit (p=0.0001), S100A8/A9 (p=0.003), TF (p=0.0002), TM (p= 0.004) in high versus low-risk patients. S100A8/A9 concentrations were also associated with venous thrombotic risk (p=0.007) when using VETS score in all MPN patients.

We then restricted our analysis to ET patients and also observed higher concentrations of H3cit (p=0.0001), S100A8/A9 (p=0.003), TF (p=0.015), TM (p= 0.001) in high versus low-risk patients using the conventional thrombotic risk score. These markers also correlated with the IPSET-Thrombosis score, specific to ET.

Analysis in PV patients revealed higher concentrations of S100A8/A9 in patients with previous venous thrombosis (p=0.017). Patients at high risk of venous thrombosis using VETS score also had higher S100A8/A9 (p=0.017) and VWF (p=0.017) concentrations, suggesting a specific role of thromboinflammation in venous thrombosis during PV.

Interestingly, JAK2V617F allele burden was correlated with H3cit (r=0.17, p=0.05), S100A8/A9 (r=0.26, p=0.00007), TF (r=0.13, p=0.048), TM (r=0.19, p=0.027) and sCD62P (r=0.21, p=0.001) concentrations, suggesting that the presence of JAK2V617F is directly involved in the activation of neutrophils, monocytes, endothelial cells, and platelets therefore promoting thromboinflammation.

Conclusion

This study is the largest analysis demonstrating the association of thromboinflammation markers with thrombotic risk in MPN patients. Future research should aim to integrate these markers into clinico-biological scores to better refine thrombosis risk assessments in MPN patients.