High Frequency of Circulating Endothelial Colony Forming Cells (ECFCs) in Myeloproliferative Neoplasms (MPNs) Is Associated with Diagnosis of Prefibrotic Myelofibrosis, History of Splanchnic Vein Thrombosis, and Vascular Splenomegaly.

High mobilization into peripheral blood of putative endothelial progenitor cells (EPCs) has been proposed as a characteristic of primary myelofibrosis (PMF) among the myeloproliferative neoplasms (MPNs). Endothelial colony forming cells (ECFCs) are a distinct colony type with robust proliferative potential and vessel forming capacity in vivo, and candidates to represent true EPCs. Low number of circulating ECFCs, however, imposes analytical work out in order to use this measurement as a feasible biological indicator of EPC mobilization. Here we report the assessment of ECFCs in patients with MPNs, based on a high number of patients in order to characterize the MPN categories and patients' clinical phenotype that better correlate with such measurement.One hundred thirty nine patients with PMF, 32 patients with essential thrombocythemia (ET) or polycythemia vera (PV) and 22 normal subjects were studied. ECFCs were grown in vitro by plating mononuclear cells, obtained from 40 ml of peripheral blood, onto collagen treated culture plates in the presence of EGM-2MV medium, according to Ingram et al (Blood, 2004). Medium was changed every 3 days. Plates were scored for the presence of ECFC every 2 days, starting from 10 days of culture and up to 28 days. JAK-2V617F mutation was assessed on patients granulocytes by semi-nested PCR.Patients with PMF had a significant higher (p<0.5) median frequency of ECFCs in peripheral blood (0.48 ECFC/10 ml, range 0–6) with respect to normal subjects (0.14, range 0–0.5) and patients with ET (0.25, range 0–1) or PV (0.13, range 0–0.57). Frequency was higher (p<0.05) in patients with prefibrotic PMF (0.86 ECFC/10 ml; n=36) than in patients with fibrotic PMF (0.41/10 ml; n=103). A history of previous splanchnic vein thrombosis (SVT) was associated to statistically higher (p<0.01) ECFC frequency (1.45 ECFC/10 ml; n=21) compared to patients without thrombosis (0.46/10 ml; n=118), as well as to healthy subjects, either with or without history of thrombosis (0.29 and 0.14, respectively, p<0.05 for both). At univariate analysis, increased expression of circulating ECFCs in PMF was associated with younger age (r = − 0.22, p<0.02), diagnosis of prefibrotic myelofibrosis (r = 0.72, p<0.01) and history of splanchnic vein thrombosis (r = 0.68, p<0.001). No correlation was found with the JAK-2 mutational status. At multivariate analysis all the three characteristics maintained an independent association. Since there was no association between the frequency of circulating ECFCs and the time from the onset of SVT, and since patients with an idiopathic SVT had marginal increase in ECFCs frequency compared to healthy subjects, we favored the hypothesis that an increased frequency of circulating ECFC is associated with the myeloproliferative phenotype more than with thrombosis per se. Finally, in the whole population of patients, a direct relationship (r= 0.71, p< 0.02) between the frequency of circulating ECFCs and the index of vascular splenomegaly (spleen index / number of circulating hematopoietic CD34+ cells) was evidenced.Increased mobilization into peripheral blood of ECFCs is a distinctive and new biomarker of prefibrotic myelofibrosis. In addition, this study addresses to the hypothesis that mobilization of ECFCs in MPNs is associated with high risk of SVT and with increased vasculature in the spleen, suggesting a potential role for EPCs in neoangiogenetic processes in this organ.No relevant conflicts of interest to declare.