Purpose:

Thromboembolic events (TE) are the most common complications of myeloproliferative neoplasms (MPN). Clinical parameters, including patient age and mutation status, are used to risk-stratify patients with MPN, but a true biomarker of TE risk is lacking. Protein disulfide isomerase (PDI), an endoplasmic reticulum protein vital for protein folding, also possesses essential extracellular functions, including regulation of thrombus formation. Pharmacologic PDI inhibition prevents thrombus formation, but whether pathologic increases in PDI increase TE risk remains unknown.

Experimental Design:

We evaluated the association of plasma PDI levels and risk of TE in a cohort of patients with MPN with established diagnosis of polycythemia vera (PV) or essential thrombocythemia (ET), compared with healthy controls. Plasma PDI was measured at enrollment and subjects followed prospectively for development of TE.

Results:

A subset of patients, primarily those with JAK2-mutated MPN, had significantly elevated plasma PDI levels as compared with controls. Plasma PDI was functionally active. There was no association between PDI levels and clinical parameters typically used to risk-stratify patients with MPN. The risk of TE was 8-fold greater in those with PDI levels above 2.5 ng/mL. Circulating endothelial cells from JAK2-mutated MPN patients, but not platelets, demonstrated augmented PDI release, suggesting endothelial activation as a source of increased plasma PDI in MPN.

Conclusions:

The observed association between plasma PDI levels and increased risk of TE in patients with JAK2-mutated MPN has both prognostic and therapeutic implications.

Translational Relevance

Although many scoring systems, which incorporate various clinical and laboratory parameters to identify patients with myeloproliferative neoplasms (MPN) at elevated risk of thromboembolic events (TE) are in clinical use, a biomarker that can identify patients with elevated risk of thromboembolic complications, particularly short-term risk, is lacking. Protein disulfide isomerase (PDI) plays a regulatory role in thrombosis and its inhibitors have been shown to prevent TEs in patients with cancer. Pathologic elevations in circulating protein PDI in a subset of patients with MPN and its association with an elevated risk of thromboembolic complications, therefore, opens an entirely new area of investigation for prognostication and prevention of patients with MPN.

Myeloproliferative neoplasms (MPN) are disorders of the bone marrow characterized by excess clonal hematopoiesis resulting in elevated peripheral blood counts. Among seven distinct clinicopathologic entities defined by the 2016 World Health Organization classification system of tumors of the hematopoietic and lymphoid tissues, polycythemia vera (PV), and essential thrombocythemia (ET) are the most common (1, 2). Acquired JAK2 mutations define PV, particularly V617F mutation in exon 14 of JAK2, whereas CALR and MPL mutations are specific to JAK2-unmutated ET (1).

The most common complication associated with PV or ET is an elevated risk of thromboembolic events (TE; refs. 3–6). The risk of TE is about 5- to 7-fold elevated compared with the general population and can involve both arterial and venous circulations (7, 8). TEs are associated with increased morbidity and mortality in the context of ET and PV. Although the association of significantly elevated risk of TE in these disorders is well recognized, the pathophysiology remains poorly defined. Multiple factors, including platelet hyperreactivity, monocytic and endothelial activation, and extracellular vesicles have been reported as potential contributors to TE in MPNs (9–11).

Many clinical parameters, such as patient age, blood cell counts [such as total white blood cell (WBC) count], prior history of TE, and presence of JAK2 mutations, have been found to be associated with elevated risk of TE (8, 12, 13). Consequently, many risk-stratification models, such as the three-step International Prognostic Score for ET (IPSET), which takes into consideration age, prior history of TE, and mutation status, have been validated and commonly used to identify those at higher risk of TE (14, 15). However, a biomarker that can reliably identify patients with MPN with elevated risk of TE, particularly short-term risk of TE, is lacking.

Protein disulfide isomerase (PDI), an endoplasmic reticulum protein critical for protein folding and chaperoning functions, also plays essential roles in the extracellular milieu. We and others have established a critical role of PDI in the pathophysiology of thrombus formation (16–18). PDI, released by stimulated platelets and endothelial cells, regulates thrombus formation and the mechanism by which PDI promotes thrombosis is an area of active research. In a phase 2 clinical study, we observed that inhibitors of PDI reduced thrombin generation and prevented venous thromboembolism, and larger clinical trials are underway in patients with cancer (19, 20). Given the essential role of PDI in thrombosis, we hypothesized that pathologic elevations in circulating PDI are prothrombotic. Measurable plasma PDI has been previously described but has never been prospectively evaluated in any particular disease state, particularly those associated with elevated risk of TEs (21–23). Because of high occurrence of TE, as well as associated platelet and endothelial dysfunction, we chose to explore our hypothesis in MPN.

Study approval

The protocol was approved by the Institutional Review Board at Beth Israel Deaconess Medical Center. Studies were conducted in accordance with the Declaration of Helsinki. Written informed consent was obtained from all study participants before inclusion in the study.

Study design and patients

The study was a prospective, standard-of-care, observational cohort study conducted between 2016 and 2020. Patients were eligible if they were 18 years of age or older and had a prior diagnosis of PV (with a positive JAK2 mutation) or ET (history of platelet count >450 K/μL with either positive JAK2/CALR/MPL mutation or characteristic bone marrow biopsy). Controls were individuals 18 years of age or older, considered to be in good health, and without significant cardiac disease, disorders of hemostasis or thrombosis, active cancer, liver disease (bilirubin >2 mg/dl), or chronic kidney disease (creatinine >2 mg/dl). Clinical and laboratory parameters were recorded for all study subjects and PDI levels estimated at enrollment. Subsequent sample collection was performed at routine 3- to 6-month follow-up visit and patients were followed prospectively for development of TE.

Reagents

The following antibodies were used: mouse monoclonal anti-PDI (RL77; Abcam Cat#ab5484, RRID:AB_304927), rabbit polyclonal anti-PDI (DL-11; Sigma-Aldrich Cat# P7122, RRID:AB_477395), rabbit polyclonal anti–phospho-STAT3 (Cell Signaling Technology Cat# 9134, RRID:AB_331589), rabbit polyclonal anti-STAT3 (Cell Signaling Technology Cat# 9132, RRID:AB_331588), rabbit polyclonal anti-JAK2 (Cell Signaling Technology Cat# 3230, RRID:AB_2128522), mouse monoclonal anti-GAPDH antibodies (Cell Signaling Technology Cat# 97166, RRID:AB_2756824), and FITC-labeled mouse monoclonal anti-PECAM1 (WM59; Bio-Rad Cat# MCA1738T, RRID:AB_1101905). Platelet factor 4 ELISA kit was obtained from R&D Technologies (Cat# DPF40). Human umbilical vein endothelial cells (HUVEC) and endothelial growth medium were obtained from Lonza (Cat# C2519A and CC3156, respectively). All other reagents were obtained from Sigma.

ELISA

For PDI ELISA, 96-well ELISA plates were coated with rabbit anti-PDI (DL-11) at a concentration of 1 μg/mL for 1 hour at room temperature. Plates were then blocked with 3% BSA in PBS for 30 minutes at room temperature, followed by incubation with PDI standards or samples (plasma samples diluted 1/4 and 1/8 with PBS) overnight at 4°C. Biotinylated mouse anti-PDI RL77 was then used as a detection antibody at a concentration of 1 μg/mL for 1 hour at room temperature, followed by streptavidin–horseradish peroxidase at a concentration of 0.1 μg/mL for 45 minutes. Finally, platelets were incubated with TMB substrate for 15–30 minutes and absorbance estimated using a plate reader. PDI ELISA was highly specific to PDI and did not detect other vascular thiol isomerases (Supplementary Fig. S1). vWF ELISA was carried out as previously described (24).

Di-eosin-GSSG reductase assay

A previously described assay (25) was modified to use plasma (1/10 final dilution) as sample. Plasma samples were assayed in duplicates with and without PDI inhibitor quercetin-3-rutinoside (10 μmol/L final concentration) to estimate PDI-dependent reductase activity of the plasma. In addition, controls with recombinant PDI were assayed with and without quercetin-3-rutinoside. The reductase activity obtained from recombinant PDI was used to normalize plasma PDI-dependent reductase activity for each sample to obtain plasma PDI reductase activity in arbitrary units.

Platelet isolation and stimulation

Platelets were isolated from whole blood, washed, and stimulated with 0.1 U/mL of thrombin using a previously published protocol (26). PDI antigen in platelet releasates was measured using PDI ELISA described above.

Cell culture

HUVECs were obtained from Lonza and cultivated using previously published protocol (26). Mycoplasma testing was not performed. Blood outgrowth endothelial cells (BOEC) were cultivated as previously described (27). Mycoplasma testing was not performed. Only passages 2–4 were used for experiments described. For estimation of basal PDI release, cells were incubated with serum-free media for 6 hours followed by estimation of PDI antigen in the media PDI ELISA described above.

Lentiviral transduction

Lentiviral plasmid pCW107-V5-JAK2(V617F) was obtained from Addgene. pCW107-V5-JAK2(WT) plasmid was obtained by mutagenesis of pCW107-V5-JAK2(V617F) plasmid using Phusion site-directed mutagenesis kit using the manufacturer's protocol. Lentiviral particles were generated and HUVEC transduced with lentiviral particles as previously described (28).

Immunoblotting

Cell lysates were treated with IP lysis buffer, resolved using SDS-PAGE, transferred onto nitrocellulose membranes using Trans-Blot transfer system (Bio-Rad), blocked with 3% BSA, and finally immunoblotted with respective antibodies.

Flow cytometry

BOECs were detached using accutase, washed with the endothelial basal media, incubated with FITC-labeled mouse anti–PECAM-1 antibody or isotype control in PBS for 15 minutes, and then PECAM-1 expression estimated using Beckman Coulter Gallios flow cytometer.

Digital PCR

Mutation testing in the BOECs was carried out using JAK2 V617F TaqMan dPCR liquid biopsy assay in the QuantStudio 3D Digital PCR System per the manufacturer's protocol.

Statistical analysis

On the basis of our preliminary data on 5 controls (mean PDI 3.6 ng/mL and SD 2.4 ng/mL) and 15 patients (mean 14.57 ng/mL and SD 13.59 ng/mL), we estimated that a cohort of 25 control individuals and 50 MPN patients would provide greater than 0.95 power to reject the null hypothesis that the PDI levels are equal based on a two-sided alpha of 0.05 using a two-sample t test with unequal variances. These subjects were included in the final analysis.

The unpaired t test assuming unequal SD was used to compare mean plasma PDI of control and MPN cohort. Pearson correlation was used to estimate correlation coefficient between plasma PDI antigen and activity. Log rank was used to compare time-to-event thrombosis rates. Statistical significance was defined as P value of <0.05.

JAK2-mutated MPN patients have high circulating levels of PDI

The baseline characteristics of the study population are shown in Table 1. A total of 65 patients with MPN and 27 controls were enrolled. Among the MPN cohort, 28 had PV and 37 ET; 55 patients had JAK2 V617F mutation, 8 had CALR mutation, and 2 were mutation negative (“triple-negative” MPN). Forty-four (68%) patients with MPN had high-risk disease defined as IPSET score >3. A total of 24 (37%) patients were receiving hydroxyurea, and 53 (81%) receiving aspirin. No patients were receiving ruxolitinib. There were 17 (29%) patients with MPN with a history of thrombosis and 9 (14%) were taking therapeutic anticoagulation. By comparison, 3 (11%) control subjects had a history of thrombosis, with 9 (33%) on aspirin, but none on therapeutic anticoagulation.

Table 1.

Clinical characteristics of the MPN and control cohorts.

CharacteristicControls (27)MPN (65)
Age median (range) 64.5 (21–84) 66 (29–93) 
Gender female (%) 16 (59) 31 (48) 
MPN type — 28 PV 
  37 ET 
Mutation type — 55 JAK2 
  8 CALR 
  2 “triple-negative” 
High-risk MPN (%)  44 (68) 
Cardiovascular disease (%) 2 (7.4) 10 (15) 
History of cancer (%) 5 (18.5) 8 (12) 
History of thrombosis (%) 3 (11.1) 19 (29) 
Aspirin (%) 9 (33.3) 53 (81) 
Hydroxyurea (%) — 24 (37) 
Anticoagulation (%) — 9 (14) 
CharacteristicControls (27)MPN (65)
Age median (range) 64.5 (21–84) 66 (29–93) 
Gender female (%) 16 (59) 31 (48) 
MPN type — 28 PV 
  37 ET 
Mutation type — 55 JAK2 
  8 CALR 
  2 “triple-negative” 
High-risk MPN (%)  44 (68) 
Cardiovascular disease (%) 2 (7.4) 10 (15) 
History of cancer (%) 5 (18.5) 8 (12) 
History of thrombosis (%) 3 (11.1) 19 (29) 
Aspirin (%) 9 (33.3) 53 (81) 
Hydroxyurea (%) — 24 (37) 
Anticoagulation (%) — 9 (14) 

Plasma PDI antigen was measured for all study subjects at time of study enrollment. Plasma PDI antigen was normally distributed in the control cohort (Pearson normality test P = 0.33) with mean PDI level 2.1 ng/mL, median 2.1 ng/mL, range 1.1–3.7 ng/mL, and SD 0.68 ng/mL (±2SD 1.9–2.4 ng/mL; Fig. 1A). The mean plasma PDI antigen in the MPN cohort was 3.3 ng/mL and was significantly higher than the controls (2.1 ng/mL vs. 3.3, P = 0.01). The subgroups of patients with elevated plasma PDI antigen were primarily JAK2-mutated MPN (Fig. 1B). When compared with controls, the JAK2-mutated patients had significantly higher plasma PDI antigen than control (2.1 ng/mL vs. 3.1 ng/mL, P = 0.03).

Figure 1.

Plasma PDI is detectable and elevated in JAK2-mutated MPN. A and B, Scatter plots showing baseline plasma PDI levels in control cohort (n = 27) and MPN cohort (n = 65) or JAK2-mutated (n = 55) and non–JAK2-mutated (n = 10) MPN patients, respectively. Bars in A show mean and ± 2SD; *, P = 0.019. C, Comparison of plasma PDI antigen versus plasma PDI activity measured in 6 MPN patients with the highest plasma PDI antigen levels and 6 patients with low plasma PDI antigen levels (R2 = 0.66; **, P = 0.001). D, Scatter plot showing fold-change in PDI levels obtained during follow-up (median follow-up 3 months) in control (n = 8) and MPN cohorts (n = 41; Bars show mean and ± 2 SD; P = 0.3).

Figure 1.

Plasma PDI is detectable and elevated in JAK2-mutated MPN. A and B, Scatter plots showing baseline plasma PDI levels in control cohort (n = 27) and MPN cohort (n = 65) or JAK2-mutated (n = 55) and non–JAK2-mutated (n = 10) MPN patients, respectively. Bars in A show mean and ± 2SD; *, P = 0.019. C, Comparison of plasma PDI antigen versus plasma PDI activity measured in 6 MPN patients with the highest plasma PDI antigen levels and 6 patients with low plasma PDI antigen levels (R2 = 0.66; **, P = 0.001). D, Scatter plot showing fold-change in PDI levels obtained during follow-up (median follow-up 3 months) in control (n = 8) and MPN cohorts (n = 41; Bars show mean and ± 2 SD; P = 0.3).

Close modal

Excluding one outlier (a “triple-negative” patient with plasma PDI 19.37 ng/dl), patients with non–JAK2-mutated MPN had mean and median plasma PDI of 2.2 ng/dl and 2 ng/dl, respectively (Fig. 1B).

To confirm that circulating PDI antigen was functionally active, we measured plasma PDI-dependent reductase activity. Using a fluorescence-based probe, di-eosin-GSSG, the PDI-dependent liberation of eosin moieties results in a detectable increase in fluorescence (18). We compared PDI activity from patients with the highest plasma PDI levels (N = 6) with those with lower PDI levels (N = 6). As shown in Fig. 1C, PDI antigen was active and correlated with plasma PDI-dependent reductase activity (R2 = 0.66, P = 0.001).

To assess whether PDI levels fluctuate over time, we measured serial PDI levels in 41 patients and 8 control subjects. PDI levels did not significantly change over time with a median follow-up period of 3 months (Fig. 1D). The mean fold-change in PDI levels in MPN group [0.2; 95% confidence interval (CI), −0.11 to 0.52] was not significantly different from that of the control group (−0.18; 95% CI, −1.13 to 0.75; P = 0.32).

Plasma PDI levels and correlation with demographic or laboratory parameters

We determined whether plasma PDI levels in patients with MPN were associated with clinical, laboratory, and thrombosis risk model parameters. Plasma PDI levels did not appear to be influenced by sex or older age (above or below 60 years; Fig. 2AB). Plasma PDI levels were similar among patients with low/intermediate IPSET scores compared with high-risk scores (Fig. 2C; ref. 14). In addition, plasma PDI antigens were not significantly different in patients on aspirin versus those not on aspirin (3.2 ng/mL vs. 3.7 ng/mL; P = 0.6).

Figure 2.

Lack of association between plasma PDI antigen and patient clinical characteristics or laboratory parameters. Scatter plots illustrating plasma PDI antigen in MPN patients according to gender (A), age (B), high or low/intermediate IPSET score (C), and prior history of thromboembolism (D). Comparison of plasma PDI antigen and WBC (E), hematocrit (F), platelet count (G), and PF4 levels (H). R2 = 0 for all comparisons.

Figure 2.

Lack of association between plasma PDI antigen and patient clinical characteristics or laboratory parameters. Scatter plots illustrating plasma PDI antigen in MPN patients according to gender (A), age (B), high or low/intermediate IPSET score (C), and prior history of thromboembolism (D). Comparison of plasma PDI antigen and WBC (E), hematocrit (F), platelet count (G), and PF4 levels (H). R2 = 0 for all comparisons.

Close modal

Similarly, plasma PDI levels were not significantly different among those with or without a prior history of TE (Fig. 2D). Neither WBC count, hematocrit, nor platelet counts correlated with plasma PDI antigen, arguing against PDI representing a surrogate marker of blood count parameters (R2 = 0; Fig. 2EG). To evaluate the possibility that plasma PDI reflected platelet activation (either in vivo or ex vivo), we compared PDI levels with plasma platelet factor 4 (PF4) antigen, a platelet-specific protein that is commonly used as a marker of platelet activation (29). Plasma PDI antigen had poor correlation with plasma PF4 antigen in MPN cohort (R2 = 0; Fig. 2H).

High plasma PDI levels predict risk of TE in MPN

We monitored patients following enrollment for the development of venous or arterial thrombotic events. In median follow-up of 11 months (range 1 to 43 months), there were 8 TEs, with a cumulative incidence of 11.4% per year. Four TEs were arterial (3 cerebrovascular accidents and 1 peripheral arterial thrombosis) and 4 venous (2 deep vein thrombosis, 1 pulmonary embolism, and 1 portal vein thrombosis). Mean plasma PDI among patients who developed TE was 5.4 ng/mL, median 2.8 ng/mL (Supplementary Table S1). To establish a PDI threshold that was most predictive of TE, we performed a receiver operating characteristic analysis and identified 2.5 ng/mL as predictive of TE with a sensitivity of 65% and specificity of 88% for TE (Supplementary Fig. S2). Alternatively, two standard deviations from the mean plasma PDI antigen in control cohort similarly resulted in a 2.5 ng/mL PDI cutoff value. Applying the 2.5 ng/mL threshold to identify high- versus low-risk MPN, the cumulative incidence of TE at 1 year for those patients with lower PDI levels was 5.5% compared with 26.6% in those with higher PDI levels, corresponding with an 8-fold increased risk of TE (HR, 8.02; 95% CI, 1.99–32.16; Log rank P = 0.01; Fig. 3A). Specifically, in JAK2 V617F–mutated MPN, high plasma PDI levels were associated with a 7-fold increased risk of TE (n = 55; HR, 7.3; 95% CI, 1.8–29.5; Log rank P = 0.02). None of the 9 patients on therapeutic anticoagulation at baseline (3 with low and 6 with high plasma PDI levels), experienced a TE during study follow-up. Excluding these patients from the analysis, the risk of TE in patients with high PDI levels (≥2.5 ng/mL) was nearly 10-fold higher (HR, 9.9; 95% CI, 2.42–40.4; Log rank P = 0.005).

Figure 3.

High plasma PDI antigen confers increased risk of thromboembolic events in MPN. A, Kaplan–Meier curves showing elevated incidence of thromboembolic events in MPN patients with high plasma PDI (>2.5 ng/mL) as compared with low plasma PDI (<2.5 ng/mL; HR, 8; 95% CI, 1.99–32.16; P = 0.01). B, Kaplan–Meier curve showing incidence of thrombosis stratified on IPSET score (HR, 3.45; 95% CI, 0.79–15.05; P = 0.36). Addition of plasma PDI level (above or below 2.5 ng/mL) to IPSET score increased the predictive specificity for thromboembolic events (HR, 6.89; 95% CI, 1.56–30.44; P = 0.03).

Figure 3.

High plasma PDI antigen confers increased risk of thromboembolic events in MPN. A, Kaplan–Meier curves showing elevated incidence of thromboembolic events in MPN patients with high plasma PDI (>2.5 ng/mL) as compared with low plasma PDI (<2.5 ng/mL; HR, 8; 95% CI, 1.99–32.16; P = 0.01). B, Kaplan–Meier curve showing incidence of thrombosis stratified on IPSET score (HR, 3.45; 95% CI, 0.79–15.05; P = 0.36). Addition of plasma PDI level (above or below 2.5 ng/mL) to IPSET score increased the predictive specificity for thromboembolic events (HR, 6.89; 95% CI, 1.56–30.44; P = 0.03).

Close modal

Time-to-event analyses were also performed to compare the risk of TE using traditional risk factors, such as WBC count (>15,000/μL vs. <15,000/μL), hematocrit (>45% vs. <45%), and platelet count (>600/μL vs. <600/μL; refs. 8, 13). Specifically, all thrombotic events occurred in patients with WBC below the 15,000/μL threshold and none of the blood count parameters were predictive of a higher TE risk (Supplementary Fig. S3A–S3C). Survival curves were also generated to compare a high IPSET score (>3) to low/intermediate IPSET score as a predictor of TE. Patients with a high IPSET score were at approximately 4-fold higher risk of TE than those with low scores (HR, 3.45; 95% CI, 0.79–15.05), but this was not statistically significant in our cohort (P = 0.36). Adding plasma PDI antigen to patients with high IPSET score improved its predictability of TE risk (HR, 6.89; 95% CI, 1.56–30.44; P = 0.03) but was less discriminating than simply applying PDI levels alone (Fig. 3B). We also measured D-dimer in the plasma samples of the MPN cohort. Patients with high PDI levels had significantly higher median D-dimer concentrations compared with patients with lower PDI levels (265 vs. 176 ng/mL, Mann–Whitney rank sum P = 0.036). These data support potential contribution of PDI in TE in MPN.

Platelet and endothelial sources of PDI in MPN

Endothelium and platelets are the principal sources of vascular thiol isomerases, including PDI (30). Because platelet hyperreactivity has been reported in MPN, we first evaluated whether platelets could be the dominant source of elevated plasma PDI levels observed in JAK2-mutated MPN. Platelets from 3 patients with JAK2 V617F mutation on treatment with aspirin and high plasma PDI (plasma PDI levels of 3.09, 4.47, and 4.77 ng/mL) were isolated and stimulated with 0.5 U/mL thrombin. Platelet-derived PDI was measured in the releasates using ELISA. Platelet-derived PDI levels in releasates of 3 control subjects (plasma PDI levels of 1.64, 1.72, and 1.66 ng/mL) were also measured. There was no difference between thrombin-stimulated platelet PDI release in JAK2-mutated MPN versus controls (P = 0.7; Fig. 4A).

Figure 4.

Elevated endothelial basal PDI release in JAK2-mutated MPN. A, Comparison of PDI antigen measured in releasates and lysates of platelets collected from JAK2-mutated MPN patients and controls. Washed platelets were stimulated with thrombin and buffer collected as releasates. Pelleted platelets then lysed with lysis buffer. PDI antigen determined in releasates and lysates using ELISA (P = 0.7; n = 3). B, Light micrograph showing characteristic endothelial cobblestone-like morphology of blood outgrowth endothelial cells (BOEC). C, BOECs were detached, washed, and then incubated with FITC-labeled anti–PECAM-1 antibody. PECAM-1 expression was then determined by flow cytometry (green, isotype control; blue, control BOEC; red, patient BOEC). D, BOEC genomic DNA was extracted using Qiagen DNeasy blood and tissue kit and digital PCR carried out using JAK2 V617F TaqMan dPCR liquid biopsy assay in the QuantStudio 3D Digital PCR System. Analysis was carried out using QuantStudio 3D Analysis Suite Cloud Software (mutation copies 1 ng/μL ± 95% CI). E, BOEC lysates were resolved by SDS-PAGE and immunoblotted with anti–phospho-STAT3, anti-STAT3, and anti-GAPDH antibodies. F, Cells were grown to confluency and incubated with serum-free media for 6 hours. Media were then collected and vWF Ag determined using vWF ELISA (n = 2; *, P = 0.04). G, Comparison of PDI antigen in releasates and lysates BOECs from JAK2-mutated MPN patients compared with controls. Cells were grown to confluency and incubated with serum-free media for 6 hours. Media were then collected as releasates and cells treated with lysis buffer to prepare lysates. PDI antigen determined in releasates and lysates using ELISA (n = 2; *, P = 0.04).

Figure 4.

Elevated endothelial basal PDI release in JAK2-mutated MPN. A, Comparison of PDI antigen measured in releasates and lysates of platelets collected from JAK2-mutated MPN patients and controls. Washed platelets were stimulated with thrombin and buffer collected as releasates. Pelleted platelets then lysed with lysis buffer. PDI antigen determined in releasates and lysates using ELISA (P = 0.7; n = 3). B, Light micrograph showing characteristic endothelial cobblestone-like morphology of blood outgrowth endothelial cells (BOEC). C, BOECs were detached, washed, and then incubated with FITC-labeled anti–PECAM-1 antibody. PECAM-1 expression was then determined by flow cytometry (green, isotype control; blue, control BOEC; red, patient BOEC). D, BOEC genomic DNA was extracted using Qiagen DNeasy blood and tissue kit and digital PCR carried out using JAK2 V617F TaqMan dPCR liquid biopsy assay in the QuantStudio 3D Digital PCR System. Analysis was carried out using QuantStudio 3D Analysis Suite Cloud Software (mutation copies 1 ng/μL ± 95% CI). E, BOEC lysates were resolved by SDS-PAGE and immunoblotted with anti–phospho-STAT3, anti-STAT3, and anti-GAPDH antibodies. F, Cells were grown to confluency and incubated with serum-free media for 6 hours. Media were then collected and vWF Ag determined using vWF ELISA (n = 2; *, P = 0.04). G, Comparison of PDI antigen in releasates and lysates BOECs from JAK2-mutated MPN patients compared with controls. Cells were grown to confluency and incubated with serum-free media for 6 hours. Media were then collected as releasates and cells treated with lysis buffer to prepare lysates. PDI antigen determined in releasates and lysates using ELISA (n = 2; *, P = 0.04).

Close modal

The endothelium is thought to contribute to vascular inflammation associated with MPN and the JAK2 mutation has been implicated in endothelial dysfunction (10, 31, 32). We isolated BOECs from 2 patients with JAK2-mutated MPN with high PDI levels (3.09 and 4.77 ng/mL) and 1 control (1.64 ng/mL). BOECs had a characteristic cobblestone-like morphology of endothelial cells and expressed endothelial-specific antigen PECAM-1 (Fig. 4BC; ref. 27). Digital PCR confirmed presence of JAK2 V617F mutation in BOECs from both patients (Fig. 4D). Immunoblot analysis of MPN BOECs demonstrated increased levels of phospho-STAT3, indicating augmented JAK-STAT signaling (Fig. 4E). In addition, basal release of vWF antigen was also elevated in MPN BOECs, indicating increased Weibel–Palade body release, as has previously been observed in MPN with the JAK2 V617F mutation (Fig. 4F; ref. 32). Unlike vWF, however, endothelium is thought to store PDI in a granule distinct from Weibel–Palade bodies (33). We therefore evaluated PDI release from BOECs from the 2 patients with JAK2-mutated MPN. Basal PDI release from MPN BOECs was 48% ± 9.4% (P = 0.035) higher than the control BOECs (Fig. 4G). PDI antigen levels in BOEC lysates were similar between the two groups (Fig. 4G).

To confirm increased basal release of PDI from endothelial cells, we expressed wild-type (WT) and mutant JAK2 in HUVECs using lentiviral transduction. The transduced HUVECs expressing mutant JAK2 V617F had higher expression of JAK2 as compared with WT JAK2 (Fig. 5A; refs. 32). Moreover, transduction was associated with JAK–STAT pathway activation in the resting state, as evidenced by STAT3 phosphorylation (Fig. 5B). Increased basal release of vWF Ag was also observed in the transfected HUVECs (Fig. 5C). Similar to the patient-derived BOECs, JAK2 V617F–expressing HUVECs had significantly elevated basal release of PDI (112% ± 26%; P = 0.002) as compared with WT HUVECs, but similar PDI content (Fig. 5D). Treatment of these cells with 100 nmol/L JAK1/2 inhibitor ruxolitinib [IC50 for peripheral blood mononuclear cells 60–130 nmol/L (refs. 34, 35) and plasma Cmax in adults following 15 mg twice-daily administration is >600 nmol/L (ref. 36)], significantly reduced PDI release from JAK2 V617F–expressing HUVECs without affecting WT cells (P = 0.035; Fig. 5E).

Figure 5.

Elevated PDI release from HUVEC-expressing JAK2 V617F; effect of pharmacologic inhibition of JAK2. A,JAK2 V617F and WT HUVEC lysates were subjected to SDS-PAGE and immunoblotted with anti-JAK2 and anti-GAPDH antibodies. B,JAK2 V617F and WT HUVEC lysates were subjected to SDS-PAGE and immunoblotted with phospho-STAT3 and anti-STAT3 antibodies. C, Cells were grown to confluency and incubated with serum-free media for 6 hours. Media were then collected and vWF Ag determined using vWF ELISA (n = 3; **, P = 0.005). D, Cells were grown to confluency and incubated with serum-free media for 6 hours. Media were then collected as releasates and cells treated with lysis buffer to prepare lysates. PDI antigen determined in releasates and lysates using PDI ELISA (n = 3; **, P = 0.002). E, Comparison of PDI antigen in releasates of HUVECs expressing JAK2 V617F mutation, as in D, with and without 100 nmol/L ruxolitinib, compared with HUVECs expressing WT JAK2 (n = 3; *, P = 0.035).

Figure 5.

Elevated PDI release from HUVEC-expressing JAK2 V617F; effect of pharmacologic inhibition of JAK2. A,JAK2 V617F and WT HUVEC lysates were subjected to SDS-PAGE and immunoblotted with anti-JAK2 and anti-GAPDH antibodies. B,JAK2 V617F and WT HUVEC lysates were subjected to SDS-PAGE and immunoblotted with phospho-STAT3 and anti-STAT3 antibodies. C, Cells were grown to confluency and incubated with serum-free media for 6 hours. Media were then collected and vWF Ag determined using vWF ELISA (n = 3; **, P = 0.005). D, Cells were grown to confluency and incubated with serum-free media for 6 hours. Media were then collected as releasates and cells treated with lysis buffer to prepare lysates. PDI antigen determined in releasates and lysates using PDI ELISA (n = 3; **, P = 0.002). E, Comparison of PDI antigen in releasates of HUVECs expressing JAK2 V617F mutation, as in D, with and without 100 nmol/L ruxolitinib, compared with HUVECs expressing WT JAK2 (n = 3; *, P = 0.035).

Close modal

Our data show that plasma PDI is detectable in patients with MPNs and a subgroup of patients with MPN, which consists primarily of those harboring the JAK2 V617F mutation, have pathologic elevations in plasma PDI that predict a higher risk of developing TEs. Moreover, we show that circulating PDI antigen possesses PDI enzyme activity and remains stable for most patients during follow-up. PDI antigen represents a novel, independent biomarker of thrombosis that does not correlate with standard clinical or laboratory parameters commonly used to stratify risk of thrombosis in MPN.

Detectable plasma PDI has been previously reported in humans. Oliveira and colleagues (23), in a cohort of 35 healthy volunteers, demonstrated that plasma PDI antigen and activity was detectable, although at lower concentrations (mean ∼ 540 pg/mL) than in our study. This difference may be secondary to the differences in the assay and antibodies used. In our hands, the commercial ELISA described above did not reliably detect recombinant PDI protein. Before this, Essex and colleagues (21) detected plasma PDI in the 250–1,000 ng/mL range in a small study of volunteers, but details relating to the sample size, subjects, or assay used are not available. PDI has also been previously reported to be present in the plasma proteome quantified using modified aptamers, but was not found to be associated with cardiovascular disease (22).

Clinical parameters, such as age, history of thrombosis, leukocytosis, mutation status, and traditional cardiovascular risk factors are currently in use to classify MPN into low- or high-risk groups for thrombosis, and also to guide cytoreductive treatment (8, 12–15). Risk-stratification models, such the IPSET score, have been devised that incorporate these variables to identify patients with ET at higher risk of TE, but a specific biomarker that can reliably predict risk of TE in MPN is lacking. Moreover, these clinical scoring systems have been developed and validated using retrospective data (12, 14, 37). Markers of hypercoagulability (such as thrombin–antithrombin complexes, prothrombin fragment 1+2, D-dimer), markers of endothelial activation (such as von Willebrand factor, E-selectin), and elevated levels of microparticles have been described in MPN; however, the role of these biomarkers in identifying patients at increased risk of thrombosis is unclear (38–40). Finding of pathologic elevations of plasma PDI particularly in patients with JAK2 V617F–mutated MPN and demonstration of its predictability to identify those at higher risk for TEs prospectively is, therefore, a major finding of this study. Furthermore, addition of plasma PDI levels to high IPSET score improved the latter's specificity to predict TE. JAK2 mutation is known to confer a higher risk of thrombosis in MPN over CALR and MPL mutations. CALR mutations have not been previously reported in the endothelial cells, which may be one factor. In addition, absence of a more generalized JAK/STAT stimulation in CALR-mutated MPN may account for its lower risk of thrombosis, as mutant calreticulin functions exclusively through MPL, the thrombopoietin receptor (41).

Our ex vivo and in vitro results implicate endothelial cells as the source of pathologic elevations of PDI in MPN. The endothelium has been previously implicated in vascular inflammation associated with MPN, and endothelium-specific JAK2-mutated murine models revealed significant endothelial dysfunction associated with abnormalities in coagulation (10, 31, 32). Our data show that endothelial exocytosis of PDI is elevated from BOECs derived from patients with JAK2-mutated MPN with high plasma PDI levels and is increased in primary endothelial cells harboring JAK2 V617F mutation. BOECs are a rare population of circulating mononuclear cells that possess endothelial markers and features, and have been widely used in vascular biology, particularly in evaluation of disorders of hemostasis and thrombosis (42–47). BOECs from patients with JAK2-mutated MPN demonstrate augmented endothelial secretion, including that of inflammatory cytokines (10). Elevated plasma PDI may therefore reflect endothelial activation known to occur in MPN thereby contributing to its prothrombotic phenotype. The premise that increased PDI release is associated with the JAK2 V617F mutation is supported by the observation that primary endothelial cells from patients with JAK2 V617F mutation, and cultured endothelial cells transduced with JAK2 V617F mutation, also demonstrate increased PDI release. In contrast, our studies indicate that the platelets, a major source of vascular thiol isomerases, do not release excessive PDI in patients with MPN and high circulating PDI levels. Although aspirin is not known to affect thrombin-mediated platelet activation, considering that the studies we performed were in patients with MPN-receiving aspirin, we cannot exclude the possibility that excess PDI is released from platelets in patients with MPN not taking aspirin (48).

Although circulating PDI has never previously been linked to a prothrombotic state in humans, the critical role that PDI serves in pathophysiology of thrombosis is well recognized (16, 30). Stimulation of endothelial cells results in PDI release, which precedes platelet accumulation and fibrin generation at the site of vessel wall injury. The exact molecular mechanism by which PDI regulates thrombosis is not known, but various downstream effectors have been reported, such as αIIbβ3, GPIbα, tissue factor, vitronectin, and platelet factor V (17, 18). Mouse models of platelet-specific knockdown or conditional expression of PDI that lacks oxidoreductase activity demonstrate impaired thrombus formation (49, 50). Inhibition of PDI with antibodies or specific inhibitors also prevents thrombus formation (20, 25). As such, PDI inhibitors are being evaluated in clinical trials as novel anticoagulants with phase 2 data demonstrating efficacy in reducing thrombin generation and preventing venous thromboembolism in patients with cancer (19).

The limitations of our study include its relatively small sample size. Yet unlike many previous studies evaluating the risk of thrombosis in MPN, patients were followed prospectively for the development of TE (12, 14, 51, 52). External validation using larger sample repositories will be required before consideration for clinical applications. We show that PDI levels stay stable over 3- to 6-month follow-up, but we do not know whether PDI levels change more significantly over time as may occur with disease progression. This study was also underpowered to determine any statistical differences between JAK2 and non–JAK2-mutated MPN, specifically CALR mutation. This study was also underpowered to determine the effect of hematocrit, which has been previously shown in clinical trials to predict thrombosis (53). We also did not explore other prothrombotic disease states such as cancer and thus cannot conclude whether PDI elevations are unique to MPN. Although PDI plays a critical role in regulating thrombus formation in vivo (16), PDI elevation may represent a state of endothelial activation associated with MPN, and additional studies, including animal models of thrombosis, will be required to more definitively establish a causal relationship between increased circulating PDI and MPN-associated hypercoagulability. Because pathologic elevations in circulating PDI are predictive of thrombosis in MPN, pharmacologic inhibition of the JAK–STAT pathway to decrease PDI release, or direct therapeutic targeting of PDI by reduction of its extracellular oxidoreductase activity would be attractive approaches toward preventing thrombosis, specifically in this high-risk population.

R. Flaumenhaft reports other support from Platelet Diagnostics and personal fees from Moderna outside the submitted work. J.I. Zwicker reports grants from Incyte during the conduct of the study, as well as grants from Quercegen and personal fees from Pfizer/BMS, Daiichi, and Parexel outside the submitted work; in addition, J.I. Zwicker has a patent for 16/696,280 pending to BIDMC and PCT/US2020/064101 pending to BIDMC. No disclosures were reported by the other authors.

A.V. Sharda: Data curation, formal analysis, investigation, writing–original draft, project administration, writing–review and editing. T. Bogue: Data curation, investigation. A. Barr: Investigation. L.M. Mendez: Investigation, writing–review and editing. R. Flaumenhaft: Conceptualization, methodology, writing–review and editing. J.I. Zwicker: Conceptualization, resources, formal analysis, supervision, funding acquisition, writing–review and editing.

This study was supported by a sponsored research grant from the Incyte Corporation (to J.I. Zwicker), and 1U01HL143365-01 (to J.I. Zwicker and R. Flaumenhaft) and R35 HL135775 (to R. Flaumenhaft) from NHLBI.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Arber
DA
,
Orazi
A
,
Hasserjian
R
,
Thiele
J
,
Borowitz
MJ
,
Le Beau
MM
, et al
The 2016 revision to the World Health Organization classification of myeloid neoplasms and acute leukemia
.
Blood
2016
;
127
:
2391
405
.
2.
Barbui
T
,
Thiele
J
,
Gisslinger
H
,
Kvasnicka
HM
,
Vannucchi
AM
,
Guglielmelli
P
, et al
The 2016 WHO classification and diagnostic criteria for myeloproliferative neoplasms: document summary and in-depth discussion
.
Blood Cancer J
2018
;
8
:
15
.
3.
Arachchillage
DR
,
Laffan
M
. 
Pathogenesis and management of thrombotic disease in myeloproliferative neoplasms
.
Semin Thromb Hemost
2019
;
45
:
604
11
.
4.
Griesshammer
M
,
Kiladjian
JJ
,
Besses
C
. 
Thromboembolic events in polycythemia vera
.
Ann Hematol
2019
;
98
:
1071
82
.
5.
Koschmieder
S
. 
How I manage thrombotic/thromboembolic complications in myeloproliferative neoplasms
.
Hamostaseologie
2020
;
40
:
47
53
.
6.
Stein
BL
,
Martin
K
. 
From Budd-Chiari syndrome to acquired von Willebrand syndrome: thrombosis and bleeding complications in the myeloproliferative neoplasms
.
Blood
2019
;
134
:
1902
11
.
7.
Hultcrantz
M
,
Bjorkholm
M
,
Landgren
O
,
Kristinsson
SY
,
Andersson
TML
. 
Risk for arterial and venous thrombosis in patients with myeloproliferative neoplasms
.
Ann Intern Med
2018
;
169
:
268
.
8.
Barbui
T
,
Carobbio
A
,
Rumi
E
,
Finazzi
G
,
Gisslinger
H
,
Rodeghiero
F
, et al
In contemporary patients with polycythemia vera, rates of thrombosis and risk factors delineate a new clinical epidemiology
.
Blood
2014
;
124
:
3021
3
.
9.
Fel
A
,
Lewandowska
AE
,
Petrides
PE
,
Wisniewski
JR
. 
Comparison of proteome composition of serum enriched in extracellular vesicles isolated from polycythemia vera patients and healthy controls
.
Proteomes
2019
;
7
:
20
.
10.
Guadall
A
,
Lesteven
E
,
Letort
G
,
Awan Toor
S
,
Delord
M
,
Pognant
D
, et al
Endothelial cells harbouring the JAK2V617F mutation display pro-adherent and pro-thrombotic features
.
Thromb Haemost
2018
;
118
:
1586
99
.
11.
Marin Oyarzun
CP
,
Glembotsky
AC
,
Goette
NP
,
Lev
PR
,
De Luca
G
,
Baroni Pietto
MC
, et al
Platelet toll-like receptors mediate thromboinflammatory responses in patients with essential thrombocythemia
.
Front Immunol
2020
;
11
:
705
.
12.
Carobbio
A
,
Ferrari
A
,
Masciulli
A
,
Ghirardi
A
,
Barosi
G
,
Barbui
T
. 
Leukocytosis and thrombosis in essential thrombocythemia and polycythemia vera: a systematic review and meta-analysis
.
Blood Adv
2019
;
3
:
1729
37
.
13.
Carobbio
A
,
Thiele
J
,
Passamonti
F
,
Rumi
E
,
Ruggeri
M
,
Rodeghiero
F
, et al
Risk factors for arterial and venous thrombosis in WHO-defined essential thrombocythemia: an international study of 891 patients
.
Blood
2011
;
117
:
5857
9
.
14.
Barbui
T
,
Finazzi
G
,
Carobbio
A
,
Thiele
J
,
Passamonti
F
,
Rumi
E
, et al
Development and validation of an international prognostic score of thrombosis in world health organization-essential thrombocythemia (IPSET-thrombosis)
.
Blood
2012
;
120
:
5128
33
.
15.
Grinfeld
J
. 
Prognostic models in the myeloproliferative neoplasms
.
Blood Rev
2020
;
42
:
100713
.
16.
Sharda
A
,
Furie
B
. 
Regulatory role of thiol isomerases in thrombus formation
.
Expert Rev Hematol
2018
;
11
:
437
48
.
17.
Bowley
SR
,
Fang
C
,
Merrill-Skoloff
G
,
Furie
BC
,
Furie
B
. 
Protein disulfide isomerase secretion following vascular injury initiates a regulatory pathway for thrombus formation
.
Nat Commun
2017
;
8
:
14151
.
18.
Stopa
JD
,
Neuberg
D
,
Puligandla
M
,
Furie
B
,
Flaumenhaft
R
,
Zwicker
JI
. 
Protein disulfide isomerase inhibition blocks thrombin generation in humans by interfering with platelet factor V activation
.
JCI Insight
2017
;
2
:
e89373
.
19.
Zwicker
JI
,
Schlechter
BL
,
Stopa
JD
,
Liebman
HA
,
Aggarwal
A
,
Puligandla
M
, et al
Targeting protein disulfide isomerase with the flavonoid isoquercetin to improve hypercoagulability in advanced cancer
.
JCI Insight
2019
;
4
:
e125851
.
20.
Jasuja
R
,
Passam
FH
,
Kennedy
DR
,
Kim
SH
,
van Hessem
L
,
Lin
L
, et al
Protein disulfide isomerase inhibitors constitute a new class of antithrombotic agents
.
J Clin Invest
2012
;
122
:
2104
13
.
21.
Essex
DW
,
Miller
A
,
Swiatkowska
M
,
Feinman
RD
. 
Protein disulfide isomerase catalyzes the formation of disulfide-linked complexes of vitronectin with thrombin-antithrombin
.
Biochemistry
1999
;
38
:
10398
405
.
22.
Ganz
P
,
Heidecker
B
,
Hveem
K
,
Jonasson
C
,
Kato
S
,
Segal
MR
, et al
Development and validation of a protein-based risk score for cardiovascular outcomes among patients with stable coronary heart disease
.
JAMA
2016
;
315
:
2532
41
.
23.
Oliveira
PVS
,
Garcia-Rosa
S
,
Sachetto
ATA
,
Moretti
AIS
,
Debbas
V
,
De Bessa
TC
, et al
Protein disulfide isomerase plasma levels in healthy humans reveal proteomic signatures involved in contrasting endothelial phenotypes
.
Redox Biol
2019
;
22
:
101142
.
24.
Sharda
AV
,
Barr
AM
,
Harrison
JA
,
Wilkie
AR
,
Fang
C
,
Mendez
LM
, et al
VWF maturation and release are controlled by 2 regulators of Weibel-Palade body biogenesis: exocyst and BLOC-2
.
Blood
2020
;
136
:
2824
37
.
25.
Lin
L
,
Gopal
S
,
Sharda
A
,
Passam
F
,
Bowley
SR
,
Stopa
J
, et al
Quercetin-3-rutinoside inhibits protein disulfide isomerase by binding to Its b'x domain
.
J Biol Chem
2015
;
290
:
23543
52
.
26.
Sharda
A
,
Kim
SH
,
Jasuja
R
,
Gopal
S
,
Flaumenhaft
R
,
Furie
BC
, et al
Defective PDI release from platelets and endothelial cells impairs thrombus formation in Hermansky-Pudlak syndrome
.
Blood
2015
;
125
:
1633
42
.
27.
Martin-Ramirez
J
,
Hofman
M
,
van den Biggelaar
M
,
Hebbel
RP
,
Voorberg
J
. 
Establishment of outgrowth endothelial cells from peripheral blood
.
Nat Protoc
2012
;
7
:
1709
15
.
28.
De Meyer
SF
,
Vanhoorelbeke
K
,
Chuah
MK
,
Pareyn
I
,
Gillijns
V
,
Hebbel
RP
, et al
Phenotypic correction of von Willebrand disease type 3 blood-derived endothelial cells with lentiviral vectors expressing von Willebrand factor
.
Blood
2006
;
107
:
4728
36
.
29.
Sharda
A
,
Flaumenhaft
R
. 
The life cycle of platelet granules
.
F1000Res
2018
;
7
:
236
.
30.
Flaumenhaft
R
,
Furie
B
. 
Vascular thiol isomerases
.
Blood
2016
;
128
:
893
901
.
31.
Etheridge
SL
,
Roh
ME
,
Cosgrove
ME
,
Sangkhae
V
,
Fox
NE
,
Chen
J
, et al
JAK2V617F-positive endothelial cells contribute to clotting abnormalities in myeloproliferative neoplasms
.
Proc Natl Acad Sci U S A
2014
;
111
:
2295
300
.
32.
Guy
A
,
Gourdou-Latyszenok
V
,
Le Lay
N
,
Peghaire
C
,
Kilani
B
,
Dias
JV
, et al
Vascular endothelial cell expression of JAK2(V617F) is sufficient to promote a pro-thrombotic state due to increased P-selectin expression
.
Haematologica
2019
;
104
:
70
81
.
33.
Jasuja
R
,
Furie
B
,
Furie
BC
. 
Endothelium-derived but not platelet-derived protein disulfide isomerase is required for thrombus formation in vivo
.
Blood
2010
;
116
:
4665
74
.
34.
Quintas-Cardama
A
,
Vaddi
K
,
Liu
P
,
Manshouri
T
,
Li
J
,
Scherle
PA
, et al
Preclinical characterization of the selective JAK1/2 inhibitor INCB018424: therapeutic implications for the treatment of myeloproliferative neoplasms
.
Blood
2010
;
115
:
3109
17
.
35.
Walker
K
. 
Inflammation Research Association–15th international conference advances in asthma and COPD and other inflammatory diseases
.
IDrugs
2008
;
11
:
863
5
.
36.
Shi
JG
,
Chen
X
,
McGee
RF
,
Landman
RR
,
Emm
T
,
Lo
Y
, et al
The pharmacokinetics, pharmacodynamics, and safety of orally dosed INCB018424 phosphate in healthy volunteers
.
J Clin Pharmacol
2011
;
51
:
1644
54
.
37.
Barbui
T
,
Carobbio
A
,
Ferrari
A
. 
Leukocytosis and thrombosis in polycythemia vera: can clinical trials settle the debate?
Blood Adv
2019
;
3
:
3951
2
.
38.
Falanga
A
,
Marchetti
M
,
Evangelista
V
,
Vignoli
A
,
Licini
M
,
Balicco
M
, et al
Polymorphonuclear leukocyte activation and hemostasis in patients with essential thrombocythemia and polycythemia vera
.
Blood
2000
;
96
:
4261
6
.
39.
Trappenburg
MC
,
van Schilfgaarde
M
,
Marchetti
M
,
Spronk
HM
,
ten Cate
H
,
Leyte
A
, et al
Elevated procoagulant microparticles expressing endothelial and platelet markers in essential thrombocythemia
.
Haematologica
2009
;
94
:
911
8
.
40.
Arellano-Rodrigo
E
,
Alvarez-Larran
A
,
Reverter
JC
,
Colomer
D
,
Villamor
N
,
Bellosillo
B
, et al
Platelet turnover, coagulation factors, and soluble markers of platelet and endothelial activation in essential thrombocythemia: relationship with thrombosis occurrence and JAK2 V617F allele burden
.
Am J Hematol
2009
;
84
:
102
8
.
41.
How
J
,
Hobbs
GS
,
Mullally
A
. 
Mutant calreticulin in myeloproliferative neoplasms
.
Blood
2019
;
134
:
2242
8
.
42.
Fuchs
S
,
Dohle
E
,
Kolbe
M
,
Kirkpatrick
CJ
. 
Outgrowth endothelial cells: sources, characteristics and potential applications in tissue engineering and regenerative medicine
.
Adv Biochem Eng Biotechnol
2010
;
123
:
201
17
.
43.
Gritti
G
,
Cortelezzi
A
,
Bucciarelli
P
,
Rezzonico
F
,
Lonati
S
,
La Marca
S
, et al
Circulating and progenitor endothelial cells are abnormal in patients with different types of von Willebrand disease and correlate with markers of angiogenesis
.
Am J Hematol
2011
;
86
:
650
6
.
44.
Lee
KW
,
Lip
GY
,
Tayebjee
M
,
Foster
W
,
Blann
AD
. 
Circulating endothelial cells, von Willebrand factor, interleukin-6, and prognosis in patients with acute coronary syndromes
.
Blood
2005
;
105
:
526
32
.
45.
Lin
Y
,
Weisdorf
DJ
,
Solovey
A
,
Hebbel
RP
. 
Origins of circulating endothelial cells and endothelial outgrowth from blood
.
J Clin Invest
2000
;
105
:
71
7
.
46.
Solovey
A
,
Lin
Y
,
Browne
P
,
Choong
S
,
Wayner
E
,
Hebbel
RP
. 
Circulating activated endothelial cells in sickle cell anemia
.
N Engl J Med
1997
;
337
:
1584
90
.
47.
Starke
RD
,
Paschalaki
KE
,
Dyer
CE
,
Harrison-Lavoie
KJ
,
Cutler
JA
,
McKinnon
TA
, et al
Cellular and molecular basis of von Willebrand disease: studies on blood outgrowth endothelial cells
.
Blood
2013
;
121
:
2773
84
.
48.
Taylor
ML
,
Ilton
MK
,
Misso
NL
,
Watkins
DN
,
Hung
J
,
Thompson
PJ
. 
The effect of aspirin on thrombin stimulated platelet adhesion receptor expression and the role of neutrophils
.
Br J Clin Pharmacol
1998
;
46
:
139
45
.
49.
Kim
K
,
Hahm
E
,
Li
J
,
Holbrook
LM
,
Sasikumar
P
,
Stanley
RG
, et al
Platelet protein disulfide isomerase is required for thrombus formation but not for hemostasis in mice
.
Blood
2013
;
122
:
1052
61
.
50.
Zhou
J
,
Wu
Y
,
Wang
L
,
Rauova
L
,
Hayes
VM
,
Poncz
M
, et al
The C-terminal CGHC motif of protein disulfide isomerase supports thrombosis
.
J Clin Invest
2015
;
125
:
4391
406
.
51.
Haider
M
,
Gangat
N
,
Lasho
T
,
Abou Hussein
AK
,
Elala
YC
,
Hanson
C
, et al
Validation of the revised international prognostic score of thrombosis for essential thrombocythemia (IPSET-thrombosis) in 585 Mayo Clinic patients
.
Am J Hematol
2016
;
91
:
390
4
.
52.
Palandri
F
,
Polverelli
N
,
Catani
L
,
Ottaviani
E
,
Baccarani
M
,
Vianelli
N
. 
Impact of leukocytosis on thrombotic risk and survival in 532 patients with essential thrombocythemia: a retrospective study
.
Ann Hematol
2011
;
90
:
933
8
.
53.
Marchioli
R
,
Finazzi
G
,
Specchia
G
,
Cacciola
R
,
Cavazzina
R
,
Cilloni
D
, et al
Cardiovascular events and intensity of treatment in polycythemia vera
.
N Engl J Med
2013
;
368
:
22
33
.