Assessment of Endothelial Injury and Pro-Coagulant Activity Using Circulating Microvesicles in Survivors of Allogeneic Hematopoietic Cell Transplantation

(1) Background: survivors of allogeneic hematopoietic cell transplantation (alloHCT) suffer from morbidity and mortality due to cardiovascular events. We hypothesized that vascular injury and pro-coagulant activity are evident in alloHCT survivors without existing alloHCT complications or relapse. (2) Methods: we enrolled consecutive adult alloHCT survivors without established cardiovascular disease and control individuals matched for traditional cardiovascular risk factors (January–December 2019). Circulating microvesicles (MVs) of different cellular origins (platelet, erythrocyte, and endothelial) were measured by a standardized flow cytometry protocol as novel markers of vascular injury and pro-coagulant activity. (3) Results: we recruited 45 survivors after a median of 2.3 (range 1.1–13.2) years from alloHCT, and 45 controls. The majority of patients suffered from acute (44%) and/or chronic (66%) graft-versus-host disease (GVHD). Although the two groups were matched for traditional cardiovascular risk factors, alloHCT survivors showed significantly increased platelet and erythrocyte MVs compared to controls. Within alloHCT survivors, erythrocyte MVs were significantly increased in patients with a previous history of thrombotic microangiopathy. Interestingly, endothelial MVs were significantly increased only in alloHCT recipients of a myeloablative conditioning. Furthermore, MVs of different origins showed a positive association with each other. (4) Conclusions: endothelial dysfunction and increased thrombotic risk are evident in alloHCT recipients long after alloHCT, independently of traditional cardiovascular risk factors. An apparent synergism of these pathophysiological processes may be strongly involved in the subsequent establishment of cardiovascular disease.


Introduction
Allogeneic hematopoietic cell transplantation (alloHCT) is the only curative option for various hematologic malignant and non-malignant diseases [1]. Although major advances in the standard

Study Population
We studied 45 patients after a median of 2.3 (range 1.1-13.2) years from alloHCT, who were free from established CV disease. Patients' pre-transplant and transplant characteristics are shown in Table 1. It should be noted that the majority of alloHCT recipients had received treatment for acute (44%) and/or chronic GVHD (66%) before enrollment in the study. When studied, alloHCT recipients were not on immunosuppressive treatment and had no signs or symptoms of severe complications, active malignancy or relapse. In addition, alloHCT recipients had full hematopoietic reconstitution. In parallel, we studied 45 non-HCT control individuals, which were matched for traditional CV risk factors (age, hypertension, diabetes, dyslipidemia, obesity, smoking) with alloHCT survivors.

Traditional Markers of Cardiovascular Disease
According to the study design, traditional cardiovascular risk factors did not differ between groups (Table 2). Interestingly, 7 out of 45 patients (16%) had been already diagnosed and treated for hypertension before the study visit (2 in the age group of 18-39 years and 5 in the group of 40-59 years). Furthermore, 4 out of 45 patients (9%) were diagnosed with hypertension based on the measurements of the present study (1 in the age group of 18-39 years and 5 in the group of 40-59 years). In addition, 10 patients suffered from dyslipidemia and 2 from diabetes mellitus.

Novel Markers of Vascular Injury and Increased Thrombotic Risk
Erythrocyte and platelet MVs were significantly increased in alloHCT recipients compared to controls (p = 0.004 and p = 0.039, respectively). Although endothelial MVs were increased in alloHCT recipients, the difference did not reach statistical significance (p = 0.133). There was also a trend towards a decrease of endothelial MVs as the interval after alloHCT increases (p = 0.094). Table 3 summarizes circulating levels of MVs in both groups. We further investigated possible associations of these novel markers in alloHCT recipients. Interestingly, endothelial MVs were significantly increased in the majority of alloHCT recipients that had received a myeloablative conditioning (n = 32/45, p = 0.012), compared to recipients of non-myeloablative conditioning. In addition, erythrocyte MVs were significantly increased in patients with a history of transplant-associated thrombotic microangiopathy (n = 3/45, p = 0.021), compared to patients with transplant-associated thrombotic microangiopathy. The latter had been treated as previously described [23,24] and resolved before a median of 13 (11-20 months) in these three patients. It is noteworthy that significant associations were found among MVs of different origins. As expected, platelet MVs were significantly associated with endothelial (r = 0.333, p = 0.002) and erythrocyte (r = 0.249, p = 0.034) MVs, suggesting an interaction between MVs of different origin

Discussion
Our findings suggest for the first time that endothelial injury and pro-coagulant activity are evident in adult alloHCT survivors, as this is reflected by the measurement of circulating MVs, independently of traditional CV risk factors. Furthermore, we were able to identify high-risk populations based on clinical features.
Over the last decades, our understanding of vascular injury has substantially improved. It is considered an early event in the pathophysiology of CV disease, contributing to subclinical target organ damage [25,26]. Biomarkers and potential treatment strategies for thrombotic events remain under investigation [27][28][29][30]. In 1992, Celermajer et al. reported the first non-invasive endothelial function test [31]. Since then, a plethora of vascular indices and biomarkers have been suggested for the assessment of endothelial dysfunction [32]. Among them, endothelial MVs are characterized by significant biological properties implicating them in several pathophysiological pathways, including the pathogenesis of CV diseases. Elevated endothelial MVs have been reported in patients with a variety of CV comorbidities, including diabetes, hypertension, acute coronary syndromes, and chronic ischemic heart disease [11,12,33]. Not only do they represent markers of endothelial dysfunction, but are also considered conveyors of biological messages between cells, which are implicated in the evolvement of vascular damage.
In alloHCT survivors, indices of endothelial injury have been evaluated in limited recent studies with heterogenous populations [34][35][36]. Using the gold standard vascular method of flow-mediated dilatation (FMD), autologous and allogeneic HCT recipients had impaired endothelial dysfunction compared to values prior HCT [34] and their siblings [36]. By contrast, when compared to age-and sex-matched controls, no difference in endothelial dysfunction was observed in childhood alloHCT survivors [35]. Regarding biomarkers of endothelial function, endothelial MVs have been scarcely studied in HCT recipients. Endothelial MVs were increased during the early post-transplant period (2-3 weeks post transplantation) [37] and in acute GVHD patients [38]. Nevertheless, none of these studies investigated a link between endothelial dysfunction and CV risk in HCT survivors. In our well-designed study of alloHCT recipients and controls, endothelial MVs were significantly increased only in patients receiving a myeloablative conditioning. Since myeloablative conditioning remains the backbone of alloHCT strategies [39,40], our finding is of great importance, since it is suggestive of sustained underlying endothelial damage, at least in this specific subgroup of patients.
In terms of pro-coagulant activity, this has been mainly studied in patients with acute endothelial syndromes, such as transplant-associated thrombotic microangiopathy [21] or GVHD [41,42]. Indeed, these recent studies have proven a crosstalk between endothelial dysfunction and thrombotic risk during acute syndromes. Nevertheless, this aspect has not been studied in alloHCT survivors, especially with regards to CV risk. In order to investigate if there is prothrombotic tendency in alloHCT survivors, after adjusting for the existence of CV risk factors, we measured promising markers of thrombosis already investigated in the field of CV disease, which is platelet and erythrocyte MVs [12,43]. Indeed, a link with thrombosis has been established not only for platelet MVs, but also recently for erythrocyte MVs [17,44]. Our study confirms for the first time this link in alloHCT recipients, suggesting that vascular injury and pro-coagulant activity may represent mutually reinforcing and interdependent processes in these patient group, independently of the presence of traditional CV risk factors.
Lastly, a really interesting finding of our study is the increase of erythrocyte MVs in patients with a history of transplant-associated thrombotic microangiopathy. This finding implicates erythrocyte MVs a potential long-lasting markers of thrombo-inflammation that is evident in these patients. Because the sample size was too small (only three patients), this result should mainly serve as an indication pointing towards future research. However, the aforementioned condition is rare and the patient's survival from this condition even rarer [45,46]. In this sense we reckon that this finding should be highlighted keeping in mind of course the above limitation.
The present study has some limitations and strengths. Firstly, the observational nature of the study does not allow for causality assumptions or for serial measurements of MVs before and after alloHCT. Since there were no similar studies in the field, the relatively small sample size was based on sample size calculations. Our group of alloHCT survivors could be considered heterogeneous regarding the disease type and phase, as well as the timing post alloHCT. It should be also noted that our study cannot assess the effect of the patients' disease itself or pre-alloHCT treatments on endothelial injury and pro-coagulant activity. Nevertheless, it is a meticulously selected population of alloHCT survivors and matched controls that provides novel data on vascular injury and pro-coagulant activity in alloHCT survivors, highlighting the clinical characteristics of high-risk patients. Further prospective studies are needed to address the role of novel markers in CV morbidity and mortality of alloHCT survivors, as well as the effects of disease type, phase, and pre-transplant treatments on these markers.

Study Population
We enrolled consecutive adult alloHCT survivors from our HCT Clinic (January-December 2019), with a follow-up from alloHCT longer than one year. We excluded patients with established cardiovascular disease, defined as stroke, angina, ischemic heart disease, heart failure, and arrhythmias; chronic graft-versus-host disease under immunosuppressive treatment (GVHD), acute or chronic inflammatory disease, active malignancy or relapse. All patients underwent allogeneic HCT at our JACIE (Joint Accreditation Committee-ISCT and EBMT) accredited Unit. Data relevant to alloHCT and post-HCT follow-up were collected retrospectively from our prospectively acquired database. AlloHCT was performed according to our standard operational procedures, as previously published [39,40,47,48].
We also enrolled consecutive control individuals from our Outpatient Hypertension Clinic who attended regular appointments, matched for traditional cardiovascular risk factors (age, hypertension, diabetes, dyslipidemia, obesity, and smoking) with alloHCT survivors. In accordance with the Helsinki Declaration, all patients have given written informed consent. All subjects were of Caucasian origin. The institutional review board of Aristotle University of Thessaloniki approved our study.

Laboratory Measurements
All measurements were performed between 9:00 and 11:00 a.m., with the participants having refrained from food, coffee, and smoking for at least 10 h. Detailed history, physical examination, and blood sampling were performed.
Office blood pressure (BP) was measured in the sitting position using a validated oscillometric device (Microlife Exact BP, Microlife AG, Widnau, Switzerland), according to standard recommendations for office BP measurement [27]. The mean of the second and third value of three consecutive measurements with a 2-min interval in the arm with the higher BP was considered as the patients' office BP. Hypertension was defined according to current guidelines [49].

Biochemical Profile
Plasma glucose, lipids (total cholesterol, LDL-cholesterol, HDL-cholesterol, triglycerides), renal and liver function were determined using routine laboratory techniques under fasting conditions.

Blood Sampling for MVs Quantitation
Blood samples for MV measurements were collected in citrated tubes (sodium citrate 3.2%) and centrifuged within 30 min with a two-step centrifugation protocol (2500× g for 15 min at room temperature, followed by a second centrifugation of the supernatant at the same condition). Supernatant was collected, and platelet poor plasma was stored at −80 • C [29]. MVs detection was then performed on thawed samples using a CyFlow Cube8 ROBBY flow cytometer (Sysmex Partec GmbH, Goerlitz, Germany). Preliminary experiments showed excellent reproducibility in samples stored less than 2 weeks (coefficient of variation <10%). Therefore, all samples were analyzed in less than 2 weeks from collection.

MVs Quantitation
Flow cytometry protocol for MVs quantitation was performed as previously published [11,12]. Background noise was checked using ultrapure water at less than 2000 events/s (flow ratẽ 10,000 events/s.

Statistical Analysis
Transplant characteristics included in the analysis were: indication for transplant (disease, stage), phase at transplant, previous lines of treatment, conditioning (myeloablative versus reduced intensity and total body irradiation/ total body irradiation (TBI)-based versus non-TBI), type of donor (sibling, unrelated, haploidentical), acute GVHD grade II-IV and extensive chronic GVHD. Timing from transplant at study sampling and patient demographics were also included in the analysis. Sample size calculation was based on previous studies of platelet MVs in patients with increased CV risk compared to controls [12]. With a 5% level of significance and 80% power [50], the estimated sample size was 40 individuals per group.
Data were analyzed using the statistical program SPSS 23.0 (IBM SPSS Statistics for Windows, Version 23.0. Armonk, NY, USA: IBM Corp.). Descriptive statistics were performed using median and range for continuous variables and frequency for categorical variables. Continuous variables were assessed for normality and compared using one-way ANOVA with the Bonferroni correction or Kruskal-Wallis test. Follow-up was measured from the date of transplantation until the date of last follow up or death. Correlation between continuous variables was assessed with the Pearson's or Spearman's correlation coefficient. The level of statistical significance was defined at 0.05.

Conclusions
In conclusion, we demonstrate for the first time a crosstalk between endothelial injury and pro-coagulant activity in alloHCT survivors long after HCT using novel biomarkers, independently of traditional cardiovascular risk assessment. Identifying high-risk clinical features may guide the role of circulating MVs as novel markers of long-lasting subclinical pathophysiology.