Endothelial Injury Syndromes after Allogeneic Hematopoietic Stem Cell Transplantation: Angiopetin-2 as a Novel Predictor of the Outcome and the Role of Functional Autoantibodies against Angiotensin II Type 1 and Endothelin A Receptor

Transplant-associated thrombotic microangiopathy (TMA) occurs in a significant percentage of patients after allogeneic stem cell transplantation (allo-SCT) and is associated with significant morbidity and mortality. The aim of the present study was to examine the association of serum angiopoetin-2 (Ang2) levels and the presence of antibodies against angiotensin II type 1 (AT1R) and ndothelin A Recreptor (ETAR) with the outcome of patients with TMA and/or graft-versus-host disease (GVHD) after allo-SCT. Analysis of our data showed that elevated serum Ang2 levels at the time of TMA diagnosis are significantly associated with increased non-relapse mortality and decreased overall survival. To our knowledge, this is the first study demonstrating an association between raised Ang2 levels and poor outcomes in patients with TMA. Antibodies against AT1R (AT1R-Abs) and ETAR (ETAR-Abs) were detected in 27% and 23% of the patients, respectively, but there was no association between the presence of autoantibodies and the outcome of patients with TMA. However, a significant finding was the strong positive correlation between the presence of AT1R-Abs with the occurrence of chronic fibrotic GVHD, such as scleroderma and cryptogenic organizing pneumonia, raising the possibility of the contribution of autoantibodies in the pathogenesis of fibrotic GVHD manifestations.


Introduction
Angiopoietin 2 (Ang2) plays a key role in the pathogenesis of endothelial cell damage in many different inflammatory diseases [1]. Ang1 and Ang2 are peptides that act as ligands for the same tyrosine kinase receptor Tie2, expressed on the surface of endothelial cells (EC). Binding of Ang1 to its receptor induces an anti-inflammatory response through downregulation of surface adhesion molecules and promotion of EC survival. On the contrary, Ang2 binding to the same receptor results in the opposing effects by inducing EC apoptosis and secretion of inflammatory cytokines. Previous studies in the setting of allogeneic stem cell transplantation have shown that Ang2 is a major factor contributing to the pathogenesis of systemic endothelial injury [2]. Therefore, in the present study we examined the association of Ang2 serum levels with the outcome of 52 patients with TMA and/or GVHD after allo-SCT. We also tested for the presence of antibodies against AT1R and ETAR, and we examined the association of antibodies with the outcome of TMA and GVHD.

Patient Characteristics
We studied 52 patients who underwent an allo-HSCT from 2016 to 2019 at two BMT Units in Greece and who were diagnosed with GVHD and/or TA-TMA. Demographic information and pretransplant data were uniformly collected on all patients. For analysis, the patients were divided into two cohorts, depending on the presence or absence of TMA, as follows: (i) Cohort 1 consisted of 20 consecutive patients with samples at the time of TMA diagnosis, (ii) Cohort 2 included 32 consecutive patients with moderate-severe acute and/or chronic GVHD, treated during the same time period, and had available stored serum samples.
The two cohorts were well balanced for basic transplant and disease characteristics, such as, sex, age, disease, type of donor, graft source, and conditioning regimen, as illustrated in Table 1. More specifically, Cohort 1 consisted of 13 males and 7 females with a median age at the time of transplant of 49 years (range, . Diagnosis of TMA was established in a median of 80 days after graft infusion (range, 45-740). TMA developed in the context of concurrent acute and chronic GVHD in 14 and 2 patients, respectively, while in 4 patients TMA was diagnosed in the absence of GVHD. Cyclosporine was dose reduced or discontinued in all patients at the time of TMA diagnosis. Plasma exchange and/or plasma infusion was administered in 8 and 5 patients, respectively, while 7 patients received treatment with eculizumab. TMA resolved in 10 out of 20 patients, while 10 patients had refractory disease and died from causes directly associated with TMA. One patient with acute myeloid leukemia (AML) died due to leukemia relapse, while another patient with Myelodysplastic syndrome (MDS) in complete remission died from infection. Complete remission was achieved in 4 out of 7 patients treated with eculizumab, and the drug was successfully discontinued in all cases. Cohort 2 consisted of 16 males and 16 females with a median age at the time of transplant of 53 years (range, 27-68). Severe acute GVHD grade III-IV and severe chronic GVHD developed in 8 and 13 patients, respectively. TMA did not develop in any of the patients in Cohort 2 during their post-transplant course.

Outcome of Patients with TMA
For patients with TMA, the median overall survival was 32 months, while the cumulative incidence of NRM and OS at 5-years was 54% (95% CI, 29-74%) and 36% (95% CI, 11-61%), respectively. In multivariate analysis, the only parameter associated with NRM was the serum levels of Ang2 ( Table 2). The cumulative incidence of NRM in the group of patients with Ang2 serum levels above the median (75%, 95% CI, 33-93%) was significantly increased in comparison with the cumulative incidence of NRM in patients with Ang2 serum levels below the median (33%, 95% CI, 7-63%), p = 0.04, (Figure 4a). Patients with Ang2 serum levels below the median had a significant trend for higher OS as compared with the group of patients with Ang2 serum levels above the median (Figure 4b). EASIX score was not statistically associated with NRM in the group of patients with TMA. patients with Ang2 serum levels above the median (75%, 95% CI, 33-93%) was significantly increased in comparison with the cumulative incidence of NRM in patients with Ang2 serum levels below the median (33%, 95% CI, 7-63%), p = 0.04, (Figure 4a). Patients with Ang2 serum levels below the median had a significant trend for higher OS as compared with the group of patients with Ang2 serum levels above the median (Figure 4b). EASIX score was not statistically associated with NRM in the group of patients with TMA.

Autoantibodies against AT1R and GVHD-Manifestations
In total, 48 patients suffered from GVHD (16 patients from Cohort 1 and 32 patients from Cohort 2). Twelve out of 48 patients had chronic GVHD associated with fibrotic manifestations (9 patients with scleroderma and 3 patients with cryptogenic organizing pneumonia). Eleven out of 12 patients with chronic fibrotic GVHD tested positive for the presence of AT1R-Abs, while among the group of 36 patients without fibrotic GVHD, only 3 tested positive for AT1R-Abs (p < 0.001). In more detail, 3 out of 3 patients with cryptogenic organizing pneumonia (COP), 8 out of 9 patients with sclerodermatous chronic-GVHD, and only 3 out of 36 patients without scleroderma and/or COP were positive for the presence of AT1R-Abs (Figure 5a). Similar findings were observed when the analysis was performed only for patients with high-titer AT1R-Abs positivity. More specifically, 5 out of 12 (42%) patients with chronic fibrotic GVHD were positive for the presence of high-titer AT1R-Abs, while only 1 out of 36 (1.8%) without fibrotic GVHD was positive for high-titer AT1R-Abs (p < 0.001).

Autoantibodies against AT1R and GVHD-Manifestations
In total, 48 patients suffered from GVHD (16 patients from Cohort 1 and 32 patients from Cohort 2). Twelve out of 48 patients had chronic GVHD associated with fibrotic manifestations (9 patients with scleroderma and 3 patients with cryptogenic organizing pneumonia). Eleven out of 12 patients with chronic fibrotic GVHD tested positive for the presence of AT1R-Abs, while among the group of 36 patients without fibrotic GVHD, only 3 tested positive for AT1R-Abs (p < 0.001). In more detail, 3 out of 3 patients with cryptogenic organizing pneumonia (COP), 8 out of 9 patients with sclerodermatous chronic-GVHD, and only 3 out of 36 patients without scleroderma and/or COP were positive for the presence of AT1R-Abs (Figure 5a). Similar findings were observed when the analysis was performed only for patients with high-titer AT1R-Abs positivity. More specifically, 5 out of 12 (42%) patients with chronic fibrotic GVHD were positive for the presence of high-titer AT1R-Abs, while only 1 out of 36 (1.8%) without fibrotic GVHD was positive for high-titer AT1R-Abs (p < 0.001).

Discussion
Emerging evidence supports the role of endothelial dysfunction in the pathogenesis of non-infectious transplant-related complications, contributing to the increased morbidity and mortality documented after allo-SCT [24]. On this basis, a number of studies have investigated the significance and contribution of various angiogenic factors, such as the vascular endothelial growth factor (VEGF) and Ang2, in the development of such complications, along with their correlation with patient outcomes. More specifically, Luft et al., [25] described a high VEGF/Ang2 ratio in patients with steroid-refractory GvHD, compared to those with sensitive disease, whereas Porkholm et al., reported a correlation with the incidence of intestinal and skin/liver GvHD with the pre-and post-transplant Ang2 levels, respectively [26]. Ueda et al., reported a strong prognostic value of serum Ang2 levels for transplant-related complications with endothelial cell damage and leukemia relapse [3].
In this study, we examined Ang2 serum levels, in patients with TMA (cohort 1) and those with acute or chronic GvHD (cohort 2). We found no significant difference between the two at the time of diagnosis. This may be attributed to the fact that the majority of patients in Cohort 1 had co-existing GvHD, which is known to be associated with high Ang2 levels [27]. It is, however, of interest that higher Ang2 levels were found to be correlated with shorter OS, which is in keeping with previous reports, documenting a significantly lower 5-year OS in patients with high Ang-2 levels [26]. Furthermore, in multivariate analysis, Ang2 levels and the co-current diagnosis of TMA were found to be the only parameters predicting high NRM, which is again consistent with previous findings [26].
Interestingly, multivariate analysis in the group of patients with TMA showed that the only parameter associated with increased NRM was increase serum Ang2 levels at the time of diagnosis. To our knowledge, this is the first study showing that serum Ang2 levels may be used as a predictive biomarker for the outcome of patients with TMA.

Discussion
Emerging evidence supports the role of endothelial dysfunction in the pathogenesis of non-infectious transplant-related complications, contributing to the increased morbidity and mortality documented after allo-SCT [24]. On this basis, a number of studies have investigated the significance and contribution of various angiogenic factors, such as the vascular endothelial growth factor (VEGF) and Ang2, in the development of such complications, along with their correlation with patient outcomes. More specifically, Luft et al., [25] described a high VEGF/Ang2 ratio in patients with steroid-refractory GvHD, compared to those with sensitive disease, whereas Porkholm et al., reported a correlation with the incidence of intestinal and skin/liver GvHD with the pre-and post-transplant Ang2 levels, respectively [26]. Ueda et al., reported a strong prognostic value of serum Ang2 levels for transplant-related complications with endothelial cell damage and leukemia relapse [3].
In this study, we examined Ang2 serum levels, in patients with TMA (cohort 1) and those with acute or chronic GvHD (cohort 2). We found no significant difference between the two at the time of diagnosis. This may be attributed to the fact that the majority of patients in Cohort 1 had co-existing GvHD, which is known to be associated with high Ang2 levels [27]. It is, however, of interest that higher Ang2 levels were found to be correlated with shorter OS, which is in keeping with previous reports, documenting a significantly lower 5-year OS in patients with high Ang-2 levels [26]. Furthermore, in multivariate analysis, Ang2 levels and the co-current diagnosis of TMA were found to be the only parameters predicting high NRM, which is again consistent with previous findings [26].
Interestingly, multivariate analysis in the group of patients with TMA showed that the only parameter associated with increased NRM was increase serum Ang2 levels at the time of diagnosis. To our knowledge, this is the first study showing that serum Ang2 levels may be used as a predictive biomarker for the outcome of patients with TMA.
The contribution of non-HLA autoantibodies, and particularly of AT1R-and ETAR-Abs, in the development of vascular inflammation and graft rejection after solid organ transplantation has been well established [9,[28][29][30]. However, there is no information on the impact of such antibodies in the context of allo-SCT, apart from a single study of 3 allo-SCT patients with refractory hypertension (3/3) and diarrhoea (2/3) who were found to have elevated AT1R-Abs [31]. Interestingly, the introduction of an angiotensin receptor blocker in these patients resulted in symptom resolution, suggesting a potential pathogenetic role for AT1R-Abs. This is the first study, to our knowledge, investigating the presence of AT1R-Abs and ETAR-Abs in the allo-SCT setting and exploring the potential association of these antibodies with the development and outcome of TMA and/or GvHD. Our results have shown that 9.6% and 5.7% of patients had high titres (>17 units/mL) of AT1R-Abs and ETAR-Abs, respectively. It is of note that different cut-off levels have been used by different studies [31][32][33][34][35], making it difficult to interpret and compare results from different studies. However, the cut-off value of >10 units/mL has been adopted by the majority of studies [31][32][33][34], and it is the same that we used in our study as well, whereas the cut-off of >17 units/mL was used to define 'high-titer' patients.
We also detected a positive correlation between AT1R-Abs and ETAR-Abs and AT1R-Abs and Ang2 levels. A similar association between ETAR-Abs and AT1R-Abs has been described in a pediatric kidney transplant cohort, prompting monitoring of such antibodies, in addition to HLA donor-specific antibodies, to allow for targeted therapeutics [34].
Significantly higher percentage of patients with fibrotic chronic GvHD were found to have high-titer levels of AT1R-Abs, as compared to those without fibrotic GvHD (42% vs. 2.5%, p = 0.001). This may be explained by the fact that AT1R has been implicated in the proliferation of fibroblasts and collagen production and in accelerated inflammatory cell infiltration [36]. Furthermore, AT1R activation has been implicated in the pathogenesis of lung and liver fibrosis, as well as in the fibrosis of chronic-GvHD in a murine chronic-GvHD model [37,38]. Raised AT1R-Abs and ETAR-Abs have also been associated with microvasculopathy and allograft loss in heart, lung, liver and multivisceral transplantation [34].
More interestingly, we demonstrated that AT1R-Abs have a positive diagnostic capacity with high sensitivity and specificity for the development of sclerodermatous GvHD and/or COP after allo-SCT. This, again, is the first time, to our knowledge, that such an association has been described, prompting further investigation.
Despite the very promising results, our study carries significant limitations, such as: (i) the retrospective nature of our data, (ii) the limited sample size, restricting the inclusion of additional factors in multivariable analyses, (iii) the absence of a control cohort group (such as healthy individuals or allo-HSCT patients with no TMA or GvHD) and (iv) the lack of serial sample collection at different time points during the disease course.
In summary, to our knowledge, this is the first study demonstrating an association between raised Ang2 levels and poor outcomes (shorter OS and higher NRM) in patients with TMA. More interestingly, raised AT1R-Abs levels were found to be positively correlated with the occurrence of fibrotic chronic GvHD after allo-HSCT, suggesting that testing for these antibodies may have clinical utility and especially in the era of multiple angiotensin and endothelin A receptor blockade agents. However, larger, prospective studies are required to validate these results prior to their clinical translation.

Patients and Study Design
The study was approved by IRB and bioethics committee of ATTIKO University Hospital (IRB: 10992). Patients gave written informed consent before entry into the study.
Diagnosis of TMA was based on the criteria proposed by the International Working Group for TMA definition criteria (IWG) [39], whereas GVHD was diagnosed and graded according to established NIH criteria [40,41]. Serum samples were collected from all patients with TMA and/or GVHD at the time of diagnosis and stored at −80 • C.
Treatment of TMA included discontinuation of immunosuppression, plasma exchange and/or plasma infusion, or administration of eculizumab. The criteria used for evaluation of response to treatment have been described previously [42]. EASIX score at the time of diagnosis was estimated for patients with TMA with the use of the following formula: [(LDH x Creatine)/PLT] [43].

Ang2 and Antibodies against AT1R and ETAR
Analysis of the Ang2 levels was performed by enzyme-linked immunosorbent-based assay (ELISA), according to the instructions of the manufacturer (ABCAM PLC, Cambridge, UK). AT1R-Abs and ETAR-Abs quantitation were determined by ELISA (CellTrend GmbH, Luckenwalde, Germany). Sera were diluted 1:100 and tested in duplicates. AT1R-Ab and ETAR-Ab concentrations were determined by a standard curve. The cutoff of ≥10 units/mL was used as 'positive' for the presence of autoantibodies [31][32][33], whereas patients with levels >17 units/mL were considered as 'positive with high-titers'. Serum and plasma samples from patients collected at the time of diagnosis of TMA or GVHD were stored at the biobanking facilities of the two hospitals.

Statistical Analysis
Statistical analysis aimed to identify pre-treatment factors associated with NRM in patients with TMA and/or GVHD after allo-SCT, and to estimate the proportion of patients tested positive for the presence of autoantibodies against AT1R and ETAR. Overall survival (OS) was defined as the time from transplantation to last follow-up or death due to any cause, while NRM was defined as the time from transplantation to last follow-up, or death due to any cause in patients free of disease. Categorical variables between groups were compared using the two-sided Fisher's exact test. Competing risk analysis was used for estimating the cumulative incidence of NRM. Gray's test was used for univariate analysis of GVHD and NRM, while Fine and Gray proportional hazards regression model was used for multivariate analysis. The following variables entered in the multivariate analysis model: (1) Age at the time of allo-SCT (above vs. below the median), (2) Sex (male vs. female), (3) Presence of TMA (yes vs. no), (4) Ang2 serum levels (above vs. below the median), (5) Presence of Abs against AT1R and ETAR (yes vs. no), (6) EASIX score (above vs. below the median). Receiver operating characteristic (ROC) curve analysis was performed in order to identify the predictive accuracy of anti-AT1R-Abs titers for the diagnosis of specific GVHD manifestations. Statistical analysis was performed with the use of easy R and Medcalc statistical software [44].