Endothelin-1 is Increased in the Plasma of Patients Hospitalized with Covid-19

Importance The coronavirus disease 2019 (Covid-19) pandemic continues to place a devastating strain on healthcare services worldwide and there remains an ongoing requirement for new treatments. A key mechanism recognised in progressive severe disease is virus-induced endothelial dysregulation. Endothelin-1 (ET-1), being the most highly expressed peptide in endothelial cells and potent vasoconstrictor of human blood vessels, represents a potential therapeutic target through the use of Endothelin receptor antagonists. Objective To investigate the association of plasma ET-1 with Covid-19 disease severity Design Retrospective longitudinal cohort study of Covid-19 patients divided into Group A (asymptomatic or symptoms not requiring hospitalisation), Group B (symptoms requiring hospitalisation) and Group C (symptoms requiring supplemental oxygen therapy or assisted ventilation) recruited between March and July 2020 (the first wave of the Covid-19 pandemic in the UK). Data were compared with a contemporaneous cross-section of non-infected volunteers (Controls). Setting Single Tertiary National Health Service Hospital. Participants Tissue banked plasma samples were obtained from 194 patients. Exposures Quantitation of ET-1 in plasma by specific enzyme linked immunosorbent assay. Main outcome and measures Pairwise comparison of ET-1 levels (median [IQR]) between patient categories, and subgroups defined by clinical outcomes. Results Baseline ET-1 plasma levels (pg/ml) were elevated in patients requiring hospitalisation compared with controls and patients with asymptomatic or mild infection (Group B: 1.59 [1.13-1.98], and Group C: 1.65 [1.02-2.32] versus controls: 0.68 [0.47-0.87], p=<0.001 and Group A: 0.72 [0.57-1.10], p=<0.001). ET-1 levels were also elevated in patients that died (2.09 [1.66-3.15]), developed acute kidney (1.70 [1.07-2.36]) or myocardial injury (1.50 [0.92-2.28]) compared with patients with an uncomplicated infection (1.00 [0.61-1.57], p=<0.01). Amongst surviving hospitalised patients, ET-1 concentrations decreased when measured at 28 days (Group B: 0.86 [0.60-1.61] and Group C: 1.17 [0.66-1.62] versus baseline, p=<0.05) and 90 days (Group B: 0.69 [0.59-1.38] and Group C: 1.01 [0.64-1.21] versus baseline, p=<0.05). Conclusions and relevance Hospitalised Covid-19 patients demonstrate elevated ET-1 levels during the acute phase of infection and this is associated with increasing clinical severity of the disease. The results support the hypothesis that endothelin receptor antagonists may be beneficial for certain Covid-19 patients.


Main outcome and measures
Pairwise comparison of ET-1 levels (median [IQR]) between patient categories, and subgroups defined by clinical outcomes.

Results
Baseline ET-1 plasma levels (pg/ml) were elevated in patients requiring

Conclusions and relevance
Hospitalised Covid-19 patients demonstrate elevated ET-1 levels during the acute phase of infection and this is associated with increasing clinical severity of the disease. The results support the hypothesis that endothelin receptor antagonists may be beneficial for certain Covid-19 patients. All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted January 1, 2022. ; https://doi.org/10.1101/2021. 12 Endothelin-1 (ET-1), being the most highly expressed peptide in endothelial cells and potent vasoconstrictor of human blood vessels 9,10 , represents a potential therapeutic target. The benefit of endothelin receptor antagonists is already well established in pulmonary arterial hypertension 11 hence these medications may be suitable for accelerated regulatory approval. ET-1 is released from endothelial cells via a continuous constitutive pathway and supplemented by ET-1 release from Weibel-Palade bodies (the unique storage granules of endothelial cells) in response to extracellular stimuli including inflammatory cytokines 10 . Serialised measurement of plasma ET-1 during Covid-19 infection has not previously been undertaken.
We hypothesised that elevated levels of plasma ET-1 in the acute phase of Covid-19 infection would be associated with more severe disease. All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted January 1, 2022. All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. levels between patient categories A-C and controls, between different time-points for categories A-C, and between subgroups defined by clinical endpoints was undertaken using the Independent Samples Kruskal-Wallis Test. To assess for potential confounding effects when comparing ET-1 levels between patient categories A-C and controls, univariate analysis of co-variance was performed for each of: age, gender, hypertension, ischemic heart disease, diabetes, congestive cardiac failure and chronic kidney disease.

Results
All data that support the findings of this study are available from the corresponding author upon reasonable request. Plasma samples were obtained from 194 patients. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted January 1, 2022. ; https://doi.org/10.1101/2021.12.30.21268236 doi: medRxiv preprint in Group B and 32.1% in Group C. The frequency of underlying comorbidities was higher in Groups B and C compared with Group A and controls ( Table 1). Despite heterogeneity in clinical and demographic characteristics, differences in baseline ET-1 between all patient categories remained significant (p<0.05) in corrected models for age, gender, hypertension, ischemic heart disease, diabetes, congestive cardiac failure and chronic kidney disease using between-subjects effect analysis of covariance (supplemental information).
Baseline ET-1 levels (pg/ml) were also significantly elevated in subgroups of patients  Figure 1A). (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted January 1, 2022.  15 . In our Covid-19 cohort, the highest ET-1 levels were seen in dying patients and those whose infection was complicated by respiratory failure (indicated by a requirement for supplemental oxygen or assisted ventilation), acute kidney or myocardial injury suggesting that increased circulating ET-1 may also be an important contributor to the pathogenesis of these complications.

Limitations
Our study is limited by its retrospective observational design with all patients enrolled during the first wave of the pandemic, when no specific therapies for Covid-19 were known and prior to the availability of vaccines. Further exploration of the pathophysiological changes resulting from elevated ET-1 levels in critically unwell patients was not feasible in this period. All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

Conclusion
Our results demonstrate that elevated ET-1 in the acute phase of Covid-19 infection is a clinical feature unique to severe disease, supporting the hypothesis 8 that endothelin receptor antagonists, used to treat pulmonary arterial hypertension, may be beneficial in a subset of Covid-19 patients. All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted January 1, 2022. ; https://doi.org/10.1101/2021.12.30.21268236 doi: medRxiv preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted January 1, 2022. ; https://doi.org/10.1101/2021.12.30.21268236 doi: medRxiv preprint Table 1 Legend  (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted January 1, 2022. ; https://doi.org/10.1101/2021.12.30.21268236 doi: medRxiv preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted January 1, 2022.  (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

Figure 2
All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

Note to Referees/supplemental information:
All data that support the findings of this study are available from the corresponding author upon request. To assist the review process, we have provided below further details of the statistical analyses performed as well as the following details of the possible confounding variables considered in this study and Univariate Analyses of Co-variance.
All statistical analysis was performed using SPSS version 27 (IBM Corp., USA) for Windows. Normality of continuous variable distributions was tested by one-sample Kolmogorov-Smirnov test. Non-normally distributed continuous variables are presented as Median (inter-quartile range [Q1 -Q3]). Due to the skewed distribution of ET-1, no potential or extreme outliers have been excluded from analysis. Outliers were noted to mostly occur in patient category C (hospitalised and requiring supplemental oxygen/assisted ventilation) or amongst subgroups with clinical complications of Covid-19 infection therefore we have assumed these values reflect true biological variability in the study population. Comparison of ET-1 between patient categories was performed using the Independent Samples Kruskal-Wallis Test and also between time-points for each patient category. In the latter case, an independent samples non-parametric test was chosen to increase power given the low number of paired corresponding samples available due to patient drop-out at 28 and 90 days. Reasons for loss of patients to follow up were in many cases unavoidable including patient death, persisting disability and patients being repatriated outside our local area having been referred to our centre for specialist tertiary care during their initial illness.

Demographic details of patient categories and univariate analyses of co-variance.
Covariate variables are presented as n (%) unless otherwise stated. No demographic or clinical endpoint data was available for 3/39 patients in category B otherwise clinical variables were available for all other enrolled patients. Test statistic: difference in mean square ET-1 at time 0 between patient categories (controls, A, B and C) after adjustment for covariate variables; p value after univariate adjustment refers to probability that differences in baseline ET-1 are not significantly different between patient categories (controls, A, B and C) after adjustment for differences in covariate variables. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.