High value of mid‐regional proadrenomedullin in COVID‐19: A marker of widespread endothelial damage, disease severity, and mortality

Abstract The widespread endothelial damage due to severe acute respiratory syndrome coronavirus 2 (SARS‐CoV‐2) may lead to a disruption of the adrenomedullin (ADM) system responsible for vascular leakage, increased inflammatory status, and microvascular alteration with multi‐organs dysfunction. The aim of this study was to evaluate the role of mid‐regional proadrenomedullin (MR‐proADM) as a marker of SARS‐CoV2 related widespread endothelial damage, clinically identified by organs damage, disease severity and mortality. Patients with SARS‐CoV‐2 infection has been prospectively enrolled and demographic characteristic, clinical and laboratory data has been evaluated. In the overall population, 58% developed acute respiratory distress syndrome (ARDS), 23.3% of patients died, 6.5% acute cardiac injury, 1.4% of patients developed acute ischemic stroke, 21.2% acute kidney injury, 11.8% acute liver damage, and 5.4% septic shock. The best MR‐proADM cut‐off values for ARDS development and mortality prediction were 3.04 and 2 nmol/L, respectively. Patients presenting with MR‐proADM values ≥2 nmol/L showed a significantly higher mortality risk. In conclusion, MR‐proADM values ≥2 nmol/L identify those patients with high mortality risk related to a multiorgan dysfunction syndrome. These patients must be carefully evaluated and considered for an intensive therapeutic approach.


| INTRODUCTION
The coronavirus disease 2019 , caused by the severe acute respiratory distress syndrome coronavirus 2 (SARS-CoV-2), was responsible for an unprecedent threat to global health. 1 The clinical manifestations of the disease range from asymptomatic cases to severe pneumonia with high mortality rates (4%-13%), mainly in the case of acute respiratory distress syndrome (ARDS) development. 2 Trying to stratify disease severity, the World Health Organization classified patients in four classes (mild to critical) basing on clinical and radiological characteristics. 3 The use of biomarkers, however, may help clinicians identifying those patients with a severe disease and a higher risk of death. 4,5 Up to date, no markers of endothelial damage in COVID-19 have been validated in clinical practice.
Adrenomedullin (ADM), a 6 kDa protein with a 22 min half-life, is produced by endothelial and vascular smooth muscle cells due to volume overload to maintain endothelial barrier function, freely diffuses through the blood and interstitium, and binds to specific widespread receptors, mainly located in cardiovascular and pulmonary tissues. 6,7 The leading function of ADM is the vasodilatation in both vascular resistance and capacitance vessels resulting in a blood flow increase. ADM further reduce vasoconstriction through an inhibition of the renin-angiotensin-aldosterone system and maintains endothelial integrity reducing vascular permeability. 7 A disruption of ADM system results in vascular leakage that represents the first step of inflammation and of coagulation cascade activation. 8,9 As derived from ADM in a 1:1 ratio, mid-regional proadrenomedullin (MR-proADM) values directly reflect the effects of its less stable and easily detectable precursor and has been recently introduced in clinical practice as a prognostic marker in patients with bacterial infection. 10 A significant relation between MR-proADM values and bacterial pneumonia severity index score, indeed, has been highlighted. 11 Healthy individuals showed MR-proADM values of about 0.33 nmol/L. 10 MR-proADM values of 0.8 nmol/L, conversely, are diagnostic of bacterial infection with higher values indicative of a higher infection severity from 1.2 to 1.9 and 3.7 nmol/L in localized infections, sepsis or septic shock, respectively-. 12 In patients with sepsis and septic shock, MR-proADM values more than 3.4 and 4.3 nmol/L, respectively, were significantly associated with 90-day mortality. 13 Despite the vast majority of studies evaluated the role of MR-proADM in bacterial infections leading to sepsis, scant evidence is available in patients with viral infections without any information on COVID-19. 12,[14][15][16] Knowing that COVID-19 related damage resemble the alteration occurring during sepsis, however, a disruption of ADM pathway during SARS-CoV-2 infection may be hypothesized. 8,9,17 The aim of this study was to evaluate the role of MR-proADM as a marker of SARS-CoV-2 related widespread endothelial damage, clinically identified by organs damage, disease sevrity, and mortality.

| MATERIALS AND METHODS
This study has been approved by the Ethical Committee of the University Campus Bio-Medico of Rome and all patients provided informed consent before the enrollment within the study.

| Patient selection and characteristics
All patients hospitalized for SARS-CoV-2 infection at COVID Center of the Campus Bio-Medico of Rome University, were prospectively included between 1st April and 30th June 2020. The COVID center included both medicine department and intensive care unit (ICU). Pregnancy and lack of informed consent represented exclusion criteria.
The following data were collected at inclusion: demographic characteristics (age and gender); onset symptoms; relevant comorbidities; immune status (active malignancy or other causes of immunosuppression); concomitant antimicrobial, antiviral, or immunosuppressive treatments administration; clinical presentation. Furthermore, all patients received a complete physical examination including body temperature, blood pressure, heart and respiratory rate, cardiac, pulmonary, abdominal, and neurological evaluation. Laboratory values at inclusion comprehended complete blood counts, MR-proADM, C-reactive protein (CRP), ferritin, procalcitonin (PCT), coagulation (D-dimer, international normalized ratio, activated partial thromboplastin time), liver (aspartate aminotransferase, alanine aminotransferase, albumin, bilirubin) and kidney (creatinine) functionality tests, serum lactate, arterial blood gas examination.
All patients received standard of care basing on disease severity and comprehending oxygen support, anticoagulant therapy, hydroxychloroquine, and tocilizumab whether indicated.  Abbreviations: ARDS, acute respiratory distress syndrome; CRP, C-reactive protein; ICU, intensive care unit; IQR, interquartile range; MR-proADM, mid-regional proadrenomedullin; PCT, procalcitonin; SOFA, sequential organ failure assessment. a These outcomes are not available for all included patients (please refer to the text).

| Laboratory values measurement
All included patients were followed until death or 30-day follow-up, whichever came first.

| Statistical analysis
Continuous variables were expressed as mean (standard deviation) or median (interquartile ranges), according to data distribution, and were compared using the Student's t test or the Mann-Whitney U test; categorical variables were expressed as counts and percentages and compared using the χ 2 or Fisher's exact tests, as appropriate.

| Laboratory markers values in SARS-CoV-2 infection
AUCs values resulting from ROC curve analysis for MRproADM, CRP, PCT, and ferritin in patients with SARS-CoV-2 infection are showed in Table S2. ROC curves and AUC values resulted statistically significant for all variable, but PCT ( Figure S1, Table S2).

ROC curves comparison between the different variables has been
reported in Table S3 and schematized in Figure S1. AUC value for MR-proADM (0.78) was significantly higher than PCT (0.55; p < .0001), smaller than CRP (0.91, p < .0001) and similar than ferritin (0.86,

| ARDS prediction during SARS-CoV-2 infection
Median values with interquartile ranges and Mann-Whitney's comparison for MR-proADM, CRP, ferritin and SOFA score for patients with or without ARDS development during follow-up are reported in Table S4. All these variables resulted significantly higher in patients with ARDS.
ROC curves and AUC values resulted statistically significant for all considered variables despite only CRP presented significantly higher AUC values than MR-proADM (p = .030) ( Figure 1A, Figure S2, and Table S5 and S6). Furthermore, the best cut-off for ARDS development prediction were 3.04 nmol/L for MR-proADM, 3.88 mg/dl for CRP, 165.58 ng/ml for ferritin, and 1 for SOFA, respectively.

| 30-Day mortality prediction during SARS-CoV-2 infection
Median values with interquartile ranges and Mann-Whitney's comparison for MR-proADM and SOFA score in survivors and non survivors at 30-day follow-up are reported in Table S7.
ROC curve and AUC values for MR-proADM and SOFA score resulted statistically significant (p < .0001) without differences between the variables ( Figure 1B, Figure S3, and Tables S8 and S9). The best cut-off for 30-day mortality prediction were ≥2 nmol/L for MR-proADM, 2.91 mg/dl for CRP, 635.86 ng/ml for ferritin, and ≥3 for SOFA score.

| DISCUSSION
The results of this study showed that MR-proADM may be used Being responsible for endothelial integrity, an alteration of ADM system during sepsis causes vascular leakage and organ dysfunction ( Figure 3). 31 Recent observational studies confirmed these pathological

| CONCLUSION
MR-proADM values ≥2 nmol/L identify those patients with high mortality risk related to a multiple organ dysfunction syndrome.
These patients must be carefully evaluated and considered for an intensive therapeutic approach. Further studies in larger populations will be warranted to confirm these data.

ACKNOWLEDGMENT
The authors thank Stefano Spoto for his help in collecting clinical data.

CONFLICT OF INTERESTS
The authors declare that there are no conflict of interests.

AUTHOR CONTRIBUTIONS
Silvia Spoto led the study design, data collection, data analysis, data interpretation, and manuscript writing. Felice E. Agrò, Federica Sambuco, and Francesco Travaglino assisted with data collection and analysis of the validation dataset. Emanuele Valeriani and Marta Fogolari assisted with computer queries, data analysis and manuscript preparation. F.M. assisted with data collection and analysis of the validation dataset. Massimo Ciccozzi assisted with data collection and analysis of the development dataset as well as study design, data interpretation and manuscript writing. Sebastiano Costantino and Silvia Angeletti assisted with chart review and data analysis and supervised all aspects of the investigation, as well as assisting with study design, data interpretation, and manuscript writing.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.