Impact of High-Dose Intravenous Vitamin C on Cell-Free DNA and Syndecan-1 in Patients with Sepsis-Associated ARDS


 Background:We set out to examine the effects of high dosage intravenous vitamin C (HDIVC) infusion on plasma cell-free DNA and Syndecan-1, two mortality biomarkers that represent neutrophil extracellular trap (NET) formation and degradation of the endothelial glycocalyx.Methods: Post-hoc analysis of plasma cell-free DNA and syndecan-1 in patients enrolled in the randomized placebo-controlled trial: Vitamin C Infusion for Treatment in Sepsis-Induced Acute Lung Injury, the CITRIS-ALI trial. Setting: Seven intensive care units in hospitals located in five different states in the U.S. Patients: Septic adults with ARDS between September 2014 to November 2017, final follow-up January 2018.Results: In 167 study patients, baseline plasma cfDNA levels in HDIVC (84 patients) and placebo (83 patients) were 2.18 ng/µL (SD 4.20 ng/µL) and 2.65 ng/µL (SD 3.87 ng/µL), respectively, p=0.45. At 48-hours, the cfDNA reduction was 1.02 ng/µL greater in HDIVC, compared to placebo, p=0.05. Mean baseline plasma syndecan-1 levels in HDIVC and placebo were 9.49 ng/mL (SD 5.57 ng/mL) and 10.83 ng/mL (SD 5.95 ng/mL) respectively, p=0.14. At 48 hours, patients in the placebo arm exhibited a 1.53 ng/mL (95% CI, 0.96 to 2.11) increase in syndecan-1 vs. 0.75 ng/mL (95% CI, 0.21 to 1.29), in HDIVC patients, p=0.05. The 48-hour plasma syndecan-1 levels in patients treated with HDIVC exhibited a linear association with improved oxygenation (PaO2/FiO2, beta= -18.9, p=0.004).Conclusions:HDIVC infusion significantly attenuated plasma cell-free DNA and syndecan-1, biomarkers known to be elevated in sepsis-induced ARDS. These results support the conclusion that high dosage intravenous vitamin C infusion reduces sepsis-induced vascular injury.Trial Registration: ClinicalTrials.gov identifier: NCT02106975


Introduction
Acute respiratory distress syndrome (ARDS) is an in ammatory lung disease with a mortality rate of 35-46%. [1][2][3] Sepsis can trigger vascular injury that leads to ARDS through systemic and local in ammation, damaging lung barrier function by both alveolar epithelial cell and lung capillary endothelial cell injury.
Loss of barrier integrity leads to interstitial and alveolar ooding, surfactant damage, and collapsed lung units. The cumulative result is severe lung function impairment characterized by diminished compliance, increased shunt, and severe hypoxemia.
Clinical and laboratory evidence suggest that High-Dose Intravenous Vitamin C (HDIVC) may have a role in the ARDS treatment. [4][5][6][7] A recently completed double-blind, randomized placebo-controlled trial revealed that HDIVC signi cantly reduced 28-day mortality and organ failure in sepsis-associated ARDS. [8,9] We hypothesized that important mechanisms of action of HDIVC are the regulation of Neutrophil Extracellular Trap (NET) formation and preservation of the endothelial glycocalyx. [10][11][12] A key biomarker of glycocalyceal integrity is the proteoglycan syndecan-1, an important structural component of the glycocalyx which lines luminal endothelial vascular surfaces, including alveolar capillaries. [13] Endotoxemia and bacteremia that lead to sepsis disrupt the glycocalyx, one of the earliest and most signi cant injury sites. [14] Early phase injury leads to degradation of the glycocalyx barrier and shedding of syndecan-1 into the circulation. [15] Syndecan-1, a biomarker of the degree of glycocalyx damage, is associated with ARDS development. [10,16] Loss of glycocalyx integrity results in movement of cells, protein, and uid from the vascular space into lung perivascular interstitial and alveolar spaces, a hallmark of ARDS. Therapies that target protection or restoration of the glycocalyx may bene t septic patients and theoretically reduce ARDS-associated mortality. [17] NETs, a recent discovery in innate immunity, are composed of granules and nuclear content extruded from neutrophils which kill bacteria extracellularly. [18] A key plasma biomarker of NET formation is cellfree DNA (cfDNA), which is elevated in sepsis and ARDS. [19][20][21] Prior studies have investigated the prognostic utility of cfDNA and syndecan-1 at the onset of sepsis and ARDS. [15,22,23] To date, no study has investigated an intervention in septic individuals that directly reduces NET formation and glycocalyx degradation. Patients receiving HDIVC in the CITRIS-ALI trial exhibited improved mortality and organ failure. [8,9] Given these results, we hypothesized that HDIVC would reduce plasma cfDNA and syndecan-1 when compared to placebo patients, receiving only standard of care for sepsis and ARDS. We further hypothesized that cfDNA and syndecan-1 are associated with objective clinical oxygenation indices in patients with ARDS.

Plasma Syndecan-1 Quanti cation
Plasma syndecan-1 levels were analyzed using a human magnetic bead Luminex assay system (LXSAHM), according to the manufacturer's instructions (R&D Systems, Minneapolis MN) and quanti ed using a Luminex LX200 instrument with xPONENT 3.1 software (Luminex Corporation, Austin, TX). Biomarker concentrations were calculated from standard curves of Median Fluorescence Intensity (MFI) by generating a ve-parameter logistic (5-PL) curve-t and multiplying by the dilution factor. Specimens outside the standard range were further diluted and assays repeated.

Statistical Analysis
Analyses were conducted using Stata Statistical Software (Rel.16.1, TX StataCorp LP). Multiple linear regressions were applied to assess biomarker differences at 48 hours, adjusting for baseline biomarker levels, comparing them among the two randomized groups (HDIVC and Placebo). Post-estimation plots represent the ndings graphically. Regression residuals for normalcy were then assessed. In instances where residuals were not normally distributed the non-parametric Wilcoxon test was applied. [24] Multiple logistic regression was applied to evaluate the adjusted effect of biomarkers on mortality. We evaluated the models with the area under the receiver operator characteristic curve, and the Hosmer-Lemeshow goodness-of-t test. [25] Results

Study Participants
Patients with sepsis-induced ARDS were enrolled at the time of ARDS onset (n = 167). Patients were randomized to receive HDIVC (n = 84) or placebo (dextrose 5% in water, n = 83). Baseline plasma samples were analyzed for cfDNA (n = 167) and syndecan-1 (n = 166). At 48 hours, 82 (97.6%) HDIVC and 72 (86.8%) of the placebo patients survived and remained in the ICU (P = 0.009). Plasma specimens from the survivors were analyzed. A summary of study patients and specimens can be found in Table 1. The study population is described extensively elsewhere. [8] (Fig. 1).

Association of cfDNA and Syndecan-1 with Mortality
Baseline plasma syndecan-1 levels and both 48-hour ΔcfDNA and Δsyndecan-1 levels predicted 28-day all-cause hospital mortality. Table 2 outlines the Odds Ratio (OR) of death for every incremental unit increase of the corresponding biomarker.

Discussion
The present study reports that a 96-hour HDIVC infusion in patients with sepsis-associated ARDS attenuated increases in 48-hour cfDNA and syndecan-1 plasma levels. Attenuated syndecan-1 levels correlated with improved lung function, as gaged by improved 48-hour P a O 2 /F i O 2 ratios (Fig. 3). HDIVC's impact on syndecan-1 and cfDNA levels independently predicted lower 28-day all-cause mortality. Elevated cfDNA and syndecan-1 levels in the plasma of septic patients with ARDS provides fresh insight into the extent of systemic in ammation and the molecular mechanisms that produce vascular injury, leading to ARDS onset.
Neutrophil extracellular traps are highly linked to endothelial damage and organ failure, crucial events in sepsis. [19] NET formation is a neutrophil effector mechanism whereby neutrophils extrude a web of chromatin bers complexed to granule-derived antimicrobial peptides and enzymes. This process occurs following neutrophil activation and is implicated in producing endothelial damage. [26] Hirose et al. identi ed NETs in peripheral blood smears of critically ill patients. [27] In septic patients, circulating cfDNA levels correlated with the degree of lung injury, as higher concentrations were found in patients who developed moderate or severe ARDS than septic patients without ARDS. [28] LeFrancais et al found that attenuating NET formation in an acute lung injury mouse model led to increased survival. [21] Activated endothelial cells induce neutrophil NET formation and are themselves susceptible to NETosismediated cell death, [29] thus, promoting a self-perpetuating damage that ultimately leads to hypercoagulable states. [30] The association of syndecan-1 with endothelial damage and neutrophilic in ammation has focused attention on a biomarker indicative of vascular injury. [15,17,31] Plasma syndecan-1 levels in septic patients are increased at baseline and may remain elevated for up to 72 hours. [32] Further, in these septic patients, elevations of syndecan-1 are associated with heightened risks of developing respiratory failure and increased mortality. [33] Both cfDNA and syndecan-1 levels are reported for mortality predictions in septic patients and are associated with adverse clinical outcomes (e.g., development of multiple organ failure, ARDS). [14,34] Correlations between the two biomarkers to clinical outcomes pertain to their roles as surrogates for NET formation and glycocalyx integrity. Plasma syndecan-1 elevations are a robust marker of glycocalyx degradation and development of ARDS. [31] To our knowledge, this is the rst human randomized placebo-controlled study of sepsis-associated ARDS to examine an interventional therapy's effect on these biomarkers.
CITRIS-ALI is the rst study to show that a 96-hour infusion of HDIVC decreased human plasma cfDNA and attenuated the rise in syndecan-1 levels at 48-hours (Figs. 1 and 2). [8] A re-analysis of the CITRIS-ALI data, accounting for the missing SOFA scores due to the large survival differences among the two arms, (i.e., Survivorship Bias), disclosed improved overall organ-function (modi ed SOFA Scores) in ARDS patient who received HDIVC infusion. [9] Murine models of polymicrobial sepsis using high dose vitamin C have demonstrated attenuation of lung NET formation, and circulating cfDNA. [5] Other studies reveal that high dose vitamin C reduced multiple organ failure, neutrophilic capillaritis and increased extravascular lung water in septic mice. [6] Thus, decreased circulating cfDNA may represent a surrogate marker of high dose vitamin C's ability to reduce neutrophil cell death, thereby reducing NET formation and the ensuing in ammatory vascular injury. The present study found signi cant correlations between decreased plasma syndecan-1 and improved P a O 2 /F i O 2 ratios (Fig. 3) and 28-day all-cause hospital mortality (Table 1).
Taken together, these ndings suggest that a 96-hour infusion HDIVC may improve ARDS recovery by protecting or restoring glycocalyx integrity.

Limitations
The study has several limitations. First, the speci c origin of circulating cfDNA was not determined. One study suggests that surges in cfDNA in sepsis results from cellular necrosis. [35] However, other studies show that cfDNA in septic patients is host derived and the cfDNA base pair length is consistent with neutrophil NET formation and not cellular necrosis. [36] Second, syndecan-1 was measured, but not other glycocalyceal structures (i.e., endocan, heparan sulfate, hyaluronan). Multiple studies show that syndecan-1 levels in sepsis correlate strongly with other markers of glycocalyx degradation. [14,15] Third, missing data points at 48-hours due to early deaths in placebo patients (13.3%) vs. HDIVC patients (2.4%) may have biased the ability to detect an even greater difference in biomarker levels.
Conclusion HDIVC treatment reduced 48-hour cfDNA and syndecan-1 plasma levels in patients with sepsisassociated ARDS. The dynamic changes of these biomarkers were strongly associated with lung oxygenation and 28-day all-cause mortality. These results suggest that HDIVC reduces the severity of illness by decreasing neutrophil activation and glycocalyx degradation. Syndecan-1 and cfDNA signal pathophysiological processes that lead to vascular injury in sepsis-associated ARDS. Future studies will clarify the role of these biomarkers in directing the care of patients with sepsis-associated ARDS.