Endothelial glycocalyx degradation is more severe in patients with non-pulmonary sepsis compared to pulmonary sepsis and associates with risk of ARDS and other organ dysfunction

The endothelial glycocalyx is a complex layer of glycoproteins, proteoglycans, and glycosaminoglycans that coats the luminal surface of the microvascular endothelium. Hydrated glycosaminoglycans form a thick and rigid endothelial surface layer (ESL) that plays a key role in limiting vascular permeability and regulating leukocyte adhesion [1, 2]. Shedding of the ESL occurs in response to a variety of insults and results in hyperpermeability, inappropriate leukocyte adhesion [3], and loss of capillary autoregulation [4].

The major endothelial cell surface proteoglycans are syndecans, which are heparan sulfate proteoglycans. Syndecan-1 is abundant in the endothelial surface layer, and circulating syndecan-1 is a marker of endothelial glycocalyx degradation [5‐11]. In mice, plasma syndecan-1 levels are negatively correlated with ESL thickness and positively correlated with microvascular permeability [12]. Glycocalyx dysfunction and syndecan-1 shedding have been described in a variety of clinical pathophysiologic processes, including sepsis [13], hemorrhagic shock [14], atherosclerosis [15], acute coronary syndrome [16], renal disease [17], diabetes [10], and hypervolemia [18]. Furthermore, elevated plasma syndecan-1 levels have been associated with increased mortality in patients with trauma and sepsis [8, 11]. However, despite the known role of endothelial injury and activation in the pathogenesis of ARDS [19‐23], the association between syndecan-1 and development of ARDS or other organ dysfunction in sepsis has not been well studied.

Both pulmonary and non-pulmonary sepsis can lead to ARDS with pulmonary sepsis (due to pneumonia or aspiration of gastric contents) resulting in direct injury to the lung and non-pulmonary sepsis resulting in indirect injury to the lung [22]. In both animal models and patients, direct injury to the lung is characterized by more severe lung epithelial damage, and an indirect insult to the lung is characterized by more severe systemic endothelial injury [24‐26]. However, whether there is greater disruption of the glycocalyx in indirect versus direct ARDS has not been well studied. In a mouse model of non-pulmonary sepsis, glycocalyx degradation contributed to acute lung injury, and human studies have shown an association between glycocalyx degradation and development of pulmonary edema [27, 28], findings that support an association between glycocalyx degradation and ARDS. In addition, a study of 17 patients with acute respiratory failure demonstrated differences in the overall pattern of circulating glycosaminoglycans between those with direct versus indirect causes of acute respiratory failure, raising the question of whether glycocalyx degradation differs depending on the mechanism of lung injury. However, syndecan-1 was not measured in that study and it was not reported whether the patients met clinical criteria for ARDS [29].

To address these gaps in knowledge, we designed the current study to test the hypothesis that in critically ill patients with sepsis, the extent of glycocalyx disruption as measured by plasma syndecan-1 levels is associated with development of ARDS and that glycocalyx degradation is more extensive in non-pulmonary sepsis compared to pulmonary sepsis. We further hypothesized that the degree of glycocalyx degradation is associated with other organ dysfunction and adverse clinical outcomes including in-hospital mortality. Finally, we hypothesized that neutrophil activation may contribute to enzymatic cleavage of syndecan-1 from the surface of endothelial cells, and we tested this hypothesis by measuring myeloperoxidase (MPO), a circulating marker of neutrophil activation [26].

Degradation of the endothelial glycocalyx has been increasingly recognized as an important contributor to the pathophysiology of sepsis [13]. Several studies have demonstrated elevated syndecan-1 levels as a marker of glycocalyx degradation in patients with sepsis. In a study of 150 patients, Steppan et al. [9] compared patients with sepsis and patients after major abdominal surgery to controls and found highest levels of syndecan-1 and inflammatory markers in sepsis patients. Two other small observational studies of eighteen [42] and twenty [11] patients with septic shock demonstrated elevated syndecan-1 levels compared to controls. However, to our knowledge, this is the first large-scale study of syndecan-1 as a biomarker of risk of ARDS and other organ dysfunction in patients with sepsis. Similarly, it is the first study to assess the independent association of this biomarker with sepsis mortality. Our findings suggest that in patients with sepsis, the severity of glycocalyx degradation, as measured by syndecan-1, is strongly associated with organ dysfunction and mortality, but contrary to our primary hypothesis, is only associated with development of ARDS in patients with non-pulmonary sepsis.

The finding of higher syndecan-1 levels in patients with ARDS due to non-pulmonary sepsis compared to ARDS due to pulmonary sepsis suggests that degradation of the endothelial glycocalyx may be more prominent in the pathophysiology of non-pulmonary sepsis. Concordant with this finding, elevated syndecan-1 levels were only associated with development of ARDS in patients with non-pulmonary sepsis. These observations are concordant with our prior observation that other biomarkers of endothelial injury such as angiopoietin-2 and von Willebrand factor antigen are more elevated in non-pulmonary sepsis patients with indirect ARDS compared to pulmonary sepsis patients with direct ARDS [24]. Similarly, the predominance of heparan sulfate fragments in indirect respiratory failure reported by Schmidt et al. [29] likely indicates a greater degree of glycocalyx degradation in that patient group. Taken together, these findings provide strong evidence that injury to the endothelial barrier is more severe in indirect mechanisms of acute lung injury than in direct mechanisms. These findings may explain, in part, why a recent study of a large group of patients found that in contrast to direct ARDS, where age and the severity of ARDS were independent predictors of mortality, in indirect ARDS, only the number of non-pulmonary organ failures was independently associated with mortality [43].

The diagnosis of ARDS is still based on clinical criteria and does not rely on underlying pathophysiology [33]. As a result, current definitions identify a highly heterogeneous group of patients that may have different underlying mechanisms of acute lung injury and different responses to therapy. Using a latent class analysis approach, Calfee and colleagues reported that distinct subphenotypes of ARDS can be identified that respond differently to experimental therapies [44, 45]. The current findings provide further evidence of the heterogeneity of ARDS in humans as defined by the extent of endothelial glycocalyx degradation. If validated, syndecan-1 might be useful as a biomarker to help further distinguish molecular phenotypes of ARDS.

Alternatively, plasma syndecan-1 levels might be used to identify subgroups of sepsis patients for therapy targeted at protection or restoration of the endothelial glycocalyx. In patients with severe trauma and hemorrhagic shock, glycocalyx shedding causes coagulopathy and perturbations in fibrinolysis [8]. Furthermore, recent evidence suggests that transfusion of plasma reduces syndecan-1 shedding and reconstitutes the glycocalyx in hemorrhagic shock [14]. Only 22 patients in the current study received transfusion of fresh frozen plasma prior to blood sampling, and we did not find any relationship between receipt of plasma and syndecan-1 levels (data not shown); however, our power for this analysis was low. Another therapy that may affect the glycocalyx is sevoflurane, which has been shown to reduce glycocalyx shedding and leukocyte adhesion in animal models of ischemia–reperfusion injury [46‐48].

This study has several strengths including the large sample size and detailed prospective phenotyping for ARDS and other organ dysfunction as part of the parent VALID cohort study. There are also some limitations. First, the observed association between syndecan-1 shedding and organ dysfunction and clinical outcomes does not prove causation, nor does it elucidate the underlying mechanisms by which glycocalyx shedding contributes to organ dysfunction. Although we hypothesized that neutrophil activation contributes to enzymatic cleavage of syndecan-1 from the endothelial cell surface, we did not find a strong correlation between MPO, a circulating marker of neutrophil activation and syndecan-1 levels. Second, we intentionally selected patients who were more severely ill and more likely to develop ARDS by requiring an APACHE II score of 25 or greater and mechanical ventilation. These inclusion criteria were selected to maximize the likelihood of studying patients with a significant degree of glycocalyx degradation. However, the findings may not be generalizable to a less severely ill patient population, nor are the findings necessarily applicable to patients at risk of ARDS from non-septic causes or to patients with surgical critical illness. Finally, it is not possible to know from the current study what fraction of elevated syndecan-1 levels is due to increased shedding of the glycocalyx, versus impaired clearance of syndecan-1. If syndecan-1 were cleared primarily in the liver or kidney, then the association between elevated levels and kidney and liver dysfunction might simply reflect impaired clearance. However, very little is known about the clearance of syndecan-1 from the circulation. In one study of patients with chronic kidney disease, plasma syndecan-1 clearance was not a function of creatinine clearance [49].

In summary, the current findings highlight the potential importance of disruption of the endothelial surface layer in the pathogenesis of organ dysfunction in sepsis. Further studies are warranted to validate the current findings in other patient populations and to determine the precise mechanisms of organ injury that occurs in association with glycocalyx degradation. In addition, this study adds to the growing body of evidence that sepsis and ARDS are not homogenous disease states. Further characterization of molecular phenotypes of organ dysfunction and acute lung injury in sepsis may help to better target therapies in the future, including therapies targeted at protection and restoration of the endothelial glycocalyx.

In conclusion, our study evaluated the severity of endothelial glycocalyx degradation and its impact on ARDS, non-pulmonary organ dysfunction, and mortality in a cohort of critically ill patients with sepsis. Contrary to our initial hypothesis, elevated syndecan-1 levels were associated with ARDS only in a subgroup of patients with non-pulmonary sepsis, suggesting that degradation of the glycocalyx is more severe in patients with non-pulmonary sepsis, and adding to the growing body of evidence that the mechanisms underlying direct and indirect causes of ARDS are distinct. Regardless of etiology of sepsis, elevated syndecan-1 levels were associated with non-pulmonary organ dysfunction and in-hospital mortality. Together, these findings suggest that measurement of syndecan-1 levels in patients with sepsis may be useful for identifying patients at high risk of organ dysfunction and mortality and those who may benefit from therapies targeted at protecting or restoring the glycocalyx.

Ethics approval and consent to participate: The VALID study was approved by the Vanderbilt Institutional Review Board (#051065). Informed consent was obtained from patients or their surrogates prior to enrollment. Due to the minimal risk of this observational study, a waiver of informed consent was granted by the Institutional Review Boards for patients who were unable to participate in the informed consent process and for whom no surrogate decision maker was available.