Human glycocalyx shedding: Systematic review and critical appraisal

The number of studies measuring breakdown products of the glycocalyx in plasma has increased rapidly during the past decade. The purpose of the present systematic review was to assess the current knowledge concerning the association between plasma concentrations of glycocalyx components and structural assessment of the endothelium.


| INTRODUC TI ON
The glycocalyx is a 0.2-5 µm thick layer of glycosylated proteins that covers the luminal side of the endothelium throughout the cardiovascular system ( Figure 1). The layer is believed to have great functional importance to local vasodilatation, coagulation, and inflammation. 1,2 Degradation of the endothelial glycocalyx layer ("shedding") occurs in inflammatory states, during ischemia, and after vigorous volume loading. Such shedding is claimed to quickly change the physiology of the endothelium by obstructing local adaptation of blood flow and increasing the capillary permeability for macromolecules, which promotes hypovolemia. 3 Given these properties, there is no wonder that the glycocalyx has received considerable attention over the past decade as demonstrated by the significant increase in publications. The majority of studies on glycocalyx shedding have been performed in laboratory animals, mostly in rats. Here, shedding of the endothelial glycocalyx seems to result primarily from metalloproteinase (MMP) activity, as inhibition of MMPs reduces endothelial glycocalyx degradation in response to inflammatory stimulation. [4][5][6] Increases in circulating heparan sulfates, hyaluronan, and syndecan-1 have been reported in rat models of sepsis 7,8 and hemorrhagic shock. [9][10][11] In terms of human investigations, the most common clinical study design has been to identify situations where the glycocalyx is degraded by measuring glycocalyx breakdown products in the plasma of patients. This has been done for a wide variety of conditions with the hypothesis being that the integrity of the glycocalyx can be related to changes in plasma concentrations of glycocalyx shed products.
In the discussion, we review confounding issues related to measuring plasma components assumed to represent cell-surface constituents but that have not been correlated to induce meaningful structural implications of clinical relevance.

| ME THODS
To obtain a comprehensive dataset on all of the publications evaluating the human endothelial glycocalyx, a search of PubMed using the term "glycocalyx" yielded 3454 publications, returned from the earliest date in the database record, October 1958 to August 2020. The presentation adhered to the PRISMA Statement, whenever applicable. Figure 2 represents a summary of our methodology. The first search was automated using the search terms "glycocalyx," then all abstracts were manually reviewed by one person (VP) to ensure the Methods indicated that the study was: (a) human data, (b) recorded a glycocalyx component: syndecan-1 (sdc-1), syndecan-4 (sdc-4), glypican (Gpc), heparan sulfate (HS), chondroitin sulfate (CS) or hyaluronan (HA), and (c) recorded the source (plasma, CSF, urine). Every abstract published that included the word "glycocalyx" was reviewed in detail. The methods sections were reviewed for the type of study that was performed. Review articles and manuscripts without an Abstract in English or else did not evaluate the "human" glycocalyx, "glycocalyx integrity," and/or "glycocalyx damage" were excluded.
Structural assessments of the glycocalyx were categorized as sidestream dark field (SDF) imaging, orthogonal polarization spectral (OPS) imaging, incidental dark field (IDF) imaging, bright-field imaging and erythrocyte sodium sensitivity studies. A brief explanation of each method is presented in the Appendix. SDF, OPS, and IDF are quite similar methods but with incremental refinement to improve image clarity and resolution.

Editorial Comment
Glycocalyx shedding during inflammatory and ischemic states, and as a response to volume loading, is an very active area of research in recent years in anesthesiology and critical care medicine. This systematic review of glycocalyx shedding provides a detailed update on the current state of knowledge on quantification of glycocalyx shedding as assessed by measurements from plasma and in vivo structural measurements. Discussion is included regarding uncertainties when interpreting plasma measurements of breakdown products from glycocalyx shedding.
We then categorized glycocalyx evaluation in common disease processes and organ systems studied by the authors. In doing so, we categorized common disease and syndromes as: critical illness; endocrine disease; pregnancy; surgery; healthy experimental patients; cardiovascular disease; renal disease; infectious disease; miscellaneous. From there, we identified subcategories in each category based on the models from the reviewed publications. These subcategories are given in the Tables that we chose to present our results. After placing the reviewed publications into one of the major categories and subcategories, the publications were divided into three groups based on whether the publication investigated breakdown products, performed a structural assessment, or performed both.
The publications are listed in four tables using the following methods. We first assigned reference numbers in an ascending fashion to the models that performed an investigation of breakdown products as there were more publications that investigated this.
These were assigned to Tables 1-3. We continued our reference list numerically to include the studies that performed a structural assessment. These were assigned to Table 4.
If a study performed both analysis of breakdown products and structural assessment, it was given the reference number that was labeled in Tables 1-3 and was also placed in Table 4 with the same numerical reference. We indicate an increase (+), decrease (-), or no change (0) in the glycocalyx breakdown product(s) depending on the individual study findings. Methods showing glycocalyx thickness were references as showing decrease (-), increase (+), or no change (0).

| RE SULTS
After reviewing the 3,454 articles, secondary filters narrowed that list to 228 publications that met inclusion criteria of human studies.

| Commonly studied glycocalyx biomarkers
Sepsis and trauma are the most frequently studied conditions and comprise about 40 studies. They usually report 3-4-foldt increased levels of glycocalyx degradation products. Exceptions are quite rare (Table 1, (Table 2).
Fourteen studies have measured glycocalyx degradation substances in cardiovascular disease, which is few when considering that the endothelium is part of the vascular system (Table 3,

| Rarely studied glycocalyx biomarkers
Two studies included data on plasma Gpc; one study 21 measured plasma from septic vs control patients and the second study 164 examined plasma glycocalyx markers (glypican and sdc-1) in relationship to forearm arterio-venous fistula failure. In sepsis, Gpc was increased along with C-reactive protein, lactate, pro-calcitonin, sdc-1, and heparin-binding protein. In the second study, there was no relationship between plasma Gpc-1 or sdc-1 vs fistula failure; however, HA was positively correlated with failure of the arterio-venous fistula.
The five studies that reported on plasma concentrations of CS found increasing plasma concentrations during sepsis and acute respiratory failure, 33,36 trauma, 54 ischemic stroke, 61 and gestational diabetes. 67 The results for sdc-4 in plasma are more variable. One study found no change in sdc-4 during sepsis. 34 In a cohort of ICU patients

| Structural measurements
The most widely used method for structural studies of the endothelial glycocalyx layer is to assess its thickness in the sublingual area by side Stream Dark Field (SDF) imaging. The assumption is that shedding of the glycocalyx layer both elevates the plasma concentration of glycocalyx degradation products and causes a thinning or complete absence of this layer on the SDF image.
When sepsis has been studied by this approach, some studies have found a thinning of the glycocalyx layer while others have found no change (Table 4, top). There are two studies in stroke, and both also show a thinning. Assessment of the glycocalyx thickness was made along with biomarker measurements in 15 studies of which 9 showed a correlation (60%). These studies can be found in Table 4. The common pattern is that several glycocalyx biomarkers become elevated, but in some studies only sdc-1 was increased and not HS. 101,122,126,178,189 128 Elevation of the sdc-1 and HA levels has been reported after correction for plasma dilution, 126 but the validity of such corrections is unproven due to a lack of pharmacokinetic characteristics for these substances.

| D ISCUSS I ON
Animal studies show that the glycocalyx layer might become degraded within 10 minutes, while restoration requires up to one week provided the shedding stimulus is removed. 241 In the rat lung restoration even occurs within 24 hours, but the recovery is dependent on the expression of the fibroblast growth factor receptor which is inhibited by sepsis. 242 These rates cannot be uncritically extrapolated to humans due to interspecies differences in metabolic rate and substrate turnover time. 243 The metabolism of these compounds is very complicated, and with much of the elimination normally taking place in the liver. The kidneys are usually not considered to be of importance, but urinary concentrations actually do not deviate much from the plasma concentrations. A recent report highlighted the kidney as a source of elimination by measuring sdc-1, HS, and HA over 5 hours in healthy volunteers and post-surgical patients. 170 The renal clearance of sdc-1 suggested that the entire plasma pool of free sdc-1 would be completely excreted within 15 hours and that a 6-fold variability in plasma concentration could be explained by acute changes in renal function and not due to increased shedding.
Glycocalyx constituents are assumed to stem from the endothelium despite being widely expressed in the body. Protein expression of sdc-1 is abundant in the liver, digestive tract, kidney, urinary bladder, and bone marrow, but hardly at all in muscle, adipose tissue, tongue, and skin. Protein expression for HS is found in muscle and occurs in the cytoplasma of many cell types, as well as in the interstitial matrix. 244 It is often assumed that measured plasma constituents are uniformly derived from the luminal side of the endothelium throughout the vascular tree but, in fact, we do not know from where in the body they originate. This fact constitutes a bias in the present review, as authors being aware of the widespread distribution of glycosaminoglycans in the body may not always use manuscript titles and search terms that refer to the glycocalyx. 235 We believe that direct measurement of perfused boundary region (PBR) in the microcirculation is, at present, the best method to assess degradation of the glycocalyx. The sublingual vessels have been used as an easily accessible vascular bed to determine the PBR, which is used a surrogate for glycocalyx thickness. In one study, sdc-1 was elevated but there were no changes in the PBR and They also demonstrated that normovolemic hemodilution-induced glycocalyx shedding in rats does not alter vascular permeability to dextrans, albumin or plasma. 248 In chronic diseases, compensatory responses may offset changes in shedding or expression. For example, in one study assessment of the glycocalyx by SDF imaging showed no significant difference in glycocalyx dimensions between patients with and without cardiovascular disease. 227 However, two studies have reported elevations in sdc-1 and HA in patients with cardiovascular disease. 159 If we were to assume that all the data in Table 1 was free from publication bias and issues of renal clearance, the preponderance of data suggests that the glycocalyx is damaged during a wide variety of insults, as shedding increased in 96% of all published reports. If this percentage is valid, then what is the usefulness of a biomarker that increases ubiquitously across many disease states? Although uncertain, we suggest that an increase of five times the baseline in acute disease or trauma might be accepted as evidence of release of glycocalyx degradation products from somewhere in the body.
Smaller changes might be due to short-term fluctuations in metabolism and urinary excretion.

| CON CLUS ION
The utility of measuring glycocalyx breakdown products in the plasma as biomarkers with a predictive value for determining a specific disease or the progression of a disease is unproven. Shedding of glycocalyx components is a ubiquitous process that occurs during both acute and chronic inflammation with no sensitivity or specificity for a specific disease or condition. Uncertainties related to proteoglycan expression levels, turnover rate, shedding, renal clearance and lack of correction for hemodilution cast doubts on many of the reported alterations in glycocalyx breakdown products measured in plasma. There is only a moderately good correlation between plasma concentration and the structural assessment of glycocalyx thickness (60% agreement), which further questions the utility of measuring plasma glycocalyx components as a surrogate for structural and functional alterations. Finally, compensatory expression of glycocalyx constituents may offset the loss of specific components in order to maintain structural integrity.

None.
O