Hyaluronic acid plasma levels during high versus low tidal volume ventilation in a porcine sepsis model

Introduction

The endothelial glycocalyx (EG) is a thin gel-like layer composed of proteoglycans and glycoproteins found on the luminal surface of healthy blood vessels (Ruane-O’Hora, Ahmeda & Markos, 2020; Uchimido, Schmidt & Shapiro, 2019). First described in the 1960s, many physiological interactions between the glycocalyx and endothelial cells have been discovered which influence physiological and pathological processes. It has been postulated that the EG plays an important role as part of the vascular barrier (Sieve, Münster-Kühnel & Hilfiker-Kleiner, 2018), in mechanotransduction (Tarbell & Pahakis, 2006), in the development of diabetes mellitus (Dogné, Flamion & Caron, 2018) and atherosclerosis (Sieve, Münster-Kühnel & Hilfiker-Kleiner, 2018), in ischemia/reperfusion injury (Dogné & Flamion, 2020), as well as in inflammation and sepsis (Dogné & Flamion, 2020; Uchimido, Schmidt & Shapiro, 2019).

In the last years, several experimental and clinical studies have shown that sepsis damages the EG. This results in endothelial dysfunction and may initiate the development of edema. The resulting impaired lung function eventually leads to the development of acute respiratory distress syndrome (ARDS) (Huang et al., 2018; LaRivière & Schmidt, 2018). The same applies to the brain: Shedding of the EG leads to an increased capillary permeability resulting in edema (Jin et al., 2021; Zhu et al., 2018). In these studies, EG degradation was measured indirectly by increased plasma concentrations of EG components, such as heparan sulfate, syndecan-1, or hyaluronic acid (HA).

One of the cornerstones in the management of ARDS during sepsis is protective mechanical ventilation with low tidal volumes (V_t 6 ml kg⁻¹ body weight), which reduces mortality and barotrauma and provides a higher weaning rate than conventional ventilation strategies (Maccagnan Pinheiro Besen, Tomazini & Pontes Azevedo, 2021).

Mechanical ventilation itself can induce an inflammatory response, thus inducing shedding of the EG as it is happening during sepsis. Several studies describe effects of different ventilation strategies on systemic inflammation. It has been demonstrated that a protective ventilation strategy during major surgery leads to an attenuated inflammatory response compared with non-protective ventilation (Michelet et al., 2006; Wrigge et al., 2004). It has been shown, that protective ventilation in ARDS reduces cytokine response compared to ventilation with higher tidal volumes and less PEEP (Ranieri et al., 1999).

Despite these findings, it remains unknown whether different ventilatory strategies influence the shedding of the EG by reducing or aggravating an inflammatory response. Therefore, we proposed two hypotheses:

• – Sepsis leads to depletion of EG as measured by plasma HA levels.

• – The selected V_t influences the extent of pulmonary but not cerebral EG depletion, measured by increased plasma levels in the pulmonary circulation, but not in the cerebral circulation.

Methods

Experimental protocol

Only animals that achieved ARDS criteria were included in the intervention groups (Tiba et al., 2021).

After instrumentation and a brief stabilization period of 15 min, baseline measurements were performed. The animals were then randomly divided into three groups:

• (1) LPS administration with low tidal volume ventilation with (LTV) (7 ml kg⁻¹; n = 7)

• (2) LPS administration with high tidal volume ventilation (HTV) (15 ml kg⁻¹; n = 7)

• (3) No LPS administration, with low tidal volume ventilation (no LPS; n = 6)

In groups (1) and (2) systemic septic shock-like conditions were simulated by intravenous administration of lipopolysaccharides (LPS; Echerichia coli O111:B4, Sigma-Aldrich, St. Louis, MO, USA) starting with 160 µg kg⁻¹ h⁻¹ for 30 min and continuous administration of 16 µg kg⁻¹ h⁻¹ for the remainder of the experiment. Premature euthanasia of the animals was foreseen if animals did not achieve ARDS criteria, unpredictable stress and pain were observed or if uncontrollable complications such as excessive bleeding or intubation problems occurred during the preparation. Six animals did not achieve ARDS and were euthanized like described below. Major complications never occurred.

All the animals were ventilated with a positive end-expiratory pressure (PEEP) of 5 mbar, an inspired oxygen fraction of 0.4, and a variable respiratory rate to maintain normocapnia. When peripheral oxygen saturation fell below 93% for more than 5 min, the ventilator settings were changed according to a defined schedule (Table 1). The no LPS group was observed for 6 h and received the same ventilation as the LTV group, but with no administration of LPS (Fig. 1).

Table 1:

Intervention scheme.

	LTV		HTV
Step 1	FiO2 0.4	PEEP 5 cmH2O	FiO2 0.4	PEEP 5 cmH2O
Step 2	FiO2 0.4	PEEP 5-8 cmH2O	FiO2 0.45	PEEP 5 cmH2O
Step 3	FiO2 0.4	PEEP 8-10cmH2O	FiO2 0.5	PEEP 5 cmH2O
Step 4	FiO2 0.5	PEEP 10 cmH2O	FiO2 0.55	PEEP 5 cmH2O
Step 5	FiO2 0.6	PEEP 10 cmH2O	FiO2 0.6	PEEP 5 cmH2O
Step 6	FiO2 0.6	PEEP10-14cmH2O	FiO2 0.65	PEEP 5 cmH2O
Step 7	FiO2 0.6	PEEP 15 cmH2O	FiO2 0.7	PEEP 5 cmH2O
Step 8	FiO2 0.7	PEEP 15 cmH2O	FiO2 0.75	PEEP 5 cmH2O
Step 9	FiO2 0.8	PEEP 15 cmH2O	FiO2 0.8	PEEP 5 cmH2O
Step 10	FiO2 0.9	PEEP 15 cmH2O	FiO2 0.85	PEEP 5 cmH2O
Step11	FiO2 1.0	PEEP 15 cmH2O	FiO2 0.9	PEEP 5 cmH2O
Step 12			FiO2 0.95	PEEP 5 cmH2O
Step 13			FiO2 1.0	PEEP 5 cmH2O

DOI: 10.7717/peerj.12649/table-1

Note:

If SpO₂ dropped below 93% for > 5 min.

For fluid replacement, Sterofundin ISO (Braun, Melsungen, Germany) was infused continuously at a rate of 10 ml kg⁻¹ h⁻¹. To avoid hemodynamic deterioration, a standardized fluid bolus (250 ml hydroxyethyl starch over 30 min; Volulyte 6% 130/0.4, Fresenius Kabi, Bad Homburg, Germany) was administered during preparation, followed by 90 ml h⁻¹ hydroxyethyl starch during the experiment. To ensure stable hemodynamic conditions, a continuous infusion of norepinephrine was administered during the experiment to maintain mean arterial pressure above 60 mmHg.

At the end of observation, the animals were euthanized under deep anesthesia by central venous injection of propofol (200 mg) and potassium chloride (40 mmol).

Results

Hyaluronic acid concentrations

Analysis of the HA concentration showed no significant difference between the three study groups at baseline. After 3 and 6 h, the HA acid concentrations were found to be significantly higher in the LTV and HTV groups compared to their baseline concentrations and the no LPS group (Fig. 2). In the no LPS group, the 6-h values were higher than the baseline concentrations, but they were significantly lower than the concentrations in the LTV and HTV groups at the same time points. No significantdifferences between the LTV and HTV groups were measurable before and after the brain or lung passage (Fig. 2).

Post-mortem assessment

The pulmonary and cerebral wet-dry ratios showed no significant difference between the experimental groups. The global scoring values and the regional distribution showed no significant difference between the three groups due to the high variances (Fig. 3).

Discussion

The present study found significantly increased plasma HA levels in the central venous, pulmonary, and cerebral circulation in a porcine model of sepsis-induced ARDS. Furthermore, the pulmonary HA release was not specifically influenced by the different tidal volumes in the short run. Therefore, the first hypothesis of the study was confirmed, but the second hypothesis was not.

All animals in the LTV and HTV groups developed ARDS according to the current definitions (Ranieri et al., 2012). The fact that no significant differences in the oxygenation ratio were found when comparing the two sepsis groups is surprising only at first glance. It is well known that the benefits of a protective ventilation regime become apparent only after a period of time that is longer than the 6 h, which was the duration of our experiment (Biener et al., 2013). In contrast, higher tidal volumes are able to recruit atelectatic lung tissue and improve gas exchange, but they represent a double-edged sword by inducing ventilator-induced lung injury in the long run. Furthermore, a previous study demonstrated that LPS impairs hypoxic pulmonary vasoconstriction, which may lead to a greater decrease in the oxygenation rate even with a protective ventilator regime (Davieds et al., 2016; Turzo et al., 2018). HA is an established serum marker for shedding of the EG. A closer look at the EG suggests that HA is the first component of EG shedding because it is closest to the vessel lumen compared with heparan sulfates or even deeper transmembrane proteins (Dogné & Flamion, 2020; Martin et al., 2016).

Consequently, the significant increase in plasma HA levels provides evidence of damage to the EG in both sepsis groups. Previous studies confirm that inflammatory processes and sepsis lead to shedding of EG (Martin et al., 2016).

In the present study, the ventilator regime of LTV versus HTV had no relevant effect on the amount of EG shedding over the 6 h of the experiment. Considering that protective ventilation also requires more time to show positive effects in comparison to a high tidal ventilation regime, it is conceivable that the corresponding effects on EG could also be seen later. Also it must be noted, that the pathophysiology of ARDS is not only explained by shedding of the EG but also the alveolar epithelial barrier. A recent study showed elevated levels of heparan sulfate,a consituent of the alveolar epithelial glycocalyx and EG, in LPS-induced ARDS, which was not collected in our study. (Li et al., 2021). Maybe this parameter would have showed a difference between our experimental groups.

Finally, it is conceivable that the precision of the ELISA used was not sufficient to detect minor changes in the plasma HA levels that could indicate the differential effects of LTV and HTV on EG. The smaller but also significant increase in plasma HA concentration in the no LPS group in comparison to the LTV and HTV groups could be explained by the animal preparation. Furthermore, general anesthesia and mechanical ventilation have been found to lead to an impairment of pulmonary integrity, which could be seen especially in the lung injury score of animals that were mechanically ventilated (Kamuf et al., 2018; Kamuf et al., 2020).

Interestingly, data published by Choi et al. (2006) suggest that mechanical ventilation with higher tidal volumes and without PEEP during major abdominal surgery leads to increased plasma leakage and disruption of the endothelial barrier of the pulmonary vasculature compared with ventilation using lower tidal volumes and PEEP. If this observation is a consequence of EG depletion, one would assume that a difference in EG levels had to be detected at least when comparing the animals in the HTV and LTV groups. It can only be assumed that the expected effect was masked by septic shock. However, Murphy et al. (2017) observed that non-pulmonary sepsis resulted in greater degradation of EG than pulmonary sepsis. This may also suggest a superimposed effect of the septic shock model on subtle differences in EG degradation between the HTV and LTV groups that might otherwise have been detected.

The main limitation of our study is the missing direct measurement of EG. However, the indirect method of serum concentration of HA is widely accepted and has been used in several previous studies (Dogné & Flamion, 2020; Smart et al., 2018). Attempts were made to measure additional parameters, such as heparan sulfate and Syndecan-1, with commercially available ELISA kits. Unfortunately, no valid results could be measured with these kits in our model.

The septic shock model used for the LTV and HTV groups required the use of norepinephrine to ensure stable circulatory conditions. In this regard, MAP was maintained above 60 mmHg throughout the 6-h time period of the experiment. Higher MAP values might have been achievable with more generous fluid therapy. Considering potential damage to the EG from atrial natriuretic peptide release or volume overload, this option was discarded (Jacob et al., 2013).

The significantly higher dosage of norepinephrine in the LTV group in comparison to the HTV group can be explained by the significantly higher PEEP values in the LTV group; a correlation between the use of higher PEEP and lower MAP due to reduced preload caused by high intrathoracic pressures has been scientifically recognized (Luecke & Pelosi, 2005). The same mechanism serves as an explanation for the significantly reduced cardiac output in the LTV group after 6 h of the experiment. The blood count results showed a significant decrease in leukocytes and platelets in both the LTV and HTV groups; this phenomenon is known to be an immunological response to the administration of LPS and also a criterion for the diagnosis of sepsis. An interesting observation was made regarding hematocrit. It was significantly increased in both sepsis groups, which could be interpreted as a sign of hemoconcentration due to fluid loss across the endothelial barrier impaired by sepsis (Rehm et al., 2004).

Finally, the increased serum lactate levels in the LTV and HTV groups can be explained by impaired microcirculation resulting from septic shock. The small but significant decrease in the serum lactate levels in the no LPS group could be the result of slightly elevated levels at baseline after animal preparation. Neither the pulmonary wet-to-dry ratios nor the cerebral wet-to-dry ratios showed significant differences among the three experimental groups. The chosen time period of 6 h was based on experience from previous animal studies in which LPS was used to induce septic shock, although we used higher concentration of LPS to achieve ARDS based on experiences from preliminary prepared animals (Lipcsey et al., 2010). However, there is a possibility that a longer study period would be required to analyze a capillary leak in order to observe pulmonary edema and, more importantly, cerebral edema In retrospect, the choice of hydroxyethyl starch for hemodynamic stabilization proved to be less than optimal because it has beneficial effects on glycocalyx (Zhao et al., 2020). Additionally, hydroxyethyl starch is no longer used in sepsis due to its negative effects on various organ systems (Rhodes et al., 2017). This can be explained by the fact that, in the present study, this knowledge was not available at the time the experiments were conducted.

Although there were no significant differences in the lung injury score among the three groups, the parameters assessed generally showed high variances. However, it is important to note that the present study was not primarily designed or powered for this analysis. Moreover, several aspects need to be considered to further elucidate the results. A septic shock model with high doses of norepinephrine induces microcirculatory disturbances and, at the same time, LPS attenuates the mechanism of hypoxic pulmonary vasoconstriction (Easley et al., 2009). A sustained effect of pulmonary perfusion on the decrease in the oxygenation rate may explain these discrepancies. Additionally, the reported baseline oxygenation ratio values under baseline conditions were significantly lower than data from recent studies from our laboratory with comparable animals (breed, age, weight, breeder). Therefore, a preexisting or contributory environmental influence cannot be excluded.

In conclusion, this study demonstrated that LPS-induced systemic sepsis caused EG degradation as measured by the plasma HA levels. In this porcine model, different ventilation regimes like LTV and HTV had no influence on EG degradation in the pulmonary or cerebral circulation. Further studies focusing on more EC-specific parameters and a longer study period are needed.

Supplemental Information