The study was performed at the Hedenstierna Laboratory, Uppsala University, Sweden. Twenty male pigs (Sus scrofa domesticus) of mixed Swedish Hampshire and Yorkshire breeds (mean weight 30.4 ± 1.8 kg) received premedication with Zoletil Forte® (tiletamine and zolazepam) 6 mg/kg and Rompun® (xylazine) 2.2 mg/kg i.m. The animals were placed in supine position after adequate sedation was obtained. A peripheral intravenous catheter was placed in an auricular vein and a bolus of fentanyl of 5–10 µg/kg administered i.v. Anaesthesia was then maintained with ketamine 30 mg/kg/h, midazolam 0.1–0.2 mg/kg/h and fentanyl 4 µg/kg/h, in glucose 2.5% during the experiment. Rocuronium 2.5 mg/kg/h was added as muscle relaxant after adequate depth of anaesthesia was assured by absence of reaction to pain stimulus between the front hooves. Ringer’s acetate was infused i.v. at a rate of 10 ml/kg/h during the first 30 min of the protocol. The animals were under deep anaesthesia during the whole experiment (6 h of peritonitis/sepsis), including during euthanasia (100 mmol KCl i.v.). Bolus doses of 100 µg fentanyl i.v. were administered if signs of distress or reaction to pain stimulus were noted.

The airway of the animals was secured via tracheostomy. A tube of an internal diameter of eight mm (Mallinckrodt Medical, Athlone, Ireland) was inserted in the trachea. Thereafter, volume-controlled ventilation (Servo I, Maquet, Solna, Sweden) was maintained as follows: respiratory rate (RR) 25/min, tidal volume (VT) 8 ml/kg, positive end-expiratory pressure (PEEP) 8 cmH2O and inspired oxygen concentration (FIO2) 0.3. The settings of VT and PEEP were maintained throughout the protocol, while RR was adjusted aiming at a PaCO2 < 6.5 kPa, and FIO2 was adjusted to keep PaO2 > 10 kPa.

A pulmonary artery catheter for measurement of pulmonary artery pressures and cardiac output (CO) and a triple lumen central venous catheter for fluid infusions were inserted via the right jugular vein. An arterial catheter was inserted via the right carotid artery for blood pressure measurement and blood sampling. A PiCCO (Pulse index continuous cardiac output) catheter (Pulsion, Munich, Germany) was inserted via the right femoral artery for estimation of stroke volume variation (SVV) and extravascular lung water (EVLW). Blood gases were analysed immediately after sampling and executed on an ABL 3 analyser (Radiometer, Copenhagen, Denmark). Hemoglobin (hgb) and hemoglobin oxygen saturation were analysed with a hemoximeter OSM 3 calibrated for porcine hemoglobin (Radiometer, Copenhagen, Denmark).

A midline laparotomy was performed. The bladder was catheterized for urinary drainage and an incision was made in the caecum, feces were collected and thereafter the incision in the cecal wall was closed. After insertion of a large-bore intra-peritoneal drain, the abdominal incision was closed.

Study protocol

To yield a stock solution of 1% (10 mg/ml), five grams of HMW-HA 1560 kDa (Sodium hyaluronate Lot# 027362 HA15M-5, Lifecore Biomedical LCC, Chaska, MN, USA) was dissolved in 500 ml 0.9% saline. The solution of 1% HMW-HA was produced under sterile condition in laminar airflow and stored as 50 ml aliquots at − 20 °C. On the day of experiment, aliquots were diluted 1:10 in 0.9% saline, to yield 0.1% concentration.

After the laparotomy and collection of feces, baseline measurements were performed, after which peritonitis was induced via peritoneal instillation of autologous feces (2 g/kg body weight in 200 ml warmed 5% glucose solution). Thereafter the abdominal wall was closed.

Experimental design

The experimental time line is presented in Fig. 1. The animals were randomized in two steps (block randomization, sealed opaque envelope), first to peritonitis (n = 16) or time control (n = 4), then into two treatment groups: intervention with HMW-HA (n = 8) or control group (n = 8). The study was prospective and the researchers were blinded for the group allocation until a master file (Additional file 2) for the whole experiment was produced.

Fig. 1

Experimental timeline. Induction of peritonitis, followed by intervention (2-h infusion) and a total of 6 h of observation period after peritonitis induction (end of protocol)

The intervention was started at the time of the laparotomy with 0.1% solution HMW-HA, administered with a rate of 6 mg/kg/h (6 ml/kg/h) for 2 h [22]. The control group received the same volume of vehicle (0.9% saline, 6 ml/kg/h) as an infusion over 2 h. After 2 h duration of peritonitis 2 g of Piperacillin/Tazobactam in 10 ml of 0.9% saline was administered i.v.

If the animals developed circulatory instability (defined as MAP < 55 mmHg > 5 min) an infusion of Norepinephrine (40 µg/ml) was started with the rate of 5 ml/h and increased stepwise, aiming at maintaining MAP > 55 mmHg. No additional fluids were administered.

Analyses and physiologic parameters

The primary endpoint parameter, SVV, was measured at baseline and every hour for the following 6 h duration of the experiment, simultaneously with EVLW and arterial blood gas analysis. Concomitantly, hemodynamic parameters (systemic arterial and pulmonary arterial pressures, CO, heart rate), respiratory parameters (FIO2, SaO2, ETCO2, peak pressure, plateau pressure, dynamic and static compliance) and urine output were measured. Mixed venous blood gas analysis, collection of plasma samples and arterial blood for bacterial cultures were drawn at baseline and at peritonitis duration of 1, 2, 3 and 6 h.

Cytokine and HA analyses, VAP1, syndecan 1, heparan sulphate

Porcine-specific sandwich ELISAs were used for the determination of TNF-α and interleukin-6 (IL-6) in plasma (DY690B (TNF-α), DY686 (IL-6), R&D Systems, Minneapolis, MN, USA). The ELISAs had total coefficient of variations (CV) of approximately 6%. A commercial ELISA kit (Hyaluronan DuoSet, DY3614, R&D Systems, Minneapolis, MN, USA) was used to measure the hyaluronan concentration. Porcine specific ELISA kits were used to analyze plasma concentration of vascular adhesion protein 1 (VAP 1) (MyBioSource cat. No. MBS9364679), syndecan 1 (MyBioSource cat. No. MBS2703970) and heparan sulphate (MyBioSource cat. No. MBS265068).

Total protein and osmolality

Total protein was analysed using a Mindray BS380 (Mindray, Shenzhen, China) with reagents from Abbott Laboratories (7D73-22, Abbott Park, IL, USA). Osmolality was measured using an OsmoPRO Multi-Sample Micro-Osmometer (I&L Biosystems, Königswinter, Germany).

Bacterial investigations

From a sterile arterial catheter, 0.5 ml blood was drawn for quantitative blood cultures. 100 µl was cultured on three separate cysteine lactose electrolyte deficient (CLED) agar plates and cultured at 37 °C overnight. Colony forming units (CFU) were quantified with viable count technique the following day. The median of counted CFU/mL was calculated. CFU on one of three CLED plates from a timepoint was interpreted as a contamination. More than 1 CFU/mL were considered a positive blood culture.

Statistical analysis

To determine sample size, we used data from a previous peritonitis protocol, where the stroke volume variation was used as a guide to fluid therapy [24]. The control group had a standard deviation of ± 2% at baseline. Aiming at detecting a difference of 3 per cent units of SVV between groups, a power of 0.8 and a significance level of < 0.05 justified a sample size of eight animals in each group.

The Shapiro–Wilk’s test was used to test data for normality. The two-tailed Student’s t test or the Mann–Whitney U test were used to compare the two groups, pending distribution of data. To compare the two groups throughout the experiment we used a mixed model, with the animal as random effect. The Bonferroni correction was applied.

Data are expressed as mean ± SD or median (IQR) according to distribution of data. We conducted the statistical analyses using SPSS v. 28.0.0 software (SPSS, Inc., Chicago, IL, USA). A p value of < 0.05 was considered statistically significant.