Animal housing and all experimental procedures were approved by the local Committee for Experimental Animal Research (Cantonal Veterinary Office Zurich, Switzerland) under the license number ZH119/2019 in conformity with the European Directive 2010/63/EU of the European Parliament and the Council on the Protection of Animals used for Scientific Purposes, and the Guide for the Care and Use of Laboratory Animals procedure (National Research Council [US] Committee for the Update of the Guide for the Care and Use of Laboratory Animals, 2011).

Six female white alpine sheep (Ovis gmelini aries) were included in this acute in vivo trial. All sheep were mature (2–5 years of age) and had a mean body weight of 75.6 ± 12.1 kg. After arrival, the sheep were housed in groups, with at least two animals in one pen, and allowed an adaptation period to the new environment of at least 1 week. The health status of the animals was guaranteed by the conventional livestock supplier through regular screening and subsequent official attestation according to the guidelines of the Swiss Federal Food Safety and Veterinary Office (FSVO) and determined via clinical examination by an on-site veterinarian. The animals had ad libitum access to water and hay. A salt licking stone was provided for all sheep. Room temperature was kept at a constant 19°C with a relative humidity of 45–55%. Experiments within the scope of this study were performed during the winter months in the northern hemisphere. Lights were on in all animal rooms from 7:00 am to 5:00 pm, and each room in the facility was equipped with an additional skylight to provide adequate daylight of approximately 10 h. The general condition of each animal was checked twice daily by the animal caretakers and regularly by a veterinarian.

One day before surgery, sheep were deprived of food for approximately 16–18 h. They received water ad libitum during the fasting period. On the day of surgery, an intravenous catheter (Braunüle^®MT Luer Lock, 14 G, B. Braun Medical, Melsungen, Germany) was placed in the jugular vein. The animal was then pre-medicated with 3 mg/kg BM ketamine hydrochloride (Ketasol^®100, Dr. E. Graeub AG, Bern, Switzerland) and 0.3 mg/kg BM midazolam (Dormicum^® Roche Pharma, Reinach, Switzerland) intravenously. The animal was placed in a sternal position, and, prior to orotracheal intubation, anesthesia was induced with 2–5 mg/kg BM of propofol (Propofol-^®Lipuro 1%, B. Braun Medical AG, Sempach, Switzerland) intravenously. A double-lumen urinary catheter (Rüsch Gold 3way, 18 F, Teleflex Medical Enterprise, Athlone, Ireland) was inserted into the bladder under sterile conditions. The double-lumen catheter provided another port for continuous intravesical pressure (IVP) measurements. An orogastric tube (Rüsch 12 mm, Teleflex Medical Enterprise, Athlone, Ireland) was placed to drain excess gastric fluid and relieve ruminal gases. A naso-esophageal temperature probe was inserted through the ventral nasal meatus for continuous body temperature monitoring. Anesthesia was maintained throughout the experiment by inhalation of 1–2% isoflurane (Attane™ Isoflrane ad.us.vet., Piramal Enterpr. India, Lyssach, Switzerland) (fresh gas flow 1–1.5 L/min, 18 breaths/min, tidal volume 10–15 mL/kg BM, FiO2 0.7, Pmax 30 mmHg) in an oxygen/air mixture via volume-controlled ventilation through a semi-closed breathing circuit (Dräger Primus^®, Dräger Medical, Lübeck, Germany). Anesthesia was balanced with a constant rate infusion of propofol (2–4 mg/kg/h). Intraoperative analgesia was provided by a constant rate infusion of sufentanil (Sufenta^®forte, Janssen-Cilag AG, Zug, Switzerland; 2.5 μg/kg BM/h, i.v.). Ringer solution (Ringerfundin^®, B. Braun Medical AG, Sempach, Switzerland) was administered intravenously at an infusion rate of 5 mL/kg/h.

Corresponding pressures were measured at 11 body locations by inserting catheters endovascularly or surgically in their respective positions (Figure 1). After implantation, the catheters were connected to a pressure acquisition device.

Intraventricular and intrathecal transducers were zeroed to atmospheric pressure at the level of the lateral ventricles to obtain hydrostatic equivalence between both. The arterial and central venous sensors were zeroed at the right atrium level, and the jugular venous sensor was zeroed at the tip of the jugular pressure probe. The sheep remained in sternal recumbence throughout the experiment. This posture corresponds to the horizontal position in humans, where ITP and ICP are assumed to be unaffected by hydrostatic fluctuations. IVP was measured via a two-lumen urinary Foley catheter, with one opening connected to the pressure transducer and the second opening serving as a urinary outlet. Urine was removed from the bladder before each intervention. After urine removal, a standard volume of 20 mL saline solution was infused into the bladder via the catheter to standardize IVP measurement.

Pressure data were acquired during an undisturbed 10-min baseline, using the commercially available software Ponemah v5.1 (Data Science International, ST. Paul, MN, United States) combined with the ACQ-7700 acquisition unit. To amplify the signal, the Universal XE and ABCD 4 were used. The catheters were connected to DTXPlus pressure transducers (Argon Medical Devices, The Hague, Netherlands). Data acquisition was performed at a sampling frequency of 1 kHz, then discriminated to 100 Hz and analyzed using custom scripts written in Python 3.7.10 (Open Source, Python Software Foundation, Wilmington, DE, United States). Baseline mean pressures and pulse amplitudes were calculated as 5-min arithmetic means, and values are described as mean ± SD. For each sheep, the 10-min cABP baseline data was separated into its respective frequency components via Fast Fourier Transform (FFT), followed by the separation of the cardiac band. The heart rate was then calculated individually for each sheep as the peak of the cardiac band within the FFT spectrum. Finally, the arithmetic means and standard deviation between sheep were calculated in Microsoft Excel (Microsoft Corporation, Redmond, WA, United States).

At the end of the experiment, sheep were humanely euthanized under deep general anesthesia by administering pentobarbital (Eskonarkon^®ad.us.vet., Streubli Pharma AG, Uznach, Switzerland) (75 mg/kg BM) intravenously.

Table 1 lists the pressures for each sheep. The average mean ICP for all sheep was 13.9 ± 9.4 mmHg, and the average mean ITP was 10.6 ± 12.0 mmHg. Sheep B and F had an ICP and ITP above 20 mmHg during baseline measurements.

Sheep A and D had ITP values leveling around 0 mmHg but displayed physiologic ICP levels in a range of 2.2 and 10 mmHg. Sheep C showed comparable ICP and ITP values in a physiologic range of 6.5 and 6.2 mmHg. Sheep E had a greater difference between ICP with 16.7 mmHg and ITP of only 5.6 mmHg. The ITP in sheep E showed the most marked fluctuation, as evidenced by the relatively high standard deviation of 4.6 mmHg compared to the other sheep.

A complete list of abdominal pressures of each sheep can be seen in Table 3. The mean IVP for all sheep was 12.2 ± 2.2 mmHg. The mean IAP_cr was 13.5 ± 11.8 mmHg, whereas the IAP_cd was 11.7 ± 14.2. For the IAP_ds, the mean was calculated at 12.9 ± 14.2, and the mean IAP_ventr was calculated at 11.7 ± 11.5. IVP measurements presented comparable values for all sheep, with the lowest IVP at 9.4 ± 9.5 mmHg and the highest at 15.6 ± 8.1 mmHg. The IAP_cr showed higher pressure value variability than the IAP_cd, where only sheep F had an average increased IAP_cd of 40.4 ± 3.4 mmHg. The IAP_ds behaved similarly to the IAP_cd, whereas the IAP_ventr does not show consistency.

In 3 out of 6 sheep, a baseline measurement correlation between ICP and ITP was found and, during the subsequent perfusion tests, communication existed in 4 out of 6 sheep. ICP and ITP did not correlate in Sheep A, D, and E during baseline measurements, but Sheep D displayed communication between ICP and ITP during tap and infusion tests. In 3 out of 6 sheep, ICP was high when arterial blood pressure was low, compared with the remaining animals. When comparing mcABP and mfABP, it was found that both pressures agreed in all of the four sheep in which these pressures were measured (Sheep C, D, E, and F). In all four sheep, in which JVP was measured (Sheep C, D, E, and F), JVP displayed higher pressure levels where there was increased ICP. ICP was higher in sheep B, D, and E and, in these three sheep also displayed a higher ABP compared with the other sheep. In Sheep F, where ICP was markedly more elevated with 27.2 mmHg, ABP remained lower (64.4 mmHg), although an increased JVP of 7.3 mmHg was observed. During the baseline measurements, no correlation between IVP and IAP in any of the four quadrants was found. The same conclusion was drawn when comparing abdominal pressures with all hemodynamic parameters and CSF pressures. Figure 4 presents a visualization of the pressure correlation results.

In humans, craniospinal pressure is commonly measured while the subject is lying in a horizontal recumbent position. The pressure is assumed to be the same along the spinal subarachnoid space and inside the cranium (Davson et al., 1987; Lenfeldt et al., 2007). Taking a simplified view, it is assumed that the pressure gradients between the intracranial and along the subarachnoid space are formed based solely on hydrostatic differences. Although this simplistic mechanism of regulating craniospinal pressure gradients has been questioned, at least in horizontally positioned cats, a uniform CSF pressure behavior has proven true (Klarica et al., 2014). We aimed for a comparable baseline situation by placing all sheep in a sternal position. However, the different body constitutions of each individual sheep may have altered the horizontal line between the cerebral ventricle and the lumbar subarachnoid space to a rather convex shape, causing posture-dependent hydrostatic pressure gradients. Although the data were corrected for this hydrostatic pressure, ICP and ITP pressure were aligned in only 3/6 sheep (B, C, and F), while in the other 3 animals (A, D, and E), ITP was markedly lower than ICP. Volume infusions and extractions performed within the same series of experiments showed non-communication between the ITP in the subarachnoid space and the intracranial CSF space in sheep A and E (Podgoršak et al., 2022). Considering the simplified bulk flow theory, CSF is thought to be produced by the choroid plexus throughout the ventricular system, from where it flows through the interventricular foramina of Monro into the third ventricle in the midbrain, before continuing through the cerebral aqueduct, to the fourth ventricle. Afterward, it exits through the foramina of Luschka and Magendie into the central spinal canal and subarachnoid space (Brodbelt and Stoodley, 2007; Brinker et al., 2014; Yamada, 2014). One can only speculate about the underlying causes of non-communication between the craniospinal space and the right lateral ventricle, where our ICP pressure probe was placed. In the non-communicating sheep, the intracranial baroreflex, described by Vari et al. (2021), could be triggered by intrathecal volume infusion, as shown recently by our research group by performing infusion and extraction tests in this present sheep model (Podgoršak et al., 2022). It might be assumed that the induced change in volume and pressure still reached the intracranial space, but not the right lateral ventricle.

Previous studies reported 4.0 ± 1.1 mmHg (Guild et al., 2018), 6.4 ± 1.9 mmHg (Wells et al., 2015), and 7.0 ± 1.0 mmHg (Vari et al., 2021), respectively, as physiological ICP values in sheep. While 4/6 sheep in this study had ICP levels following normal physiologic values, 2/6 (Sheep B and F) had an ICP value > 20 mmHg, which is considered pathologic (Jantzen, 2007). An increased ICP in sheep can have a variety of underlying causes, such as traumatic brain injury (Vink et al., 2008), ischemic stroke (Wells et al., 2015), or infection (Finnie, 2003). Furthermore, anesthesia-related causes such as hypoxia (Iwamoto et al., 1991) or hypercapnia (Cassaglia et al., 2008), anesthetic agents such as isoflurane or desflurane (Artru et al., 1994; Brosnan et al., 2002), and animal positioning (Brosnan et al., 2002) are known contributors to intracranial hypertension (ICH). As ventilation was controlled in all sheep in the present study, hypoxia or hypercapnia can be widely excluded as a cause for increased ICP. The decreased mean ABP measured in Sheep F might indicate a deeper anesthetic plane, possibly induced by increased isoflurane concentrations. A low mean ABP combined with an increased ICP can reduce CPP, thus leading to the assumption that cerebral autoregulation was hampered by a deep anesthetic plane in this sheep (Dagal and Lam, 2009). Another plausible explanation for ICH in Sheep F might be the increased IAP measured in the caudal and dorsal quadrant. An increase in IAP has been shown to lead to a cranial displacement of the diaphragm, increased intrathoracic pressure, and compression of the vena cava, which in turn will increase sagittal sinus pressure and ICP (Rosenthal et al., 1998). As sagittal sinus pressure was not assessed, and JVP is thought to increase only at ICP levels of > 40 mmHg (Johnston and Rowan, 1974), no conclusive statement can be made at this point. Sheep B, on the other hand, showed an increased MAP and an increased heart rate, indicating an autoregulatory response to an increased ICP to sustain CPP. All IAPs were in a physiological range between 0 and 13 mmHg (Hartman, 1973) in sheep B. Therefore, the elevation in ICP is suspected of having been induced by catheter placement or an injury to the cribriform plate through the naso-esophageal insertion of the temperature probe. Disturbance of the lymphatic drainage at the cribriform plate reduces CSF drainage in sheep (Silver et al., 2002) and consequently increases ICP.

Jugular venous pressure is not commonly measured during clinical interventions in humans because there is no exact correlation with the right atrial pressure of the heart and it is, furthermore, often mismeasured. However, some authors believe that JVP measurement is less invasive than CVP measurements and yields fewer complications (Abdullah et al., 2011; Chua Chiaco et al., 2013). According to clinical studies performed in humans, there is no significant pressure difference between the external JVP and the CVP in the supine position (Parker et al., 2002; Leonard et al., 2008). As seen in humans during the perpendicular posture, the collapse of the external jugular vein acts as a preventive mechanism by constricting blood flow from the cranium to the heart. This mechanism maintains the CPP and ensures a stable ICP (Holmlund et al., 2018). This collapse can lead, in turn, to a difference between JVP and CVP. To our knowledge, in sheep, JVP and the total collapse of the common jugular vein in an upright position have not yet been investigated. These data would have to be addressed in a future study.

In horses, it was found that CVP correlated well with JVP in the lateral recumbency (Tam et al., 2011), whereas one study in dogs and cats found a poor correlation between JVP and CVP measurements (Chow et al., 2006). In our present study, one animal had a negative JVP, likely associated with a blockage in the vascular sheath or valves lying proximal to the pressure probe tip. Elevations in intrathoracic pressure also influence the variability between CVP and JVP and can vary considerably during spontaneous breathing (Leonard et al., 2008). In 3/4 animals, JVP was higher than CVP. Since our sheep were ventilated via volume-controlled ventilation during baseline measurements, we suspect that ventilation caused the high variability in JVP and CVP measurements. In the present study, JVP was measured to get as close as possible to sagittal sinus pressure levels, which we have not measured directly.

Nevertheless, as JVP measurements were not reliable in this setting, since both catheters were placed in the same jugular vein, no conclusions regarding sagittal sinus pressure can be drawn. Interference between both pressure probes cannot be excluded.

The two pressures that behaved similarly throughout the baseline measurements in all sheep, were both mcABP and mfABP. Even in sheep with elevated intra-abdominal pressures, comparable mcABP and mfABP following previous experiments in pigs were found (Aksakal et al., 2012). We therefore conclude that both locations can be used for systemic ABP measurements equally if the surgical setting or study design allows it.

The IAP is an essential, and often underestimated or neglected, factor for optimal VP shunt function (Sahuquillo et al., 2008). The indirect measurement of IAP via the bladder is still the standard method in humans and animals (Lee et al., 2002; Al-Abassi et al., 2018), completely neglecting gravitational and shear-associated abdominal pressure gradients (Loring et al., 1994; Hurcombe and Scott, 2012). Aside from a gravitational gradient directly affected by body positioning, IAP fluctuates markedly with respiration, abdominal muscle activity, and shear deformation of the visceral organs due to intra-abdominal fat depositions and changes in the filling status of the hollow organs (Loring et al., 1994; De Keulenaer et al., 2009; Wilson et al., 2010). Sheep have a digestive tract consisting of four stomachs: the rumen, the reticulum, the omasum, and the abomasum. The rumen is the first and largest digestive organ, occupying the entire left part of the abdomen and accounting for 50–60% of the abdominal volume (Medjekal and Ghadbane, 2021). The ruminal content is partitioned into three primary zones based on their specific gravity. Gas rises to fill the upper regions, grain and fluid-saturated roughage sink to the bottom, and newly arrived roughage floats in a middle layer. IAP measurements in goats have been shown to coincide with the fluid surface level within the rumen. Higher pressures were recorded on ventral positions in the abdomen, and were shown to be directly proportional to vertical distance to the fluid surface (Hartman, 1973).

Interestingly, pressures recorded at identical points on the two sides of the abdomen were comparable (Hartman, 1973). The striking effect of the circular (from ventral to dorsal) ruminal contractions mixing up the ruminal contents once or twice per minute makes the sheep an intriguing animal model in abdominal pressure gradient assessment (Ruckebusch and Tomov, 1973). While IVP measurements in our setting delivered comparable values between all sheep, inter-individual variability of measurements in the four quadrants was much higher. While all sheep were fasted for 16–18 h prior to surgery, individual filling levels of the rumen and different nutritional status are plausible explanations for the increased variance. In 2/6 sheep (sheep A and D), IAP measurements followed gravitational gradients described by Hartman (1973) (Dagal and Lam, 2009). Namely, increased ventral and cranial pressures and lowered dorsal and caudal pressures. Increased dorsal pressure was measured in 3/6 sheep (Sheep C, E, and F). Reticulo-ruminal hypomotility caused by general anesthesia subsequently leads to a suppression of gas eructation and the onset of bloat (Dunlop et al., 1997). Although an orogastric tube was placed in all sheep to avoid any increased accumulation of gas, obstruction or dislodgement of the tube might have caused ruminal bloat and increased dorsal pressure in these three sheep. Abdominal pressures in the ventilated sheep showed different curve forms than in the spontaneous breathing sheep, which needs to be considered in VP shunt development using sheep as an animal model.