Safety evaluation of the Guardian device on the common carotid artery in sheep

Abstract

Background

Pulse pressure intensity in middle-aged adults is a risk factor for dementia. The Guardian device (The Brain Protection Company, Sydney, Australia) has been developed to reduce pulse pressure, as a potential therapy.

Objectives

The aim of this study was to evaluate the safety of the Guardian, a novel pulse modulation device designed to reduce the intensity of the pulse pressure that penetrates into the cerebral small vessels. The Guardian is a helix that gently wraps around the common carotid artery (CCA) to slightly change its shape, to absorb pulsatility, without lowering flow.

Methods

The Guardian was implanted bilaterally on the CCAs of 10 mature sheep for chronic implant periods of 3, 6 or 8 months. The ratio of internal device diameter to outer diameter of the CCA varied from 63% to 92% (n = 20). The implant position on the vessel was marked surgically at implant. Gross pathology and histopathology of the CCA were examined at 3- and 6-months post explant. Most devices were explanted using open surgery, however minimally invasive surgical explant techniques were examined in 2 animals to assess the potential of this approach for explant in humans if required.

Results

The Guardian was successfully implanted with no adverse events, and minimally invasive explant appeared to be viable for removal. Following implant, the device was surrounded by a thin fibrous capsule, with similar pathology at 3- and 6-months. Minimal or no movement was observed. CCA sections appeared histologically normal, with no evidence of thrombosis, stenosis, fibrosis, chronic inflammatory response, or vessel degeneration.

Conclusions

The feasibility of surgical implantation and biomaterial safety of the Guardian was confirmed over 8 months. Minimally invasive explant of the Guardian has the potential to be viable. Further work is required to demonstrate efficacy in vitro and/or in vivo before evaluation in humans.

Highlights

• •Guardian device biomaterial safety was demonstrated in a sheep model.
• •Guardian devices were implanted bilaterally on common carotid arteries for up to 8 months.
• •Potential for surgical explant via minimally invasive surgery if required.

1. Introduction

Although the aetiology of dementia is understood to be multifactorial, numerous recent studies have shed light on a major new risk factor for cognitive deterioration: strong pressure pulsations in the blood vessels to the brain. Population studies in middle-aged adults have linked high pulse pressure to subsequent deterioration in brain function [1,2], and scientific research is revealing the pathways that link cardiovascular risk factors with neurodegeneration [3]. With aging of the vasculature, arteries become stiffer, which leads to greater penetration of high pulse pressure into the cerebral vascular network. The effect of this increased pulse pressure on the brain vasculature has been linked with functional, structural, and metabolic changes, as well as hemodynamic deterioration that could contribute to neuronal dysfunction and cognitive decline [4].

Blood pressure lowering by pharmacologic agents alone may not be sufficient or practical to protect against cognitive decline and dementia. In the SPRINT-MIND study [5], aggressive blood pressure reduction decreased the new onset of mild cognitive impairment, but dementia was not decreased. There was, however, a significantly higher rate of medication-related adverse effects including hypotension, syncope, electrolyte abnormalities and kidney injury [6].

The Brain Protection Company has developed a small device for modulation of cerebral pulse pressure. It can be placed around the common carotid artery (CCA), with the aim of limiting the intensity of the pulse pressure without reducing blood flow. This device, called the Guardian, is a Nitinol-in-Silicone helix that is designed to gently wrap around the CCA to slightly change its shape, since the internal diameter (i.d.) of the device is less than the outer diameter (o.d.) of the vessel on which it is placed. The intention of the device is to absorb excess pulse pressure intensity before it reaches the high-flow, low-resistance blood vessels in the brain. By selectively reducing pulse pressure, the aim of the Guardian is to prevent the downstream microvascular damage and inflammatory neurodegenerative processes that can lead to cognitive decline and progression to dementia.

Sheep (Ovis aries) are an established model for translational research in cardiovascular surgery because they are large animals with similar size, weight, anatomy and physiology to humans [7,8]. This model has been used routinely for cardiovascular surgical device testing, including assessment of ventricular assist devices and cardiac valves [9,10]. Sheep weighing 50+ kg have a comparable total blood volume to that of humans and the dimensions and structure of relevant blood vessels, including the CCAs and the aorta, are similar to those of humans, although the anatomy of the aortic arch and branching is slightly different in sheep from that in humans [11].

The purpose of this study was to examine the safety of the Guardian in a chronic implant large animal model with a cardiovascular system similar to humans. The safety of acute placement of the Guardian was previously assessed in preclinical studies in sheep (unpublished).

2. Methods

2.1. Guardian device

The Guardian is a small implantable extravascular device for placement around the CCA. The Guardian is constructed with composite layers of implantable medical-grade Nitinol, Silicone adhesive, and Silicone elastomer. It is a single-piece device with no articulated parts, constructed as an open spiral helix of 45 mm total length. The Guardian tested in this study (model ESA-1, Sydney, Australia) is being evaluated in an ongoing First-In-Human Acute study in Australia.

The Guardian (Fig. 1a) fits around the outside of the CCA to deflect the artery smoothly and continuously along the length of the artery segment (Fig. 1b). There is gentle deflection because the i.d. of the device is slightly smaller than the o.d. of the vessel. The vessel cross-sectional area is preserved, to maintain flow. The artery resumes its native path on exiting the device. The helix inner diameter of the Guardian is smaller than the artery diameter, but at no point does the device completely encircle the vessel. Thus, the Guardian deflects the artery but does not substantially alter its cross-sectional area. The eventual intention is for long term Guardian implantation in humans. In this experiment, prespecified periods of implantation were chosen to explore the effects of the Guardian and the potential methods of explant, in the unexpected case where this might be needed, in clinical utilization.

Fig. 1.

The Guardian (model ESA-1) schematic illustration (a); and as implanted around an artery (b). Mild deflection of the vessel is evident, due to the pitch and internal diameter of the Guardian. Blue non-resorbable sutures (blue arrows) were placed on adjacent tissue planes to mark the position of the device on the CCA. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

2.2. Animal ethics

This study was approved by the Animal Care and Ethics Committee of Charles Sturt University (CSU), Australia, according to protocol numbers A19048 and A19392. The studies were conducted humanely and in compliance with the Australian Code of Practice for the Care and Use of Animals for Scientific Purposes 8th edition (2013).

Mature (5–6 years old) Border Leister cross bred ewes (Ovis aries, n = 10) weighing 60–72 kg were used for this study.

2.3. Presurgical measurement of common carotid artery dimensions

The o.d. of the CCA vessels were measured using percutaneous ultrasound prior to the day of surgery. Ultrasonic measurement rather than intra-operative measurement was used because the surgical positioning, anaesthesia and surgical intervention could all dynamically impact vessel diameter. The CCA i.d. and wall thickness were measured, to determine the o.d. of the CCAs, for sizing of the Guardian for implant. Measurements were conducted with B-mode ultrasound using a GE Medical Systems Vivid e, coupled with a Logiq e 8L-RS transducer probe (GE Healthcare, Chicago, IL, USA) on non-sedated animals. The sheep were held in the shearing position and the ultrasound transducer was positioned to acquire transverse images for measurement of the average i.d. over the cardiac cycle (from lateral to medial at the widest point in each image) in triplicate, at 3 sites along the CCA (caudal, middle, rostral) on right and left CCAs. Triplicate measurements of wall thickness were also acquired. Average i.d. and vessel wall thickness from these measurements was used to determine the vessel o.d.

2.4. Study design

Ten sheep were implanted with bilateral Guardian devices on the CCAs for 3 months (n = 3; sheep numbers 1, 2 and 4), 6 months (n = 3; sheep numbers 3, 5 and 6) and 8 months (n = 4; sheep numbers 7, 8, 9 and 10). A range of device i.d. to the o.d. of the CCA from 63% to 92% was examined in this study (n = 20). These sizes were chosen based on mathematical models suggesting optimal device sizing to reduce pulse pressure without reducing flow.

2.5. Surgical procedure

2.5.1. Anaesthesia and surgical monitoring

The sheep were fasted, premedicated, induced and maintained with anaesthesia as described previously [12]. A wide bore stomach tube was placed orally to prevent aspiration of ruminal contents and a warming blanket under the sheep was used to maintain body temperature. An intravenous infusion of physiological saline was given via the cephalic vein catheter at surgical maintenance rates (5–10 ml/kg/h) for the duration of the procedure. The sheep were ventilated and throughout the surgical procedure heart rate, respiratory rate, blood oxygen (SpO₂) and carbon dioxide (ETCO₂) levels, arterial blood pressure and ECG were continually monitored and recorded regularly (every 5 min at least), together with the drugs administered. The sheep received prophylactic intravenous injection of the antibiotic cephazolin (20 mg/kg) at induction.

2.5.2. Surgery

The sheep were placed in dorsal recumbency for access to the CCAs. The ventral neck of the sheep and the skin were prepared with serial applications of antiseptic and detergent solution as described previously [12]. The prepared area was isolated with drapes prior to commencement of the surgical procedure.

Analgesia was further maintained throughout the surgery by the intravenous administration of buprenorphine (0.01 mg/kg). An incision via electrocautery in cutting mode was used to open the skin using a standard ventral cervical midline approach extending from the larynx to just rostral to the thoracic inlet. The subcutis was sharply incised and tissues were separated along native tissue planes to access the middle section of the right and left CCAs on each side of the trachea, proximal to the bifurcation of the CCA. The CCA was gently dissected to separate a 5–6 cm length from the internal jugular vein and the vagus nerve for placement of the Guardian.

2.5.3. Device placement

Guardian devices were positioned on the vessels so that the distal end of the device was ≥6 cm proximal to the bifurcation of the CCA. To assess the stability of the positioning of the device on the CCA, the position of the device was marked using a non-absorbable suture placed into adjacent soft tissue at each end of the device. Marking sutures were not placed on the CCA, to ensure that the histological response to the device on the CCA was not confounded by the response to the presence of the sutures. The positioning of the device with the marker sutures was photographed, see Fig. 1b.

The wound was sutured closed. At cessation of anaesthesia, carprofen (4 mg/kg) was administered intravenously, providing 72 h post-operative analgesia. The sheep were transferred to a recovery pen adjacent to the surgical suite. The endotracheal tube was removed when the sheep could breathe independently, and the animal was monitored by veterinary support staff until it could stand and was eating and drinking. Additional buprenorphine (0.01–0.1 mg/kg) was administered post-operatively for analgesia management. Each animal's recovery from surgery was observed and recorded on a monitoring chart.

2.6. Animal monitoring

The sheep were held individually in indoor pens for a minimum of 3 days post operatively and monitored further in small group indoor pens for at least one week for recovery and healing of the wound site. They were then moved to outdoor group pens.

The sheep were fed a 50:50 mix of lucerne and oaten hay per sheep per day, with water provided ad libitum in troughs. They had twice daily monitoring for at least 3 months, and daily monitoring until the end of the implant period, with activity, urination, defecation, appetite, and other clinical observations recorded.

2.7. Explant procedures and observations

At the end of the implant period, the animals were re-anaesthetised as described previously and placed on the surgical table in dorsal recumbency for access to the CCAs. The ventral neck of the sheep was clipped, and the skin was prepared as for the implant procedure. Analgesia was provided as required by the subcutaneous administration of buprenorphine (0.01 mg/kg).

2.7.1. Explant procedures

Most of the implants (13 of 20) were retrieved using open dissection to expose the device on the CCA, with en-bloc resection of the device in position on the vessel. Two additional techniques were used to retrieve the remainder of the implants (n = 7), to develop and explore minimally invasive explant procedures. The first stage involved open dissection to expose the device with retrieval of the device from within its fibrous capsule (n = 5); this procedure was conducted on right and left implants on sheep number 4 (3 month), the right implants on sheep numbers 7 and 8 and the left implant on sheep number 10 (8 month). The further stage involved minimally invasive surgery (n = 2) which was conducted on the left implants on sheep numbers 7 and 9 (8 month).

2.7.1.1. Open resection

For the open dissection, an incision via electrocautery in cutting mode was used to open the skin using a standard ventral cervical midline approach extending from the larynx to just rostral to the thoracic inlet. The middle section of the right and left CCA was isolated by blunt dissection proximal to the bifurcation of the CCA. The CCA was gently dissected to fully expose the vessel, the implant, and the markers of the original position of the implant. The devices were photographed in position on the CCAs following exposure (Fig. 2a). An en-bloc resection comprised the vessel with the device in position and ∼6 cm length of CCA proximal and distal to the implant, to be used as control tissue for histopathology assessment.

Fig. 2.

Open explant of Guardian at 6 months (a); and 8 months (b–e). Arterial forceps indicate the ends of the device (white arrows) and show the position of sutures on adjacent tissue planes that marked the position of the device on the CCA at implant (blue arrows) (a). Explant with exposure of the CCA for development of the minimally invasive surgical explant procedure (b–e). Open dissection to expose the implant (arrows) (b); the fibrous capsule is incised at one end of the device (c); gripping the device with forceps to withdraw it from the capsule (d); and further stage of gentle retrieval of the device from the encapsulating fibrous tissue (e). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

2.7.1.2. Surgical retrieval with exposure of the CCA

The first stage of development of the minimally invasive surgical retrieval procedure involved open dissection to expose the CCA. The open dissection procedure was followed to the point at which devices were exposed and had been photographed in position on the vessel. The device with a thin layer of fibrous encapsulation was dissected off the CCA (n = 2). After explant, for one of these devices the capsule tissue was carefully dissected, to demonstrate that the device could be readily removed from the thin fibrous capsule that surrounded it.

For subsequent explants in this group (n = 3), the fibrous encapsulation at one end of the device was incised and the device was rotated off the vessel out of the capsule. In the initial assessment of this technique, the capsule along the upper surface of the device was exposed (Fig. 2b–e). In subsequent development of this technique, it was found that the device could be rotated out of the capsule by incising it at one end, without full exposure of the capsule along the length of the device.

2.7.1.3. Minimally invasive retrieval of the guardian device

Two devices from two different animals were explanted using a minimally invasive procedure. B-mode ultrasonic imaging was used to assist estimation of the position of the devices along the neck prior to the surgery. The tissue to the depth of the device was sharply dissected in a transverse orientation using a skin incision of 25 mm. Once the device was located, the capsule was incised, the end of the device was gripped with forceps, and it was withdrawn from the capsule envelope by twisting to flatten its profile (Fig. 3a–d). Open dissection was subsequently conducted to retrieve the vessels on which the Guardian had been placed. The fibrous tissue which had been around the Guardian before its explant was used as an indicator of device position and for comparison with the tissue marker position. The CCA vessels were examined grossly and histologically to assess whether device retrieval had caused tissue damage.

Fig. 3.

Explant of a Guardian at 8 months using minimally invasive surgery. Transverse dissection to the depth of the implant (a); gripping the device with forceps (b); initial stage of retrieval of device (c); and the former implant position is indicated above the surgical incision used to retrieve it (d). Open dissection was subsequently conducted to retrieve the vessel on which the Guardian had been placed.

2.8. Identification and gross observations

At explant, after the vessel tissue had been isolated on all sides, dual ligatures were placed at each end of the vessel. The sample was cut between the ligatures at each end and the sample was retrieved. The ligatures on the sample were incised and saline was flushed through the sample. A suture was tied at each end to identify the proximal versus the distal end of the CCA sample; a suture with one tail was used to identify the proximal end of the sample and a suture with two tails was used to identify the distal end of the sample. The vessels (with or without devices) were immersed in neutral buffered 10% formalin solution. The observations at explant comprised: the position with respect to surgical markers placed at implant, the tissue response to the implant, including the degree of fibrosis, and the ease of device explant, if applicable.

2.9. Euthanasia

Following explant of the CCA samples, the animal was euthanised by administration of a lethal intravenous infusion of barbiturate (sodium pentobarbitone 200 mg/kg).

2.10. Histopathology

Histology was conducted on the left and right CCAs at 3- and 6-months post explant.

2.10.1. Histological assessment of the common carotid artery

Where the Guardian remained in position on the CCA at explant, for 13 of 20 implanted devices, the positioning of the device was marked on the vessel tissue and the device was removed for sectioning. The histology sectioning was performed at several sites along the carotid artery as presented in Fig. 4, which shows the location and orientation of Sections A–G. Oblique sections (B–F) aligned with the pitch of the device were used to assess the interface with the device along an extended length (C–E) compared with the control locations (B and F).

Fig. 4.

Histology sectioning plan for Guardian implanted on vessel. Section A: orthogonal control; section B: oblique control (angle aligned with pitch of the device); section C: oblique sample at one end of the device; section D: oblique sample mid-device; section E oblique sample at other end of the device; section F: oblique control; section G: orthogonal control.

2.10.1.1. Sectioning and staining

Sections from positions A–G were embedded in paraffin, cut at 4 μm thickness and stained with haematoxylin and eosin, Masson's trichrome or immunohistochemistry (IHC) for elastin fibres. Elastin was identified using monoclonal anti-elastin antibody produced in mouse (Sigma-Aldrich E4013, Merck, Darmstadt, Germany). The orthogonal and oblique sections at A, B, G and F from the proximal and distal ends were used as controls for assessment of changes to the vessel in the regions C, D and E.

2.10.1.2. Histopathological assessment

Histopathological assessment of the biocompatibility of the device on the vessel comprised examination of the sections for evidence of: thrombosis; structural changes to the vessel wall including hyperplasia and stenosis; and fibrosis, as well as the peri-vascular response to the implanted Guardian, including evidence of any acute inflammatory response; chronic inflammatory response; foreign body multinucleate giant cell response; and necrosis.

3. Results

All (n = 10) animals remained healthy throughout the implant periods of 3–8 months. Wound healing was good and there were no surgical infections evident in any of the animals; there were no adverse events identified, and no abnormal behaviour was observed in any of the animals.

The mean preoperative ultrasound measure of CCA i.d. was 5.6 mm ± 0.8 mm s d. (n = 20). The mean CCA wall thickness was 0.9 mm ± 0.1 mm s d. (n = 20). The o.d. was 7.3 ± 0.9 mm s d. (n = 20).

The ratio of device i.d. to the o.d. of the CCA was varied from 63% to 92% for the CCA vessels implanted (n = 20), see Fig. 5.

Fig. 5.

Ratio of device i.d. to o.d. of CCA for Guardian implants on right and left sides of each sheep in the study (n = 20 implants). Sheep numbers 1, 2 and 4 were implanted for 3 months; sheep numbers 3, 5 and 6 were implanted for 6 months and sheep 7–10 were implanted for 8 months. Right and left implants in sheep 2 and 6 had equivalent ratios of device i.d. to o.d. of CCA.

At implantation, mild deflection of the CCA by the device was evident, as shown in Fig. 1b.

3.1. Gross observations at explant including stability of implants

Dissection along the original surgical sites showed no evidence of ongoing inflammation or infection. The Guardian elicited a benign tissue response as evidenced by the development of a thin fibrous encapsulation surrounding all devices. Gentle palpation indicated that Guardian devices were stable and immobile in position around the CCA. The thin, pliable fibrous encapsulation was similar in all Guardian devices at 3-, 6- and 8-months post-implant, as shown in Fig. 2.

The tissue markers indicated that the position of the device on the vessel was stable over the implant period. Eighteen of the 20 implanted devices did not show any evidence of significant movement. This was harder to assess for the two devices that were removed via minimally invasive surgery and could only be done by observing the fibrous encapsulation remaining on the vessel with respect to the tissue markers. One 3-month implant moved approximately 5 mm cranially and one 8-month implant moved approximately 5 mm caudally.

3.2. Histopathology of common carotid artery

All CCA sections, in all samples examined, appeared histologically normal with no evidence of thrombosis, stenosis, fibrosis, scar tissue, chronic inflammatory response, or necrosis. Whilst there was evidence of a mild inflammatory reaction immediately adjacent to the implants, there was no evidence of disruption, distortion, inflammation, or degeneration in any part of the arteries in the 3-month explants (Fig. 6a–f), and 6-month explants (Fig. 7a–f). Collagen and elastin fibres appeared in normal orientation throughout the tunica intima, tunica media and tunica adventitia.

Fig. 6.

Three-month implants showing oblique control CCA sections (B or F) and oblique device interface sections (C or E). Top panel haematoxylin and eosin sections control site B (a); interface site E (b). Centre panel Masson's Trichrome control site F (c); interface site C (d). Bottom panel elastin immunohistochemistry (IHC) control site F (e); interface site C (f). Labels: tunica intima *, tunica media +, tunica adventitia #. Scale bars: 100 μm.

Fig. 7.

Six-month implants showing oblique control CCA sections (B or F) and oblique device interface sections (C, D or E). Top panel haematoxylin and eosin sections control site B (a); interface site D (b). Centre panel Masson's Trichrome control site F (c); interface site C (d). Bottom panel elastin IHC control site B (e); interface site E (f). Labels: tunica intima *, tunica media +, tunica adventitia #. Scale bars 100 μm.

Evaluation of the interface between the vessel and the device showed that the vessel was intact and normal, with well vascularised connective tissue between the tunica adventitia of the CCA and the device, as is evident in sections at the interface at 6 months (Fig. 8a–d).

Fig. 8.

Six-month implants showing examples from two animals (sheep numbers 3 and 5) of the interface between the device and the CCA in mid-device oblique (D) sections, stained using haematoxylin and eosin (a) and (c); or Masson's Trichrome (b) and (d). Arrows show normal well vascularised connective tissue between the tunica adventitia of the CCA and the device. Scale bar: 200 μm.

Around the midpoint of the implants (section D) on one 3-month and one 6-month explant there were foci of scattered haemosiderophages in the tunica adventitia and adjacent connective tissue. This was interpreted to be a normal response to previous haemorrhage caused by the surgical implantation. Some moderate diffuse congestion of the outer layers of the tunica adventitia was also observed and is understood to have occurred at collection.

4. Discussion

This study demonstrated the safety of the Guardian in a sheep carotid artery study extending to 8 months post-implant. Devices were straightforward to implant on the CCA and stable in position, with minimal movement in an animal model with a longer neck than humans.

The gross pathology revealed benign fibrous encapsulation of the Guardian. Histopathology confirmed that the sections of the CCA vessels that were covered by the Guardian were normal with no evidence of disruption, distortion, or degeneration of the vessel walls and no thrombosis, stenosis or necrosis. Guardian devices were implanted with a range of ratio of device i.d. to CCA o.d. from 63% to 92%, based on prior mathematical modelling for achievement of pulse pressure reduction, distal to the device, without reduction in cerebral blood flow. There was no apparent difference in the gross pathology or histopathology of the CCAs for the Guardian sizes over this range.

The i.d. in the sheep compares well with human CCA end diastolic i.d. of 5.59 mm ± 0.69 (n = 43); for a normotensive population with age range from 29 to 76 years old, where 65% of the group was male [13]. This human cohort had end diastolic wall thickness (defined as combined intimal-medial wall thickness) of 0.32 mm ± 0.09 mm s d., in a group that included subjects with atherosclerosis. The wall thickness values of the sheep in this study and the human cohort in Roman et al. [13], are not directly comparable because the measurements in the sheep included adventitia, which may explain the greater thickness values obtained, whereas those in the human study comprised the combined intima and media layers only.

The minimally invasive surgery for device removal conducted on two devices indicated the potential for explant using this approach in humans (noting that explant is not planned routinely as the device will be designed to remain in situ unless unexpected complications occur). Once the end of the device was isolated, it could readily be withdrawn without damage to the CCA or adjacent tissues, aided by grasping the end of the device and twisting to flatten the helix that wrapped around the CCA. The surgeons conducting this procedure highlighted the value of accurately imaging the position of the device before surgical retrieval. In this study, B-mode ultrasound was used, but X-ray with a skin marker may be more accurate. Further studies would be required to confirm this approach.

Whilst the sheep is a good model for preclinical safety studies and for demonstration and refinement of surgical procedures for implant and retrieval of the Guardian, the ovine model is less applicable for evaluation of the efficacy of the device. A limitation is that the sheep CCA is associated with significant damping of pulse pressure [12] and stiffening of the arteries with age, as occurs in humans, has not been reported in sheep. There are anatomical differences in the cerebral vasculature of the brain which would further confound efficacy assessment of damping of pressure pulse in the vessels of the brain by the Guardian. In humans, the CCA bifurcates in the neck; the internal carotid arteries enter the skull, whereas the external carotid arteries provide blood supply to the face. The internal carotid artery branches into the anterior cerebral artery and continues to form the middle cerebral artery. In sheep, blood supply to the anterior portion of the cerebral vasculature arises from the CCA via the external carotid artery and internal maxillary artery. Blood flows from the external carotid artery via branches of the maxillary artery to the rete mirabile [14]. From the rete mirabile blood flows via the internal carotid artery to the Circle of Willis. The rete mirabile is a complex network of arterioles and venules running in close proximity; it uses counter-current flow to exchange heat, ions or gases. In the sheep, the rete mirabile protects the brain against heat stress. It is likely to also act as pressure and flow-damping structure, as described for the goat [15] and the dolphin [16]. Furthermore, the sheep is not as dependent on carotid artery blood flow as humans, due to adaptations in blood supply via the vertebral artery that can occur in the case of bilateral occlusion and/or ligation of the common carotid arteries [17].

5. Conclusions

This study confirms the feasibility of surgical implantation and biomaterial safety of the Guardian over an implant period of up to 8 months in sheep and demonstrates the potential for surgical explant via a minimally invasive approach. Further work is required to demonstrate efficacy in vitro and/or in vivo before evaluation in humans, including development of minimally invasive surgical implant procedures for humans.

Author contribution statement

Natalie L James: Performed the experiments; Analyzed and interpreted the data; Wrote the paper.

Zoran Milijasevic; Anthony Ujhazy; Kieri Jermyn: Performed the experiments.

David Huber; Randi Rotne: Performed the experiments; Contributed reagents, materials, analysis tools or data.

Glenn Edwards: Performed the experiments; Wrote the paper.

David S Celermajer: Conceived and designed the experiments; Wrote the paper.

Funding statement

This work was supported by The Brain Protection Company.

Data availability statement

Data included in article/supplementary material/referenced in article.

Declaration of interests statement

The authors declare the following conflict of interests: Any potential conflicts of interest, including related consultancies, patent applications, shareholdings and funding grants.

Natalie L James, Zoran Milijasevic, Anthony Ujhazy, David Huber are consultants to The Brain Protection Company (BPCo).

David S Celermajer is Chief Medical Officer, Director and a shareholder of BPCo.

Zoran Milijasevic, Anthony Ujhazy and David S Celermajer are named on patent applications/registrations for the Guardian device.

Randi Rotne, Glenn Edwards and Kieri Jermyn were funded to conduct surgery via a contract between Charles Sturt University and BPCo.

This research was funded by BPCo.