Description of a novel ultrasound guided peribulbar block in horses: a cadaveric study

In equine practice, the mortality rate for healthy horses undergoing general anaesthesia remains approximately 0.9% [1]. The risks of general anaesthesia can be avoided by performing surgery in the standing horse with combinations of intravenous sedatives, analgesics [2, 3] and local anaesthetic techniques. For standing enucleations, a retrobulbar nerve block (RB) is routinely used to provide analgesia and akinesia, the loss of voluntary eye movement [4]. The use of a “blind” technique to perform a RB for procedures other than enucleation is controversial due to the potential risk of iatrogenic damage to the globe, retrobulbar haemorrhage and nerve penetration.

Motor and sensory innervation of the equine eye and eyelids is supplied by the auriculopalpebral, oculomotor, abducens, trochlear and ophthalmic nerves (or cranial nerves VII, III, VI, IV, V), which all (with the exception of the auriculopalpebral nerve) pass through the orbital fissure at the back of the orbit [5, 6]. The maxillary nerve (which supplies innervation to the lower eyelid) is in close proximity to the orbital fissure in the horse as it exits the alar canal [5]; therefore administration of local anaesthetic to the level of the orbital fissure results in widespread desensitisation of the eye, eyelids and periocular area.

Peribulbar nerve block (PB) has been described in dog cadavers [7, 8, 9], live dogs [10, 11] and cats [12] but not in horses. With this technique, the local anaesthetic is deposited outside of the extraocular muscle cone, avoiding key structures such as the optic nerve and ophthalmic blood vessels. Due to the peribulbar compartment being a potential space between the extraocular muscle cone and surrounding temporalis muscle, an injection here allows injectate to track between fascial planes to the orbital foramen, as shown in canine cadavers [9]. This approach may have advantages over the RB technique due to a theoretical lower complication rate [13, 14, 15], and as a consequence a wider application for use.

The use of ultrasonography to guide local anaesthetic techniques is becoming established in small animal anaesthesia [16], and several ultrasound guided (USG) techniques are now described in horses for various procedures [17, 18, 19]. Human evidence shows improved blockade and reduced complication rates such as vascular puncture when an USG technique is used [20]. An USG RB nerve block has been described in equine cadavers [21] but has yet to be evaluated clinically.

The primary aim of this study was to describe a novel method for USG PB in equine head cadavers, using computed tomography (CT) imaging to confirm correct location of contrast medium (Phase 1). The secondary aims were to compare success rates of both USG PB and “blind” RB with two naïve operators, and to assess spread of contrast medium for each technique (Phase 2). We hypothesise that the USG technique will have a higher injection site success rate than the RB technique.

The study was performed on equine cadaver heads of horses euthanised for reasons unrelated to this study. The European Union directive and the Royal Decree on the protection of animals for scientific purposes and teaching (2010/63/UE, RD 53/2013) were adhered to.

Phase 1 - nerve block description

Two thawed equine cadaver heads were used for description of the technique (one Thoroughbred, one Warmblood). Hair was removed bilaterally around the area of the supraorbital fossa and the area cleaned with alcohol before applying ultrasound (US) gel. The globe was injected with US gel to restore globe size, increase intraocular pressure and optimise sonoanatomy of the cadaver heads. Similar techniques using saline have been described elsewhere [21, 22]. A 5-1 MHz Phased US probe was placed over the supraorbital fossa perpendicular to the orbital rim (Fig. 1). A 19-gauge 3.5-inch spinal needle was primed with saline and a 2.5 mL syringe of saline attached.

Fig. 1
Placement and orientation of the US probe on the supraorbital fossa.

• Click for larger image
• Download as PowerPoint slide

The needle was inserted alongside the US probe in the supraorbital fossa using an in-plane technique, approximately 2 cm behind the probe in a caudal-cranial direction (Fig. 2). It was then advanced until the tip was adjacent to but not inside the extraocular muscle cone. At this point 1 mL of saline was injected to confirm correct needle placement, which was defined as ventral movement of the border of the extraocular muscle cone upon injection. Once this was confirmed, 10 mL of 50% saline 50% iodinated contrast (Iohexol [Omnipaque 300 mg·I/mL], GE Healthcare, Spain) mixture was injected. The opposite eye was scanned in the same manner; however, this time the needle was purposely advanced inside the extraocular muscle cone and 10 mL of the saline-contrast mixture injected. Location and spread of injectate were confirmed using CT imaging with both techniques.

Fig. 2
Illustration of the lateral view of the equine orbit, with needle adjacent to the myofascial extraocular muscle cone.
A, US probe; B, orbital rim; C, myofascial extraocular muscle cone; D, retrobulbar space/intraconal compartment; E, globe; F, needle; G, orbital fissure.

• Click for larger image
• Download as PowerPoint slide

Phase 1: orientation of the US probe as depicted in Fig. 1 allowed imaging of the retrobulbar region. The US image showed the orbital rim with acoustic shadowing cranially, and the myofascial extraocular muscle cone visible as a well-defined cone shape caudal to this shadowing (Fig. 3, Supplementary Video 1). Injection of contrast-saline mixture just outside of the myofascial extraocular muscle cone resulted in contrast spread within the peribulbar space, seen on CT imaging. When the needle was purposely advanced into the extraocular muscle cone under US guidance, CT images confirmed contrast within the retrobulbar space.

Fig. 3
US image achieved with placement of the US probe as depicted in Fig. 1.
Yellow arrow heads, border of the myofascial extraocular muscle cone; Cr, cranial; Cd, caudal.

• Click for larger image
• Download as PowerPoint slide

Phase 2: within this phase, a total of twenty injections were performed, ten by each naïve operator, and ten of each technique. In two cases (2/10 USG PB), the operators described difficulty in identifying the ultrasound landmarks. Despite this, 10/10 USG PB nerve blocks were successfully injected in their intended location compared to none in the RB group (p < 0.001). Two injections of the twenty resulted in injection of contrast into the masseter muscle (Fig. 4). Both were in the RB group; however, this difference between techniques was not statistically different (p = 0.473). No significant difference was found between operators as regards injection site success rate (OpA 5/10 vs. OpS 5/10, p > 0.99) or number of IM injections (OpA 0/10 vs. OpS 2/10, p = 0.473). Contrast spread to the orbital fissure (Fig. 5) was seen with six injections (6/18, discounting 2 IM injections): 4 in the RB group, 2 in USG PB group, all within the peribulbar space. However, this difference was not significantly different between groups (p = 0.629). No contrast was seen at the other deeper locations monitored (rostral alar foramen, oval foramen, mandibular nerve and skull). No evidence of globe penetration or intraneural injection was seen with any technique.

Fig. 4
Computed tomography image showing IM contrast, following unsuccessful retrobulbar injection.

• Click for larger image
• Download as PowerPoint slide

Fig. 5
Computed tomography image showing a transverse view of both orbits and bilateral presence of contrast in the peribulbar space. Contrast can be seen at the level of the R orbital fissure (green arrowhead).

• Click for larger image
• Download as PowerPoint slide

This is the first study where USG PB has been described and demonstrated successfully in horses. The approach described in Phase 1 is an easy way to perform the USG PB with excellent injection site success demonstrated in Phase 2 in the hands of naïve operators. In contrast, none of the “blind” RB injections were in the retrobulbar space, with the majority (8/10) in fact in the peribulbar space. This highlights the advantages of ultrasound guidance, to enable needle visualisation and injection of local anaesthetic in the intended location with confidence. It also concurrently allows identification and avoidance of key structures, therefore reducing the risk of inadvertent intravascular injection, globe penetration, haematoma and intraneural injection.

In Phase 1 we described a novel technique to perform an USG PB in horses. Injection of local anaesthetic in the peribulbar space can result in blockade of the oculomotor, abducens, trochlear and ophthalmic nerves as they pass through the orbital fissure [5, 23, 24], desensitising the eye and periocular region. Furthermore, local anaesthetic has been shown to migrate into the intraconal space from the peribulbar space in other species [12, 25], resulting in a complete blockade of the eye and its adnexa.

A similar USG PB technique in dog cadavers reported maxillary nerve staining [8] and a clinical study in cats showed the PB providing corneal and periocular analgesia [12]. The use of PB providing analgesia and a central eye for cataract surgery is reported in dogs [10] and is used routinely in humans [14]. In horses it could be an alternative to systemic neuromuscular blocking agents for achieving a central eye under general anaesthesia for ocular surgery, eliminating the risks associated with their use and antagonism [26]. A clinical study assessing the area of sensory and motor blockade with a single PB injection in live horses is required for confirmation.

This study is the first describing an USG PB in horses. A technique for USG RB has been described by Morath et al. [21] but uses a different approach. An “out of plane” technique is used, with the US probe and needle approximately perpendicular to one another. This requires the US probe to be placed over the eyelid to scan transbulbar, with the needle inserted via the supraorbital fossa. Difficulty in visualising the optic nerve and needle was described in some cases, with one instance (1/40) of optic nerve sheath puncture. The median distance of needle tip to the optic nerve was 3 mm (range 1–7) [21]. This highlights that even with USG, serious complications are still possible with the RB technique, especially with any unexpected movement in a sedated horse.

Compared to the USG RB technique [21], the USG PB technique described here has several advantages. First, an “in plane” technique improves needle visualisation. Second, the US probe and needle are placed at the supraorbital fossa, which would be out of the horse's vision in the live animal. Finally, the target for the PB is distant from vital structures inside the extraocular muscle cone such as the ophthalmic artery and branches, and the optic nerve.

In Phase 2, operators with similar levels of experience with both the RB technique and the use of US, but naïve to the USG PB, were used. This was to enable direct comparison of technique success, and to avoid experience level being a confounding factor. The results strongly support the USG PB as being easy to learn and perform correctly, after only one demonstration and with little prior experience. The large difference seen in injection site success between techniques was surprising, certainly the fact that none of the RB injections were in the retrobulbar space. The majority were actually in the peribulbar space. More experienced operators may have increased RB success rate, and the use of cadavers as opposed to live animals could also have made RB more challenging.

Clinically, although these injections were not successfully injected at the target site, they may still have had good effect due to diffusion of local anaesthetic into intraconal and extraconal compartments. This, however, was beyond the scope of this study. A potential surrogate measure of clinical effectiveness could be spread to the orbital fissure, although this is theoretical. In Phase 2, contrast spread to the orbital fissure was seen in 6/18 injections into the peribulbar space (4 in the RB group, 2 in the USG PB group). No statistical difference was seen between groups; however, the sample size may be too small to detect one. Spread of injectate to the orbital fissure may be affected by many factors, including injection location (rostral vs. caudal in the peribulbar space), volume used and viscosity of the injectate. The RB group would have had a more caudal injection location than the USG PB group, which may explain the trend of orbital fissure spread seen in our results.

A standard volume of injection (10 mL) was used for both techniques to enable direct comparison between techniques, contrast location and spread. Due to a lack of research into PB in horses, the ideal injection volume to achieve adequate spread is not known. Studies in small animals have used a wide range of volumes from 0.5 mL [8], 3 mL [12], 0.3 mL/kg [11], with one study using an allometric equation to calculate injection volume in dogs (2.33 × BW^0.33) [7]. In humans, a volume 2–4 times that used for RB anaesthesia is suggested for PB [14]. It is possible that had a larger injection volume been used, more frequent spread to the orbital fissure may have been seen.

In horses an injection volume of 10 mL is suggested for the RB [6]; however, in this study 8/10 RB were in fact PB. In the clinical setting this may have resulted in incomplete or failed nerve blocks due to inadequate volume to achieve appropriate spread. One disadvantage to increasing injection volume is the association with a short-lived increase in intraocular pressure seen in small animals [7, 11, 12]. Further investigation into the optimal injection volume for PB to achieve adequate spread in horses is required.

This study has several limitations. First, it is a cadaveric study and results cannot be directly extrapolated to live animals. However, the frozen thawed cadaver model used has been described previously in multiple papers in this area [8, 9, 17]. For instance, no evidence of movement of contrast from the peribulbar space to the retrobulbar space was seen in accordance with other studies in canine cadavers [7, 8]. However, evidence in live subjects shows free movement of injectate between the extraconal and intraconal space [12, 25]. Second, the same volume of injection was used for both techniques, despite evidence in other species suggesting a higher volume being required for PB. Therefore, the effect of a higher volume on contrast spread is not known. Finally, a mixture of 1:1 radiopaque contrast medium (Iohexol [Omnipaque 300 mg·I/mL], GE Healthcare) and saline was used for injection. This dilution of contrast agent was done to reduce viscosity of the injectate; however, the spread of local anaesthetic in a live animal may differ to that of the contrast–saline mixture used.

In conclusion, the USG PB described here was easy to perform, with no apparent complications aside from the difficulty in optimising ultrasound images in two cases. The USG PB could be a superior alternative to RB in horses due to the potential for fewer complications giving a wider scope for use – for example, not only for enucleations, but also facial, eyelid and corneal surgeries, and ocular analgesia. Further studies are required to evaluate optimal injection volume and clinical effect in live horses.

Supplementary Video 1

Ultrasound video highlighting the key ultrasound anatomy seen when performing this nerve block.