History

An approximately 2-year-old 23.9-kg (52.6-lb) spayed female mixed-breed dog was referred to the orthopedic service of the University of California-Davis Veterinary Medical Teaching Hospital (VMTH) for evaluation of lameness in the left thoracic limb. The dog had a history of trauma about 1 year earlier and was presumed at the time to have been hit by a car. The referring veterinarian performed radiography at that time and diagnosed left elbow joint luxation.

On physical examination at the VMTH, the dog was bright, alert, and responsive and had a heart rate of 120 beats/min (reference range, 70 to 120 beats/min), respiration rate of 24 breaths/min (reference range, 18 to 34 breaths/min), and a body condition score of 4 (on a scale of 1 to 9). The dog had pink mucous membranes and a capillary refill time of < 2 seconds. Because the dog had a spay surgery scar in an early phase of healing, the dog's abdomen was not palpated. The dog had toe-touching lameness in the left thoracic limb, which had muscular atrophy (presumed from disuse) and lateral luxation and low range of motion of the elbow joint. No other abnormalities were noted on physical examination.

Results of radiographic evaluation indicated chronic elbow joint luxation in the left thoracic limb, with lateral displacement of the radius and ulna in relation to the humerus and mild osseous remodeling of the articulation between the distal aspect of the humerus and the proximal aspects of the radius and ulna. The dog was admitted to the VMTH for surgical reduction of the luxation.

On the following day, the dog had a heart rate of 102 beats/min, respiratory rate of 24 breaths/min, PCV of 41% (reference range, 40% to 55%), plasma total protein concentration of 7.4 g/dL (reference range, 6.0 to 8.0 g/dL), blood glucose concentration of 103 mg/dL (reference range, 86 to 118 mg/dL), and BUN concentration of 5 to 15 mg/dL (measured with a reagent strip^a; reference range, 5 to 21 mg/dL). The dog was assigned an American Society of Anesthesiologists physical status of 1 (on a scale from 1 [healthy] to 5 [moribund]) and was premedicated with hydromorphone (0.05 mg/kg [0.02 mg/lb], IM) and atropine (0.02 mg/kg [0.01 mg/lb], SC). Approximately 20 minutes later, a 20-gauge, 4.8-cm IV catheter was placed in the dog's right cephalic vein, and the dog was pre-oxygenated with 100% oxygen delivered by mask for approximately 5 minutes before anesthetic induction with propofol (1.92 mg/kg [0.87 mg/lb], IV, titrated to effect) and midazolam hydrochloride (0.20 mg/kg [0.09 mg/lb], IV). The dog was intubated with an 11-mm-internal-diameter endotracheal tube under direct visualization with a laryngoscope. General anesthesia was maintained with isoflurane delivered in oxygen with an out-of-circle precision vaporizer through a semiclosed circle breathing system, and lactated Ringer solution (5.0 mL/kg/h [2.3 mL/lb/h], IV) was administered during the procedure. A multiparameter monitor^b was used to monitor the dog's end-tidal concentrations of carbon dioxide and isoflurane, oxygen saturation of hemoglobin (pulse oximetry), body temperature (with an esophageal thermometer), and ECG. In addition, a Doppler ultrasonic flow detector combined with sphygmomanometry was used to measure indirect blood pressure. An ultrasonographically guided,^c subscalenic approach for a brachial plexus nerve block was performed to provide locoregional anesthesia of the left elbow joint and distal humoral region as described.^d After aspiration to rule out intravascular drug administration, bupivacaine hydrochloride 0.5% (0.30 mL/kg [0.14 mL/lb]) was injected through the stimulation needle^e used with a peripheral nerve stimulator.^f

During surgery, minimal variation was noticed in the dog's heart rate, blood pressure, or respiratory rate, indicating that the locoregional block was effective. The anesthetic depth of the patient also stayed appropriate throughout the procedure, with no substantial change noticed irrespective of the varying nociceptive input from the surgical procedure. After surgery, the dog was recovered from anesthesia (175 minutes of general anesthesia) and extubated (110 minutes after the nerve block was performed). The dog was moved to a cage in the postoperative recovery room, where staff noticed that the dog's left eye had ptosis of the upper eyelid, protrusion of the third eyelid, enophthalmos, and miosis, consistent with Horner syndrome (Figures 1 and 2). Carprofen (2.2 mg/kg [1.0 mg/lb], IV) was administered in recovery for postoperative analgesia as planned, and the dog remained hospitalized overnight.

View Full Size

Photograph of an approximately 2-year-old 23.9-kg (52.6-lb) spayed female mixed-breed dog during recovery from general anesthesia that was augmented with an ultrasonographically guided, subscalenic approach to a brachial plexus nerve block for surgical reduction of a luxated elbow joint in the left thoracic limb. The dog's left eye has ptosis of the upper eyelid, protrusion of the third eyelid, enophthalmos, and miosis, consistent with Horner syndrome.

Citation: Journal of the American Veterinary Medical Association 255, 9; 10.2460/javma.255.9.1016

• Download Figure
• Download figure as PowerPoint slide

View Full Size

A close-up of the dog in Figure 1 showing signs of Horner syndrome affecting the left eye during recovery from anesthesia.

Citation: Journal of the American Veterinary Medical Association 255, 9; 10.2460/javma.255.9.1016

• Download Figure
• Download figure as PowerPoint slide

Question

Why did the dog have Horner syndrome after the procedure?

Answer

Horner syndrome in the dog of the present report was likely a complication of the brachial plexus nerve block performed.

Discussion

Because the dog otherwise appeared normal after the procedure and did not show signs of discomfort, no interventions for Horner syndrome were performed. By the following morning, the dog had regained motor and sensory function in the left thoracic limb, indicative that the effects of the brachial plexus nerve block had worn off. In addition, the dog no longer had signs of Horner syndrome, consistent with a transient complication of the brachial plexus nerve block.

In companion animal practice, brachial plexus nerve blocks are commonly performed as an adjunct to general anesthesia and as part of a multimodal analgesic regimen to help reduce patient requirements for general anesthetic agents. In addition, by prolonging the duration of locoregional anesthesia into the postoperative period, the requirement for systemic analgesics is also reduced.¹ Indications for the ultrasonographically guided, subscalenic approach for a brachial plexus nerve block include but are not limited to thoracic limb amputation; humeral, radial, or ulnar surgeries; and elbow joint surgery, alone or in combination.²

Because the surgical treatment of the dog in the present report involved manipulation of the dog's left thoracic limb at the level of the humerus and elbow joint, an ultrasonographically guided, subscalenic approach for a brachial plexus nerve block (which involved blocking the ventral branches of spinal nerves C6 through T1 where they course between the scalenus medius and the longus colli muscles before reaching the axillary space^d) was used. Other common approaches to administer thoracic limb locoregional anesthesia include cervical paravertebral nerve block, axillary brachial plexus nerve block, and radius-ulna-median-musculocutaneous nerve block.²

Although a study^d in dog cadavers shows that dye injected at a dose of 0.3 mL/kg only successfully stained the ventral rami of spinal nerves C7 through T1 and that dye injected at a dose of 0.4 mL/kg (0.2 mL/lb) was needed to successfully stain the ventral rami of spinal nerves C6 through T1, it has been suggested that when the volume dose of local anesthetic used is high, blockade of untargeted nerves could occur.³ In the dog of the present report, 0.3 mL/kg of bupivacaine 0.5% was administered for the nerve block. This volume dose was consistent with that administered by a subscalenic approach to successfully induce blockade of brachial plexus nerves (evidenced by lack of any cardiovascular response during surgery) but not result in Horner syndrome.^g

Horner syndrome is a rare complication of brachial plexus nerve block. The author was not aware of any published report of Horner syndrome following a subscalenic approach for a brachial plexus nerve block; however, the syndrome has been reported following a cervical paravertebral approach.^4,5 Horner syndrome results from loss of sympathetic innervation to the dilator muscle of the iris, Muller muscles of the upper and lower eyelids, and extraocular muscles in the orbit. The loss results in miosis, ptosis, enophthalmos, and third eyelid protrusion.

A study⁶ shows that the brachial plexus in dogs is formed most commonly (34/58 [59%]) by the ventral rami of spinal nerves C6 through T1, but could also be comprised of the ventral rami of spinal nerves C5 through T1 (12 [21%]), C6 through T2 (10 [17%]), or C5 through T2 (2 [3%]). In addition, contributions from spinal nerves C5 through T2 in the formation of brachial plexuses are also described in dogs.⁷ Sympathetic nerves that innervate the eye are in close anatomic proximity with nerves that form the brachial plexus. Briefly, sympathetic innervation to the eye consists of first-order (central), second-order (preganglionic), and third-order (postganglionic) neurons. The first-order neurons originate in the hypothalamus and rostral midbrain and travel distally through the tectotegmental spinal tract to synapse with second-order neurons in the lateral horn of the spinal cord gray matter at the level of spinal cord segments T1 through T3. The second-order axons exit the spinal cord with spinal nerve roots T1 through T3. The sympathetic axons separate from the spinal nerve roots as the ramus communicans and form the thoracic sympathetic trunk that courses cranially within the carotid sheath in close apposition with the vagus nerve, forming the vagosympathetic trunk, from which the sympathetic axons continue rostrally through the cervicothoracic ganglion and middle cervical ganglion before synapsing in the cranial cervical ganglion adjacent to the tympanic bulla. The third-order neurons originate in the cranial cervical ganglion, pass through the middle ear, enter the cranial cavity with the glossopharyngeal nerve, and pass close to the cavernous sinus before leaving the cranial cavity through the orbital fissure.

Given the close anatomic proximity between the sympathetic nerves that innervate the eye and the nerves that form the brachial plexus, it was hypothesized that because of diffusion of the bupivacaine injected close to the first rib to block spinal nerves C8 and T1 in the dog of the present report, blockade of the cervicothoracic ganglion or the middle cervical ganglion, alone or in combination, occurred and resulted in development of Horner syndrome. Local diffusion of bupivacaine also could have caused blockade of the phrenic nerve and resulted in hemiparalysis of the diaphragm^{d, g}; however, the dog's respiratory function seemed clinically normal as evaluated by end-tidal carbon dioxide monitoring before the dog was extubated. A study¹ shows that when lidocaine hydrochloride with new methylene blue added as a marker was used to perform cervical paravertebral brachial plexus blockade by 3 different techniques, staining of the spinal cord occurred in up to 9 of 23 (39%) procedures, indicating that the anesthetic agent marked with the dye similarly reached the spinal cord and could have resulted in respiratory anomalies. This complication was less likely with the subscalenic approach used in the dog of the present report because no respiratory anomaly was detected, consistent with a previous study^d in dog cadavers that shows no epidural distribution of dye with the subscalenic approach for brachial plexus nerve blockade.

At the time the dog was treated, Horner syndrome was not considered a common complication with the subscalenic approach for brachial plexus nerve blockade. However, Horner syndrome occurred in 2 of 8 (25%) dogs following the cervical paravertebral approach for brachial plexus nerve blockade.⁵ Findings in the dog of the present report suggested that the potential also exists for Horner syndrome to occur with the subscalenic approach and that veterinarians should be aware of this potential complication. The clinical relevance of Horner syndrome in scenarios similar to that of the dog in the present report is questioned because the abnormal signs are usually transient, do not produce any systemic implications, and usually do not require treatment.^4,5