Prolonged Horner Syndrome Due to Neck Hematoma after Continuous Interscalene Block

A 48-yr-old woman, 162 cm tall, 85 kg in weight, was admitted for treatment of a persistent complex regional pain syndrome type I of the right hand. Except for obesity, routine physical and laboratory examinations yielded normal results. Blood coagulation tests were within the normal range. The patient was not using nonsteroidal antiinflammatory drugs or aspirin treatment. We decided to start treatment by means of continuous interscalene analgesia. After identification of the interscalene groove by palpation, a 21-gauge, short-bevel stimulating needle (Stimuplex-A; B. Braun, Melsungen, Germany) connected to a nerve stimulator (Stimuplex-DIG; B. Braun) was introduced through the skin and advanced in the direction of the interscalene groove. With the first attempt, muscle twitches of the deltoid and triceps muscle could be elicited by a current of 1.4 mA with a pulse duration of 0.1 ms. The placement of the needle was adjusted so that the muscle twitches were still present with a threshold stimulation of 0.35 mA. Then a 20-gauge, 50-mm cannula was advanced over the stimulating needle into the interscalene space, and a 22-gauge catheter with a stylet (Polyplex N30-T; Polymedic, Bondy, France) was introduced through the cannula and pushed forward 3 cm into the interscalene space (cannula-over-needle technique). After tunneling the catheter subcutaneously with the aid of an 18-gauge intravenous cannula and fixing it with dressing tapes, 4interscalene block was performed by injecting 40 ml ropivacaine, 0.6%, through the indwelling catheter. Within 15 min, the block was complete, and the patient experienced no more pain in her hand. Six hours after the initial bolus, a continuous interscalene infusion of 0.2% ropivacaine at a rate of 8–10 ml/h was started through the interscalene catheter, with excellent pain relief. On the third day after interscalene catheter placement, the patient reported blurred vision and painful swelling on the right lateral aspect of her neck. Inspection revealed a swelling around the site of the catheter insertion and concomitant Horner syndrome on the ipsilateral side, including myosis, ptosis, enophthalmia, anhydrosis on the whole ipsilateral side of the face, and conjunctival hyperemia, which was not present at the time of the initial block. The interscalene catheter was removed immediately. Subsequent investigation of the neck by means of ultrasound showed a hematoma (4 × 5 cm) expanding from behind the anterior scalene muscle toward the prevertebral longus colli muscle. No compression of the interscalene brachial plexus within the interscalene groove could be shown. Eight hours after removing the interscalene catheter, sensory block of the right arm resolved, but the Horner syndrome remained. Neurologic investigation confirmed the Horner syndrome, but electroneuromyography and the sympathetic skin response did not show neurologic damage of the right upper extremity, including the sympathetic fibers.

The pain associated with the complex regional pain syndrome type I returned and was further treated conservatively. After 3 months, the Horner syndrome was still present. Six months after its first appearance, the symptomatology of the Horner syndrome started to improve. A light residual ptosis was the only remaining symptom. After 1 yr, the Horner syndrome resolved completely.

A 20-yr-old woman, 170 cm tall, 65 kg in weight, with anterior instability of the right shoulder was admitted for Bankart repair. At the time of admission to the hospital, physical examination was unremarkable, and the results of routine laboratory investigations were normal. The patient was not to take any medication.

The surgical procedure was performed during general anesthesia in the beach chair position using propofol and fentanyl. After an uneventful surgery that lasted 2 h, the patient underwent extubation and brought to the recovery room. The presence of severe postoperative pain led to the placement of an interscalene catheter to provide continuous analgesia. The same technique described in case 1 for catheter placement was used, but two attempts were necessary. An initial bolus of 30 ml ropivacaine, 0.2%, was administered through the catheter, and 20 min later, the patient was pain-free. Analgesia was maintained with a continuous infusion of 0.2% ropivacaine at a rate of 9 ml/h. The patient left the recovery room 3 h after the initial bolus and was pain-free and comfortable until the next day. Then, the patient reported visual disturbances and a swelling on the neck. The clinical examination showed edema at the site of the catheter insertion and Horner syndrome similar to that of the first case (myosis, ptosis, enophthalmia, anhydrosis, and conjunctival hyperemia) on the same side. The catheter was withdrawn immediately, and ultrasound investigation of the neck was performed. A hematoma (3 x 4 cm) was found, expanding between the prevertebral and scalene muscles, without evident compression of the interscalene brachial plexus. Three hours after removal of the catheter, shoulder pain returned and was further treated with 0.1 mg/kg subcutaneous morphine and 2 g intravenous propacetamol. Further investigations showed no neurologic deficit in her right upper extremity. The only abnormality was the persistence of Horner syndrome.

Two days later, the ptosis improved slightly, and the patient was discharged to her home. Two weeks later, the patient showed no further amelioration. At this time, neurologic examination of the right upper extremity with the aid of electromyography, magnetic resonance imaging of the neck, and duplex sonography of the major cervical vessels were performed. The results of these investigations were all normal. After 3 months, improvement of the ptosis and myosis was evidenced, and 6 months later, only slight ptosis was present, without visual disturbances. The symptoms resolved completely after 1 yr.

Transient Horner syndrome is a well-known side effect of stellate ganglion block, interscalene block of the brachial plexus, and occasionally epidural analgesia. 5,6However, prolonged Horner syndrome occurring after interscalene block is rare and is a matter of significant concern because it may represent traumatic interruption of the sympathetic chain of the neck. Sukhani et al.  3described the only prolonged Horner syndrome after single-shot interscalene block. They did not find a direct cause of this complication, and the presence of a hematoma was not mentioned.

In the current cases, a hematoma expanding between the prevertebral longus colli muscle and the anterior scalene muscle was present and may explain the persistence of Horner syndrome. Indeed, such a hematoma may have compressed and damaged the thin preganglionic cervical sympathetic fibers, which are located in this area.

It is known that the sympathetic preganglionic fibers have the potential to regenerate after denervation. Langley 7was the first to show the regenerating potential of the cervical sympathetic trunk. In this study, the cervical sympathetic trunk of six anesthetized cats was cut below the superior cervical ganglion, and the author noted that the time needed for full regeneration of the sympathetic nerves was between 1 and 15 months. In another study, McLachlan 8resected the cervical sympathetic trunk of guinea pigs below the superior cervical ganglion and observed that 3 months were necessary after section to obtain a nearly complete reinnervation. Purves 9also demonstrated in guinea pigs the ability of sympathetic preganglionic nerve fibers to regenerate. Using the same electrophysiologic method as McLachlan, 8he showed that 6 months after cervical sympathetic trunk section, the degree of innervation was again almost normal.

According to these animal findings, we believe that a prevertebral hematoma, as described in the current cases, may have compressed and damaged preganglionic cervical sympathetic fibers, resulting in prolonged Horner syndrome. The presence of anhydrosis on the whole ipsilateral side of the face—not only limited to the forehead—is a characteristic sign of preganglionic Horner syndrome, which strongly suggests preganglionic sympathetic fiber damage. The fact that Horner syndrome resolved completely in both of our cases reflects the potential of human sympathetic fibers to regenerate spontaneously. The time needed for complete regeneration and resolution of the clinical symptomatology was approximately 1 yr, which is in accordance with the results obtained from experimental studies. The absence of disturbances of the sympathetic supply of the upper limb in both cases, as shown by electroneuromyographic investigation and sympathetic skin response, suggest the hematoma was located above the stellate ganglion and damaged the sympathetic nerves at this level, sparing those going through the stellate ganglion.

In summary, we described two cases of prolonged Horner syndrome occurring during continuous interscalene analgesia. A complete remission of the symptoms occurred spontaneously after 1 yr. In both cases, a prevertebral hematoma was diagnosed and is likely to be the cause of the preganglionic sympathetic trunk damage. In case of persistent Horner syndrome after interscalene block, examination of the neck by means of ultrasound to exclude the presence of a hematoma is recommended. It is still not known whether the placement of the catheter itself potentially increases the risk of hematoma, and prospective studies of complications associated with the interscalene catheter are welcomed.