Severe Brachial Plexopathy after an Ultrasound-guided Single-injection Nerve Block for Total Shoulder Arthroplasty in a Patient with Multiple Sclerosis

A 65-yr-old right-hand-dominant man, American Society of Anesthesiologists physical status III, presented for a right total shoulder arthroplasty secondary to osteoarthritis. The patient’s medical history was significant for hypothyroidism, benign prostatic hypertrophy, mitral valve prolapse, and MS. His medications included 40 mg pravastatin by mouth daily, 75 μg levothyroxine by mouth daily, and 15 mg oxybutynin (extended release) by mouth daily. The patient was allergic only to oysters, which had caused anaphylaxis in the past. Although without clinical changes for 2 yr, his MS was remarkable for bilateral lower extremity weakness (walker needed for ambulation) and the requirement for self–urethral catheterization.

After informed consent, the patient underwent an interscalene nerve block and general anesthetic. In the preanesthetic block room, sedation was provided with 50 μg intravenous fentanyl and 2 mg intravenous midazolam. An ultrasound-guided “single-shot” injection using an in-plane needle approach and nerve stimulation was performed. The injection was made at the mid-neck level at the nerve roots of the brachial plexus. The needle direction was in reference to the middle scalene muscle from the lateral toward medial direction. Three injections were made starting laterally on C5, then anteriorly on C5, and then medially to C5. The injections were made to create circumferential spread around the roots of the brachial plexus. The injection was performed as previously described.3,4A 50-mm, 22-gauge b-bevel (B. Braun Medical, Bethlehem, PA) was inserted in plane with the ultrasound beam during visualization of the roots of the brachial plexus on short axis. The needle (stimulating at 0.45 mA, 0.1-ms pulse duration, 2 Hz) was directed until it approached the outer edge of the C5 nerve root. The needle was not seen to penetrate the epineurium by our ultrasound image (fig. 1). After the demonstration of biceps contraction, an injection of 30 ml bupivacaine (0.5%), 1:400,000 epinephrine, and 50 μg clonidine was injected using a 10-ml Luer-Lok controlled stroke syringe. The local anesthetic was noted to surround the C5–C6 nerve roots. The needle was repositioned three times to generate complete coverage of the C5–C6–C7 roots. During the procedure, the patient experienced no discomfort, and there was not resistance to injection. The block was checked for success by the senior regional resident. This patient was noted to have partial sensory (to ice) blockade over the anterior shoulder (axillary nerve distribution C5) and partial motor (by strength testing) and sensory (to ice) blockade of the musculocutaneous nerve distribution 10 min after regional blockade.

After the induction and maintenance of general endotracheal anesthesia, the patient was placed in the beach chair position. Consistent with the sitting position, an episode of hypotension (77/46 mmHg, mean arterial pressure 56 mmHg) was noted after induction and patient repositioning to the sitting position. The patient initially required a total administration of 2 l incrementally of lactated Ringer’s solution and a total of 15 mg ephedrine (in 5-mg dosing increments) to return to a mean arterial pressure greater than 70 mmHg. Intraoperatively, the patient’s temperature ranged from approximately 35 to 36.4 centigrade. The arm was held in place by the Spider Arm Retractor (Tenet Medical Engineering, Calgary, Alberta, Canada). A Zimmer anatomic total shoulder system was used (Zimmer Inc., Warsaw, IN). During placement of the glenoid component, the arm was positioned in 35° of external rotation and 45° of abduction. The estimated blood loss was 400 ml, and the patient received 2,800 ml lactated Ringer’s solution. Surgical time was 3 h 45 min. After emergence in the postanesthesia recovery unit, the patient was noted by nursing staff to have a dense motor and sensory block and was also noted to be comfortable for the first hour. The patient then began to report right arm pain that was described as burning in quality. It was rated as 5 out of 10 on a visual analog scale. A neurologic examination was performed by the operating orthopedic resident within 4 h postoperatively. At that time, the patient was again noted to have a dense motor and sensory block of the operative extremity, as would be expected 10 h after a successful regional blockade.

On postoperative day 1, the patient continued to have shoulder pain with a persistent flaccid motor block of his entire right upper extremity. This pain was exacerbated by shoulder and arm movement and not by neck movement or a Valsalva maneuver, as can be seen in cervical radiculopathy. A consultation by the neurology service on postoperative day 2 found sensation to temperature throughout dermatomes C4–T1, with absent light touch sensation in C6–T1. Vibration and joint position perception were absent throughout. A magnetic resonance image of the chest was performed on postoperative day 3, which demonstrated postsurgical changes without any evidence for compressive or avulsive pathology. However, it was diagnostic for brachial neuritis (fig. 2). High-dose methylprednisolone was initiated to treat a presumed autoimmune brachial neuritis. An electromyelogram performed on postoperative day 4 showed loss of the median and ulnar F waves. In addition, there was no voluntary recruitment of the following muscles: deltoid, triceps, biceps, brachioradialis, wrist extensors, and first dorsal interosseous. At this time, there was no evidence of active denervation in any of the muscles examined. On postoperative day 11, a complete paresis of the patient’s entire arm persisted; an electromyelogram demonstrated active denervation of all muscles and no voluntary motor recruitment. This study demonstrated low-amplitude compound muscle action potentials of the median and ulnar motor nerves. Median ulnar and radial sensory nerve action potentials were absent. Electromyographic examination revealed active denervation in all of the muscles previously examined, with no voluntary motor recruitment (table 1).

A follow-up electromyelogram 3 months from the date of surgery showed improvement. There was reduced voluntary motor recruitment with evidence of reinnervation in all of the muscles that were previously examined. The patient’s unaffected limbs were tested, and studies of the radial and sural sensory nerves and ulnar and peroneal motor nerves with F waves yielded normal results. Nerve fiber loss can still be significant despite normal nerve conduction study results. Therefore, a normal electromyelogram does not completely rule out subclinical peripheral neuropathy.

At 8 months postoperatively, the patient continued to have significant range of motion and strength deficits. His distal hand function remained limited secondary to stiffness from the prolonged neurologic recovery. Range of motion at the wrist, metacarpophalangeal joints, proximal interphalangeal joints, and distal interphalangeal joints were significantly limited, with approximately 50% loss of motion at each level. The patient also continued to have visible isolated muscle atrophy of proximal musculature, including the pectoralis major and posterior deltoid. His final diagnosis was an inflammatory brachial neuritis (IBN).

Multiple sclerosis is described as a chronic disease of the central nervous system that usually begins in young adults. Pathologically, MS is characterized by multiple areas of central nervous system white matter inflammation, demyelination, and glial scarring or sclerosis.5The clinical course of MS varies from a benign, largely symptom-free disease to a rapidly progressive and disabling disorder. The etiology of MS is likely due to autoimmune mechanisms, possibly triggered by infectious and other environmental factors in genetically susceptible individuals.6 

Controversy exists in providing regional anesthesia to patients with neurologic diseases. The “double-crush” phenomenon suggests that patients with preexisting neural compromise may be more susceptible to injury at another site when exposed to a secondary injury.7The performance of a neuraxial technique in patients with preexisting central nervous system disorders may increase the risk of a double-crush phenomenon.8In contrast to a spinal or epidural block, a peripheral nerve block in MS patients is theoretically attractive because the neural pathology is presumed to be located in the central nervous system. However, this association seems to be incomplete and is based on the fact that the clinical involvement of the peripheral nervous system in MS patients has traditionally been ignored by modern textbooks. This is despite the fact that the description of this link dates back a half century.9Importantly, this conventional teaching is also present in the anesthesia literature.1,2,10–13Careful assessment of the literature reveals that multiple recent studies have shown the existence of subclinical peripheral neuropathy in some patients with MS.14–18Pogorzelski et al.  14noted both sensory and peripheral motor nerve lesions of a demyelinating-axonal character. They also noted that sensory abnormalities were more pronounced than motor ones. Another study found electrophysiologic abnormalities in the 14.7% of all peripheral nerves examined (n = 244) in patients with MS.17This is well above the reported prevalence of 2.4% in the general population. In the elderly, the prevalence is reported to be as high as 8%, mostly due to diabetes mellitus.19Hughes et al.  20described an association of a demyelinating peripheral neuropathy in MS patients. Other inflammatory demyelinating diseases exist that have both central and peripheral components, such as chronic inflammatory demyelinating polyneuropathy.21 

Patients with underlying peripheral neurologic disorders may be more susceptible to nerve injury with the use of regional techniques.22Despite testing modalities such as electromyography and magnetic resonance imaging, it may be difficult to differentiate between multiple etiologies, including direct trauma during the regional procedure, neurotoxicity from local anesthetics (and additives), and patient positioning, such as extreme abduction and external rotation, which has been implicated in surgical stretch injury of the brachial plexus. All of these could occur in a patient undergoing total shoulder replacement. The other confounding variable in diagnosing the etiology of a postoperative neurologic deterioration is that the clinical course of MS may be exacerbated from many nontraumatic-related reasons, such as hyperthermia, electrolyte abnormalities, stress, and pain.

Brachial plexus injury after total shoulder arthroplasty has been estimated at 2.8%.23To our knowledge, this is the first report of an IBN after total shoulder replacement in a patient with MS. This is also the first report of IBN in a patient using an ultrasound-guided regional anesthesia technique. Brachial plexus injury after interscalene nerve blockade has been previously described.24IBN has also been reported to occur in patients during treatment for MS.18IBN is a well-recognized clinical syndrome characterized by brachial pain followed by patchy atrophy of muscles in the shoulder girdle and arm innervated by individual branches of the brachial plexus.25–27Postsurgical IBN has not been widely recognized since Parsonage and Turner’s original description.27 

In summary, we report a case of a severe brachial plexus injury that occurred in a patient with MS after a total shoulder replacement during combined general anesthesia and interscalene nerve block. Although the mechanisms of this injury are unclear, the potential preexisting pathology of the peripheral nervous system may have contributed. It is possible that this patient preoperatively had an occult peripheral neuropathy, and his underlying MS predisposed him to development of a peripheral autoimmune injury leading to a brachial neuritis. The individual decision to perform peripheral regional anesthesia in a patient with MS must rest on the perceived benefits of avoiding non–opioid-based analgesia and/or avoiding general anesthesia. Anesthesiologists should recognize that the peripheral nervous system may also be abnormal in patients with MS.