STELLATE ganglion blockade (SGB) is an accepted method to diagnose and treat patients with complex regional pain syndromes (CRPS) and other sympathetically mediated pain states. 1At our institution, we use the anterior paratracheal technique described by Carron and Litwiller 2commonly. By establishing contact with the transverse process of C6 (the Chassaignac turbercle), the risk of damage to the pleura and the vascular structures is reduced. Procedural complications can occur and include recurrent laryngeal nerve block, hematoma, brachial plexus motor blockade, phrenic nerve block, pneumothorax, vertebral artery injection, and subarachnoid and epidural injection and osteitis. 1 

We report a case of a young woman with CRPS type 1 (or reflex sympathetic dystrophy) who developed neurologic symptoms of an incomplete spinal anesthetic 1 h after SGB.

A 34-yr-old woman with CRPS type I of her right upper extremity presented for elective SGB. She had been treated since 1996 for CRPS, which followed an industrial injury in 1993. At the time of her initial examination, she had limited function of her right upper extremity, and her treatment during a 2-yr period of time included tricyclic antidepressants, intravenous lidocaine, and opiates, but little functional improvement was made. She did, however, obtain significant subjective pain relief with decreased swelling of her hand and arm for 3–5 months from intermittent SGBs using 0.25–0.5% bupivacaine (12–15 ml). Her medical history was unremarkable, and her medications included amitriptyline, hydrocodone–acetaminophen (5/500), and naproxen. Her weight and height were 89 kg and 86.5 cm, respectively. Preoperative blood pressure was 115/72 mmHg, and heart rate was 75 beats/min. Physical examination was notable for an anxious appearance, her right arm held close to her chest, and her elbow flexed 90°. Allodynia to light touch was present from her fingers to her elbow in a glove distribution.

After informed consent, a peripheral intravenous catheter was inserted in her left hand. Blood pressure, oxygen saturation (Spo2), electrocardiography, and temperatures of the dorsum of each hand were monitored, and verbal feedback was maintained throughout the procedure. Sedation with 50 μg intravenous fentanyl and 2 mg intravenous midazolam was provided.

After sterile preparation of her neck with povidone–iodine (swab sticks), the transverse process of C6 was identified easily by palpation, using the cricoid cartilage as a landmark. A 22-gauge A-beveled needle was inserted using the anterior paratracheal approach until contact with the periosteum was made. The needle was then withdrawn 1–2 mm, and, after negative aspiration, 3 ml bupivacaine, 0.25%, was injected. After 2 min and no hemodynamic or neurologic changes, needle contact with the periosteum was reestablished. After again withdrawing 1–2 mm and after negative aspiration, 5 ml bupivacaine, 0.25%, was injected. This maneuver was repeated once more until 13 ml bupivacaine, 0.25%, was injected. The patient was placed in a sitting position, and the onset of right-sided Horner syndrome, a decrease in pain in the right hand and arm, and a 2°C increase in the right hand temperature from a baseline of 35°C were noted within approximately 10 min after completion of the procedure. During the next 45 min, blood pressure remained stable (systolic blood pressure, 120–137 mmHg; diastolic blood pressure, 60–75 mmHg); however, her pulse rate slowly increased from 65 to 102 beats/min. She noted no subjective changes. In anticipation for discharge from the hospital, the patient walked to the restroom and changed into her street clothes. When walking back to check out of the postanesthesia care unit (65 min after SGB), she reported dizziness and weakness in her right leg. She was placed on a gurney, and standard monitors were applied. Her initial blood pressure was 140/80 mmHg, and her pulse was 85 beats/min. During the next 3 min, she became progressively more confused, unresponsive to verbal commands, and diaphoretic. An episode of hypotension with a mean arterial pressure of 55 mmHg occurred, which was treated with 500 ml NaCl and intravenous ephedrine (10 mg). Her mental status cleared over the next several minutes. During this period, serial physical examinations revealed a progressive loss of strength and sensation in both lower extremities, the trunk, and the right upper extremity. The lower extremities were both flaccid, with loss of deep tendon reflexes. Her left upper extremity strength, sensation, and reflexes were normal. She reported mild dyspnea and being unable to move anything other than her left upper extremity, of which she had complete control. Her dystrophic right upper extremity was ranged gently, without evidence of discomfort. Spontaneous ventilation was maintained with adequate but shallow tidal volumes. A chest radiograph was obtained to evaluate for possible pneumothorax, revealing hypoventilated lungs but no other acute changes. Neurosurgical evaluation was obtained and agreed with our preliminary diagnosis of inadvertent centriaxial blockade from the SGB. A computed tomography scan of the head and a cervical spine magnetic resonance image were obtained to evaluate for possible intracranial process, vertebral artery vasospasm or injury, or cervical epidural hematoma. These images were evaluated as normal.

The patient was closely monitored for 3 h, with no changes in the level of the presumed block and with stable vital signs. She was admitted to an intermediate care hospital bed. Approximately 10 h after the SGB, she noted a gradual return of sensation and motor function in her extremities. The next morning, or 16 h after the SGB, her physical examination results were at baseline, and she was discharged home. She noted significant reduction in her CRPS for the next 4 months.

Neurologic changes beginning 1 h after procedure have not been reported previously. After an extensive workup, including diagnostic imaging studies, the possibility of a conversion reaction was considered. This diagnosis seemed unlikely because the patient allowed her affected arm to be moved throughout a normal range of motion, without any evidenced pain behavior. Further, she had no spontaneous movement of her lower extremities, absent deep tendon reflexes, and no response to noxious stimuli.

Few references have been published regarding conversion reactions in chronic pain patients. One case report involves a conversion reaction initially diagnosed as trigeminal neuralgia. 3The diagnosis of conversion reaction was determined after a response to edrophonium and after taking a global view of the total clinical course. Because our patient never had these responses to stellate blocks in the past and her neurologic symptoms resolved without any sequelae, the diagnosis of conversion reaction seems even more unlikely.

We believe this patient sustained a subdural injection. Mechanisms proposed by Bruyns et al.  4include the following: (1) improper needle placement directed intrathecally through the intervertebral foramen, (2) a dural cuff that may accompany a nerve root distal to the intervertebral foramen, and (3) a perineural injection that may diffuse back into the subarachnoid space. This last mechanism could account for a long onset time, would involve a depot of local anesthetic, and would necessitate a large dose to be clinically important. Wulf et al.  5have reported that 3 of 10 patients who were administered a stellate ganglion block with 10 ml bupivacaine, 0.5%, had maximum concentrations exceeding 2 μg/ml, which is more than the reported toxic concentration of 1.5 μg/ml. The dosage used in this case was 13 ml bupivacaine, 0.25%, which is well below a dosage that would normally cause toxic blood concentrations.

Subdural block may account for the time course and unusual symptoms consistent with a high central neuraxial block. 6–8The subdural space is a potential space between the dura and arachnoid membranes that contains a small amount of serous fluid and has been confirmed with clinical presentation via  computed tomography and magnetic resonance imaging. 8,9Dorsal and lateral spread is seem more commonly with pathologic examination using intravenous contrast. Subsequently, more sensory and sympathetic block is noted than motor blockade. The onset of subdural blockade is 5–30 min after injection. Hypotension may be associated with subdural blockade and may be profound but is usually easier to treat than subarachnoid blockade.

The incidence of these late neurologic complications after SGB is unknown. It is unclear what signs or symptoms are predictive of possible neurologic complications. There has been a previous description of an inadvertent subdural stellate block with symptoms occurring suddenly, within 5 min of the procedure. 4Another case report involving subdural anesthesia after an interscalene block had a similar presentation but started 20 min after completion of injection. 10In our case, the symptoms occurred 45 min after the conclusion of the procedure.

Currently, our postrecovery period for SGB is 1 h. Standard monitoring, such as continuous electrocardiography, pulse oximetry, and intermittent noninvasive blood pressure monitoring, is essential. Nevertheless, no predictors were present for the profound sequelae this patient experienced. Patients who are administered stellate ganglion blocks are not discharged earlier than 1 h secondary to this incident. Each institution should make its own determination as to whether the added costs associated with extra postoperative recovery time is justified by the low incidence of this occurrence.

It is not known whether repeat stellate blockade can predispose a patient to the risk of subdural injections. Further, for difficult patients with redundant subcutaneous tissue, should more specific techniques, such as real-time magnetic resonance imaging, be used to track the accuracy of placement and spread of local anesthetics? 11–15 

Breivik H, Cousins M, Löfström J: Sympathetic neural blockade of upper and lower extremity, Neural Blockade in Clinical Anesthesia and Management of Pain, 3rd edition. Edited by Cousins MJ, Bridenbaugh PO. Philadelphia, JB Lippincott, 1998, pp 411–47
Carron H, Litwiller R: Stellate ganglion block. Anesth Analg 1975; 54: 567–70
Yokono A, Yokono S, Ogli K, Yamanobe K, Knodo A: A case of hysterical conversion manifested by pain in face and head. Masui 1991; 40: 306–12
Bruyns T, Devulder J, Vermeulen H, De Colvenaer L, Rolly G: Possible inadvertent subdural block following attempted stellate ganglion blockade. Anaesthesia 1991; 46: 747–9
Wulf H, Maier C, Schele HA, Wabbel W: Plasma concentration of bupivacaine after stellate ganglion blockade. Anesth Analg 1991; 71: 546–8
Reynolds F, Speedy H: The subdural space: The third place to go astray. Anaesthesia 1990; 45: 20–3
Asato F, Nakatani K, Matayoshi Y, Katekawa Y, Chinen K: Development of a subdural motor blockade. Anaesthesia 1993; 48: 46–9
Gershon R: Surgical anaesthesia for Caesarean section with a subdural catheter. Can J Anaesth 1996; 43: 1068–71
McMenemin I, Sissons G, Brownridge P: Accidental subdural catheterization: Radiological evidence of a possible mechanism for spinal cord damage. Br J Anaesth 1992; 69: 417–9
Tetzlaff J, Yoon H, Dilger J, Brems J: Subdural anesthesia as a complication of an interscalene brachial plexus block. Reg Anesth 1994; 19: 357–9
Ralph C, Williams M: Subdural or epidural? Confirmation with magnetic resonance imaging. Anaesthesia 1996; 51: 175–7
Hogan QH, Erickson SJ, Haddox JD, Abram SE: The spread of solutions during stellate ganglion block. Reg Anesth 1992; 17: 78–83
Slappendel R, Thijssen, HO, Crul, BJ, Merx JL: The stellate ganglion in magnetic resonance imaging: A quantification of the anatomic variability. A nesthesiology 1995; 83: 424–6
Wallace MS, Milholland AV: Contralateral spread of local anesthetic with stellate ganglion block. Reg Anesth 1993; 18: 55–9
Hogan QH, Erickson SJ: MR imaging of the stellate ganglion: Normal appearance. Am J Roentgenol 1992; 158: 655–9