Spinal Anesthesia as a Complication of Brachial Plexus Block Using the Posterior Approach

For surgical procedures of the upper extremities, the brachial plexus block is a suitable technique and offers several advantages for the patient, surgeon, and anesthesiologist (1). The brachial plexus block can be performed at several sites, but the most frequently used are the axillary, interscalene, infraclavicular, and supraclavicular approaches (2). Pippa et al. (3) described an alternative, posterior approach to the plexus with a similar field of analgesia. In this article, we report a patient who developed a life-threatening paralysis without loss of consciousness during a brachial plexus block with the posterior approach. Case Report A 52-yr-old man (weight, 90 kg; height, 180 cm) was scheduled for acromioplasty of the left shoulder. His medical history was remarkable for gastric ulcerations, herniated L2-3 disc, and previous arthroscopy of the same shoulder under a brachial plexus block with the posterior approach. The patient requested the same anesthetic procedure. Before brachial plexus blockade, a 16-gauge IV cannula was inserted. Routine monitoring included electrocardiogram, pulse oxymetry, and automated blood pressure measurement. The patient was placed in a sitting position with the head bent forward. A hollow Teflon-coated needle (100-mm, 21-gauge, short-bevel needle, Stimuplex® A; B. Braun, Melsungen, Germany) was introduced 3 cm left of the midline at the intervertebral space C6-7 and was advanced perpendicular to the skin in a sagittal plane. The introduction was smooth, without contacting lamina C6 or the transverse process of C7. Contractions of the triceps muscles were noted on the first attempt, and the needle was immobilized at a depth of approximately 6 cm. After negative aspiration, bupivacaine 0.37% (without epinephrine) was slowly injected. The aspiration test was repeated after every 5 mL. Injection was painless, and the muscle contractions, which could be evoked at a current of 0.44 mA, disappeared after the first bolus of local anesthetic. A few minutes after the start of injection (18 mL bupivacaine), a slight acceleration in heart rate (from 75 to 90 bpm) was noted, and the patient reported feeling unwell. Injection was stopped immediately. His upper body drifted slowly backward, and he was unable to vocalize. The needle was removed, and the patient was placed in the supine position. Within a few seconds, total flaccid paralysis of all extremities and apnea were noted. The blood pressure decreased to 80/40 mm Hg; the patient was still able to move his eyes on command. He was told that ventilation would be assisted. In the following minutes he was ventilated with 100% oxygen by use of a face mask, and general anesthesia was induced with etomidate (0.25 mg/kg), succinylcholine (1 mg/kg), and fentanyl (2 μg/kg) IV. Anesthesia was maintained with isoflurane (0.7%–1.5% end-tidal concentration) in an oxygen/nitrous oxide mixture (fraction of inspired oxygen, 0.3). Acromioplasty was initiated under general anesthesia and was performed without complications. During the 2-h operation, ephedrine (5 mg) was administered three times because of hypotension and bradycardia. After the operation, the patient awoke and was able to breathe spontaneously. There was full motor control and normal sensibility of the left upper extremity. Further neurologic examination showed a left-sided miosis and ptosis. The patient reported that, until oxygenation through a face mask was introduced, he was able to understand the verbal commands, but was unable to move his extremities, talk, or breathe. The patient was told that paralysis had occurred during brachial plexus blockade, and questions and fears were discussed. Professional psychological support was offered, but the patient considered this unnecessary. Discussion Our patient developed a total paralysis and Horner’s syndrome during brachial plexus block with bupivacaine with the posterior approach. The technique of the posterior approach was first described by Pippa et al. in 1990 (3). It was brought forward as a safe alternative to the more frequently used interscalene techniques (3), with possibly a slightly different distribution of analgesia (4). In the original article by Pippa et al. (3), a needle with an air-filled syringe was inserted at the intersection of the external superior edge of the trapezius muscle with the horizontal intervertebral line C6-7 (usually 3 cm lateral to the midline) and was advanced perpendicular to all planes until resistance to injection was lost (4). Instead of loss of resistance, we made use of a neurostimulator, requiring twitches in the arm or shoulder at <0.5 mA, and determined the site of injection 3 cm lateral to midpoint of spines C6 and C7 (Fig. 1).Figure 1: Landmarks and site of injection for brachial plexus block with the posterior approach (see text for details).In theory, this posterior technique has two advantages. First, it is easy to teach, because the anatomical landmarks (the cervical spinal processes) are easily identifiable and the direction of the needle is in square angles to anatomical planes. Second, it seems safer because most delicate structures in the neck either are hidden by transverse laminae (e.g., the vertebral artery) or lie anterior to the brachial plexus (e.g., the carotid artery). After interscalene blocks, various complications and side effects have been reported (5,6), including total spinal anesthesia and Horner’s syndrome. Inadvertent injection of local anesthetic into the vertebral artery causes a transient locked-in-like syndrome after a brachial block by a lateral approach (7). In our case, the possibility of intrathecal injection of anesthetic should be considered. Local anesthetics can enter the spinal space through at least three different routes. First, the drug may be injected directly intrathecally. Especially with use of a flexible needle, it must be kept in mind that the position of the tip can deviate from its original direction. Second, a dural cuff may accompany a nerve root some distance distal to the intervertebral foramen through which it passes and then may accidentally be punctured, making direct intrathecal injection possible (8). All authors who reported spinal anesthesia complicating paravertebral techniques claimed to have had negative aspiration tests, which therefore did not guarantee absolute safety. Finally, local anesthetics injected intraneurally could spread in a central direction to the spinal space. This should be suspected if there is a marked resistance to injection accompanied by pain (8). The sudden onset and rapid progression of the signs and symptoms, the absence of seizures or pain on injection, and the complete recovery of symptoms after two hours favor the assumption that the block in our patient was complicated by a spinal spread of local anesthetic (9). The incidence of complications with this technique is unknown. Pippa et al. (3) reported no intrathecal injections in 50 brachial plexus blocks with the posterior approach, but with the interscalene approach they found a frequency of 2 in 50 (4). In our institution, this complication has occurred twice in an estimated number of 60 blocks with the posterior approach over the past two years. We may still be on the ascending limb of the learning curve, but as yet we see no advantage of this technique as being safer. Our case shows that total spinal anesthesia may complicate brachial plexus blockade during the posterior approach and that the patient might be fully awake and conscious during this life-threatening event. The perceived safety of the posterior approach may be more related to the fact that this approach is rarely performed than to the anatomical advantages of this technique.