Continuous Interscalene Brachial Plexus Blocks: An Anatomical Challenge between Scylla and Charybdis?

Abstract

Brachial plexus blocks at the interscalene level are frequently chosen by physicians and recommended by textbooks for providing regional anesthesia and analgesia to patients scheduled for shoulder surgery. Published data concerning interscalene single-injection or continuous brachial plexus blocks report good analgesic effects. The principle of interscalene catheters is to extend analgesia beyond the duration of the local anesthetic’s effect through continuous infusion, as opposed to a single injection. However, in addition to the recognized beneficial effects of interscalene blocks, whether administered as a single injection or through a catheter, there have been reports of consequences ranging from minor side effects to severe, life-threatening complications. Both can be simply explained by direct mispuncture, as well as undesired local anesthetic spread or misplaced catheters. In particular, catheters pose a high risk when advanced or placed uncontrollably, a fact confirmed by reports of fatal outcomes. Secondary catheter dislocations explain side effects or loss of effectiveness that may occur hours or days after the initial correct function has been observed. From an anatomical and physiological perspective, this appears logical: the catheter tip must be placed near the plexus in an anatomically tight and confined space. Thus, the catheter’s position may be altered with the movement of the neck or shoulder, e.g., during physiotherapy. The safe use of interscalene catheters is therefore a balance between high analgesia quality and the control of side effects and complications, much like the passage between Scylla and Charybdis. We are convinced that the anatomical basis crucial for the brachial plexus block procedure at the interscalene level is not sufficiently depicted in the common regional anesthesia literature or textbooks. We would like to provide a comprehensive anatomical survey of the lateral neck, with special attention paid to the safe placement of interscalene catheters.

1. Introduction

The use of interscalene brachial plexus blocks has been an established procedure for decades to provide intra- and postoperative analgesia for patients undergoing shoulder surgery. Although often regarded as a simple procedure [1], this technique is associated with specific side effects and complications. The underlying reasons are certainly not known to all users [2,3]. Typical undesired side effects such as phrenic nerve palsy, Horner’s syndrome, or recurrent nerve palsy are commonly observed [3]. Among them, the frequently observed ipsilateral phrenic nerve block and hemidiaphragmatic paralysis (HDP) are considered the most undesirable side effects. The use of ultrasound allows clinicians to reduce local anesthetic (LA) volumes [4,5,6], concentrations [7,8], and/or LA injections [9] outside the interscalene gap. All these strategies can reduce the incidence of HDP but not completely negate it. Consequently, numerous alternative techniques, such as C7 nerve root [10], suprascapular [11], upper-trunk [12], and costoclavicular brachial plexus block [13], have been described and extensively discussed [14,15]. They all provide acceptable postoperative analgesia but not to the same degree as interscalene blocks (ISBs), which are conceptional from an anatomical view. Furthermore, they cannot reduce the risk of HDP to zero, which is in demand for patients with severe pre-existing pulmonary disease. The side effects of a single-injection interscalene block (ISB) can be explained by uncontrolled LA spread, especially with larger LA volumes, high injection pressure, or a medially directed canula placement toward the neuroforamina. Besides relatively mild side effects, serious complications such as intrathecal [16] or intravascular injections [17] or damage to the spinal cord [18] have been reported in relation to single-injection ISB. Catheter techniques for continuous postoperative analgesia are even more challenging. Despite uneventful initial LA injection before catheter placement, life-threatening and fatal complications have been reported following intrathecal [2,19,20], epidural [21], or intravascular [22,23] catheter positioning. Symptom onset for misplaced catheters can occur immediately after injection but also with a delay in non-monitored patients [24]. The underlying mechanisms can be explained by extensive catheter threading beyond the correctly placed canula [2] or by late dislocation [25,26]. Because of the good analgesic effects of interscalene brachial plexus blocks, whether administered as a single injection or continuously, they remain the standard of care for many anesthesiologists. To understand the mechanisms of potentially severe side effects, profound anatomical knowledge of the posterior triangle and the adjacent space of the neck is mandatory. For example, the epineurium of the ventral rami of the spinal nerves is the continuation of the dura mater. One could argue that the interscalene brachial plexus block may be considered a central neuraxial rather than a peripheral nerve block (Figure 1). The reports of life-threatening cases of neuraxially misplaced interscalene catheters confirm these propositions. We would like to discuss, from an anatomical basis, why the safe use of interscalene catheters is a balance between high analgesia quality and complications, much like the passage between Scylla and Charybdis.

2. Relevant Anatomy for Interscalene Brachial Plexus Blocks

2.1. Innervation and Formation

The brachial plexus originates from spinal segments C5 to Th1 and innervates the upper extremity and parts of the trunk wall. The plexus runs through four defined anatomical regions according to its course from proximal to distal directions: (1) the scalene hiatus (the hiatus scalenorum posterius), (2) the posterior or lateral neck triangle (trigonum colli posterius sive laterale, the posterior cervical triangle), (3) the infraclavicular region (trigonum deltoideopectorale sive, Mohrenheim’s fossa or groove), and (4) the axillary fossa.

The fila radicularia arise from the spinal cord and form the radix anterior and radix posterior, which then form the actual spinal nerves in the spinal canal. In the cervical region, the spinal nerves emerge, after their formation, from the spinal canal and on through the neuroforamina (intervertebral foramina). Immediately after their ascent from the neuroforamen, they split into a larger ventral ramus and a smaller dorsal ramus (Figure 2). The brachial plexus is formed in the posterior scalene gap by the ventral rami of the spinal nerves. The dorsal rami supply the neck muscles with motor activity and the skin of the neck and occiput with sensory activity. They are not involved in the formation of the brachial plexus. The distance to the plexus formation is often less than 1 cm after the splitting of the spinal nerves.

In the scalene gap, the brachial plexus is located deep to the prevertebral fascia between the anterior and middle scalene muscles in a compartment that is rich in nerves and vessels. When observed in ultrasound images, the ventral rami are often described as roots; however, the anatomical term roots describes the anterior and posterior radices occurring before the formation of the spinal nerves. They are located more centrally and are not visible through ultrasound as they are covered by the vertebral arch of the bony spinal column. After they exit the neuroforamina, the ventral rami can be visualized sonographically.

2.2. Anatomical Variations in the Scalene Gap

As a consequence of embryonic development and the principle of segmental innervation, fibers of C4 (cranial fixation) and T2 (caudal fixation) may be involved in plexus formation as well. In addition to variants of plexus formation, variants of the scalene musculature [28,29,30,31,32] and/or vessels [33] can frequently be found. There are often muscular connections between the anterior and middle scalene muscles. Another scalene muscle, the scalene minimus muscle, referred to in older nomenclature as the levator pleurae muscle [32], is frequently present. Such muscles may separate the plexus into different parts. In practice, such muscle formations, as well as the loose connective and fatty tissue located in the scalene gap (Figure 1), narrow the gap and make spatial LA distribution unpredictable or even prevent it from spreading around the plexus (Figure 3).

2.3. Variants of the Course of the Ventral Rami

In the scalene gap, the cranial plexus parts (C5 and C6) frequently disperse between the anterior and middle scalene muscles and often do not intersect the caudal parts of C7, C8, and Th1 (Figure 3). C5 may run anteriorly or through the anterior scalene muscle (Figure 3 and Figure 4), and C5 and C6 may also pierce the anterior scalene muscle together [32]. This explains the clinical observation of incomplete interscalene blocks regardless of whether neurostimulation or ultrasound guide techniques are used (Figure 3 and Figure 4).

2.4. Compartments of the Lateral Neck and Their Relevance to Side Effects

In the interscalene gap, the brachial plexus lies deep to the prevertebral fascia, as do the ventral rami of C1 to C4 that form the cervical plexus. The phrenic nerve usually arises from C4 and occasionally, in part from C3 or C5 or even lower cervical segments. The nerve then runs covered by the prevertebral fascia on the ventral surface of the anterior scalene muscle. Figure 5 demonstrates the close vicinity of the brachial plexus to the phrenic nerve.

The sympathetic trunk is located medial to the scalene gap on the surface of the longus colli muscle. It is anteriorly attached and partially enveloped by the prevertebral fascia. If LA spreads medially deep or anterior to the anterior scalene muscle, it can disperse into the scalenovertebral triangle [31,34]. The triangle is bounded laterally by the anterior scalene muscle and medially by the longus colli muscle. The tip of the triangle is marked by the carotid tubercle (the anterior tubercle) of the sixth cervical vertebra. The base is defined by the subclavian artery or the cupula pleurae [35]. The large vessels, the common carotid artery, and the internal jugular vein, as well as the vagal nerve, are enveloped in this position by the carotid vagina and mark the ventral border.

Due to the presence of the pleural dome, the scalenovertebral triangle corresponds to a pyramidally shaped three-dimensional space. The prevertebral fascia surrounds the anterior scalene and the longus colli muscle. However, the prevertebral fascia does not separate these two muscles. Consequently, there is a connection from the scalenovertebral triangle to the dorsal prevertebral, the so-called Danger Space [27], which connects to the mediastinum.

2.5. Danger Space

The Danger Space constitutes a loose connective tissue space with a craniocaudal extension from the skull base to the diaphragm. This space contains the recurrent laryngeal nerve, parts of the sympathetic trunk and the vertebral vessels, autonomic fibers of the heart, and the phrenic nerve [27] (Figure 5). Therefore, LA spreading into this space can cause unwanted side effects like Horner’s syndrome or hoarseness.

2.6. Are Brachial Plexus Blocks Sufficient for Shoulder Surgery?

ISBs are usually applied for shoulder surgery, which ranks among the most painful surgical procedures [36]. The suprascapular nerve and the axillary nerve, both originating from the ventral rami from C5 and C6, which form the superior trunk, are the main suppliers of the shoulder joint. The lateral pectoral nerve, which originates from the lateral cord, is involved, and so too are the subscapular nerves [37,38]. Since the suprascapular nerve originates from the most proximal section of the brachial plexus (the superior trunk), the target for shoulder surgery must be the brachial plexus at the level of its trunks. This explains the superior analgesic effects of an interscalene or supraclavicular block. Furthermore, the cutaneous innervation of the shoulder cape is also mediated by the supraclavicular nerves originating from C4, like the phrenic nerve. They have a short course of up to 2 cm underneath the prevertebral fascia; they then penetrate the fascia and become superficial (Figure 6). After passing through the prevertebral fascia, they run in a different compartment, superficial to the brachial plexus. These anatomical facts indicate why they are unlikely to be blocked using peripheral techniques (such as a costoclavicular block). Therefore, if shoulder surgery is carried out solely with peripheral nerve blocks, brachial plexus blocks must be combined with a supraclavicular nerve block to anesthetize the skin for incision or closure.

3. Anatomical Challenge between Scylla and Charybdis

3.1. Local Anesthetics Volume

Clinically used LA volumes can spread medially to the anterior scalene muscle and reach vascular and nervous structures in the Danger Space and possibly the mediastinum. This spreading may include the recurrent laryngeal nerve, the cervical sympathetic trunk, nerves of the autonomic innervation of the heart, or the phrenic nerve [39]. Therefore, the anatomical conditions may explain the extensive side effects that are regularly reported in the literature with high LA injection volumes of up to 40 mL or even more [3]. Reducing the LA volume to 5–7 mL may reduce the incidence of side effects but cannot preclude them completely. A further reduction in the LA volume results in incomplete analgesia [40]. However, even with such small volumes, magnetic resonance imaging has demonstrated that the LA can reach the neuroforamina and the cervical epidural space [5]. These results were also shown in an anatomical study conducted on body donors [41].

Since the scalene gap is not an empty space but is filled with connective and fatty tissue between the scalene muscle fascias and the prevertebral fascia, as well as around the spinal nerves, the spread of LA volumes injected in this area is unpredictable or uncontrollable. An LA distribution inside a “brachial plexus sheath”, as historically cited repeatedly, is more a wishful thought than an anatomically reproducible fact [42].

3.2. Interscalene Access

Ultrasound guidance is increasingly being used for ISB performance, whether through an in-plane (IP) or out-of-plane (OOP) technique. The OOP puncture technique, which is frequently used in practice, comes very close to the original landmark-based trajectory, according to Winnie [43]. In the age of ultrasound, clinicians should be aware that because of the round nature of the neck, an ultrasound beam targets medially, consequently resembling the pathway of Winnie if the canula is advanced perpendicularly to the beam. Therefore, the complications known from Winnie’s technique (subarachnoid, epidural, or vertebral artery injection) can also occur despite the use of sonography. This can be avoided if the needle is advanced in a slide-down technique from medial to lateral directions while shifting the probe along the course of the spinal nerves.

3.3. The Risks of Interscalene Catheters

If catheters are advanced, they can also deviate medially toward the neuroforamen or caudally in the direction of the scalenovertebral triangle and thus reach the pleura (Figure 4) [44]. This can explain inter-/intrapleural catheter malpositions if the catheter is being threaded too far or by a secondary catheter shift. On the other hand, if the catheter is advanced under the prevertebral fascia in a lateral direction, it can follow an anatomically preformed space with loose connective tissue in an infraclavicular/axillary direction (Figure 7) [39]. This corresponds to the landmark-based techniques described by Borgeat [45] and Meier [46]. Both techniques aim in a lateral direction and not toward the neuroforamina like Winnie’s original technique. They are therefore less prone to complications.

The connective tissue, the ventral rami, and vessels in the scalene gap are obstacles during catheter advancement, making it so that a catheter cannot always be reliably placed in the target position. Catheter advancement several centimeters beyond the canula tip is therefore frequently enforced in this area. If doing so, the definite course and final position of the catheter are speculative. A straight-line course within a predefined plexus sheath is anatomically impossible. Hence, a course consisting of moving the catheter in the direction of the neuroforamina is almost to be expected if the catheter has advanced too far [2,19].

3.4. Awake or Asleep Interscalene Blocks?

For decades, anesthetists have believed that performing blocks and inserting catheters are safer for awake patients. This is underlined by the case reports of Benumof and recommendations by anesthesia societies [18,47]. We have been performing ultrasound-guided blocks on thousands of anesthetized patients for decades. We have observed that patients more often agree to regional blocks if they are carried out under deep sedation or general anesthesia. For children, a puncture under deep sedation or anesthesia is generally accepted [48], whereas for adults, only awake punctures have been recommended in line with the relevant guidelines [47]. On closer inspection, the pediatric recommendations of the ASRA apply to “children” up to 18 years of age. In this context, it is inexplicable to anxious adult patients why different “safety standards” should apply to them compared to children, whose far smaller anatomical proportions and resulting shorter distances place even greater demands on the skills of a clinician [49]. The fact that touching nerves with the cannula does not necessarily trigger paresthesia has been repeatedly reported and can be regularly observed in routine sonography. Therefore, the absence of paresthesia cannot be taken as evidence of an extraneural and, thus, supposedly safe injection. If the basic rule “Without canula visualization: no needle advancement, no LA injection” is maintained, an increase in complications is not to be expected among asleep patients. According to the data available to date, it cannot be proven that there is a higher safety margin for awake patients [49].

In his excellent review, Marhofer pointed out myths and facts about regional blocks carried out under sedation or general anesthesia [49]. He concluded that with the use of ultrasound guidance in skilled hands, it is a reasonable option to carry out blocks in anesthetized or sedated patients. We fully support this statement, especially since interscalene catheter placements are often perceived to be unpleasant by awake patients.

3.5. Late Onset of Side Effects and Complications

It must be considered that side effects and complications can also occur late and at any time during continuous LA application. It is well known that, in principle, catheters can dislocate during the postoperative course, particularly during movement or physiotherapy [25,26,44,50]. If a reduction in effectiveness occurs over time during continuous catheter analgesia, the diagnosis is clinically simple. However, sonographic examinations have also shown that changes in catheter position occur in situ, even without a loss of efficacy [26]. As recommended by Gaus et al. [2], in order to detect such malpositions in a timely manner, continuous care must be ensured until the end of catheter therapy. In addition, patients must be informed about such unpredictable catheter dislocations and the possibility of the late occurrence of side effects or complications.

In a study involving 1505 outpatients, 27% of patients reported shortness of breath at home, while 13% and 7% reported hoarseness and dysphagia, respectively [51]. This clearly shows that side effects are almost inevitable, even in the late course, whether due to unnoticed catheter dislocation or simply the cumulatively applied LA volumes. In an anatomical study, a mean volume of only 4 mL was determined for the scalene gap [52]. This volume is higher than in the living, as connective tissue, nerves, and vessels are removed. In addition, the prevertebral fascia was removed, too. The fascia acts as the ventral boundary of the scalene gap and is therefore important for LA distribution. For example, a continuous rate of only 4 mL/h of LA results in 96 mL/day. Additional bolus function increases the volume even more. This amounts to a huge volume in the small scalene gap. The route the LA will take over time depends crucially on the anatomical pathways, the consecutive pressure conditions in the compartment, and the resorption rate at the application site. However, this highly individual resorption rate is ultimately unknown.

3.6. The Role of Ultrasound

Modern ultrasound probes, such as those used for ISB, offer an axial and lateral resolution of approximately 0.4 mm and 0.7 mm, respectively, at a depth of 2 cm and a transmission frequency of 12–15 MHz. They allow extremely precise needle guidance and positioning. However, such high puncture precision can only be achieved by experienced operators using an excellent sonography technique. It is important to move the ultrasound probe in small steps in order to generate a sonogram that appears almost fixed. One way to train in the scanning of the lateral region of the neck is the so-called traceback technique [53]. Ultrasound helps to visualize LA spread, too. Physicians can stop the injection if LA spreads toward an undesired direction or proceed if the distribution is correct (Figure 8).

As discussed by Gaus et al. [2], the sonographic visualization of a catheter in situ is not easy, but it is certainly possible [26]. Catheters are difficult to visualize in situ in a single acoustic plane. Due to their anatomically determined, usually non-linear course, they cause significantly lower reflection than a straight cannula. The use of ultrasound-optimized catheters, which have recently become available (Figure 9), also facilitates catheter identification. However, if a catheter is advanced in an uncontrolled manner several centimeters beyond the visible cannula tip, identification of the catheter tip is almost impossible if the catheter tip has already reached the neuroforamen or the scalenovertebral triangle. In a study evaluating the dislocation rates of ISC among volunteers, Marhofer et al. advanced a catheter 3 cm beyond the needle tip and retracted it until a slow injection of saline confirmed an optimal spread pattern at the ventral rami C5–7, thus excluding intrathecal placement [25]. Such a procedure seems to be the most reasonable technique but is rarely addressed in the literature; thus, it is probably not performed by most practicians. Many anesthetists believe that catheter advancement several centimeters beyond the cannula tip prevents dislocations. In our own study using self-coiling catheters, catheters were advanced 2 cm beyond the tip until the coil was outside of the cannula. The correct position was confirmed sonographically [26]. The technique of using catheter depiction to sonographically confirm correct positioning was also used by others [54]. Despite an initially correct placement, it was demonstrated that catheters dislocated subcutaneously in the postoperative course without presenting externally visible signs. To give the catheter a certain stability in situ with a short puncture path to the target, subcutaneous tunneling can be performed instead of pushing the catheter beyond the target area in an uncontrolled manner [25]. A flat puncture angle with a lateral and caudal puncture direction corresponding to the anatomical course of the spinal nerves can also lead to a more stable catheter position in situ (Figure 9) [39]. However, clinical confirmation with larger case numbers is still lacking in this regard.

4. Limitations

The limitations of this review include the insufficient availability of data on prospective and blinded studies on interscalene catheters that were sonographically monitored postoperatively and over time. While the methodologies of individual studies occasionally provide more or less detailed descriptions of ultrasound-guided punctures, there is a lack of anatomical richness in the details. The anatomical foundation with which the complications and side effects of interscalene catheters can be explained was the motivation behind writing this article, aiming to encourage the consideration of such details in the future.

5. Summary

Interscalene analgesia is still considered the gold standard for the treatment of postoperative pain after shoulder surgery. This technique provides excellent analgesia but is associated with frequently occurring side effects and even life-threatening complications. To date, the evidence of alternative techniques for continuous postoperative analgesia is sparse. Profound anatomical knowledge is the key to understanding why side effects and complications of brachial plexus blocks at the interscalene level may occur. Currently, ultrasound is the best regional anesthesia strategy with which to face the complex demands of regional anesthesia of the lateral neck; identify individual variable anatomy; perform safe punctures; observe correct LA spread; detect (mis-)placed catheters; and therefore provide best practices to patients scheduled for shoulder surgery. Profound manual skills and knowledge of the relevant spaces and potential pathways of local anesthetics are the indispensable basis for safe and effective blocking techniques, thus avoiding travel between Scylla and Charybdis.