Fluoroscopy and Ultrasound-Guided Sympathetic Blocks: Stellate Ganglion, Lumbar Sympathetic Blocks, and Visceral Sympathetic Blocks


The paravertebral sympathetic chain extends from the neck down to the coccyx where it fuses to become a single ganglion. Besides its role in the defense of the body, it has been shown to be involved in the propagation and chronicity of certain types of pain. Pain practitioners have targeted the chain at various locations, including the stellate, celiac plexus, lumbar sympathetic chain, superior hypogastric plexus, and ganglion impar, depending on the location of pain, by performing image-guided diagnostic and therapeutic blocks. This chapter will discuss the relevant anatomy, technique, evidence, and complications of these sympathetic blocks.


celiac plexus, ganglion impar, image guidance, lumbar sympathetic ganglia, stellate, sympathetic blocks



Analgesic and neurolytic blockade of the autonomic (sympathetic) nervous system was introduced in the beginning of the previous century. The frequently observed autonomic phenomena in different acute and chronic painful conditions prompted physicians to investigate and invent methods of interruption of the sympathetic pathways with the purpose of alleviating chronic pain. Though the use of the paravertebral analgesic blockade was primarily developed for surgical anesthesia, the method was introduced for the diagnosis and treatment of visceral pain. Eventually, the definition of “paravertebral blockade” evolved. The interventions targeting sympathetic ganglia and nerves were christened as “the sympathetic nerve blocks.”

The sympathetic blocks are anatomically subdivided according to the corresponding spinal segments, vis à vis cervical, thoracic, lumbar, and sacral. The abdominal autonomic ganglia (e.g., celiac, mesenteric) are surrounded by visceral and vascular structures and situated ventrally from the vertebral column; special analgesics and neurolytic techniques have been developed.

There has been confusion regarding the type and structure of nerve cells and fibers—the blockade of which produces an analgesic effect. While the efferents have strict dichotomy to preganglionic and postganglionic fibers, the afferents, bearing pain signals may or may not make ganglionic connections. Aside from the purely autonomic pathways, sensory sympathetic unmyelinated fibers accompany somatic nerves making connections with the prevertebral or cranial ganglia. Therefore, blockade of a large mixed somatic nerve would inevitably result in certain autonomic phenomena. The detailed discussion of the autonomic nervous system anatomy and taxonomy is beyond the scope of this chapter. It is sufficient to mention that either autonomic ganglia or traversing fibers are blocked; the end result is abolition of the sympathetic tone (e.g., vasodilation, dry skin or mucosa, increased temperature) and alleviation of the sympathetically maintained pain that is typically described in terms of burning, diffuse, and annoying.

Stellate Ganglion and Cervical Sympathetic Chain Blockade

Cervical sympathetic analgesic and neurolytic blockade are commonly used in the diagnosis and management of sympathetically mediated pain and vascular insufficiency of the upper extremities. In addition, the stellate ganglion block has been advocated for the treatment of a variety of medical conditions, such as phantom pain, postherpetic neuralgia, cancer pain, cardiac arrhythmias, orofacial pain, and vascular headache. Recently, cervical sympathetic blockade has been suggested as an effective method for the prevention and treatment of the cerebral vasospasm, as well as the posttraumatic stress disorder.

The stellate ganglion, also known as the cervicothoracic ganglion, represents a fusion of the inferior cervical and first thoracic ganglions of the sympathetic trunk. It can be found in about 80% of the population. Anatomy and position of the stellate ganglion have been investigated by dissection, magnetic resonance imaging (MRI), and computed tomography (CT). It is usually situated at the lateral border of the longus colli muscle anterior to the neck of the first rib. It lies posteriorly to the vertebral vessels and is separated from the cervical pleura by the suprapleural membrane inferiorly. It measures 1–2.5 cm long, about 1 cm wide, and 0.5 cm thick, and may be fusiform, triangular, or globular.

Although a C7 approach to stellate ganglion has been described, the blockade is routinely performed at the C6 level according to the following anatomic landmarks: prominent anterior tubercle of the transverse process (the Chassaignac’s tubercle), cricoid cartilage, and carotid artery. Given that only traversing sympathetic fibers or middle cervical ganglia can be found at the C6 level, the procedure should more accurately be named the cervical sympathetic block. The middle cervical ganglion or traversing sympathetic fibers are located anterolateral to the belly of the longus colli muscle. Conceivably, such a “convenient” position makes it easily accessible for either the diagnostic or the therapeutic blockade.

Cervical sympathetic block has been traditionally performed as a surface anatomy-based (also known as a “blind”) injection, even though fluoroscopic verification has been strongly recommended. Practitioners are typically taught to palpate Chassaignac’s tubercle, gently retract the carotid artery, and then insert the needle paratracheally until it contacts bone surface. The needle is then withdrawn 1–5 mm, and a solution injected. This maneuver was assumed to be sufficient to position the needle outside the longus colli muscle, where the stellate ganglion was thought to be situated. However, this “blind” paratracheal injection technique has been associated with unreliable results and multiple side effects and complications, including intravascular injection, bleeding, temporary paralysis of the recurrent laryngeal nerve, discitis, and esophageal injury.

A “blind” injection at the C6 level on the left side may cause inadvertent esophageal puncture. Hematoma formation is likely related to penetration of the vascular thyroid gland or damage of the inferior thyroid artery. Postprocedural hoarseness is attributable to an injury, although fortunately transient, of the recurrent laryngeal nerve.

Fluoroscopic guidance reduces overall risk associated with the “blind” technique. It has the advantage of identifying bony anatomy, although the anatomic position of the cervical sympathetic trunk (CST) may be only approximately correlated with the X-ray image. CST is completely surrounded by soft prevertebral tissues (e.g., longus colli muscle, fascia, carotid artery). Preliminary injection of a contrast agent should produce a “honey-comb” longitudinal spread along the longus colli muscle. Nevertheless, often the spread is rather intramuscular and inconsistent. Ostensibly, neither “blind” nor fluoroscopy-guided injection can guarantee a desirable result and prevent adverse outcomes.

Injection of local anesthetics at the C6 level has a long history and evolution, but the reliability of achieving blockade of the stellate ganglion has only recently been tested. The success or failure of the cervical sympathetic block is contingent on precise needle placement, and is therefore entirely dependent on the anatomic location of the CST and the thickness of the longus colli muscle. Several clinical and cadaver trials have been performed in an attempt to elucidate the pattern of spread when solutions are injected at the C6 level. These studies reported conflicting results. The discrepancies were probably related to differences in research design: cadavers versus live subjects, low volume versus high volume of injectate, and CT versus fluoroscopy control. One cadaver experiment was suggestive that only deposition of a solution into the prevertebral “interlaminar space” provides reliable spread to the stellate ganglion. The cervical prevertebral fascia is attached to the base of the skull and extends over the prevertebral muscles (longus capitis, rectus capitis, and longus colli muscles) to the T4 vertebra, just beyond the longus colli muscle. This positioning of the fascia forms a plane along which the injected fluid can flow.

Although some anatomic and imaging studies indicate a subfascial position of the sympathetic chain, other sources relate the path of the CST to the suprafascial plane. Two published studies unequivocally demonstrated the true anatomical pathway of CST. One utilized cadaver dissections and showed subfascial position of the stellate ganglion. This study described a highly variable thickness of the longus colli muscle, which may lead to failed “blind” injection. The second study was designed as a step-by-step methodology validating of a new ultrasound-guided approach (described below); a subfascial position of the sympathetic trunk was discovered by three-dimensional (3D) ultrasonography and confirmed by cadaver dissections. In addition, this study measured the thickness of the longus colli muscle at the C6 level, and proved that the muscle is 2–10 times thicker than was previously suggested in the regional anesthesia literature. As such, routine injection by the traditional method most certainly ends up as an intramuscular injection, although CST will be eventually blocked by an overflow or diffusion of the injectate. Even the CT guidance may not assure the precise needle position between fascial layers ( Fig. 84.1 ).

FIG. 84.1

Three-dimensional reconstruction of the stellate ganglion block. The contrast agent spread is intramuscular.

Ultrasound guidance is a logical solution to ensure accurate injection when soft tissues are involved. Conspicuous muscles, fasciae, blood vessels, viscera, and bone surface make ultrasonography superior to fluoroscopy for image-guided CST block. In 1995, Kapral and colleagues described an ultrasound-guided technique in a case series. Compared with “blind” injection, the authors found that the ultrasound-guided stellate ganglion block required a smaller quantity of local anesthetic (5 mL vs. 8 mL), was not associated with the formation of a hematoma (whereas three patients in the blind injection group had a hematoma), and produced a more rapid onset of Horner’s syndrome. However, because tissue visualization was probably not feasible below the C7, the authors concluded that a local anesthetic depot was limited to the C4–C7 levels and speculated that the abolished sympathetic response in the upper extremity was not related to the blockade of the stellate ganglion per se. Their findings agreed with those published by Hogan and Erickson, but these observations and conclusions have been refuted by Gofeld and colleagues, who observed the contrast agent spread between the C4 and T1 levels in all patients, occasionally reaching the T2 level.

Shibata and colleagues were the first to suggest that subfascial injection would result in a better spread of the injectate and more reliable sympathetic blockade ; however, the published image in that study was more consistent with an intramuscular injection. Such an injection may be a limiting factor in the onset and spread of the blockade. In a study by Gofeld and colleagues, 5 mL of local anesthetic injected beneath the fascia, but superficial to the longus colli muscle, ensured consistent spread of the solution to the stellate ganglion. More recently, experimental clinical studies endorsed the ultrasound-guided technique and found it superior to the fluoroscopy-controlled technique.


There are two ultrasound-guided approaches to the CST: the modified “anterior” paratracheal out-of-plane approach, and the newer “lateral” in-plane method. Both techniques can be performed using either low-frequency curvilinear or high-frequency linear ultrasound transducers. Low-frequency sonography provides a wide-angle field of view and facilitates needle entry planning, while a high-frequency transducer allows better resolution of the pertinent anatomy and fascial planes.

Anterior Approach

The patient is positioned supine. A pillow can be placed under the lower neck to achieve an extension. The head may be slightly rotated contralaterally to the injection side, thereby increasing the distance between the carotid artery and trachea and improving the sonographic view. After skin preparation and dressing, sterile ultrasonic gel is applied. The transducer is covered by a sterile adhesive transparent dressing or sleeve. Ultrasonography of the anterior neck is performed with initial transducer placement at the level of the cricoid cartilage, anterior to the sternocleidomastoid muscle. Short-axis ultrasonography reveals the typical appearance of the C6 transverse process—the prominent anterior tubercle, the short posterior tubercle, and the exiting C6 nerve root ( Fig. 84.2 ). Scanning caudally and dorsally brings the C7 transverse process into the view. The C7 transverse process has no anterior tubercle. The C7 nerve root is situated just anterior to the posterior tubercle ( Fig. 84.3 ). At the C6 level, the longus colli muscle is seen as an oval structure adjacent to the base of the transverse process and vertebral body (see Fig. 84.2 ). Sometimes the caudal portion of the longus capitis muscle can be seen as well. The CST is visualized as a spindle-shaped structure (the mid-cervical ganglion), which is typically situated at posterolateral surface of the longus colli muscle; if the CST cannot be identified, some widening of the tissue plane underneath the prevertebral fascia is typically seen. Once the desired target is localized, the surrounding anatomical structures are identified and the feasibility of the “anterior” approach should be determined. Often the distance flanked by the carotid artery, and the trachea is wide enough and, therefore, only thyroid tissue and superficial neck muscles are seen between the needle entry and the surface of the longus colli muscle. Gentle pressure may effectively decrease the skin-to-target distance and visually further separate the carotid artery from the trachea. Additional scanning should be performed to confirm that the inferior thyroid artery is not traversing the planned needle trajectory. The injection is performed as a short-axis out-of-plane approach ( Fig. 84.4 ). The skin is anesthetized immediately caudad to the transducer. The injection is performed using a spinal needle (22–25 gauge and 2–3.5 inches long) with a three-way stopcock and extension tubing connecting two syringes, one with NaCl 0.9% and another with a local anesthetic. The needle is inserted under continuous ultrasound guidance, directed to the anterior surface of the longus colli muscle using a short-axis out-of-plane approach. When the needle tip is visualized, either directly or indirectly (tissue movement), approaching the target, 1–2 mL of saline is injected to confirm placement of the needle under the prevertebral fascia, which is appreciated as a separation of the tissue planes by the anechoic fluid ( Fig. 84.5 ). If the injectate is observed above the fascia or within the muscle, the needle must be carefully repositioned. If the spread is appropriate, 5 mL of local anesthetic is injected, and the needle is withdrawn.

FIG. 84.2

Linear transducer positioned at the cricoid cartilage level (right); sonogram. Asterisk, longus capitis muscle; C, carotid artery; C6, C6 vertebra; double asterisk, longus colli muscle; N, exiting C6 nerve root; SCM, sternocleidomastoid muscle; T, trachea.

FIG. 84.3

Sonogram at C7 vertebral level. C7, C7 vertebral body; N, exiting C7 nerve root; VA, vertebral artery.

FIG. 84.4

Sonogram of anterior neck obtained with gentle transducer pressure. C, Carotid artery; C6, C6 vertebral body; E, esophagus; LCM, longus colli muscle; Th, thyroid gland; T, trachea; white dotted line, skin to target distance (2.2 cm).

FIG. 84.5

The local anesthetic (LA) injected superficially to the longus colli muscle (LCM) and underneath of the prevertebral fascia. The longus colli muscle is compressed by the injectate and appears “hyperechoic.” C, Carotid artery; Th, thyroid gland.

The “anterior” approach must be abandoned and an alternative “lateral” approach may be attempted if any of the following conditions are present: (1) the anterior sonogram shows narrow distance between the carotid artery and the thyroid; (2) the inferior thyroid artery cannot be eliminated from the view; (3) the esophagus is seen above the longus colli muscle (left side); or (4) thyroid cysts are evident.

Lateral Approach

The patient is placed in the semilateral decubitus position, with the side to be treated uppermost. Preparation and ultrasonography is performed as previously described. However, the transducer is centered at the C6 transverse process and not at the anterior neck. It is of utmost importance to localize the C6 nerve root and the anterior process. With the transducer properly placed, only the anterior tubercle of the C6 transverse process is visible adjacent to the projected entry point of the needle without visceral or neural elements between the entry site and the anterolateral surface of the longus colli muscle. The needle tract should be entirely intramuscular, passing through the sternocleidomastoid muscle or the anterior scalene muscle, or both. Occasionally the internal jugular vein is seen within the projected needle tract, but it can be readily collapsed by light pressure on the transducer. Skin anesthesia is performed immediately posterior to the ultrasound transducer. Under continuous ultrasound guidance, the needle is inserted using the short-axis in-plane approach ( Fig. 84.6 ). The advantage of the lateral approach, in addition to avoiding the trespass through the thyroid, is in the in-plane control of the needle insertion from the skin entry point to the target. Verification of the needle position and the rest of the procedure are the same as that for the anterior approach. An injection of 5 mL of a local anesthetic typically results in C3–T1 prevertebral spread and the complete blockade of the CST and the stellate ganglion ( Fig. 84.7 ). If anesthetic blockade of the upper cervical ganglion is not desirable, it will be prudent to limit volume of the injectate to 3 mL.

FIG. 84.6

In-plane insertion of the block needle (arrowheads) superficially to the longus colli muscle (LCM). Injectate (blue) is wrapping LCM. C6, C6 vertebra (note: only the anterior tubercle is seen); CA, carotid artery; SCM, sternocleidomastoid muscle; T, thyroid gland.

FIG. 84.7

Anteroposterior X-ray of the cervical spine. The contrast dye is clearly seen between C5 and T2 spinal levels as an interfascial (nonmuscular) pattern.

Fluoroscopy-Guided Approach

The patient is positioned supine. A pillow can be placed under the lower neck to achieve some extension. The head may be slightly rotated contralaterally to the injection side increasing the distance between the carotid artery and the trachea. The image intensifier is positioned in such way that the cervical spine anteroposterior view is obtained. Additional adjustments are made to eliminate misalignment of the lower cervical intervertebral discs and to correct orientation of the spinous processes. After skin preparation and dressing, the target—junction between C6 transverse process and the vertebral body—is localized. Skin anesthesia is performed and the block needle inserted utilizing “tunnel view” until the bone surface is contacted. Prior to the needle placement, the anterior neck is palpated to verify absence of the carotid pulse at the entry point. If the carotid pulse is appreciated, the image intensifier may be slightly axially rotated or “tunnel view” may be abandoned. Once the bone surface is contacted, the needle is gently retracted for 3–5 mm and injection of a contrast agent is performed. Although, a “honey-combed” longitudinal spread has been reported as a confirmation of interfascial flow along the CST, most commonly the spread is intramuscular ( Fig. 84.8 ). Although the needle tip is not directly located at the CST, injection of 10 mL of bupivacaine 0.25% or another local anesthetic will typically result in the facial and upper extremity sympathetic blockade in 10–15 minutes. This characteristic delay is attributable to a secondary spread onto the CST.

FIG. 84.8

Anteroposterior fluoroscopy demonstrates correct needle placement, and an intramuscular spread of injected contrast agent.


The blind paratracheal technique produces unreliable results, and is associated with a variety of side effects and complications, such as intravascular injection, formation of hematomas, temporary paralysis of the recurrent laryngeal nerve, discitis, and esophageal injury. Fluoroscopy guidance may prevent adverse outcomes related to intravascular, nerve root or neuroaxial injections. However, soft tissue, parenchymal, and vascular injury may occur.


Current literature and trends in clinical practice have made ultrasonography the preferred imaging guidance for CST block. A 5-mL injection into the correct fascial plane may reliably produce the sympathetic blockade. Ultrasound guidance may prevent complications and adverse outcomes associated with either blind or fluoroscopy-guided techniques.

Key Points

  • The stellate ganglion is usually situated at the lateral border of the longus colli muscle anterior to the neck of the first rib.

  • Given that only traversing sympathetic fibers or middle cervical ganglia can be found at the C6 level, the stellate ganglion block should more accurately be named the cervical sympathetic block.

  • At the lower neck, the sympathetic trunk is typically situated underneath the prevertebral fascia at posterolateral surface of the longus colli muscle.

  • Ultrasound-guided block is a standard approach as it allows real-time visualization of pertinent anatomy, needle insertion, and spread of local anesthetic.

Lumbar Sympathetic Block

Historically the first reported lumbar sympathetic block (LSB) was done by Brunn and Mandl, who, in 1924, described a technique of injecting the lumbar sympathetic nerves as a component of their paravertebral approach blocking the mixed spinal outflow in the lumbar region. Kappis also described a technique of the LSB and surgical resection of the lumbar sympathetic nerves around that time. During the 1950s, Bonica described in detail the importance of the lumbar sympathetic blockade, particularly in the treatment of the causalgia and posttraumatic reflex dystrophies in servicemen after World War II.

There are two paravertebral sympathetic trunks located symmetrically at each side of the vertebral column. The thoracolumbar sympathetic fibers originate in the dorso-lateral region of the anterior column and extend within the ventral roots of all the thoracic and the upper two or three lumbar spinal nerves. These preganglionic fibers form the rami communicantes that connect to the sympathetic chain. Some rami communicantes bypass the prevertebral plexuses and connect to the collateral ganglia (e.g., the celiac plexus). The cell bodies responsible for vasoconstriction in the lower limbs are in the lower three thoracic and first three lumbar segments. Each lumbar sympathetic chain enters the retroperitoneal space under the right and left crus of the diaphragm, continuing inferiorly in the interval between the anterolateral aspect of the vertebral bodies and the origin of the psoas muscle to enter the pelvis and the L5–S1 disc. Posteriorly, the periosteum overlies the vertebral bodies and the fibro-aponeurotic origin of the psoas muscles and their fascial coverings. The parietal reflection of the peritoneum is anterior, the aorta lying anteromedial to the left trunk, and the vena cava anterior to the right trunk. It should be noted that the white and gray rami communicantes pass to their respective ganglia beneath the fibrous arcades of the psoas attachments alongside the middle of the vertebral bodies.

The sympathetic ganglia of the lumbar sympathetic chain are variable in both numbers and position. Rarely, all five ganglia are found on each side in the same individual. In most cases, only four are found. The L1 and L2 ganglia are usually fused in most patients, and ganglia are aggregated at the L2–L3 and L4–L5 discs. Also, there is considerable variability in the size of the ganglia, some being fusiform and as long as 10–15 mm, and others being round and approximately 5 mm long.

The indications for LSB may be divided into three broad categories:

  • 1.

    Circulatory insufficiency in the leg, including arteriosclerotic vascular disease, diabetic gangrene, Buerger disease, Raynaud’s phenomenon and disease, and reconstructive vascular surgery after arterial embolic occlusion

  • 2.

    Pain from renal colic, complex regional pain syndrome types I and II, intractable urogenital pain, postamputation stump pain, phantom pain, and frostbite

  • 3.

    Other conditions, such as hyperhidrosis, erythromelalgia, and acrocyanosis.

Although the mechanism contributing to pain relief remains unclear, blocks of the sympathetic nervous system may have two actions: (1) interruption of preganglionic and postganglionic sympathetic efferents may influence the function of the primary afferent neuron; or (2) the afferents from deep structures in the leg that travel with the sympathetic nerves may be blocked.

Fluoroscopy-Guided Approach

With the patient in a prone position, the C-arm is placed first in the anteroposterior and then ipsilateral to the injection side oblique view. The injection is typically performed at L3 level. The degree of the obliquity is judged based on appearance of the L3 spinous process. It has to overlay the contralateral facet joint. When the L3 transverse process is aligned with the center of lateral L3 vertebral body, the C-arm is tilted to eliminate the transverse process from the view and assure an unobstructed needle path. Skin and subcutaneous tissues are anesthetized. A 22-gauge 5- to 7-inch needle is then introduced in a tunnel view and advanced until it contacts the vertebral body ( Fig. 84.9 ). Fluoroscopy in two planes should confirm its position and the angle to guide redirection of the needle to its final position at the anterolateral aspect of the vertebral body. With the fluoroscopy positioned laterally, final adjustments can be made to ensure the needle tip lies exactly at the anterolateral edge of the vertebral body. A contrast dye of 3–5 mL should confirm the location of the correct tissue plane ( Fig. 84.10 ). If the contrast spread was limited to one segment, the procedure is then repeated with the second needle at the L2 or L4 vertebral body. The dye should spread to form a line conforming to the anterolateral margin of the vertebral bodies.

FIG. 84.9

Utilizing a steep oblique view, the block needle is inserted in tunnel view aiming for the anterolateral L3 vertebral body.

FIG. 84.10

Fluoroscopy performed in lateral (A) and anteroposterior (B) views confirmed prevertebral spread of the contrast dye (black arrow) .

A short-acting local anesthetic is commonly used for diagnostic sympathetic blocks. A long-acting agent, such as bupivacaine or ropivacaine, is advantageous for both therapy and prognosis, because it gives the patient a longer time to evaluate the effects of sympatholysis and any effect this might have on the pain. A concentration of 0.375% bupivacaine or 0.5% ropivacaine gives optimal duration without the need for an added vasoconstrictor.

Ultrasound-Guided Approach

Fluoroscopic guidance remains the standard approach to LSB with the sympathetic chain lying in an anterolateral location to the lumbar vertebra. While ultrasound-guided interventional procedures are in their infancy, they have benefits over the established fluoroscopically and CT-guided procedures in the reduction of radiation exposure and the ability to visualize soft tissue structures. Ultrasound-guided LSB was described by Kirvelä and colleagues in 1992. Unfortunately, the description of the technique was unclear and the image quality was poor.


The patient is positioned prone and spinal sonography, with a curved array transducer, is performed for identification of spinal level and target. The injection target is the most ventral and medial part of the L3 vertebral body. After sterile preparation and administration of local anesthesia, a 20-g or 18-g Touhy needle is inserted utilizing an in-plane approach contacting the lateral L3 vertebral body ( Fig. 84.11 ). A loss of resistance technique is used to advance the needle deeper. Prevertebral space is reached usually after 1–2 cm, and an injection of 5–10 mL of 1% lidocaine is performed. Due to the limited visualization, ultrasound guidance for the neurolytic LSB cannot be currently recommended.

Sep 21, 2019 | Posted by in PAIN MEDICINE | Comments Off on Fluoroscopy and Ultrasound-Guided Sympathetic Blocks: Stellate Ganglion, Lumbar Sympathetic Blocks, and Visceral Sympathetic Blocks

Full access? Get Clinical Tree

Get Clinical Tree app for offline access