The paravertebral space (Fig. 12.3) is the shape of a four-sided pyramid with its apex facing posteriorly into the neural foramen and its base bordered anteriorly by the parietal pleura. The thoracic paravertebral space is defined by the following four borders: (1) the bone and articular capsules of the rib and transverse process above, (2) the rib below, (3) medially by the vertebral body, and (4) laterally by the intercostal space and the costotransverse ligament. The costotransverse ligament runs from the transverse process to the superior root and its continuation, the intercostal nerve. The intercostal nerve branches into dorsal and ventral rami in the paravertebral space. Gray and white rami communicates course through the space to and from the respective sympathetic ganglion at that level, which is also contained within the paravertebral space. Other contents include areolar tissue, fat, and blood vessels. It is important to keep in mind that the paravertebral space is contiguous with the epidural and intercostal spaces as it lies between these two other spaces. Any substance injected into the paravertebral space may potentially spread cephalad and caudad to adjacent paravertebral spaces as well as medially and laterally to the epidural and intercostal spaces, respectively [56]. Rarely, an injection into the paravertebral space will spread to the contralateral space, and this has been demonstrated radiologically [19–21, 57–59].

In general, topographic spread is variable and difficult to predict [60, 61]. Naja et al. performed a series of paravertebral blocks using nerve stimulator guidance to determine the effect of varying injection points on spread of solution [61]. Their findings indicated that injection in the more ventral aspect of the thoracic paravertebral space resulted in a multisegmental longitudinal spreading pattern. Injecting at the dorsal aspect of the space showed a cloud-like spread with limited distribution to adjacent segments (Fig. 12.4).

The similarity in the anatomic distribution and density of block produced by continuous paravertebral block and continuous epidural infusion would seem to indicate that some cases of unilateral “epidural” block might be attributable to inadvertent continuous paravertebral blockade. This phenomenon has been confirmed radiologically [62].

Patient comfort during performance of a paravertebral block is improved by good technique, the use of small-gauge needles, and the avoidance of paresthesias while performing the block [63]. Generous infiltration of local anesthetic also makes the procedure more tolerable. Sedation before the procedure is strongly recommended and adds greatly to patient comfort.

A linear high-frequency ultrasound transducer is used in the transverse plane with its medial border just lateral to the previously identified spinous process of the vertebral body at the level of the facet joint, and an ultrasound survey scan is obtained. Once the transverse process is identified, the transducer is slowly moved superiorly or inferiorly until the space between the two adjacent transverse processes is identified. One can identify the pleura that appears as a bright hyperechoic downward curving line, which can be seen to slide back and forth with respiration. Just above the hyperechoic pleural line as it curves down as it moves medial toward the vertebral body is the triangular-shaped thoracic paravertebral space. Just above the paravertebral space is the linear hyperechoic internal intercostal membrane. The depth of the posterior border of the paravertebral space is noted. When these anatomic structures are clearly identified on the transverse ultrasound scan, the skin is prepped with anesthetic solution, and a 3 ½ in. needle with stylet is advanced from the middle of the inferior border of the ultrasound transducer using an out-of-plane approach with the trajectory being adjusted under real-time ultrasound guidance until the needle tip is at the previously identified depth of the posterior border of the paravertebral space. After careful aspiration, a small amount of solution can be injected to aid in identification of the needle tip position. The needle is then advanced slowly with attention paid to the relative location of the bright hyperechoic pleura line until the needle tip is seen to be within the paravertebral space. After careful aspiration, the remainder of the solution is slowly injected. The needle is then removed and sterile pressure dressing placed at the injection site [67].

The most important factors for safe performance of paravertebral neural blockade are a solid knowledge of pertinent anatomy, meticulous attention to injection technique, and anticipation of all possible physiologic changes associated with the block. The clinician must have a comprehensive understanding of the potential complications. Early recognition facilitates rapid treatment, thus minimizing more serious sequelae. Utilization of a nerve stimulator-guided technique is associated with a higher success rate and fewer complications than standard techniques [68]. It is strongly suggested that an intravenous line be in place before performing the block.

It is imperative that low-osmolarity contrast agents be used when performing these blocks, because spread of high-osmolarity solutions into the subarachnoid space can lead to significant neurologic harm. The proximity of the paravertebral space to the central nervous system creates the obvious potential for needle entrance into either the epidural or subarachnoid space. Iodinated contrast has been injected into the epidural space with and without spread into the paravertebral space. In performing 45 paravertebral blocks, Purcell-Jones et al. showed contrast confined to the paravertebral space in only 18 % of procedures. There was epidural extravasation in 70 % and exclusive epidural spread in 31 % of cases [56].

In addition to epidural [13] and subdural [69] injection, unrecognized subarachnoid puncture can occur. Headaches not associated with obvious dural puncture occurred in 3 of 24 cases in one series of paravertebral blocks. Aspiration was negative for CSF before injection [70]. Negative aspiration for CSF is not an absolute guarantee of proper needle placement, especially with small-gauge needles or long, small-bore catheters. The headaches resolved with conservative management within 5–14 days postoperatively. The medial approach proposed by Shaw and modified by Tenicela and Pollan has shown excellent results with low complication rates [17, 65]. Of the 384 blocks performed by Tenicela and Pollan, there was one incident of pneumothorax (0.26 %), one recognized dural puncture, two intrathecal injections of the test dose, 18 incidents of hypotension (4.6 %), five bilateral blockades (1.3 %), and 27 incidents of fair to poor block (7.0 %). Poor results were attributed to centralized pain disorders. There were no incidents of serious or permanent sequelae.

Intravenous, intra-arterial, and intraneural injection can occur using any approach to the paravertebral space [68]. In addition, infection, hematoma formation, or damage to the neural fascicle may occur from dry needling. The type of needle can also affect the incidence of sequelae. Short-beveled needles have been shown to cause less nerve damage than long-beveled needles [71].

Aspiration will not reveal the presence of an intrafascicular needle tip. Injectate can dissect back through an epineural injection to the contiguous pia mater [72, 73]. This is driven by the occurrence of severe sequelae (death, paraplegia, transverse myelitis) from injection of a long-acting formulation of procaine [74–76]. The diffuse tissue necrosis was attributed to the carrier solution [77]. For this reason, the use of fluoroscopy and injection of low-osmolarity iodinated contrast to confirm proper needle placement are recommended when performing paravertebral blockade.

When using a continuous technique, there is always a risk of shearing the catheter if it is withdrawn back through the needle. Predictably, there will almost always be some pain at the site of needle insertion. Infection and hematoma are also possible risks. Monoplatythela (unilateral flat nipple) may occur with a successful block [78].

Other potential complications involve interpleural or intrapulmonary injections. If the tip of the needle is in the interpleural space, aspiration should reveal air. Injection of a small volume of radiocontrast under live fluoroscopy can quickly and easily detect an interpleural or intrapulmonary injection.

Prolonged anesthesia and motor block after inguinal hernia repair under general anesthesia with paravertebral blockade was observed in a patient with multiple sclerosis [79]. Abnormal uptake of local anesthetics into the spinal cord secondary to the presence of demyelination was proposed as the mechanism.

Contraindications to paravertebral block are infection at the site, patient refusal, and allergy to any of the solutions to be injected.

Intercostal nerve blocks are an alternative to thoracic epidural analgesia. Intercostal nerve blocks can be used in the acute setting for rib fractures, postthoracotomy pain, and trauma. Potential pitfalls of intercostal nerve blocks include difficulty in placement, high rate of local anesthetic absorption, and risk of pneumothorax. Ultrasound guidance has diminished these risks.

The anatomy of the intercostal nerves and spaces is depicted in Figs. 12.1, 12.2, 12.7, and 12.8. Intercostal nerves are derived from the spinal roots of the respective thoracic segments. They are composed of dorsal horn sensory afferent fibers, ventral horn motor efferent fibers, and postganglionic sympathetic nerves that join the nerve via the paravertebral gray rami communicantes. Thus, each intercostal nerve has autonomic and somatic sensory and motor functions. Soon after the sympathetic contribution occurs within the paravertebral space, the intercostal nerve divides into ventral and dorsal rami. The dorsal ramus provides sensory innervation to the posteromedial structures of the back (synovium, periosteum, fascia, muscles, and skin) and motor innervation to the erector spinae muscles. The ventral ramus travels between the ribs. It is protected within the subcostal groove by the rib and two layers of intercostal muscle.

Each intercostal nerve (ventral ramus) is associated with a vein and artery. The intercostal vein is derived from the confluence of venules along the thoracic cage and empties into the azygous vein on the right and the hemiazygous vein on the left. The most cephalad intercostal veins join and empty into the respective brachiocephalic veins bilaterally. The intercostal arteries are derived directly from the aorta.

The neurovascular structures are always superficial to the parietal pleura and thin aponeurotic-areolar tissue called the intercostalis intimus muscle. The aponeurotic-areolar tissue has muscle fibers embedded within its substance, and despite its name, its classification as a true muscle is a matter of debate among anatomists. There is various cutaneous branching of the ventral rami. In general, there are anterior and lateral branches, which divide and innervate skin and intercostal muscles of an individual segment along with variable collateral innervation of the adjacent segments. Because of this collateral innervation, it is necessary to block a level above and below the desired level. Because there is minimal adhesion of the aponeurosis to the parietal pleura, and the intercostalis intimus muscle is a rather flimsy structure, cephalad and caudad spread of injected solution to the adjacent intercostal spaces is not impeded (Fig. 12.7 and 12.8). It is important to keep in mind that the intercostal and paravertebral spaces are contiguous at all levels. Spread of local anesthetic to the paravertebral space produces unilateral segmental sympathetic blockade.

Intercostal neural blockade can be achieved intermittently or continuously in one or several segments depending on the technique used. Careful attention to technique decreases the rate of complication. Percutaneous injection of 2–5 mL of local anesthetic in at least three adjacent levels will ensure anesthesia/analgesia in the distribution of the middle intercostal nerve because of collateral innervation. Although relief is temporary, this technique is very effective in alleviating somatic pain in the chest wall and abdominal wall. Prolonged blockade requires either multiple reinsertions with the attendant risk of pneumothorax, placement of a catheter for bolus dosing or continuous infusion [80], injection with a neurolytic agent [81], or cryoablation [82].

Another important risk to keep in mind is local anesthetic toxicity. Blood levels of local anesthetic after intercostal blockade and interpleural analgesia are significantly greater than after any other frequently performed regional anesthetic techniques. Tucker et al. performed epidural, caudal, intercostal, brachial plexus, and sciatic/femoral nerve blocks with a single injection of mepivacaine 500 mg (1 % and 2 % solutions) with and without epinephrine [83]. When measuring arterial plasma levels, the highest levels were found after intercostal nerve blocks without epinephrine (5–10 μg/mL). When epinephrine was added to the solution (1:200,000 concentration), the plasma level decreased to 2–5 μg/mL. Epinephrine should be uniformly added to local anesthetic for performance of intercostal nerve block to minimize the potential for systemic toxicity.

The rib at the level to be blocked is by palpation and traced posterior to the posterior angulation of the affected rib. A linear high-frequency ultrasound transducer is then placed in the longitudinal plane with the superior aspect of the ultrasound transducer rotated about 15° laterally over the affected rib at the posterior angulation of the ribs. The rib can be seen as a hyperechoic curvilinear line with an acoustic shadow underneath it. The three layers of intercostal muscle, the external, internal, and innermost, are identified in the intercostal space between adjacent ribs. Color Doppler helps identify beneath the adjacent intercostal artery and vein. This space between adjacent ribs provides an excellent acoustic window, which allows easy identification of the intercostal space and the pleura beneath it. The depth of the pleura is noted. Usually, both the rib inferior to the targeted rib and the targeted rib can be visualized in the same window. 22-gauge echogenic needle is advanced from the inferior border of the ultrasound transducer using an in-plane approach with the trajectory being adjusted under real-time ultrasound guidance until the needle tip is resting in the internal layer of the intercostal muscle. At that point, after careful aspiration, a small amount of solution is injected under real-time ultrasound imaging to utilize hydrodissection to reconfirm the position of the needle tip. Once the position of the needle tip is reconfirmed, the needle is carefully advanced into the innermost layer of the intercostal muscle just short of the previously identified depth of the pleura. After careful aspiration, a small amount of solution is again injected to aid in identification of the position of the needle tip with attention paid to the relative location of the bright hyperechoic pleural line. After careful aspiration, the remainder of the solution is slowly injected. There should be minimal resistance to injection. The needle is then removed and a sterile dressing is applied at the injection site [88].

The most common complications of intercostal nerve block are associated with the aberrant needle placement (pneumothorax, hemothorax, hemoptysis, hematoma, intravascular injection, neuritis, subarachnoid block, failed block) or problems associated with the injectate (allergic reaction, toxic reaction, epinephrine reaction, tissue necrosis, respiratory insufficiency).

The actual incidence of pneumothorax secondary to intercostal nerve block is quite small. A large, retrospective study reporting 50,097 intercostal nerve blocks in 4333 patients undergoing surgery or therapeutic nerve blocks revealed only four clinically significant pneumothoraces (0.092 %) and no other significant complications [89]. The technique for intercostal neural blockade was similar to the posterior approach described by Nunn and Slavin [84]. There was some minor discomfort at the injection sites in 5 % of patients. A prospective study by the same authors in 200 consecutive patients undergoing intercostal nerve block compared pre- and postinjection films to evaluate for pneumothorax [90]. There were only four pneumothoraces in a total of 2610 needle punctures, of which three pneumothoraces were attributed to the actual surgical procedure itself and not performance of the blocks. In the largest retrospective study with more than 100,000 needle punctures, Moore reported an incidence of pneumothorax of 0.073 % without any other serious complications [90]. It is important to note that residents still in training performed most of these blocks [91].

There are sporadic case reports of other types of complications. Hematoma has occurred in a heparinized patient [92]. Bilateral intercostal nerve blocks have resulted in postoperative respiratory failure in patients with preoperative pulmonary compromise [93, 94]. Motor blockade and the loss of accessory respiratory muscle function were the hypothesized etiologic mechanisms. In a study looking at the efficacy of continuous epidural versus intercostal analgesia, one intercostal catheter led to rib osteomyelitis which had to be treated surgically [80]. Local anesthetic toxicity can occur due to higher absorption due to the close proximity of the intercostal vasculature.