Anatomical basis of erector spinae plane block: a dissection and histotopographic pilot study

Erector spinae plane (ESP) block is an interfascial blockade, first described by Forero in 2016 [1]. It consists of an injection of the local anesthetic in a fascial plane placed between erector spinae muscles and the tip of the transverse vertebral process [1]. The anesthetic spreads over the fascial plane both in the cranial and caudal direction, also diffusing anteriorly and laterally at several levels by one dermatome per 3.4 mL of injected liquid [2]. It provides analgesia in a wide range of different clinical scenarios [3, 4].

However, the effectiveness of the ESP block is not entirely reliable and reproducible [3], and some questions still remain unaddressed. The ongoing debate is mainly focused on the real spread of anesthetics, as by anatomical dissections, radiological imaging, and paravertebral endoscopy, some authors described conflicting evidence on the issue. Sometimes, no spreading to the anterior paravertebral space was reported at all both in humans and pigs [5‐8], or on the contrary, the spread to the neural foramina [1, 9, 10], to the anterior paravertebral space and intercostal nerves [1, 9‐13], or the epidural space was mentioned [9, 13]. All the anatomical studies reported a wide dye spread on the dorsal rami [1, 5‐7, 9‐14].

On these bases, the present study was aimed to investigate the spread of the dye to the anterior paravertebral space and its pathway of diffusion, by both an anatomical morphometric and histologic study on corpses.

Two fresh-frozen corpses within the body donation program “Donation to Science” of Padua University [15] were used for the present study. All the procedures performed in this study involved human bodies from the Veneto Region Reference Center for the preservation and use of gifted bodies (Deliberation of the Regional Council of the Veneto Region No. 245, Mar 8th, 2019; No. 389897), in accordance with the national laws, and the ethical standards of the regional/national research committees, as well as with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. Written informed consent was provided to join the body donation program.

The donors were a 78 year old male affected by atrial fibrillation (Corpse A), and a 93 year old female affected by mild hypertension (Corpse B), both died of heart disease. The sorted donors had no significant past clinical condition affecting the present study. In particular, they were not long-term bedridden, which would have potentially led to changes in the trophism and integrity of the posterior muscle planes due to the prolonged supine position.

The experimental procedure and dissection were compliant with a dedicated internal protocol implemented to standardize the operative steps.

Four sides from two corpses were injected according to an ESP block and subsequently dissected. The cadaveric section did not reveal the significant presence of the dye in the subcutaneous tissue, except for minor leakage along the injection pathway. Posteriorly, there was extensive dye spread observed in planes both superficial and deep to the erector spinae muscles group and the deepest muscles of the posterior wall (multifidus thoracis, rotatores thoracis breves and longi, levatores costarum breves and longi, and thoracic intertrasversarii). The craniocaudal spread of the dye was over many segments and differed across the muscular planes, as much as for left or right sides. Macroscopically, once removed all the dorsal muscles as mentioned above, the dye spread ranged on the left side from T2/3–T9/10 and T2/3–T8/9, on the right side from T2/3–T9/10 and T2/3–T10/11 for corpse A and B, respectively (Figs. 6, 7). Overall, a significant lateral spread was documented as reaching the maximal extension (9–10 cm) at the costal angle of T5/6–T8/9 level to progressively narrow at the contiguous levels  (Figs. 6, 7).

On the ventral surface of the chest, the dye was detected at the intervertebral foramen on the left side from the level T3/4–T9/10 and T2/3–T8/9 for corpse A and B, respectively, showing an extension comparable as for the dorsal surface  (Figs. 5, 7). On the right side, the dye spreading was narrower than on the dorsal surface, ranging from the level T3/4–T9/10 and T3/4–T8/9 for corpse A and B, respectively (Fig. 5).

Microscopically, the presence of dye on the left side levels resembled its macroscopic distribution for both corpses. However, while for the corpse A, the staining generally involved only the adipose tissue at a distance from the vascular and nervous structures, being noted near the adventitia of the intercostal vessels or nerve only at the levels T5/6 (lateral sample) and T8/9 (whole specimen); for corpse B, the dye near the adventitia of the vessels was almost ubiquitous for the positive samples, with evidence for staining next to the intercostal nerve at T5/6 (intermediate sample) level, and the only adipose tissue involvement for the intermediate and lateral samples of T3/4 level (Figs. 4, 5) . On the other hand, the presence of dye on the right-side levels appeared to be narrower than its macroscopic distribution for both corpses. Notably, while for the corpse A, the staining involved only the adipose tissue at a distance from the vascular and nervous structures in positive samples, the same was true for the corpse B except for the T4/5 level (medial sample), for which the dye was nearby the adventitia of the vessels, even if is still far away from the intercostal nerve (Figs. 4, 5). Overall, the dye diffused ventrally to the intercostal spaces (2–3 spaces on the right and 5–6 spaces on the left) by following the blood vessels coupled to the dorsal nerve passing through the costotransverse foramen (Fig. 5).

The present study suggested the pattern of anterior diffusion of the local anesthetic in ESP block by gross anatomical dissection with histological examinations.

In ESP block, the mechanism of pharmacodynamics action was hypothesized to be in part due to the involvement of the dorsal rami of spinal nerves by superficial and deep spreading among muscles, and in part due to the penetration of the local anesthetic anteriorly to reach the intercostal nerves, both coherently to the distribution of cutaneous dermatomes [1, 6, 9]. Notwithstanding, even in the original description of ESP block, there was no definitive confirmation and characterization of dye spreading to involve the intercostal nerves [1], although if the paravertebral diffusion or their sporadic involvement was reported in some studies [1, 9‐13], while others failed to prove it as much in humans as in pigs [5‐8]. This contradictory evidence on the issue matched with the practical observation that the effectiveness of ESP block is not entirely reliable and reproducible in daily clinical practice [16].

In the present study, the number and location of the intercostal spaces showing macroscopic signs of anterior dye diffusion were approximately symmetric between both sides from T2/3–T9/10, substantially centered on the injection site according to previous data [1, 5, 7, 9‐12].

Histologically, dye staining of the medial soft tissue forming the intercostal space on samples retrieved by anatomical microdissection was here demonstrated. A discrepancy was detected between the macroscopically detected coloring of the intercostal spaces, and the histological confirmation on the same sample, more evident on the right side. This finding could be explained by the possible discoloration of weakly dye-stained samples.

The overall evidence provides a further basis for the analgesic effect reported in daily clinical practice, at the same time, confirming the presence of an anatomical pathway for the anterior spread of local anesthetic [1, 17‐19]. A possible pathway through the costotransverse foramen and its costotransverse ligaments was suggested [1, 9, 11, 12], as demonstrated by recent evidence [10] supporting the introduction of the costotransverse foramen block technique [20]. The costotransverse foramen (Fig. 3) is an anatomical space bordered cranially by the transverse process, caudally by the underlying rib, laterally by the medial-free edge of the superior costotransverse ligament, and medially by the lamina and facet joints [1]. The costotransverse ligament consists of both the superior costotransverse ligament and the lateral costotransverse ligament [12, 21, 22]. The medial-free edge of the superior costotransverse ligament constitutes the lateral border of the costotransverse foramen and the lateral portion of the superior costotransverse ligament joins with the internal intercostal membrane [22, 23]. The superior costotransverse ligament has anterior and posterior layers and presents gaps in its structure [21, 24, 25] (Fig. 3). At each thoracic segmental level, the dorsal ramus of the spinal nerve and the accompanying blood vessels pass through the costotransverse foramen between the anterior and posterior layers of the superior costotransverse ligament [12, 21, 23, 24] (Fig. 3). However, the hypothesis of an anterior diffusion path of the dye through the costotransverse foramen has been somehow criticized by some authors, as the complex attachment of the erector spinae muscle group to the transverse process and the costotransverse ligaments may prevent the anterior anesthetic diffusion [5, 7, 10‐12].

The ventral pathway of the dye through the costotransverse foramen to reach the intercostal nerves is suggested by the dye spread in the intercostal soft tissue through the direct infiltration demonstrated at histotopographic examination. In particular, four aspects converge on this consideration. First, the macroscopic involvement of the dorsal rami of spinal nerves was detected. It is known that they pass through the costotransverse foramen. Second, the medial portion of the intercostal soft tissue is placed anteriorly to the costotransverse foramen.

Third, whenever a harvested intercostal soft tissue showed dye staining at histology, the dye was detected steadily as infiltrating its medial portion, with a variable degree of further extension of the dye laterally. Fourth, the dye was demonstrated near the adventitia of the intercostal vessels, which are initially coupled with the dorsal rami of the spinal nerves and respective blood vessels passing through the costotransverse foramen as mentioned [10, 12] (Fig. 3). As a consequence, the anterior histotopographic diffusion of dye through the erector spinae plane block resembles a paravertebral block [26, 27].

Overall, the variability of shapes and dimension of the dye spreading reported in the literature for corpses may be referred to some factors, as follows. The costotransverse ligaments showed different anatomical configuration and relationships along the vertebral column [6, 28]. Indeed, the posterior costotransverse ligament is absent/rudimentary between T1 and T6, very well developed at the T7–T10, and undetectable at the T11–T12, which could negatively influence the diffusion within the paravertebral spaces at these lower levels for dye injection at the T5 level [9, 11, 12, 21, 22]. However, other authors showed as dye injection at T10 level only shifted the area of tissue spread, without limiting it downward [7]. Also, the volume and viscosity of injected dye may play a role by generating a different pressure-driven mechanism, which differs from study to study [11]. Even the time awaited for the distribution of the injected dye before dissection may have played a role in its different distribution. In the present paper, the ESP block was performed as per daily clinical practice, for both procedures and times. In literature, the time elapsed between the injection of dye and the cadaveric dissection was widely ranging up to 2 h [5, 6, 9, 10].

There are limitations in using a cadaveric model to represent the spread of local anesthetic during real-life clinical practice. First, postmortem changes in the integrity and permeability of tissues might influence dye dispersion. Second, in the living, there may be a more extended spread of local anesthetic related to intrathoracic pressure changes during breathing, or muscle tone and muscular contraction. Third, the different placement of the patient may has a role. Fourth, the assumption that observing the dye spread one can infer the clinical effect of local anesthetic has to be demonstrated. Furthermore, in this pilot study, the procedure was performed just on two corpses and variability of observations has to be considered.

In conclusion, the present study performing an ESP block in corpses confirmed the extensive anterior spread of the dye involving the intercostal nerves, both at the macroscopic and microscopic examination. The histotopographic evidence suggests that the dye’s anterior diffusion takes place through the costotransverse foramen and seems to follow the vessels and dorsal rami of spinal nerves. The dye diffusion pattern resembles a paravertebral block, hinting at a complex action mechanism to explain the ESP block clinical effect.

Although these results do not make yet clear what the definitive answer is regarding the mechanism of ESP block, it helps to understand what its path of diffusion among the anatomical planes is and which structures interest it more anteriorly. It will be of interest to investigate the anatomo-clinical causes underlying the variable effect of the ESP block observed in the daily practice, which could be related at least in part to individual morphometric differences of the structures involved by the passage of the anesthetic.