Introduction

The extent of the sympathectomy undertaken for upper limb hyperhidrosis is controversial and ranges from isolated resection of the second thoracic ganglion[4] (i.e. a second thoracic ganglionectomy) to excision of the sympathetic chain from the second to fourth thoracic ganglia [10, 15]. In recent years the second thoracic ganglion has come to be regarded as the key to sympathetic innervation of the upper extremity. Thoracoscopic sympathectomy is now acknowledged as the most effective surgical option in obtaining an enduring treatment for palmar hyperhidrosis [1, 13].

Accurate identification of the second thoracic ganglion at thoracoscopy is essential for successful upper limb sympathetic denervation. If missed or misinterpreted at thoracoscopy, this can lead to failure [2]. To date the localization of this ganglion at thoracoscopy remains unclear. Lin [12] reported that the second thoracic ganglion lies consistently between the second and third ribs. However, Lai et al. [11] reported that this ganglion lies in the second intercostal space near the upper border of the third rib, but in some cases the main body of this ganglion lies embedded, behind the pleura, in front of the third rib.

Since the second thoracic ganglion is pivotal to upper limb sympathectomy, it is crucial that it is correctly identified; misinterpretation is conceivable when the second ganglion is fused to the first thoracic or stellate ganglia. Misinterpretation of the second thoracic ganglion may also result when the first thoracic ganglion (which commonly fuses with the inferior cervical ganglion to form a stellate ganglion) is not fused and extends over the neck of the second rib to beyond its lower border. Appreciation of the location of the second thoracic and stellate ganglia will therefore be of surgical relevance. In this context the location of the second thoracic ganglion in the second intercostal space and the spectrum of additional neural connections associated with it will be of value to the thoracic surgeon.

Of significance is the relationship of the second thoracic ganglion to the intrathoracic ramus (between the first and second thoracic nerves) described by Kuntz [8] in 1927. This ramus was postulated to act as an alternative pathway conveying sympathetic fibres to the brachial plexus [8]. Variations associated with the second thoracic ganglion include a communicating branch from the ganglion or the sympathetic trunk traced directly into the intrathoracic ramus connecting the first and second thoracic nerves [5, 8]. Subsequent studies by Kirgis [6], Kirgis and Kuntz [7] and Kuntz [9] reported a similar intrathoracic ramus between the second and third intercostal nerves lying in front of the third rib. Kirgis and Kuntz [7] recorded an incidence of 54.5% of this intrathoracic ramus in their series of 88 dissections. This intrathoracic ramus can divide into two branches to join the second intercostal nerve at more than one place. Communicating branches have also been reported from the third intercostal nerve to the second thoracic ganglion and to the sympathetic trunk just below the second thoracic ganglion.

Kirgis [6] depicted a “recurrent ramus” from the third intercostal nerve which joined either the second thoracic ganglion or the sympathetic trunk just inferior to it. This “recurrent ramus” arose either from the grey communicating branch of the third intercostal nerve or from the nerve just distal to the point at which it was joined by the grey ramus. The “recurrent ramus” also joined the second intercostal nerve or its grey communicating branch. This “recurrent ramus” constitutes an additional neural connection through which fibres emerging from the sympathetic trunk via the grey communicating branch of the third intercostal nerve may reach the second intercostal nerve, from which they enter the brachial plexus via the intrathoracic ramus between the second intercostal nerve and the first thoracic ventral ramus.

This study aimed to record the incidence, location and associated additional neural connections of the second thoracic ganglion.

Materials and methods

Twenty cadavers were dissected bilaterally to define the incidence and location of the second thoracic ganglion and alternate neural pathways associated with this ganglion. Cadavers displaying cervical trauma, evidence of previous surgical exploration, or macroscopic pathology were excluded. The thoracic and abdominal cavities were eviscerated to expose their posterior walls. Parietal pleura was gently stripped bilaterally to reveal the sympathetic chain. Dissections were undertaken to display all neural connections from the second thoracic ganglion (including ganglionic and interganglionic segments) to adjacent ventral rami. A total of 40 sides were dissected and were available for evaluation. The incidence of ganglionic fusion (i.e. second to first thoracic ganglion, second to stellate ganglion and second to third thoracic ganglion) was recorded.

Results

Incidence of alternate neural connections

Of the 20 cadavers, 19 (95%) demonstrated alternate neural connections between the first and second thoracic ventral rami. In one cadaver, two alternate neural connections were noted bilaterally; one cadaver had noted two neural connections unilaterally. A spectrum of additional neural connections were demonstrated. These were categorized as either type A or type B using the intrathoracic ramus (nerve of Kuntz) between the second intercostal nerve and the ventral ramus of the first thoracic nerve as a basis on both right and left sides [14].

  • Type A: This group displayed demonstrable sympathetic connections to either stellate, second thoracic ganglia or the interganglionic portion of the sympathetic chain (47.5%, n=19/40)

  • Type B: This group included somatic connections between the second intercostal nerve to the first thoracic ventral ramus (27.5%, n=11/40; classical nerve of Kuntz) and the first intercostal nerve or its lateral cutaneous branch (17.5%, n=7/40).

Proximity of alternate neural connections to the second thoracic ganglion

The following mean distances laterally were made: 10.15 mm on the right (range 3.0–30.0), 7.18 mm on the left (2.3–12.3), overall 8.66 mm (2.3–30.0).

Second thoracic ganglion: topography and incidence of ganglionic fusion

In the 40 sides analysed the second thoracic ganglion was consistently located in the second intercostal space between first and second ribs in 37 dissections (92.5%); in 4 dissections (10%) the second thoracic ganglion was noted to be as follows: (a) fused first thoracic (separate inferior cervical ganglion) and second thoracic ganglia lying over the neck of the second rib and first intercostal space (2.5%; Fig. 1), (b) elongated, flattened second thoracic ganglion fused to first thoracic ganglion (separate inferior cervical ganglion) and extending to upper border of third rib (2.5%), and (c) fused directly to the stellate ganglion, lying in first and second intercostal spaces (5%; Fig. 2).

Fig. 1
figure 1

Right superior oblique view demonstrating fused first and second thoracic ganglia (Note the presence of a separate inferior cervical ganglion). C 8 , T 1 VR Eighth cervical, first thoracic ventral rami; ICG inferior cervical ganglion; ITR intrathoracic ramus; SA subclavian artery; SG stellate ganglion; T 1 G first thoracic ganglion; T 2 G second thoracic ganglion; T 2 ICN second intercostal nerve

Fig. 2
figure 2

Right superolateral view demonstrating fused stellate and second thoracic ganglia. Note the presence of the classical nerve of Kuntz. Abbreviations See Fig. 1

Discussion

The safety and ease of undertaking upper limb sympathectomy by the thoracoscopic route has culminated in large numbers of this procedure being undertaken. This has rekindled interest in the anatomy of the upper limb sympathetic pathways. The invariable successful clinical outcome to the excision of the second thoracic ganglion has suggested that this ganglion is crucial with respect to the upper limb sympathetic supply. It is therefore incumbent upon the surgeon to be facile with the morphological variations and the spectrum of neural connections associated with the second thoracic ganglion. There are few reports describing the surgical landmarks of the second thoracic ganglion; this is not surprising since at thoracoscopy the sympathetic chain and the second intercostal space (the invariable location of the second thoracic ganglion) are readily visualized.

The failure to appreciate the morphological anomalies associated with the second thoracic ganglion, whilst not consistently affecting upper limb sympathectomy, may predispose to the development of Horner’s syndrome. This complication may ensue in those instances of second thoracic and stellate or first thoracic ganglion fusion; in these situations intemperate surgical endeavour may predispose to injury of the proximal sympathetic chain leading to the development of Horner’s syndrome or inappropriate ganglion excision, resulting in an incomplete sympathectomy.

In this study the main body of the second thoracic ganglion was noted to be isolated and lay consistently (92.5%) in the second intercostal space, similar to the incidence reported by Chung et al. [3]. Ganglionic fusion was noted between the second thoracic ganglion and the stellate (5%) and first thoracic (5%) ganglia extending between the first intercostal space and upper border of the third rib. Thus, even in these uncommon states of ganglionic fusion, provided the surgical endeavour is just proximal to the superior aspect of the third, injury to the stellate or the first thoracic ganglion may be avoided; furthermore, an upper limb sympathectomy effect is guaranteed.

Several variations in the additional neural connections made by communicating branches between the second intercostal nerve and second thoracic ganglion are evident. Historically, these variations were considered responsible for an unsuccessful sympathectomy causing persistent hyperhidrosis after a seemingly accurate sympathectomy. Type A of additional neural connections was the most prevalent (47.5%) with the intrathoracic ramus displaying sympathetic connections to either stellate, second thoracic ganglia or the interganglionic portion of the sympathetic chain (Fig. 3). In the cadaveric dissections undertaken the additional neural connections were noted to be located at a mean distance of 8.66 mm (range 2.3–30.0) lateral to the sympathetic chain along the inferior border of the second rib. Our data support the findings of Chung et al. [3] who reported a mean distance of 7.3 mm (range 10–15) in their series.

Fig. 3
figure 3

Right superolateral view demonstrating the intrathoracic ramus with its neural connections (type A). Wire demonstrates rami connumicantes to second intercostal nerve. Abbreviations See Fig. 1

Given the sevenfold optical magnification afforded by thoracoscopy it is surprising that additional neural connections are not readily encountered when the dissection is confined to the second thoracic ganglion. The operating surgeon needs to bear this in mind at thoracoscopy and should resect parietal pleura over the second rib transversely for at least 3 cm laterally in order to transect an under appreciated intrathoracic ramus; this would obviate the development of recurrent symptoms.

Whilst abnormalities in the location of the second thoracic ganglion are uncommon, it is the fusion of this ganglion with the stellate and first thoracic ganglion that may predispose to complications such as Horner’s syndrome and to the inappropriate ganglion being surgically removed. A spectrum of alternate neural pathways issue from the second thoracic ganglion. These, although not readily evident at thoracoscopy, may be addressed by transecting the pleura laterally along the second rib.