Histologic evidence of brachial plexus compression sites at the thoracic inlet and variations in formation of the lower trunk in cadavers

In thoracic “outlet” syndrome (TOS), pathologic evidence is well documented for vascular but not neurologic compression. We hypothesized that histologic evidence of compression would be identified at sites where the upper trunk was impacted by the anterior scalene muscle and the lower trunk by anatomic anomalies or the first rib. The purpose of this study was to investigate this hypothesis in human cadavers.


| INTRODUCTION
The concept that the brachial plexus can be compressed chronically as it passes from the neck into the arm seems clear enough, but the existence of chronic compression in the thoracic inlet, of this complex arrangement and re-arrangement of five cervical nerve roots, has been difficult to prove.As an example, how do we know the median can be compressed within the carpal tunnel?Reasoning that a person's parenthesis of the thumb, index and middle finger are relieved when the transverse carpal ligament is divided is circumstantial, subjective, evidence at best.Historically, (Amadio, 1992;Learmonth, 1993) the first patient to have this ligament divided had the diagnosis made at the Mayo Clinic, with the transverse carpal ligament being divided by surgeon James R. Learmonth, MD, a Neurosurgeon, also at the Mayo Clinic.Learmonth has been called the First Peripheral Nerve Surgeon (Dellon et al., 2000).While proof of median nerve compression can come from a histologic evaluation of the median nerve, this never occurs clinically because we do not excise the median nerve for pathologic evaluation.Yet, the situation can arise, as it did for Marie and Foix in 1913, when they evaluated, histologically, compression of the median nerve identified in an otherwise routine cadaver dissection, a subclinical presentation (Marie & Foix, 1913).This type of investigation, documenting subclinical nerve compressions, now has been reported for the median nerve at the wrist (Neary et al., 1975;Thomas & Fullerton, 1963), the ulnar nerve at the elbow (Neary et al., 1975), the lateral femoral cutaneous nerve (Jefferson & Eames, 1979), and the lower trunk of the brachial plexus related to a cervical rib (Tubbs et al., 2008), and in clinical entrapments in which the nerves were excised due to neuroma formation [tibial nerve in tarsal tunnel (Mackinnon, Dellon, & Daneshvar, 1984), radial sensory nerve in forearm (Mackinnon et al., 1986)] and patient death (ulnar nerve at elbow; Dellon & Mackinnon, 1988).These pioneering studies in humans foretold the same histologic findings as the experimental models of chronic nerve compression did in the rat (Mackinnon, Dellon, Hudson, & Hunter, 1984) and subhuman primate model (Mackinnon et al., 1985): myelin thinning in fascicular regions closest to the site of compression; presence of Renaut bodies which are loosely textured, whorled, cell-sparse structures found in the sub-perineurial space of peripheral nerves affected by compression (Jefferson et al., 1981); subperineurial edema; interfascicular epineurial thickening; and, in the more severe cases, loss of large myelinated fibers.
The hypothesis of this study was that histologic evidence of compression would be identified by analysis of sites at which the upper trunk was impacted by the anterior scalene muscle and the lower trunk impacted by anatomic anomalies or the first rib.
A second outcome measure of this study was to evaluate in cadavers an intra-operative observation (made by the senior author) that often the anatomic structure deep to the middle trunk (C7 nerve root) was not the lower trunk (C8 joined to T1 nerve roots), but rather was the C8 nerve root, and that when a muscle was deep to the middle trunk, it was the scalenus minimus (muscle of Albinus), which itself could compress the lower trunk, or C8, and would therefore require resection to complete a neurolysis of the brachial plexus.

| Surgical dissection
Forty-eight hour post-mortem cadavers (n = 25) were obtained from the Anatomy Board of the State of Maryland.These were unidentified bodies and approval from an IRB was not required.
A total of 50 brachial plexuses were examined.A 12-cm incision was made paralleling the clavicle at 1.5 cm superior to the clavicle.The clavicular head of the sternocleidomastoid muscle was excised.The superficial cervical fascia was opened.The omohyoid muscle was excised.The phrenic nerve was identified on the anterior surface of the anterior scalene muscle.At this point, the upper trunk of the brachial plexus and its contributing C5 and C6 nerve roots could be observed in relationship to the anterior scalene muscle.A suture was placed at this site.The anterior scalene was resected and the upper trunk and its divisions were inspected for signs of compression/adherence to the anterior scalene muscle.The C7 nerve root and the middle trunk could then be observed, as could the middle scalene muscle.The dorsal scapular artery, arising from the subclavian artery, could be evaluated with respect to compression of the middle trunk.The dissection was then carried medially and deeper until the lower trunk of the plexus could be identified.A careful medial dissection was done until the contributing C8 and T1 nerve roots could be identified, and the relationship to the formation of the lower trunk of the plexus and the border of the first rib was recorded as in Figure 1.The presence of the scalenus minimus was noted.Once the dissection of the brachial plexus was complete, nerve specimens at (1) the anterior scalene and upper trunk and (2) C8 and T1 nerve roots and posterior border of first rib, potential compression sites, were excised and stored in formalin for histology.

| Histochemical staining and imaging of brachial plexuses
Specimens were prepared for histology with toluidine blue staining for myelin in concordance with previous protocols performed by our research team (Mackinnon et al., 1985;Mackinnon, Dellon, Hudson, & Hunter, 1984).
Slides were viewed under light microscopy and qualitatively evaluated for signs of chronic compression such as intraneural fibrosis and myelin thinning (National Research Council (US) Steering Committee for the Workshop on Work-Related Musculoskeletal Injuries: The Research Base, 1999).

| Data collection and statistical analysis
The RStudio 4.1.2program was used to analyze data with descriptive statistics.

| Histology results
Histologic analysis demonstrated epineurial and perineurial fibrosis, myelin thinning, and Renaut bodies at the anterior scalene/upper trunk sites of potential compression (Figures 2 and 3), and at the C8 and T1 root/posterior first rib junction sites (Figures 4 and 5).In general, changes in myelin thinning were more apparent in the upper trunk sections than in the lower trunk sections.

| Anatomical relationships of brachial plexus components to compression sites
With regard to the middle trunk, a large dorsal scapular artery crossed C7 in 32% of the specimens and was related to a high arched subclavian artery (Table 1).
With regard to the lower trunk, 34% of the specimens had T1 join C8 quite medially so that a dissection would demonstrate a lower trunk directly over the first rib.In 36% of the specimens, T1 joined C8 quite laterally, so that a dissection would demonstrate C8 and not the lower trunk crossing the first rib.In the remaining 30%, both T1 and C8 crossed the first rib together.These results are illustrated in Figure 6.
The scalenus minimus was present in 18%, where it compressed the lower trunk or the C8 nerve root against the first rib.
Bilateral brachial plexus asymmetry was documented in 32% of cadavers.

| DISCUSSION
Among the reasons that brachial plexus compression remains controversial is the nomenclature change when physical therapists, Peet et al. (1956) described the exercises to stretch the anterior scalene muscle and strengthen the upper trapezius and the rhomboids (Craig & Knepper, 1937).Those exercises were designed to treat patients with scalenus anticus syndrome (Kirgis & Reed, 1947;Stowell, 1956), but that paper (Roos & Owens, 1966) introduced the name "thoracic outlet syndrome" (TOS).This led the way for thoracic surgeons to do transaxillary first rib resection (Urschel Jr et al., 1968;Wilbourn, 1999), and led neurologists to denounce the diagnosis (Carroll & Hurst, 1982;Sanders et al., 2008), as most "neurologic" cases of TOS could not be confirmed electrodiagnostically.Hand surgeons began to write that these neurologic symptoms were related to distal compression of the median nerve in the carpal tunnel (Howard et al., 2003), and not to compression of the brachial plexus in the thoracic outlet.This is an additional source of confusion, as the plexus is not compressed at the thoracic outlet but at the thoracic inlet anatomically (Dellon, 1993).Owens, 1966;Stowell, 1956).The subclavian artery and vein are clearly, anatomically, located such as to be compressed by the anterior scalene muscle at its insertion and the 1st rib at the thoracic inlet.Such vascular compression accounts for up to 5% of patients (Gilliatt et al., 1978).Fully 95% of patients have neurologic symptoms at the thoracic inlet (Gilliatt et al., 1978), where the brachial plexus travels between the clavicle and the 2nd rib, not the 1st rib.
Because electrodiagnostic testing can identify only compression of the lower trunk (abnormal EMG of both median and ulnarinnervated intrinsic muscles and decreased sensory amplitude/ conduction velocity in the little finger), this most rare presentation has been termed true neurologic TOS (Renaut, 1881;Sanders et al., 2008).
Based upon the histologic findings of the present study, it can now be confirmed histopathological findings consistent with chronic nerve compression exist at brachial plexus compression sites.These findings also confirm those of Tubbs et al. (Tubbs et al., 2008).The Renaut bodies (Renaut, 1881; Figure 5) identified at the biopsied sites of chronic compression in this study are identical to those identified in the lateral plantar nerves of rats that walk on wire cages (Ortman et al., 1983), in the human median and lateral femoral cutaneous nerve (Jefferson et al., 1981), in the suprascapular nerves of horses (Duncan et al., 1987), the tibial branch in the popliteal fossa of Beagles (Elcock et al., 2001), and in the trunk of elephants (Witter et al., 2007).These studies all conclude that the fibrous tissue that is the Renaut body occurs at sites of stretch/traction of a peripheral F I G U R E 6 Percentage of lower trunks formed with T1 joining C8 quite medially, centrally, and laterally.
F I G U R E 7 Renaut's original drawing showing intraneural hyaline retaining structure, 1881, in public domain (Duncan et al., 1987).
As a short aside, Bernard Siegfried Albinus is a little-known anatomist who was born in Frankfurt, Germany on February 24, 1697, and died in Leiden, Netherlands on September 9, 1770 (Bernard Siegfried albinus, 2022).At the time of his death, he was Chairman of Anatomy, Surgery and Medicine at the University of Leiden.He described both the scalenus minimus and the rhizorus muscle.He is remembered most for his 1747 book and its engravings (Figure 7) Tabulae sceleti et musculorum corporis humani ("Tables of the Skeleton and Muscles of the Human Body"; Bernard Siegfried albinus, 2022).
Along with the demonstration of subclinical, chronic compression at brachial plexus compression sites, we show that only around one-third of cadavers were observed to have the lower trunk pass over the anterior border of the first rib (Figure 6).Therefore, our findings suggest that this structure may actually be the C8 nerve root instead.Additionally, there may be around a one-third chance that a surgeon may find a different variation of lower trunk formation on the contralateral side.We hope that surgeons take this information into account for their preoperative planning, as well as a significant chance that a scalenus minimus may be compressing brachial plexus structures.
Our approach to plexus neurolysis for compression in the thoracic inlet is to make a 4 cm incision in the supraclavicular region, divide the platysma, spare the supraclavicular nerve, divide the sternocleidomastoid clavicular head, identify and preserve the phrenic nerve, resect a 2 cm segment of the anterior scalene muscle, and then neurolyse the C5 and C6 roots and branches of the upper trunk.We respect the medial scalene if it is involved in the scarring and subsequently neurolyse C7 and the lower trunk, which may require resection of the scalenus minimus and other fibrous bands.
Finally, if the long thoracic nerve is involved, we neurolyse this nerve at its entrapment site in the posterior/middle scalene muscles, which is posterior to the plexus.
To decompress brachial plexus structures in patients with neurologic TOS, current literature suggests that neither supraclavicular, transaxillary, or video-assisted thoracic surgery approaches demonstrate superiority over the other (Davoli et al., 2021).Though first rib resections with anterior scalenectomies have proven to be an effective method to treat neurologic TOS, a rib-sparing approach has been described in the literature and may reduce length of hospital stay and other 30-day complications (Jubbal et al., 2019;Maqbool et al., 2019).Recent literature by Ransom et al. show no difference in intermediate-and long-term outcomes in adolescents who underwent first rib resection and those who did not (Ransom et al., 2022).However, rib resection may be indicated in those with recurrent neurologic TOS symptoms (Annest et al., 2021).We would like to suggest that other structures besides the first rib may be sources of compression such as the scalenus minimus and that surgeons should examine for compression by that muscle before continuing with a first rib resection.
We personally do not resect the first rib unless there has been specific deformity in this structure.
Last, though our findings suggest that one-fifth to one-third of patients may have an anatomic variation predisposing them to neurologic TOS, previous studies have estimated the prevalence of neurologic TOS to range from 25 per year in a metropolitan area of 1,000,000 (Illig et al., 2021) to three to 80 in 1000 people for all forms of TOS (Gkikas et al., 2022).This syndrome is rarer than what our results suggest (Illig et al., 2021).Additionally, predisposing risk factors for TOS are known, as it affects young adults 20-50 years old and women 3-4 times more than men (Gkikas et al., 2022).The risk of TOS may be increased in those working occupations requiring prolonged elevation or repetitive use of the arms or with a history of trauma, masses, or other anatomical abnormalities in the thoracic inlet as well (Laulan et al., 2011).In this context, anatomic variations found in our studies in combination with risk factors may increase the chance of chronic compression at brachial plexus sites and subsequent symptoms of pain, weakness, and parenthesis seen in TOS (Laulan et al., 2011).
A limitation of this study is that histology was not obtained from the C7 nerve root/middle trunk in the location where the dorsal scapular artery can cause chronic nerve compression.Another limitation of this study is that the upper trunk of each plexus was not biopsied but just the ones in which a fibrous edge of the anterior scalene was in close proximity to the upper trunk.Additional specimens were not able to be obtained as well for further sites of comparison and with further histologic analysis with different staining techniques.Neither demographic information of cadavers nor their history of TOS symptoms were able to be obtained as well.Though we show a percentage of our cadavers with anatomical variations, we acknowledge that this may not reflect the general population due to our sample size and urge further studies to advance our understanding of neurologic TOS.

| CONCLUSIONS
We report the histologic evidence of compression in the thoracic inlet that correlate well with the clinical neck/shoulder (upper trunk) and "ulnar nerve-like" (C8-T1/lower trunk) symptoms of brachial plexus compression in the thoracic inlet, the so called "thoracic outlet syndrome".The pattern of anatomic variations described may help surgeons doing supraclavicular brachial plexus neurolysis.

F
I G U R E 1 Example of excised upper trunk (a) and lower trunk (b) from a cadaver.Sutures mark sites for biopsy.F I G U R E 2 Upper trunk histology.(a) Normal, not compressed, 40Â, (b) site of anterior scalene compression, showing interfascicular epineural and perineurial thickening, at 40Â, (c) normal, at 550Â, (d) site of anterior scalene compression showing myelin thinning, and decreased number of large, myelinated fibers, at 550Â.
This study helps us to understand some of the confusing aspects of what the literature currently, mistakenly, calls "thoracic outlet syndrome".First, we now understand that this clinical syndrome was called originally "scalenus anticus syndrome" and encompassed both vascular and nerve compression sites(Kirgis & Reed, 1947; Roos & F I G U R E 3 Upper trunk histology.(a) Normal, not compressed, 40Â, (b) site of anterior scalene compression, with arrow at a Renaut body, 25Â, and (c) higher power view of Renaut body with adjacent perineurial (P) thickening 250Â.F I G U R E 4 Lower trunk histology.(a) Normal, 40Â, (b) site of compression at 1st rib with Renaut bodies (R), 40Â, (c) another example at site of compression at first rib, with Renaut bodies (R), 40Â.

F
I G U R E 5 Lower trunk histology.(a) Normal, at 25Â, (b) and (c) higher power fascicular views, with normal myelination, 40Â and 150Â, (d), at site of first rib compression, arrow at a Renaut body, 40Â, (e) with higher views at a Renaut body, 100Â, and (f) even higher view, 250Â.T A B L E 1 Accessory muscle details. usually in locations associated with chronic nerve compression.From our study, these sites include (1) the upper trunk and the C5 & C6 nerve roots related to compression by the anterior scalene muscle and (2) the lower trunk and C8 & T1 nerve roots between the scalenus minimus and the first rib at the thoracic inlet.Of course, there can be other causes of brachial plexus compression in the thoracic outlet, such as direct trauma, tumor, or other congenital anomalies.