The concept of the axillary "sheath" has been a central tenet of brachial plexus regional anesthesia for many years. Recent investigations have cast doubt on its nature and existence. This study further examines the issue.


Computerized axial tomographic dye studies were performed using continuous catheter systems for the sciatic nerve and the brachial plexus. The resultant images were compared and contrasted.


The images of the two catheter systems were the same, with the exception that one was of the upper extremity and the other was of the lower extremity.


The sciatic nerve is not surrounded or enveloped by a "sheath"--it lies in the tissue plane between rigid anatomical structures. Similarly, the brachial plexus lies in the tissue plane between the rigid anatomy of the chest wall, scapula, humerus, and pectoral fascia. This finding is inconsistent with the concept of the axillary sheath.

THE axillary or brachial plexus “sheath” has been a fundamental tenet of regional anesthesia applied anatomy for many years.1–4More recently, however, some doubt has been cast on the precise nature and existence of the sheath.5–8In the context of this controversy, we sought to find an alternative way to try to further the debate.

Using computerized axial tomography dye studies (CTDSs), we compared and contrasted the images of two peripheral nerve catheter systems: brachial plexus and sciatic nerve.

With the approval of the Nelson-Marlborough Ethics Committee (Nelson, New Zealand) and written informed consent, three patients with functioning brachial plexus catheters and two patients with functioning sciatic nerve catheters were enrolled in the study.

The three patients with brachial plexus catheters had these inserted using the “bent needle” technique of supraclavicular block,9and all had undergone major wrist, elbow, or forearm surgery during regional anesthesia, followed by regional analgesia using their catheters. The catheters were inserted to a depth of 10–11 cm from skin to catheter tip. The CTDS was performed on the second postoperative day. For the purposes of the study, a functioning brachial plexus catheter block was defined as absence of cold sensation in the territories of the ulnar, median, radial, medial and lateral antebrachial cutaneous, and axillary nerves 1 h before CTDS.

The two patients with sciatic nerve catheters had these inserted in similar fashion to the technique described by Sutherland.10Both had undergone major ankle surgery using combined general anesthesia and regional nerve blocks. Postoperative analgesia was provided via  their sciatic nerve catheters. The catheters were inserted to a depth of 12–15 cm from skin to catheter tip. The CTDS was performed on the second postoperative day. For the purposes of the study, a functioning sciatic nerve catheter was defined as absence of cold sensation in the territories of the common peroneal and tibial nerves 1 h before CTDS.

The position of the catheter tips was determined by a radiologist (C.L.) by injecting 1 ml of 50% diluted 300 mg/ml Omnipaque (iohexol; Amersham Health Limited, Auckland, New Zealand) dye down the respective catheters and then performing a preliminary computerized tomography scan. Then, 20 ml of 50% diluted Omnipaque dye was injected down each catheter, and the patients were rescanned at a slice thickness of 3 mm and a pitch of 2. The brachial catheter CTDS was performed from the top of the clavicle to 30 mm below the glenoid fossa, and the sciatic catheter CTDS was performed from the top of the sacroiliac joints to 20 mm below the lesser trochanter.

The images were then visually compared for similarities and differences. Particular attention was paid to the direction of distribution of the dye from the catheter tip, the contours of the dye-enhanced neural tissues, and the nature of the anatomical structures immediately adjacent to the nerves. Distortion or displacement of the tissues surrounding the nerves was noted.

The brachial plexus catheter tips lay just inferolateral to the coracoid process of the scapula, whereas the sciatic nerve catheter tips lay between the tip of the ischial tuberosity and the femur.

The following observations relate to figures 1–4. Both systems showed substantial anterograde and retrograde flow from the catheter tip, with clear limits of outward spread determined by solid anatomical structures, and clear lines demarcating the limits of spread (figs. 1–4). The pattern of distribution of dye within the neural tissues was similar between the two systems, both in the region of the catheter tip and at the extremes of dye distribution. There was neither distortion nor displacement of the anatomical structures surrounding the nerves (figs. 1B, 2B, 3, and 4). Indeed, frequently the line of dye conformed to the shape of the surrounding anatomy (figs. 2B, 3, and 4). The major difference between the two systems lay beyond the limits of spread of the dye, where the anatomy was clearly different, i.e. , upper extremity versus  lower extremity (figs. 1–4).

We believe that the similarities between the two different sets of CTDSs are not coincidental and that there are sound anatomical reasons to explain this.

It is generally agreed that there is no “sheath” surrounding or containing the sciatic nerve.11Rather, the sciatic nerve lies in a “tissue plane” closely surrounded by the gluteal muscles, the ischium of the pelvis, and the femur. A tissue plane is a potential space of embryologic origin that separates muscular and/or visceral compartments and that provides space for transmission of arteries, veins, lymphatics, and nerves between these compartments. Tissue planes are fundamental to surgical dissection technique,12but perhaps they are not sufficiently emphasized in anesthetic practice, given the number of publications in the literature that describe them.5,6,13–16 

Similar to the sciatic nerve then, the brachial plexus lies in a tissue plane closely surrounded by the clavicle, scapula, chest wall, and humerus.

There are several reasons why the tissue plane is important anatomy to understand. First, there may be minimal space for soft tissue expansion at any one point,17and therefore, flow must occur along the tissue plane, according to resistances encountered along the way.18As observed in the images from this study, one of the paths of lesser resistance within the tissue plane is along the line of the nerve, and in both directions from the point of injection. The dynamics of spread have the potential to be quite different18from those one might expect from a simple tubular structure such as a sheath. A series of plain x-ray film dye studies3,19,20have shown containment of solution, but plain x-ray films do not adequately define nonbony structures, and they lack the three-dimensional perspective of computerized tomography scanning. Accordingly, it is possible to misinterpret the images from these studies. Second, compartmentalization may occur because the layers of connective tissue within the tissue plane are not homogeneous, do not necessarily interconnect, and may hinder or prevent diffusion.8Therefore, injection at one point does not guarantee spread elsewhere. Third, it is said that “supraclavicular” techniques are more effective than “axillary” techniques because the nerves are closer together. This is a less than convincing argument because the nerves are not far apart at either level. At the supraclavicular level, however, the elements of the plexus—trunks, divisions, and cords—interlace and interlink, and hence the connective tissues containing these nerves interconnect, a feature observed when they are dissected.8This interconnection would allow for a more even spread of injected solution, a phenomenon that is observed clinically.21This is distinctly different at the axillary level, where the terminal nerves do not interconnect and the connective tissues surrounding them create distinct compartments for each.16Fourth, side effect profiles may be explained by the interconnection, via  tissue planes, of the anatomical compartments across which the nerves travel. Hence, an injected solution can spread to unwanted places with unwanted effects, e.g. , phrenic nerve block and interscalene22or subclavian perivascular block.23 

In conclusion, this study suggests that there may not be an enveloping and encasing soft tissue structure, or “sheath,” around the brachial plexus. The brachial plexus instead lies in the tissue plane between the rigid anatomical structures that form the boundaries of the anatomical axilla.17This rigid anatomy and the tissue plane within its borders combine to contain and direct the flow of an injected solution. The concept of the brachial plexus “sheath” seems to describe the anatomy inaccurately.

Reding M: Nouvelle méthode d'anesthésie régionale du membre supérieur. La presse médicale 1921; 30:294–6
Burnham PJ: Regional block of the great nerves of the upper arm. Anesthesiology 1958; 19:281–4
Winnie AP, Collins VJ: The subclavian perivascular technique of brachial plexus anesthesia. Anesthesiology 1964; 25:353–63
Davies DV: Gray's Anatomy, 34th edition. London, Longmans, Green, 1967, p 605
Longmans, Green
Thompson GE, Rorie DK: Functional anatomy of the brachial plexus sheaths. Anesthesiology 1983; 59:117–22
Partridge BL, Katz J, Benirschke K: Functional anatomy of the brachial plexus sheath: Implications for anesthesia. Anesthesiology 1987; 66:743–7
Hogan QH, Erickson SJ: MR imaging of the stellate ganglion: Normal appearance. AJR Am J Roentgenol 1992; 158:655–9
Cornish PB, Greenfield LJ: Brachial plexus anatomy. Reg Anesth 1997; 22:106–7
Cornish PB: Supraclavicular regional anaesthesia revisited: The bent needle technique. Anaesth Intensive Care 2000; 28:676–9
Sutherland ID: Continuous sciatic nerve infusion: Expanded case report describing a new approach. Reg Anesth Pain Med 1998; 23:496–501
Davies DV: Gray's Anatomy, 34th edition. London, Longmans, Green, 1967, p 1227
Longmans, Green
Carlson GW: Surgical anatomy of the neck. Surg Clin North Am 1993; 73:837–52
Vester-Andersen T, Broby-Johansen U, Bro-Rasmussen F: Perivascular axillary block VI: The distribution of gelatine solution injected into the axillary neurovascular sheath of cadavers. Acta Anaesthesiol Scand 1986; 30:18–22
Pashchuk AIu, Shtutin AA: Topographical and anatomic substantiation of sheath block of the brachial plexus [in Russian]. Ortop Travmatol Protez 1990; 5:13–7
Pippa P, Rucci FS: Preferential channelling of anaesthetic solution injected within the perivascular axillary sheath. Eur J Anaesthesiol 1994; 11:391–6
Klaastad O, Smedby O, Thompson GE, Tillung T, Hol PK, Rotnes JS, Brodal P, Breivik H, Hetland KR, Fosse ET: Distribution of local anesthetic in axillary brachial plexus block: A clinical and magnetic resonance imaging study. Anesthesiology 2002; 96:1315–24
Cornish PB, Leaper CJ: The axillary tunnel: Redefining the limits of spread for brachial plexus blockade (abstract). Reg Anesth Pain Med 2004; 29:24
Cornish PB, Leaper CJ, Hahn JL: Evaluation of spread of a bolus injection administered via  a brachial catheter using contrast-enhanced computerised axial tomography (CECT) scanning (abstract). Reg Anesth Pain Med 2004; 29:25
Winnie AP, Radonjik R, Akkineni SR, Durrani Z: Factors influencing distribution of local anaesthetic into the brachial plexus sheath. Anesth Analg 1979; 58:225–34
Rodriguez J, Barcena M, Alvarez J: Restricted infraclavicular distribution of the local anesthetic solution after infraclavicular brachial plexus block. Reg Anaesth Pain Med 2003; 28:33–6
Lanz EL, Theiss D, Jankovic D: The extent of blockade following various techniques of brachial plexus block. Anesth Analg 1983; 62:55–8
Urmey WF, Talts KH, Sharrock NE: One hundred per cent incidence of hemidiaphragmatic paresis associated with interscalene brachial plexus anesthesia as diagnosed by ultrasonography. Anesth Analg 1991; 72:498–503
Knoblanche GE: The incidence and aetiology of phrenic nerve blockade associated with supraclavicular brachial plexus block. Anaesth Intensive Care 1979; 7:346–9