Variations in the formation of supraclavicular brachial plexus among Kenyans

Author Information

B.O. Olabu, BSc (Anat), J.A. Ogeng’o, BSc (Anat), MBChB, PhD, K.O. Awori, MBChB, MMed (Surg),
Dip. (SICOT), FCS (Orth), ECSA and H. Saidi, BSc (Anat), MBChB, MMed (Surg), FCS (ECSA)

Corresponding author:

B.O. Olabu, Department of Human Anatomy, University of Nairobi, P.O. Box 30197-00100, Nairobi, Kenya,


OBJECTIVE: To describe the pattern and prevalence of variations that occur in the supraclavicular part of the brachial plexus in a Kenyan population.

STUDY DESIGN: Descriptive cross-sectional study.

MATERIALS AND METHODS: Ninety-four brachial plexuses from forty-seven formalin fixed cadavers were displayed by gross dissection.

RESULTS: The presence of at least one variation from the classical anatomy was observed in 73 (77.7%) of the 94 plexuses. The roots and trunks were involved in 32 (34%) of the plexuses . Pre- and post-fixed roots were present in 23 (24.7%) and 3 (3.2%), respectively. The presence of four trunks, and trunks passing between the scalene medius and posterior were also noted. The long thoracic nerve was variant in 51 (54.3%) of the plexuses. Unusual relations of the phrenic nerve to scalene muscles and the subclavian vein were encountered.

CONCLUSION: The presence of four trunks and an accessory phrenic nerve passing through the subclavian vein are probably described and reported for the first time.  However, most of the variations of the BP among Kenyans are similar to those reported in the other populations.


The complex anatomical organization of the brachial plexus predisposes its nerves to a wide range of variations throughout their courses (1-3). Variations of the supraclavicular brachial plexus involving the roots, trunks and their branches differ in prevalence among various populations. These include prefixation (4-7) and postfixation (3,5,7). Variable relations of the plexus to the scalene muscles, absence or anomalous origin of some trunks and unusual communications have also been noted in some studies (3,8-11).

Apart from causing confusion in the assessment of upper limb nerve injuries, these anatomical variations can be responsible for failure of some brachial plexus nerve blocks or complications following these blocks (3,6,9). Literature on the variations of the supraclavicular part of the brachial plexus among Africans is scanty and lacking in Kenyans. The present study describes the variations of the brachial plexus observed in a select Kenyan population.

Materials and methods

Ninety four brachial plexuses from 47 formalin fixed cadavers from both males and females used for routine dissection by first year medical students in the Department of Human Anatomy, University of Nairobi were displayed by gross dissection to study the supraclavicular brachial plexus. The posterior triangle of the neck was exposed by making skin incisions over the sternocleidomastoid muscle and the clavicle meeting at the insertion of the former, then turning the skin flap laterally. The investing layer of deep fascia of the neck was removed, the prevertebral fascia cleared and the trunks of the brachial plexus exposed and traced proximally to their origins in the cervical and thoracic spinal nerve roots.

The origins and courses of the long thoracic, dorsal scapular, suprascapular, and the phrenic nerves were defined. The relationships of the brachial plexus and these nerves (in the supraclavicular part) to the scalene muscles were photographed and recorded using a digital camera (Sony Cybershot® P200, 7.2 megapixels).



Twenty three brachial plexuses (24.5%) were prefixed while three (3.2%) were postfixed (Figure 1). In one cadaver with bilateral cervical ribs, the roots on the left and right were prefixed.


Click to view table 1


The presence of three trunks was observed in all except one case from the left side of a body where there were four trunks. In this case C8 root failed to unite with T1 root to form the lower trunk. These two roots were separated by the scalenus medius muscle, with T1 root running anterior to the muscle (Figure 2).


The upper trunk was formed from C5 and C6 in 70 plexuses (74.5%). Twenty three plexuses (24.5%) had an additional contribution from C4 and were considered prefixed. In a single case of a postfixed plexus, the upper trunk was formed by the C5,6,7 roots while the C8 root formed the middle trunk and the lower trunk from T1,2 (Figure 1). The C7 root continued as the middle trunk in 91 cases (96.8%). In one subject with bilateral postfixation, the middle trunk formed from C7,8 (Figure 3). The lower trunk was formed by C8,T1 in 90 cases (95.7%) and T1,2 in plexuses, where the latter were postfixed.


Relation to scalene muscles

For the upper roots, both C5 and C6 nerve roots passed within the scalene gap in 75 plexuses (79.8%) and in front of scalenus anterior  in three. The two roots were separated by scalenus anterior in 26  plexuses, in which C5 penetrated the muscle in most cases. Both the middle and lower trunks were found within the scalene gap in 92 plexuses (97.9%). The middle trunk was between scalenus medius and posterior in 2 cases, either together with the lower trunk  (1 case) or with the C8 root (1 case, Figure 2).


Three modes of disposition of origin of the phrenic nerve were observed: (a) C3,4,5 roots in 61  (64.9%) plexuses; (b) C4,5 in 2 cases; and (c) C3,4 in 31 cases. Of the latter group (31), 29 (30.9%) had an accessory phrenic nerve . After its formation the phrenic nerve, in one case, lay lateral to the scalenus anterior (as opposed to anterior) and therefore it was not a content of the pyramidal space at the root of the neck. In the 29 plexuses (30.9%) the accessory phrenic nerve had three modes of presentation at the roots : (a) 20 cases from the nerve to subclavius muscle; (b) 6 cases upper trunk ; and (c) 3 cases C5 root. In relation to the subclavian vein, three courses were described. In 22 plexuses it passed anterior to the vessel and entered the thorax without joining the phrenic nerve within the neck.  In the second course it was behind the vein in 6 cases and here the accessory phrenic nerve joined the phrenic nerve at the root of the neck. A modification of this was a communication between the two nerves before their final union observed in one case (Figure 4).


The third route was observed in one case when an accessory phrenic nerve passed through the left subclavian vein (Figure 5). The long thoracic nerve was formed from C5,6,7 roots in 69.1%, C5,6 in 28.7% and C6,7 in 2.1%. All roots forming the nerve in either case passed through the scalenus medius in 41 cases.


The suprascapular nerve  originated from the upper trunk  then curved towards the suprascapular notch with no variation to this classical pattern in all cases. Two variations of origin of the dorsal scapular nerve were reported: (a) from C5 root in 91 cases (96.8%) and; (b) from C4,5 in 3 cases. The nerve classically passed through scalenus medius in 80.4% plexuses and within the scalene gap in 19.6%.


Observations of the present study have revealed an incidence of 27.7% in the variation of brachial plexus fixation, with prefixation being more common (24.5%) than postfixation (3.2%). These results are compared with those observed in various populations in Table 2.


The brachial plexus  is considered prefixed when there is a large C4 root and the contribution from T2 being generally absent and that of T1 reduced. A postfixed plexus has a reduced or absent C5 but more dominant T1 and T2 contributions (1,2). Apart from confusion in the assessment of upper limb nerve injury, a prefixation could possibly predispose the plexus to traction injuries involving the upper trunk or roots. A postfixed plexus on the other hand may increase the risk of injury to the T2 root by a cervical rib or any space occupying lesion like tumors of the apical lobe of the lung (1).

The usual three trunks were observed in 98.9% cases of the present study. Instances where C8 and T1 roots do not join but run as separate trunks, being separated by the scalenus medius muscle as observed in the present study are hitherto undescribed. Contribution of the C7 root to the upper trunk observed in one case confirms the findings of a previously reported case but unlike the present observation the middle trunk  was absent in the former case (12).

The finding that the middle trunk may be formed from C7 and C8 confirms the report of a previous study where a communicating branch from C8 (1 case) and lower trunk (1 case) joined the C7 root (3). Such a middle trunk containing C8 fibres could be indicative of the musculocutaneous nerve receiving C8 con