The pattern of innervation of the triceps brachii by the radial nerve remains controversial and sometimes complex [16].

Several studies have suggested that the axillary nerve participates in the innervation of the triceps brachii, giving the branch of the long head of the triceps brachii [13, 14, 17].

Our study consistently found four distinct motor innervation branches of the radial nerve for the triceps brachii, one for the long head, one for the lateral head and two for the medial head. No involvement of the axillary nerve in triceps innervation was found. Our results are consistent with those of previous studies [8, 10,11,12].

However, we did not systematically find a second branch for the lateral head, which was present in only 25 of our 30 dissections, and which could come from the trunk of the radial nerve or, more often, from the inferior branch of the medial head. This branch, usually called the inferior branch of the lateral head, was found in a study of 18 of the 20 dissections [12]. We also found the existence of a second and, uniquely, a third branch for the lateral head arising from the inferior medial branch.

In other studies, the inferomedial branch could give rise to the posterior cutaneous sensory branch of the forearm or branches destined for the lateral head [6, 9, 11].

Taking the inferior branch of the medial head would most often require neurolysis to spare the branches destined for the lateral head arising from it and from the posterior cutaneous branch of the forearm. If the branch is taken proximally, there is a risk of transferring a sensory rather than a purely motor contingent.

In our study, the medial branch of the superomedial branch systematically accompanied the ulnar nerve. However, no communication was found between the two, in line with the results of other studies [18].

In more than a third of our dissections, we found a high division of the radial nerve branches into a common bundle distal to the branch of the long head, in agreement with the results of another study [12].

It should be noted that the radial nerve presents a fascicular aspect on the posterior surface of the humerus, and is easily dissected, which may contribute to the anatomical variations described.

Our study provides information on the constant arrangement of branches and potential variations. We also expressed in ratio the emergence of branches and structures improving inter-individual reproducibility.

Nevertheless, our study does not allow us to conclude on the functional axonal potential of the branches.

Furthermore, it would also be interesting to study the functional loss associated with denervation of one of the triceps brachialis heads, whether in its role as an extensor and stabiliser of the shoulder for the long head, or in its role as an extensor of the elbow for each of the muscle heads.