Every dislocation or fracture of the proximal humerus with or without concomitant dislocation should be carefully examined in order to detect clinical signs of nerve injuries. Clinical exam should be performed at admittance and after conservative and/or surgical treatment in order to recognize and document every kind of lesion. This aspect is fundamental also for forensic implications [27]. In Table 14.1, upper limb muscles and their innervations are reported; it is mandatory to examine every muscle to identify the real level of nerve lesion.

Nonetheless, in early phase after trauma, clinical testing could be difficult to interpret and of low value in detecting axonal nerve lesions in these fractures just after trauma. This happens because nearly all patients show muscle weakness at testing in the first weeks after the trauma due to pain making it difficult to distinguish between this condition and paresis due to a nerve lesion [14].

Clinical exam requires competence and knowledge of motor and sensory innervations of brachial plexus primary and secondary trunks and terminal nerves for the upper limb. Clinical signs of axillary nerve lesions may depend on the nerve branches that have been compromised. They may range from paresthesia and hyperesthesia around the shoulder and upper arm to deltoid atrophy manifested by contour changes around the shoulder (this becomes evident at least 3 weeks after the palsy). Active deltoid muscle contraction should be evaluated. Compensatory activity of the supraspinatus muscle in conjunction with the long head of the biceps should be checked. The sensibility has a minor role in initial diagnosis because clinical sensory loss in deltoid area was present only in 7 % of all the patients with lesion of the axillary nerve detected by means of EMG. A similar pattern of clinical presentation is observed also for the musculocutaneous nerve. Clinical exam includes examination of active elbow flexion in order to test biceps brachii and brachialis muscle. Sensory exam of the lateral cutaneous antebrachial nerve shows loss of sensory in only 4 % of all the patients in which a lesion of musculocutaneous nerve was detected by means of EMG [14].

The suprascapular and radial nerves have common root origins, so painful stimulation may overlap. Pain is more commonly appreciated over the posterior portion of the shoulder along the border of the trapezius muscle, with pain worsening at night. Loss of abduction and external rotation of the arm may be observed in such cases depending on to the extent of nerve impairment. Radial, median, and ulnar nerve injuries are rarely isolated and usually observed in case of stretching of infraclavicular terminal branches of brachial plexus. Diagnosis of nerve lesions can be suspected by examining antebrachial motor activity and activity and sensitivity of the hand.

In the presence of clinical suspicion of nerve injury, the first instrumental evaluation is electrophysiologic study by EMG. EMG is the initial investigation of choice, and it is recommended about 3–4 weeks after the nerve injuries (and not earlier) in order to detect site and entity of the lesions [6]. The exam can be subsequently repeated in order to follow the progression of nerve regeneration/healing or to plan an eventual surgical treatment when the absence of nerve recovery is documented after an adequate waiting time. Improved results of EMG without voluntary muscle activity warrant further conservative treatment. In most cases, the clinical evidence of a recovery is detected earlier than the electrophysiologic detection. Operative treatment of shoulder girdle neuropathy can be considered if no clinical and electrophysiologic recovery is present within 3–6 months after injury (9–18 cm of axonal regeneration proceeding 1 mm/day). This approach is indicated also after axillary isolated lesions and also after dislocations, fractures, or iatrogenic injuries without evidence of neurotmesis [28].

In every “closed” nerve lesion, the Tinel sign [29] (“tingling” feeling or “pins and needles” felt at the lesion site or more distally along the course of a nerve when it is tapped) should be tested. This test is one of the most important physical exam maneuvers in peripheral nerve injury evaluation to assess regeneration progress in the site of lesion and along the nerve distally to the lesion. If the Tinel sign progresses distally along the nerve over time, it means that something is recovering (1 mm/day) and the regeneration is possible and occurs; if the Tinel sign is stationary (stays in one place over time), a fourth, fifth, or sixth degree of injury is to be suspected. Unfortunately at the shoulder region, a Tinel sign is very difficult to arouse because of the anatomic depth of nerves around the shoulder (median and ulnar nerves at the axilla and distally are easy to test and radial nerve posteriorly and at the arm level, but suprascapular and axillary nerves at the shoulder are very difficult to test).

A magnetic resonance imaging (MRI) and/or an ultrasound (US) examination of the rotator cuff is always a fundamental and complementary diagnostic step studying a paralytic shoulder to verify the integrity of the rotator cuff [30].

Prognosis of nerve lesions in proximal humeral fractures depends on the grade of axonal and nerve damage. According to Seddon’s [31] classification, we can describe neurapraxia when there is an alteration in nerve conduction, axonotmesis when axons have been cut, and neurotmesis when the nerve continuity is lost. In case of axonotmesis and neurotmesis, a Wallerian degeneration occurs losing the connection among proximal axons and the distal part of the nerve. Subsequently, Sunderland subclassified [32] these injuries into five types in order to better define the entity of injury and the optimal therapeutic approach. Besides neurapraxia (I degree), the axonotmesis was divided into axonal injuries with or without an intact basil lamina (II and III degree) or with complete scar block (IV degree). Neurotmesis is the complete transection (V degree) of the nerve trunk. A combination of conduction block and transection (VI degree) was added later by Mackinnon. From a therapeutic and practical point of view, according to the Sunderland classification, (1) I and II degree injuries will recover spontaneously and should be treated conservatively, (2) V and VI degree injuries should be surgically repaired, and (3) III and IV degree injuries have partial recovery and will likely need surgery.

Posttraumatic nerve injuries following shoulder fractures and dislocations commonly have a good prognosis, and recovery of function may be expected in a period of 3–6 months [6]. In the clinical series reported by Visser including 96 cases of nerve lesions, 73 % of patients reported no or only slight limitations in general function and activities of daily life, 19 % had limitations in such situations, and 6 % presented severe disabilities in daily life activities [14]. Regarding axillary nerve lesions, that is, the most common kind of nerve injury occurring in proximal humeral fractures, recovery is observed in the vast majority of cases [33]. According to Steinmann baseline, EMG should be obtained within 4 weeks after injury with a follow-up evaluation at 12 weeks. When no clinical or EMG improvement is noted, then operative treatment should be performed within 3–6 months from injury [33]. Recently, Hems reported a large study in which he defined four patterns of injuries of the terminal branches of the infraclavicular brachial plexus. In case of anterior glenohumeral dislocation, axillary and ulnar nerves are commonly injured, but rupture is rare. Instead, axillary nerve lesions without dislocation have the highest probability of rupture. In displaced humeral fractures, nerve injury may be due to direct compression of the proximal aspect of the humeral shaft. Last but not least in hyperextension injuries, the musculocutaneous nerve is commonly injured and disrupted. Therefore, the author suggested conservative treatment in most cases of infraclavicular injuries of the terminal branches of brachial plexus with some exceptions. Early exploration was indicated for axillary and musculocutaneous nerve injury without dislocation, and urgent operation was indicated for dislocated proximal humeral fractures in order to relieve pressure on suffering nerves [7].

Regarding iatrogenic injuries, some important prognostic data are given by intraoperative electrodiagnostic analysis [17]. Warrenders et al. [17] reported that the majority of alerts (65 % of the nerve events) occurred during fracture reduction, whereas 31 % occurred during plate application. During that phase of surgical procedure, the humeral shaft and tuberosities were subjected to traction in order to aid fracture reduction. Therefore, removing traction and positioning the arm in neutral position could be useful when an intraoperative delay or a prolonged procedure occurred [17]. Alert signal started when the arm is positioned in flexion abduction and external rotation in 58 % of cases, and positioning of the arm in neutral position and removing of retractors turned off alert signal. In 64 % of cases, there was a return at baseline at the end of surgery, but in 36 % there was no return of MEP at the baseline, and these patients corresponded clinical weakness of muscles especially in radial and axillary nerve territories. All of these patients recovered muscle function by 3 weeks postoperatively [17].

The quality of recovery after a nerve repair depends on many local and general factors. Regarding the age of the patient, nerve function recovery begins to decline after the second decade (75 % good results in children vs 50 % good results in adults) and may be poor after the fifth/sixth decade. Other factors considered are the mechanism of injury, elapsed time from injury, level of injury, type of nerve, specific nerve injured, associated injuries, and tension across the repair.

Some anatomic considerations are useful to prevent iatrogenic nerve injuries especially concerning the axillary nerve. Landmarks of the location of the axillary nerve are taken from the acromion or from the greater tuberosity. When the acromion is considered as landmark, the mean distance from the superior border of the axillary nerve to the anteroinferior border of the acromion is about 6.3 ± 0.5 cm [34]. When the greater tuberosity is considered as reference, the distance between this landmark and the axillary nerve is relatively constant from 3.5 ± 0.2 to 4.6 cm [35]. The surgeon must take into account that the greater tuberosity may be displaced after a fracture. In this case, the course of the axillary nerve can be estimated from the acromion [36].