Adult Brachial Plexus Injury

The brachial plexus is the network of nerves that carries motor and sensory signals from the lower cervical and first thoracic spinal nerves out to the whole upper limb. An adult traumatic brachial plexus injury is usually a high-energy traction injury: the classic patient is a young man thrown from a motorcycle, whose head and shoulder are forced apart so violently that the nerves are stretched, ruptured, or pulled out of the spinal cord altogether.

What it costs. A partly or completely paralysed, numb, and often painful arm in a young, working person. Three questions run through everything that follows: where and how badly the plexus is injured, above all whether each root is preganglionic or postganglionic; whether there is still time to reinnervate the muscle; and in what order to rebuild the limb.

Not birth palsy. Brachial plexus birth palsy is a different obstetric traction problem in neonates, with its own classification (Narakas) and a high rate of spontaneous recovery. The adult injury is higher-energy, recovers far less spontaneously, and its surgical decisions are driven hard by timing and by the preganglionic/postganglionic distinction.

The five sections. Roots, trunks, divisions, cords and branches, proximal to distal ("Read Those Damn Cadaver Books"). What matters clinically is which section is injured and what that predicts.

• Roots (C5, C6, C7, C8, T1): the spinal nerves emerging from the neck. The dorsal root ganglion (DRG) sits on the sensory root, and its position is the anatomical basis of the preganglionic/postganglionic split.
• Trunks: C5+C6 form the upper trunk, C7 alone the middle trunk, C8+T1 the lower trunk, all in the supraclavicular region.
• Divisions: each trunk splits into an anterior and a posterior division behind the clavicle.
• Cords (lateral, posterior, medial): named for their relation to the axillary artery, in the infraclavicular region.
• Branches: the terminal nerves - musculocutaneous, axillary, radial, median, ulnar.

View larger

Branches that leave early are diagnostic flags. The dorsal scapular nerve (rhomboids) and the long thoracic nerve (serratus anterior) arise from the roots themselves, and the suprascapular nerve leaves the upper trunk. Paralysis of these muscles places the lesion very proximally, at root level, and points to a preganglionic injury.

The dorsal root ganglion. Sensory cell bodies live in the DRG. An injury proximal to it (preganglionic, a root avulsion) leaves the peripheral sensory axon attached to its living cell body, so the axon survives and the sensory nerve action potential (SNAP) is preserved even though no signal reaches the cord and the limb is anaesthetic. An injury distal to it (postganglionic, a rupture) disconnects the peripheral axon from its cell body, so the axon degenerates and the SNAP is lost. That one anatomical fact explains a whole cluster of findings, and it decides the operation: an avulsed root has no proximal stump to sew to and needs a nerve transfer, whereas a rupture leaves a proximal stump for direct repair or a nerve graft.

View larger

How the injury happens. Adult plexus injury is overwhelmingly a traction (stretch) injury. When the head and shoulder are forced apart, the classic motorcyclist thrown over the handlebars, the angle between neck and shoulder widens violently and the nerves are stretched beyond their limit.

Where the nerve fails depends on the force and the anatomy. The upper roots (C5-C6) are tethered and tend to rupture distal to the DRG (postganglionic), whereas the lower roots (C8-T1) are less supported and are more prone to being avulsed from the cord (preganglionic). This is why a Horner's sign, a T1 sign, signals a severe, proximal, lower-root injury.

The clock starts at injury. Once an axon is disconnected from its cell body it undergoes Wallerian degeneration, and the denervated muscle it supplied begins an irreversible clock toward fibrosis. Regenerating axons advance only slowly, on the order of about a millimetre a day, which is why reinnervation distance and timing dominate the surgical strategy: the closer and the sooner new axons can be delivered to a muscle, the better.

Other mechanisms. A sharp or open laceration divides the plexus cleanly, which favours early direct repair. An infraclavicular injury is often produced by a shoulder dislocation or a proximal humerus fracture dragging on the cords and branches.

View larger

Adult plexus injuries are classified in several overlapping ways, and they are used together: where along the plexus, how severe the nerve lesion, and which functional pattern.

The assessment is structured: history, then a systematic limb examination, then the special signs that localise the level.

History and resuscitation. These are high-energy poly-trauma patients, so assess and resuscitate the whole patient first. Then note the mechanism (motorcycle, fall, traction), any associated fractures and vascular injury, and the time since injury, which drives the surgical clock.

Look and feel. Inspect for posture (waiter's tip in upper injuries, a flail arm in pan-plexus), wasting, a winged scapula (serratus/long thoracic) and swelling or bruising. Map the sensory loss by dermatome.

Motor and reflex map. Grade every key muscle on the MRC scale by root and nerve so that the level can be plotted:

• Deltoid and supraspinatus - C5
• Biceps - C5-C6
• Wrist and finger extensors - C7
• Finger flexors - C8
• Intrinsics - T1
• Rhomboids and serratus - the proximal-branch flags

Vascular and special signs. Check the distal pulses and perfusion. Look hard for a Horner's sign: ptosis, miosis and facial anhidrosis on the injured side mean the T1 sympathetic outflow has been disrupted, a marker of a lower root avulsion and a severe, proximal injury that changes both the prognosis and the plan. Document the Tinel sign in the supraclavicular fossa; an advancing Tinel suggests a postganglionic lesion with regenerating axons.

Investigations answer one question: for each root, is the lesion preganglionic or postganglionic, and is there anything to repair?

NeurophysiologyMRICT myelographyPlain films and CT

Nerve conduction studies and EMG are central, and are usually most informative from about 3 to 4 weeks, once Wallerian degeneration has declared itself. A preserved SNAP in an anaesthetic dermatome is the classic marker of a preganglionic lesion, for the reason set out in the anatomy: the DRG cell body is still attached to its peripheral axon.

EMG maps which muscles are denervated and, on serial testing, can show early reinnervation (nascent motor units) that argues for watchful waiting rather than surgery. Loss of paraspinal muscle (erector spinae) activity points to a very proximal, root-level lesion.

MRI of the brachial plexus, with high-resolution, heavily T2-weighted myelographic sequences, shows the roots, the cord and pseudomeningoceles, and increasingly visualises the postganglionic plexus directly. It is non-invasive and also shows associated cord and soft-tissue injury. A pseudomeningocele, a CSF-filled outpouching where the dural sleeve has torn as the root was avulsed, is a strong but imperfect sign of root avulsion.

"Imperfect" has a number. In a prospective series of 24 patients taken to reconstruction with operative exploration as the reference standard, pseudomeningoceles were present in only about 68% of avulsions, so roughly one avulsed root in three has none and their absence is no reassurance whatever. Direct visualisation of the rootlets on high-resolution axial T2-DRIVE MR myelography performed far better, identifying the specific avulsed roots with sensitivity 100%, specificity 97.1% and accuracy 98.1%; the absent or pauci-rootlet appearance was the single most reliable sign.

The practical rule is to make the diagnosis on the rootlets and treat the pseudomeningocele as ancillary, not the other way round. The evidence is level II (a prospective, small, single-centre cohort with surgical correlation), so the precise figures are indicative, but the direction is clear and it inverts the order in which these signs are usually taught.

View larger

View larger

View larger

CT myelography has long been a reference investigation for root avulsion: contrast in the thecal sac outlines the rootlets, and an absent rootlet or a pseudomeningocele indicates avulsion. It is invasive, but remains highly regarded for defining the rootlets at each level and is often used alongside MRI.

Plain radiographs and CT assess the bony injuries that travel with a plexus lesion, and those injuries carry clues of their own. Transverse process fractures suggest a violent root-level traction force, a clavicle or first-rib fracture raises concern for the supraclavicular plexus and vessels, and a fractured cervical spine must be excluded. An elevated, non-moving hemidiaphragm on a chest film implies phrenic nerve (C3-C5) palsy and a proximal injury.

View larger

View larger

Management is structured as who to operate on, when, and what to do, with the non-surgical care that runs alongside.

TimingReconstructive ladderNerve transfersSupportive care

Timing dominates outcome. A denervated muscle has a limited time before it becomes fibrotic and can no longer be re-animated by reinnervation, so the whole strategy is to reconnect the nerve before that happens.

Sharp / open laceration

Timing

Immediate or early (primary repair)

Reasoning

Clean-cut ends can be repaired directly while they are identifiable

Associated vascular injury

Timing

Urgent (with the vascular repair)

Reasoning

Limb perfusion takes priority; the plexus can be tagged/explored at the same time

Closed traction, clear root avulsion

Timing

Early (around 6-12 weeks)

Reasoning

No spontaneous recovery is possible - do not waste the muscle's time window

Closed traction, uncertain

Timing

Explore/reconstruct by 3-6 months if no recovery

Reasoning

Allows neurapraxic/recovering lesions to declare themselves while staying inside the window

Late presentation (beyond about 12 months)

Timing

Reconstruct function, not nerves

Reasoning

Denervated muscle is lost - shift to tendon transfer, fusion, or free muscle transfer

Timing of Surgery by Injury Type

Injury type	Timing	Reasoning
Sharp / open laceration	Immediate or early (primary repair)	Clean-cut ends can be repaired directly while they are identifiable
Associated vascular injury	Urgent (with the vascular repair)	Limb perfusion takes priority; the plexus can be tagged/explored at the same time
Closed traction, clear root avulsion	Early (around 6-12 weeks)	No spontaneous recovery is possible - do not waste the muscle's time window
Closed traction, uncertain	Explore/reconstruct by 3-6 months if no recovery	Allows neurapraxic/recovering lesions to declare themselves while staying inside the window
Late presentation (beyond about 12 months)	Reconstruct function, not nerves	Denervated muscle is lost - shift to tendon transfer, fusion, or free muscle transfer

Referral is a clinical decision. Delay is the commonest avoidable failure, and it is common. Across 84 adults with traumatic plexus injury assessed at two tertiary centres, 44% reached a plexus surgeon more than 3 months after injury, that is, only once the 3-6 month reconstruction window set above had already opened. Global palsy was the commonest pattern (39%), and a motorcycle mechanism and insurance or system factors predicted delay. The surgeon who first sees the patient rarely does the reconstruction, so the threshold for onward referral to a specialist plexus centre should be low and immediate rather than contingent on watching for recovery.

The surgical options form a ladder, chosen by how usable the proximal nerve is.

Neurolysis. For a neuroma-in-continuity that still conducts, with a positive intra-operative nerve action potential across it, the nerve is simply freed from scar; resecting it would do more harm than good.

View larger

Nerve repair or graft. A postganglionic rupture with a healthy proximal stump is repaired directly if this can be done without tension, or bridged with an interposition nerve graft, for example sural nerve. Either needs a viable proximal source of axons.

View larger

Nerve transfer (neurotisation). The answer when the root is avulsed and there is no proximal stump. A nearby expendable working donor is re-routed to power a key target close to its muscle, giving a short reinnervation distance.

Free functioning muscle transfer (FFMT). For late presentation or after failed reconstruction: a whole muscle, commonly gracilis, is transplanted with its vessels and nerve and then powered by a donor nerve, importing new contractile tissue when the native muscle is gone.

View larger

Priorities. Restore elbow flexion first, then shoulder stability and abduction, then hand function; the prognostic factors below explain why the hand comes last. Know a few workhorse transfers well, organised by the function they restore:

• Elbow flexion: the Oberlin transfer, a fascicle of the ulnar nerve to the biceps motor branch, often with a median fascicle to brachialis (the "double fascicular" transfer). Used when C8-T1 is intact, that is, in an upper plexus injury.
• Shoulder abduction and external rotation: spinal accessory nerve to suprascapular nerve, plus a triceps motor branch (radial) to axillary nerve transfer.
• Global or pan-plexus injury: extraplexal donors such as the intercostal nerves, spinal accessory and the contralateral C7 are recruited; goals are scaled back to elbow flexion and shoulder stability, and hand function is rarely restored.

What "elbow flexion first" delivers, and why age alone should not stop a referral. In 58 patients older than 50 reconstructed for elbow flexion, mean flexion grade rose from 0.26 to 2.63 and 60% achieved functional flexion of M3 or better, against none preoperatively. Nerve transfers outperformed free functioning muscle transfer (69% versus 43% reaching M3 or better). High-energy injuries did worse, but age itself did not correlate with the flexion grade achieved.

That last clause is the one to carry. Age is a common and largely unexamined reason for not referring. The factors that genuinely predict a poor result are injury energy, the preganglionic/postganglionic pattern and time to reconstruction, not the birth date. The evidence is level IV (a retrospective single-centre cohort), so the percentages are indicative; the direction is what matters for the referral decision.

View larger

View larger

View larger

Reconstruction is only part of the package; these run alongside throughout.

• Rehabilitation: physiotherapy to keep the joints supple and prevent contractures while awaiting reinnervation, then motor re-education afterwards.
• Expectation-setting: a frank early conversation about realistic goals, a useful "helper arm" rather than a normal arm, is part of good care and is explicitly recommended for the difficult and the failed case.
• Pain: deafferentation pain, especially with avulsions, is severe and is managed actively with neuropathic agents and a multidisciplinary approach; its treatment is set out in the next section.

Deafferentation (avulsion) pain is one of the most disabling features of a preganglionic injury and a major determinant of quality of life. It is a crushing or burning pain in the anaesthetic limb, generated by hyperactive deafferented neurons in the dorsal horn after the root is torn from the cord, and it is treated in tiers.

Outcome turns on a handful of factors that are easy to list separately and hard to weigh together; the table pulls them into one framework.

Why the hand comes last. Whatever the pattern, hand reinnervation is the hardest and least reliable goal, because the regenerating axons must travel the greatest distance and must arrive before the intrinsic muscles fibrose. That is why the reconstructive priorities run elbow flexion, then shoulder, then hand, and why honest counselling aims for a useful "helper arm" rather than a normal hand. High energy and delay are the two most modifiable predictors of a poor result, hence the emphasis on early referral.

• Specialist, multidisciplinary care is the consistent global message: traumatic adult plexus injuries do best when managed early in dedicated centres combining orthopaedic/hand, plastic, and peripheral-nerve neurosurgery, with physiotherapy and pain services. Delayed referral is documented as common and as a driver of poorer results.
• Timing principles are broadly agreed: explore sharp/open and vascular injuries early; reconstruct closed traction injuries inside roughly a 3 to 6 month window before denervated muscle is lost; shift to reconstructive (tendon transfer, fusion, free muscle transfer) options for late presentations.
• Nerve transfers have shifted global practice over recent decades: distal nerve transfers (such as the Oberlin transfer for elbow flexion and spinal-accessory-to-suprascapular transfer for the shoulder) are now widely used, shortening reinnervation distance and improving outcomes, while nerve grafting remains the answer for graftable postganglionic ruptures.
• Outcome variation reflects injury severity and access: upper-plexus injuries reliably regain elbow and shoulder function, whereas hand recovery after lower-plexus or pan-plexus injury remains poor everywhere - so realistic goal-setting (a useful helper arm) is a universal part of consent and care.
• Mechanism and demographics differ by region: motorcycle and road-traffic trauma dominate in many low- and middle-income settings, shaping a younger, predominantly male, higher-severity caseload - a global epidemiological pattern rather than any single country's experience.

High-yield, single-best-answer style points that examiners turn into questions.

The following studies support the key teaching points above.