Tendon transfer is the surgical rerouting of an expendable muscle-tendon unit from its normal insertion to a new insertion to restore a lost motor function following irreversible nerve injury. In the hand, it is the definitive treatment for radial, median and ulnar nerve palsies that have not recovered following nerve repair, grafting, or observed spontaneous regeneration.

The modern practice of tendon transfer in the upper limb was established by Sterling Bunnell, who codified the principles that remain the foundation of hand surgery. The goal is not to recreate normal anatomy but to restore functional grasp, release, pinch and opposition using remaining functional motors.

The six principles of tendon transfer (Bunnell):

1. One tendon, one function: each transferred tendon restores only one function. A transfer split between two different functions produces weak and unpredictable movement in both.
2. Expendable donor: the donor muscle must be genuinely dispensable. Removing it must not create a new functional deficit. For example, the FDS of the ring finger can be sacrificed because the FDP of the same finger provides independent flexion.
3. Synergy: muscles that normally contract together should be preferentially used. A wrist flexor transferred to restore finger extension fires naturally during the grasp-release cycle because wrist flexion and finger extension are synergistic.
4. Straight line of pull: the transfer must run as directly as possible from origin to new insertion. Acute angulations waste power and excursion.
5. Adequate excursion: the donor tendon must have sufficient amplitude of glide to produce the required range of motion. Wrist-level muscles (FCR, ECRL, ECRB, PL, BR) have approximately 30-35 mm excursion; finger-level muscles (FDS, FDP) have approximately 70 mm.
6. Supple joints: full passive range of motion must be present before transfer. A transferred tendon cannot mobilise a stiff joint - Bunnell's dictum.

Motor anatomy of the hand by nerve:

• Radial nerve (C5-C8, posterior cord): supplies all extensor compartments of the forearm. Above the supinator: ECRL, brachioradialis, supinator (via PIN after piercing supinator). The posterior interosseous nerve (PIN) branches at the supinator arcade of Frohse and supplies ECRB, ECU, EDC, EIP, EDQ, EPL, APL, EPB. The superficial radial nerve is purely sensory.
• Median nerve (C5-T1, lateral and medial cords): motor to PT, FCR, PL, FDS (all four), FDP to index and middle (via AIN: FPL, pronator quadratus, FDP index and middle). In the hand: LOAF (Lateral two lumbricals, Opponens pollicis, Abductor pollicis brevis, superficial head of Flexor pollicis brevis).
• Ulnar nerve (C8-T1, medial cord): motor to FCU, FDP to ring and little. In the hand: hypothenar (opponens digiti minimi, abductor digiti minimi, flexor digiti minimi), all palmar and dorsal interossei, medial two lumbricals, adductor pollicis, deep head of FPB.

Biomechanical considerations:

• Tensioning: the transfer must be sutured at appropriate tension. Bunnell recommended suturing with the wrist and digit in the position the transfer would produce, then adjusting so the resting posture is correct. Overtension restricts motion; undertension produces a lag.
• Pulleys and routing: the natural flexor and extensor sheaths, retinacula and fascial slings act as pulleys. When routing a transfer, surgeons may create artificial pulleys (e.g., Guyon's canal for opponensplasty, or the distal edge of the dorsal retinaculum for EDC transfers) to maintain the line of pull.
• Excursion matching: the donor and recipient excursions should be matched as closely as possible. If the donor has less excursion than required, a tenodesis effect (wrist motion amplifying digit motion) can compensate - approximately 10 degrees of wrist flexion produces about 1-2 cm of apparent finger extension.

Assessment before tendon transfer:

• Document the precise nerve deficit using the Medical Research Council (MRC) grading for each muscle group. Distinguish high from low palsy for each nerve, as the transfer set differs.
• Assess joint mobility in every joint the transfer will cross. Use the Bunnell principle: if you cannot passively achieve full range, the transfer will fail. Institute intensive therapy, dynamic splinting, and serial casting before surgery.
• Evaluate donor muscles for strength (MRC grade 4 or better is required), excursion, and expendability. Test each proposed donor independently.
• Assess soft tissue status: scars, skin grafts, and contractures along the planned route may interfere with gliding. Staged soft tissue cover or scar release may be needed.
• Electrodiagnostic studies (EMG and nerve conduction): confirm the absence of reinnervation at the appropriate interval. Serial studies at 3-month intervals document recovery or its absence.
• Sensory assessment: a tendon transfer restores motor function only. Protective sensation (at minimum) in the target territory is needed for functional use of the reconstructed hand. If sensation is absent, consider sensory nerve reconstruction or educate the patient in protective strategies.

• Adhesions and restricted glide: the most common complication. Transferred tendons passing through scarred tissue beds or subcutaneous tunnels may develop adhesions that restrict excursion. Meticulous tissue handling, early protected mobilisation, and avoiding surgery through heavily scarred zones reduce this risk.
• Rupture of the transfer: occurs when the Pulvertaft weave is insufficient (fewer than three passes), when the tendon is of poor quality, or when early active motion is started too aggressively. Re-exploration and re-weaving is required.
• Overtension or undertension: overtension creates a fixed contracture (e.g., a wrist held in excessive extension); undertension produces a persistent lag. Both are avoided by careful intra-operative tensioning and checking the tenodesis effect.
• Loss of donor function: if a non-expendable donor is used, the patient trades one deficit for another. For example, harvesting FCR for EPL transfer in a patient who also needs FCR for wrist flexion will weaken wrist flexion. Planning transfers that use truly expendable donors avoids this.
• Nerve injury during routing: the superficial radial nerve, lateral antebrachial cutaneous nerve, and dorsal sensory branch of the ulnar nerve are all at risk during tendon harvest and routing. Identification and protection prevents neuroma formation.
• Infection: as with any surgical procedure, superficial or deep infection can jeopardise the transfer. Prophylactic antibiotics and meticulous wound management are standard.
• Persistent claw or inadequate correction: in ulnar palsy, inadequate correction occurs when the joint was stiff before surgery, when the transfer was undertensioned, or when the FDS lasso slips. Revision with an alternative transfer may be needed.
• Synergy mismatch: a transfer that violates the synergy principle (e.g., using an extensor to replace a flexor function without considering the relearning required) produces a movement pattern that is difficult to rehabilitate and may never become automatic.

International guidance:

• ASSH (American Society for Surgery of the Hand) and BSSH (British Society for Surgery of the Hand): no single formal guideline exists for tendon transfers; practice follows the principles established by Bunnell, Brand, Riordan and Zancolli as described in standard hand surgery texts. The ASSH manual and BSSH education curricula teach the six principles as the non-negotiable framework.
• FESSH (Federation of European Societies for Surgery of the Hand): European hand surgery training mandates competency in tendon transfers for all three nerve palsies. The FESSH examination typically includes a viva station on tendon transfer principles and specific transfer selection.
• Nerve transfer vs tendon transfer debate: in recent years, nerve transfers (e.g., median-to-radial nerve transfer for radial palsy, anterior interosseous nerve-to-ulnar motor nerve transfer) have emerged as alternatives or adjuncts to traditional tendon transfers. Nerve transfers restore original muscle function but must be performed within 6-12 months of injury for best results. Current practice at major centres often combines nerve transfers (for reinnervation of critical muscles) with tendon transfers (for functions that cannot be restored by nerve transfer).

Global practice variation:

• Leprosy-endemic regions (India, Brazil, parts of Africa and Southeast Asia): tendon transfers for combined nerve palsy are performed in enormous numbers by specialised centres (e.g., the Leprosy Mission hospitals, LKH Karigiri). The Brand transfers were developed in this setting and remain the standard of care. Surgeons in these regions have some of the world's highest volumes and best-reported outcomes for intrinsic replacement.
• High-income countries: the aetiologic spectrum shifts to traumatic nerve injuries (lacerations, fractures, gunshot wounds) and compressive neuropathies. Nerve transfers are increasingly performed alongside or instead of tendon transfers, particularly in brachial plexus reconstruction.
• Paediatric practice: in obstetric brachial plexus injury and congenital conditions, tendon transfers are adapted for growth (e.g., using the Steindler flexorplasty for elbow flexion, or modified opponentsplasties in the growing child). Timing is critical to prevent fixed deformities.

Rehabilitation:

• Post-operative rehabilitation is as important as the surgery itself. A structured hand therapy programme typically involves 4-6 weeks of protected immobilisation followed by progressive active and then resisted exercises. Biofeedback and electrical stimulation of the transferred muscle may aid motor relearning. Total rehabilitation takes 3-6 months.