Hand Surgery
1st Edition

Reconstruction for Radial Nerve Palsy
Reid A. Abrams
The radial nerve arises from the posterior cord of the brachial plexus. As all three trunks contribute posterior divisions to the posterior cord, all roots of the plexus contribute to the various radially innervated muscles. For example, the brachialis and supinator are innervated by C5 and C6, whereas the extensor indicis proprius (EIP) and extensor pollicis longus (EPL) are innervated by the C7 and T1 roots.
The radial nerve enters the upper posterior brachium and passes through a triangular space that is bordered superiorly by the teres major muscle, laterally by the humerus, and medially by the long head of the triceps muscle. Accompanied by the profunda brachii artery, the nerve enters the musculospiral groove between the medial and lateral heads of the triceps. The nerve crosses the humerus posteriorly from proximal medial to distal lateral. It is closely adherent to bone for a distance of approximately 6.5 cm. The nerve is coincident with the medial aspect of the humerus at approximately 20 cm proximal to the medial humeral epicondyle or at approximately the junction of the proximal and the next 25% of the length of the humerus. It traverses the lateral aspect of the humerus at approximately midhumeral length, which is approximately 14 cm proximal to the lateral humeral epicondyle (1). Throughout its course posterior to the humerus, the nerve gives off branches to the lateral and medial heads of the triceps and to the lower lateral brachial cutaneous nerve (1). After passing around the lateral aspect of the humerus and piercing the lateral intermuscular septum, the radial nerve enters the interval between the brachialis and brachioradialis muscles. The brachialis muscle is usually dually innervated and receives motor supply from the musculocutaneous and radial nerves (2,3). Proximal to the lateral epicondyle, branches go to the brachioradialis and the extensor carpi radialis longus (ECRL). The origin of the branch to the extensor carpi radialis brevis (ECRB) is variable, arising from the main nerve trunk, the superficial branch of the radial nerve, or the posterior interosseous nerve (PIN). The incidence of each of the ECRB nerve branch morphologies is highly variable in the literature (2,4,5). Near the level of the lateral epicondyle of the humerus, the nerve bifurcates into the superficial and deep branches (2,6). The superficial branch of the radial nerve continues distally beneath the brachioradialis muscle, until it passes between the tendons of the brachioradialis and the ECRL, approximately 9 cm proximal to the radial styloid. Distally, it arborizes to provide sensory innervation to the dorsoradial aspect of the hand, thumb, index finger, and, variably, the long and ring fingers (6). Distal to the elbow joint, the deep radial nerve branch, or PIN, passes beneath recurrent vessels from the radial artery and then, approximately 5 cm distal to the lateral humeral epicondyle, enters the supinator muscle underneath the arcade of Frohse (2). The arcade of Frohse is the proximal margin of the supinator. Its morphology can vary from a muscular to a tendinous quality (7,8). Multiple branches innervate the supinator as the nerve traverses beneath it. As the PIN emanates from the distal margin of the supinator, branches exit to supply the extensor digitorum communis (EDC), the extensor digiti quinti (EDQ), and the extensor carpi ulnaris (ECU). The remaining trunk of the PIN continues distally in the interval between the EPL and the abductor pollicis longus (APL). One branch exits the nerve to branch again to innervate the APL and extensor pollicis brevis (EPB), another branch emanates from the nerve and subdivides to supply the EIP and EPL, and the remainder of the nerve continues distally to the wrist joint (Fig. 1).
Knowledge of proximal-to-distal muscle innervation order is practical information when observing for spontaneous radial nerve recovery or recovery after neurorrhaphy. In different cadaver specimens, this order is variable (2,3). Average innervation distances from a proximal landmark indicates the average innervation order from proximal to distal for the extensor forearm muscles; this order is brachioradialis, ECRL, ECRB, supinator, EDC, ECU, EDQ, APL, EPL, EPB, and EIP (2,3,9) (Fig. 1). Measuring from a landmark

100 mm proximal to the lateral humeral epicondyle, the mean distance to the point at which the most proximal branch reaches the brachioradialis was 97.2 mm. The analogous distance to the EIP (usually the distal-most radial nerve–innervated muscle) was 299.8 mm (2). Using the estimation by Seddon and Medawar (10) that nerve regeneration proceeds at a rate of approximately 1 mm per day, a theoretic prediction of time to recovery of muscles that are supplied by an injured radial nerve can be made. For example, if the nerve was injured 100 mm proximal to the lateral humeral epicondyle, recovery would begin with the brachioradialis at approximately 3 to 4 months and would be complete, ending with the EIP, by approximately 10 months.
FIGURE 1. A: Schematic of the typical branch pattern sequence (1,2,3,4,5,6,7,8,9,10,11 and 12). Numbers with decimals represent distances in millimeters along the nerve and branches from a landmark on the humerus that is 100 mm proximal to the lateral epicondyle (asterisk). The lateral epicondyle is near the branch point of the posterior interosseous nerve (PIN) and the superficial branch of the radial nerve (SBRN). B: Artist’s drawing of the extensor forearm with an overlay showing PIN anatomy. APL, abductor pollicis longus; BC, brachialis; BR, brachioradialis; ECRB, extensor carpi radialis brevis; ECRL, extensor carpi radialis longus; ECU, extensor carpi ulnaris; EDC, extensor digitorum communis; EDQ, extensor digiti quinti; EIP, extensor indicis proprius; EPB, extensor pollicis brevis; EPL, extensor pollicis longus; SUP, supinator. (From Abrams RA, Zeits RJ, Leiber RL, et al. Anatomy of the radial nerve motor branches in the forearm. J Hand Surg 1997;22:232–237, with permission.)
Injury to the radial nerve in the upper extremity rarely occurs proximal to the innervation of the triceps; consequently, lesions at this level are not considered in this chapter. The functional deficits in the upper extremity after a radial nerve injury involve the wrist, finger, and thumb extensors. Traditionally, radial nerve palsy has been categorized into high and low types. High palsy refers to a radial nerve injury that is proximal to the elbow and that results in deficits in wrist and digital extension. Low palsy refers to denervation of muscles that are innervated by the PIN, thus sparing wrist extension, with deficits in thumb and finger extension. As the ECU is denervated in a low palsy, wrist extension occurs in a dorsoradial direction. Because the innervation to the ECRB has a variable origin from the radial nerve, there may be a subtle intermediate palsy, which relates to the integrity of the ECRB branch, between the high and low varieties. As the ECRB insertion at the wrist is relatively central, being at the base of the long finger meta-carpal, its action is to extend the wrist with minimal deviation in the coronal plane. If the ECU is denervated, the ECRL remains innervated, and the branch to the ECRB is spared, there is only mild radial deviation with wrist extension. If the ECU and ECRB are denervated, dramatic radial deviation with wrist extension occurs due to the isolated function of the ECRL. In a high radial nerve palsy, aside from the obvious inability to extend the digits or wrist, poor grasp is a major functional complaint. After radial nerve block in volunteers, grip strength decreased by 77% (11). Because there are no active extensor muscles that cross the flexor-extensor axis of the wrist to balance the flexor moment, any active digital flexion results in wrist flexion (Fig. 2). Not only does this cause the digital flexors to work against the resistance of the passive tension of their denervated antagonists, but also they work at suboptimally shortened sarcomere lengths (on the ascending limb of the length–tension curve) (12).
The etiology of radial nerve palsy in the brachium is usually direct or indirect trauma. Penetrating trauma can variably affect a nerve. Sharp lacerations usually result in nerve

transection, whereas gunshot wounds often result in neuropraxia or axonotmesis and, rarely, neurotmesis (13). High radial nerve palsy that is associated with humeral shaft fractures can be secondary to nerve contusion or nerve entrapment in the fracture site at the time of injury or after closed reduction. Late entrapment of the nerve can occur within fracture callus (14,15,16,17,18,19,20,21,22 and 23). The reported incidence of radial nerve injury that is associated with fractures of the humeral shaft ranges from 1.8% to 16.0% (15,18,19,22,24). In most series of supracondylar humeral fractures in children, the radial nerve is the most commonly injured nerve, especially when the distal fragment is displaced posteromedially (25,26).
FIGURE 2. The effect of radial nerve palsy at the wrist. Each tendon that crosses the wrist is represented by a group of circles, each of which denotes one unit on a scale of relative tension. Each tendon lies in its mechanical relationship to the axes of flexion-extension and radial-ulnar deviation. APL, abductor pollicis longus; ECRB, extensor carpi radialis brevis; ECRL, extensor carpi radialis longus; ECU, extensor carpi ulnaris; EDCI, extensor digitorum communis index; EDCL, extensor digitorum communis long; EDCM, extensor digitorum communis middle; EDCR, extensor digitorum communis ring; EDQ, extensor digiti quinti; EIP, extensor indicis proprius; EPB, extensor pollicis brevis; EPL, extensor pollicis longus; FCR, flexor carpi radialis; FCU, flexor carpi ulnaris; FDPI, flexor digitorum profundus index; FDPL, flexor digitorum profundus long; FDPM, flexor digitorum profundus middle; FDPR, flexor digitorum profundus ring; FDSI, flexor digitorum superficialis index; FDSL, flexor digitorum superficialis long; FDSM, flexor digitorum superficialis middle; FDSR, flexor digitorum superficialis ring; FPL, flexor pollicis longus; PL, palmaris longus. (From Brand PW, Hollister A. Operations to restore muscle balance to the hand. In: Brand PW, Hollister A, eds. Clinical mechanics of the hand, 2nd ed. St. Louis: Mosby–Year Book, 1993:180–189, with permission.)
PIN palsy can arise from penetrating trauma or indirect trauma from fractures and dislocations about the elbow. Iatrogenic trauma during open reduction of radial head or neck fractures, repair of distal biceps ruptures, procedures for the treatment of lateral epicondylitis, removal of proximal radioulnar synostosis, and other procedures around the elbow have been implicated in PIN palsy (27,28,29,30,31 and 32). Palsy from compression of the PIN by a mass, such as a ganglion cyst or synovitis that arises from the elbow joint, has also been reported (33,34). Nerve entrapment can occur in the radial tunnel in which the nerve underlies unyielding fascial structures, most commonly, the arcade of Frohse or bands within the supinator muscle (7).
Evaluation of a patient in the acute phase of a potential peripheral nerve injury can be difficult due to pain or cumbersome dressings or splints. After more urgent priorities have been addressed, a careful history and physical examination that focus on nerve function are indicated.
As a general rule, when evaluating motor function, the muscle that is being tested should be palpated to appreciate its activation. To screen radial nerve sensory function, assess sensibility over the dorsal first web space. Testing a relatively distally innervated muscle, such as the EPL, serves as a good motor screening test, because, if it works, the remaining proximal nerve is most likely intact. It is important not to misinterpret thumb interphalangeal (IP) joint extension, necessarily, as an indication that the EPL is functional. Thumb IP joint extension can occur via the EPL or by median- and ulnar-innervated thumb intrinsic muscles that affect terminal joint extension via the extensor hood. IP extension in the latter instance occurs concomitantly with thumb metacarpal palmar abduction, adduction, or flexion, whereas with activation of the EPL, the thumb extends dorsal to the plane of the hand.
As the clinical situation permits, a more detailed examination should follow, especially if the initial screening examination indicates a deficit. With the patient’s forearm in neutral rotation, the brachioradialis is tested by palpating and visualizing the muscle during active elbow flexion. Manual wrist extensor muscle testing is best performed with the digits in a fist to isolate wrist extensor muscle function. Frequently, the ECRB and ECRL tendons can be palpated separately near their insertions at the bases of the long and index finger metacarpals, respectively. Also, when both muscles are functional, a sulcus can be appreciated between them just distal to their origins, with activation.
Evaluation of digital extension can be confounding to the inexperienced. When testing extrinsic digital extensor function, the wrist should be held in neutral. If wrist flexion is allowed, tenodesis of the EDC, EIP, and EDQ can extend the finger MCP joints and simulate functional extensors, despite muscle inactivity. (The wrist flexion tenodesis

maneuver is often useful to distinguish EDC tendon rupture from PIN palsy.) Intact IP joint extension can be misinterpreted as an indication of intact radial nerve function. Proximal IP joint extension occurs not only through the extrinsic extensor contribution to the central slip, but also via interosseous and lumbrical muscles, which are innervated by the ulnar and median nerves. The examiner should focus on MCP extension when testing extrinsic digital extensor function. Asking the patient to extend the MCP joints with the IP joints flexed (in a “claw hand” posture) makes the examination less confusing.
In pronation, the ECU is an ulnar deviator of the wrist (35). Having the patient deviate the wrist ulnarly while it is held in pronation and extension can isolate ECU function. If the ECU is denervated, active ulnar deviation occurs only with wrist flexion [from unopposed flexor carpi ulnaris (FCU) activation]. When evaluating a patient with a chronic radial nerve palsy, in whom treatment likely involves tendon transfer, evaluation must include potential donor muscles.
In general, I have not found electrodiagnostic studies to be a useful adjunct to a careful physical examination (or serial examinations) in the diagnosis and decision making during the treatment of radial nerve palsy.
Radial Nerve Palsy Associated with Penetrating Trauma
Nerve injuries that are associated with sharp penetrating trauma, such as glass or knife lacerations, usually result in nerve transection. Thus, after management of emergent associated injuries, nerve exploration and repair should proceed as early as possible. Nerve palsy from high- and low-velocity gunshot wounds results in spontaneous recovery in 69% of cases (13). Unless early operative treatment is indicated for associated injuries, observation for a reasonable period of time can be justified. This period can be calculated based on available data (2). However, if recovery does not progress in a timely and anatomic fashion, exploration should proceed expeditiously. If the muscles are not reinnervated within 1 to 2 years, they become fibrotic, lose motor end plates, and become less receptive to reinnervation (10,36).
Radial Nerve Palsy Associated with Fractures and Dislocations
Considerable controversy exists regarding the management of radial nerve palsy that is associated with humeral shaft fractures (14,15,16,17,18,19,20,21,22 and 23). Although some authors have advocated early exploration (16,18), others have recommended exploration only if recovery is delayed (14,17,19,20,21 and 22). Although most authors support delayed exploration, there is no agreement for the optimal duration of observation (37). Consensus on the superiority of early versus delayed exploration is contingent on (a) whether there is a significant incidence of remediable lesions (nerve laceration or entrapment), and (b) whether the prognosis for correcting the problem early is better than addressing it later.
Omer observed spontaneous recovery in 83% of nerve injuries that were associated with upper extremity fractures, with 72% recovering in the first 4 months. No appreciable recovery was noted after 7 months (13). Eighty percent to 95% of radial nerve palsies that were associated with closed humeral fractures recovered spontaneously (19,24,38). Closed diaphyseal fractures that are associated with radial nerve palsy are rarely associated with remediable lesions (22,38,39). Of 15 humeral fractures that were associated with radial nerve palsy that were treated by acute nerve exploration and open reduction and internal fixation, only three nerves (20%) were lacerated, two irreparably, and one nerve was repaired with recovery. All intact nerves recovered spontaneously (38). In a series of 111 humeral shaft fractures, 17 (15%) were associated with radial nerve palsy (22). Fourteen radial nerves were explored early, and 13 (93%) were uninjured (22). In another series of humeral shaft fractures that were associated with radial nerve palsy, 85% of explored nerves were no more than slightly bruised (39). Delay of nerve exploration and neurolysis, liberation from fracture callus, or repair has not appeared to compromise recovery (20,22,40). These reports support radial nerve exploration, only if recovery is delayed. Due to reports of occasional fracture site incarceration, intramedullary fixation of humeral diaphyseal fractures in the setting of radial nerve palsy is contraindicated (41).
Special fracture circumstances may require particular consideration. In a series of 14 open humeral fractures, seven nerves were found to be lacerated partially or completely, and two more nerves were interposed between fracture fragments (15). Of the nine (64%) remediable lesions out of 14 radial nerve palsies, six nerves were repaired, and two interposed nerves were freed. All but one patient, who was lost in follow-up, had full or nearly full motor recovery (15). A high incidence of radial nerve laceration or fracture interposition has been reported to be associated with closed longitudinal or oblique fractures of the distal one-third of the humerus (16,18,39). However, other investigators have not corroborated this (19,24). Some authors believe secondary palsy after closed reduction warrants early exploration (42,43), whereas others have demonstrated that secondary palsies have a high enough incidence of spontaneous recovery not to warrant early exploration (24).
Author’s Preferred Method
There is strong evidence that spontaneous recovery usually follows closed humeral shaft fractures. Furthermore, delay of remediation does not appear to compromise the outcome if

nerve exploration with repair or liberation from scar or fracture callus is performed within a reasonable period. To discern how long this period should be, the level of nerve injury is estimated, and a predicted time to recovery is calculated, based on available data in the literature (2,3,9). If initial recovery is delayed beyond 4 to 5 months for mid-shaft fractures, or if the delay is beyond the predicted recovery duration, exploration is indicated. Although the literature is divided with regard to management of the spiral oblique fracture at the junction between the middle and distal one-third of the humerus (Holstein-Lewis lesion), there is sound support for delayed exploration (which is my practice, unless open fracture treatment is elected).
As the incidence of radial nerve laceration or fracture site entrapment appears to be quite significant with open humeral fractures, I prefer to explore the nerve when treating open fractures. It is also prudent to explore a dysfunctional radial nerve if it is elected for any reason to treat a closed fracture by open means (e.g., irreducible fracture, fracture that is associated with vascular injury, multiple trauma patient). I also support early exploration of a dysfunctional radial nerve that was fully functional before fracture manipulation.
Radial Nerve Palsy Associated with Supracondylar Fractures in Children
The incidence of neurologic injury that is associated with supracondylar humeral fractures in children is reported between 9% and 49% (25,44,45 and 46). The most frequently involved nerve is variable in the literature, but the radial nerve has been affected in approximately 25% of cases with neurologic involvement (44,45 and 46). A high association of radial nerve palsy with posteromedially displaced fractures has been shown (44). There are case reports and series that include the rare occurrence of radial nerve transection noted at delayed exploration (25,46,47 and 48). In these cases, the results of delayed radial nerve repair have ranged from highly successful (47,48) to poor, which required tendon transfer reconstruction (25). For the most part, the literature supports nonsurgical treatment of nerve palsy that is associated with supracondylar humerus fractures. Although the main indication for acute nerve exploration is neurovascular injury that is associated with an irreducible fracture, usually it is not the radial nerve that is found entrapped in the fracture but rather the brachial artery or ulnar or median nerves (26). Delayed radial nerve exploration is indicated if recovery is delayed beyond 4 to 5 months (25,45).
Radial Nerve Palsy Associated with Monteggia’s Fractures
The incidence of radial nerve injury that occurs with Monteggia’s lesions varies in the literature from 3% to 43% (49,50,51,52 and 53). Tardy palsy has been reported in cases of old Monteggia’s lesions in which the radial head remains dislocated (54,55). There have been reports of PIN palsy that accompanied irreducible Monteggia’s lesions in which, on exploration, the nerve was found wrapped around the radial neck, thus hampering the reduction (56). Full recovery followed open reduction and liberation of the entrapped nerve (56). An irreducible radial head or a secondary palsy are indications for early exploration (56,57). Otherwise, late exploration is undertaken only if recovery is delayed beyond what would be anticipated, based on calculations after estimation of the level of the radial nerve lesion. If the lesion is approximately at the level of the arcade of Frohse, a reasonable estimated time to recovery is approximately 3 months.
Nonoperative treatment of radial nerve palsy is most appropriate for those patients who await spontaneous nerve recovery or after radial nerve repair. In some cases of combined palsies, in which there may be a paucity of suitable donors, wrist splints can enhance hand function definitively. Otherwise, the anticipated outcome of tendon transfer reconstruction is so rewarding that, unless there are contraindications, nonsurgical reconstruction is not the treatment of choice for radial nerve palsy in the setting of an irreparable nerve.
Maintaining a supple hand is of paramount importance, whether nerve recovery is expected or tendon transfers are anticipated. If a wrist drop is neglected, a wrist flexion contracture ensues. Due to the EDC tenodesis effect, the MCP joints are maintained in extension, and secondary MCP joint extension contractures result. Clearly, there are poor wrist extension and grasp if these contractures persist after motor recovery or tendon transfer reconstruction. Early after radial nerve injury, patients should be taught range-of-motion and stretching exercises that focus on wrist extension and MCP joint flexion. A simple wrist extension splint may assist in this endeavor. Patients find frequently that a splint that stabilizes the wrist in slight extension improves grasp, and, although the splint prevents the wrist flexion tenodesis effect that assists with hand opening, the extension that is provided by the intrinsic muscles at the proximal and distal IP joints may be adequate. If hand opening is a problem, then other functional braces that passively extend the MCP joints are available (43,58). However, these higher-profile braces may be cumbersome or may be perceived as unsightly compared to a simple wrist splint.
Wrist fusion in the setting of radial nerve palsy is generally ill advised, as it sacrifices tenodesis hand opening with wrist flexion. Common transfer options for digital extension are

the wrist flexor muscles that do not have as much excursion as the EDC. (See Chapter 52.) Consequently, if the wrist motors are used to replace the EDC, incomplete digital extension occurs if wrist flexion tenodesis is not available to compensate. If wrist fusion is a selected part of the reconstruction, the flexor digitorum superficialis (FDS) transfer, if available, is probably the best motor to reconstruct digital extension, due to more adequate excursion in the absence of wrist tenodesis (59,60). When radial nerve palsy is combined with paralysis of other nonradially innervated muscles (e.g., concomitant median nerve palsy, brachial plexopathy, or injured potential donor muscles), there may be a paucity of donors, and wrist fusion to stabilize the wrist in extension may improve hand function. Fusion can be simulated preoperatively with functional wrist splints, and patients can decide if fusion can be useful for them. It is important for patients to confirm that they are capable of adequate hand opening (via intrinsic muscle function at the proximal and distal IP joints) with the wrist stabilized in extension. If fusion is elected, positioning in excessive wrist extension should be avoided. Neutral position to no more than 10 degrees of extension is preferable.
Internal Splint Tendon Transfer
The concept of early internal splint tendon transfers was championed by Burkhalter (61), who advocated pronator teres (PT) transfer to the ECRB early after radial nerve repair or when awaiting spontaneous nerve recovery. Important principles for early internal splint transfers are that the transfer should not detract from remaining hand function, and it should not create a deformity if and when the recipient muscle recovers. During the nerve recovery period, the transfer serves to improve hand function by providing wrist extension during grasp. In contrast to static external splints, the transferred muscle behaves phasically, turning off during release, allowing wrist flexion to occur, and thus permitting hand opening via tenodesis. The transfer is performed end to side, and the ECRB tendon is not detached from its muscle belly in the event that motor function recovers. I have no personal experience with early internal splint transfers but believe that the idea has merit, especially when the prognosis for recovery is poor, when incomplete recovery is predicted, or when an exceedingly prolonged time to recovery is anticipated. For example, it may be particularly appropriate in the setting of a high radial nerve injury or if nerve grafting is required.
Reconstruction of Wrist Extension
Reconstruction of wrist extension is necessary in high radial nerve palsy. In a low palsy, the radial wrist extensors are functional. Reconstruction of wrist extension or stabilization of the wrist in extension can preferably be accomplished by tendon transfer but can also be achieved by tenodesis or arthrodesis if there is a paucity of donors.
The PT, if available, has been accepted virtually universally as the best donor muscle for reconstruction of wrist extension. Controversy exists in the preferred insertion. Some surgeons have recommended insertion into both radial wrist extensors (62,63,64 and 65). Others prefer insertion into only the ECRB, citing that this limits the excessive radial wrist deviation that occurs when the PT is inserted into both radial wrist extensors, especially if the FCU is used for digital extension (37,43,60,66,67,68 and 69). Brand has pointed out that the ECRB insertion still has a radial deviation moment and recommends one of two options to balance the wrist in extension (12). One option is to attach the PT into the ECRL and ECRB but to detach the ECRL from its usual insertion and reinsert it into the base of the fourth metacarpal. The other option is to insert the PT into the ECRB and also create an ulnar-balancing yoke with the FCU and ECU. This is accomplished by transecting the ECU tendon at the musculotendinous junction and attaching the distal stump to the FCU tendon. In effect, the FCU muscle becomes the motor for the FCU and ECU. Although most authors describe an end-to-side insertion of the PT into the wrist extensors, Smith (60) criticized this approach and stressed that a more efficient transfer arises when its action is in a straight line. He preferred to detach the ECRB at the musculotendinous junction and to perform an end-to-end juncture to the PT.
On occasion, a radial nerve injury spares the branch to the ECRL but involves the innervation to the ECRB. This type of injury occurs due to anatomic variation of the branch to the ECRB. In this instance, a severe radial deviation posture results during wrist extension. Smith (60) recommended against using the FCU as a donor for the reconstruction of digital extension, especially in this circumstance. He also advocated transecting the denervated ECRB at the musculotendinous junction and performing tenodesis of the stump to the innervated ECRL tendon with sufficient tension to correct the radial deviation deformity.
Reconstruction of Thumb and Finger Extension
The palmaris longus (PL) is the most broadly accepted donor to reconstruct thumb extension (12,60,64,68,69 and 70). When the PL is vestigial or unavailable, transfer of the ring FDS (12) or by using one FDS to motor the EPL and EIP (59) has been recommended. Although some authors have advocated when the PL is missing to include the EPL in the FCU or flexor carpi radialis (FCR) insertion with the rest of the digital extensors (63,65,71,72), the disadvantage is lack of independent thumb extension.

Smith (60) and Brand and Hollister (12) discussed the importance of thumb metacarpal extension, which is provided by the APL. Brand and Hollister (12) asserted that the APL is truly an extensor, not an abductor, of the thumb metacarpal and that APL function is important for arching of the thumb in pinch. I believe that Smith (60) expressed the same concept by stating that the APL prevents the collapse of the thumb into a swan-neck posture during pinch. Brand, Hollister, and Smith believed that this function should be reconstructed in radial nerve palsy. Smith (60) preferred APL tenodesis to the brachioradialis, which was tensioned to hold the thumb metacarpal in extension. Brand and Hollister (12) recommended transfer of the FCR into the APL, stating that the two muscles are architecturally similar and that the APL insertion has a flexor moment at the wrist, such that wrist flexion would be preserved with the transfer. Although Brand and Hollister (12) have also promoted using the FCR for reconstruction of digital extension, in cases in which the FCR is used for thumb metacarpal extension, they have recommended that the middle-finger FDS be used for digital extension.
Most accounts of PL transfer to the EPL suggest rerouting of the EPL through the subcutaneous tissue on the radial aspect of the wrist to provide the dual effect of thumb extension and metacarpal abduction (12,37,68,69). Tajima has recommended rerouting the EPL through the first dorsal wrist compartment to prevent volar translation and bowstringing of the transfer through the subcutaneous tissue (70). I have no experience with this technique, but it seems rational.
The FCR, FCU, and FDS to the middle and ring fingers have been the most widely recommended donors to reconstruct finger extension (37,43,60,62,64,68,69,73,74,75 and 76). Some of the early tendon transfer surgeons promoted the use of both wrist flexors. Jones advocated a transfer of (a) FCR to EPL; EPB, APL, and EDC of the index finger; and (b) FCU to EDC of the ulnar three digits (63). Bunnell (71) and others suggested a transfer of the FCR to the EPL and the EDC of the index finger and a transfer of the FCU to the ulnar three digital extensors (72). Another recommendation was to transfer the FCR to the EIP, EPL, EPB, and APL and to use the FCU for the EDC (65). Although transferring both wrist flexors to the extensor side improves wrist dorsiflexion, inability to adequately flex the wrist results (60). This causes dysfunction of the digital extension transfers by elimination of the wrist flexion tenodesis effect. With progression of a wrist hyperextension deformity, there is an increasing extensor moment at the wrist, compared to the MCP joints, and the MCP joints fall into flexion. Furthermore, activities such as putting the hand in a pocket or picking up objects from a table top are awkward without active wrist flexion (60).
Contemporary transfers use only one of the two wrist flexors or the FDS for digital extension (Table 1). The FCU transfer for digital extension has been dubbed part of the standard set of transfers for radial nerve palsy (37,69). Considerable controversy exists, however, as to which wrist flexor is better to transfer. I have been convinced by the principles that have been disseminated by Brand and Hollister (12) and others that the FCU is not as expendable as the FCR (or the FDS) (37,60,64). Using the FCU is particularly problematic in low radial nerve palsy when the radial wrist extensors are functioning. With sacrifice of the FCU, a radial-deviation wrist posture can result. Furthermore, strong grasp, which requires ulnar deviation of the wrist, and use of the wrist in a common functional axis (extension-radial deviation and flexion-ulnar deviation) can be disturbed by loss of the FCU. Finally, the tenet that architecturally similar muscles should be used as donors supports the use of the FCR over the FCU for replacement of the EDC (77,78).
FDS transfers
   Brand and Hollister (12) and Chotigavanich (88)
      FDS middle to EDC index to small
      PT to ECRB
      PL (or FDS ring) to EPL
      FCR to APL
   Chuinard et al. (59)
      FDS ring to EPL, EIP
      FDS middle to EDC
      PT to ECRB and ECRL
      FCR to APL and EPB
FCU transfers
   Riordan (67,68), Skoll (86), Schneider (69), Green (37), and Chotigavanich (88)
      FCU to EDC
      PL to EPL
      PT to ECRB (and ECRL)
   Kruft et al. (87) and Dunnet et al. (85)
      FCU to EDC and EPL; PL to EPB (and APL)
      PT to ECRB (and ECRL)
FCR transfers
   Skoll (86), Brand and Hollister (12), and Green (37)
      FCR to EDC
      PL to EPL
      PT to ECRB
   Dunnet et al. (85)
      FCR to EDC and EPL
      PL to EPB
      PT to ECRB
APL, abductor pollicis longus; ECRB, extensor carpi radialis brevis; ECRL, extensor carpi radialis longus; EDC, extensor digitorum communis; EIP, extensor indicis proprius; EPB, extensor pollicis brevis; EPL, extensor pollicis longus; FCR, flexor carpi radialis; FCU, flexor carpi ulnaris; FDS, flexor digitorum superficialis; PL, palmaris longus; PT, pronator teres.
Chuinard et al. (59) popularized transfer of the FDS for restoration of digital extension. The recommended transfer was FDS of the ring finger to the EPL and EIP, with the FDS of the middle finger transferred to the EDC (59). The proposed course of the transfer was through the interosseous

membrane. Some arguments for using the FDS for reconstruction of digital extension include the following:
  • The FDS has longer fiber lengths than the wrist flexors and can accommodate the necessary excursion to substitute for the digital extensors (77,78). This is especially important in the presence of wrist stiffness, which eliminates the tenodesis effect.
  • The course of the transfer through the interosseous membrane is straight, thereby avoiding deviation of the fingers and providing for a more efficient transfer.
  • The FDS transfer avoids sacrifice of the function of the less expendable FCU and allows use of the FCR for another function [such as transfer to the APL (12)].
In my limited experience with this transfer, patients have had difficulty reeducating the digital flexors to be used as extensors, a drawback which Smith (60) highlighted. Brand and Hollister (12) suggest that this phenomenon may be age dependent and is less of a problem in younger patients. Other potential disadvantages of the FDS transfer are possible grip weakness and technical difficulties in passing the transfers through the interosseous membrane, with associated scarring. In the instance in which the FDS is used for digital extension and neither wrist flexor is used as a donor, the transfer for wrist extension (usually the PT muscle) may be overpowered, thus diminishing its efficacy (60).
Brand and Hollister (12) proposed a different combination of FDS transfers than Chuinard et al. (59). Because the FDS to the middle finger is architecturally similar to the EDC, they proposed using it alone for extension of all four fingers. They preferred to have independent thumb extension, transferring the PL or, if absent, the ring FDS to the EPL (12).
Author’s Preferred Method
My preferred transfers for high radial nerve palsy in which wrist, finger, and thumb extension must be restored are (a) PT transferred to the ECRB; (b) PL to the EPL (if PL is absent, I use the ring FDS); (c) and FCR to the EDC.
When restoring wrist extension, my preference is to detach the ECRB at the musculotendinous junction and weave the PT and ECRB together, providing a straight line of action. If I were to consider PT transfer as an internal splint, I would leave the ECRB intact and transfer the PT to the ECRB end to side, as Burkhalter (61) recommended.
I concur with the observations of Tajima (70) that there can be gradual palmar shift of the PL to EPL transfer. I have observed this to lead to excessive palmar abduction of the thumb. In an extreme case, in addition to palmar migration of the transfer at the radial side of the wrist, the EPL eventually subluxated radially over the MCP joint. The transfer became a thumb palmar abductor and MCP joint flexor, and, with the force of the transfer thus dissipated, thumb IP joint extension was poor. After transecting it at the musculotendinous junction, I have preferred to leave the EPL stump in the third dorsal wrist compartment and to insert the PL into it just proximal to Lister’s tubercle. This has provided excellent thumb extension that has not deteriorated over time. Admittedly, I have noted one patient to have difficulty with movements that required simultaneous thumb palmar abduction and IP joint extension. This is because when the EPL remains in the third dorsal wrist compartment, it not only acts as an extensor, but it also retains some function as a thumb adductor. Tajima’s (70) idea of rerouting the EPL transfer through the first dorsal wrist compartment may be a solution to this problem. I have also observed recently the swan-neck thumb posture or metacarpal adduction collapse deformity, as pointed out by Smith, Brand, and Hollister, in a patient with a hyperextensible thumb MCP joint. In the future, I may try APL tenodesis to address this deformity (60).
Although it is well established that muscles in the upper extremity can be trained to “change phase,” my experience is that patients (especially adults) have difficulty with learning to use the FDS for digital extension. For this reason, provided that there are no contraindications or other preferred uses, when a synergistic donor is available and expendable, it is my favored muscle to transfer. Wrist flexion is synergistic with digital extension. The FCR is more expendable and has architecture that is more similar to the EDC than to the FCU. For these reasons, the FCR is my preferred donor for digital extension.
Flexor Carpi Radialis Transfer
Under tourniquet hemostasis, a curvilinear incision is made starting proximally on the radial aspect of the mid-forearm. The incision is carried distally to approximately 2 cm proximal to the radial styloid and then is curved dorsally, distally, and ulnarly to the mid-dorsal wrist at the proximal margin of the extensor retinaculum (Fig. 3A). The sensory radial nerve is identified and protected. The PT insertion beneath the radial wrist extensors is detached, along with an extension of periosteum, to prolong the length of the tendon. The PT fascia is dissected proximally to enhance excursion and to facilitate rerouting of the PT muscle. A transverse 1- to 2-cm incision is made in the proximal wrist crease, extending from the FCR to the PL (Fig. 3B). Care is taken not to injure the palmar cutaneous branch of the median nerve, which lies between these two tendons. The FCR and PL are transected and passed subcutaneously into the radiodorsal forearm incision (Fig. 3C).
The wrist extension transfer is performed first. The ECRB is divided at the musculotendinous junction. The PT is then rerouted superficial to the radial artery, the brachioradialis, and the ECRL and is attached to the ECRB by using a Pulvertaft (79) weave (Fig. 3D). (I use a 3-0 braided nonabsorbable suture.) Tensioning the transfer is critical.

Common wisdom and practice is to err in making tension too tight as opposed to too loose (37,80). I tension the transfer, so that with gravity the wrist is held in neutral or slight extension. Other recommendations include putting maximum tension on the PT and suturing the juncture to the ECRB, with the wrist in 45 degrees of extension (37), or suturing the PT into the ECRB without tension, with the wrist held in 60 degrees of extension. Using these parameters, gravity should allow the wrist to rest in 30 degrees of extension, and 30 degrees of passive flexion should be possible (60). The drawback to making the wrist extension transfer too tight is that it diminishes the efficacy

of the digital extension transfers. With the wrist in too much extension, the MCP joints fall into flexion.
FIGURE 3. A: Intraoperative photo of dorsal forearm incision that is drawn for exposure of the thumb, digital, and wrist extensor tendons, as well as the pronator teres (PT) insertion. B: Intraoperative drawing of incision for exposure of the palmaris longus (PL) and flexor carpi radialis (FCR) tendons. Extreme care is taken not to injure the palmar cutaneous branch of the median nerve that lies between the two tendons. C: Intraoperative photo that shows harvested PT (the most proximal and dorsal muscle), FCR (the muscle in the middle), and PL (the most distal and volar muscle). The latter two muscles have been passed from the volar wrist wound into the dorsal forearm wound. D: Drawing of FCR tendon transfer (the author’s preferred variation). PT is transferred to the detached extensor carpi radialis brevis (ECRB) tendon; FCR is transferred to the extensor digitorum communis (EDC) around the radial side of the forearm; PL is transferred to the extensor pollicis longus (EPL). The EPL tendon remains in situ in the third dorsal wrist compartment or can be rerouted subcutaneously (37) or into the first dorsal compartment (70). E: Intraoperative photo of the FCR-to-EDC juncture (large and dorsal-most muscle transfer) and the PL-to-EPL transfer (small volar muscle). The PT-to-ECRB transfer is deep to the FCR-to-EDC transfer.
Next, attention is directed to restoration of digital extension. If traction on the EDC provides good extension of the small finger, the EDQ does not need to be included in the transfer. However, the most common EDC tendon slip pattern has minimal contribution to the small finger, and, if traction on the EDC does not provide adequate small finger extension, the EDQ should be included (37,81). The EDC and, if necessary, the EDQ tendon slips are transected at their musculotendinous junctions. The index EDC slip is woven into that of the long finger, and the EDC slip is woven into the small finger slip, if present, or the EDQ slip is woven into the ring finger EDC slip, taking care to maintain a balanced cascade. (The relative at rest position of the MCP joints should be progressively smaller in extension in the adjacent ulnar digit, proceeding from the index to the small finger.) The FCR tendon and the slips of the EDC for the long and ring fingers are woven together (Fig. 3D,E). I tension the transfer, so that, with the wrist in neutral, the MCP joints are held in full extension and full composite passive flexion is possible, with the wrist in extension. Others have recommended tensioning the digital extensors such that, with the wrist in 30 degrees of extension, the MCP joints fall to 20 degrees of flexion (60,43).
Lastly, restoration of thumb extension is provided by transferring the PL into the EPL. The EPL is divided at its musculotendinous junction. My practice has been to leave the EPL in situ in the third dorsal wrist compartment and to weave the PL into it just proximal to Lister’s tubercle (Fig. 3D,E). I have tensioned the transfer such that, with the wrist in neutral, the thumb stays in 0 degrees of extension with gravity. With the wrist in extension, the thumb should be able passively to reach the small finger tip. Green (37) recommends suturing the PL into the EPL (rerouted out of the extensor retinaculum in the subcutaneous tissue on the radial side of the wrist) with maximal tension, while the wrist is in neutral position.
Flexor Carpi Ulnaris Transfer
A 6-cm-long incision is made on the volar radial mid-forearm over the insertion of the PT (Fig. 4). The PT is harvested, as was described previously. A volar ulnar incision is made from the mid-forearm and extends distally to the proximal wrist crease at which point the incision curves radially and ends at the PL. The FCU is divided just proximal to the pisiform. The FCU fascia is dissected to the proximal extent of the wound. The PL is transected at the wrist crease. A counter incision is made over the FCU muscle belly in the proximal forearm to further facilitate fascial release of the FCU. A fourth curvilinear incision is made on the middorsal forearm, exposing the EDC and EPL just proximal to the dorsal wrist retinaculum. [The volar radial mid-forearm and dorsal forearm incisions can be incorporated into one incision if large exposure is desired (37).] The PT is transferred to the ECRB, as described previously. The FCU is passed subcutaneously around the ulnar forearm from the volar ulnar wound to the middorsal distal forearm wound. Tendon juncture can be end to side (68,69) or end to end (75). With end-to-side juncture, the FCU can be split longitudinally, with the sides of the split sandwiching the EDC slips. Alternatively, the FCU tendon can be directly woven into the EDC slips in a proximal-to-distal oblique orientation, with tension progressively increasing to each EDC slip from small to index fingers, maintaining the cascade, as was described previously (43,69). End-to-end juncture is similar to that described for the FCR transfer. The EPL transfer is completed, as was described previously. Tension considerations are similar to those that were discussed previously for the wrist and thumb extension transfers. The FCU transfer is tensioned such that the MCP joints are held in 0 degrees of extension with the wrist in 25 degrees of extension (69).
FIGURE 4. Drawing of incisions for flexor carpi ulnaris (FCU) tendon transfers. A: Incision 1 exposes the pronator teres insertion and the juncture site for pronator teres transfer to the extensor carpi radialis brevis. Incision 2 is for harvest and partial fascial release of the FCU. Incision 3 is for more proximal fascial release of the FCU. B: Incision 4 is for the extensor digitorum communis and extensor pollicis longus juncture.
Flexor Digitorum Superficialis Transfer
In this transfer, the usual recommendation is to use the FDS of the middle finger to all slips of the EDC (59).

Although some surgeons have recommended using the ring FDS for the EIP and EPL (59), others have combined the middle FDS to EDC transfer with the PL transfer for thumb extension (12). If the FDS of the middle finger is to be used to extend all four fingers, the transfers to reconstruct thumb and wrist extension are performed as they were described previously.
The middle and, if desired, the ring FDS tendons are harvested through transverse incisions at the MCP joint flexion creases (59). A midvolar forearm incision that is proximal to the pronator quadratus allows retrieval of the FDS tendons by proximal traction. A single fenestration is made in the interosseous membrane, taking care to protect the median nerve proper and the anterior and posterior interosseous neurovascular bundles, and is 4 cm long and as wide as the interosseous space (60). Alternatively, two windows can be made on either side of the anterior and posterior interosseous neurovascular structures (59). After making a dorsal counter incision to expose the EDC, the EIP, and the EPL proximal to the wrist retinaculum, the FDS of the middle finger is routed dorsally through the interosseous space radial to the profundus tendons. If used, the ring finger FDS is passed to the ulnar side of the profundus muscle (59,60). The middle-finger FDS is woven end-to-side into the index finger through small EDC slips (59), or the EDC slips are transected, and an end-to-end juncture is performed (60). The EDC transfer is tensioned, with the wrist in 30 degrees of extension, the fingers fully extended, and no tension on the FDS (60). If the ring finger FDS has been transferred, it can also be joined end to side or end to end to the EPL and EIP with the previously mentioned tension parameters.
Postoperatively, a short arm bulky splint is applied with the wrist in 15 to 30 degrees of extension, the fingers fully extended, and the thumb in radial abduction and extension. The splint is changed to a cast at 1 week postoperatively, after swelling has resolved. At 4 weeks postoperatively, the cast is removed, and full-time Orthoplast splint wear is initiated, with the patient removing the splint during exercise periods. We prefer a static Orthoplast splint with the thumb in radial abduction and full extension, with the wrist at 20 to 10 degrees of extension.
Reeducation is begun at 4 weeks postoperatively. This process can be facilitated if patients have had the opportunity to have preoperative therapy sessions during which they can be educated regarding the postoperative rehabilitation program. If possible, patients can learn to isolate and activate the proposed donors. After transfer of the PT to the ECRB, activation of forearm pronation can initiate wrist extension. The PL-to-EPL transfer is reeducated with activation of wrist flexion and facilitates thumb extension with place-and-hold passive assistance. These transfers, in my experience, are usually rapidly learned. Wrist flexor transfers to reconstruct digital extension are synergistic and, for this reason, are also easily reeducated. Initiation of wrist flexion activates digital extension. Advancement to functional activities potentiates the reeducation process. In fact, Smith advocated functional reeducation over encouragement of the patient to contract donor muscles by trying to use them for their previous function (82). Biofeedback and electrical stimulation can be adjunctive in patients who have difficulty isolating and activating the transferred muscle and seem to be more necessary in those who have been immobilized for long periods after surgery.
Complications of tendon transfers for radial nerve palsy have included rupture of tendon juncture, inappropriate tensioning (too tight or too loose), tendon transfer adhesions, and MCP or wrist joint extension contracture (59). Overtightening of the wrist extension transfer leads to difficulty with MCP joint extension, thus necessitating a lengthening of the transfer (43,60).
Inability to stabilize the wrist in extension because of poor wrist extensor transfer function results in poor grasp and requires transfer revision. If an adequate motor is nonexistent, salvage by wrist extension tenodesis or fusion may be necessary. However, the tenodesis effect that is enabled by wrist motion is lost if revision is elected with wrist fusion or tenodesis, and digital extension may be rendered suboptimally if a wrist flexor was used in the original reconstruction. Revising the digital extension transfer with the FDS, which has a better amplitude match to the digital extensors, may also be necessary (12,43,59,60).
Overtightening or adhesions of the digital extension transfer cause extrinsic extensor tightness, and composite digital flexion is limited. If the MCP joints remain in extension, hook grip is possible, but when the MCP joints are allowed to flex, the IP joints have poor flexion. Secondary MCP joint extension contractures may result. If extrinsic digital extensor tightness persists after transfer, tenolysis or lengthening of the transfer may be necessary. MCP joint capsular release may also be necessary.
The use of outcome assessment questionnaires in the evaluation of radial nerve palsy tendon transfers is lacking. For the most part, the literature has used the elusive terms excellent, good, fair, and poor. Zachary (83) attempted to grade results more objectively by using a demerit system in which percentage points were subtracted for increments of limitations in finger extension, thumb extension, wrist extension,

wrist flexion, and finger flexion. Although several authors subsequently have used his system, which has allowed some comparison of results between studies, Zachary’s criteria have been felt to be too rigid and arbitrary, with little attention to the patient’s assessment of functional outcome (59,75,83,84). Chuinard et al. (59) developed a method to objectify results based on flexion and extension of the fingers and the wrist and abduction and extension of the thumb. This grading scale, like Zachary’s, emphasizes active motion and does not incorporate assessment of functional task performance or patient satisfaction.
Chuinard et al. (59) reviewed 22 patients who had the FDS of the middle finger transferred to the EDC slips and the FDS of the ring finger transferred to the EPL and EIP. PT was transferred to the ECRL and ECRB. The FCR was inserted into the APL. By their assessment criteria, 16 (73%) patients achieved a good or excellent result, and, by Zachary’s grading scheme, the results ranged from 60% to 100%, with an average of 90%.
In a series of 14 patients in whom the FCU was used for finger extension (with a variety of transfers used for thumb and wrist extension), one-half of the patients achieved 100%, by Zachary’s criteria, and the rest were more than 80% (75). All patients achieved useful grip. One-half of the patients had slight radial deviation of the hand that did not interfere with function. In a study that compared transfer of the FCU versus the FDS for digital extension, there was no difference in the Zachary score, which averaged approximately 85% (84).
Raskin and Wilgis (76) compared wrist function in six patients, who had FCU transfer to reconstruct digital extension, to ten healthy volunteers. Range of motion, dynamic power of wrist motion, and work simulation tasks were assessed. Functional wrist power and range of motion were maintained, and there was no significant difference in the ability to perform the work simulation tasks (hammer, saw, pliers, and screwdriver) between the transfer and control groups. Transfer patients were able to ulnarly deviate the wrist, despite loss of the FCU, presumably, via the digital flexors.
In a subset of a larger series, there were 22 patients with radial nerve palsy who were treated with FCU or FCR transfer to the EDC and EPL, PT to the ECRB, and PL to the EPB (85). All patients had improvement in active motion by Chuinard’s criteria, with good to excellent wrist extension in 78%, finger extension in 57%, thumb extension in 50%, and thumb abduction in 87%. Hand function was considered good by most patients; however, 64% reported impaired fine manipulation, 55% had difficulty grasping or releasing large objects, and 77% complained of early fatigue. Power grip and pinch were approximately one-half of the contralateral side. Although the numbers were too small for fair comparison, the authors did not note any differences in function of the FCU versus the FCR transfers, except a trend for better power grasp in those who had the FCU transfer.
Skoll et al., in a South African cohort, retrospectively reviewed 22 patients who were treated for high radial nerve palsy with transfers of the FCU to the EDC, the PT to the ECRB, and the PL to the rerouted EPL (86). For posterior interosseous nerve palsy, the FCR was used instead of the FCU. Patients in this study had what the authors believed to be limited follow-up (an average of eight follow-up visits) and rehabilitation, due to poor patient compliance. The authors believed that their study was unique in this regard, but the intensity of postoperative aftercare in their patients was not unlike what is available to patients in the current state of our third-party payer system in the United States. It is interesting to note that the results did not appear to be compromised when compared to other reports. By Chuinard’s criteria, there were 19 good to excellent results for wrist extension, with 20 for wrist flexion, 20 for finger extension, 20 for thumb extension, and 21 for thumb abduction (86). Power grip was approximately one-half that of the contralateral side. Ten of the 15 patients with the FCU transfer and two of the seven patients with the FCR transfer had a wrist radial deviation posture at rest. All but one patient were able to perform activities of daily living. The Jebson test took an average of 1.46 times longer than was expected. Patient satisfaction was graded at 6.5 out of 10, with the most common complaints being stiffness and lack of dexterity and power. Thirteen of 17 patients who were employed preinjury were able to return to their previous work. However, only one of the seven heavy manual laborers returned to his preinjury occupation.
In a series of 43 patients with radial nerve palsy who were treated with a variation of the FCU transfer (d—Aubigne procedure), 31 good to excellent results, by Zachary’s criteria, were observed, with 41 patients (95%) expressing sufficient satisfaction that they would undergo the procedure again (87). The authors claimed that close to one-half of the patients had some independent finger extension, with selective recruitment noted on electromyographic examination. It is difficult to understand the mechanism of this because the transferred motor only has a single tendon that is joined to the multiple slips of the EDC.
Chotigavanich (88) reviewed 50 patients (43 high radial nerve palsies and seven low palsies) who underwent FCU or middle finger FDS transfer to the EDC. In all patients, the PL was transferred to the EPL, and, in 40 patients, the PT was transferred to the ECRL and ECRB. All regained useful hand function. Manual muscle testing demonstrated grade 4 or 5 wrist extension strength in 82%, finger MCP joint extension in 90%, and thumb extension in 92%. Good or excellent results were achieved by Tajima’s finger, thumb, and wrist motion criteria (88) in 43 patients. Five cases with FCU transfer had radial wrist deviation. No problems with grip weakness were noted in patients in whom the FDS was used as a donor, but the author stressed the use of only one FDS.
In summary, most reports have generally demonstrated improved thumb, finger, and wrist extension. Although measurement of subjective and functional criteria has been lacking, in those studies that include some discussion of subjective or functional evaluation, hand function was generally

considered by the patients to be good (85). Useful hand function was regained (88), and patients were able to perform activities of daily living (86). Most regained useful grasp (75). Patients were sufficiently satisfied to undergo the procedure again (87). My experience has been that tendon transfers for radial nerve palsy are rewarding, with patients generally realizing much improved function (Fig. 5).
FIGURE 5. Case of radial nerve palsy that was due to a humeral diaphyseal fracture (A). The patient’s radial nerve was not explored, and the fracture was allowed to heal. She presented to the author with a high radial nerve palsy 5 years after injury, beyond time that would be expected for nerve repair to have a chance for success. Tendon transfers of pronator teres to extensor carpi radialis brevis, flexor carpi radialis to extensor digitorum communis, and palmaris longus to extensor pollicis longus were performed. Several months after surgery, she was able to independently extend her thumb (B) and simultaneously extend her wrist and fingers (C).
Some limitations that have been noted by reconstructed patients have included difficulty grasping or releasing large objects and early fatigue. Power grip and pinch were approximately one-half of the contralateral side. Stiffness, loss of dexterity, and limited ability to return to heavy labor have all been reported.
1. Gerwin M, Hotchkiss RN, Weiland AJ. Alternative operative exposures of the posterior aspect of the humeral diaphysis. J Bone Joint Surg 1996;78:1690–1695.
2. Abrams RA, Zeits RJ, Leiber RL, et al. Anatomy of the radial nerve motor branches in the forearm. J Hand Surg 1997;22:232–237.
3. Sunderland S. Metrical and non-metrical features of the muscular branches of the radial nerve. J Comp Neurol 1946;85:93–111.
4. Fuss FK, Wurzl GH. Radial nerve entrapment at the elbow: surgical anatomy. J Hand Surg 1991;16:742–747.
5. Prasartritha T, Liupolvanish P, Rojanakit A. A study of the posterior interosseous nerve (PIN) and the radial tunnel in 30 Thai cadavers. J Hand Surg 1993;18:107–112.
6. Abrams RA, Brown RA, Botte MJ. The superficial branch of the radial nerve: an anatomical study with surgical implications. J Hand Surg 1992;17:1037–1041.
7. Spinner M. The arcade of Frohse and its relationship to posterior interosseous nerve paralysis. J Bone Joint Surg 1968;50:809–812.
8. Thomas SJ, Yakin DE, Parry BR, et al. The anatomical relationship between the posterior interosseous nerve and the supinator muscle. J Hand Surg 2000;25:936–941.
9. Linell EA. The distribution of nerves in the upper limb, with reference to variabilities and their clinical significance. J Anat 1921;55:79–112.
10. Seddon HJ, Medawar PB. Rate of regeneration of peripheral nerves in man. J Physiol 1943;102:191–215.
11. Labosky DA, Waggy CA. Apparent weakness of median and ulnar motors in radial nerve palsy. J Hand Surg 1986;11:528–533.

12. Brand PW, Hollister A. Operations to restore muscle balance to the hand. In: Brand PW, Hollister A, eds. Clinical mechanics of the hand, 2 ed. St. Louis: Mosby–Year Book, 1993:180–189.
13. Omer GE. Injuries of the nerves to the upper extremities. J Bone Joint Surg 1974;56:1615–1624.
14. Bleeker WA, Nijsten MW, Ten Duis HJ. The treatment of humeral shaft fractures related to associated injuries: a retrospective study of 237 patients. Acta Orthop Scand 1991;62:148–153.
15. Foster RJ, Swiontkowski MF, Back AW, et al. Radial nerve palsy caused by open humeral shaft fractures. J Hand Surg 1993;18:121–124.
16. Holstein A, Lewis GB. Fractures of the humerus with radial nerve paralysis. J Bone Joint Surg 1963;45:1382–1388.
17. Kettlecamp DB, Alexander H. Clinical review of radial nerve injury. J Trauma 1967;7:424–432.
18. Packer JW, Foster RR, Garcia A, et al. The humeral fracture with radial nerve palsy: is exploration warranted? Clin Orthop 1972;88:34–38.
19. Pollock FH, Drake D, Bovill EG, et al. Treatment of radial neuropathy associated with fractures of the humerus. J Bone Joint Surg 1981;63:239–243.
20. Samardzic M, Grujicic D, Mikinkovic ZB. Radial nerve lesions associated with fractures of the humeral shaft. Injury 1990;21:220–222.
21. Shaw JL, Sakellarides H. Radial nerve paralysis associated with fractures of the humerus: a review of forty-five cases. J Bone Joint Surg 1967;49:899–902.
22. Sonneveld GL, Patka P, van Mourik LC, et al. Treatment of fractures of the shaft of the humerus accompanied by paralysis of the radial nerve. Injury 1987;18:404–406.
23. Szalay EA, Rockwood CA. The Holstein-Lewis fracture revisited. Orthop Trans 1983;7:516.
24. Shah JJ, Bhatti NA. Radial nerve paralysis associated with fractures of the humerus. Clin Orthop 1983;172:171–176.
25. Culp RW, Osterman AL, Davidson RS, et al. Neural injuries associated with supracondylar fractures of the humerus in children. J Bone Joint Surg 1990;72:1211–1215.
26. Dutkowsky JP, Kasser JR. Nerve injury associated with fractures in children. In: Gelberman RH, ed. Operative nerve repair and reconstruction. Philadelphia: JB Lippincott Co, 1991:635–653.
27. Dobbie RP. Avulsion of the lower biceps brachii tendon: analysis of fifty-one previously reported cases. Am J Surg 1941;51:662–683.
28. Failla JM, Amadio PC, Morrey BF. Post-traumatic proximal radio-ulnar synostosis: results of surgical treatment. J Bone Joint Surg 1989;71:1208–1213.
29. Louis DS, Hankin FM, Eckenrode JF, et al. Distal biceps brachii tendon avulsion: a simplified method of operative repair. Am J Sports Med 1986;14:234–236.
30. Mekhail A, Ebraheim N, Jackson W, et al. Vulnerability of the posterior interosseous nerve during proximal radius exposures. Clin Orthop 1995;315:199–208.
31. Strachan JC, Ellis BW. Vulnerability of the posterior interosseous nerve during radial head resection. J Bone Joint Surg 1971;53:320–323.
32. Young C, Hudson A, Richards R. Operative treatment of palsy of the posterior interosseous nerve of the forearm. J Bone Joint Surg 1990;72:1215–1219.
33. Ishikawa H, Hirohata K. Posterior interosseous nerve syndrome associated with rheumatoid synovial cysts of the elbow joint. Clin Orthop 1990;254:134–139.
34. Millender LH, Nalebuff EA, Holdsworth DE. Posterior interosseous nerve syndrome secondary to rheumatoid synovitis. J Bone Joint Surg 1973;46:375–377.
35. Loren GJ, Shoemaker SD, Burkholder TJ, et al. Human wrist motors: biomechanical design and application to tendon transfers. J Biomech 1996;29:331–342.
36. Sunderland S. Decision making in the clinical management of nerve injury and repair. In: Sunderland S, ed. Nerve injuries and their repair: a critical appraisal. Edinburgh, UK: Churchill Livingstone, 1991:421–426.
37. Green DP. Radial nerve palsy. In: Green DP, Hotchkiss RN, Pederson WC, eds. Green’s operative hand surgery. New York: Churchill Livingstone, 1999:1481–1493.
38. Dabezies EJ, Banta CJ, Murphy CP, et al. Plate fixation of the humeral shaft for acute fractures, with and without radial nerve injuries. J Orthop Trauma 1992;6:1–13.
39. Bostman O, Bakalim G, Vainionpaa S, et al. Immediate radial nerve palsy complicating fracture of the shaft of the humerus: when is early exploration justified? Injury 1985;16:499–502.
40. Amillo S, Barrios RH, Martinez-Peric R, et al. Surgical treatment of the radial nerve lesions associated with fractures of the humerus. J Orthop Trauma 1993;7:211–215.
41. Yang KH, Han DY, Kim HJ. Intramedullary entrapment of the radial nerve associated with humeral shaft fracture. J Orthop Trauma 1997;11:224–226.
42. Ward EF, Savoie FH, Hughes JL. Fractures of the diaphyseal humerus. In: Browner BD, Jupiter JB, Levine AM, et al., eds. Skeletal trauma. Philadelphia: WB Saunders, 1998:1523–1547.
43. Wheeler DR. Reconstruction for radial nerve palsy. In: Peimer CA, ed. Surgery of the hand and upper extremity. New York: McGraw-Hill, 1996:1363–1379.
44. Campbell CC, Waters PM, Emans JB, et al. Neurovascular injury and displacement in type III supracondylar humerus fractures. J Pediatr Orthop 1995;15:47–52.
45. Dormans JP, Squillante R, Sharf H. Acute neurovascular complications with supracondylar humerus fractures in children. J Hand Surg 1995;20:1–4.
46. McGraw JJ, Akbarnia BA, Hanel DP, et al. Neurological complications resulting from supracondylar fractures of the humerus in children. J Pediatr Orthop 1986;6:647–650.
47. Martin DF, Tolo VT, Sellers DS, et al. Radial nerve laceration and retraction associated with a supracondylar fracture of the humerus. J Hand Surg 1989;14:542–545.
48. Sairyo K, Henmi T, Kanematsu Y, et al. Radial nerve palsy associated with slightly angulated pediatric supracondylar humerus fracture. J Orthop Trauma 1997;11:227–229.
49. Bado JL. The Monteggia lesion. Clin Orthop 1967;50:71–80.
50. Boyd DH, Boals JC. The Monteggia lesion. Clin Orthop 1969;66:94–100.
51. Bruce HE, Harvey JP, Wilson JC. Monteggia fractures. J Bone Joint Surg 1974;56:1563–1576.
52. Jessing P. Monteggia lesions and their complicating nerve damage. Acta Orthop Scand 1975;46:601–609.
53. Ring D, Jupiter JB, Simpson NS. Monteggia fractures in adults. J Bone Joint Surg 1998;80:1733–1744.

54. Austin R. Tardy palsy of the radial nerve from a Monteggia fracture. Injury 1975;7:202–204.
55. Lichter RL, Jacobsen T. Tardy palsy of the posterior interosseous nerve with a Monteggia fracture. J Bone Joint Surg 1975;57:124–125.
56. Morris AH. Irreducible Monteggia lesions with radial nerve entrapment. J Bone Joint Surg 1974;56:1744–1746.
57. Siegel DB, Gelberman RH. Peripheral nerve injuries associated with fractures and dislocations. In: Gelberman RH, ed. Operative nerve repair and reconstruction. Philadelphia: JB Lippincott Co, 1991:619–633.
58. Colditz JC. Splinting for radial nerve palsy. J Hand Ther 1987;1:18–23.
59. Chuinard RG, Boyes JH, Stark HH, et al. Tendon transfers for radial nerve palsy: use of superficialis tendons for digital extension. J Hand Surg 1978;3:560–570.
60. Smith RJ. Tendon transfers to restore wrist and digit extension. In: Smith RJ, ed. Tendon transfers of the hand and forearm. Boston: Little, Brown and Company, 1987, p. 35–56.
61. Burkhalter WE. Early tendon transfer in upper extremity peripheral nerve injury. Clin Orthop 1974;104:68–79.
62. Chuinard RG, Boyes JH, Stark HH, et al. Tendon transfers for radial nerve palsy: use of superficialis tendons for digital extension. J Hand Surg 1978;3:560–570.
63. Jones R. Tendon transplantation in cases of musculospiral injuries not amenable to suture. Am J Surg 1921;35:333–335.
64. Starr CL. Army experiences with tendon transference. J Bone Joint Surg 1922;4:3–21.
65. White WL. Restoration of function and balance of the wrist and hand by tendon transfers. Surg Clin North Am 1960;40:427–459.
66. Reid RL. Radial nerve palsy. Hand Clin 1988;4:179–185.
67. Riordan DC. Tendon transfers for median, ulnar or radial nerve palsy. Hand 1969;1:42–46.
68. Riordan DC. Tendon transfers in hand surgery. J Hand Surg 1983;8:748–753.
69. Schneider LH. Tendon transfer for radial nerve palsy. In: Gelberman RH, ed. Operative nerve repair and reconstruction. Philadelphia: JB Lippincott Co, 1991:697–705.
70. Tajima T. Tendon transfer in radial nerve palsy. Recommended choices based on retrospective analysis of methods used and their follow-up. In: Hunter JM, Schneider LA, Mackin EJ. Tendon surgery of the hand. St. Louis: CV Mosby, 1987:413–418.
71. Bunnell S. Tendons. In: Bunnell S, ed. Surgery of the hand. Philadelphia: JB Lippincott Co, 1944:295.
72. Lucky CA, McPherson SR. Tendinous reconstruction of the hand following irreparable injury to the peripheral nerves and brachial plexus. J Bone Joint Surg 1947;29:560.
73. Adams J, Wood VE. Tendon transfers of irreparable nerve damage in the hand. Orthop Clin North Am 1981;12:403–432.
74. Beasley RW. Tendon transfers for radial nerve palsy. Orthop Clin North Am 1970;1:439–445.
75. Moberg E, Nachemson A. Tendon transfers for defective long extensors of the wrist and fingers. Acta Chir Scand 1967;133:31–34.
76. Raskin KB, Wilgis S. Flexor carpi ulnaris transfer for radial nerve palsy: functional testing of long-term results. J Hand Surg 1995;20:737–742.
77. Brand PW, Hollister A. Muscles: the motors of the hand. In: Brand PW, Hollister A, eds. Clinical mechanics of the hand, 2 ed. St. Louis: Mosby–Year Book, 1993:13–33.
78. Lieber RL, Jacobson MD, Fazeli BM, et al. Architecture of selected muscles of the arm and forearm: anatomy and implications for tendon transfer. J Hand Surg 1992;17:787–798.
79. Pulvertaft RG. Suture materials and tendon junctures. Am J Surg 1965;109:346–352.
80. Friden J, Lieber RL. Evidence for muscle attachment at relatively long lengths in tendon transfer surgery. J Hand Surg 1998;23:105–110.
81. von Schroeder HP, Botte MJ. Anatomy of the extensor tendons of the fingers: variations and multiplicity. J Hand Surg 1995;20:27–34.
82. Smith RJ. Tendon transfers: post-operative care. In: Smith RJ, ed. Tendon transfers of the hand and forearm. Boston: Little, Brown and Company, 1987:314–317.
83. Zachary RB. Tendon transplantation for radial paralysis. Br J Surg 1946;23:350–364.
84. Thomsen M, Rasmussen KB. Tendon transfers for defective long extensors of the wrist and fingers. Scand J Plast Reconstr Surg 1969;3:71–78.
85. Dunnet WJ, Housden PL, Birch R. Flexor to extensor transfers in the hand. J Hand Surg 1995;20:26–28.
86. Skoll PJ, Hudson DA, de Jager W, et al. Long term results of tendon transfers for radial nerve palsy in patients with limited rehabilitation. Ann Plast Surg 2000;45:122–126.
87. Kruft S, von Heimburg D, Reill P. Treatment of irreversible lesion of the radial nerve by tendon transfer: indication and long-term results of the Merle d’Aubigne procedure. Plast Reconstr Surg 1997;100:610–616; discussion, 617–618.
88. Chotigavanich C. Tendon transfer for radial nerve palsy. Bull Hosp Joint Dis Orthop Inst 1990;50:1–10.