CHAPTER 80 PRINCIPLES OF TENDON TRANSFERS
DOUGLAS M. SAMMER
A tendon transfer is the re-routing of a functioning muscle–tendon unit (MTU) (Table 80.1) to a new insertion, in order to restore a function that has been lost. Tendon transfers were first developed in the 19th century to restore ambulation in patients with poliomyelitis. During the subsequent World Wars, thousands of soldiers returned home with upper extremity nerve injuries.1 This influx of patients with upper extremity nerve palsies coincided with the development of hand surgery as a surgical specialty. Over the ensuing decades, tendon transfer techniques were adapted and refined for use in the upper extremity by pioneers Bunnell, Boyes, Brand, Burkhalter, Riordan, Zancolli, and others.
Today, the most common indication for a tendon transfer in the upper extremity is a nerve injury that has no potential for recovery.2 This includes irreparable injuries like nerve root avulsions, failed nerve repairs or reconstructions, or nerve injuries that present too late for recovery due to motor end-plate fibrosis. Other common indications include loss of tendon or muscle substance from trauma, tendon rupture (such as in rheumatoid arthritis), and central neurologic deficits (e.g., spinal cord injuries, stroke, and cerebral palsy).3 Although historically important, leprosy, a mycobacterial granulomatous disease that affects the peripheral nerves, is a rare indication for tendon transfer today.4
PRINCIPLES OF TENDON TRANSFER
Over the last century, a number of principles have been established to guide the performance of tendon transfers. Although adhering to these guidelines does not guarantee success, to ignore them inevitably results in failure.
Often, patients with upper extremity nerve palsies will present with stiff joints. Performing a tendon transfer in this setting will not improve function. Joints must be supple prior to tendon transfer.Hand therapy or surgical release may be required to maximize passive motion in preparation for a tendon transfer. It should be noted that a joint or contracture release should never be performed at the same time as the tendon transfer. The postoperative management of a joint release includes immediate mobilization and prolonged rehabilitation in order to restore passive motion, whereas a tendon transfer must be immobilized for 3 to 4 weeks to allow tendon healing.
Soft Tissue Equilibrium
The soft tissue bed through which a tendon transfer will be routed should reach “equilibrium” prior to performing the transfer.4 This means that it should be free of edema, inflammation, or scar, so that the tendon transfer can glide freely. A tendon transfer that passes through an inflamed or scarred bed will develop adhesions, reducing the effectiveness of the transfer. At times, it is necessary to route the transfer along a non-standard path in order to avoid an area of scar. If the area of scar is extensive and cannot be avoided, it may be necessary to resurface this area with a fasciocutaneous flap prior to performing the tendon transfer.
The donor MTU should have enough excursion, or linear movement, to achieve the desired motion at the target joint. In other words, the excursion of the donor MTU should be equal to or greater than that of the MTU it is replacing. A good rule of thumb is that the extrinsic finger flexors have approximately 70 mm of excursion, the extrinsic finger extensors have approximately 50 mm of excursion, and the extrinsic wrist motors have approximately 30 mm of excursion.5
Unfortunately, it is not always possible to match donor and recipient excursion. In some cases, the tenodesis effect can be employed by the patient to augment the effective excursion of the donor MTU. For example, the flexor carpi radialis (FCR) is often transferred to the extensor digitorum communis (EDC) to restore finger metacarpophalangeal (MCP) extension. However, the FCR only has 30 mm of excursion, whereas the EDC normally has 50 mm of excursion. The excursion of the FCR is therefore inadequate to fully extend the MCP joints. The patient can overcome this lack of excursion via the tenodesis effect, by simultaneously flexing the wrist and extending the fingers. Wrist flexion increases the distance between the origin and insertion of the tendon transfer, resulting in greater effective excursion and full MCP extension.
Appropriate Strength of Donor
The strength of the donor MTU should be matched to that of the MTU whose function is being restored. This means that the donor MTU must be strong enough to achieve the desired movement. A weak donor MTU will not be able to move the target joint through a functional range of motion, particularly if there is any stiffness, or if there is a strong antagonist. The palmaris longus (PL), for example, does not have adequate strength to power wrist extension. On the other hand, an excessively strong donor MTU results in muscle imbalance and abnormal posture. A transfer of the brachioradialis (BR) to the extensor pollicis longus (EPL), for example, would result in an extension contracture or an abnormal position of the thumb at rest.
When considering potential donor MTUs for transfer, it is more practical to compare relative muscle strength than it is to compare absolute muscle strength.2 The strongest donor MTUs are the BR and the flexor carpi ulnaris (FCU), which each have a relative strength of 2 units. The FCR, the wrist extensors, the finger flexors, and the pronator teres (PT) each have a relative strength of 1 unit. The finger extensors are weaker, with a relative strength of 0.5 units each. The weakest donor MTUs are the polmaris longus (PL) and the thumb extensors and abductors, each of which has a relative strength of 0.1 units. Ideally, the relative strength of the donor MTU should match that of the recipient MTU.
It is also important to choose a donor MTU that has not been weakened by injury or denervation. In general, a donor MTU will lose up to one grade of motor strength simply by being transferred.6 In some situations, such as brachial plexus palsy, the availability of donor MTUs may be severely limited. Although it can be tempting to use a donor MTU that has recovered function after initial denervation or injury, this should be avoided if at all possible.
It is essential to consider the potential functional deficit that will be created by a tendon transfer. It is of little use to restore one function but lose another equally important function. Fortunately, there is ample redundancy built into the hand and forearm. For example, there are two wrist flexors and three wrist extensors. The FCR or the FCU can be transferred without losing wrist flexion, and two of the three wrist extensors can be transferred without compromising wrist extension. The PL is completely redundant, and the extensor indicis proprius (EIP) and extensor digiti minimi (EDM) are excellent donors whose harvest results in minimal donor deficit. In addition, each finger has two flexors, the flexor digitorum profundus (FDP) and flexor digitorum superficialis (FDS). The FDS is often used as a donor MTU, and the finger retains flexion via the intact FDP.
Straight Line of Pull
A tendon transfer that has a direct path to its insertion is most effective. Any direction change or pulley decreases the force of the transfer. However, there are instances in which a direct line of pull is not ideal. For example, a PT to extensor carpi radialis brevis (ECRB) transfer is commonly performed to restore wrist extension in patients with radial nerve palsy. The transfer is typically performed in an end-to-end fashion, which creates a straight line of pull. However, if there is a possibility of ECRB recovery in the future, the PT is transferred in an end-to-side fashion into the ECRB tendon. Although this results in an indirect line of pull, the end-to-side insertion allows the ECRB to participate in wrist extension if it recovers function in the future.7 In other situations, the required line of pull cannot be achieved without a direction change. For example, opponensplasties are routed from the level of the pisiform toward the abductor pollicis brevis (APB) insertion, a line of pull that produces thumb opposition. In some cases, this line of pull cannot be achieved without routing the tendon transfer around a pulley. Although this direction change weakens the transfer, it is necessary to achieve opposition.
The original function of the donor MTU should be synergistic with the function that is being restored.8 A tendon transfer that is synergistic, as opposed to antagonistic, is easier for the patient to learn to use. Synergy refers to certain movements that are typically combined during routine hand use. For example, wrist extension and finger flexion are synergistic for grasping, whereas wrist flexion and finger extension are synergistic. When a wrist flexor is transferred to restore finger extension (FCR to EDC transfer), the patient can learn to use the transfer without much difficulty. On the other hand, if a wrist extensor were to be transferred to the finger extensors, the patient may have difficulty learning to use the tendon transfer in a natural manner. Although a synergistic transfer is ideal, it is not always possible. Furthermore, it should be noted that certain donor MTUs, such as the FDS, are able to adapt to a new function readily, whether that function is synergistic or not.
Single Transfer, Single Function
Finally, a single tendon transfer should only perform a single function. Attempting to restore multiple functions with a single donor MTU will result in loss of strength and motion. The exception to this rule is that a single donor MTU may be used to restore the same movement in multiple digits. For example, it is acceptable to use the FDS or FCR to restore MCP extension for all four fingers. However, the FDS or FCR would be inadequate to restore both wrist and finger extensions.
The moment arm of a tendon transfer affects how much rotation will occur at the joint, and will affect the torque generated. The moment arm is determined by the distance between the joint axis of rotation and the tendon that crosses the joint. A tendon that passes far from the joint axis of rotation or inserts far from the joint will have a large moment arm, whereas a tendon that lies close to the joint axis of rotation and inserts close to the joint will have a small moment arm. A tendon transfer with a large moment arm will generate greater torque, but at the expense of the arc of motion (greater muscle excursion will be required for a given degree of rotation). A tendon transfer with a smaller moment arm will have an increased arc of motion (less muscle excursion is required for a given degree of rotation), but the transfer will not generate as much torque. In many cases, the insertion point of a tendon transfer is determined by the normal insertion of the recipient tendon. However, there are instances in which the surgeon can choose the insertion point of the tendon transfer. Understanding the concept of the moment arm will help the surgeon determine the optimal insertion point, balancing the needs for joint rotation and generation of torque.
Setting the tension of the tendon transfer is the most critical and difficult part of the operation. Ideally, a tendon transfer should be tensioned in such a way as to maximize actin– myosin overlap. Unfortunately, it is impossible to determine this intraoperatively, although research is being conducted into using laser diffraction intraoperatively to determine the optimum tension for a tendon transfer.9,10 The pragmatic solution is that the tendon transfer should be set at a tension as close as possible to the donor MTUs preoperative resting tension. The donor muscle belly is marked at regular intervals before dividing its insertion, and the tendon transfer is tensioned in such a way as to restore the distance between the intervals. On the other hand, many authors recommend tensioning a tendon transfer substantially tighter than the donor MTUs resting tension. This is because a tendon transfer tends to loosen or stretch out during rehabilitation. However, a tendon transfer that is set too loosely will not tighten postoperatively.
RADIAL NERVE PALSY
Radial nerve palsy results in loss of wrist extension, finger MCP extension, thumb abduction, thumb extension, and thumb retropulsion. In addition, patients notice a substantial decrease in grip strength due to the inability to stabilize the wrist during grip. Loss of supinator function is compensated by the biceps brachii and by shoulder rotation. From a sensory standpoint, the loss of sensibility is not critical. Furthermore, there is some overlap with the lateral antebrachial cutaneous nerve. Radial nerve palsy can be categorized as high or low. High radial nerve palsy is a nerve injury proximal to the elbow, in which all of the above functions are lost. Low radial nerve palsy, on the other hand, occurs with an injury distal to the elbow, in which only the posterior interosseous nerve (PIN) is injured. The BR remains intact, and wrist extension is preserved because the branch to the extensor carpi radialis longus (ECRL) arises proximal to the take-off of the PIN (innervation to the ECRB is variable). The goals of tendon transfer are restoration of finger MCP extension, thumb extension and radial abduction, and wrist extension in cases of high radial nerve palsy.
All of the median and ulnar nerve innervated MTUs may be considered potential donors for tendon transfer in patients with radial nerve palsy. Although many different sets of tendon transfers have been described for radial nerve palsy, the PT to ECRB transfer for restoration of wrist extension is nearly universal (Figure 80.1). This transfer results in minimal, if any, donor deficit, because the transferred PT continues to act as a forearm pronator after transfer. Insertion on the ECRB as opposed to the ECRL is preferred, in order to minimize radial deviation of the wrist. This transfer is typically performed in an end-to-end fashion. However, if there is potential for recovery of radial nerve function, the transfer is performed in an end-to-side fashion, allowing the ECRB to contribute to wrist extension should it become reinnervated (Figure 80.1). In fact, the end-to-side PT to ECRB transfer can be used as an “internal splint” to restore wrist extension while the radial nerve is recovering.7,11,12
Finger MCP extension is restored by transfer of the FCR, FCU, or FDS to the EDC. Although the FCU transfer was one of the first described and is still used today (Figure 80.2), the FCR and FDS transfers are preferred over the FCU by many surgeons. This is because transfer of the FCU results in loss of the functionally critical “dart-throwing motion.” In addition, in the wrist without a functioning ECU, transfer of the FCU removes the only remaining ulnar-sided wrist motor, leading to wrist imbalance and radial deviation. The FCR is a good donor for restoration of finger MCP extension. It is expendable (wrist flexion is maintained by the FCU), and its use does not result in loss of the dart-throwing motion or in radial deviation. The primary limitation of the FCR is that its excursion (approximately 33 mm) is inadequate to provide full MCP extension. However, the tenodesis effect (wrist flexion with concomitant MCP extension) can be used to bring the MCP joints into full extension after an FCR to EDC transfer and is easily learned by the patient. The FDS is also a good donor for restoration of finger MCP extension. It has excellent excursion (approximately 70 mm), and flexion of the donor finger is preserved by the remaining intact FDP. The primary disadvantage of the FDS to EDC transfer is that some grip strength is lost. In the patient with a fused wrist who cannot employ the tenodesis effect, the FDS is the preferred donor MTU.
For restoration of thumb extension and radial abduction, the PL or an FDS can be transferred to the EPL (Figure 80.3). The EPL is usually re-routed and allowed to lie in a more radial and volar position. This results in restoration of radial abduction as well as extension, at the expense of retropulsion. Alternatively, the FDS can be transferred to both the EPL and the EIP in order to restore simultaneous thumb and index finger extension, a functionally useful combination of movements.
FIGURE 80.1. PT to ECRB transfer, end-to-side, for restoration of wrist extension.
Over the years, three “standard” sets of tendon transfers have been established for reconstruction of radial nerve palsy: the FCR transfer,12–14 the FCU transfer,15,16 and the superficialis transfer.17,18All three sets of transfers employ the PT to ECRB transfer for restoration of wrist extension. The FCR transfer involves FCR to EDC transfer for MCP extension, and PL to re-routed EPL transfer for thumb extension. The FCU transfer is the same, except the FCU is used in place of the FCR to restore MCP extension. Finally, in the superficialis transfer, the ring FDS is transferred to the EPL and EIP for simultaneous thumb and index extension, and the long FDS is transferred to the remaining digital extensors. The FCR is transferred to the abductor pollicis longus and EPB to restore thumb MCP extension and radial abduction. The author’s preference is to use the FCR transfer in patients with intact wrist flexion and the superficialis transfer in patients who have undergone a wrist arthrodesis.
FIGURE 80.2. FCU to EDC transfer, end-to-side, for restoration of finger MCP extension.
For the FCR set of transfers, a curvilinear longitudinal incision is made along the course of the FCR in the distal two thirds of the forearm. Skin flaps are elevated, protecting the palmar cutaneous branch of the median nerve. The fascia is incised and the PL and FCR are identified and mobilized along their tendinous portion. Proximal retraction is applied to the PL and FCR tendons with the wrist in flexion, and they are divided at the distal wrist flexion crease (Figure 80.4A). Dissection continues proximally, mobilizing the distal half of the muscle bellies while preserving their innervation and blood supply. In the proximal aspect of the incision, the PT insertion on the radius is identified. The PT tendon, which is short, is harvested in continuity with a 4 cm cuff of periosteum in order to have sufficient length for the Pulvertaft weave.
A dorsal midline longitudinal incision is made over the distal half of the forearm. Skin flaps are elevated exposing the extensor retinaculum and the musculotendinous junctions of the extrinsic extensors, while protecting the superficial sensory branch of the radial nerve. The EPL and the EDC tendons are identified. The EPL is divided proximal to the extensor retinaculum and is mobilized in a proximal to distal direction, releasing it from the third extensor compartment. A subcutaneous tunnel is created around the radial border of the wrist, deep to the subcutaneous adipose tissue, and directly on the antebrachial fascia. The EPL is passed through this tunnel to the volar incision in preparation for Pulvertaft weave to the PL (Figure 80.4B). The tension on the PL to EPL transfer is set with the wrist in neutral position and with maximum tension on both the PL and EPL tendons. Next, a more proximal subcutaneous tunnel is made along the radial border of the forearm, and the FCR is passed to the dorsal incision. The FCR to EDC weave is performed proximal to the extensor retinaculum in an end-to-side fashion (Figure 80.4C). Tension is set with the wrist in neutral and the MCP joints in full extension. The proximal half of the extensor retinaculum can be divided if it inhibits full excursion of the Pulvertaft weave. Alternatively, the EDC tendons can be divided proximally, released from the extensor retinaculum and weaved end-to-end to the FCR in order to create a straighter line of pull. Although a straight line of pull is advantageous, it can result in undesired radial deviation. Finally, the PT is passed through a subcutaneous tunnel, superficial to the BR and ECRL, and weaved into the ECRB just distal to its musculotendinous junction. If the periosteal extension is thin and the Pulvertaft weave is of questionable strength, it can be reinforced with a strip of ECRL (preferred) or BR tendon. The PT to ECRB transfer is tensioned with the wrist in full extension. The PT to ECRB transfer should be tensioned last, so that the surgeon can passively flex and extend the wrist, using the tenodesis effect to evaluate the tension of the FCR to EDC transfer. With the wrist in flexion, the MCP joints should move into full extension or slight hyperextension. With the wrist in extension, the surgeon should be able to passively flex the MCP and interphalangeal (IP) joints without difficulty, touching the fingertips to the palm. If the MCP joints do not move into full extension with wrist flexion, the finger extensions are too loose. If the fingers cannot be passively flexed into the palm with the wrist extended, the transfer is too tight. A sugar-tong splint with a thumb spica extension is placed in the operating room. The wrist should be extended, the MCPs slightly flexed (15°, without placing tension on the transfer), and the IPs left free. The thumb spica extension should maintain the thumb in extension and radial abduction.
FIGURE 80.3. PL to re-routed EPL transfer, for restoration of thumb extension and radial abduction.
MEDIAN NERVE PALSY
Median nerve injuries are categorized as high or low, depending on whether the injury is proximal or distal to the innervation of the forearm muscles. Low median nerve palsy usually results in a loss of thenar function and opposition. However, even when the median nerve has been completely transected, it is not uncommon to see preserved thenar function via a Riche-Cannieu connection.19,20 High median nerve palsy results not only in lost thenar function but also in loss of the FDS to all four fingers, and loss of flexor pollicis longus (FPL) and index FDP function. This causes severe impairment of fine motor control and prehension, loss of oppositional and appositional pinch, and diminished grip strength. Although forearm pronation is lost, the patient compensates with shoulder rotation. FCR function is also lost, but wrist flexion is maintained via the ulnar nerve innervated FCU. Median nerve palsy is a devastating motor injury and is compounded by the loss of critical median nerve distribution sensibility. Even if motor recovery is not possible and tendon transfers are required, the median nerve should be repaired or reconstructed, or sensory transfers in the hand considered to restore this critical area of sensibility.21,22 The goal of tendon transfer in low median nerve palsy is simply to restore thumb opposition. In high median nerve palsy, the goals also include restoration of FPL and index FDP function.
FIGURE 80.4. A. The FCR and PL have been released distally and mobilized proximally in the forearm in preparation for FCR to EDC and PL to EPL transfers. B. On the volar aspect of the forearm, the first pass of the Pulvertaft weave is made for the PL to EPL transfer. C.The FCR is passed through all four EDC tendons to create an end-to-side transfer.
Thumb opposition is a complex movement, which includes components of palmar abduction, pronation, and flexion. The ideal line of pull and insertion point for opponensplasties are disputed. However, most opponensplasties are routed from the level of the pisiform and inserted on the APB tendon, which essentially serves to abduct the thumb, rather than oppose.23–26 A line of pull that originates proximal and radial to the pisiform results in increased palmar abduction, whereas a line of pull that originates distal to the pisiform results in increased flexion across the palm. A tendon transfer that approaches the thumb from the level of the pisiform results in a good combination of both movements. The four most common opponensplasties are the superficialis opponensplasty (FDS),27the Huber opponensplasty (ADM),28 the Camitz opponensplasty (PL),29,30 and the EIP opponensplasty.31
In the superficialis opponensplasty, the ring FDS is transferred to the APB insertion (Figure 80.5). A zigzag incision is made in the distal palm proximal to the ring finger. The A1 pulley is divided, and the FDS is exposed, retracted proximally, and then divided. The FDS is then retrieved proximal to the carpal tunnel through a distal volar forearm incision. It is then routed around a pulley that is created at the level of the pisiform.32 Multiple pulleys have been described, including a distally based strip of FCU that is sutured to itself to form a loop at the level of the pisiform,32 the FCU tendon itself,33,34the flexor retinaculum,35,36 and Guyon’s canal.37 The tendon is then routed through a subcutaneous tunnel across the palm toward the thumb MCP joint and inserted on the APB tendon. The superficialis opponensplasty works well and is a reliable transfer. However, it cannot be used in cases of high median nerve palsy, because FDS function is lost. It should also be noted that the most common cause of low median nerve palsy is a laceration at the wrist. In this situation, the FDS tendons are often injured along with the median nerve. For these reasons, the superficialis opponensplasty is often not a viable option.
FIGURE 80.5. FDS opponensplasty, with pulley at level of pisiform, and insertion on APB tendon.
The EIP opponensplasty is almost always an option in cases of isolated median nerve palsy and has the advantages that no pulley is required, and the donor deficit is minimal (Figure 80.6). A dorsal longitudinal incision is made over the proximal phalanx and MCP joint of the index finger. The sagittal band is carefully elevated off of the EIP for later repair. The EIP is divided distal to the MCP joint and retrieved to a dorsal forearm incision proximal to the extensor retinaculum (Figure 80.7A). It is passed subcutaneously to a third incision on the ulnar border of the wrist. Finally, a fourth incision is made over the radial border of the thumb MCP joint, and a tunnel is created across the palm in the subcutaneous plane. The tendon is passed to the thumb MCP joint and weaved into the APB insertion (Figure 80.7B). The transfer is tensioned with the thumb in maximum palmar abduction and mild flexion.
The Camitz transfer (originally described by Bunnell) utilizes the PL for restoration of opposition. The primary indication for a Camitz transfer is thenar atrophy and loss of palmar abduction in patients with severe long-standing carpal tunnel syndrome.38 During carpal tunnel release, the PL is mobilized, along with a strip of superficial palmar fascia extending into the distal palm. The extended PL is then routed subcutaneously to the APB insertion. Because no pulley is created, the line of pull originates from a position that is proximal and radial to the pisiform. This results in palmar abduction, but little flexion across the palm.
The Huber transfer utilizes the abductor digiti minimus (ADM), one of the hypothenar muscles, to restore opposition (Figure 80.8). In this transfer the ADM is divided at its insertion, mobilized proximally, and turned over like the page of a book to insert on the APB tendon. This transfer is typically reserved for patients with congenital hypoplasia of the thumb, because it recreates some of the bulk of the thenar eminence. In addition, it can be used in patients in whom the FDS or EIP is not available for transfer (such as combined high median and radial nerve palsy).
FIGURE 80.6. EIP opponensplasty. Note that no pulley is required.
FIGURE 80.7. A. The EIP is identified at the index MCP joint, and at the wrist in preparation for EIP opponensplasty. B. The EIP is sutured to the APB tendon with a single suture, to check the tension of the transfer prior to performing the Pulvertaft weave.
In high median nerve palsy, thumb IP flexion is typically restored with a BR to FPL transfer. Index (and sometimes long finger) flexion is restored with a side-to-side suture of the index (and sometimes long finger) FDP tendon to the adjacent FDP tendons at the level of the distal forearm. When independent index FDP function is required, an ECRL to index FDP transfer can be performed. All of these transfers are performed through a volar incision in the distal half of the forearm. The tenodesis effect is used to evaluate the tension of the transfers. With the wrist flexed, the surgeon should be able to passively extend the index finger and radially abduct and extend the thumb. This ensures that the transfers are not too tight, reducing the incidence of flexion contracture. With the wrist passively extended, the index finger should flex into the palm, and the thumb should pinch firmly against the index finger. A sugar-tong splint with thumb spica extension is applied, taking tension off the transfers. The wrist is flexed to 20°, and the thumb is positioned in palmar abduction and flexion. If a tendon transfer for independent index FDP flexion has been performed, the index finger should be placed in the intrinsic plus position. If a side-to-side FDP suture was performed, all four fingers are immobilized in the intrinsic plus position.
FIGURE 80.8. Huber opponensplasty, with the ADM turned over and inserted on the APB tendon.
ULNAR NERVE PALSY
In low ulnar nerve palsy, the ulnar nerve is injured distal to the innervation of the forearm muscles. Adductor pollicis and first dorsal interosseous (FDI) function are lost. This is manifested by weak key pinch and a Froment sign, in which the thumb IP joint flexes during attempted key pinch as the FPL compensates for the loss of adductor pollicis function. The patient also develops clawing (MCP hyperextension and IP flexion), particularly in the ring and small fingers. Clawing is the result of unopposed pull by the extrinsic flexors and extensors due to loss of intrinsic muscle function. The lumbricals and interossei, which normally provide flexion force at the MCP joints and extension force at the IP joints, no longer oppose the pull of the EDC at the MCP joints, or the pull of the FDP and FDS at the IP joints, and clawing occurs. Bouvier test for clawing involves passively correcting MCP hyperextension and checking for extension at the IP joints. If the IP joints can extend, Bouvier test is positive, and the clawing is defined as simple. A procedure that passively maintains MCP flexion can be performed. These static procedures to keep the MCP joints flexed are volar plate advancement, MCP joint fusion or even FDS tenodesis using half of the slip of the FDS. If the IP joints remain flexed, Bouvier test is negative, and the clawing is complex.39 In this case, dynamic transfer is necessary by using the FDS to restore the intrinsic tendon function.
In addition to loss of key pinch and the development of clawing, ulnar nerve palsy causes loss of integration of finger flexion. Finger flexion normally begins at the MCP joints initiated by the intrinsic muscles, followed by flexion at the IP joints initiated by the FDP and FDS tendons. This creates a cupping motion as the fingers are folded into the palm. Without intrinsic function, flexion begins at the IP joints, and the fingers roll into the palm pushing objects away during attempted grasp. Finally, finger abduction and adduction are lost, and the patient loses the ability to spread or cross the fingers.
In high ulnar nerve palsy, in addition to the above findings, the FCU and the FDP to the ring and small fingers are lost, further weakening grip strength. More importantly, absence of FCU function results in loss of the critical “dart-throwing motion.” It should be noted that clawing is less pronounced in high than in low ulnar nerve palsy. Because clawing is the result of unopposed extrinsic function, the loss of small and ring finger FDP function makes clawing less severe. This has two clinical implications. A patient with a recovering high ulnar nerve injury will experience worsening of clawing as the small and ring finger FDP muscles become reinnervated. Therefore, worsening of clawing after a high ulnar nerve repair should be considered a positive finding indicating nerve regeneration. A second consideration is that tendon transfers designed to restore FDP function will result in more pronounced clawing unless an anti-clawing procedure is performed. The goals of tendon transfer in ulnar nerve palsy are restoration of key pinch, correction of clawing, integration of finger flexion, and restoration of small and ring finger distal interphalangeal flexion in cases of high ulnar nerve palsy.
In high ulnar nerve palsy, restoration of small and ring finger FDP function is fairly straightforward and is achieved by side-to-side suturing of the FDP tendons of the ring and small finger to the adjacent functioning long finger FDP tendon at the level of the distal forearm (Figure 80.9). The index FDP should not be included in the transfer, in order to maintain independent index FDP function. It should be remembered, however, that clawing will worsen after this transfer is performed, and the benefits of this transfer should be weighed against the increase in clawing that will occur. After tendon transfer, the fingers are immobilized in the intrinsic plus position, with the wrist flexed to take tension off the transfer.
Restoration of key pinch may or may not be required. Many patients are able to function well because of compensation by the FPL, or because the adductor pollicis receives aberrant innervation from the median nerve. No tendon transfer is necessary unless the patient notices a significant loss of pinch function. In addition, when key pinch is required, it is usually only necessary to restore adductor pollicis function. Although transfers have been described to restore FDI function, the function of the FDI is compensated by stabilization of the index finger against the long finger. Multiple transfers have been described for the restoration of key pinch, including the use of the ECRB, the BR, the EIP, and the FDS.40–42 When the ECRB or BR is used to restore key pinch, it must be elongated with a tendon graft (Figure 80.10). The tendon graft is then routed through the second or third intermetacarpal spaces from dorsal to volar, and into the palm. The graft is then routed radially across the palm, deep to the flexor tendons, digital nerves, and digital arteries, and toward the adductor pollicis insertion. The border of the second or third metacarpal acts as a pulley to achieve the correct line of pull for key pinch. If an FDS transfer is performed, the long FDS is divided in the finger proximal to the PIP joint and retrieved into the palm. It is then passed radially across the palm deep to the flexor tendons and neurovascular bundles and inserted on the adductor pollicis insertion. The advantage of this transfer is that a tendon graft is not required. However, harvest of the FDS can further weaken grip strength. In addition, care must be taken in choosing which FDS to use. In high ulnar nerve palsy, the ring and small finger FDPs have lost innervation, and those fingers rely on the median nerve innervated FDS for flexion. Therefore, in high ulnar nerve palsy the long finger FDS should be used instead of the ring FDS. In all cases, the surgeon should confirm normal FDP function prior to using the FDS in a tendon transfer. A final option for restoring key pinch is the EIP transfer. In this transfer, the EIP is divided distal to the MCP joint of the index finger, as for EIP opponensplasty. It is mobilized up to the extensor retinaculum, and then directed through the second or third intermetacarpal space and across the deep palm to the adductor pollicis insertion. This tendon is available for transfer in high and low ulnar nerve palsy, its harvest does not weaken grip strength, and it does not usually require elongation with a tendon graft. After a transfer to restore key pinch, the thumb is immobilized in a position of pinch. Wrist position depends upon whether the FDS or another donor MTU was used. If the FDS was used, tension is taken off the transfer by immobilizing the wrist in flexion. If another donor MTU was used and passed through the intermetacarpal space, the wrist is immobilized in extension.
FIGURE 80.9. Side-to-side suture of small and ring FDP to the long FDP.
FIGURE 80.10. Transfer of elongated ECRB to adductor pollicis insertion for restoration of key pinch.
Correction of clawing is the other major goal in ulnar nerve palsy and relies on correcting MCP hyperextension. This can be achieved with many different operations, which can be divided into static and dynamic procedures. If Bouvier test is positive, and the IP joints can extend with the MCPs passively flexed, then a simple block to MCP hyperextension is considered. Creation of a bony block on the dorsal aspect of the metacarpal head has been described, but is of primarily historical interest.43 A volar plate capsulodesis is another static option, in which a distally based flap of the volar plate is advanced proximally, tightening the volar aspect of the joint and effectively preventing MCP hyperextension.44 Tenodesis with a tendon graft that limits MCP hyperextension can be performed as well.45All of these static procedures tend to weaken over time, however, and are best employed in patients with mild, simple clawing.
Many dynamic tendon transfers to correct clawing have been described. These include the use of the FDS, the BR, the FCR, the ECRL, and the ECRB.46–52 Because all of these transfers restore active MCP flexion, they integrate finger flexion in addition to correcting clawing. These transfers differ in whether tendon grafts are required, whether grip strength is augmented or weakened, and whether active IP extension is provided. In the FDS transfer, the ring or long finger FDS is divided proximal to the PIP joint and retrieved into the palm. There, it is split into two slips, one for the ring finger and one for the small finger. If clawing is present in all four fingers, it is possible to divide the FDS into four slips. Alternatively, the long FDS can be used to correct clawing in the index and long fingers, and the ring FDS can be used to correct clawing in the ring and small fingers. Various insertions have been described (Figure 80.11). The FDS slips can be inserted on the flexor tendon sheath in a “lasso” fashion as described by Zancolli, providing active MCP flexion.50 Alternatively, they can be passed volar to the deep transverse metacarpal ligament along the path of the lumbrical and inserted on the radial lateral band or into the bone of the proximal phalanx. Insertion on the lateral band provides both MCP flexion and IP extension. However, because the FDS normally restrains PIP hyperextension, transfer of the FDS to the lateral band can lead to PIP hyperextension.49 The main drawback of the FDS transfer is that it results in further weakening of grip strength. As noted above, it is important to ensure that the FDP of a given finger functions normally prior to using the FDS for tendon transfer.
FIGURE 80.11. Three classic insertions for transfers to correct clawing: lateral band, bone of proximal phalanx, and the flexor tendon sheath.
Wrist level motors such as the BR, the FCR, the ECRL, and the ECRB can be used to correct clawing, but all require elongation with a tendon graft. (It should be noted that the FCR should only be used in cases of low ulnar nerve palsy, in which the FCU is intact.) The elongated tendon is routed to the dorsal wrist, and the graft is split into the appropriate number of slips. The slips are then routed through the intermetacarpal spaces of the corresponding fingers and then passed along the course of the lumbrical, volar to the deep intermetacarpal ligament, and out to the finger (Figure 80.12). Again, the insertion can be on the lateral band or into the bone of the proximal phalanx, depending on whether active IP extension is required. If insertion on the pulley system is preferred, the tendon graft is not passed through the lumbrical canal, but rather inserted on the flexor tendon sheath in a “lasso” fashion. In addition to requiring tendon grafts for elongation, another drawback to these transfers is the possibility of adhesions within the intermetacarpal space. The primary advantage of using a wrist level motor instead of the FDS is that these transfers augment rather than diminish grip strength.
For active transfers to correct clawing, whether the FDS or a wrist level motor is used, the tendon transfer is tensioned with the wrist in neutral, the MCPs maximally flexed, and the IPs extended. The hand is immobilized with the MCPs flexed and the IPs extended, in order to take tension off the transfer. The position of wrist immobilization depends upon whether the FDS or wrist level motor was used.
FIGURE 80.12. Transfer of elongated ECRL to the radial lateral bands of the ring and small fingers, for dynamic correction of clawing.
The therapist should be involved in the care of the patient well in advance of a planned tendon transfer. The therapist will focus on maximizing passive joint motion and on strengthening the planned donor MTUs. During this time period the therapist and the patient get to know each other. This allows the therapist to gauge the patient’s motivation and compliance and to communicate with the patient about the postoperative rehabilitation protocol. In general, tendon transfers are immobilized for 4 weeks postoperatively prior to initiating therapy. During these 4 weeks, it is crucial to maintain motion in the uninvolved joints in order to prevent stiffness. The splint applied in the operating room should be removed at 2 weeks and changed to a cast which is worn for 2 more weeks. Sutures are removed at the 2-week visit. At 4 weeks rehabilitation is initiated. A thermoplast splint is made and is worn when the patient is not performing prescribed exercises. Gentle active and assisted active range of motion is initiated. Synergistic movements are taught. Electrical stimulation and biofeedback can be useful in teaching the patient to activate the transfer. Passive stretching of the transfer is gradually introduced, usually at 6 weeks. At 8 weeks postoperatively, strengthening is begun, and the splint is gradually weaned off for light hand use. Full unrestricted activity is allowed at 3 months postoperatively.
1. Green DP. Radial nerve palsy. In: Green DP, Hotchkiss RN, Pederson WC, eds. Green’s Operative Hand Surgery. Vol 2. 4th ed. New York: Churchill Livingstone; 1999:1481-1496.
2. Richards RR. Tendon transfers for failed nerve reconstruction. Clin Plast Surg. April 2003;30(2):223-245, vi.
3. Van Heest AE, House JH, Cariello C. Upper extremity surgical treatment of cerebral palsy. J Hand Surg Am. March 1999;24(2):323-330.
4. Boyes JH. Tendon transfers in the hand. Proceedings of the 15th General Assembly of the Japan Medical Congress. Medicine in Japan in 1959. 1959;5:958.
5. Boyes JH. Selection of a donor muscle for tendon transfer. Bull Hosp Joint Dis. April 1962;23:1-4.
6. Omer GE Jr. Reconstructive procedures for extremities with peripheral nerve defects. Clin Orthop Relat Res. March 1982(163):80-91.
7. Burkhalter WE. Early tendon transfer in upper extremity peripheral nerve injury. Clin Orthop Relat Res. October 1974(104):68-79.
8. Littler J. Restoration of power and stability in the partially paralyzed hand. In: Converse J, ed. Reconstructive Plastic Surgery. Philadelphia, PA: WB Saunders; 1977:3266-3280.
9. Lieber RL, Jacobson MD, Fazeli BM, Abrams RA, Botte MJ. Architecture of selected muscles of the arm and forearm: anatomy and implications for tendon transfer. J Hand Surg Am. September 1992;17(5):787-798.
10. Lieber RL, Ponten E, Burkholder TJ, Friden J. Sarcomere length changes after flexor carpi ulnaris to extensor digitorum communis tendon transfer. J Hand Surg Am. July 1996;21(4):612-618.
11. Omer GJ. Tendon transfers for reconstruction of the forearm and hand following peripheral nerve injuries. In: Omer GJ, Spinner M, eds. Management of Peripheral Nerve Problems. Philadelphia, PA: WB Saunders; 1980: 817-846.
12. Brand P. Tendon transfers in the forearm. In: Flynn JE, ed. Hand Surgery. Baltimore, MD: Williams & Wilkins; 1975:189-200.
13. Starr C. Army experiences with tendon transference. J Bone Joint Surg Am. 1922;4:3-21.
14. Tsuge K, Adachi N. Tendon transfer for extensor palsy of forearm. Hiroshima J Med Sci. December 1969;18(4):219-232.
15. Jones R. II. On suture of nerves, and alternative methods of treatment by transplantation of tendon. Br Med J. May 13 1916;1(2889):679-682.
16. Jones R. Tendon transplantation in cases of musculospiral injuries not amenable to suture. Am J Surg. 1921;35:333-335.
17. Chuinard RG, Boyes JH, Stark HH, Ashworth CR. Tendon transfers for radial nerve palsy: use of superficialis tendons for digital extension. J Hand Surg Am. November 1978;3(6):560-570.
18. Boyes J. Tendon transfers for radial palsy. Bull Hosp Joint Dis. 1960;21: 97-105.
19. Riche P. Le nerf cubital et les muscles de l’eminence thenar. Bull Mem Anat Paris. 1897:251-252.
20. Cannieu A. Note sur une anastomose entre le branche profunde de cubital et le median. Bull Soc Anat Physiol Horm Path Bordeaux. 1897;17: 339-342.
21. Ozkan T, Ozer K, Gulgonen A. Restoration of sensibility in irreparable ulnar and median nerve lesions with use of sensory nerve transfer: long-term follow-up of 20 cases. J Hand Surg Am. January 2001;26(1):44-51.
22. Brand PW. Clinical Mechanics of the Hand. St. Louis, MO: Mosby; 1985.
23. Cooney WP, Linscheid RL, An KN. Opposition of the thumb: an anatomic and biomechanical study of tendon transfers. J Hand Surg Am. November 1984;9(6):777-786.
24. Curtis RM. Opposition of the thumb. Orthop Clin North Am. April 1974;5(2):305-321.
25. Lee DH, Oakes JE, Ferlic RJ. Tendon transfers for thumb opposition: a biomechanical study of pulley location and two insertion sites. J Hand Surg Am. November 2003;28(6):1002-1008.
26. Roach SS, Short WH, Werner FW, Fortino MD. Biomechanical evaluation of thumb opposition transfer insertion sites. J Hand Surg Am. March 2001;26(2):354-361.
27. Royle ND. An operation for paralysis of the intrinsic muscles of the thumb. JAMA. 1938;111:612-613.
28. Huber E. Hilfsoperation bei median Uhlahmung. Dtsch Arch Klin Med. 1921;136:271.
29. Bunnell S. Reconstructive surgery of the hand. Surg Gynecol Obstet. 1924;34:259-274.
30. Camitz H. Uber die behandlung der oppositionslahmung. Acta Orthop Scand. 1929;65:77-81.
31. Aguirre SC, Caplan S. Sobre secuelas de lesion alta e irreparable de nervios mediano y cubital, y su tratamiento. Prensa Med Arg. 1956;43:2341-2346.
32. Bunnell S. Opposition of the thumb. J Bone Joint Surg Am. 1938;20: 269-284.
33. Boswick JA Jr, Stromberg WB Jr. Isolated injury to the median nerve above the elbow. A review of thirteen cases. J Bone Joint Surg Am. June 1967;49(4):653-658.
34. Mangus DJ. Flexor pollicis longus tendon transfer for restoration of opposition of the thumb. Plast Reconstr Surg. August 1973;52(2):155-159.
35. Michelinakis E, Vourexakis H. Tendon transfer for intrinsic-muscle paralysis of the thumb in Charcot-Marie Tooth neuropathy. Hand. October 1981;13(3):276-278.
36. Bohr HH. Tendon transposition in paralysis of the opposition of the thumb. Acta Chir Scand. 1953;105(1-4):45-54.
37. Brand PW. Tendon transfers for median and ulnar nerve paralysis. Orthop Clin North Am. November 1970;1(2):447-454.
38. Littler JW, Li CS. Primary restoration of thumb opposition with median nerve decompression. Plast Reconstr Surg. January 1967;39(1):74-75.
39. Bouvier. Note sur une paralysie partielle des muscles de la main. Bull Acad Nat Med Paris. 1851;18:125.
40. Smith RJ. Extensor carpi radialis brevis tendon transfer for thumb adduction—a study of power pinch. J Hand Surg Am. January 1983;8(1):4-15.
41. Boyes JH. Bunnell’s Surgery of the Hand. 5th ed. Philadelphia, PA: J.B. Lippincott Company; 1970.
42. Hamlin C, Littler JW. Restoration of power pinch. J Hand Surg Am. July 1980;5(4):396-401.
43. Mikhail IK. Bone block operation for clawhand. Surg Gynecol Obstet. May 1964;118:1077-1079.
44. Zancolli EA. Claw-hand caused by paralysis of the intrinsic muscles: a simple surgical procedure for its correction. J Bone Joint Surg Am. October 1957;39-A(5):1076-1080.
45. Parkes A. Paralytic claw fingers—a graft tenodesis operation. Hand. October 1973;5(3):192-199.
46. Bunnell S. Surgery of the intrinsic muscles of the hand other than those producing opposition of the thumb. J Bone Joint Surg Am. 1942;24:1-3.
47. Riordan DC. Tendon transplantations in median-nerve and ulnar-nerve paralysis. J Bone Joint Surg Am. April 1953;35-A(2):312-320; passim.
48. Stiles HJ. Treatment of Injuries of Peripheral Spinal Nerves. London: Frowde H, Hodder and Stoughton; M F Forrester-Brown.
49. Burkhalter WE. Restoration of power grip in ulnar nerve paralysis. Orthop Clin North Am. April 1974;5(2):289-303.
50. Hastings H 2nd, McCollam SM. Flexor digitorum superficialis lasso tendon transfer in isolated ulnar nerve palsy: a functional evaluation. J Hand Surg Am. March 1994;19(2):275-280.
51. Brand PW. Tendon grafting illustrated by a new operation for intrinsic paralysis of the fingers. J Bone Joint Surg Br. 1961;43B:444-453.
52. Enna CD, Riordan DC. The Fowler procedure for correction of the paralytic claw hand. Plast Reconstr Surg. October 1973;52(4):352-360.