Moment arm and force-generating capacity of the extensor carpi ulnaris after transfer to the extensor carpi radialis brevis.

Tendon transfers to the extensor carpi radialis brevis (ECRB) are often performed to augment wrist extension. This study was conducted to analyze how transfer of the extensor carpi ulnaris (ECU) to the ECRB affects the moment arms, force-generating capacity, and moment-generating capacity of the ECU over a range of wrist flexion-extension. A graphics-based computer model was developed from anatomic measurements of the muscle-tendon paths before and after transfer. This model calculates the lengths and moment arms of the muscles over a range of wrist flexion-extension and represents the muscles' force-generating characteristics from previous measurements of their physiologic cross-sectional areas, fiber lengths, and pennation angles. Analysis of the computer model revealed that the maximum isometric extension moment of the ECU at the neutral wrist position increased from 0.50 N-m to 1.72 N-m after transfer to the ECRB. The deviation moment shifted from 2.72 N-m ulnar deviation to 1.42 N-m radial deviation. The extension moment generated by the ECU varied more with wrist flexion angle after transfer due to its broadened operating range on the muscle force-length relationship. The simulations highlight the need for proper intraoperative tensioning of the ECU to maximize the force-generating potential of the transferred muscle over the functional range of motion.     

Architecture of selected muscles of the arm and forearm: anatomy and implications for tendon transfer.

The architectural features of twenty-one different forearm muscles (n = 154 total muscles) were studied. Muscles included the extensor digitorum communis to the index, middle, ring, and small fingers, the extensor digit quinti, the extensor indicis proprius, the extensor pollicis longus, the flexor digitorum superficialis, the flexor digitorum profundus, the flexor pollicis longus, the pronator quadratus, the palmaris longus, the pronator teres, and the brachioradialis. Muscle length, mass, fiber pennation angle, fiber length, and sarcomere length were determined with the use of laser diffraction techniques. From these values, physiologic cross-sectional area and fiber length/muscle length ratio were calculated. The individual digital extensor muscles were found to be relatively similar in architectural structure. Similarly, the deep and superficial digital flexors were very similar architecturally, with the exception of the small finger flexor digitorum superficialis, which was much smaller and shorter than the rest of the digital flexors. The brachioradialis and the pronator teres had dramatically different architectural properties. While the masses of the two muscles were nearly identical, the muscles had significantly different predicted contractile properties based on their different fiber arrangement. The brachioradialis, with its long fibers arranged at a small pennation angle, had a physiologic cross-sectional area that was only one third that of the pronator teres, with its short fibers that were more highly pennated. Using these architectural data and the statistical method of discriminant analysis, we provide additional information that might be useful in the selection of potential donor muscles to restore thumb flexion, thumb extension, finger extension, and finger flexion.     

Architecture of selected wrist flexor and extensor muscles

The architectural features of 25 wrist flexor and extensor muscles were studied. Muscles included the flexor carpi ulnaris, the flexor carpi radialis, the extensor carpi ulnaris, the extensor capri radialis brevis, and the extensor carpi radialis longus. Muscle length, mass, fiber pennation angle, fiber length, and sarcomere length (by use of laser diffraction techniques) were determined. In addition, physiological cross-sectional area and fiber length/muscle length ratio were calculated. The muscles were found to be highly specialized, with architectural features of same muscles very similar. The fiber length/muscle length ratio, muscle length, and pennation angle represented the major differences between muscles. Thus using these parameters in discriminant analysis permitted correct identification of each of the 25 muscles. In terms of size and intrinsic design, these individual muscles were highly specialized for their function.     

Architectural design of the human intrinsic hand muscles.

The architectural features of twenty different muscles (18 intrinsics and 2 thumb extrinsics, n = 180 total muscles) were studied. Muscle length, mass, fiber pennation angle, fiber length, and sarcomere length were determined. From these values, physiologic cross-sectional area and fiber length/muscle length ratio were calculated. Intrinsic muscle lengths were relatively similar to one another, which we interpreted as representing a space constraint within the hand. However, several specialized architectural designs were observed: lumbrical muscles had an extremely high fiber length/muscle length ratio, implying a design toward high excursion. The first dorsal interosseous and adductor pollicis had physiologic cross-sectional areas comparable to those of extrinsic muscles and much greater than those of the other intrinsic muscles. The interosseous muscles had relatively high physiologic cross-sectional areas with low fiber length/muscle length ratios, suggesting their adaptation for high force production and low excursion. Taken together, these observations illustrate the underlying structural basis for the functional capacities of the intrinsic muscles.     

The supination effect of tendon transfer of the flexor carpi ulnaris to the extensor carpi radialis brevis or longus: a cadaveric study.

Flexor carpi ulnaris (FCU) transfer to the extensor carpi radialis brevis (ECRB) and/or the extensor carpi radialis longus (ECRL) has been commonly used to provide wrist extension. The ability of this wrist extension transfer to also provide forearm supination has been inferred but not formally investigated. This laboratory study investigated the forearm supination effect of FCU transfer to the ECRB and to the ECRL in a cadaveric model. Two vectors of pull were investigated: freeing either the distal one third or the distal two thirds of the FCU ulnar origin. Five fresh-frozen, above-elbow, non-matched cadaveric specimens placed in a mounting device that allowed the arm to rotate about its ulnar axis starting from a full pronated position were measured for resultant supination after tendon transfer and loading. This study showed that the transfer of the FCU into either the ECRB or the ECRL resulted in no significant difference in maximum supination. The vector of origin, however, did significantly affect the maximum supination obtained. Releasing the distal two thirds of the FCU ulnar origin resulted in a mean supination that was significantly greater than the mean supination achieved with releasing the distal one third of the FCU ulnar origin. We concluded that in the cadaveric model, transfer of the FCU into either the ECRB or ECRL provided similar resultant supination and that freeing the distal two thirds of the FCU ulnar origin provided significantly more supination than freeing only the distal one third. For the hand surgeon treating wrist flexion in combination with forearm pronation deformity, transfer of the FCU into the ECRB and/or the ECRL can be used to concomitantly provide wrist extension and forearm supination.     

拇长伸肌腱自发性断裂桡侧腕短伸肌腱替代术的临床应用

目的 探讨腕部拇长伸肌腱自发性断裂,采用桡侧腕短伸肌腱替代术的临床疗效.方法 1996年2月-2005年7月,对12例拇长伸肌腱自发性断裂的患者采用桡侧腕短伸肌腱移位替代术.其中桡骨远端骨折愈合后肌腱断裂6例,类风湿骨关节炎4例,不明原因2例.术后均不用石膏固定,术后2d开始练习伸手、伸腕、伸拇动作,2～3次/d;术后7d停止练习;术后4～5周可缓慢伸手持物.结果 12例术后随访均在1年以上,12个月后平均伸拇肌力已达4～5级,腕背肌腱滑动时无障碍.术后无肌腱再断裂者,伸腕、伸指时拇指末节过伸2°～5°,平均3°,拇对掌、对指功能正常.根据中华医学会手外科学会手部肌腱修复后功能评定标准评价,优良率达100%.全部患者对术后功能满意.结论 拇长伸肌腱自发性断裂,应用桡侧腕短伸肌腱移位替代术,对供区损伤小,操作方便,肌腱缝合牢固,有利早期功能恢复.     

Quantitative evaluation of the posterior deltoid to triceps tendon transfer based on muscle architectural properties

The architectural properties of the posterior deltoid muscle and the 3 heads of the triceps were measured using microdissection techniques to determine whether substitution of triceps function by the posterior deltoid is architecturally appropriate. Muscles from 10 fresh cadaver specimens were fixed by high-pressure perfusion using buffered formaldehyde. Muscle architectural properties, including pennation angle, fiber bundle length, sarcomere length, and physiologic cross-sectional area, were determined. Fiber bundle length varied significantly among the deltoid (123.1 +/- 7.8 mm), medial (64.5 +/- 3.8 mm), lateral (66.5 +/- 5.4 mm), and long (85.3 +/- 9.5) heads of the triceps. The physiologic cross-sectional area of the posterior deltoid was significantly less than the total triceps area and was predicted to provide only approximately 20% of the maximum isometric tension of the combined triceps heads. These data demonstrate that the long fibers of the posterior deltoid render it a very suitable transfer to provide elbow extension because of its tremendous excursion and also show why useful functional results seem relatively independent of posterior deltoid tension at the time of surgery.     

Measurement of gliding resistance of the extensor pollicis longus and extensor digitorum communis II tendons within the extensor retinaculum

PURPOSE: The etiology of spontaneous extensor pollicis longus (EPL) tendon rupture is still largely unknown. It is possible that friction within the sheath may play a role. The purposes of this study were to compare gliding resistance of the EPL tendon with that of the extensor digitorum communis tendon of the index finger (EDC II) and to find the wrist position that gives the EPL tendon the lowest gliding resistance. METHODS: Fifteen fresh-frozen cadavers were used. Gliding resistance was measured directly in 7 different wrist positions. RESULTS: The mean gliding resistance of the EPL tendon was 0.16 +/- 0.08 N and that of the EDC II tendon was 0.11 +/- 0.06 N. This difference was significant. There was also a significant effect on gliding resistance due to wrist position. For the EPL tendon, the gliding resistance was significantly greater in 60 degrees wrist flexion compared with all other wrist positions tested. Additionally the gliding resistance of the EPL in 30 degrees flexion, 60 degrees extension, and 15 degrees radial deviation was significantly higher than wrist positions of 30 degrees extension, neutral, and 30 degrees ulnar deviation. CONCLUSIONS: Positioning the wrist close to neutral flexion/extension and in some ulnar deviation minimizes the friction within the EPL sheath. Such positions may be advantageous for splinting patients at risk for EPL rupture.     

食指固有伸肌腱移位重建拇长伸肌功能及评价

目的评价采用食指固有伸肌腱移位重建拇长伸肌功能的临床疗效. 方法对1978年8月～2003年3月以食指固有伸肌移位重建拇长伸肌功能的46例患者进行随访、评价.其中男32例,女14例.年龄16～51岁,平均36岁.外伤陈旧性断裂24例,继发性断裂22例.病程2天～5个月,平均74天. 结果 41例获7个月～23年随访,平均9年3个月.术后拇指抬高丢失0～2.2 cm,平均1.8 cm;拇指屈曲丢失0～3 cm,平均1.6 cm;食指均能单独背伸,背伸丢失0～8度,平均5度.按SEEM评分标准:优29例,良10例,可2例,优良率达95%. 结论食指固有伸肌腱移位重建拇长伸肌功能是一种简便、有效的方法,采用SEEM评价标准使评价结果更加客观和合理.     

Lateral epicondylitis: anatomic relationships of the extensor tendon origins and implications for arthroscopic treatment

The purpose this study was to provide an understanding of the location of the extensor carpi radialis brevis (ECRB) and its relationship to surgically relevant landmarks. In part I, 20 fresh-frozen cadaveric elbows were dissected, and the anatomic boundaries of the ECRB origin were determined, including its relationship to the extensor carpi radialis longus, the extensor digitorum communis, the lateral epicondyle, the radiocapitellar joint, the lateral collateral ligament, and the elbow capsule. All measurements were made by 3 separate investigators. Whereas the ECRB and extensor digitorum communis converged distal to the radiocapitellar joint, the bony origin of the ECRB was reliably identified beneath the distal-most aspect of the supracondylar ridge, with a mean dimension of 13 +/- 2 mm in length by 7 +/- 2 mm in width. Part II was performed on 10 separate specimens to document whether the ECRB tendon could be identified and released arthroscopically. In all specimens, the ECRB origin was safely and completely released, by use of the landmarks defined in part I. When the ECRB is released arthroscopically, the elbow capsule must be resected and the tendinous origin of the ECRB released from the top of the capitellum to the midline of the radiocapitellar joint.     

The anatomy of the posterior interosseous nerve as a graft

Thirty upper limbs from skeletally mature embalmed cadavers were studied to determine the anatomic reliability of the posterior interosseous nerve as a donor nerve graft. The posterior interosseous nerve branches 0.43 +/- 0.52 cm from the distal edge of the superficial head of the supinator and 8 +/- 1.6 cm from the lateral epicondyle form a common leash. There are 6 branches, which are arranged from the ulnar to the radial side at their origin from this leash. The first and second branches supply the extensor digitorum communis, the third branch supplies the extensor carpi ulnaris, the fourth branch supplies the extensor digiti minimi, and the fifth branch arises from the undersurface of the common leash and divides into 2 sub-branches (medial and lateral) 10.1 +/- 3.2 cm distal to the lateral epicondyle and 12.8 +/- 2.2 cm proximal to Lister's tubercle. These 2 sub-branches make an inverted V shape around the extensor pollicis longus. The medial branch supplies the extensor pollicis longus and extensor indicis proprius. The lateral branch supplies the extensor pollicis longus and extensor pollicis brevis and ends at the wrist capsule. At a mean distance of 8.1 +/- 1.2 cm proximal to Lister's tubercle the lateral sub-branch gives off its last muscular branch to the extensor pollicis longus and becomes a pure sensory terminus. As the terminal part of the lateral sub-branch approaches the wrist capsule it expands at a mean distance of 1.9 +/- 0.5 cm proximal to Lister's tubercle. The sixth branch arises from the radial side of the common leash and divides into 3 sub-branches. The first sub-branch supplies the abductor pollicis longus and extensor pollicis brevis, the second supplies the abductor pollicis longus, and the third supplies the superficial head of the supinator. This study showed that the mean length obtainable for harvesting the lateral sub-branch of the fifth branch of the posterior interosseous nerve is 6.2 +/- 0.7 cm, which represents the length of the nerve between the last muscular branch to the extensor pollicis longus to the point at which the nerve expands.     

Extensor hallucis longus innervation: an anatomic study

Thirty legs from skeletally mature embalmed cadavers were dissected to define the most common pattern and the variants of innervation of the extensor hallucis longus muscle and its clinical significance. Twenty-seven muscles had only one innervating branch (90%). Only three muscles had two innervating branches (10%). Twenty-one of the branches entered the muscles from the fibular side (63.6%), six entered the muscles from the tibial side (18.2%), and six entered the muscles from the anterior edge (18.2%). The branches innervating the extensor hallucis longus from the fibular side had a closer relation with the fibular periosteum than those entering the muscle from the tibial side or the anterior edge. The mean length of these branches between their points of origin and entry in the extensor hallucis longus was 5.0 +/- 1.5 cm. The high risk zone for the iatrogenic injury to the muscular branch of the extensor hallucis longus was located between 5.9 +/- 1.7 and 10.9 +/- 1.7 cm inferior to the most distal palpable point of the fibular head. The current study confirmed that the extensor hallucis longus was supplied mostly by one nerve that usually entered the muscle from the fibular side and had a close relation to the fibular periosteum in the dangerous zone.     

How musculotendon architecture and joint geometry affect the capacity of muscles to move and exert force on objects: A review with application to arm and forearm tendon transfer design

This commentary reviews musculotendon architecture and the relation between architectural parameters and the force, speed, and excursion capacity of musculotendon units. It is hoped that this review will help provide the framework within which to appreciate the importance of the data presented by Lieber et al. Muscle fiber pennation hardly affects musculotendon output of forearm and hand muscles. Instead, physiologic cross-sectional area and muscle fiber length affect force capacity and speed and excursion capacity, respectively. How muscles with equal mass can have different force, speed, and excursion capacities is explained. Since the moment arm of a muscle (the shortest distance from the musculotendon unit to the joint center of rotation) transforms muscle output into musculotendon output, it is shown why the capacity for a muscle to exert force on an object, as during grasping, is directly proportional to its moment arm and why the range of joint movement and speed over which muscles exert force is inversely proportional to the moment arm. Finally, tendon, being not stiff in forearm and hand musculotendon units, also affects their output. Criteria are given for designing tendon transfer reconstructions from architectural data and moment arm data to best replicate the biomechanical function of the replaced muscle. To have the same capacity for imparting movement to objects and exerting force on them, the donor muscle should have the same moment arm/physiologic cross-sectional area product, the same fiber length/moment arm ratio, and the same tendon length/muscle fiber length ratio as the replaced muscle.     

Strain reduction of the extensor carpi radialis brevis tendon proximal origin following the application of a forearm support band

STUDY DESIGN: Experimental laboratory design. OBJECTIVES: To measure the strain at the proximal origin of the extensor carpi radialis brevis (ECRB), and to determine the influence of a forearm support band. BACKGROUND: A forearm support band is often used with the intent to decrease stresses around the origin of the wrist extensors. However, the influence of the location of the band has not been studied. METHODS AND MEASURES: The forearm support band was applied on 8 cadaver arms (mean +/- SD age, 78.4 +/- 10.3 years) and 2 experimental conditions were performed. First, strain measurements were made without applying tension to the distal ECRB tendon, then strain measurements were made with a traction force of 21.5 N being applied to the distal ECRB tendon. Strain of the proximal origin of the ECRB, 1.0 cm distal from the lateral epicondyle, was recorded using a strain gauge. The band was mounted on the forearm at distances equal to 80%, 70%, 60%, 50%, 40%, 30%, and 20% of the forearm length as measured from the wrist. Testing order was randomized. Tension applied to the band was 19.6 N. RESULTS: When no tension was applied to the ECRB, there was no statistically significant difference (P>.05) in strain values at the ECRB origin by mounting the band at any of the forearm positions. In the tension condition, the average (SD) strain with no band was 2.40% (1.40%). The average strain value of 0.85% (0.65%), when the band was mounted 80% of the forearm length proximal to the wrist, was statistically smaller than that obtained without the band (P<.05). CONCLUSIONS: The strain on the ECRB origin was less when the forearm support band was applied 80% proximal from the wrist joint. LEVEL OF EVIDENCE: Therapy, level 5.     

Palliative tendon transfer for reanimation of the wrist and finger extension lag. Report of 14 transfers for radial nerve palsies and ten transfers for brachial plexus lesions

This retrospective study is based on 23 males and one female, of an average age of 36.2 years that presented to us between 1982 and 2000 with an average follow up of 61 months, with fully established paralysis of wrist and fingers extension. Fourteen patients had isolated radial nerve palsy, while ten patients had brachial plexus lesions. 1) The tendon transfer for radial nerve palsy was: PT to ECRB, FCU to ED + EPL and PL to APL + EPB; 2) for brachial plexus injury, the tendon transfer was: PT (n = 4) or FDS III or IV (n = 5) to ECRB, FCU (n = 8) or FDS IV (n = 1) to ED + EPL, PL to APL + EPB and wrist arthrodesis with transfer of FDS IV to ED + EPL and PL to APL + EPB. The results were evaluated according to the degree of wrist movement, MP extension of long fingers, opening of first commissure, thumb opposition, grip power and the subjective evaluation of results. Concerning the radial nerve palsy: results are excellent in nine cases and good in one case. An active extension of the wrist of 38 degrees was obtained as well as MP extension of 0 degree with the wrist straightened. Thumb oppositioned was conserved (Kapandji = 8.2), opening of the first commissure 40 degrees and grip power was 20 kg. Concerning the brachial plexus lesions: results are excellent in five cases and good in the other five. An active wrist extension of 32 degrees was obtained, as well as MP extension deficit of 16 degrees with wrist straightened. Opposition was concerned (Kapandji = 7.2), opening of first commissure of 38 degrees and grip power of 13 kg. The functional results are satisfactory, but the analytic study shows some effect of tenodesis of MP extension.     

An analysis of results after selective tendon transfers through the interosseous membrane to provide selective finger and thumb extension in chronic irreparable radial nerve lesions

PURPOSE: We present the results of a modified tendon transfer for the restoration of wrist and finger extension in irreparable radial nerve lesions. METHODS: Restoration of wrist extension, finger extension, thumb extension, and thumb abduction was done in 29 patients (20 males and 9 females; age range: 10-58 years) with isolated, irreparable radial nerve palsy. We used a modified tendon transfer technique using the flexor digitorum superficialis (FDS) 3 (to extensor indicis proprius [EIP] and extensor pollicis longus [EPL]) and FDS 4 (to extensor digitorum communis 2-4 [EDC]) as donors for the reconstruction of selective finger and thumb extension (all patients) and pronator teres (PT) for wrist extension (25 patients). Thumb abduction was achieved by transferring the palmaris longus (PL) tendon to the abductor pollicis longus (APL) (all patients). RESULTS: Results show that near-normal wrist extension was achieved in 22 of 25 patients with extension strength of M4+. In the other 3 patients, wrist extension strength did not exceed M3 (1 patient) or M4 (2 patients). Extension of long fingers with a completely extended wrist joint was achieved in 12 of 29 patients. In the remaining 17 patients, full-range finger extension was possible only with the wrist in neutral. The advantage of the selective tendon transfer (FDS 3 to EIP and EPL and FDS 4 to EDC 2-4) resulted in selective extension of the index finger and thumb, as well as other digits, in all patients. Thumb abduction and rotation was achieved in all. CONCLUSIONS: Tendon transfers are indicated in longstanding, irreparable, isolated radial nerve lesions. Selective tendon transfer of FDS 3 to EIP and EPL and FDS 4 to EDC through the interosseous membrane results in reliable selective extension of these digits. The sacrifice of FDS 3 and 4 to reconstruct finger extension results in bowing of the donor digits. TYPE OF STUDY/LEVEL OF EVIDENCE: Therapeutic III.     

Transfer of extensor carpi radialis longus or brevis for opponensplasty

For the restoration of thumb opposition many types of tendon transfer techniques have been described. The flexor digitorum superficialis (FDS) of the ring finger is commonly selected as a motor. On occasion, however, the quality of the flexor muscles of the fingers or wrist is not good enough for tendon transfer and another available muscle must be selected. In this situation, we have preferred to use an extensor carpi radialis longus (ECRL) or brevis (ECRB) transfer to restore opposition of the thumb. Follow-up examination, at an average 5 years and 10 months after operation, showed that the results of ten of 11 transfers were excellent and the other was good.     

Fiber length variability within the flexor carpi ulnaris and flexor carpi radialis muscles: implications for surgical tendon transfer

PURPOSE: The purpose of this study was to understand the detailed architectural properties of the human flexor carpi radialis (FCR) and flexor carpi ulnaris (FCU) muscles and their implications for tendon transfer surgery. METHODS: Muscle fiber length was measured in 6 separate regions of the FCU and FCR from 10 cadaveric specimens. Sarcomere length was measured by laser diffraction for normalization. Moment arms were estimated by measuring tendon excursion with respect to joint angle. The position of entry of the motor nerve branches into each muscle also was measured to establish limits for the safe length of muscle mobilization. RESULTS: Muscle fiber length varied significantly along both the FCU and FCR. Fiber length variability in the FCU was twice that of the FCR. Although the average fiber length for both muscles across all regions was similar (62.6 +/- 2.1 mm for the FCR and 63.1 +/- 4.0 mm for the FCU), the proximal fibers of the FCU were longer compared with the proximal fibers of the FCR and the distal fibers of the FCU were shorter compared with the distal fibers of the FCR. The 99% confidence interval for the second nerve branch entry into the muscles was located approximately 69 mm distal to the medial epicondyle for the FCU and approximately 73 mm distal for the FCR. CONCLUSIONS: These data show different designs of both the FCU and the FCR. The functional significance of fiber length variability is not clear but imply that, when used in tendon transfer, the properly mobilized FCU has a much greater excursion.     

Effects of extensor pollicis longus transposition and extensor indicis proprius transfer to extensor pollicis longus on thumb mechanics

PURPOSE: The spatial relationship of the extensor pollicis longus (EPL) to the thumb carpometacarpal (CMC) joint may be altered by its transposition from the third dorsal wrist compartment and by subcutaneous extensor indicis proprius (EIP) to EPL tendon transfer. Changes in tendon position could alter thumb function. This study examined changes in the EPL adduction moment arm after EPL tendon transposition from its extensor compartment or EIP transfer. METHODS: The EPL adduction moment arm at the thumb carpometacarpal joint was determined under 4 tendon conditions: (1) intact extensor pollicis longus, (2) transposed extensor pollicis longus, (3) extensor indicis proprius to extensor pollicis longus tendon transfer through an extensor retinacular pulley, and (4) extensor indicis proprius tendon transfer through a subcutaneous route. Each tendon condition was tested in 2 wrist positions: neutral and 40 degrees of flexion. RESULTS: The wrist neutral/flexion moment arms for the 4 tendon conditions, in millimeters, were 9.2/7.3, 3.6/1.2, 8.3/5.1, and 4.8/1.0. CONCLUSIONS: EPL transposition produces a significant decrease of its adduction moment arm at the thumb CMC joint, an effect exacerbated by wrist flexion. The moment arm mechanics of the pulley and subcutaneous EIP tendon transfer resemble those of the intact and transposed EPL, respectively. Diminution of the adduction moment arm could impair thumb function, especially adduction.     

Common variations of the radial writs extensors

An anatomical study of the extensor carpi radialis longus (ECRL) and brevis (ECRB) in 173 upper limbs demonstrated abnormalities in 50%. Interconnecting tendons between the longus and brevis were found in 35% of limbs, the most common variation (26%) being a tendon arising proximally from the ECRL and inserting distally with the ECRB. In 42 limbs (24%) an extra muscle (the extensor carpi radialis intermedius) was present; many were large enough to activate a tendon transfer. Two case reports illustrate the use of extra muscles in the surgical rehabilitation of patients with paralytic disorders.