You are on page 1of 13

Available online at

ScienceDirect
www.sciencedirect.com
Hand Surgery and Rehabilitation 35 (2016) 68–80

Recent advance

Principles of tendon transfers
Principes des transferts tendineux
B. Coulet
Service de chirurgie de la main et du membre supérieur, chirurgie des paralysies, faculté de médecine, université Montpellier 1, hôpital Lapeyronie,
CHU de Montpellier, avenue du Doyen-Gaston-Giraud, 34295 Montpellier cedex 5, France
Received 10 August 2015; received in revised form 19 December 2015; accepted 21 December 2015
Available online 14 March 2016

Abstract
Tendon transfers are carried out to restore functional deficits by rerouting the remaining intact muscles. Transfers are highly attractive in the
context of hand surgery because of the possibility of restoring the patient’s ability to grip. In palsy cases, tendon transfers are only used when a
neurological procedure is contraindicated or has failed. The strategy used to restore function follows a common set of principles, no matter the
nature of the deficit. The first step is to clearly distinguish between deficient muscles and muscles that could be transferred. Next, the type of palsy
will dictate the scope of the program and the complexity of the gripping movements that can be restored. Based on this reasoning, a surgical
strategy that matches the means (transferable muscles) with the objectives (functions to restore) will be established and clearly explained to the
patient. Every paralyzed hand can be described using three parameters. 1) Deficient segments: wrist, thumb and long fingers; 2) mechanical
performance of muscles groups being revived: high energy–wrist extension and finger flexion that require strong transfers with long excursion; low
energy–wrist flexion and finger extension movements that are less demanding mechanically, because they can be accomplished through gravity
alone in some cases; 3) condition of the two primary motors in the hand: extrinsics (flexors and extensors) and intrinsics (facilitator). No matter the
type of palsy, the transfer surgery follows the same technical principles: exposure, release, fixation, tensioning and rehabilitation. By performing an
in-depth analysis of each case and by following strict technical principles, tendon transfer surgery leads to reproducible results; this allows the
surgeon to establish clear objectives for the patient preoperatively.
# 2016 SFCM. Published by Elsevier Masson SAS. All rights reserved.
Keywords: Tendon transfer; Hand; Palsy; Principles

Résumé
Les transferts tendineux permettent de restaurer une fonction déficitaire en détournant des muscles restés intacts. À la main, ils sont d’autant
plus intéressants qu’ils ont pour but de restituer la préhension. En cas de paralysies, les transferts tendineux ne vivent que des échecs ou contreindications d’un geste à visée neurologique. Quelle que soit la nature du déficit, la stratégie de restauration fonctionnelle est établie suivant les
mêmes principes. Dans un premier temps, la distinction précise entre muscles déficitaires et potentiellement transférables constitue la base de la
réflexion. Ensuite, le type de paralysie dictera l’ambition du programme et la complexité des prises pouvant être restituées. À l’issue de cette
réflexion, une stratégie chirurgicale en adéquation entre les moyens (muscles transférables) et les objectifs (fonctions à restaurer) sera établie et
clairement explicitée au patient. Chaque main paralysée peut être schématisée selon 3 paramètres : 1) les segments déficitaires : poignet, pouce et
doigts longs ; 2) les performances mécaniques de groupes ou systèmes musculaires à réanimer : a) haute énergie : extension du poignet/flexion des
doigts, nécessitant des transferts forts avec une course importante ; b) basse énergie : flexion du poignet/extension des doigts mécaniquement moins
exigeants pouvant parfois être activés par la seule pesanteur ; 3) le statut des deux grands systèmes moteurs de la main, extrinsèque (fléchisseurs/
extenseurs), et intrinsèque facilitateur. Quel que soit le type de paralysie, cette chirurgie de transfert suit les mêmes principes techniques,
d’exposition, de libération, de fixation, de réglage de la tension et de rééducation. À partir d’une analyse précise de chaque cas et en suivant des

E-mail addresses: b-coulet@chu-montpellier.fr, bertrand-coulet@wanadoo.fr.
http://dx.doi.org/10.1016/j.hansur.2015.12.011
2468-1229/# 2016 SFCM. Published by Elsevier Masson SAS. All rights reserved.

B. Coulet / Hand Surgery and Rehabilitation 35 (2016) 68–80

69

principes techniques stricts, la chirurgie de transfert conduit à des résultats reproductibles, permettant d’établir avec le patient des objectifs
préopératoires clairs.
# 2016 SFCM. Publié par Elsevier Masson SAS. Tous droits réservés.
Mots clés : Transfert tendineux ; Main ; Paralysie ; Principes

1. Introduction
Tendon transfers are carried out to restore functional deficits
by using the remaining intact muscles. This is one of the most
interesting fields within hand surgery because it aims to restore
the hand’s primary function—the ability to grip.
For a long time [1,2], surgeons have been fascinated by
tendon transfers, with multiple techniques having been
developed to address various types of palsy. In contrast to
primary procedures aimed at neurological function, tendon
transfers provide more reproducible results, but require a more
detailed clinical analysis of the deficit to establish an ‘‘a la
carte’’ strategy to restore function.
Articles on this topic are as plentiful as the surgical
techniques, making it difficult and tedious to describe each in
detail. The objective of this review is to provide a framework as to
how to construct a tendon transfer strategy appropriate for each
type of case through a systematic analysis of each type of palsy.

Fig. 1. Basic muscle biomechanics. The muscle’s mechanical outputs are its
excursion and its force-generating ability. Excursion is a function of the length
of the muscle fibers and the pennation angle (a). Force is a function of the
muscle’s volume (V) and its length (‘). The cross-sectional area (CSA) is equal
to V/‘. The physiologic CSA (PCSA) takes into account the pennation angle
(a): PCSA = V.cos a / l.

2. Review of anatomy and biomechanics
2.1. Motor innervation of hand

2.2. Muscle biomechanics

The motor innervation of the hand involves three major
nerves with specific territories (Table 1):

Like a cylinder in a motor, the mechanical performance of a
muscle is determined by its volume. It corresponds to the
product of the piston diameter or muscle cross-section area
(CSA) and its excursion or the shortening of its fibers during
contraction, with the fibers themselves being proportional to
muscle length [3,4]. Thus, a larger muscle can produce more
force and a longer muscle will have a greater excursion. Other
parameters such as muscle fiber type must be added to this
simplistic model, as it affects contraction speed and the
muscle’s ability to withstand prolonged efforts [5]. These
concepts are modulated by pennation angle, which is the angle
between the axis of the muscle fibers and that of the terminal
tendon. A larger pennation angle increases the contraction
strength but reduces the excursion, similar to a gearbox (Fig. 1).
The force-generating ability of muscles in the upper limb
have been studied extensively, particularly by Brand [6], who 

the radial nerve opens the hand through extension of the
elbow, wrist and fingers; 
the median nerve allows precise thumb-finger pinching
through motricity of the thumb and index, along with
independence of the long fingers through the flexor digitorum
superficialis (FDS); 
the ulnar nerve is the nerve of forceful grasping as it controls
the flexor digitorum profundus (FDP), intrinsic muscles in the
long fingers and deep thenar muscles.
Although this schematization only applies to trunk palsy
cases, it helps to frame the broad objectives and limitations of a
functional restoration program.
Table 1
Best muscles to transfer.
Nerves

Motor territories

Main transferable muscles in territory

Radial

High (humerus): BR, ECRL, ECR
BLow (PIN): ECU, EDC, EPL, EPB, EIP, LAP, EPD
High (elbow): FCR, PL, FDS MII V, FPL, FDPII III
Low (wrist): Opp, CF, Lb
High: FCU, FDPII IV V
Low: Add, IO

BR, ECRL, EPB, EIP

Median
Ulnar

FCR, PL, FDS MII
FCU

V

B. Coulet / Hand Surgery and Rehabilitation 35 (2016) 68–80

70

Fig. 2. Representation of the mechanical properties of hand muscles as a function of their excursion and force according to Brand [6].

described them in terms of their mechanical performance
(Fig. 2). In this diagram, each muscle is represented by its
mechanical performance (force and excursion).
On the whole, three separate muscle systems can be defined: 
finger flexors with high force generation and greater than
70 mm excursion; 
finger extensors, which are weaker and have less than 50 mm
excursion; 
wrist flexors and extensors, which have less than 30 mm
excursion but have significant force-generating ability.
As for the thumb, its muscles have the same mechanical
properties as those of the long fingers, but with less excursion.
2.3. Muscle synergy
The motor muscles of the long fingers and those of the wrist
are synergistic, with finger flexion associated with wrist
extension and vice-versa—this is called the wrist tenodesis
effect. As a consequence, the motricity of the wrist reinforces
the action of the finger flexors and extensors due to
simultaneous contraction. This synergy in certain muscles in
the motor pattern implies that they can easily be substituted one
for another during tendon transfers.
In the context of tendon transfers, functional synergies are
supplemented by the summation of the excursion of the various
transfers, bringing into play the wrist’s tenodesis effect. The

Fig. 3. Results of the summation of tendon excursions due to the wrist tenodesis
effect when restoring the finger flexors. ECRL: extensor carpi radialis longus;
FDP: flexor digitorum profundus. Tenodesis effect  30 mm = 50 to 60 mm.

flexors require 70 mm excursion for full finger curling; if a
wrist extensor is used to revive them, its 30 mm excursion is not
enough, but when combined with the additional 30 mm
excursion that occurs with tenodesis during wrist extension,
the fingers can be curled almost fully. The missing 10 mm
requires that excessive tension be placed on the transferred
tendon, which often causes a residual flexion deformity (Fig. 3).
To revive the extensors, 50 mm of excursion is needed; a wrist
flexor can only provide 30 mm excursion but when combined
with the 30 mm excursion that occurs with tenodesis during
wrist flexion, full extension is possible (Fig. 4).
Brand’s representation of the force-generating ability of hand
and wrist muscles [6] (Fig. 2) can be used to identify two systems
within the wrist and finger muscles: a high energy system that
requires a long excursion and high force production, which
corresponds to the finger flexors and wrist extensors, and a low
energy system that requires less excursion and force, which
corresponds to wrist flexion and finger extension. While the first
system always needs to be restored using a strong muscle
transfer, in some cases, the second can simply be activated by
gravity during wrist flexion and a weaker muscle transfer or
simple tenodesis effect for the extensors (Fig. 5).

Fig. 4. Results of summation of tendon excursions due to the wrist tenodesis
effect when restoring the finger extensors. EDC: extensor digitorum communis;
FCR: flexor carpi radialis; FCU: flexor carpi ulnaris. Tenodesis
effect  30 mm > 50 to 60 mm.

B. Coulet / Hand Surgery and Rehabilitation 35 (2016) 68–80

71

Fig. 5. Diagram of the various motor systems of the hand muscles. There are two types of extrinsic systems (high energy and low energy) based on their excursion and
contractile force and an intrinsic (facilitator) system. EDC: extensor digitorum communis; EIP: extensor indicis proprius; FCR: flexor carpi radialis; FCU: flexor carpi
ulnaris; FDS: flexor digitorum superficialis; FDP: flexor digitorum profundus.

3. Functional systematization of the paralyzed hand
The significant number of muscular motors needed to
activate the various degrees of freedom in the joints of the hand
is not compatible with the small space in the hand. The solution
is to use two systems: an extrinsic one where the motors are
located further away in the forearm and an intrinsic system that
is located in the hand itself. The two are cleverly synergistic and
complementary.
For the thumb, the main function of the intrinsic system is to
position of the first column; activation of the pinching motion is
mainly controlled by the extrinsic system.
For the long fingers, the two motors of flexion and extension,
extrinsic and intrinsic, have offsetting application points, which
allows them to be complementary and synergistic: 
the extrinsic system activates the flexion of the proximal
interphalangeal (PIP) and distal interphalangeal (DIP) joints,

and the extension of the metacarpophalangeal (MCP) joints,
while the intrinsic system controls MCP flexion and PIP and
DIP flexion. This implies that the extrinsic system is
essential for opening and closing the fingers, but that without
the action of the intrinsic system, their extension is
incomplete, stopping at the MCP joints, and their flexion
is uneven with early MCP flexion inhibiting proper gripping
(Fig. 6); 
the extrinsic flexors are powerful motors with significant
excursion, making them a high-energy system; conversely,
the extensors require less force and excursion and are lowenergy motors (Fig. 5).
Wrist motricity is the final parameter to take into
consideration as it reinforces the action of the extrinsic system.
Flexion belongs to the low-energy system and can mostly be
ensured through gravity alone. In contrast, active wrist
extension is essential for strong grasping actions; any
deficiency here must be addressed first.

Fig. 6. Finger flexion and extension in patients with palsy of the intrinsics: note the uneven curling of the fingers with the PIP and DIP joints flexing first, incomplete
PI extension and MCP hyperextension.

72

B. Coulet / Hand Surgery and Rehabilitation 35 (2016) 68–80

With this in mind, all hand palsy conditions can be grouped
into one of three segments and one of three systems to establish
a functional restoration strategy using tendon transfers, as
follows: 
segments: wrist, thumb and long fingers; 
systems: high-energy, low-energy, and adjuvant intrinsic
(facilitator) system: 
wrist extension and finger flexion make up the high-energy
system; as a consequence, they are the priority for
movement restoration, 
finger extension and wrist flexion make up the low-energy
system; restoring these movements involves tendon
transfers on the finger extensors, or even a simple tenodesis
of the extensors on the radius, with wrist flexors being
restored (only) very occasionally, 
the intrinsic system harmonizes and reinforces the above
two systems.

Each type of palsy can be represented by a diagram that
integrates these different segments and systems; this diagram
can be used to establish the reconstruction strategy (Fig. 7).
4. When should tendon transfer be proposed?
Tendon transfers are secondary procedures when nerve
recovery is no longer possible. There are two main indications:
after nerve repair or transfer in which recovery does not
conform to the time frame of axonal regrowth, or because such
procedures are not feasible because of the time elapsed after the
nerve injury or because the patient’s medical condition is
particularly poor.
Progress of nerve regeneration (one millimeter per day) can
be assessed by locating the axonal regrowth front with Tinel’s
sign. The correspondence between theoretical recovery and
clinical recovery can be evaluated (Fig. 8). An electromyogram
can be used to detect early signs of muscle re-innervation
before muscle contraction can be detected clinically.

Fig. 7. Systematization of basic hand functions according to anatomical
segments (wrist, thumb and fingers) and motor systems: high energy and
low energy extrinsics (flexors and extensors), intrinsic facilitators (adjuvants).

Fig. 8. Method used to determine the progress of axonal regrowth during nerve
recovery to help establish a prognosis. The brachioradialis (BR) is the first
muscle to be re-innervated.

The time elapsed after nerve injury is very important; a muscle
that has been denervated for too long cannot be re-innervated
even if the nerve is repaired successfully. It is typically said
that motor recovery is unrealistic beyond 12–18 months
of denervation (cumulative time for treatment and nerve
regrowth) [7,8].
The patient’s predisposition is also an important factor:
axonal regrowth is limited after 40 years of age; continued
smoking will also be very detrimental. Other patient-specific
factors must be considered such as motivation, free time to
undergo rehabilitation, and access to specialized rehabilitation
support.
Lastly, as with any surgery to address a disability, a contract
must be established with the patient that clearly outlines the
goals. The contribution of tendon transfer surgery must be
clearly specified, especially with higher disability levels. To
facilitate the implementation of such a contract, it may be
useful to specify which movements that are used in daily life
will be gained or improved through surgery.
If re-innervation is impossible given these parameters, tendon
transfer becomes the only alternative to restore the ability to grip,
with its extent being clearly laid out for the patient.

B. Coulet / Hand Surgery and Rehabilitation 35 (2016) 68–80

5. Performing a tendon transfer
Although each type of transfer has its own features, they
must all be performed by following common rules detailed
below [9].
5.1. Surgical approach
Selecting the surgical approach is essential when performing
a tendon transfer as it must follow a strict set of requirements. It
must allow release of the muscle being transferred and exposure
of the distal tendon attachment, while keeping the functional
cutaneous units intact and avoiding any restrictive bridle scars.
The transferred muscle must pass through the fewest possible
incisions to reduce the potential for adhesions as much as
possible. Lastly, the incision must avoid cutaneous regions with
precarious vascularization and the development of bridle scars.
Globally, there are two broad opposing strategies for surgical
approaches: 
a large incision provides access to all of the areas needed to
release and reattach the transferred tendon. The best example
is the lateral approach in the forearm used for transferring the
extensors or flexors; 
multiple smaller incisions over the working areas that are
joined by subcutaneous tunnels. The best example is the
transfer around the ulna of the extensor indicis proprius (EIP)
to restore thumb opposition [10,11].
5.2. Release
Release of the muscle being transferred is an essential step.
It helps to increase its excursion, reorient some of the
tendon and muscle fascicles, and allows for a straighter line of
pull, without the angulation that causes excessive resistance.
The best example of the importance of this release is the
brachioradialis (BR). This elbow flexor has fascial attachments
all over the forearm. If the BR is only released distally, its
excursion is extremely limited—5 mm at the most. The
maximum excursion can only be obtained after completely
releasing its tendon and muscle body over the entire forearm.
The muscle–tendon unit has been released sufficiently when
pulling on the tendon results in 25–30 mm of lengthening.

73

The trajectory can also confer an accessory function to the
transfer; for example, transferring the flexor carpi ulnaris (FCU)
on the finger or wrist extensors will add a pronation component if
it is passed around the radius, supination if passed around the ulna
and be neutral if passed through the interosseous membrane.
Furthermore, in the first two scenarios, it will be reinforced by the
pronation–supination movements of the forearm.
5.4. Distal fixation
Special attention must be paid to distal fixation of the
transferred tendon for several reasons: 
first and foremost because it determines the action of the
transfer: when fixed to the tendon of the muscle being
revived, it will not alter its function, but when fixed to a bone
segment, its action is altered or even amplified. This means
that when the fixation point is further away from the
rotational axis of the joint being moved, the moment arm of
the transfer will be larger; as a consequence, a larger force
will be applied to the joint, but a larger excursion will be
needed for the same movement amplitude (Fig. 9). During a
thumb opposition transfer, the pronation component will be
larger when it is fixed more dorsally on the MCP; 
the fixation method must allow for early mobilization, at least
before the tendon is theoretically healed in 6–8 weeks; 
lastly, the fixation method must allow the tension on the
transfer to be set and adjusted as needed.
Whenever possible, a Pulvertaft-type weave (Fig. 10) should
be used as it combines solid fixation with the ability to set the
tension. Transosseous bone fixation with a tunnel or an anchor
may be needed in some cases, but it limits the surgeon’s ability
to set the tension.
It is also important to make sure that the tendon will be long
enough when harvesting the transfer; one extra centimeter can
largely determine the quality of the fixation and tensioning of
the transplant.

5.3. Trajectory
The trajectory of a transferred muscle–tendon unit greatly
determines its action. A straighter trajectory results in the
transferred muscle having greater activity, and more importantly,
less energy lost to friction. The trajectory of a transferred tendon
is also a means to alter its action. A good example is transferring
the EIP around the ulna to counter deficiency in the thenar
muscles. The action of the transferred muscle–tendon unit will
vary greatly depending on the location of the simple pulley
relative to the pisiform: when it is proximal to the pisiform, the
transfer has a larger abduction component; when it is distal to the
pisiform, the adduction component will be greater.

Fig. 9. Consequences of distal fixation of a transferred tendon relative to the
rotational axis of the mobilized joint on the flexion moment arm and the joint’s
range of motion.

Fig. 10. Distal fixation of a tendon transfer using a Pulvertaft weave.

74

B. Coulet / Hand Surgery and Rehabilitation 35 (2016) 68–80

5.5. Setting the tension
Tensioning the transplant is probably the most important
step and the most delicate when performing a tendon transfer.
The surgeon must keep the following broad principles in mind: 
contrary to excessive tension, rehabilitation can never make
up for inadequate tension; 
excessive tension can cause muscle dysfunction.
Two methods are used to set the tension: 
for postural transfers such as restoring elbow or wrist
extension [12], these tendons must have maximum tension on
them so that only 50% of the movement amplitude against
gravity is restored. This means that a wrist extension transfer
[13] must be tensioned so that flexion against gravity does not
exceed the neutral position; 
transfers aimed at restoring finger actions have special
considerations. The tension must be such that, when
restoring the flexors, the fingers are practically fully closed
by wrist extension only, while maintaining full extension of
the fingers when the wrist is flexed. The same principles
apply to restoring finger extension: the transfer must be
tensioned so that it produces MCP extension up to the
neutral position of the wrist when straight, while
maintaining full flexion of the fingers when the wrist is
extended.
These tension adjustments must take into account the joints
above when transferring polyarticular muscles. In fact, the
tension on a transferred extensor carpi radialis longus (ECRL)
or BR, which has its proximal attachment on the humerus, will
be determined by the elbow’s position: fixation with the elbow
extended will place excessive tension on the transfer and cause
it to release when the elbow is flexed. Thus, it is better to set the
tension on these transfers with the elbow flexed, which is not
obvious on the surgery table.
5.6. Adjuvant procedures
An adjuvant procedure is one that will increase the
effectiveness of the transfer. Without going into too much
detail, these include supplemental arthrodesis that simplifies
the polyarticular chain, or an associated tenodesis that helps to
neutralize or shift the transferred tendon’s action. A very
specific example of these adjuvant procedures exists in the
thumb: crossing tenodesis of the flexor on the extensor such as
the New Zealand procedure [13].

will negatively affect the outcome because of the risk of suture
failure and adhesions.
No matter which rehabilitation protocol is used, the sutures
must be protected outside of rehabilitation work throughout the
6–8 weeks tendon healing period. This is accomplished by
positioning the upper limb and the various joint segments so
that the sutures are under the least amount of tension possible.
In certain multiple transfer cases, some of the positions are not
compatible with each other, requiring that compromises be
made.
Early mobilization of the tendon transfers is a debated
concept; it reduces the development of adhesions and facilitates
the integration of the transfer, however it could cause the
sutures to release if excessive loads are applied. A tendon repair
remains plastic up to the 8th week. Early mobilization is
feasible as long as it is initially passive and protected by using
the tenodesis action of the joints (wrist and elbow) crossed by
the transfer. Active mobilization must be performed very
carefully at first, with greater loads being applied up to the 6th
week and continued protection used up to the 8th week.
It is typical to immobilize transfers for 3 weeks to obtain
complete healing of the soft tissues; adhesions will not have
developed yet during this time period. Starting in the 4th week,
the transfers will be ‘‘awakened’’ mainly through passive
motion using the tenodesis effect in the wrist and elbow. Active
loading can be initiated during the 6th week; patients will only
be allowed to perform resisted contractions starting in the 8th
week.
Cortical integration of the transfers is generally not a
problem, especially when the transferred muscles are synergistic with the restored functions. Nevertheless, nonsynergistic
transfers will also lead to satisfactory results. In some more
challenging cases, muscle stimulation and biofeedback are
needed to improve integration of the transfer.
In the particular case of central paralysis in brain-damaged
patients, or of chronic deficits observed following obstetric
brachial plexus palsy, integration of transfers into a new
function is more difficult and uncertain. In such cases, it is
essential to evaluate the patient’s ability to control the muscle
being transferred before the surgery. Even so, these transfers are
less effective, particularly in terms of excursion and integration
into the motor pattern. They act in two ways: suppression of
their deleterious action that generate abnormal postures, but
also by restoring function in part through tenodesis effect.
6. Broad clinical forms
Using the framework outlined above, we will look broadly at
the main types of upper limb palsy and build a strategy to treat
them. The treatment strategy is based on the same principles:

5.7. Postoperative phase and rehabilitation
The postoperative and rehabilitation phases are very
important. They aim to allow the tendon sutures to heal while
limiting adhesions, and to ensure that the newly restored
functions are integrated into the overall motor pattern. Any
complication, such as hematoma, infection or wound dehiscence, 

which functions are deficient? 
which muscles can be transferred? 
which are the most important functions to restore given the
segments and systems? 
segments: wrist, thumb and fingers, 
systems:

B. Coulet / Hand Surgery and Rehabilitation 35 (2016) 68–80

75

– extrinsic high-energy (1st priority): wrist extension and
finger flexion,
– extrinsic low-energy: wrist flexion and finger extension,
– intrinsic facilitator: thumb, positioning/stabilizing of
finger MCP joints.
6.1. High radial palsy
High radial palsy is the likely the most satisfying indication
for tendon transfers [14,15] (Fig. 11, Table 2). Both wrist and
finger extension are paralyzed. Two of the systems are deficient,
the ‘‘high-energy’’ one, which is the priority for restoring wrist
extension, and the ‘‘low-energy’’ one that controls finger
extension. The hand’s intrinsic function is intact and there are
no major sensory deficits in the hand.
The priority in this scenario is restoring wrist extension,
which is a strong, postural transfer that determines grasping
actions. The absence of active finger extension is particularly
problematic when an object must be held deep in the palm; but
this is a low-energy transfer that is largely compensated by the
tenodesis effect of wrist flexion, as long as some resistance is
applied to the extensors.
There are several, powerful muscles that can be transferred:
pronator teres (PT), the wrist flexors (FCU, flexor carpi radialis
[FCR] and palmaris longus [PL]), or even FDS. Independent of
the restoration strategy, the PT is always used: its distal tendon
is shaped into a tube and transferred to the extensor carpi
radialis brevis (ECRB) in a manner that does not induce to
much radial deviation, as this would negatively affect grasping
strength.
Three muscles can be used to restore finger extension [15–
17]: 
FCR: it can be passed through the interosseous membrane or
around the radius; this is a synergistic muscle; harvesting it
will not result in frontal plane destabilization of the wrist; 
FCU: it is passed on the ulnar side of the forearm; this is a
powerful muscle that, because of its trajectory, will be
reinforced by forearm pronation, which is often used during
grasping. The major drawback is that transferring the FCU will
destabilize the wrist, with greater tendency to radial deviation
during extension, a factor that alters grasping strength [18]; 
the FDS in the 4th finger can be used to restore finger
extension; this muscle is not synergistic but it is powerful and

Fig. 11. Systematized representation of the various nerve trunk palsies affecting the hand to identify the functions that need to be restored and the muscles
that can be transferred. This can help the surgeon establish a tendon transfer
strategy for each type of palsy. High radial nerve palsy.

more importantly has a long excursion, which could be
particularly useful in cases of wrist contracture.
To restore thumb extension, the PL will be used if present in
the first two scenarios; otherwise, all of the thumb and long
finger extensors will be restored using the same motor. It may
be relevant to restore one of the functions of the abductor
pollicis longus (APL); some surgeons advocate using the PL; in
some cases, APL tenodesis on the dorsal radius will allow better
opening of the first web space during wrist flexion.
The outcomes of tendon transfers for patients with high
radial palsy are probably one of the most reproducible and
satisfying. The patients can expect to regain quasi-normal
function of their hand.
6.2. Low radial palsy
This palsy can be superimposed on the previous one, but the
fact that wrist extension is still intact greatly improves the
prognosis (Fig. 12, Table 3).
The strategies to restore the extension of the long fingers are
superimposable; the only exception is that the extensor carpi
ulnaris (ECU) should not be used because the FCU is defective
in these patients; the radial extensors of the carpus are present,
which makes radial deviation of the wrist even worse during its
extension.

Table 2
Proposition of strategy while facing a high radial nerve palsy.
Nerves

Functions or muscles to restore

Transferable muscles

Strategy

High radial

Wrist extension

RP +++
FCU 
FDS III 
FCR +++
FCU+++
FDS IV +++
PL ++

Transfer RP onto ECRB

Wrist extension: EDC, EPL,  APL

Transfer FCR onto EDC, PL onto EPL, tenodesis APL to radius
Transfer FCU onto EDC, PL onto EPL, tenodesis APL to radius
Transfer FDSIV onto EDC, PL onto EPL, tenodesis APL to radius

B. Coulet / Hand Surgery and Rehabilitation 35 (2016) 68–80

76

Fig. 12. Low radial nerve palsy.

Fig. 13. Low median nerve palsy.

The BR can also be used to restore finger extension because
it is not paralyzed, however this muscle’s excursion is limited
and its activation is greatly affected by the elbow’s position,
hence why it was nicknamed ‘‘bad boy’’ by Möberg.
6.3. Low median palsy
Palsy of the median nerve at the wrist mainly affects sensory
aspects because the thumb–finger pinch is done blindly [19]
(Fig. 13, Table 4). From a motor standpoint, the thumb column
is mainly affected due to paralysis of the superficial thenar
muscles, especially since the median nerve’s territory is more
important than the ulnar nerve’s territory. In general, abduction
and antepulsion are deficient. Many different muscles could be
transferred but the two main ones are: 
the PL, when present, using Camitz’s technique: the PL
tendon is extended using the volar aponeurosis, then attached
to the distal insertion of the abductor pollicis brevis; 
the EIP, which is likely more effective [20]: the transfer can
have two trajectories, either through the interosseous
membrane proximally to the pronator quadratus, which
places emphasis on abduction, or around the ulna, but with a

Fig. 14. Result of transferring the extensor indicis proprius around the ulna
based on its trajectory relative to the pisiform.

pulley proximal to the pisiform to provide both antepulsion
and abduction (Fig. 14).
6.4. High median palsy
In this scenario, intrinsic deficit of the thenar muscles due to
low median palsy is associated with deficits in the extrinsic

Table 3
Proposition of strategy while facing a low radial nerve palsy.
Nerves

Functions or muscles to restore

Transferable muscles

Strategy

Low radial

Finger extension: EDC, EPL,  APL

FCR +++
FCU+++
FDS IV +++
PL ++

Transfer FCR onto EDC, PL onto EPL, tenodesis APL to radius
Transfer FDSIV onto EDC, PL onto EPL, tenodesis APL to radius

Table 4
Proposition of strategy while facing a low median nerve palsy.
Nerves

Functions or muscles to restore

Transferable muscles

Strategy

Low median

Thumb positioning

EIP +++ PL ++
FCU 
FDS IV 

Transfer EIP around ulna or through interosseous membrane Transfer PL

B. Coulet / Hand Surgery and Rehabilitation 35 (2016) 68–80

77

Fig. 16. Low ulnar nerve palsy.

Fig. 15. High median nerve palsy.

flexors: flexor pollicis longus (FPL), the FDP and FDS of digit
II, and the FDS of digits III to V (Fig. 15, Table 5).
Thumb opposition will be restored using the previously
described techniques. Lateral anastomosis of the FDP for digits
III–V to that of digit II will restore index flexion; the BR can be
transferred to the FPL to restore thumb flexion. All of these
transfers are carried out during the same surgery session.
Cases of anterior interosseous nerve palsy with associated
FDP-II and FPL palsy will be treated using the same strategy.
6.5. Low ulnar palsy
High ulnar palsy mainly affects the intrinsic muscles of the
hand (Fig. 16, Table 6). Thumb adduction is affected the most,
along with a pronation deficit of the thumb column that
prevents pulp-to-pulp contact. Correction of this deformity
rests on transferring the EIP around the ulna, but this time
passing it distally relative to the pisiform to enhance its
adduction and pronation functions.

The hallmark of this palsy is a deficit of the intrinsics in the
long fingers. The lack of MCP stabilization causes uneven
rolling of the finger chain during flexion, since the PI joints flex
before the MCP, making palm contact difficult to achieve
during grasping; similarly, finger extension leads to a PI
deficiency and MCP hyperflexion (Fig. 6). This deformity is
more prominent in the two most ulnar-sided fingers, leading to a
claw hand deformity.
Numerous methods have been described to stabilize the
MCP joints. Among the passive methods, the capsulodesis
described by Zancolli and Cozzi [21] is the most widely used; it
consists of shortening the volar plate to induce a flexion
deformity of the MCP joints. The MCP joints can be stabilized
actively. In the previous case, since the two flexors are
functional, the superficial one is detached distally, turned over
the A1 pulley and then fixed to itself. In this configuration, the
FDS becomes a flexion motor for the MCP joints. Other
techniques use a wrist extensor that is extended with tendon
grafts to revive the interossei muscles [22]. The tension on these

Table 5
Proposition of strategy while facing a high median nerve palsy.
Nerves

Functions or muscles to restore

Transferable muscles

Strategy

High median

Thumb positioning:
Thumb flexion: FPL

EIP +++
BR +++
ECRL 
FDP IV V
ECRL 

Transfer EIP around ulna or through interosseous membrane
Transfer BR onto FPL

Finger flexion: FDP

IIIII

Anastomosis FDP

IVV

onto FDPII

III

Table 6
Proposition of strategy while facing a low ulnar nerve palsy.
Nerves

Functions or muscles to restore

Transferable muscles

Strategy

Low ulnar

Thumb positioning: adductor

EIP +++
PL +
FCU 
FDS IV
FDS
ECRL + graft

Transfer EIP around ulna or through interosseous membrane

Stabilization of MP in digits IV and V

Active: lasso FDS (Zancolli II)
Transfer ECRL+ Graft onto IO
Passive: capsulodesis (Zancolli I)
Tenodesis cross-over (House)

B. Coulet / Hand Surgery and Rehabilitation 35 (2016) 68–80

78

Fig. 17. High ulnar nerve palsy.

Fig. 18. Low median-ulnar nerve palsy.

transfers is more difficult to set because the same motor acts on
several rays; excessive tension can cause swan-neck deformity.
One specific subtype of this palsy is the Wartenberg
deformity. It presents as a paradoxical adduction of the 5th
finger due to a deficit of the intrinsics in the 4th space. Several
techniques have been described to correct this deformity, in
particular use of the extensor digiti minimi, which is rerouted
under the intermetacarpal ligament in the 4th web space. One
must be aware that in some cases, stabilization alone of the
MCP joints will be sufficient to correct the deformity.

For the correction of the intrinsics deficit, the same
techniques can be used as in the previous case, except for
the Zancolli lasso [21], because the FDS muscles would then be
the only actors in the last two rays to provide finger curling. In
certain cases, the independence of the pinch grip can be
maintained by transferring the ECRL on the FDP of digits III–
V. But this transfer has limited indications; as seen above, the
excursion of the wrist extensors cannot fully compensate for
that of the flexors.
6.7. Low median–ulnar nerve palsy

6.6. High ulnar palsy
Complex low median and ulnar nerve palsy cases combine
deficits of the intrinsic muscles of the thumb and long fingers
(Fig. 18, Table 8). According to the principles outlined
previously, these functions will need to be restored by a
positional transfer on the thumb column and active or passive

High ulnar palsy adds a flexion deficit of the two ulnar
fingers to the previous clinical picture (Fig. 17, Table 7). Lateral
anastomosis of the ulnar FDP tendons to that of the index will
generally result in better curling of the finger chain.
Table 7
Proposition of strategy while facing a high ulnar nerve palsy.
Nerves

Functions or muscles to restore

Transferable muscles

Strategy

High ulnar

Thumb positioning: adductor

EIP +++
PL +
FCU 
FDS IV
FDS
ECRL + graft

Transfer EIP around ulna or through interosseous membrane

Stabilization of MP in digits IV and V

Finger flexion: FDP

III IV V

FDP II
ECRL

III

++

Active: lasso FDS (Zancolli II)
Transfer ECRL+ Graft onto IO
Passive: capsulodesis (Zancolli I)
Tenodesis cross-over (House)
Anastomosis II III FDP onto FDPIV

V

Table 8
Proposition of strategy while facing a low median-ulnar nerve palsy.
Nerves

Functions or muscles to restore

Transferable muscles

Strategy

Low
MedianUlnar

Thumb positioning

EIP +++
PL +
FCU 
FDS IV
FDS
ECRL + graft

Transfer EIP around ulna or through interosseous membrane

Stabilization of MP in digits II to V

Active: lasso FDS (Zancolli II)
Transfer ECRL+ Graft onto IO
Passive: capsulodesis (Zancolli I)
Tenodesis cross-over (House)

B. Coulet / Hand Surgery and Rehabilitation 35 (2016) 68–80

79

Table 9
Proposition of strategy while facing a high median-ulnar nerve palsy.
Nerves

Functions or muscles to restore

Transferable muscles

Strategy

High MedianUlnar

Thumb positioning

EIP +++

Active: transfer EIP around ulna Passive: trapeziometacarpal fusion

Stabilization of MP in digits II to V

FDS
ECRL + graft

Active: Lasso FDS (Zancolli II)
Transfer ECRL+ Graft onto IO
Passive: capsulodesis (Zancolli I)
Tenodesis cross-over (House)
Transfer ECRL onto FDPV V

Finger flexion: FDP

III IV V

ECRL

MCP stabilization for the long fingers. In these patients, who
also have significant sensory deficits in the hand, the extrinsic
function is still present.

In these cases of severe palsy, it is evident that even after
tendon transfer, the hand function will still be abnormal, in part
because of a deep sensory deficit, but also because these
patients have somewhat closed hands.

6.8. High median–ulnar nerve palsy

7. Conclusion

High median and ulnar nerve palsy cases are the most severe
nerve trunk deficit; they also correspond to the clinical picture
described for low-level tetraplegia [13,23,24] (Fig. 19, Table 9).
As in cases of low palsy, these patients have completely
deficient intrinsic muscles and more importantly, the extrinsic
flexor muscles.
Two surgical sessions are often needed: the first to restore
the extrinsic system and the second to perform an intrinsic
procedure to position the thumb column and stabilize the MCP
joints, as described above.
When it comes to restoring the extrinsic functions, any of the
muscles innervated by the radial nerve can be transferred. In
general, the ECRL is transferred on the FDP tendons after
synchronization, and the BR transferred on the FPL to restore
the index–thumb pinch. Synchronization of tendons receiving a
transfer consists of suturing them together so that the transfer
acts the same way on the various muscles being revived. This
also makes it possible to set a different tension on each ray
(physiological digital cascade). For the flexors, a suture will be
placed around the four tendons beforehand, distal to the transfer
fixation area.

Tendon transfer surgery in the hand is generally a ‘‘winner’’
as long as the initial deficit is analyzed correctly to establish a
case-specific surgical strategy and to enter into a contract with
the patient that has clear, realistic objectives. Every palsy case
can be classified in a simple, precise system that uses three
segments (wrist, thumb, fingers) and three systems:

Fig. 19. High median-ulnar nerve palsy. 

extrinsic high-energy (1st priority): wrist extension and finger
flexion; 
extrinsic low-energy: wrist flexion and finger extension; 
intrinsic facilitator: thumb, positioning/stabilizing of finger
MCP joints.

Disclosure of interest
The author declares that he has no competing interest.
References
[1] Smith RJ. History of tendon transfers. In: Smith RJ, editor. Tendon
transfers of the hand and forearm. Boston: Little Brown Co; 1987.
[2] Steindler A. Tendon transplantation in the upper extremity. Am J Surg
1939;44:260–71.
[3] Lieber RL, Fridén J. Clinical significance of skeletal muscle architecture.
Clin Orthop Relat Res 2001;383:140–51.
[4] Lieber RL, Fridén J. Implications of muscle design on surgical reconstruction of upper extremities. Clin Orthop Relat Res 2004;419:267–79.
[5] Henriksson-Larsén K, Fridén J, Wretling ML. Distribution of fibre sizes in
human skeletal muscle. An enzyme histochemical study in m tibialis
anterior. Acta Physiol Scand 1985;123:171–7.
[6] Brand PW. Biomechanics of tendon transfers. Hand Clin 1988;4:137–54.
[7] Alnot JY, Chammas M. Lésions traumatiques des nerfs périphériques.
Cahiers d’enseignement de la SOFCOT. Paris: Elsevier; 2007.
[8] Coulet B, Boretto JG, Lazerges C, Chammas M. A comparison of
intercostal and partial ulnar nerve transfers in restoring elbow flexion
following upper brachial plexus injury (C5-C6C7). J Hand Surg Am
1297–303.
2010;35:
[9] Tubiana R, Masquelet AC. Functional classification of the upper limb
paralyses. Hand Clin 1988;4:557–62.

80

B. Coulet / Hand Surgery and Rehabilitation 35 (2016) 68–80

[10] Burkhalter W, Christensen RC, Brown P. Extensor indicis proprius
opponens plasty. J Bone Joint Surg Am 1973;55:725–32.
[11] Bourrel P. New technic for restoration of opposition of the thumb. Med
Top (Mars) 1967;27:403–7.
[12] Chammas M, Coulet B. Paralysies de la flexion et de l’extension du coude.
EMC Tech Chir Orthop Traumatol 2014;9(3)1–22 [Article 44-3302014.
Elsevier 2014].
[13] Allieu Y, Coulet B, Chammas M. Chirurgie fonctionnelle du membre
supérieur chez le tétraplégique. Techniques chirurgicales - OrthopédieTraumatologie. Paris: Elsevier; 2003: 44–460 [16 p.].
[14] Ingari JV, Green DP. D.P. Radial nerve palsy. In: Green DP, Hotchkiss RN,
Pederson WC, editors. 6th ed., Green’s Operative hand surgery, Vol. 1, 6th
ed. Philidelphia: Elsevier; 2011.
[15] Merle M, Dautel G. La main traumatique. Tome 2. Chirurgie secondaire,
le poignet traumatique. Paris: Elsevier Masson; 1994 [488 p].
[16] Tubiana R. Problems and solutions in palliative tendon transfer surgery for
radial nerve palsy. Tech Hand Up Extrem Surg 2002;6:104–13.

[17] Seiler JG 3rd, Desai MJ, Payne SH. Tendon transfers for radial, median,
and ulnar nerve palsy. J Am Acad Orthop Surg 2013;21:675–84.
[18] Tubiana R. Tendon transfer for radial paralysis. Chir Main 2002;21:157–65.
[19] Davis TR. Median and ulnar nerve palsy. In: Green DP, Hotchkiss RN,
Pederson WC, editors. Green’s operative hand surgery, Vol. 2. Philadelphia: Elsevier; 2011.
[20] Tubiana R. Palliative treatment of paralysis of intrinsic muscles of the
thumb with tendon transfer. Ann Chir 1971;25:971–9.
[21] Zancolli EA, Cozzi EP. Atlas of surgical anatomy of the hand. Philadelphia: Churchill Livingstone; 1992 [745 p.].
[22] McCarthy CK, House JH, Van Heest A, Kawiecki JA, Dahl A, Hanson D.
Intrinsic balancing in reconstruction of the tetraplegic hand. J Hand Surg
Am 1997;22:596–604.
[23] Allieu Y, Coulet B, Chammas M. Functional surgery of the upper limb in
high-level tetraplegia: part I. Tech Hand Up Extrem Surg 2000;4:50–63.
[24] Allieu Y, Coulet B, Chammas M. Functional surgery of the upper limb in
high-level tetraplegia: part II. Tech Hand Up Extrem Surg 2000;4:64–8.