HAND/PERIPHERAL NERVE

Angiosome Territories of the Nerves of the

Upper Limbs
Matthew K.-H. Hong,
B.Med.Sc.
Michael K.-Y. Hong,
B.Med.Sc.
G. Ian Taylor, M.D.
Parkville, Victoria, Australia

Background: The use of free vascularized nerve grafts requires intimate knowledge of the blood supply of peripheral nerves. The authors aimed to demonstrate radiographically the topography of the upper limb nerves with their blood
supply, and to examine them as an application of the angiosome concept. An
angiosome is a three-dimensional block of composite tissue supplied by a single
source artery.
Methods: This anatomical study involved the meticulous dissection of four fresh
upper limb specimens injected intraarterially with a gelatinlead oxide mixture.
The nerves were tagged circumferentially with copper wire and radiographs
were taken of the nerves with their arterial blood supply. The median, ulnar,
radial, musculocutaneous, and axillary nerves were examined.
Results: The authors showed that the nerves of the upper limb were supplied
segmentally by source vessels, which reinforced the angiosome concept. The
suitability of each nerve for harvest in free vascularized nerve transfer was
assessed according to its pattern of blood supply.
Conclusions: The authors work has a wide range of clinical applications and
provides an anatomical basis for neurovascular and neurocutaneous flaps and
free vascularized nerve grafting. (Plast. Reconstr. Surg. 118: 148, 2006.)

he blood supply of the peripheral nerves

has been well studied over the centuries.
Earlier works occurred when interest in the
peripheral nervous system was dominated by
processes of nerve degeneration and regeneration. Later, the overflow of complicated nerve
injuries from the two world wars demanded a
reappraisal of the blood supply in the context of
nerve injury and repair. Significant advances
were made in our understanding of the topography and morphology of the minute vasa nervorum by the work of Sir Sydney Sunderland.13
He concluded that although the precise locations of the vasa nervorum were mostly highly
variable, any artery favorably placed to do so
would send branches to the nerve. He added
that these source arteries were reasonably constant, given that the gross neurovascular relationships of the upper limb are reasonably constant.
The introduction of the free vascularized
nerve graft by Taylor and Ham in 1976 4
From the Jack Brockhoff Reconstructive Plastic Surgery Research Unit, Department of Anatomy and Cell Biology, University of Melbourne.
Received for publication August 28, 2005; accepted December 29, 2005.
Copyright 2006 by the American Society of Plastic Surgeons
DOI: 10.1097/01.prs.0000221075.91038.08

148

prompted further study of the blood supply of

peripheral nerves. This technique has been reserved for exceptional cases of nerve injury when
(1) the recipient bed is heavily scarred as a result
of ischemia or previous radiotherapy; (2) the
defect is exceptionally large; and (3) the distant
transfer of a thick nerve is desired. To aid in the
description of donor nerves, Taylor developed a
system that classified peripheral nerves according to their blood supply, with special reference
to their suitability for microvascular free transfer
(Fig. 1).5,6
In 1987, Taylor and Palmer introduced the
angiosome concept, whereby the body is considered to consist of three-dimensional blocks of
tissue supplied by particular source arteries.7
From their study of the relationship of vessels to
nerves, they noted that the vessels hitchhiked
with the nerves. Their observation led to the
expectation that peripheral nerves would also be
supplied segmentally by source arteries.
More recently, Taylors group reexamined the
blood supply of each lower limb nerve and assessed the potential of each segment of each
nerve for vascularized transfer.8 They demonstrated the nerves with their blood supply radiographically and applied the angiosome concept
to the nerves, providing the anatomical basis not

www.PRSJournal.com

Volume 118, Number 1 Angiosomes of the Upper Limb Nerves

Fig. 1. Classification of peripheral nerves according to their suitability for microvascular free transfer, with type A being the best and type E being the worst. Type A indicates an unbranched nerve
supplied segmentally by a vessel in parallel; type B is similar but the nerve has branches. Type C has
a long vessel coursing in the epineurium of an unbranched nerve. In types D and E, the nerve has a
fragmented blood supply or many branches. Reprinted with permission from Taylor, G. I. Free vascularized nerve transfer in the upper extremity. Hand Clin. 15: 673, 1999.

only for vascularized nerve grafts from the region but also virtually any composite tissue free
flap that might include a vascularized nerve. It is
the intention of this present study to perform a
similar assessment of the upper limb nerves.

MATERIALS AND METHODS

Four upper limb specimens from three fresh
cadavers were used in this study. Although a series
of 20 specimens would have been ideal, the number of specimens studied was severely restricted by
the availability of fresh cadavers suitable for total
body perfusion, the requirement for adequate
perfusion, and the extensive work required to dissect each specimen fully. The first and second
authors performed the dissections, and the senior
author assessed progress regularly. In each cadaver, total body perfusion of the arterial system
was performed with a gelatinlead oxide mixture
made up according to the protocol described by
Rees and Taylor in 1986.9 The upper limbs were
disarticulated and the skin was carefully removed.
The radial, axillary, musculocutaneous, median,
and ulnar nerves were exposed with their source
vessels and meticulously dissected to their terminal muscles or cutaneous branches. The nerves

were tagged circumferentially with fine electrical

copper wiring at 3-mm intervals along their entire
length. The wire allowed the outline of the nerves
to be identified on radiographs and would allow
simultaneous demonstration of nerves and
vessels.8 With minimal disruption to the tissues
around the nerves, the bones were removed to
enable radiography of the injected vessels and
tagged nerves without bone shadows (Fig. 2). The
muscles were then removed and a dissecting microscope used to meticulously dissect and examine the fine vasa nervorum.
Photographs were taken through the microscope to document these tiny vessels along the
entire length of each nerve. Radiographs were
taken at each stage of dissection for all specimens.
Every radiograph was reviewed and, by consensus
of all authors, those radiographs considered to
demonstrate the extrinsic blood supply of individual nerves most representatively were digitally
traced to produce schematic diagrams of the
nerves with their blood supply. Similarly, conclusions about the general nature of the topography
of the extrinsic blood supply of the upper limb
nerves were drawn from observations made of the
radiographs and photographs across the speci-

149

Plastic and Reconstructive Surgery July 2006

Fig. 2. Dissected specimen of the median and ulnar nerves with their blood supply (left). Radiographs were taken (center) with the nerves being delineated by circumferential copper wires (right).
An arterial lead oxide injection has allowed the arteries to be demonstrated simultaneously.

mens, again by general consensus. This included

classification of the neuronal vascular supply according to the five categories devised by Taylor.5,6
Review of the radiographs before and after dissection allowed us to apply the angiosome concept
to the upper limb nerves.

RESULTS
Previous work on the neurovascular relationships in the skin10 showed the inductive effect
nerves have on nearby blood vessels, and our investigations confirmed this in the deep tissues.
Two observations about the arterial supply of upper limb nerves are worthy of mention.
First, an epineurial longitudinal arterial system accompanies each peripheral nerve and may
have choke zones and true anastomoses between
long ascending and descending branches of Yshaped arteriae nervorum. Furthermore, branch-

150

ing of this system matches the branching of the

nerve. Sometimes, the arterial branches are recurrent and opposite in direction to the nerve
branching. Interestingly, these vascular branch
patterns often reflect the choke zones between
consecutive angiosomes (Figs. 3 through 5).
Second, nerves unaccompanied by prominent
vessels receive important supply from muscles.
Where a nerve is closely related to a muscle, it may
receive small branches from the vessels within that
muscle. This occurs even where large nutrient arteries exist nearby, with tiny vessels connecting the
longitudinal anastomotic system of the nerve and
the vasculature in the muscle. For example, despite the presence of a median artery, the vasa
nervorum of the median nerve appears to freely
share arterial connections with the vascular network within the flexor digitorum superficialis muscle. These vessels from within a muscle are espe-

Volume 118, Number 1 Angiosomes of the Upper Limb Nerves

cially important when no major arteries are in
close accompaniment with a nerve. Of note is the
musculocutaneous nerve, which may directly receive a large artery from within the biceps muscle.
In general, nerve branches to muscles are supplied
by vessels traveling retrogradely from within the
muscles.
Median Nerve
The median nerve is supplied consecutively by
the brachial, ulnar, and radial angiosomes, and
then the ulnar angiosome again. In the arm, the
median nerve has a type A pattern of blood supply.
It is characterized by well-developed longitudinal
arterial channels on its surface and receives
Fig. 4. Magnified view of the longitudinal arterial system of the
superficial radial nerve showing a long recurrent branch. The artery is colored by a lead oxide injection.

Fig. 5. The ulnar nerve with circumferential copper wires before

radiography. A typical Y-shaped arteriae nervorum supplies the
nerve from the adjacent superior ulnar collateral artery.

Fig. 3. Fine radiograph and corresponding schematic diagram

of the median nerve from the upper arm to the wrist demonstrating the main features of the longitudinal arterial system of a peripheral nerve. The system traverses the entire length of the
nerve, and includes (1) choke anastomosis, (2) true anastomosis
between adjacent nutrient vessels, (3) branch point of nerve, and
(4) segmental supply of the median artery to the median nerve.

branches directly from the brachial artery or from

branches to nearby muscles such as the brachialis.
In the cubital and proximal forearm regions, the
median nerve is type E. After the inferior ulnar
collateral artery nourishes it, the median nerve
branches extensively and its blood supply is fragmented with nutrient vessels from second- and
third-order arterial branches. Nerve branches to
the pronator teres and flexor carpi radialis are
supplied by twigs from the anastomotic system

151

Plastic and Reconstructive Surgery July 2006

between the inferior ulnar collateral and recurrent ulnar arteries.
In the forearm, the median nerve is a type C
nerve, as it may have a well-developed median
artery on its surface feeding branches to the nerve
segmentally. The nerve branches to the flexor
digitorum superficialis and the median nerve in
the immediate vicinity are supplied by vessels from
the proximal arterial pedicle of the muscle, which
may give rise to the median artery. As the nerve
progresses down the forearm, the median artery
occasionally sends branches to the overlying flexor
digitorum superficialis muscle and forms an anastomotic network with the vasculature within this
muscle. The median nerve is supplied by long
branches directly from the radial artery as the
nerve leaves the cover of the flexor digitorum
superficialis. Notably, at the level of the proximal
border of the pronator quadratus, a large, prominent direct branch from the radial artery may
supply the nerve. A branch of this vessel may supply the distal portion of the flexor pollicis longus.
Beyond this, the blood supply becomes fragmented again and the median nerve gains a type
E pattern. In the carpal tunnel, the median nerve
receives no branches, but as it exits, it receives long
retrograde vessels from the superficial palmar
arch. The common digital nerves and branches to
the muscles of the thenar eminence are accompanied also by arterial branches from the superficial arch (Fig. 6).
The anterior interosseous nerve originates
from between the two heads of the pronator teres
and dives deep to run anterior to the interosseous
membrane. It has a type A pattern overall as it
receives supply from the intimately related anterior interosseous artery.
Ulnar Nerve
The ulnar nerve is supplied by the brachial
and the ulnar angiosomes. The deep branch in the
hand appears to be supplied by the radial angiosome. In the arm, the ulnar nerve is type C as it
courses distally with the superior ulnar collateral
artery on its surface. In the posterior compartment
of the arm, the ulnar nerve receives some vessels
from within the long head of the triceps muscle
(seen in all specimens).
Around the elbow, the blood supply becomes
somewhat fragmented, with a type E pattern. As it
passes behind the medial epicondyle, the ulnar
nerve is nourished by the anastomotic system
formed by the superior ulnar collateral and posterior ulnar recurrent arteries. Arterial branches

152

from the ulnar artery destined for nearby muscles

contribute to the fragmented blood supply of the
nerve as they cross it. In the forearm, the ulnar
nerve receives a type A supply as it accompanies
the ulnar artery, receiving many characteristic Yshaped arteriae nervorum.
In the hand, the ulnar nerve has a type E blood
supply pattern again. At first, the nerve is met by
long retrograde branches from the superficial
palmar arch. From then on, the ulnar nerve and
its branches receive a fragmented and complex
blood supply, with contributions from both superficial and deep palmar arches (Fig. 6).
Radial Nerve
The radial nerve is supplied by the brachial,
profunda brachii, and radial angiosomes. The posterior interosseous nerve begins in the radial angiosome and is then supplied by the posterior
interosseous artery and finally by the anterior interosseous artery.
Throughout most of its course, the radial
nerve is characterized by early and prolific branching. It first receives supply from the axillary artery
and then the brachial artery. The nerve wraps
around the humerus posteriorly in the spiral
groove, intimately related to the profunda brachii.
Here, the nerve is type B, and the extensive
branching of the nerve before and in the spiral
groove results in branches from the profunda
brachii supplying several nerve trunks simultaneously. Despite this branching, the radial collateral branch of the profunda brachii gains intimate
relation to the radial nerve and eventually appears
to be part of the longitudinal arterial system on the
nerve, making the nerve segment type C as it descends to the elbow. The middle collateral branch
of the profunda brachii artery accompanies the
branches to the medial head of the triceps and
anconeus, and the lower lateral brachial cutaneous nerve pierces the deep fascia in company with
an arterial perforator.
Around the elbow, the radial nerve gains a type
E pattern as it branches extensively and receives a
fragmented blood supply. The nerve is accompanied and nourished by the radial collateral branch
of the profunda brachii artery and the anastomosis
with the radial recurrent artery. The latter vessel
from the radial artery runs along the nerve proximally, sending Y-shaped branches to it and the
nerve branches to the brachioradialis, extensor
carpi radialis longus, and brachialis. The radial
nerve continues to receive supply from the radial
recurrent artery as it divides into its terminal

Volume 118, Number 1 Angiosomes of the Upper Limb Nerves

Fig. 6. Diagram of the blood supply of the median and ulnar nerves (left) traced from the radiograph in Figure 2. The radial nerve (right)
and its blood supply have been traced from similar radiographs. The arteries (center) are colored to demonstrate the angiosomes.

branches, the superficial radial and posterior interosseous nerves.

The posterior interosseous nerve is predominantly a type E nerve. As it exits from between the
two heads of supinator, the nerve showers into a

multitude of branches and gains a loose association with the posterior interosseous artery before
slipping deep to the extensor pollicis longus. A
nerve branch to the extensor indicis may continue
with the artery (one specimen).

153

Plastic and Reconstructive Surgery July 2006

Distally, the posterior interosseous nerve and
its terminal branches are met by vessels from the
anterior interosseous artery penetrating posteriorly through the interosseous membrane. The anterior interosseous artery turns dorsally to meet
and accompany the terminal portion of the posterior interosseous nerve, which it now supplies
directly to give a type A pattern of blood supply.
In contrast, the superficial radial nerve is a
type A nerve. It gains a close relationship with the
radial artery early and receives many short, direct
vasa nervorum. As it begins to move away from the
radial artery, the superficial radial nerve gains a
vascular pedicle that accompanies the nerve distally into the hand for a considerable distance
(Fig. 6).
Musculocutaneous Nerve
The musculocutaneous nerve runs entirely in
the brachial angiosome. It continues as the lateral
antebrachial cutaneous nerve in the radial angiosome. The musculocutaneous nerve receives a
fragmented blood supply for its entire length and
is therefore predominantly type E. Near its origin
in the brachial plexus, the musculocutaneous
nerve receives branches from the axillary artery.
From then on, in the absence of a close accompanying axial artery, this nerve relies on direct vasa
nervorum from the vascular pedicles of the coracobrachialis, biceps, and brachialis. The longitudinal system also receives contributions by retrograde vessels arising from the vasculature within
the muscles and running back along the nerve
branches. More distally, the musculocutaneous
nerve gains a close relationship with the biceps
muscle and is nourished by vessels that arise from
within the muscle. These vessels may be relatively
large should the proximal blood supply to the
nerve prove meager. The rest of the nerve is sustained by its intrinsic system reinforced at irregular intervals by small vessels arising from networks in the surrounding connective tissue.
The lateral antebrachial cutaneous nerve is
the continuation of the musculocutaneous nerve,
and more proximally is a type D nerve. It becomes
closely related to the cephalic vein and receives
arterial supply from the small chain-linked systems
of the superficial fascia, which are in turn supplied
by the radial artery (Fig. 7).
Axillary Nerve
The axillary nerve originates in the brachial
angiosome but rapidly gains its supply from the
posterior circumflex humeral angiosome. The

154

nerve is accompanied through the quadrangular

space by the posterior circumflex humeral artery.
The association between nerve and artery here
appears to be looser than other nerveartery relationships. As such, instead of short direct arteriae nervorum from the posterior circumflex humeral artery, the axillary nerve receives long direct
vessels. It also receives indirect vessels by means of
the surrounding connective tissue. Nevertheless,
the nerve and artery become more intimately related as they approach the deltoid. The branching
of the nerve and its blood supply result in a type
E nerve.
Of note is the close accompaniment of a musculocutaneous skin perforator from the posterior
circumflex humeral artery with the upper lateral
brachial cutaneous nerve. The nerveartery relationship occurs throughout from within the deltoid muscle to the skin (Fig. 7).
Cutaneous Nerves
The neurovascular relationships in the integument has been the subject of previous work.10
The authors found that the arterial relationship
with the cutaneous nerves was obvious in the region of the shoulder and palm of the hand,
whereas veins dominated the picture in the arm,
forearm, and dorsum of the hand. Our work extends this to emphasize that when cutaneous
nerves pierce the deep fascia accompanied by a
dedicated perforating artery, this artery would
contribute to a separate longitudinal epineurial
vessel observed on the nerve. When closely related
to a vein, the nerve would be followed by a chainlinked system of arteries that would segmentally
send short branches to the cutaneous nerve. The
system of arteries usually shared many connections with the vasa vasorum around the adjacent
superficial vein, and often sent branches to the
surrounding skin and superficial fascia. In addition to a close venous relationship, a prominent
companion artery would at times accompany a
nerve.
Indeed, the upper and lower lateral brachial
cutaneous nerves are accompanied by dedicated
arteries, and the relationship is formed early,
some distance before piercing the deep fascia.
The medial brachial and medial, lateral, and posterior antebrachial nerves were accompanied by
large veins and their tributaries and received arterial supply segmentally from a less conspicuous
chain-linked system.
As a cutaneous nerve passes through different
vascular territories of the skin, it comes to be sup-

Volume 118, Number 1 Angiosomes of the Upper Limb Nerves

Fig. 7. Diagrams of the blood supply of the musculocutaneous (left) and axillary (right) nerves.

plied by the same source vessels. The angiosomes

of the cutaneous nerves match the angiosomes of
the skin.

DISCUSSION
We have radiographically demonstrated the
upper limb nerves with their arterial blood supply
and shown that the arteries supplying each nerve
are derived from the source vessels of consecutive
angiosomes as the nerve crosses them. The angiosomes of the upper limb define the various tissues
that can be combined together or raised separately on the various source arteries. As certain
thick nerves with a vasculature suitable for vascularized nerve grafting may be available in rare
situations, our study included all major nerves of

the upper limb. We traced the nerves with their

vessels from the deep tissues to the skin and were
able to determine the patterns of blood supply to
the various nerve segments according to Taylors
classification system (Fig. 8). We aimed to explore
the general patterns of the extrinsic blood supply,
rather than count and tally the individual arteriae
nervorum. This work was completed by Sunderland, who aptly stated: a study of the number of
vessels alone can give no true conception of the
blood supply of a nerve.11
Several general observations were made in regard to the arterial blood supply of the upper limb
nerves. When an axial artery accompanies them,
the nerves tend to be supplied with direct arteriae
nervorum. In the absence of such an axial artery,

155

Plastic and Reconstructive Surgery July 2006

Fig. 8. Simplified diagram illustrating blood supply to the upper limb nerves. The nerve segments
are classified as types A to E according to their suitability for vascularized transfer. From left to right:
ulnar, median, radial, musculocutaneous, and axillary nerves. Some branches have been omitted for
clarity.

nerves receive blood supply from either a smaller

artery skirting the surface or a fragmented blood
supply from second- and third-order arterial
branches. Around the elbow and proximal part of
the forearm there is a tendency for the blood
supply to fragment as nerves and arteries branch
extensively to the numerous forearm muscles.
Previous work has demonstrated the various
patterns of venous drainage of the peripheral
nerves (Fig. 9).12 Perhaps of note is the striking
consistency with which the arterial patterns are
mirrored in the patterns of venous drainage to
these nerves. It is often said that the venous drainage of nerves parallels the arterial supply.
In type A segments of nerves, the venae nervorum tend to drain to the periarterial venous plexuses, whereas in type C segments there is drainage to
venae comitantes or a vein accompanying the artery
on the surface of the nerve. For example, in the arm,
the superior ulnar collateral artery runs on the ulnar
nerve together with the corresponding vein. In the
forearm, the ulnar nerve is supplied directly by the
ulnar artery, and its venous drainage is to the periarterial venous plexus.
Around the elbow, the situation appears different. Although on the arterial side the tendency

156

is for fragmentation, the venous drainage appears

to be directly to nearby veins. Cutaneous nerves
are generally followed by a chain-linked system of
arteries, and their venous drainage is to a
perivenous plexus around the usual accompanying vein.
Intrinsic Blood Supply
Sir Sydney Sunderland noted that the nature
of the intraneural vascular pattern appeared to
provide a blood supply in excess of that required
for the exclusive needs of the supporting connective tissue of peripheral nerves.3 This would explain the resilience of the neural blood supply to
gross interference. Indeed, Lundborg described
at a microscopic level the way vessels would open
or close in response to vascular challenges and
how flow in individual vessels could rapidly change
direction.13
More recent investigations have focused on
the interaction between the intrinsic blood supply
and the extrinsic system. Breidenbachs group
demonstrated that a long length of nerve can
maintain its flow based only on its intrinsic blood
supply.14 They also showed that even with reliance

Volume 118, Number 1 Angiosomes of the Upper Limb Nerves

Fig. 9. Schematic diagram of venous drainage of (A) ulnar; (B) median; (C) radial; (D) sciatic; and (E)
cutaneous nerves. Arrows designate the level of the elbow and knee. Compare this with the arterial
supply of the upper limb nerves in Figure 8. The lateral antebrachial cutaneous nerve, the continuation of the musculocutaneous nerve, is best compared with the cutaneous nerve. Reprinted with
permission from Del Pinal, F., and Taylor, G. I. The venous drainage of nerves: Anatomical study and
clinical implications. Br. J. Plast. Surg. 43: 511, 1990.

on a single extrinsic vessel, a significant length of

nerve could be maintained. This would suggest
that a long free vascularized nerve segment would
have its intrinsic arterial system sufficiently supplied by its pedicle.
Another interesting point raised by the study
was the significant blood flow achieved by vessels
of the nerve branches. This appears to correspond
with our observation of the significant number of
vessels from the muscles supplying the nerve
branches and traveling retrogradely back toward
the main nerve.
Studies by Best and colleagues have highlighted the importance of inosculation as the
mechanism for the revascularization of small-diameter nerve grafts15 and the rapidity with which
this occurs.16 Although it took 7 days for revascularization of a nerve graft from surrounding tissues, ingrowth of epineurial vessels from the proximal stump was observed after 2 days. These
findings suggest that small-diameter nerve grafts
spontaneously revascularize and microvascular
techniques are unnecessary. However, the studies
examined relatively short segments of nerve, and
it is unclear whether the early inosculation alone
is sufficient to allow adequate perfusion of the

middle segments of longer nerve grafts. In addition, the thicker peroneal nerve in the ewe in one
study failed to vascularize well, suggesting a role
for free vascularized nerve grafting.15 This highlights well the fact that rather than to replace
conventional techniques, the free vascularized
nerve graft was introduced for specific indications:
where the nerve gap is large, the recipient bed is
heavily scarred, or when the transfer or a thick
nerve is desired. More studies comparing vascularized and nonvascularized nerve grafts are required in a setting where there is considerable
challenge to the reestablishment of an adequate
blood supply.
Free Vascularized Nerve Grafting
The free vascularized nerve graft was developed not to replace conventional nerve grafting
techniques but to provide an alternative operation
and a potential solution to the problem of graft
survival in exceptional circumstances. These include situations where (1) the nerve gap is very
large (e.g., a 20-cm defect in the median nerve of
the forearm); (2) transfer of a thick nerve is desired, such as a median or ulnar nerve, which may

157

Plastic and Reconstructive Surgery July 2006

become available in an amputation stump5 or in
some cases of brachial plexus injury17; and (3) the
recipient bed is poor (e.g., a bony vascular bed, in
areas of previous radiotherapy, or where there is
extensive scarring).
In the search for donor sites, several systems
have been devised to classify the pattern of blood
supply to peripheral nerves with regard to their
suitability for free vascularized grafting. Most appear to have been developed around only those
nerves seen as potential donors, as the assumption has been that the donor site is always completely normal. In exceptional circumstances, donor nerves may become available without
morbidity. For example, an upper limb amputation stump may yield the ulnar and median nerves
of the arm as donors.5 To account for these situations, each major nerve segment of the upper
limb must be assessed for its potential in vascularized nerve grafting.
Taylors system allows all peripheral nerves to
be classified and studied, some as potential donors
in special circumstances. Under this system, type
A and C patterns of blood supply indicate the most
ideal donor nerves. Type A represents a nerve
supplied segmentally by a long unbranching artery, whereas type C is similar except the artery
courses on the surface of the nerve instead of in
parallel. Type B is similar to type A, but the nerve
divides early, and is suitable if the graft can be
reversed.
Our study examined all the nerves of the upper limb. According to our findings, we identified
the following nerves as suitable for microsurgical
transfer, being of type A or C: (1) the ulnar nerve
in the upper arm and in the forearm; (2) the
median nerve in the upper arm and in the forearm; (3) the segment of anterior interosseous
nerve distal to the flexor pollicis longus branch;
(4) the upper lateral brachial nerve; (5) the lower
lateral brachial nerve; (6) the superficial radial
nerve; (7) the terminal branch of posterior interosseous nerve; and (8) a branch to the extensor
indicis following the posterior interosseous artery
(when present).
In normal clinical situations, nerves 1 and 2
cannot be used because of their functional importance. Harvest of nerve 3 results in lost function of pronator quadratus, which may be acceptable. This leaves nerves 4 through 8 as donor
nerves for vascularized nerve transfer, and potentially nerve 3 in normal situations. It is notable that
most of these are sensory or cutaneous nerves. Our
results confirm the anatomical basis for the su-

158

perficial radial and ulnar nerves as vascularized

nerve grafts.
The following donor sites have been identified
previously for grafting: the superficial radial nerve
based on the radial artery, the ulnar nerve based
on the superior ulnar collateral artery, the median
nerve based on the brachial artery,5 the sural nerve
based on the superficial sural artery, the anterior
tibial nerve based on the anterior tibial artery, the
superficial peroneal nerve based on the peroneal
artery, the saphenous nerve based on the saphenous artery, the anterior interosseous nerve based
on the anterior interosseous artery, and the lateral
branch of the posterior interosseous nerve with
the posterior interosseous artery.18 20
The suitability of donor sites for free vascularized nerve grafts varies enormously with the demands of each clinical situation. For example,
whereas the terminal portion of the posterior interosseous nerve lacks the length to bridge long
nerve gaps, its length is adequate in most cases of
single digital nerve grafting. It may be possible to
harvest this nerve segment with the distal portion
of the anterior interosseous artery for simultaneous reconstruction of both digital artery and
nerve, which may occur in avulsion injuries of the
thumb and fingers.19 In the event of a scarred or
bony vascular bed, it is conceivable that some donor sites currently used in conventional free nerve
grafts can yield a vascularized nerve graft (e.g., the
anterior interosseous nerve transferred as a vascularized nerve graft with the anterior interosseous artery as the pedicle).
In clinical practice, the free vascularized nerve
graft has led to good functional outcomes. These
cases include large defects of the median nerve
repaired with the free vascularized superficial radial nerve, with follow-up as long as 23 years showing very good functional results.5 The free vascularized ulnar nerve based on the superior ulnar
collateral artery has also found a place in brachial
plexus reconstruction.17 However, the technique
of free vascularized nerve grafting is still reserved
for exceptional cases where conventional techniques are deemed insufficient.
Neurovascular Flaps
Knowledge of the blood supply of the upper
limb nerves demonstrates the basis for many flaps
already in use. Mathes and Nahai present a myriad
of musculocutaneous and fasciocutaneous flaps
that can be modified to make them sensate or
functional.21 For example, the deltoid flap based
on the posterior deltoid subcutaneous artery can

Volume 118, Number 1 Angiosomes of the Upper Limb Nerves

be a sensory flap based on the upper lateral brachial cutaneous nerve. The lateral arm flap has the
lower lateral brachial cutaneous nerve in company
with the cutaneous perforator of the profunda
brachii artery. In another example, the radial forearm flap, based on the radial artery, may include
the medial or lateral antebrachial nerves for sensation. In addition, the superficial radial nerve
may be harvested as a vascularized nerve graft.21
Our observation that the nerves and skin are in the
same angiosomes supports the notion that nerves
incorporated in these flaps are vascularized on
transfer. If a skin flap is based distally on a perforating artery, and its longitudinal axis incorporates one of the upper or lower lateral brachial
nerves or the superficial radial nerve, it is likely to
have a superior blood supply. The detailed description of the blood supply to the peripheral
nerves of the upper limb suggests that there is
potential for refinements to be made to existing
techniques for local or free transfer of composite
tissue.
Neurocutaneous Flaps
Since the introduction of neurocutaneous
flaps,22,23 there has been much interest in their
anatomical basis, with many reports of flaps based
on the dorsal branch of the ulnar nerve and the
saphenous and sural nerves. It is well known that
arteries always accompany nerves and that the surrounding chain-linked arterial networks in the
subcutaneous tissue are oriented in the direction
of nerves.10 Given the strength of the longitudinal
arterial systems in and around a nerve, it is not
surprising that small- to medium-sized flaps can
almost be supported by these blood vessels alone.
The flap relies on the chain-linked arterial systems
that follow in the same direction as the nerve. The
reliability of the neural blood supply inherently
allows the design of distally based neurocutaneous
flaps. Gurunluoglu et al. have demonstrated the
importance of neural vessels in these flaps in rats.24
It is worth reemphasizing that the course of
the nerve should provide the longitudinal axis of
a robust skin flap10 and that neurocutaneous flaps
are a direct application of the concept. Nerves
provide a link between adjacent vascular territories of arterial perforators by means of their continuous longitudinal systems. The effectiveness of
these connections led Sir Sydney Sunderland to
state that The anastomoses, on and within the
nerve, between nutrient arteries derived from different and widely separated major arteries form
the basis for the development of collateral circu-

lations when the major arterial channel to a limb

has been interrupted.25
Functional Muscle Transfer
Free vascularized muscle transfers taken with
their nerve branches to restore function have become increasingly popular. Our finding that vessels from within a muscle and the muscles vascular
pedicle all supply its nerve branches illustrates that
the particular nerve segment is already vascularized on transfer in this case.
Various free composite grafts that include vascularized nerves have also been described. For
example, the radial nerve has been taken together
with the brachioradialis to provide a functional
muscle in a free composite graft.5
Vascularized Nerve and Composite Tissue
Allotransplantation
The use of nerve autografts will always be limited by the availability of suitable donor sites. The
prospect of allografting in reconstructive surgery
has become more promising with recent advances
in immunosuppression therapy.26 When considering nerve allotransplantation, both the nerve
regeneration process and rejection response to
allogeneic tissue must be considered. Mackinnon
has pioneered the technique of nerve allografting,
with encouraging results.27
Vascularized nerve allografts offer several theoretical advantages: (1) to allow en bloc reconstruction of nerve plexi; (2) to enhance nerve
regeneration rate; and (3) to permit the use of
larger trunk grafts without the problem of central necrosis.28 Should they find their place in reconstructive surgery, our work has direct application to the selection of suitable donor sites.
Composite tissue allografts have also made
their way into clinical practice,26 with the first human hand transplant occurring in 1998.29 Virtually
any combination of tissues, matched to the needs
of the recipient, may be harvested. The angiosome
concept finds its place here, as it allows the reconstructive surgeon to understand what tissues or
parts of tissues, including nerves, can be taken in
combination with others based on one or more
source vessels. Such an understanding of blood
supply is important for the clinical success of these
flaps.

CONCLUSIONS
Our work has radiographically depicted the
blood supply of the upper limb nerves and shown
them to have angiosome territories that match

159

Plastic and Reconstructive Surgery July 2006

those of the surrounding composite tissues. Furthermore, we have classified each nerve segment
according to its suitability for vascularized nerve
transfer, confirming the anatomical basis for donor sites already in use and showing other sites
whose availability depends on the prevailing clinical circumstances. This work has important clinical applications in the reconstruction of nerves of
the upper limb.
G. Ian Taylor, M.D.
Jack Brockhoff Reconstructive Plastic Surgery Research
Unit
Department of Anatomy and Cell Biology
University of Melbourne
Parkville, Victoria 3010, Australia
g.taylor@medicine.unimelb.edu.au

ACKNOWLEDGMENTS

This work was supported in part by The Jack Brockhoff Foundation and The Colonial Foundation. The
authors thank the staff of the Jack Brockhoff Reconstructive Plastic Surgery Research Unit and in particular Dr.
Hiroo Suami and Dr. Wei-Ren Pan for their invaluable
support of this study. They are also indebted to the Department of Anatomy and Cell Biology of the University
of Melbourne for their assistance.
REFERENCES
1. Sunderland, S. Nerves and Nerve Injuries. Edinburgh: E. & S.
Livingstone, 1968. P. 758.
2. Sunderland, S. Nerves and Nerve Injuries, 2nd Ed. New York:
Churchill Livingstone, 1978. P. 46.
3. Sunderland, S. Nerve Injuries and Their Repair: A Critical Appraisal. New York: Churchill Livingstone, 1991. P. 53.
4. Taylor, G. I., and Ham, F. J. The free vascularized nerve graft:
A further experimental and clinical application of microvascular techniques. Plast. Reconstr. Surg. 57: 413, 1976.
5. Taylor, G. I. Free vascularized nerve transfer in the upper
extremity. Hand Clin. 15: 673, 1999.
6. Taylor, G. I. Nerve grafting with simultaneous microvascular
reconstruction. Clin. Orthop. 56, 1978.
7. Taylor, G. I., and Palmer, J. H. The vascular territories (angiosomes) of the body: Experimental study and clinical applications. Br. J. Plast. Surg. 40: 113, 1987.
8. Suami, H., Taylor, G. I., and Pan, W. R. Angiosome territories
of the nerves of the lower limbs. Plast. Reconstr. Surg. 112:
1790, 2003.
9. Rees, M. J., and Taylor, G. I. A simplified lead oxide cadaver
injection technique. Plast. Reconstr. Surg. 77: 141, 1986.

160

10. Taylor, G. I., Gianoutsos, M. P., and Morris, S. F. The neurovascular territories of the skin and muscles: Anatomic study
and clinical implications. Plast. Reconstr. Surg. 94: 1, 1994.
11. Sunderland, S. Blood supply of the nerves of the upper limb
in man. Arch. Neurol. Psychiatry 53: 91, 1945.
12. Del Pinal, F., and Taylor, G. I. The venous drainage of nerves:
Anatomical study and clinical implications. Br. J. Plast. Surg.
43: 511, 1990.
13. Lundborg, G. The microcirculation in rabbit tendon: In vivo
studies after mobilisation and transection. Hand 7: 1, 1975.
14. Maki, Y., Firrell, J. C., and Breidenbach, W. C. Blood flow in
mobilized nerves: Results in a rabbit sciatic nerve model.
Plast. Reconstr. Surg. 100: 627, 1997.
15. Best, T. J., Mackinnon, S. E., Evans, P. J., et al. Peripheral
nerve revascularization: Histomorphometric study of smalland large-caliber grafts. J. Reconstr. Microsurg. 15: 183, 1999.
16. Best, T. J., Mackinnon, S. E., Midha, R., et al. Revascularization of peripheral nerve autografts and allografts. Plast. Reconstr. Surg. 104: 152, 1999.
17. Terzis, J. K., Vekris, M. D., and Soucacos, P. N. Outcomes of
brachial plexus reconstruction in 204 patients with devastating paralysis. Plast. Reconstr. Surg. 104: 1221, 1999.
18. Terzis, J. K., Skoulis, T. G., and Soucacos, P. N. Vascularized
nerve grafts: A review. Int. Angiol. 14: 264, 1995.
19. Arnez, Z. M., and Lister, G. D. The posterior interosseous
arterial graft. Plast. Reconstr. Surg. 94: 202, 1994.
20. Merle, M., and Dautel, G. Vascularised nerve grafts. J. Hand
Surg. (Br.) 16: 483, 1991.
21. Mathes, S. J., and Nahai, F. Upper Limb. New York: Churchill
Livingstone, 1988.
22. Masquelet, A. C., Romana, M. C., and Wolf, G. Skin island
flaps supplied by the vascular axis of the sensitive superficial
nerves: Anatomic study and clinical experience in the leg.
Plast. Reconstr. Surg. 89: 1115, 1992.
23. Bertelli, J. A., and Khoury, Z. Neurocutaneous island flaps in
the hand: Anatomical basis and preliminary results. Br. J.
Plast. Surg. 45: 586, 1992.
24. Gurunluoglu, R., Bayramicli, M., Sonmez, A., et al. A neurocutaneous island flap model: An experimental study in
rats. Ann. Plast. Surg. 44: 53, 2000.
25. Sunderland, S. Blood supply of peripheral nerves. Arch. Neurol. Psychiatry 54: 280, 1945.
26. Petit, F., Minns, A. B., Dubernard, J. M., et al. Composite
tissue allotransplantation and reconstructive surgery: First
clinical applications. Ann. Surg. 237: 19, 2003.
27. Mackinnon, S. E., Doolabh, V. B., Novak, C. B., et al. Clinical
outcome following nerve allograft transplantation. Plast. Reconstr. Surg. 107: 1419, 2001.
28. Bain, J. R. Peripheral nerve and neuromuscular allotransplantation: Current status. Microsurgery 20: 384, 2000.
29. Dubernard, J. M., Owen, E., Herzberg, G., et al. Human hand
allograft: Report on first 6 months. Lancet 353: 1315, 1999.