I3riri.

d Gourd ofP/asric Surgery (1987),40, 113-141

Q 1987 The Trustees of British Association of Plastic Surgeons

The vascular territories (angiosomes) of the body:

experimental study and clinical applications
G. I. TAYLOR and J. H. PALMER
Department of Plastic and Reconstructive Surgery, Royal Melbourne Hospital and Department of Anatomy,
~n~~~ers~yof ~efbo~rne

Summary-The blood supply to the skin and underlying tissues was investigated by ink injection
studies, dissection, perforator mapping and radiographic analysis of fresh cadavers and isolated
limbs. The results were correlated with previous regional studies done in this department.
The blood supply is shown to be a continuous three-dimensional network of vessels not only in
the skin but in all tissue layers. The anatomical territory of a source artery in the skin and deep tissues
was found to correspond in most cases, giving rise to the angiosome concept.
Arteries follow closely the connective tissue framework of the body. The primary supply to the skin
is by direct cutaneous arteries which vary in calibre, length and density in different regions. This
primary supply is reinforced by numerous small indirect vessels, which are “spent” terminal branches
of arteries supplying the deep tissues.
An average of 374 major perforators was plotted in each subject, revealing that there are still many
more potential skin flaps. Our arterial roadmap of the body provides the basis for the logical planning
of incisions and flaps. The angiosomes defined the tissues available for composite transfer.

Sir Harold Gillies wrote that “Plastic Surgery
is a constant battle between blood supply and
beauty’* (Gillies and Millard, 1957)-a struggle
which has not yet been resolved. The speciality of
Plastic Surgery evolved in Europe and North
America to improve the lot of the victims of the
two Great Wars. With artistic flair and geometrical
precision tissues were advanced and rotated. They
were transposed locally and dispatched to a distant
site on a limb carrier-only to be rebuffed on
occasions by necrosis. Gradually rigid length to
breadth ratios were formulated, because the major-
ity o’f these flaps were designed without a precise
knowledge of the vessels upon which they were
based.
This anatomical info~ation was available but
as is often the case it was hidden in texts in foreign
languages, In 1889 Manchot, at the age of 23
completed his treatise, Die Hautarterien des Men-
schli~he~ Korpens-the Skin Arteries of”the Body, in
the incredible time of 6 months. This anatomical
landmark has recently been made available in
English (1983). Since Roentgen was not to make
his discovery until several years later, Manchot did
not have the advantage of radiography. Neverthe-
less, with an accuracy which has mostly stood the
test of time, Manchot identified the cutaneous
113
perforators, assigned them to their underlying
source vessels and charted the cutaneous vascular
territories of the body.
The next major work was performed by Michel
Salmon in the 1930s and was published in French
(1936a). With the aid of X-rays he reappraised the
work of Manchot. He injected entire cadavers with
a mixture containing lead oxide and mapped not
only the entire cutaneous circulation but also the
blood supply of every muscle in the body (1936b).
His work has recently been rediscovered and will
soon be available in English (Salmon, 1987, in
press).
Early this century advances on the clinical front
gave significance to the work of these great
anatomists. Tansini reported a latissimus dorsi flap
supplied by the thoracodorsal artery in 1906. In
1921 Blair described a forehead flap based on the
superficial temporal vessels and in 1929 Esser
published his book, Artery Flaps. In 1937 Webster
cited the work of Manchot when he described a
long bipedicled thoraco-epigastric flap based on
named arteries which extended from the groin to
the axilla. Shaw and Payne (1946) utilised the
information in wartime to provide one stage direct
flaps for hand reconstruction. Then in 1965,
Bakamjian drew attention to the long paramedian
114 BRITISH JOURNAL OF PLASTIC SURGERY

perforators of the internal thoracic system. The (iii) Integration of this information with that of
significance of the cutaneous circulation was about other workers.
to emerge into a new era.
Considering the results of our own work and that
The 1970s saw the beginning of the “anatomical
of others, some important anatomical concepts
revolution”. In 1973 McGregor and Morgan differ-
have emerged. They will be listed at this stage as
entiated between large flaps based on a known axial
they provide the basis for a better understanding of
blood supply and those based on random vessels in
the arterial network.
the area. Daniel and Williams (1973) reappraised
the work of Manchot and others, and on an (9 The blood supply of the body courses within
embryological and anatomical basis segregated the or adjacent to the connective tissue framework,
cutaneous arteries into direct cutaneous and mus- whether it is bone, septa or fascia.
culocutaneous vessels. (ii) Vessels course from fixed loci to mobile areas.
“The free flap” (Taylor and Daniel; Daniel and (iii) The vascular outflow is a continuous system of
Taylor) was published in 1973 and in the same year arteries linked predominantly by reduced
the musculocutaneous flap was revived by McGraw calibre vessels-the choke arteries and arter-
and co-workers (McGraw, 1980). Both procedures ioles.
demanded a precise knowledge of the cutaneous (iv) The body is a three-dimensional jigsaw made
vasculature. In the search for new donor sites for up of composite blocks of tissue supplied by
tissue transfer, surgeons scurried back to the named source arteries. The arteries supplying
dissection room. Many new and exciting techniques these blocks of tissue are responsible for the
were developed. However, there was a tendency for supply of the skin and the underlying struc-
these new techniques to neglect the aesthetic side tures. These composite units we have named
of Plastic Surgery. The results could sometimes be ANGIOSOMES.
what the late Frank McDowell (1979) has described
It is hoped that this paper will provide the reader
as “globs and blobs”.
with a basic understanding of the vascular anatomy
To escape from the hamburger of muscle and
which will help clarify nomenclature, aid in the
skin, surgeons soon rediscovered that blood vessels
planning of flaps and incisions and provide the
follow fascial planes. In the 1980s a sleek new
basis for explaining certain pathological processes,
model emerged-the fasciocutaneous flap (Ponten,
1981). Together with this there has been an
explosion of new terms and new classifications of
the cutaneous circulation. These are still in a state The background
of flux, and the thesaurus of flaps includes a The research of this department into the vascular
bewildering array of terms such as axial, random, territories of the body commenced in 1972. Since
direct cutaneous, musculocutaneous, fasciocuta- then over 2000 fresh cadaver studies have been
neous, super and septocutaneous. Indeed there has performed involving dissection, injection and ra-
been a recent attempt to classify flaps into no less diography of various regions, tissues and combina-
than 10 types and subtypes based on the origin of tions of tissues. In each case our anatomical studies
the cutaneous perforators (Nakajima ef al., 1986). were problem orientated. They were designed to
Are these terms all justified, or are they just a provide a surgical solution to patients’ needs.
different expression of the basic cutaneous vascu- Initially the investigations involved an analysis
lature that Manchot and Salmon told us about a of various arteriovenous systems to define possible
century and 50 years ago? donor sites for free skin flap transfer (Taylor and
This paper attempts to answer these questions. Daniel, 1975). Next the studies focused on other
It presents a review of the vascular territories of the tissues and included the anatomical basis for the
body based on : transfer of bone (Taylor et al., 1975) nerve (Taylor
and Ham, 1976) and certain muscles (Taylor et al.,
(9 A reappraisal of the works of Manchot and 1983). Encouraged by the success of some of the
Salmon by a detailed series of dissections and resulting clinical procedures, the research expanded
radiographic studies, performed after total to investigate composite units of tissue, supplied by
body injections in fresh cadavers. a single vascular system. Units of skin and tendon
(ii) A correlation with our previous detailed (Taylor and Townsend, 1979), muscle with nerve
studies of isolated areas of the body. (Taylor 1978) and skin, muscle and bone (Taylor
THE VASCULARTERRITORIES L~NOI~~~MES)OFTHE BODY 115

and Watson, 1978; Taylor et al., 1979a and b) were neous perforators. Initially we used barium sulphate
analysed. It was jiom this work that the angiosome but often the results were unsuitable for publication.
concept germinated. Various regions including the Since 1983 our radiography has been greatly
anterior abdominal wall (Boyd et al., 1984; Taylor improved by using a modification of Salmon’s
et a/., 1984), the anterior thorax (Reid and Taylor. injection mixture (Rees and Taylor, 1986). Both
1984; Palmer and Taylor, 1986), the lower limb and regional and total body studies have produced
the upper limb were then studied. The results added excellent results.
strength to the angiosome concept of the blood Our results in general agree with those of Salmon
supply and revealed the interconnections that exist and Manchot. However, our method differed in
at all levels between adjacent vascular territories, a many respects and our investigations were more
relationship which is evident throughout the body. extensive.
Our methods evolved in three stages. Initially
dissection was used to define vascular anatomy. It
also showed the considerable variability that exists Materials and methods
between individuals and between opposite sides in
This study involved dye injections in 50 fresh
the same individual. It revealed a reciprocity
cadavers, and radiographic investigations in 6
between adjacent arterial territories. When the
whole bodies, 2 amputated upper limbs and 4
vessels in one area were small those in the adjacent
area were large. This has been referred to by amputated lower limbs after each had been injected
in the fresh state with a mixture containing lead
Salmon (1936a) as the “law of equilibrium”.
oxide. Particular attention was paid to those regions
Next, techniques of dye injection were employed.
not previously studied.
The:y were validated in the experimental pig and
then used extensively in fresh cadavers (Taylor and
Dye injection studies
Corlett, 1981). They defined the cutaneous terri-
torie:s and the supply to deeper tissues. Macroscopic An average of four vascular territories were
staining of the tissues and microscopic extent of examined in each cadaver with a range of 1 to 30.
tissue perfusion were assessed. When the vessel of Usually regions were investigated by injecting the
a known flap was injected in the intact body, the source artery (the main distributing artery) of
adjacent territories alternately with red, black or
perimeter of cutaneous staining never extended
purple ink (Fig. 1). The skin staining was recorded
beyond the expected area of survival of the potential
photographically at various stages of the injection
flap. This is explained as follows by “the overflow
and then the deeper tissues were dissected, noting
phenomenon” (Taylor and Corlett, 1981; Taylor,
the course of the vessels and the staining of the
1983).
underlying muscle and bone.
When large volumes of ink were used, the dye
would appear often from the cannula inserted into
Radiographic and perforator mapping studies
an adjacent artery or be seen to escape into the
This was a time consuming exercise and, unlike
major arteries and appear at a remote site such as
Manchot, it took us 2 years to complete. It evolved
the viscera. The reason for this phenomenon is
in four stages : preliminary investigations, whole
revealed in our radiographic studies which illustrate
body studies, examination of isolated limbs and
the reduced calibre choke arterial connections that
cross section assays.
link adjacent vascular territories (Taylor, 1983).
Beca.use there is no pressure buffer effect (McGregor 1. Preliminary investigations. Because of the limited
and Morgan. 1973) from adjacent arteries in the availability of fresh cadavers and the length of each
cadaver, the dye simply follows the line of least study, three important problems had to be solved.
resistance. It is distributed via the branches of the The first was to find a suitable preservative which
artery under investigation until it reaches these could be combined with our lead oxide-gelatin
choke connections. Then it overflows into the mixture and would retain the normal colour and
branches of the adjacent vessels and escapes in texture of the tissues. Various substances were
retrograde fashion via their main trunk. rather than tried, including Salmon’s original mixture, and
perfusing a successive series of linked vascular ultimately chlorocresol was selected. The details of
territories. this study are to be published separately.
Finally, radiography was used to define the exact Next a large radiolucent platform was required
course, distribution and connections of the cuta- which would support the weight of the body and
 BRITISH JOURNAL OF PLASTIC SURGERY

Fig. I
Figure l-Fresh cadaver ink injection studies of various source arteries: (A) Internal thoracic and deep inferior epigastric systems.
(B) The same territories on the opposite side injected with different colours plus the deep circumflex iliacs. (C) Two intercostal
vessels injected. (D) Muitiple territories stained (compare with Fig. 16). (E) Right suprascapular, thpraco-dorsal and profunda
territories. (F) Same subject with left posterior circum~ex humeral and circumflex scapular vessels injected.

beneath which X-ray cassettes could be passed
without moving the subject. This was necessary in
order to produce an accurate montage of the area
under investigation. Various materials were tested
and 5 mm thick perspex was selected. A supporting
platform was constructed by the authors with
longitudinal parallel rails to guide the X-ray
cassettes accurately beneath a grid system marked
on the perspex (Fig. 2A). Transverse removable
rods were included to interrupt the travel of the
cassettes at regular intervals. The grid system was
designed so that the radiographs overlapped by
5 cm around their perimeter. The distance between
the tube of the X-ray machine and the specimen
was determined to give maximum definition and
minimum magnification. It was fixed at 57 inches.
Most radiographs of the integument were exposed
at 55 to 60 KV, 100 mA for 0.3 seconds.
Finally, when planning our incisions, it was
important that the skin and attached subcutaneous
THE VASCULAR TERRITORIES (ANGIOSOMES) OF THE BODY

Fig. 2
Figure l--(A) Authors‘ designed piatfor~l with sliding perspex grid to allow radiography without disturbing the subject. Note
parallel rail systems, x-ray cassette and transverse removable rods. (B) Model used to plan one of the methods of skin removal.

tissue could be laid fiat without overlap and that as we wanted to preserve as many of them as possible.
wide an expanse of integument as possible could be We experimented first by removing simulated skin
examined. Previous workers, including Salmon, from plastic models coated with a latex rubber
had made topographical boundary incisions to solution (Fig. 2B). Each pattern of incisions was
remove areas of skin, particularly in the lines of the tested on such a model before the fresh cadaver was
groins, axillae, neck and limb joints. These junc- dissected.
tional regions are of great clinical importance and 2. Total hod~~~t~~~~~.Three adult males and three
118 BRITISH JOURNAL OF PLASTIC SURGERY

females were examined. Those who were very

elderly, had extensive cardiovascular or metastatic
disease, or who had more than 2 surgical scars,
were rejected. Each subject took approximately 3
months to examine. An average of 109 radiographs
were taken in each case and an average of 374
perforators were mapped and traced to their
underlying source vessels. In every study the body
was injected within 24 hours of death and cooled.
The important body landmarks, previous scars
and our proposed incisions were marked with
indelible inks of different colours and photographed
(Fig. 3). The subjects were chosen and the incisions
were planned: to provide a comparison between
sexes; to obtain anterior, posterior, lateral, “bird’s-
eye” and “worm’s-eye” views of the blood supply
to the integument (skin and subcutaneous fat); and
finally, in the last subject, cross-sections of the body
were made to retain continuity of the vessels
between the deep tissues and the skin.
The body was injected via a T-tube placed in one
femoral artery sited at the apex of the femoral
triangle. This left undisturbed the cartwheel of
cutaneous vessels that radiate from this artery in
the groin. The substance was injected in a pulsatile
manner to avoid small vessel rupture (Reid and
Taylor, 1984).
At first 3 to 4 litres of 0.3 per cent chlorocresol
preservative was injected until the superficial veins
became distended. The venous sytem was then
decompressed via cannulae inserted into the exter-
nal jugular veins. Next the lead oxide gelatin
mixture was introduced until its orange colour was
noted in the conjunctival vessels, as mottled patches
in the skin, and within small incisions made in the
pulps of the toes and finger tips. This took
approximately 1 hour. The most difficult areas to
perfuse were the pressure points in contact with the Fig. 3
dissection table. In the first 3 cadavers we massaged
Figure 3-Cadaver with body landmarks and incision lines
these areas during injection but the perfusion was marked.
not complete. Subsequently we adopted the tech-
nique described by Last and Tompsett (1962) and
floated the bodies in a bath. Finally 10% formalin to its outer layer on the other. It must be remembered
was introduced into the cranial, thoracic and that the palmar and plantar aponeuroses are part
abdominal cavities since chlorocresol is relatively of the deep fascia.
ineffective in preserving the viscera. The cutaneous perforators were measured as they
Dissection was deferred until the next day to pierced the deep fascia. In keeping with our
allow the orange radio-opaque mixture to set to a previous studies (Boyd et al., 1984; Palmer and
firm rubbery consistency. The integument was Taylor, 1986) those vessels of 0.5 mm or greater
removed usually as 2 halves in order to preserve in were tagged with lead beads and divided to leave
continuity as wide an area of cutaneous vasculature the lead markers on the surface of the deep fascia.
as possible. In the limbs the integument was raised The perforators were not tagged on the undersurface
superficial to the deep fascia on one side and deep of the integument as their stumps were easily
THE VA~C~.JI_ARTERRITORIES(ANGI~S~MES) OF THEBODY

Fig.4
Figure I---Perforator mapping pins on dorsal aspect of subject.

identified on X-ray. When a long vessel was noted possible and always within S days. While the
it was sutured in the correct orientation. The integument was being studied the body was placed
integument was spread on the perspex platform. in a deep freeze or an alcohol bath. Additional
The cutaneous “carpet” was X-rayed and a montage radiographic studies were made of the pancreas,
of its arterial circulation was constructed using liver, spleen, renal tract, gall bladder with bile
contact photographic prints of the carefully labelled ducts, and representative samplesof bowel, individ-
radiographs (Figs 5,6,7 and 8). ual muscles and nerves. This was done to compare
Next the deep tissues were X-rayed and dissected, the arterial framework of these tissues with each
tracing the tagged cutaneous perforators to their other and with the skin.
underlying source vessels (Fig. 9). The lead beads
3. Isolated limb studies. Four upper limbs and two
were removed and replaced with coloured mapping
lower limbs, amputated for trauma and tumour in
pins, assigning a different colour to the perforators
young individuals, also became available for
of each source artery (Fig. 4). The results were
analysis in the fresh state. They were injected and
photographed and recorded on a master diagram
analysed as described. Because of the age of the
of the body in each case (Fig. 15). The coloured
patients there was no evidence of vascular disease
pins were then matched to the perforators in the
and excellent studies were obtained, especially of
radiographic studies of the integument. The vascu-
the hands and the feet (Fig. 8).
lar territory of the perforator, or cluster of perfora-
tors, arising from each source artery was charted. 4. Cross-section studies. Our radiographs of the
This was done by drawing a line around the integument provided us with a two-dimensional
perimeter of the territories, across the choke vessels “plan” picture of the cutaneous vascular network.
which linked them together (Figs 11, 12, 14 and To define the course of vessels between the deep
16). fascia and the skin, strips were taken from the
Because of the lengthy nature of the investiga- integument of each subject along the axes of
tions, certain steps were taken to deiay tissue decay. selected cutaneous arteries. Particular attention
The body cavities were eviscerated as early as was paid to those areas where the skin was mobile
 120 BRITISH JOURNAL OF PLASTIC SURGERY

Fig. 5
Figure 5--Gne of the montages of the cutaneous circulation revealing the continuous network of arteries and arterioles, their origin,
size, orientation and density in different regions of the body. Several pressure areas on the back did not perfuse in this subject. Note
the large perforators in the head, neck, torso and proximal limbs compared with the small closely packed perforators in the forearms
and legs. Note also the long axial vessels radiating from the groins and popliteal fossae.
 THE VASCULAR TERRITORIES (ANGIOSOMES) OF THE BODY 121

Fig. 6
Figure &Lateral view of one female subject (A) and anterior view of another (B). The arm has been removed in study (A). Note
the nebvork of large vessels which sweep laterally from the ventral and dorsal midlines, ascend from the groins, descend from the
shoulder girdle and converge on the summits of the scalp and the breasts. This demonstrates the principle that vessels radiate from
fixed concave zones and radiate to mobile convex areas, A lower midline scar interrupts the vessels in subject(B) with compensatory
opening of a large choke vessel above the umbilicus (arrow) to re-establish the flow across the midline.
 122 BRITISH JOURNAL OF PLASTIC SURGERY

Fig. 7
Figure ?--“Birds-eye” view of scalp and shoulder epaulette. Note rich choke interconne~ti(~ns between vessels which converge on
the convexity of the scalp (A) and radiate from the concavity at the root of the neck (B). Arrows indicate the elbow joints. and the
clavicle is outlined. The platysma muscle was not included on the right side of the subject which explains the poverty of perforators
in this area.
A

D
Fig. 8
Figure X--Radiographs of the integument of the hands removed with the palmar aponeurosis (A) and without it (B). The blood
supply around the right knee. the distal legs and the feet (C and DJ are shown. The position of the patella is represented (shaded
area). NOZP The high density of perforators in the palm and sole skin. the convergence of vessels over the loose-skinned areas of the
extensor surface of the knee and interphalangeal joints of the fingers. and longitudinal orientation of large vessels which accompanv
the sensory nerves (arrows).
 Fig. 9
Figure 9-Radiographs of deep tissues with lead beads locating the dominant cutaneous perforators. Note: (A) The perpendicular
supply to the quadriceps of the thigh versus the oblique supply to the hamstrings. This is due to the different range of movement
between these muscle groups. (B) The continuous arch formed between the anterior and posterior tibia1 arteries. (C) The choke
vessels above the umbilicus which connect the internal thoracic-superior epigastric arteries with the deep inferior systems to
establish the ventral vascular railroad between the clavicles and the groins. Branches of these systems radiate laterally to connect-in
approximately the mid-clavicular line from above down with the acromiothoracic, lateral thoracic, posterior intercostals, lumbar
and ascending branches of the deep circumflex iliac arteries. The territory of one DIEA is defined.
 n-

Fig. 1U
Figure IO--- Sectional strip radiographic studies from above down of the breast (A), thigh (B), sole of foot (C) and buttock (D). The fatter includes the ynderlying gluteus
maximus muscle. Schematic diagram illustrates the dominant horizontal axis of vessels which provides the primary supply to the skm m each case and tts relationshtp to
the deep fascia. In Type A they predominate in the subdermal plexus. Note from left to right the internal thoracic perforator and lateral thoracic artery converging on
nipple (arrow) rn the radiograph of this loose skin region of the torso. fn Type B they are seen coursing on the surface of the deep fascia m this relatively tixed skin area. In
Type C the source artery itself is the dominant horizontal vessel supplying the skin. coursing beneath the deep fascia in this rigidly fixed skin region. In Type D the
horizontal vessel is again the source artery (inferior glutealf but this time its branches have to pierce muscle directly to reach this tixed skin region. Smalt arrows define the
deep fascia and the large arrow indicates the large fasciocutaneous branch of the &teal artery which descends with the posterior cutaneous nerve of the thigh. E
126 BRITISH JOURNAL OF PLASTIC SURGERY

Fig. 11

Figure 1 l-Sectional studies selected from the abdomen, thorax and thigh revealing the course of the cutaneous and muscular
branches derived from the source arteries in the deep tissues, Note: (top left) The section is taken obliquely along the intercostal-
femoral artery axis showing 2 angiosomes linked by choke vessels at corresponding levels in the integument and deep tissues
(arrows). The cutaneous (i) and muscular branches of the posterior intercostal (I) angiosome connect with the superficial inferior
epigastric (e) and ascending branch of the deep circumflex iliac artery (D). The source artery on the right is the external iliac-
femoral artery trunk. (Bottom left) The internal thoracic artery (arrow) with its anterior intercostal, muscular and cutaneous
branches which sweep laterally to meet the posterior intercostal artery and its lateral cutaneous perforator (arrow). Posteriorly
branches of the posterior intercostal artery anastomose with the circumflex scapular artery at this level (arrows). (Top and centre
right) The dominant cutaneous perforators which arise from the superficial femoral (F), profunda (P) and descending branch of the
lateral circumflex femoral (L) arteries. They reach the skin via the intermuscular septa. (Bottom right) The angiosomes of these 3
source arteries which supply skin, muscle and bone (compare with Fig. 14).
THE VASCULAR TERRITORIES (AN~IOSOMES) OF THE BODY

Fig. 12
Figure I ?---Radiographs of selected regions of the integument and muscles. More: (A) Dotted line drawn around perimeter of choke
connecting vessels of a large acromiothoracic perforator to define its anatomical territory {compare with Figure 5 front view, left
side of body). (B) Chain-linked system of perforators anatomising without change in calibre (arrows). They accompany lateral
cutaneous nerve of arm (compare also with Figure 5 front view, right arm). (C) Trapezius muscle with transverse cervical (above)
and posterior intercostal (below) territories defined. (Compare with their angiosomes in Figure 16.) (D) Latissimus dorsi muscle
with thoracodorsal (above), posterior intercostal (middle) and lumbar territories (below) outlined which again correspond to their
respec%ive angiosomes.
 128 BRITISH JOURNAL OF PLASTIC!SURGERY

Fig. 13
Figure 134chematic diagram of a source artery with direct and indirect cutaneous perforators. The direct vessel (D) courses here
in the intermuscular septum to reach the integument, supplying muscle and other deep tissues en route. The smaller indirect
perforators (i) arise as terminal twigs of large arteries to the muscles. Both systems communicate within the connective tissue
framework of the muscles, on the surface of the deep fascia (arrow) and within the integument to form a continuous three-
dimensional network of vessels.

over the deep fascia and to where it was fixed. The
strips were cooled until rigid, placed on their side
and X-rayed to give an elevation view of the vessels
(Fig. 10).
Finally, in one body composite blocks of integu-
ment and underlying deep tissue were sectioned
from the torso, the buttocks and from the limbs at
5 cm intervals. This was done to define the course
of the cutaneous perforators from their source
arteries as well as the supply to the deeper tissues.
The specimens were again X-rayed (Fig. 11).
Results
Our ink injection, dissection, perforator mapping
and radiographic studies complemented each other.
The results were correlated and then compared
with our previous anatomical studies in fresh
cadavers. This provided an overview of the blood
supply of the entire body. Some results verify the
observations of other workers, others vary from
published material, some confirm our previous
concepts, others offer new ideas. Therefore only the
important anatomical observations which have
particular surgical significance will be presented.
Injection studies
These outlined the cutaneous territories of the
source arteries (compare Figs 1 and 16) as well as
their supply to the deep tissues. This information
was coupled with our dissection and radiographic
studies in order to define the angiosomes of the
body (vide infra). The ink also outlined the course
of the arteries during dissection. This was of
particular benefit when tracing the course of the
vessels in the integument, as the dye penetrated
further than the lead oxide-gelatin mixture and
reached the capillary bed.
Dissection and perforator mapping
The bright orange colour of the lead oxide-gelatin
mixture facilitated dissection. Perhaps the most
striking feature was the relationship ofthe vessels to
the connective tissue framework of the body. They
followed it closely as if for support and protection.
Wherever the connective tissue was rigid they
hugged its surface. Where the connective tissue was
loose they travelled within it.
No matter which part of the body was studied,
certain principles became clear. Few vessels crossed
mobile tissue planes. They crossed where those
planes were fused. The vessels radiated from fixed
to mobile areas. Where mobility existed between
tissues the vessels coursed parallel to the surface for
long distances, usually separated from the plane of
mobility by a sheet of connective tissue of variable
thickness. Sometimes the vessels travelled in a
fibrous sheath, but always this tunnel contained
loose areolar tissue. Gray’s Anatomy (1980) states
the reason for this-it is to allow the veins to dilate
and the arteries to pulsate.
The pattern of supply is well illustrated if the
THE VASCULAR TERRITORIES (ANGIOSOMES) OF THE BODY 129

-._ ._*
/

Fig. 14

Figure 14.- Schematic diagrams showing

how we detined the angiosom~s. IA) The cutaneous perforators with their choke
connectionsare depicted on the left. The origin of these perforators from their underlying source arteries and their muscular
branches is shown on the right. (Bj The vascular territories of each source artery are illustrated in the integument (left) and deep
tissues fright) by lines drawn through the choke connecting vessels. Note that the territories correspond in these 2 layers and ho%
they appear as sectors in the limbs.

arterial network is traced from the heart to the
periphery. The major arteries are closely related to
the blones of the axial skeleton, a fact which is
cIearl:y illustrated by the corosion cast studies of
Last and Tompsett (Fig. ISA). Their branches at
first radiate from fixed locations to reach the
viscera, or follow the intermuscular septa to supply
the musculo-skeletal tissues and the integument.
In the deep tissues the branches penetrate the
muscles, usually on their deep surface, the tendons,
the bones, the nerves and the deep fat deposits, As
these vessels divide and subdivide within the
specialised tissues, their branches follow the con-
nective tissue framework to reflect the architecture
of the tissue in question,
Classification of the cutaneous blood supply
The course of the cutaneous perforators depends on
the proximity of the source artery to the undersur-

Figure 1S-The dominant cutaneous perforators emerging from the deep fascia with different colours identifying their corresponding
source arteries in the deep lavers (compare with Figure 16). Note their relationship to the intermuscular and the intramuscular
septa. This result was aveiageh from all-our studies. -
face of the deep fascia. In each case they follow the
inte~us~ular and the intramuscular septa, supply-
ing branches to each tissue as they pass by. They
fall broadly into two groups: (i) direct and
(ii) indirect (Fig. 13). This classification was de-
scribed originally by Spalteholtz (1893) and has
been reviewed recently by Timmons (1985).
Direct vessels
The direct vessels constitute the primary cutaneous
supply. Whether they follow the intermuscular
septa, or pierce muscles en route, their main
destination is the skin.
I~d~reet vessels
The indirect vessels constitute the secondary cuta-
neous supply. They emerge from the deep fascia as
terminal spent branches of arteries whose main
BRITISH JOURNAL OF PLASTIC SURGERY
purpose is to supply the muscles and other deep
tissues. They reinforce and interconnect with the
primary supply to the skin. They are sparse where
the muscle pistons beneath the deep fascia and are
plentiful where the muscle is fixed to that structure.
The direct cutaneous vessels arise:
(i) From the source artery as it courses just
beneath the deep fascia-for example the
superior ulnar collateral artery; the superficial
inferior epigastric artery; and the cutaneous
perforators of the transverse cervical, radial
and plantar arteries.
(ii) As a direct continuation of the source artery-
for example the cutaneous branches of the
internai and external carotid vessels, such as
the superficial temporal and the supratrochlear
arteries.
 3 (Anti :OSOMES) OF THE BODY

Figure L&-The angiosomes of the source arteries of the body coloured to correspond with Figure 14. They are: (1) Thyroid.
(2) Facial, (3) Buccal (internal maxillary), (4) Ophthalmic, (5) Superficial temporal, (6) Occipital, (7) Deep cervical. (8) Transverse
cervical. (9) Acromiothoracic, (10) Suprascapular, (I I) Posterior circumflex humeral. (12) Circumflex scapular. (13) Profunda
brachii, t 14) Brachial. (15) Ulnar, (16) Radial, (17) Posterior intercostals, (18) Lumbar, (19)Superior gluteal. (20) inferior gluteal.
(21) Profunda femoris. (22) Popliteal, (22a) Descending geniculate (saphenous), (23) Surai, (24) Peroneal. (25) Lateral plantar,
(26) Anterior tibial, (27) Lateral femoral circumflex, (28) Adductor (profunda), (29) Medial plantar. (30) Posterior tibia],
(31) Superficial femoral, (32) Common femoral, (33) Deep circumflex iliac, (34) Deep inferior epigastric, (35) Internal thorncic.
(36) Lareral thoracic, (37) Thoraco-dorsaf, (38) Posterior interosseous, (39) Anterior interosseous, (40) Internal pudendal

{iii) From a deeply situated source artery, or one of
its branches to a muscle. They follow the
intermuscular septa to the surface-for exam-
ple the cutaneous branch or branches of the
circumflex scapular and lateral circumflex
femoral vessels.
(iv) From the source artery as it courses on the
undersurface of a muscle. A single large vessel
olr a number of vessels are given off and each
pierces the muscle vertically or obliquely to
reach and penetrate the deep fascia. For
example the cutaneous perforators of the
internal thoracic, intercostals, deep inferior
epigastric and the gluteal vessels.
The direct cutaneous perforators pierce the outer
layer of the deep fascia near where it is anchored to
bone or to the intermuscular and large intramuscu-
lar septa (Figs 4, 5 and 15). These lines and zones
of fixation also correspond to the fixed skin areas of
the body. They are easily seen in a thin, well
muscled individual as grooves and valleys around
the perimeter of muscles-especially where they
interdigitate, over well developed intermuscular
and intramuscular septa, over the flexor surface of
joints, adjacent to the dorsal and ventral midlines
of the body, around the base of the skull and in the
region of some bony prominences.
From these grooves and valleys in the deep fascia
132
E branch in all directions in stellate form or course in
Fig. 17
Figure 17-Diagram showing how the size and course of the
direct cutaneous perforators X and Y. which emerge from fixed
points in the deep fascia. couId be rn~~~ed by growth either
before or after birth. In (A) the perforators, which are fixed in
number and position, form a major connecting network on the
surface of the deep fascia in the “resting state”. In (B) they are
stretched with the deep fascia by the expansion of underlying
tissues-for example the scalp vessels as the brain and skull
expand during foetal development. In (C)as the breast develops
~ifhj~ the integument the vessels are displaced towards the
dermis and lengthened as they converge on the nipple. In (D)
they are stretched apart in the limbs as the long bones graw, but
stili retain their original relationship to the deep fascia. In (E)
the vessels are again stretched apart by growth but the mobile
relationship between the under surface of the integument and
the deep fascia is responsible for their oblique course. This is
characteristic of the loose skin areas of the torso.
the vessels flow towards the convexities of the body
surface, branching within the integument. The
wider the distance between the concavities and the
higher the summit, the longer the vessel. For
example: the blood supply of the breast, nose, ears,
genitalia, the extensor surface of the joints and over
the surface of bulging muscles. This pattern of
supply is clearly shown in our radiographic studies
(Figs 5 to 11).
BRITISH JOURNAL OF PLASTIC SURGERY
The size and density of the direct perforators
vary in different regions. In the head, neck, torso,
arm and thigh the vessels are larger, longer and less
numerous. In the forearm, leg, dorsum of the hands
and feet, the vessels are generally smaller, shorter
and more numerous. In the palms of the hands and
the soles of the feet, perhaps the most fixed skin
areas of the body, the density of perforators is at a
maximum. Thus the primary supply to each
cutaneous territory varies between source arteries.
It ranges from a small number of large direct
perforators to a large number of smaller direct
vessels (Figs 5, 6, 7 and 8). Each territory is
reinforced by indirect perforators.
The course of the cutaneous vessels between the
deep fascia and the skin also varies in different
regions. Nevertheless they follow the connective
tissue framework of the superficial fascia, intercon-
necting at al1 levels. They ramify on the undersur-
face of the subcutaneous fat adjacent to the deep
fascia and then branch and course towards the
subdermal plexus, worming their way between the
fat locules. The smaller vessels in general course
vertically towards the skin. The larger vessels either
a particular axis, branching as they pass parallel to
the skin surface.
In the scalp and in the extremities where the skin
is relatively or rigidly fixed to the deep fascia the
larger vessels hug that surface. They course on the
deep fascia for a considerable distance in the loose
areofar Iayer that separates it from the subcutaneous
fat (Figs IO and 11). This is evident particularly
where the perforator accompanies a cutaneous
nerve. Frequently several vessels interconnect in
chain link fashion to follow a nerve, anastomosing
without change in calibre (Figs 8 and 12).
In the loose skin areas of the body, especially
over the pectoralis major muscle, the iliac fossa and
the extensor surface of joints, the direct cutaneous
vessels course for a variable distance parallel to the
deep fascia. However, they are more intimately
related to the undersurface of the su~utaneous fat
and are plastered to it by a thin glistening sheet of
fascia which separates them on their deep surface
from a plexus of smaller vessels. This plexus of
smaller vessels lies in loose areolar tissue on the
surface of the deep fascia. It is formed by branches
which arise from the direct perforators as they
pierce the deep fascia, and the connections these
branches make with smaller indirect perforators
(Fig 17). The large direct perforators then pierce
the subcutaneous layer. They ascend obliquely
THE VASCULAR TERRITORIES (ANGIOSOMES) OF THE BODY

Fig. 18
Figure I8-_( A) Tompsett‘s corrosion cast of the arteries in a stillborn baby. Note how the major vessels relate to the bony skeleton
of the body. Reproduced by kind permission from the Department of Anatomy, Royal College of Surgeons of England. (B) The
wing of a moth with its superstructure of “veins” which interconnect and nourish the part. Note the resemblance to the mesenteric
vessels in the human gut

within the superficial fascia to reach the rich
subdermal plexus, where they travel for considera-
ble distances (Figs 10 and 11).
Radiographic studies
These investigations provide the basis for a three-
dimensional arterial atlas of the body (Figs 5 to 12).
They confirm the origin and course of the direct
cutaneous perforators (vide supra) as well as
demonstrating their density in different areas of the
body. The most striking observation, however, is
that adjacent source arteries and their branches are
linked to form a continuous network of t)essels
regardless of the tissue. Some connections are true
anastomoses without change in calibre, for example
between the dorsalis pedis and the posterior tibia1
artery, between some posterior and anterior inter-
costal arteries (Fig. 9) and to a lesser extent in the
muscles and in the integument. More commonly
the connections are formed by reduced calibre
vessels-the choke arteries and arterioles (Fig. 12).
As the calibre of the arteries decreases the number
of connections increases.
The blood supply to the integument is revealed
134
as a continuous three-dimensional mesh of vessels,
contributed to at fixed skin sites by the entry of
cutaneous perforators. It spans the entire body like
a national map of roadways and subways made up
of interconnecting channels of all sizes. It is
subdivided into anatomical vascular territories by
the perimeter of choke arteries which connect the
perforators of a source artery with those of its
neighbour (Figs 12 and 14).
In the deep tissues a similar pattern exists. The
branches of the source arteries are linked within
the various tissues, usually by choke vessels. In the
muscles they often exhibit a characteristic cork-
screw appearance (Figs 9, 11 and 14).
The angiosome concept
We compared the territory supplied by a source
artery in the deep tissues with its territory in the
overlying integument and found that their per-
imeters often coincided (Figs 11 and 14). This
composite unit of skin and underlying deep tissue,
supplied by a source artery, we have named an
ANGIOSOME*. The angiosomes of the body were
then plotted, based upon the combined results of
our ink injection, dissection, perforator mapping
and radiographic studies (Figs 14, 15 and 16).
Though the boundaries of these anatomical
territories are shown as lines, they represent the
middle of choke zones. They outline the vascular
territories of the integument and the deep structures,
e.g. muscles, nerves and peritoneum. It is clear
from our studies that some muscles will be wholly
contained within one angiosome, e.g. thenar; many
span at least two angiosomes, e.g. rectus abdominis
and gluteus maximus; but the majority traverse
more than two, e.g. the latissimus dorsi, pectoralis
major and sartorius muscles.
There is a close correlation between the angio-
somes and the neurological dermatomes on the
torso and in the head. In the extremities this is less
evident where the angiosomes appear as sectors
(Figs 11 and 14). Sometimes the angiosomes concept
is clouded by nomenclature. For example :
(i) The territory of the superficial temporal artery.
This fan-shaped territory overlies a similar
territory supplied by the deep temporal vessels.
This in turn overlies a comparable territory in
the dura supplied by the middle meningeal
*From the Greek angio- meaning vessel, and somite meaning
segment or sector of the body (derived from soma, body)
BRITISH JOURNAL OF PLASTIC SURGERY
artery. The latter 2 groups of arteries are
derived from the internal maxillary artery but
all 3 are derived directly or indirectly from the
terminal bifurcation of the external carotid
system. In other words, the angiosome is the
composite block of integument, muscle, peri-
cranium, bone and dura supplied by the distal
part of the external carotid artery-the terminal
external carotid artery angiosome.
(ii) The territory of the superficial inferior epigas-
tric artery (SIEA). The common femoral angio-
some (Figure 16) can be subdivided into smaller
angiosomes. The ascending branch of the deep
circumflex iliac artery (DCIA) is the deep tissue
counterpart of the SIEA (Fig 11 top left). The
continuation of the DCIA is the counterpart of
the superficial circumflex iliac artery (SCIA).
Unfortunately the deep vessels arise from the
terminal part of the external iliac artery and
the cutaneous perforators arise from the begin-
ning of the common femoral artery. Terminol-
ogy again obscures the angiosome concept.
Other important observations from our radiog-
raphy studies should be noted. The first is the dense
network of vessels in the scalp, face, axillae, breasts,
paraumbilical region, perineum, palms of the
hand and soles of the feet. Many of these corre-
spond to the areas that sweat most profusely
(Figs 5 to 8).
The second observation is that there is a dominant
axis of vessels coursing parallel to the body surface
which supplies direct vessels to the skin. The depth
of this arterial axis varies in different regions and is
fundamental to skin flap design (Figs 10 and 11). It
may be seen :
(9 In the subdermal plexus. This is characteristic
of the loose skinned areas (see Figs 10A and
11 top left).
(ii) On the surface of the deep fascia. This is
typical of the blood supply to the scalp, the
thighs and the arms, where the skin is fixed to
the deep fascia (Fig. 10B).
(iii) Beneath the deep fascia. This is apparent
where the source artery itself is the dominant
vessel which courses parallel to the skin surface
and gives off a series of perforators. It may
course just beneath the deep fascia, e.g. medial
plantar artery (Fig. lOC), beneath or within
an intermuscular septum (for example the
radial and ulnar arteries), or beneath a muscle
which intervenes between it and the deep
fascia (Fig. 1OD).
THE VASCULAR TERRITORIES (AN~IDSOMES~ OF THE BODY 135

tissues of the adjacent angiosome can be captured

We did not inject the venous system in this study. with safety. Necrosis tends to occur in the next or
However in each instance during dissection venae subsequent angiosome. This presumably is due to
commitantes were noted accompanying the cuta- the pressure gradient that must occur as blood flows
neous arteries. A further investigation is under way across the choke vessels that link adjacent arterial
to complete the picture, prompted by the work of territories. We have noted in our pig experiments
Costa et al. (1987). that it is these choke connecting arteries and
arterioles which dilate when a skin flap is delayed.
In the developing embryo the vascular system is
Discussion
the first tissue to differentiate in the mesoderm. It
Carl Manchot (1889) defined 40 cutaneous terri- commences as a syncytium of interconnecting
tories in his study of the blood supply to the skin. vascular channels. Although modified by the differ-
He did this with remarkable accuracy considering entiation, growth and migration of the specialised
that he did not have the advantage of X-rays. He tissues, this network arrangement of vessels is
excluded the head, neck, hands and feet in his retained into adulthood. Some vessels obviously
chart. Salmon (1936a) defined more than 80 skin have disappeared and others have enlarged.
territories which covered the entire body. Both GraqlS Anatomy (1958) states that the connective
gave very accurate descriptions of the origin and tissue is what is left of the mesoderm, after the
course of the cutaneous vessels and Salmon revealed specialised tissues have developed from it. It is
the interconnections that exist between ‘adjacent interesting that the connective tissue framework
cutaneous arteries. Salmon (1936b) also went a step and the vascular system co-exist as an interconnect-
further and in a separate study examined the ing network in the adult. If the specialised tissues
individual muscles. He showed that each muscle is such as muscle, nerve, fat and bone have developed
supplied by a variable number of source arteries within this mesh of connective tissue and vessels,
which interconnect within it. then this would explain why vessels have become
However. neither of these workers illustrated the captured within, and compressed between, the
course of the arteries between the deep tissues and specialised tissues. It would explain why they follow
the skin, they did not define the vascular territories the intermuscular septa, the intramuscular septa,
of the source arteries in the deep tissues, nor did the deep fascia and the connective tissue scpta of
they correlate these territories with their counter- the superficial fascia to reach the skin. As mobility
parts in the overlying integument. Salmon hinted develops between tissue planes then presumably
that a relationship existed between these layers, but some of the interconnections between vessels would
neither he nor Manchot could have envisaged the have been lost, with a compensatory enlargement
anatomical demands of modern flap surgery with of vessels at fixed tissue sites. However, this requires
the need to correlate the blood supply between the further research since most of the studies of human
integument and the deep tissues. development have been inferred from data obtained
Our studies have gone this stage further. The from the developing chick embryo.
angiosomes are the composite blocks of tissue Another concept which was suggested by John
supplied by named source arteries which span Hunter two centuries ago (1794) is that at some
between the skin and the bone. Crock (1967) has stage of foetai development, and certainly at birth,
demonstrated the connections that occur in the the body has a fixed number of arteries. This has
bones. Each tissue in an angiosome is supplied by been our impression when comparing the number
branches of the source artery. We have described of cutaneous perforators seen in children and in
40 angiosomes, but many can be subdivided into adults when raising various flaps. If this is so, and
smaller segments; for example the common fem- remembering that thecutaneousperforators emerge
oral, popliteal and posterior intercostal territories. from the deep fascia at predetermined sites. then
Adjacent angiosomes are interconnected in each the orientation of the cutaneous arteries is the
tissue layer. whether it be skin, fat, muscle, nerve product of the differential growth that has occurred
or bone, usually by reduced cafibre choke arteries. in that area from some stage in foetal development
When we consider these anatomical territories in to adulthood. It explains why vessels radiate from
the light of clinical experience it can be seen that in concavities and converge on convexities and why
most cases when a flap is based on the vessels of the vessels in some areas are small and close
one angiosome, then the corresponding tissue or together while in others they are large and spaced
136
well apart (Figs 5,6,7 and 8). There are numerous
examples which support this hypothesis (Fig. 17).
The sternomastoid and trapezius muscles split from
the same somite (Patten, 1968). The trapezius
“drags” its supplying transverse cervical artery
(and nerve) across the root of the neck to the back,
together with a large band of skin which the artery
nourishes. Manchot suggested that the long course
of the superficial superior and inferior epigastric
arteries is the result of extension of the torso of the
foetus. The expansion of the brain stretches the
scalp vessels from the base of the skull towards the
vertex. Similarly the vessels converge on the nipple
from all directions as the breast develops. The
pattern is seen also in the limbs. As the epiphyses
of the long bones grow the arteries are stretched
apart. Notably they radiate from the flexor surface
of joints, an observation well illustrated in our
studies (Figs 5 and 8).
A final concept was documented again by John
Hunter (1794). He noted the anastomoses that
occur between arteries, quoting the vascular arcades
in the hands and feet as examples, and stated that
the arcades occur more frequently as the arteries
become more distal. The arterial arcades in the
bowel mesentery are a classic example. They are
smaller and more numerous as they approach the
intestine. This basic pattern exists in all tissues and
is modified by the function of that tissue. Even
during tissue repair the pattern of vascular arcades
is reproduced in granulation tissue.
Comparison of the vascular architecture in the
human with that of other animals and other species
reveals a similar arrangement of arcades. In the
loose-skinned animals the arcades in the integu-
ment are stretched over long distances (Fig. 19). In
the wings of insects and in the leaves of plants the
“veins” assume a similar pattern of interconnecting
arcades to those of the intestinal mesentery (Fig.
18B).
Clinical applications
It would be presumptuous of us to suggest that the
anatomical information that we have provided is
completely original or a definitive work on the
blood supply to the skin and underlying tissues.
Rather it is an attempt to correlate the efforts of
many workers, to provide some concepts which will
stimulate further research, to aid in the design of
skin flaps and incisions, and to simplify nomencla-
ture. Some of the applications of the anatomical
information will be considered briefly :
BRITISH JOURNAL OF PLASTIC SURGERY
1. The axes of skin jups. Our arterial atlas of the
integument details the origin, course, size, density
and interconnections of the cutaneous perforators.
It therefore provides for the logical planning of the
base and the axis of a flap. It is obvious from our
studies that there are many more axial flaps
available for transfer than have been described to
date. Several of these are included in Figure 20 with
the known flaps.
The cross-sectional studies confirm the reason
for including the outer layer of the deep fascia with
flaps raised in the scalp and in the extremities, for
in these situations the vessels hug the fascia for
considerable distances. If a flap is designed along
the course of a cutaneous nerve, such as the
saphenous or the sural nerves, then very long safe
flaps can and have been elevated. Ponten’s (1981)
original flaps were designed in this way and we
planned our saphenous neurovascular flap in a
similar manner (Acland et al., 198 1).
In the loose-skinned areas of the torso, which can
be gauged by the pinch test, it is unnecessary to
include the deep fascia as the cutaneous arteries
course at an early stage within the integument.
Frequently they correspond with the course of a
cutaneous nerve (Badran et al., 1984). By comparing
our radiographic studies of the vessels in the
integument with the course of the cutaneous nerves
as depicted in Grant’s Atlas of Anatomy (1958), a
remarkable resemblance in orientation of the
cutaneous arteries and nerves is evident.
Finally, if it is retnembered that the large
cutaneous vessels emerge from the deep fascia
predominantly around the perimeter of muscles,
we have found in lean individuals that the Doppler
probe is a useful aid to locate these perforators. If a
line is drawn to join two such perforators and the
skin between is mobile, then in our experience a
safe flap can be planned along this axis. This is of
particular use on the torso and especially on the
back.
2. Skin flap dimensions. Since the blood supply of
the integument has been shown to be a continuous
system of linked vascular territories, the survival
length of a skin flap must depend upon :
(i) The calibre and length of the dominant vessel
upon which the flap-is based
(ii) The calibre and span of the adjacent captured
artery or arteries
(iii) The calibre and length of the connecting choke
vessels
(iv) An adequate venous return
THE VASCULAR TERRITORIES (ANGIOSOMES) OF THE BODY 13:

B
A

Fig. 19
Figure 19-The cutaneous blood supply of the rabbit. Note the large long vessels which course in the loose-skinned area of the torso,
their choke connections, and their origin from the fixed skin areas of the axillae and the groins.

In most instances the skin flap is endowed with tered, for example, with island neurovascular flaps
a rich dual system of superficial and deep veins. raised from a digit.
Problems may arise, however, with an island skin If the arterial perforators are large and widely
flap. The superficial veins are disconnected and the separated, the territory of each is big and a long
venous return may be thrust upon a diminutive flap can be raised with safety. These flaps are
deep system. This problem is sometimes encoun- characteristic of the loose-skinned areas of the torso
138 BRITISH JOURNAL OF PLASTIC SURGERY

Fig. 20
Figure 20-Some of the large axial cutaneous Aaps which have been used, or are available, as defined by our radiographic studies of
the integument. In the scalp and the limbs they should include the deep fascia. (Compare with Figures 5-8 and 14.)

and of the scalp. Conversely if the perforators are
diminutive and close together, the territory of each
is small. The viable length of the flap is short unless
the underlying source vessel is included in the
design. This is very evident in the fixed skin area of
the sole of the foot.
Where very large flaps are required, or where
vessels of a large calibre are necessary for microvas-
cular anastomoses, these requirements can be
satisfied by chasing the perforators via the inter-
muscular septaor the intramuscular septa to include
the underlying source vessels. The intelligent use-of
a delay will also allow safe capture of adjacent
vascular territories. Many studies of these septocu-
taneous vessels have been made and have resulted
in various flap designs (Shanahan and Gingrass,
1979; Song&al., 1982; Carriquicyetal., 1985). The
principles are outlined in Figure 10 which highlights
the dominant axis of vessels which course parallel
to the skin surface in different areas of the body.
However, the concept of septocutaneous vessels is
not original and is illustrated by many of the early
anatomists (Quain and Wilson, 1842; Grant, 1958).
Cormack and Lamberty (1984) have added much
to the knowledge of the anatomy of the cutaneous
perforators and have suggested a classification of
fasciocutaneous flaps.
3. Fasciocutaneousjups. It is interesting how these
flaps have evolved. In many ways we have
historically come full circle. The senior author was
taught by Sir Benjamin Rank, as a dictum from Sir
Harold Gillies, always to take the deep fascia with
a skin flap in the extremities, and especially in the
lower limb. The flap was safer if you did so. The
precise vascular basis, however, was not known.
THE VASCULAR TERRITORIES (ANGIOSO~ES) OF THE BODY
We then laboured through many bulky musculocu-
taneous flaps, some of which have been major
contributions. Now thanks to Ponten we have
returned to rediscover an old friend. However, this
time we know why the flaps are safer in the limbs
when the deep fascia is included.
Nevertheless too much emphasis has been placed
on the intrinsic value of the blood supply of the
deep fascia. Every experienced surgeon knows that
when the deep fascia is exposed it desiccates,
sloughs and granulates with reluctance. Our studies
show that the dominant vessels course adjacent to
the surface of the deep fascia, not within it. With
care they can be dissected free. Inclusion of the
deep fascia with the overlying integument simply
avoids such a tedious dissection. It also preserves
the a.djacent subfascial course of arteries in some
situations.
The term “fasciocutaneous” flap is the subject of
much debate. PontCn included the deep fascia in
his description of the flap. It is a well defined sheet
in the limbs but in certain regions is continued as
the periosteum of the tibia, the ligaments of joints,
the external oblique aponeurosis, the lumbar fascia
and the rectus sheath-to cite just a few examples
where its structure is modified. Therefore we must
be careful when we apply the term fasciocutaneous
to these regions.
We must also remember that there is a superficial
fascia, stuffed with fat pellets, which connects the
dermis to the deep fascia. Perhaps then we should
name flaps raised above the deep fascia as superjicial
fasciocutaneous flaps and those which include the
deep fascia as deep fasciocutaneous$aps. We are
reluctant to suggest new terminology but it would
certa:inly clarify the structure of the many new
“fasciocutaneous” flaps which are appearing in the
literature.
4. Musculocutaneousj?aps. It is apparent from our
studies that in many situations the cutaneous paddle
of a musculocutaneous flap is fed by small indirect
perforators. Where the skin and deep fascia are
firmly bound to the underlying muscle, for example
the gluteus maximus and the latissimus dorsi
muscles, the blood supply to the skin is ensured. At
each fixed site over the muscle vessels emerge, some
large and some small, to supply the integument.
However where the muscle is mobile beneath the
deep fascia, for example the gracilis muscle, the
cutaneous supply is at best tenuous.
In general, musculocutaneous flaps can be raised
safely if the skin paddle is placed over the
139
perforators of the feeding artery or those in the
adjacent muscular territory. Attempts to capture
territories beyond that in either muscle or skin,
without prior delay, frequently result in vascular
insufficiency. This situation may prevail for exam-
ple in the pectoralis major and the lower transverse
rectus abdominis (TRAM) flaps.
5. Composite japs. A knowledge of the supply of
all of the tissues which constitute each angiosome
provides the basis for the transfer of composite
units of skin, muscle, nerve, tendon and bone
supplied by a single arteriovenous system. This
knowledge has been applied extensively in free
composite tissue transfer.
The vessels within the angiosome interconnect
between the various layers. This interconnection is
well illustrated with the transfer of composite tissue
from the groin region. The direct cutaneous
perforators of the SCIA interconnect with the
indirect perforators of the DCIA. When a compos-
ite osteocutaneous flap is based on the deep system,
the perforators of the DCIA capture the territory
of the SCIA to perfuse the skin (Taylor et al, 1979a).
When the superficial system is utilised the reverse
applies to perfuse the anterior segment of iliac crest
and the attached muscles (Taylor and Watson,
1978).
6. Hypervascular and hypovascular planes. These
have been referred to in considerable detail by
Michel Salmon. The fixed skin areas of the body,
which mark the sites of emergence of the dominant
cutaneous perforators, constitute the hypervascular
zones. Where the skin is mobile over the deep
fascia or skull, or where muscles slide freely beneath
the deep fascia, few small vessels cross these gliding
planes. These are the hypovascular planes which
are often used by surgeons to minimise blood loss.
The placement of a breast prosthesis beneath the
mammary gland or the pectoralis major muscle
takes advantage of this fact.
7. The venous drainage ofjlaps. The anatomy of the
deep venous system was noted during our dissec-
tions, but the veins were not injected and studied
radiographically. Costa et al. (1987) have studied
the venous drainage of the TRAM flap. They have
highlighted the fact that there are choke veins as
well as arteries connecting corresponding territories
in the muscle. They have also revealed that the
initial venous return from the skin paddle is against
the valves when the skin paddle is based distally.
This is due to the fact that the venous drainage of
 140
the captured territory is in the opposite direction to
that of the vein in the ilap pedicle. This opens a
new avenue of research.
Conclusion
Our ink infusion and total body injection, dissection
and radiographic studies provide an overview of
the blood supply of the integument and the
underlying tissues. They confirm the works of
Manchot and Salmon and go a stage further to
develop the concept of the angiosome which is
defined as a block of tissue, supplied by a named
artery-a source artery and whose territories in the
integument and the underlying deep tissue corre-
spond.
The blood supply is shown as a reflection of the
connective tissue framework of the body and
consists of a continuous three-dimensional network
of linked vascular territories throughout all tissues.
Acknowledgements
We are grateful to Professor Gordon Clunie and Professor
Graham Ryan for their guidance and for the assistance given by
their respective Depa~ments of Surgery and Anatomy at The
University of Mel~ume; to Mr Douglas McManamny, Mr
Russell Corlett and Dr Gary Crossthwaite for their assistance
with the dissections; to Mr Pierre Smith and Mr Alan May for
their help with the photography; and to Dr Mirek Stranc for
proof reading this paper.
We are indebted to The Victor Hurley Research Foundation,
The Royal Australasian College of Surgeons and The Buckland
Foundation for their grants to fund our research.
References
Acland, R. D., Schustennan, M., Godiia, M., Eder, E., Taylor,
G. 1. and Car&e, I. R. (198 I). The saphenous neurovascular
free fIap. FlmrieandReconsi~ctjve Surgery, 67,763.
Badran, H. A., ~-HeIaiy, M. S. and Safe, I. (1984). The lateral
intercostal neurovascular free flap. Plastic and ~eeonst~cti~
Surgery, 73,l?.
Bakarnjian, V. Y. (1965). A two stage method for pharyngoeso-
phageal reconstruction with a primary pectoral skin flap.
Plastic and Reconstructive Surgery, 36, 173.
Blair, V. P. (1921). The delayed transfer of long pedicle flaps in
plastic surgery. Surgery, Gynecology and Obstetrics, 33,261.
Boyd, J. B., Taylor, G. I. and Corlett, R. J. (1984). The vascular
territories of the superior epigastric and the deep inferior
epigastric systems. Plastic and Reconstructive Surgery, 73, 1.
Carriquiry, C., Aparecida Costa, M. and Vasconez, L. 0. (1985).
An anatomical study of the septocutaneous vessels of the leg.
Plastic and Reeons~tac~iueSurgery, 76,354.
Cormack, G. C. aad Lamhcrty, B. G. H. (1984).A classification
of fascia-cutaneous flaps according to their patterns of
vascularisation. British JoutnulofPlasric Surgery, 37,80.
BRITISH JOURNAL OF PLASTIC SURGERY
Costa, M. A. C., Carriquiry, C., Grottiag, J. C., Herrera, R. H.,
Vaxoaez, L. 0. and Wiadle, B. H. (1987). An anatomical
studyofthevenousdrainageofthe transverserectusabdominis
musculocutaneous flap. Plastic and Reconstructive Surgery, In
Press.
Crock, H. I. (1967). The Blood Supply of the Lower Limb Bones.
Edinburgh: E. and S. Livingston Ltd.
Daniel, R. K. and Taylor, G. I. (1973). Distant transfer of an
island flap by microvascular anastomoses. Plastic and Recon-
structive Surgery, 52, III.
Daniel, R. K. and Williams, H. B. (1973). The free transfer of
skin flaps by microvascular anastomoses. Plastic and Recon-
structive Surgery, 52, 16.
Esser, J. F. S. (1929). Artery Flaps. Antwerp: De Vos-van Kleef.
GiUies, H. D. and Miiard, D. R. (1957). The Principles and Arr of
Piastie Surgery. Boston: Little Brown.
Grant’s Atlas of Anat~y (1958). Ed. Anderson, J. E. 7th
Edition. Baltimore: The Williams and Wilkins Co.
Gray’s Anatomy. (1958). Ed. Johnston, T. B. Davies, I. E. S. and
Davies, F. 32nd Edition. London: Longmans.
Gray’s Anatomy. (1980). Ed. Williams, P. and Warwick, R. 36th
Edition. Edinburgh: Churchill Livingstone.
Hunter, J. (1794). A Treatise on the Blood, Infammation and
Gunshot wounds. London: John Richardson.
Last, R. J. and Tompsctt, D. H. (1962). Corrosion casts of the
blood vessels of stillborn babies. Acta Anatomica, 51, 338.
Manchot, C. (1889). Die Hautarletien des menschlichen Kotpers.
Leipzig : Vogel.
Manchot, C. (1983). The Cutaneous Arteries of the Human Body.
Translated by Ristic, J. and Morain, W. D. New York:
Springer-Verlag.
M&raw, J. B. (1980). The recent history of my~utaneous flaps.
Clinicf in Pfastic Surgery, 7, 3.
McDowell, F. (1979). Logs vs. harpsichords, blobby flaps vs.
finished results. Plastic and Reconstructive Surgery, 64,249.
McGregor, I. A. and Morgan, G. (1973). Axial and random
pattern flaps. British JournalofPlastic Surgery, 26,202.
Nakajima, H., Fqjino, T. and Adachi, S. (1986). A new concept
of vascular supply to the skin and classification of skin flaps
according to their vascularisation. Annuls ofPlastic Surgery,
16, I.
Palmer, J. H. and Taylor, G. I. (1986). The vascular territories
of the anterior chest wall. British Journal of Plastic Surgery,
39,287.
Patten, B. M. (1968). Human Embtyo/ogy. 3rd Edition. New
York: Blakiston Division, McGraw-Hill Book Company.
PontCn, B. (1981). The fasciocutaneous flap: its use in soft tissue
defects of the lower leg. British Journa~ofP~usr~c Surgery, 34,
215.
Quaia, J. and Wilson, E. (1842). A Series of Anatomical Plates in
Lithography. Volume 2. London: Taylor and Walton.
Rees, M. J. W. and Taylor, G. I. (1986). A simplified lead oxide
cadaver injection technique. Plastic and Reconstructive Sur-
gery, 77, 14 1.
Reid, C. D. and Taylor, G. I. (1984). The vascular territory of the
acromiothoracic axis. British Journal of Plastic Surgery, 37,
194.
Salmon, M. (1936a). Art&es de la Peru. Paris: Masson et Cie.
Sahnon, M. (1936b). Art&es des Muscles des Membres et du
Tronc. Paris: Masson et Cie.
Salmon, M. (1987). Arteries of the Skin. Ed. Taylor, G. I. and
Tempest, M. London:-. Churchi Livingstone, In press.
-. __
Shanahan, H+ 6. and Gingrass, R. P. (1979). Medial plantar
THE VASCULAR TERRITORIES (ANGIOSOMES) OF THE BODY 141

sensory flap for coverage of heel defects. Plastic and Taylor, G. I., Townsend, P. L. and Corlett, R. J. (1979b).
Reconstructive Surgery, 64.295. Superiority of the deep circumflex iliac vessels as the supply
Shaw, D. T. and Payne, R. L. (1946).One-staged tubed abdominal for free groin flaps : clinical work. Plastic and Reconstructive
flaps. Surger);, Gynecology and Obstetrics, 83,205. Surgery, 64, 745.
Song, R., Gao, Y., Song, Y., Yu, Y. and Song, Y. (1982). The Taylor, G. I. and Corlett, R. J. (1981). The vascular territories of
forearm flap. Clinics in Plastic Surgery, 9,21. the body and their relation to tissue transfer. Plastic Surgery
Spalteholz, W. (1893). Die Vertheilung der Blutgefasse in der Forum, 4, I 13.
Haut. Archive .fir Anatomie und Physiologic. (Anatomische Taylor, G. I. (1983). Foreword in Manchot, C., The Cutaneous
Ab~theihmg). I, 54. Arteries of‘ the Human Body. Translated by Risic. J. and
Tans%, I. (1906). Sopra il mio nuovo process0 di amputazione Morain, W. D. New York : Springer-Verlag.
della mamella. (Coverage of the anterior chest wall following Taylor, G. I., Corlett, R. J. and Boyd, J. B. (1983). The extended
mastectomy). Gazetta Medica Italiana, 57, 141. deep inferior epigastric flap: a clinical technique. Plastic and
Taylor, G. I. and Daniel, R. K. (1973). The free flap : Composite Reconstructive Surgery, 72,75 1.
tiss,ue transfer by vascular anastomosis. Australian and New
Taylor, G. I., Corlett, R. J. and Boyd, J. B. (1984). The versatile
Zealand Journal of Surgery, 43, I. deep inferior epigastric (inferior rectus abdominis) flap.
Taylor, G. I., Miller, G. D. H. and Ham, F. J. (1975). The free British Journal of Plastic Surgery, 37. 330.
vascularised bone graft. Plastic and Reconstructive Surgery,
Timmons, M. J. (1985). Landmarks in the anatomical study of
55, 533.
the blood supply of the skin. British JournalofPlastic Surgery,
Taylor, G. 1. and Daniel, R. K. (1975). The anatomy of several
38, 197.
free flap donor sites. Plastic and Reconstructive Surgery, 56,
Webster, J. P. (1937). Thoraco-epigastric tu bed pedicles. Surgical
243.
Clinics of North America, 17, 145.
Taylor, G. Land Ham, F. J. (1976). The free vascularised nerve
graft. Plastic and Reconstructive Surgery, 57,413.
Taylor, G. 1. (1978). Nerve grafting with simultaneous microvas-
cular reconstruction. Clinics in Orthopaedics, 133, 56.
The Authors
Taylor, G. I. and Watson, N. (1978). One stage repair of G. I. Taylor, FRACS, FRCS, Consultant Plastic Surgeon, Royal
compound leg defects with revascularised flaps of groin skin Melbourne Hospital.
and iliac bone. Plastic and Reconstructive Surgery, 61,494. J. H. Palmer, FRCS, formerly Research Fellow, Department of
Taylor, G. I. and Townsend, P. L. (1979). Composite free flap Plastic and Reconstructive Surgery, Royal Melbourne Hospi-
and tendon transfer: an anatomical study and a clinical tal, and Department of Anatomy, University of Melbourne.
technique. British JournalofPlastic Surgery, 32. 170.
Taylor, G. I., Townsend, P. L. and Corlett, R. J. (1979a).
Requests for reprints to: Mr G. Ian Taylor, 766 Elizabeth Street.
Superiority of the deep circumflex iliac vessels as the supply
Melbourne 3000, Australia.
for free groin flaps: experimental work. Plasticand Reconstruc-
tive Surgery, 64, 595. Received 30 August 1986.