British Journal of Plastic Surgery (2000), 53, 51-57

9 2000 The British Association of Plastic Surgeons

Article no. BJPS. 1999.3186

I BRITISH JOURNAL OF ~ ) PLASTIC SURGERY

Generation of an autologous tissue (matrix) flap by combining an arteriovenous

shunt loop with artificial skin in rats: preliminary report

Y. Tanaka, A. Tsutsumi*, D. M. Crowe, S. Tajimat and W. A. Morrison

Bernard O'Brien Institute of Microsurgery, St Vincent's Hospital, Melbourne, Australia; *Department of
Laboratory Medicine, Division of Surgical Pathology and 1"Departmentof Plastic and Reconstructive Surgery,
Osaka Medical College, Osaka, Japan

SUMMARY. The present experiment was designed to investigate the possibility of prefabricating a tissue flap in a
rat by combining an arteriovenous (A-V) shunt loop with artificial skin dermis (AS). The A-V fistula loop was
constructed between the right femoral artery and vein by the interposition of a vein graft and the loop was
wrapped with a folded sheet of AS and buried beneath the inguinal skin. In the control group the folded sheet of
AS was inserted without a vessel loop and embedded in the inguinal region as in the experimental group. There
were three experiments. In experiment 1, the total volume of the generated tissue formed within the AS was calcu-
lated after 4 weeks in the experimental and control groups. In experiment 2, the AS in the experimental group was
harvested at 2 (group 1) and 4 (group 2) weeks after insertion to assess the change in morphology over time. In
experiment 3, full thickness skin grafts were placed over the generated tissue of the experimental groups to investi-
gate the possibility of creating skin flaps. The total volume of tissue generated in the experimental group was sig-
nificantly greater than in the control group (P < 0.01). Histological and carbon injection studies suggest that the new
capillary bed is derived from the graft loop vessels and tissue generation and organisation of the AS were further
advanced in group 2 than in group 1. The skin grafts placed over the tissues generated showed complete survival and
could be raised as island flaps in both groups. 9 2000 The British Association of Plastic Surgeons

Keywords: prefabrication tissue engineering, tissue flap, angiogenesis, arteriovenous shunt.

For reconstructive purposes a variety of synthetic
biomaterials such as artificial skin dermis (AS) and
hydroxyapatite have been developed as substitutes for
autogenous tissue sources in an endeavour to reduce
donor site morbidity and promote healing) ~ The
applicability of these synthetic biomaterials is depen-
dent on the ingrowth of vascular tissue and their use is
still restricted to areas which are reliably perfused by
the surrounding tissue. To facilitate graft take in an
unfavourable bed Erol and Spira reported the forma-
tion of a new capillary bed around an implanted arteri-
ovenous (A-V) shunt loop. This could then support an
overlying skin graft? This vascular implantation prin-
ciple has been utilised in the development of various
'prefabricated' composite free flaps to extend recon-
structive options. 6-12Although an A-V shunt loop can
generate new blood vessels it is debatable whether new
tissue is also formed. The present study was designed
to investigate the possibility of creating a new tissue
(matrix) flap by combining an A-V shunt loop with an
artificial skin dermis.
The animals were anaesthetised by intraperitoneal
administration of 30 mg/kg of nembutal in all opera-
tive procedures. Surgery was performed under sterile
conditions using an operating microscope.
Experimental design
In the experimental group, an A-V shunt loop, which
w a s created between the right femoral artery and vein
using an interpositional vein graft, was sandwiched in
a folded sheet of heparin-saline-soaked artificial skin
dermis 1.5 • 2.0 cm in size (PELNAC~M| Gunze Co
Ltd, Kyoto, Japan), and the composite was left in place
in the upper thigh (Fig. 1A,B). In a separate control
group of rats a folded sheet of AS without any vascu-
lar system was implanted into the upper thigh as in the
experimental group. AS is composed of an outer sili-
cone layer and an inner layer of collagen sponge
(porcine type I collagen - atelocollagen). The thick-
nesses of the silicone and the sponge were 150 ~tm and
3 mm, respectively, with a pore size of 70-110 ~tm as
described by Suzuki et al. 3,4

Materials and methods

Male Sprague Dawley rats weighing 250-350 g were
used in all experiments. All experiments were con-
ducted according to National Health and Medical
Research (NHMRC) guidelines for animal welfare.
Operative procedure
Through a skin incision parallel to the inguinal ligament
on both sides, bilateral femoral vessels were fully
exposed and dissected from the inguinal ligament proxi-
mally to the bifurcation of the saphenous and popliteal

51
52 British Journal of Plastic Surgery

Figure I (A) An A-V shunt loop constructedbetweenthe femoral artery and vein using an interpositionalvein graft. Artificialskin (AS)
was placed onderneath the loop vessel.(B) Loop vessel sandwichedwith a foldedsheet of AS and fixedin place.

vessels distally. The fight femoral artery and vein were
divided proximal to the origin of the epigastric vessels
and prepared for microvascular anastomoses. A vein
graft approximately 1.5 cm in length was harvested from
the left femoral vein. Interposing this vein graft in
reverse fashion, an A-V shunt loop was created between
the proximal stumps of the fight femoral artery and vein
by end-to-end anastomosis using 10-0 monofilament
nylon sutures. In the experimental groups the A-V shunt
loops were sandwiched in the folded heparin-soaked
sheet of AS, and the composites anchored in place in the
upper thigh using 5-0 nylon sutures.
In the control group animals, the AS without an
A-V shunt loop was folded and inserted in the thigh
as in the experimental group.
layer of AS was peeled off and a full thickness skin
graft harvested from the left abdomen was placed over
the generated tissue. The chamber was then closed and
kept in place in the u p p e r thigh using 54) nylon
sutures, and the wounds closed. Ten days after skin
grafting the chamber was opened and the skin graft
raised as an island skin grafted flap based on the A-V
shunt loop. In three animals of each of the 2 and 4
week groups, the skin grafted flaps were harvested for
histological study after India ink perfusion. In the
remaining three animals in each of the 2 and 4 week
groups the skin grafted flaps were transferred to the
ipsilateral inguinal skin with 6 0 nylon sutures. The
survival of the skin grafted flaps was evaluated macro-
scopically 1 week after transfer.

Experiment 1 ( Volume of tissue generated)

Assessment
Four weeks after the first operation the contents within
the AS were harvested including silicone layer both in Histological examination
the experimental and control groups. The total volume
All tissues were fixed with 10% neutral buffered forma-
o f generated tissue was assessed by planimetry of serial
lin. After fixation, tissues were embedded in paraffin
histological sections and the experimental and control
and 5 ~t thick histological sections were prepared by
groups compared (n = 5).
standard methods and stained with haematoxylin and
eosin or Masson's trichrome.
Experiment 2 (Time course of tissue generated)
The same experiment was repeated implanting the
A-V loop into the folded AS. The contents within the
AS were harvested after 2 weeks or 4 weeks (groups 1
and 2: five animals in each) and the tissues were per-
fused with India ink by direct injection into the proxi-
mal femoral artery. The nature and extent of generated
tissue was evaluated histologically.

Experiment 3 (Skin grafting and islandflap transfer)

The experiment was repeated and the AS and its tissue
content were raised at 2 weeks or 4 weeks (groups 1
and 2: six animals in each) as island flaps based on the
A-V shunt loop. This tissue matrix flap was trimmed
and placed into a cylindrical chamber (polyestercar-
bonate, 1.4 cm in inner diameter) (Fig. 2). The silicone Figure 2--The generatedtissue placed into the chamber.
Autologous flap prefabrication in rats
Volumetry of the generated tissue
After embedding in paraffin, the generated tissues were
serially cross-sectioned at a 500 ~t intersection distance
and 5 ~t histological specimens prepared at each section
and stained with Masson's trichrome. The area of gen-
erated tissue was measured at consecutive cross-sec-
tions using Video Pro 32 colour image analysis
software (Watcom C386 and 386/ASM).
To be included in the area of generated tissue the fol-
lowing criteria were considered necessary: (1) migration
and proliferation of fibroblasts and new blood vessel
formation; and (2) complete replacement of the atelo-
collagen by host-derived collagen. In the experimental
groups the area of generated tissue was calculated by
deleting the area of pre-existing loop vessels. The area
of generated tissue was multiplied by the distance
between the cross-sections and total volume was deter-
mined by summation of all contributions.
Carbon injection study
The generated tissue was completely raised on its
femoral artery and vein pedicle. India ink was perfused
by direct injection into the femoral artery after irriga-
tion with heparinised saline. The pedicle was then lig-
ated with a silk suture and the tissue fixed in 10%
neutral buffered formalin, cleared in cedar wood oil
and examined under a microscope by transmitted light.
Statistical analysis
Unpaired two-tailed Student's t-tests were used to
compare the experimental groups with the controls.
Results
All of the implanted AS were encapsulated with
fibrous tissue. Thrombosis in the A-V shunt loop
developed in two rats, one in experiment 1 and one in
experiment 3. Infection was noted in one control rat.
Therefore three additional animals were operated on to
maintain the experimental design numbers.
Figure3--Photomicrographsof the control and experimentalgroups (Masson's Trichrome;x 5). (A) Control group; new tissue was not noted
at the centre of AS. (B) Experimentalgroup; new tissue was formed surrounding the A-V shunt loop.
53
Experiment 1
In the control group, new tissue formed at the free mar-
gins of the AS but no cell migration was found cen-
trally (Fig. 3A). In the experimental group, new tissue
was formed surrounding the A-V shunt loop in addi-
tion to the generated tissue at the free margins of the
AS (Fig. 3B). A larger total volume of tissue was gen-
erated in the experimental group compared with the
control group (52.2 [13.4] mm 3 vs 18.3 [6.1] mm 3, P <
0.01) (Fig. 4). The histological appearance is described
under experiment 2.
Experiment 2
A sheet of generated tissue was observed surrounding
the A-V shunt loop in both the 2 and 4 week groups
(Fig. 5). Formation of the new tissue was further
advanced at 4 weeks macroscopically. At 2 weeks the
AS-derived collagen still remained and fibroblasts were
intricately arranged within the collagen sponge as well
as angiogenic tissue, especially in the region of the apex
of the A-V shunt loop (Fig. 6A). By 4 weeks, however,
most of the AS-derived collagen had been replaced by
host-derived fibroblasts and collagen fibres, which
were arranged at right angles to the direction of the
new capillary extensions and showed advanced organi-
sation (Fig. 6B). India ink injection studies in both
groups showed many new blood vessels originating
from the loop vessels and extending to the periphery
(Fig. 7).
Experiment 3
The chamber was encapsulated with fibrous tissue.
When the chamber was opened, yellow viscous liquid
was noted, which seemed to be derived from the seba-
ceous glands of the skin graft. However, the skin grafts
showed survival in both the 2 and 4 week groups and
this was confirmed histologically (Fig. 8A,B). In the 2
week group new capillary vessels extended into the
overlying skin dermis from the underlying generated
tissue. There were remnants of atelocollagen sur-
rounded by inflammatory cells in the generating tissue
54 British Journal of Plastic Surgery

70
~*P<O.O1

601
50

30-
= 20-
[--. 10i_

0 I
control a-v loop Figure 5--A sheet of new tissue surrounding the A-V shunt loop 4
weeks after implantation: the shape of the generated tissue
Figure 4--The total volume of generated tissues 4 weeks after corresponds with that of AS.
implantation showed significant differencebetween the
experimental group and the control group (P < 0.01, n = 5 in each).

Figure 6~Higher magnificationof generated tissue (H&E x 100). (A) Group 1; AS-derived atelocollagenremained (-o; atelocollagen)and
fibroblasts were intricately arranged. (B) Group 2; maturation of generated tissue was advanced.

ous (Fig. 9B). The skin grafted island flaps transferred

to the inguinal skin all survived even when applied at 2
weeks (Fig. 10A,B).

Discussion

In 1980, Erol and Spira reported the possibility of an

Figure 7--Carbon injection study in group 1 showingvascular
sprouts and migration from the loop vessel (x 25).
(Fig. 9A). At 4 weeks, maturation of the generated
tissue was further advanced and capillary vessel exten-
sions into the overlying skin-dermis were more numer-
A-V shunt loop in combination with a full thickness
skin graft producing sufficient vascularity to create
new flaps as well as proposing the implantation of
A-V loops to improve the circulation in replanted
fingers. 5 Since then, many experimental studies of pre-
fabricated skin flaps using an A-V shunt loop as a
vascular carrier have been reported. 8,1~ In 1990,
Suzuki et aP reported the clinical application of AS
and showed that the pores of AS were infiltrated by
cellular tufts of fibroblasts and capillaries when placed
over an open wound, and the sponge-like atelocollagen
of the AS was gradually replaced by dermis-like tissue.
The present experiment attempts to combine the
potential benefits of Erol's findings with the properties
Autologous flap prefabrication in rats 55

Figure 8~Photomicrograph of a skin grafted flap. (A) 2 week group. (B) 4 week group. The skin graft showed complete survival in both
groups (H&E; x 10). (A: artery, V: vein, S: skin graft, G: generated tissue).

A B

Figure9--High-powered photomicrograph of a skin grafted flap. (A) 2 week group; capillary extensions into the dermis of the skin graft are
indicated by the arrow (H&E; x 100). (B) 4 week group; many capillaries stained with India ink in the dermis of the skin graft (H&E; x 100).
A: artery, S: skin, G: generated tissue, B: thin band formed between the skin graft and the generated tissue,

o f the AS with a view to creating a new tissue matrix
flap.
Experiment 1 clearly demonstrated that new tissue
was generated and that it not only replaced the AS
matrix, but grew well beyond it and possibly indepen-
dent o f it, displacing it outwards (Fig. 3B). At 4 weeks
the volume o f the new tissue outstripped that o f the
original AS sheet by at least 3:1 (Fig. 4). Figure 3B
shows the original AS collagen matrix spanning the
superior and inferior surfaces o f the flap. Superiorly
the silicone sheet is still present on the surface. Matrix
replacement can be seen but the volume o f new tissue
centrally greatly exceeds the matrix volume.
The nature of the new tissue macroscopically was
that o f a soft vascular semi-formed structure which
withstood and maintained its shape on manipulation. It
was not friable and was clearly more than simple granu-
lation tissue. Histologically, it was composed o f a colla-
gen matrix which became more organised with time. The
cellular content included fibroblasts and mast cells, the
latter consistent with angiogenesis. Endothelial buds
were prevalent throughout. C a r b o n injection studies
56
Figure10--(A) Immediatelyafter transfer. (B) One week after transfer. Transferredisland skin grafted flap showing hair growth (dissecting
forceps outlining the graft).
(Fig. 7) revealed that many newly formed blood vessels
grew out of the loop vessels and migrated into the sur-
rounding tissue, which is consistent with the generally
accepted mechanism of angiogenesis that endothelial
cells in a parent vessel penetrate the vascular basal lam-
ina, migrate into the surrounding tissue in a directional
fashion and subsequently divide, and by repeated
sprouting and anastomosis, form a vascular network
characteristic of the organ. ~3,~,~7
The factors responsible for angiogenesis in the pre-
sent model are considered to be 'increased shear
stress' and 'increased wall tension' produced by
arteriovenous shunting. 15,16These factors can 'initiate
the process of angiogenesis by disturbing the integrity
of the endothelial or vascular smooth muscle cells
and making them more susceptible to the effect of
growth factors'. 15 Therefore, endothelial sprouting
from the vasa v a s o r u m and/or the lumen of the loop
vessels is the most likely source for angiogenesis in the
present model. An A-V shunt loop was chosen as a
vascular carrier rather than a ligated A-V bundle
because new vessel growth is proportional to blood
flow velocity. 14
From these results, the following mechanism of tissue
generation is postulated: the active formation of new
blood vessels and vascular network from the loop vessels
is followed by migration and proliferation of fibroblasts.
The resulting proliferation and arrangement of collagen
fibres lead to formation of new tissue, which extends
peripherally and joins the generated tissue formed at the
free margins of the AS. These generated tissues were
able to be raised as island flaps and the skin grafts
placed over them survived even in the area showing
remaining atelocollagen. Histologically, capillary exten-
sion into the dermis of the skin grafts was observed in
both groups 1 and 2. However, for creating island skin
grafted flaps, it may be safer to perform skin grafting 4
weeks after implantation when the organisation of the
AS is much more advanced. The reliable take of grafts
and their continued survival after pedicle flap transfer
were evidence that the newly generated tissue was sus-
tained and did not atrophy at least for the 6 weeks and 3
days of the experiment.
British Journal of Plastic Surgery
The flap produced in the present study is classified as
a prefabricated flap. The shape of the generated tissue
was much the same as that of the AS. This suggests that
the flap shape can be modified according to the shape
of the AS. Possible further growth of this tissue-gener-
ated flap beyond 4 weeks has not been tested in this
experiment. In 1993, Khouri et al is demonstrated tissue
generation with growth factors using arteriovenous
bundles as a vascular carrier. They also used r - P D G F
impregnated collagen disks around the vascular bun-
dles. A collagen sponge or disk has the merit of offering
an environment for tissue generation as well as a repos-
itory for various growth factors. In the future, it may be
possible to produce various prefabricated flaps of a
particular tissue differentiation by combining an A-V
shunt or other vascular carriers with AS in which
cytokines such as bFGF, TGF-[3 and BMP or cell cul-
tures are integrated. Since this flap is considered the
prototype of non-differentiated prefabricated ones, it
was termed a 'matrix flap'.
Acknowledgements
The authors would like to acknowledgethe contribution of all staff
in the ExperimentalMedical Surgical Unit, St Vincent's Hospital, for
the assistanceof the experimentalstudy. This work was supported by
a Grant in Aid for Scientific Research (C) (No. 07807119) from the
Japanese Ministry of Education, a project grant from St Vincent's
Hospital, and the Helen M. Schutt Trust.
References
1. Salyer KE, Hall CD. Porous hydroxyapatite as an onlay bone-
graft substitute for maxillofacialsurgery. Plast Reconstr Surg
1989; 84: 23644.
2. Costantino PD, Friedman CD, Jones K, Chow LC, Sisson GA.
Experimental hydroxyapatite cement cranioplasty. Plast
Reconstr Surg 1992; 90: 174-85.
3. Suzuki S, Matsuda K, Isshiki N, Tamada Y, Yoshioka K, Ikada
Y. Clinical evaluation of a new bilayer 'artificial skin' com-
posed of collagen sponge and silicone layer. Br J Plast Surg
1990; 43: 4754.
4. Suzuki S, Matsuda K, Maruguchi T, Nishimura Y, Ikada Y.
Further applicationsof 'bilayer artificial skin'. Br J Plast Surg
1995; 48: 222-9.
Autologous flap prefabrication in rats
5. Erol OO, Spira M. New capillary bed formation with a surgically
constructed arteriovenous fistula. Hast Reeonstr Surg 1980;
66: 109-15.
6. Fisher J, Yang W-Y Experimental tissue molding for soft-tissue
reconstruction: a preliminary report. Hast Reconstr Surg
1988; 82: 857-64.
7. Valauri FA, Hirase Y, Buncke HJ. Prefabricated neovascularized
free muscle flaps: pedicle variations. J Reconstr Microsurg
1988; 4: 203-7.
8. Hirase Y, Valauri FA, Buncke HJ. Creation of neovascularised
free flaps using vein grafts as pedicles: a preliminary report on
experimental models. Br J Hast Surg 1989; 42: 216-22.
9. Morrison WA, Dvir E, Doi K, Hurley JV, Hickey M J, O'Brien
BMcC. Prefabrication of thin transferable axial-pattern skin
flaps: an experimental study in rabbits. Br J Hast Surg 1990;
43: 645-54.
10. Takato T, Zuker RM, Turley CB. Prefabrication of skin flaps
using vein grafts: an experimental study in rabbits. Br J Hast
Surg 1991; 44: 593-8.
11. Khouri RK, Upton J, Shaw WW. Prefabrication of composite
free flaps through staged microvascular transfer: an experi-
mental and clinical study. Hast Reconstr Surg 1991; 87:
108-15.
12. Wilson YT, Kumta S, Hickey MJ, Hurley JV, Morrison WA. Use
of free interpositional vein grafts as pedicles for prefabrica-
tion of skin flaps. Microsurgery 1994; 15: 717-21.
13. Rhodin JAG, Fujita H. Capillary growth in the mesentery of
normal young rats. lntravital video and electron microscope
analyses. J Submicrosc Cytol Pathol 1989; 21: 1-34.
14. Risau W. Angiogenic growth factors. Prog Growth Factor Res
1990; 2: 71-9.
15. Hudlicka O, Margaret DB. Physical forces and angiogenesis.
Mechanoreception by the vascular wail. In: Rubanyi GM.
(ed), Angiogenesis. Mount Kisco, NY: Futura Publishing Co.
Inc., 1993; pp. 198-241.
16. Risau W. Mechanisms of angiogenesis. Nature 1997; 386: 6714.
57
17. Jain RK, Schlenger K, H6ckel M, Yuan E Quantitative anglo-
genesis assays: progress and problems. Nature Medicine 1997;
3: 1203-8.
18. Khouri RK, Koudsi B, Deune EG, et al. Tissue generation with
growth factors. Surgery 1993; 114: 374-9.
The Authors
u Tanaka MD, Honorary Research Fellow
Bernard O'Brien Institute of Microsurgery, St Vincent's Hospital,
Melbourne, Australia. Currently Associate Professor, Department
of PIastie and Reconstructive Surgery, Osaka Medical College,
Osaka, Japan.
Akira Tsutsumi MD, Head
Department of Laboratory Medicine, Division of Surgical
Pathology, Osaka Medical College, Osaka, Japan.
Daniel M. Crowe PhD, Research Scientist
Bernard O'Brien Institute of Microsurgery, St Vincent's Hospital,
Melbourne, Australia.
Sadao Tajima MD, Professor
Department of Plastic and Reconstructive Surgery, Osaka Medical
College, Osaka, Japan.
Wayne A. Morrison MD, BS, FRACS, Director
Bernard O'Brien Institute of Mierosurgery, St Vincent's Hospital,
Melbourne, Australia.
Correspondence to Professor W A. Morrison, Bernard O'Brien
Institute of Microsurgery, 42 Fitzroy Street, Fitzroy, Victoria 3065,
Australia.
Paper received 27 November 1998.
Accepted 15 June 1999, after revision.