1. Skin and oral mucosal substitutes
Kenji Izumi, DDS, PhDa
, Stephen E. Feinberg, DDS, MS, PhDb,*
a
Department of Tissue Regeneration and Reconstruction, Division of Reconstructive Surgery for Oral and Maxillofacial Region,
Niigata University Graduate School for Medical and Dental Sciences, Niigata City, Niigata, Japan
b
Department of Oral and Maxillofacial Surgery, University of Michigan Medical Center,
B1–A 235 UH, Box 0018, 1500 East Medical Center Drive, Ann Arbor, MI 48109, USA
Biologic substitutes of human skin and mucosa
have several prospective applications, including, but
not limited to, (1) models of skin and mucosal biol-
ogy and pathology, (2) treatment and closure of skin
and mucosal wounds, (3) alternatives to animals for
safety testing of consumer products, and (4) delivery
and expression of transfected genes [3]. This article
introduces the reader to the types of skin and mucosa
substitutes that have been and are being developed in
the area of tissue engineering for use in procedures
for trauma, ablative oncologic resections, and recon-
structive surgery.
Skin and mucosal substitutes have a common set
of requirements for the duplication of anatomic
structures and physiologic functions that they are to
emulate. For use in wound closure, the first definitive
requirement is re-establishment of the epidermal
barrier to fluid loss and microorganisms and alle-
viation of pain and enhancement of wound healing.
In full-thickness skin loss, replacement of the epi-
dermis and dermis is the preferred approach. Replace-
ment of these tissue components also must minimize
scar formation and restore acceptable function and
cosmesis. A major advantage of the use of substitutes
in wound closure is reduction or elimination of the
donor site for skin and mucosal grafts. Success in the
elimination or minimization of donor site morbidity
could shorten recovery time and reduce the length of
operative procedures.
Approaches that have been used in the fabrication,
manufacturing, and ‘‘tissue engineering’’ of skin and
mucosa substitutes can be classified as (1) in vitro
culturing of autologous and allogeneic keratinocytes,
(2) in vitro tissue engineering of dermis composed of
either artificial (collagen, glycosaminoglycans, poly-
mers of polyglycolic, and polylactic acid) or alloge-
neic and acellular dermis, and (3) a bilayer of skin
mucosa from a combination of (1) and (2). Three
essential components are known to be necessary to
engineer human skin and mucosa: cells, an extra-
cellular matrix, and cytokines [30,42].
Skin substitutes
A major milestone in the development of skin
substitutes was the introduction of the in vitro tech-
nique of Rheinwald and Green [47] that involved the
culturing of human keratinocytes into epithelial
sheets suitable for autografts. These investigators
used a combination of hydrocortisone, epidermal
growth factor, and irradiated murine 3T3 fibroblasts
to support the proliferation of keratinocytes on plastic
substrates. The following ingredients were added
to improve the culture media and to facilitate kerati-
noctye sheet formation: (1) insulin, to promote the
uptake of glucose and amino acids; (2) transferrin, to
detoxify iron; (3) hydrocortisone, to promote the
attachment of cells and cell proliferation; (4) triiodo-
thyronine, as a mitogen for keratinocytes; and (5)
cholera toxin, to upregulate cAMP levels. This
1042-3699/02/$ – see front matter D 2002, Elsevier Science (USA). All rights reserved.
PII: S1042-3699(02)00010-9
This work was supported by a Grant-in-Aid for
Scientific Research (No. 12771216) from the Ministry of
Education, Science and Culture, Japan (KI) and from Grant
No. DE13417 from the National Institute of Health,
USA (SEF).
* Corresponding author.
E-mail address: sefein@umich.edu (S.E. Feinberg).
Oral Maxillofacial Surg Clin N Am 14 (2002) 61–71

2. method allowed keratinocytes to grow into a large
epithelium sheet — 10,000 times larger than the ori-
ginal biopsy — in a short time period, enough to
cover the large body surface areas. Investigators were
then able to release sheets of keratinocytes by using
the enzyme dispase, which could digest adhesive
molecules holding the keratinocytes to the plastic
substrate without digesting the cohesive links be-
tween the keratinocytes themselves. O’Connor et al.
[39] reported the first human use of autologous
cultured keratinocyte sheet grafts for burn wounds.
Cultured keratinocyte sheets have been applied in
various clinical scenarios such as chronic skin ulcers
[51], congenital nevi [20], and junctional epider-
molysis bullosa [5]. This approach revolutionized
surgical treatment of burns and was the forerunner
to the impetus to engineer skin to enhance tissue
regeneration and repair.
Two major approaches are currently used for in
vivo tissue engineering of skin and mucosa. The older
technique of Rheinwald and Green [47] is based on
the use of a serum-containing medium and a feeder
layer of lethally irradiated transformed cell line of
mouse fibroblasts. The second approach relies on
serum-free media in the absence of a feeder layer.
The presence of irradiated feeder cells and serum can
be a confounding factor in keratinoctye cultures when
used for experimental research. This technique also
would have difficulties receiving Food and Drug
Administration (FDA) approval because of the poten-
tial transmission of unknown elements such as slow
viruses to graft recipients. For this reason, many
investigators have tried to avoid using feeder cells
and to reduce the amounts of serum and additives,
such as pituitary extract.
The various methods for in vitro engineering of
the epidermis require the use of (1) coating of
culture surfaces with molecules found in the extra-
cellular matrix, such as collagen, fibronectin, or lam-
inin, that assist in simulating in vivo conditions and
(2) variable concentrations of calcium in the culture
medium. Calcium ions play a vital role in the
growth and differentiation of keratinocytes. Increas-
ing calcium concentration is accompanied by an
increasing level of keratinocyte differentiation, as
evidenced by increasing numbers of formed desmo-
somes and the formation of multilayers and sheets of
cells. In vitro engineering of the epidermis also
requires (3) the use of complex biologic extracts
such as bovine pituitary extract and (4) the addition
of mitogens or trace elements.
For a skin or mucosa substitute to obtain FDA
approval, it is necessary to eliminate the use of serum
and xenogeneic feeder layers, undefined biologic
extracts such as pituitary extract, and products that
may be modes of transmission of diseases, such as any
bovine products that are not certified from disease-
free herds. Several skin substitutes are approved by
the FDA and are commercially available.
Epithelium
Epicel (Genzyme Tissue Repair Corp, Cam-
bridge, MA) is composed of cultured epithelial
autogenous keratinocyte sheets and has been com-
mercially available since 1988. The procedure of
generating cultured epithelial sheets follows the tech-
nique of Rheinwald and Green [47]. The epithelial
sheets are composed of several stratified layers of
keratinocytes that are formed by culturing submerged
cells (totally covered by medium) for 10 to 15 days.
An epithelial sheet produced by their culture system
has several advantages. The first advantage is the
possibility of a large expansion from a small donor
site of 2 cm2
up to 10,000-fold, which could cover
a full-surface body area of an adult, approximately
2 m2
. The second advantage is the low risk of
transmission of viral diseases such as bovine spongi-
form encephalopathy. Early clinical studies of epi-
thelial sheets used for the treatment of burn wounds
had encouraging outcomes [19,37]. Later clinical
experiences, in contrast, demonstrated low graft
‘‘take’’ rate [16,48]. Disadvantages of cultured sheets
of grafted keratinocytes include (1) time-consuming
growth of keratinocyte sheets in which, even under
optimal conditions, it would take 2 to 3 weeks to
obtain a sufficient amount of keratinocyte sheets for
grafting; (2) the use of potentially immunogenic
materials in culture, such as serum, various additives
(pituitary extract), and a xenogeneic feeder layer that
may contribute to graft loss; (3) widely reported
variations in the ‘‘take’’ of keratinocyte sheet grafts,
which is lower than that of split-thickness grafts; (4)
graft instability, such as graft fragility and blister
formation, that may be secondary to the absence of
rete ridges and inadequate formation of epidermal
keratinocytes’ anchoring fibrils; (5) wound contrac-
tion, which is secondary to a difference in specific
characteristics of the keratinocytes within the cul-
tured sheets to that of normal in situ keratinocytes
and to the lack of dermal component within cultured
epithelial sheet graft [10,44]; and (6) the cost of the
grafts production.
Dermis
Dermis plays several biologic and functional
roles in the skin. The most important roles are (1)
K. Izumi, S.E. Feinberg / Oral Maxillofacial Surg Clin N Am 14 (2002) 61–7162

3. providing mechanical support for cells involved in
skin structure formation, immunity, nutrition, and
sensation; (2) providing skin elasticity and tensile
strength; and (3) functioning as an anchor for
epithelial glands and keratinizing appendage struc-
tures of the skin (hair, nails). During skin healing,
the presence of dermis also supports faster reepithe-
lialization, inhibits wound contraction, and improves
esthetic outcome. Increased understanding of dermal
structure and composition has guided the develop-
ment of artificial dermal substitutes. Structure is
only one of the material property requirements.
Dermis also should be pliable, hemocompatible,
minimally immunogenic, and eventually degradable
and must minimize fluid loss and reduce scarring
and contractures.
In designing a dermal replacement, one must take
several clinical observations into consideration: (1)
the thicker the dermal layer of a split-thickness skin
graft, the less the graft contracts; (2) full-thickness
skin grafts contract minimally; (3) full-thickness
dermal injuries heal by contraction and hypertrophic
scarring, which produce subepithelial scar tissue that
is nothing like the original dermis; (4) partial-thick-
ness wounds with superficial dermal loss heal with
less hypertropic scarring; and (5) the length of illness
in burn cases is essentially restricted to the length of
time the burn wound is open. One might hypothesize
from these observations that the dermis provides
information to the wound that modulates the healing
process. If so, then a dermal replacement should
provide the information necessary to control the
inflammatory and contractile processes and the in-
formation necessary to evoke ordered re-creation of
autologous tissue in the form of a neodermis. The
initial replacement material also should provide
immediate physiologic wound closure and be elimi-
nated once it has provided sufficient information for
reconstitution of a neodermis.
Yannas, Burke, and colleagues [4,22,55] focused
on these observations and developed a bilayered two-
stage model that resulted in a FDA-approved product
that is marketed as Integra (Integra Life Sciences
Corp, Plainsboro, NJ). Integra has a top layer of a
silicone elastomer and a bottom layer of a porous
network of cross-linked collagen and glycosamino-
glycan. The rationale for the top silicone elastomer is
to control bacterial ingress and water evaporation and
provide additional mechanical support. The bottom
layer is designed to ensure rapid wound adherence.
Optimal porosity of this dermal analogue allows for
its slow biodegration and the induction of vascular
and cellular ingrowth that eventually replaces the
dermal matrix with neodermis. Several weeks later,
when fibrovascular ingrowth of the dermal analogue
has occurred, the epidermal silicone layer must be
removed in the operating room and the wound closed
with a thin sheet of autologous skin [4,22,55]. Clin-
ical studies with this material have been successful
[21,49] and were approved by the FDA in 1997.
Dermagraft (Advanced Tissue Science Inc, La
Jolla, CA) [15,38] is another dermal substitute that
uses a resorbable matrix material that is similar to
materials used in suture fabrication. This material is
composed of a polylactin acid on which dermal
fibroblasts are incorporated. Once incorporated within
the matrix, the fibroblasts secrete growth factors and
extracellular matrix proteins.
A dermal allograft harvested from a living or
cadaveric donor is another possible dermal substitute.
Unlike the rejection of allogeneic keratinocytes on
transplantation, allografts of dermis seem to be trans-
plantable without significant rejection because of
comparatively low immunogenicity of the dermal
components. If rejection occurs, it is usually directed
toward the passenger leukocytes and endothelial cells
that line the blood vessels. The use of a de-epider-
mized and decellularized dermis further diminishes
the allograft immunogenicity. Allogeneic, acellular
dermis prepared this way retains the structural archi-
tecture of the remaining dermal matrix. This dermal
matrix has been shown to support fibroblast in-
growth, neovascularization, and keratinoctye migra-
tion from an overlying split-thickness skin graft or
from seeded cultured keratinocytes [29]. A product
on the market, AlloDerm (LifeCell Corp, Branch-
burg, NJ), is an acellular, nonimmunogenic dermis
that retains the extracellular matrix structure and an
undamaged basement membrane complex. It also
possesses an intact vascular channel network that
allows ingrowth of fibroblasts and endothelial cells
from the underlying tissue.
Bilayers of epithelium and dermis
In contrast to the materials science and engineering
approach of Burke and Yannas et al. [4,22,55], Bell
et al. [2] took the approach of reconstituting dermal
wounds by applying a preformed tissue. They started
with the observation that fibroblasts introduced into a
collagen gel would proliferate and reorganize the
collagen into a contracted matrix containing exogen-
ous collagen and the collagen and matrix proteins
produced by the introduced fibroblasts. The rate and
final extent of contraction varied inversely with the
protein concentration and directly with cell number
introduced into the gel. The resulting product is
described as a dermal equivalent, which, unlike
K. Izumi, S.E. Feinberg / Oral Maxillofacial Surg Clin N Am 14 (2002) 61–71 63

4. Integra, relies on living cells in tissue culture to
organize the collagen network. The exact fiber struc-
ture and its relationship to normal dermis are not
known. Subsequent experiments demonstrated in
animals that these collagen gels reorganized by fibro-
blasts could be grafted onto full-thickness injuries
and that they would support the growth of keratino-
cytes into an epidermal equivalent. The dermal com-
ponent is composed of type I bovine collagen that has
been organized by introduction of human fibroblasts.
Foreskin keratinocytes were seeded onto the surface
of the dermal equivalent. After several days of sub-
merged culturing of the skin equivalents, cultures are
then air-exposed to allow the epidermis to stratify,
differentiate, and form a cornified layer [17,43]. The
total manufacturing period is approximately 20 days.
To date, clinical evaluation of this type of skin equiv-
alent has not been reported in burn patients, although
several in vivo animal studies have been conducted
[35]. A bilayered human skin equivalent, Apligraf
(Novartis Pharmaceuticals Corp, East Hanover, NJ),
already has been approved by the FDA for venous
ulcers and is likely to be commercially available for
burn wounds.
Oral mucosa substitutes
Preprosthetic and reconstructive oral and maxillo-
facial surgical procedures often produce open wounds
in the oral mucosa. These wounds should be covered
by a graft to prevent microbial infection, excessive
fluid loss, foreign material contamination, or relapse
(secondary to wound contracture) and assist in the
prosthetic reconstruction of the patient and in the
promotion of wound healing [13]. Currently, oral
mucosa or skin grafts are used for this purpose;
however, both of these grafts require a second sur-
gical procedure and have disadvantages in intraoral
use [34]. Oral mucosa is an excellent intraoral graft
material but is available in a limited supply [31,34].
Split-thickness skin grafts are available in ample
supply but may contain adnexal structures, and they
express a different pattern of surface keratinization
that can lead to the development of abnormal tissue
texture in the oral cavity that could interfere with
function [12,34,36]. The elective nature of most oral
and maxillofacial surgical procedures should allow
the flexibility and timing to develop an ex vivo tissue
engineered oral mucosa that could be used for intra-
oral grafting procedures. The recent developments of
oral keratinocyte culture techniques have paralleled
those of skin keratinocytes [24], which has enabled
the development of tissue-engineered autogenous oral
mucosa that is suitable for intraoral reconstructive
procedures [25].
Structural and functional differences between skin
and oral mucosa
The wet environment of the oral cavity compli-
cates reconstruction with skin grafts. The keratinized
surface of grafted skin tends to macerate and become
easily infected. Oral mucosa is different from skin in
that it has a moist surface and lacks adnexal struc-
tures such as hair and glandular elements. Grafting
of skin into the oral cavity can be complicated by the
presence of adnexal structures, which can be seen as
hair growth within the mouth. Oral mucosa, unlike
skin, presents three structural variations that are
located in specific anatomic locations within the
mouth. These layers are (1) masticatory mucosa
(ortho or parakeratinized; hard palate, attached gin-
giva), (2) lining or alveolar mucosa (nonkeratinized;
lip, floor of mouth, cheek), and (3) specialized
mucosa (taste buds; dorsal surface of tongue). The
keratinized and nonkeratinized mucosa differ in
the composition of their cell layers. In keratinized
mucosa, the suprabasal cell layer is divided into
three layers and designated spinous cells, granular
cells, and keratinized layers with the major cytoskel-
eton keratin of 1/10 [28]. Typical keratinized mucosa
possesses ‘‘keratohyalin granules’’ in the granular
cell layer. Tonofibrils, aggregates of keratin fila-
ments, frequently are seen in the cellular cytoplasm.
In contrast, the suprabasal layer of nonkeratinized
mucosa is less evident and ordered than that seen in
keratinized mucosa. The layers are designated as
spinous cell, intermediate cell, and surface cell layer,
in which the major intermediate filament of keratin
is 4/13. Neither keratohyalin granules nor promi-
nent aggregates of keratin are seen in nonkeratin-
ized mucosa.
In vitro culturing techniques
Most investigators and maxillofacial surgeons
have used an irradiated layer of a transformed 3T3
fibroblastic cell line as a feeder layer to propagate
and expand their oral keratinocyte population to
generate oral keratinocyte (epithelial) sheets for
intraoral grafting [11,45,52,54]. The oral mucosa
epithelial sheet grafts were placed onto the peri-
osteum of the labial aspect of the anterior mandible
to assist in performing a vestibuloplasty. All of the
studies demonstrated successful clinical outcomes
and histologic findings of postgrafting biopsies, the
longest of which was 4 months postoperatively.
K. Izumi, S.E. Feinberg / Oral Maxillofacial Surg Clin N Am 14 (2002) 61–7164

5. Normal epithelial layer was regenerated on the graft
sites. Hata et al. [21] and Ueda et al. [53] reported
that oral mucosa keratinocytes grew more rapidly
and differentiated less than skin keratinocytes. Der-
mal substitutes for burn injuries also have been
applied into the oral cavity, such as the bilayered
membranes with a collagen-GAG/silastic sheet, sim-
ilar to the Burke and Yannas’ Integra [4].
Although they showed successful postoperative
appearances and an advantage of easy sterilization
and cost effectiveness, in the authors’ clinical experi-
ence this material is difficult to handle. The silastic
sheet does not present a problem with handling, but
the collagen sponge becomes ‘‘sticky’’ when it
absorbs blood, which results in difficulty during
suturing. The environment of the oral cavity, a moist
area laden with bacteria and lytic enzymes, may not
be conducive to the collagen-rich dermal components
used in skin equivalents. An oral mucosa equivalent
not only must be anatomically similar to mucosa but
also must possess the mechanical and handling char-
acteristics of the mucosa to be useful within the
intricate confines of the oral cavity.
There have been reports of ‘‘oral mucosa’’ equi-
valent-like Apligraf [40,41]. So far, these equivalents
are still experimental and have not been used in
clinical studies. Another type of ‘‘oral mucosa’’
equivalent composed of de-epidermized dermis and
cultured oral mucosa keratinocytes from buccal
mucosa and hard palate was studied in Korea [6,7].
These oral mucosal substitutes were developed for
toxicologic and pharmacologic studies and not for
use in a clinical setting. Studies have shown that the
concurrent grafting of a dermal component aids in
enhancing the quality and time of wound healing
[26,32,33]. Parenteau et al. [43] showed that the rate
of closure of the wound and the increase in the
percentage of wound repair is enhanced with the
presence of dermis. The maturation process and bio-
logic events of skin regeneration also are accelerated
with the presence of a dermal substrate [9]. Inokuchi
et al. [23] have found that autogenous fibroblasts
within the grafted dermal matrix facilitated the long-
term maintenance of the reorganized cultured epi-
dermis by supporting self-renewal of the epithelium
in vivo. Clugston et al. [8] noted that the absence of a
grafted dermis resulted in a contracture of cultured
keratinocyte autografts on the order of 50%.
The development and grafting of a full-thickness
oral mucosal graft with a dermis can assist in epithe-
lial graft adherence, minimize wound contraction,
and assist in epithelial maturation while encouraging
the formation of a basement membrane [18]. Auger
et al. [1] showed that a dermal equivalent would be
best made out of human, rather than animal, collagen.
The human collagen (dermis) helps to promote
deposition of additional basement membrane constit-
uents, which results in a more optimal pattern of
keratinocyte differentiation and less immunogenicity
than animal collagen. The authors have been success-
ful in their own laboratory in the ex vivo production
of an oral mucosal equivalent (EVPOME) using oral
keratinocytes seeded onto a human cadaver dermal
matrix, AlloDerm [24,25]. AlloDerm is an acellular,
biocompatible, human connective tissue matrix with
an unaltered extracellular matrix and intact basement
membrane, which consistently integrates into the host
tissue. Most importantly, AlloDerm trims, adapts, and
sutures like autologous tissue. Human de-epidermized
dermis that has retained its basal lamina, consisting
of keratinocytes combined with a mesenchymal or
dermal component, has successfully shown enhanced
epithelial morphogenesis and an increase in expres-
sion of differentiation markers when it is grown at an
air-liquid interface [46].
Tissue-engineered oral mucosa
Most reconstructive procedures in oral and max-
illofacial surgery are of an elective nature. This gives
surgeons the ability to time the biopsy of autogenous
mucosa with the need of a sufficient size of tissue
necessary for the planned surgical reconstruction. In
developing a methodology to engineer any tissue, it is
necessary to abide by the requirements and restric-
tions imposed by the FDA. The cultivating technique
of Rheinwald and Green [47] uses a xenogeneic
irradiated fibroblast cell line, 3T3, as a feeder layer
to enhance keratinocyte growth. During the culturing
period to expand human cells they are exposed to a
transformed murine cell line. This contact potentially
could contribute to cross-examination or transfection
of the mutational or xenogeneic DNA into the cocul-
tured human keratinocytes. Serum and a xenogeneic
feeder layer contain undefined factors such as slow
viruses (Creutzfeld-Jakob disease, ‘‘mad cow’’ dis-
ease, or bovine spongiform encephalopathy) and
foreign contaminants [50]. The importance of not
using a feeder layer and serum to culture oral mucosal
autografts is obvious, especially in elective surgery
because of the potential of the introduction of unde-
termined risks to the patient. Other investigators also
support this point [14,27,43].
In our approach to tissue engineering an oral
mucosal equivalent we use a serum-free culture
system without a feeder layer. We also have success-
fully eliminated the use of bovine pituitary extract
in the medium, thus having a completely defined
K. Izumi, S.E. Feinberg / Oral Maxillofacial Surg Clin N Am 14 (2002) 61–71 65

6. K. Izumi, S.E. Feinberg / Oral Maxillofacial Surg Clin N Am 14 (2002) 61–7166

7. culture medium for the manufacture of their ex
vivo produced oral mucosal equivalent (EVPOME)
[24,25]. The authors’ EVPOME is composed of
autogenous oral keratinocytes and a cadaver acellu-
lar, AlloDerm (Fig. 1 A–C). Electron microscopic
evaluation of the EVPOME shows that the AlloDerm
retains an intact basement membrane and anchoring
fibrils on the papillary surface [29]. After being
cultured 4 days submerged, the authors’ EVPOME
shows several layers of keratinocytes adherent to
one another via desmosomal attachments (Fig. 2),
whereas specific junctional structures between basal
cells and the basement membrane of the AlloDerm
were not seen at that time (Fig. 3). At day 11
EVPOMEs, cultured 4 days submerged and 7 days
at an air-liquid interface, numerous rudimentary
hemidesmosome-like structures were seen incorpo-
rated into anchoring fibrils of the basement mem-
brane of the AlloDerm (Fig. 4). This finding seems to
indicate that the basal cell layer was attached firmly
to the underlying dermal equivalent of the day 11
EVPOME, suggesting an ability of the epithelial
layer to withstanding shear stress.
From a 4 Â 4 mm2
punch biopsy of the palate it
would take approximately 40 days to fabricate an
EVPOME the size of one US dollar bill. This size
EVPOME should be large enough to cover most
mucosal defects. Approved human clinical trials
were initiated in the Fall of 2000 at the Dental
School Hospital of Niigata University, Niigata City,
Japan. Our group at the University of Michigan
also is in the process of obtaining FDA approval
for a tissue-engineered oral mucosa for use in human
clinical trials.
The clinical protocol that was used for the first
patients in the study performed at Niigata University
in Japan was first to take a 5 Â 5 mm2
punch biopsy
of the retromolar trigonal mucosa in an outpatient
setting under local anesthesia. The biopsy is planned
sufficiently before the surgical procedure to ensure
that an adequate piece of EVPOME is available for
grafting. In most cases, to date, a period of 4 weeks
has been sufficient. Oral keratinocytes are dissociated
from the biopsy and expanded in a standard, serum-
free defined culture medium. Once a sufficient num-
ber of oral keratinocytes has been harvested, 1.25 Â
Fig. 2. Transmission electron micrograph of keratinocytes in D4E. Numerous desmosomes (arrows) are formed between
keratinocytes, while abundant tonofibrils are seen in the cytoplasm of the keratinocytes (osmium tetroxide postfixation and
uranyl acetate/lead citrate, original magnification Â17.000).
Fig. 1. (A) Ex vivo produced oral mucosa equivalent (EVPOME) cultured 4 days submerged (D4E). Continuous epithelial
monolayer has developed over dermal component, AlloDerm (Life Cell Corporation, Branchberg, NJ; H&E, original
magnification Âl25). (B) EVPOME cultured 4 days submerged and 7 days at an air-liquid interface (D11E). Epithelial layer of
D4E has started to stratify and differentiate. Keratinocytes in superficial layer are flattened and eosinophilic (H&E staining,
original magnification Â250). (C) EVPOME cultured 4 days submerged and 14 days at an air-liquid interface (D18E). An
increase in stratification of the layers is noted that is consistent with a more fully differentiated epithelium. Epithelial layer
demonstrates parakeratinization (H&E, original magnification Â150).
K. Izumi, S.E. Feinberg / Oral Maxillofacial Surg Clin N Am 14 (2002) 61–71 67

8. 105
cells/cm2
are seeded onto the acellular cadaver
dermal equivalent, AlloDerm. The protocol outlined
by Izumi et al. [25] is then followed. Briefly, the
composites of oral keratinocytes and AlloDerm are
cultured submerged for 4 days and at an air-liquid
interface for 7 days to encourage epithelial stratifica-
tion (Fig. 1 B). This protocol was determined to be
optimal through in vivo grafting studies performed in
SCID mice (Izumi et al., Tissue engineering, 2002,
manuscript accepted for publication).
In patients, the EVPOME is produced and trans-
planted on day 11 after the oral keratinocytes have
been seeded onto the AlloDerm. A gauze bolster or
stent is then used to stabilize the EVPOMEs at the
time of surgery. Surgical stents or bolsters are
removed at 6 days postoperatively, and the surface
of the transplanted EVPOME at the time is scraped
with a swab for cytologic examination. The presence
of small, round-shape cells suggests the presence of
basal cell-like characteristics. Transnasal feeding is
Fig. 4. Transmission electron micrograph of dermal-epithelial junction in D11E. Hemidesmosomal-like structures (arrows)
incorporated into anchoring fibrils are well developed. Note anchoring fibrils (arrowheads) newly integrated within the
hemidesomosomal-like structures (osmium tetroxide postfixation and uranyl acetate/lead citrate, original magnification Â30.000).
Fig. 3. Transmission electron micrograph of dermal-epithelial junction in D4E. There are no specific junctional apparatus seen
between the basal cells and basement membrane. Original, retained, anchoring fibrils (arrowheads) in the papillary surface of
AlloDerm (osmium tetroxide postfixation and uranyl acetate/lead citrate, original magnification Â17.000).
K. Izumi, S.E. Feinberg / Oral Maxillofacial Surg Clin N Am 14 (2002) 61–7168

9. used until removal of the stitches to minimize disrup-
tion of the grafted EVPOME. A soft diet is begun at
postoperative day 8 at the time of removal of the
transnasal feeding tube. At 4 weeks postoperatively, a
punch biopsy is performed for histologic examination.
In some cases, AlloDerm without autogenous ker-
atinocytes as a control also has been transplanted onto
oral mucosal defects. In contrast to the EVPOME,
the AlloDerm graft without an epithelium showed
more shrinkage over time postoperatively, which
resulted in a greater degree of wound contraction.
The AlloDerm graft without an epithelium caused an
indurated wound, which could impair soft tissue
mobility. On histopathologic examination of 4 weeks
after surgery, the epithelial layers of the EVPOME
and AlloDerm without epithelium demonstrated a
regenerative, well-stratified epithelial layer. The pres-
ence of endothelial cells was evident as was a marked
vascular ingrowth and cellular infiltration into the
underlying dermal component of the EVPOME and
AlloDerm alone. Because the presence of the forma-
tion of intense granulation tissue may lead to addi-
tional scarring, the grafted AlloDerm without an
epithelium might result in a functional compromise
within the oral cavity. The histopathologic features of
grafted EVPOMEs showed a favorable remodeling
and incorporation within the host tissue during the
healing phase.
Studies are in progress using tissue-engineered
oral mucosa as a vehicle for the use of gene therapy
to enhance wound healing and/or transmucosally
administer systemically needed growth factors.
Summary
To date, successfully developed EVPOMEs in a
serum-free culture system without a feeder layer are
the most acceptable and promising oral mucosal
substitutes for human intraoral grafting because of
minimal risk of foreign contaminants, easy handling
and stitching, subsequent rapid revascularization into
dermal component after transplantation, and contri-
bution to favorable open wound closure without
functional compromise, although several types of oral
mucosal substitutes described in this article have been
used in patients.
Acknowledgment
The authors thank Masaaki Hoshino for his
technical assistance, Dr. Michiko Yoshizawa for her
input and involvement in the development of our
tissue-engineered human oral mucosa, and Dr.
Cynthia Marcelo for many fruitful discussions.
References
[1] Auger FA, Lopez Valle CA, Guignard R, Tremblay N,
Noel B, Goulet F, et al. Skin equivalent produced with
human collagen. In Vitro Cell Dev Biol Anim 1995;
31:432–9.
[2] Bell E, Ehrlich HP, Buttle DJ, Nakatsuji T. Living
tissue formed in vitro and accepted as skin-equivalent
tissue of full thickness. Science 1981;211:1052–4.
[3] Boyce ST, Supp AP, Harriger MD, Pickens WL, Wick-
ett RR, Hoath SB. Surface electrical capacitance as a
noninvasive index of epidermal barrier in cultured bar-
rier in cultured skin substitutes in athymic mice. J In-
vest Dermatol 1996;107:82–7.
[4] Burke JF, Yannas IV, Quinby WC Jr, Bondoc CC, Jung
WK. Successful use of a physiologically acceptable
artificial skin in the treatment of extensive burn injury.
Ann Surg 1981;194:413–28.
[5] Carter DM, Lin AN, Varghese MC, Caldwell D, Pratt
LA, Eisinger M. Treatment of junctional epidermolysis
bullosa with epidermal autografts. J Am Acad Derma-
tol 1987;17:246–50.
[6] Cho KH, Ahn HT, Park KC, Chung JH, Kim SW, Sung
MW, et al. Reconstruction of human hard-palate mu-
cosal epithelium on de-epidermized dermis. J Dermatol
Sci 2000;22:117–24.
[7] Chung JH, Cho KH, Lee DY, Kwon OS, Sung MW,
Kim KH, et al. Human oral buccal mucosa recon-
structed on dermal substrates: a model for oral epi-
thelial differentiation. Arch Dermatol Res 1997;289:
677–85.
[8] Clugston PA, Snelling CF, MacDonald IB, Maledy HL,
Boyle JC, Germann E, et al. Cultured epithelial auto-
grafts: three years of clinical experience with eighteen
patients. J Burn Care Rehabil 1991;12:533–9.
[9] Compton CC. Acceleration of skin regeneration from
cultured epithelial autografts by transplantation to
homograft dermis. J Burn Care Rehabil 1993;14:
653–62.
[10] Cuono CB, Langdon R, Birchall N, Barttelbort S,
McGuire J. Composite autologous allogeneic skin re-
placement: development and clinical application. Plast
Reconstr Surg 1987;80:625–35.
[11] De Luca M, Albanese E, Megna M, Cancedda R, Man-
giante PE, Cadoni A, et al. Evidence that human oral
epithelium reconstituted in vitro and transplanted onto
patients with defects in the oral mucosa retains proper-
ties of the original donor site. Transplantation 1990;50:
454–9.
[12] Dellon AL, Tarpley TM, Chretien P. Histologic evalu-
ation of intraoral skin grafts and pedicled flaps in hu-
mans. J Oral Surg 1976;34:789–94.
[13] Donoff RB. Biological basis for vestibuloplasty proce-
dures. J Oral Surg 1976;34:890–6.
K. Izumi, S.E. Feinberg / Oral Maxillofacial Surg Clin N Am 14 (2002) 61–71 69

10. [14] Donoff RB. Ex vivo developlment of a composite hu-
man oral mucosal equivalent [discussion]. J Oral Max-
illofac Surg 1999;57:577–8.
[15] Edmonds ME, Foster AV, McColgan M. Dermagraft: a
new treatment for diabetic foot ulcers. Diabet Med
1997;14:1010–1.
[16] Eldad A, Burt A, Clarke JA. Cultured epithelium as
skin substitutes. Burns 1987;13:173–80.
[17] Falanga V, Margolis D, Alvarez O, Auletta M, Mag-
giacomo F, Altman M, et al. Rapid healing of venous
ulcers and lack of clinical rejection with an allogeneic
cultured human skin equivalent. Arch Dermatol 1998;
134:293–300.
[18] Gallico GG, O’Connor NE. Engineering a skin re-
placement. Tissue Eng 1995;1:231–40.
[19] Gallico GG, O’Connor NE, Compton CC, Kehinde O,
Green H. Permanent coverage of large burn wounds
with autologous cultured human epithelium. N Engl J
Med 1984;311:448–51.
[20] Gallico GG, O’Connor NE, Compton CC, Remens-
nyder JP, Kehinde O, Green H. Cultured epithelial
autografts for giant congenital nevi. Plast Reconstr
Surg 1989;84:1–9.
[21] Hata K, Kagami H, Ueda M, Torii S, Matsuyama M.
The characteristics of cultured mucosal cell sheet as a
material for grafting: comparison with cultured epider-
mal cell sheet. Ann Plast Surg 1995;34:530–8.
[22] Heimbach D, Luterman A, Burke J, Cram A, Herndon
D, Hunt J, et al. Artificial dermis for major burns: a
multi-center randomized clinical trial. Ann Surg 1998;
208:313–20.
[23] Inokuchi S, Shimamura K, Tohya H, Kidokoro M,
Tanaka M, Ueyama Y, et al. Effects of fibroblasts of
different origin on long-term maintenance of xeno-
transplanted human epidermal keratinocytes in immu-
nodeficient mice. Cell Tissue Res 1995;281:223–9.
[24] Izumi K, Takacs G, Terashi H, Feinberg SE. Ex vivo
development of a composite human oral mucosal
equivalent. J Oral Maxillofac Surg 1999;57:571–7.
[25] Izumi K, Terashi H, Marcelo CL, Feinberg SE. Devel-
opment and characterization of a tissue-engineered hu-
man oral mucosa equivalent produced in a serum-free
culture system. J Dent Res 2000;79:798–805.
[26] Kangesu T, Navsaria HA, Manek S, Fryer PR, Leigh
IM, Green CJ. Kerato-dermal grafts: the importance of
dermis for the in vivo growth of cultured keratinocytes.
Br J Plast Surg 1993;46:401–9.
[27] Lauer G. Use of cultured mucosal grafts to cover de-
fects cause by vestibuloplasty: an in vitro study [dis-
cussion]. J Oral Maxillofac Surg 1995;53:878–9.
[28] Lindberg K, Rheinwald JG. Three distinct keratino-
cytes subtypes identified in human oral epithelium by
their patterns of keratin expression in culture and in
xenografts. Differentiation 1990;45:230–41.
[29] Livesey SA, Herndon DN, Hollyoak MA, Atkinson
YH, Nag A. Transplanted acellular allograft dermal
matrix. Transplantation 1995;60:1–9.
[30] Lorenz HP, Hedrick MH, Chang J, Mehrara BJ, Long-
aker MT. The impact of biomolecular medicine and
tissue engineering on plastic surgery in the 21st cen-
tury. Plast Reconstr Surg 2000;105:2467–81.
[31] Maloney PL, Shepherd N, Doku C, Murnane T. Free
buccal mucosa grafts for vestibuloplasty. J Oral Surg
1972;30:716–21.
[32] Marks MG, Doillon C, Silver FH. Effects of fibroblasts
and basic fibroblast growth factor on facilitation of
dermal wound healing by type I collagen matrices.
J Biomed Mater Res 1991;25:683–96.
[33] Martin G, Bene MC, Mole N, Penaud J, Faure GC,
Miller NA. Synthetic extracellular matrix supports
healing of mucogingival surgery donor sites. Tissue
Eng 1995;1:279–88.
[34] Martis C. Mucosa versus skin grafts. In: Stoelinga PW,
editor. Proceedings consensus conference: the relative
roles of vestibuloplasty and ridge augmentation in the
management of atrophic mandible. Chicago: Quintes-
sence; 1984. p. 41–44.
[35] Medalie DA, Eming SA, Collins ME, Tompkins RG,
Yarmush ML, Morgan JR. Differences in dermal ana-
logs influence subsequent pigmentation, epidermal dif-
ferentiation, basement membrane, and rete ridge
formation of transplanted composite skin grafts. Trans-
plantation 1997;64:454–65.
[36] Moller JF, Jolst O. A histologic follow-up study of free
autogenous skin grafts to the alveolar ridge in humans.
Int J Oral Surg 1972;1:283–9.
[37] Munster AM, Weiner SH, Spence RJ. Cultured epider-
mis for the coverage of massive burn wounds. Ann
Surg 1990;211:676–80.
[38] Naughton G, Mansbridge J, Gentzkow G. A metaboli-
cally active human dermal replacement for the treat-
ment of diabetic foot ulcers. Artif Organs 1997;21:
1203–10.
[39] O’Connor NE, Mulliken JB, Banks-Schlegel S, Ke-
hinde O, Green H. Grafting of burns with cultured
epithelium prepared from autologous epidermal cells.
Lancet 1981;1:75–8.
[40] Oda D, Savard CE, Eng L, Sekijima J, Haigh G, Lee
SP. Reconstituted human oral and esophageal mucosa
in culture. In Vitro Cell Dev Biol Anim 1998;34:
46–52.
[41] Odioso LL, Doyle MJ, Quinn KW, Bartel RL, Zimber
MP, Stevens-Burns D. Development and characteriza-
tion of an in vitro gingival epithelial model. J Perio-
dontal Res 1995;30:210–9.
[42] Omura S, Mizuki N, Horimoto S, Kawabe R, Fujita K.
A newly developed collagen/silicone bilayer membrane
as a mucosal substitute: a preliminary report. Br J
Oral Maxillofac Surg 1997;35:85–91.
[43] Parenteau NL, Nolte CM, Bilbo P, Rosenberg M, Wil-
kins LM, Johnson EW, et al. Epidermis generated in
vitro: practical considerations and applications. J Cell
Biochem 1991;45:245–51.
[44] Phillips T, Bhawan J, Leigh IM, Baum HJ, Gilchrest
BA. Cultured epidermal allografts: a study of differ-
entiation and allograft survival. J Am Acad Dermatol
1990;23:189–98.
[45] Raghoebar GM, Tompson AM, Scholma J, Blaauw EH,
K. Izumi, S.E. Feinberg / Oral Maxillofacial Surg Clin N Am 14 (2002) 61–7170

11. Witjes MJ, Vissink A. Use of cultured mucosal grafts to
cover defects cause by vestibuloplasty: an in vitro study.
J Oral Maxillofac Surg 1995;53:872–8.
[46] Regnier M, Schweizer J, Michel S, Bailly C, Prunieras
M. Expression of high molecular weight (67K) keratin
in human keratinocytes cultured on dead de-epider-
mized dermis. Exp Cell Res 1986;165:63–72.
[47] Rheinwald JG, Green H. Serial cultivation of strains
of human epidermal keratinocytes: the formation of
keratinizing colonies from single cells. Cell 1975;6:
331–44.
[48] Rue LW III, Cioffi WG, McManus WF, Pruitt BA Jr.
Wound closure and outcome in extensively burned pa-
tients treated with cultured autologous keratinocytes.
J Trauma 1993;34:662–8.
[49] Sheridan RL, Hegarty M, Tompkins RG, Burke JF.
Artificial skin in massive burns: results to ten years.
European Journal of Plastic Surgery 1994;17:91–3.
[50] Smola H, Reinke M, Shephard P, Krieg T, Hess S. The
use of autologous patient serum for the culture of ker-
atinocyte transplants can reduce the risk of a transmit-
table disease. Lancet 1999;353:641–2.
[51] Teepe RG, Koebrugge EJ, Ponec M, Vermeer BJ.
Fresh versus cryopreserved allografts for the treatment
of chronic skin ulcers. Br J Dermatol 1990;122:81–9.
[52] Tsai C, Ueda M, Hata K, Horie K, Hibino Y, Sugi-
mura Y, et al. Clinical results of cultured epithelial
cell grafting in the oral and maxillofacial region. J
Craniomaxillofac Surg 1997;25:4–8.
[53] Ueda M, Hata K, Horie K, Torii S. The potential of oral
mucosal cells for cultured epithelium: a preliminary
report. Ann Plast Surg 1995;35:498–504.
[54] Ueda M, Hata K, Sumi Y, Mizuno H, Niimi A. Peri-
implant soft tissue management through use of cul-
tured mucsal epithelium. Oral Surg Oral Med Oral
Pathol Oral Radiol Endod 1998;86:393–400.
[55] Yannas IV, Hansbrough JF, Ehrlich HP. What criteria
should be used for designing artificial skin replace-
ments and how well do the current grafting materials
meet these criteria? J Trauma 1984;24:S29–39.
K. Izumi, S.E. Feinberg / Oral Maxillofacial Surg Clin N Am 14 (2002) 61–71 71