CH02CH19-Yarmush

ARI

12 May 2011

ANNUAL
REVIEWS

15:43

Further

Annu. Rev. Chem. Biomol. Eng. 2011.2:403-430. Downloaded from www.annualreviews.org
Access provided by Imperial College London on 01/18/16. For personal use only.

Click here for quick links to
Annual Reviews content online, including: • Other articles in this volume
• Top cited articles
• Top downloaded articles
• Our comprehensive search

Tissue Engineering and
Regenerative Medicine:
History, Progress, and Challenges
Francois Berthiaume,1 Timothy J. Maguire,1
¸
and Martin L. Yarmush1,2
1
Department of Biomedical Engineering, Rutgers, The State University of New Jersey,
Piscataway, New Jersey 08854; email: ireis@sbi.org
2
Center for Engineering in Medicine, Massachusetts General Hospital, Boston,
Massachusetts 02114

Annu. Rev. Chem. Biomol. Eng. 2011. 2:403–30

Keywords

First published online as a Review in Advance on
March 17, 2011

artificial organs, skin, cartilage, liver, stem cells

The Annual Review of Chemical and Biomolecular
Engineering is online at chembioeng.annualreviews.org

Abstract

This article’s doi:
10.1146/annurev-chembioeng-061010-114257
Copyright c 2011 by Annual Reviews.
All rights reserved
1947-5438/11/0715-0403$20.00

The past three decades have seen the emergence of an endeavor called tissue engineering and regenerative medicine in which scientists, engineers, and physicians apply tools from a variety of fields to construct biological substitutes that can mimic tissues for diagnostic and research purposes and can replace (or help regenerate) diseased and injured tissues. A significant portion of this effort has been translated to actual therapies, especially in the areas of skin replacement and, to a lesser extent, cartilage repair. A good amount of thoughtful work has also yielded prototypes of other tissue substitutes such as nerve conduits, blood vessels, liver, and even heart. Forward movement to clinical product, however, has been slow. Another offshoot of these efforts has been the incorporation of some new exciting technologies
(e.g., microfabrication, 3D printing) that may enable future breakthroughs.
In this review we highlight the modest beginnings of the field and then describe three application examples that are in various stages of development, ranging from relatively mature (skin) to ongoing proof-of-concept (cartilage) to early stage (liver). We then discuss some of the major issues that limit the development of complex tissues, some of which are fundamentalsbased, whereas others stem from the needs of the end users.

403

CH02CH19-Yarmush

ARI

12 May 2011

15:43

Annu. Rev. Chem. Biomol. Eng. 2011.2:403-430. Downloaded from www.annualreviews.org
Access provided by Imperial College London on 01/18/16. For personal use only.

A BRIEF HISTORY OF TISSUE ENGINEERING
Tissue engineering is a relatively new field that uses living cells, biocompatible materials, and suitable biochemical (e.g., growth factors) and physical (e.g., cyclic mechanical loading) factors, as well as combinations thereof, to create tissue-like structures. Most frequently, the ultimate goal is implantation of these tissue constructs into the body to repair an injury or replace the function of a failing organ. The critical functions may be structural (e.g., bone, cartilage), barrier- and transport-related (e.g., skin, blood vessels), or biochemical and secretory (e.g., liver and pancreas).
Tissue engineering also applies to the development of specialized extracorporeal life support systems containing cells (e.g., bioartificial liver and kidney) as well as tissue units that may be used for diagnostic screening. In addition to clinical applications, other uses include drug testing for efficacy and toxicology as well as basic studies on tissue development and morphogenesis. The term regenerative medicine is often used synonymously with tissue engineering, although regenerative medicine often implies the use of stem cells as a cell source.
Some historical highlights related to tissue engineering and regenerative medicine are shown in Table 1. The first tissue-based therapies developed were skin grafting techniques. Then came techniques to preserve cells and tissues that enabled allograft skin banking, making these skin grafts an off-the-shelf product. The first synthetic skin substitute reportedly used by more than one investigator was developed in 1962; however, the first successful tissue-engineered skin products were made in the late 1970s and early 1980s. Most would agree that this is when modern tissue engineering really started, although the term “tissue engineering” was apparently coined later, around 1987.
Among the first tissue-engineered skin constructs was the product developed by Howard Green and colleagues (1–3) at Harvard Medical School, who described techniques to grow skin epidermis starting with a skin biopsy harvested from a patient. Keratinocytes isolated from the biopsy could be proliferated by coculturing with a feeder layer of mouse mesenchymal cells, thus expanding

Table 1 Some historical landmarks in tissue engineering
Year

Technology/accomplishment

Reference

3000 BCE

Skin grafting described in Sanskrit texts of India

1794

Autologous skin grafting in Europe by Bunger, Reverdin, and Baronio

(25)

1881

Cadaveric skin allograft by Girdner

(25)

1944

Refrigerated skin allografts by Webster

(129)

1949

Cell cryopreservation at subzero temperatures developed by Polge

(130)

1952

Skin cryopreservation developed by Billingham

(131)

1962

Ivalon sponge developed as “synthetic substitute for skin” by Chardack

(132)

1975

In vitro cultivation of keratinocytes by Rheinwald and Green

(1)

1979

Cultured autologous epithelium, later commercialized as Epicel by Genzyme

(2)

1981

Composite living skin equivalent by Bell, later commercialized as Apligraf by Organogenesis

(6)

1982

Collagen-glycosaminoglycans (GAG)-based dermal matrix by Yannas, later commercialized as Dermal
Regeneration Template by Integra Lifesciences

(5)

1987

“Tissue engineering” term coined

(133)

1988

Cell transplantation in synthetic biodegradable polymers

(134)

1994

Chondrocyte culture and transplantation by Brittberg, later commercialized as Carticel by Genzyme

(54)

2006

Bioartificial bladder cultured in vitro and implanted in vivo

(135)

2008

Engineered trachea from decellularized matrix seeded with human cells derived from stem cells

(136)

404

Berthiaume

·

Maguire

·

Yarmush

(25)

Annu. Rev. Chem. Biomol. Eng. 2011.2:403-430. Downloaded from www.annualreviews.org
Access provided by Imperial College London on 01/18/16. For personal use only.

CH02CH19-Yarmush

ARI

12 May 2011

15:43

the coverage area several thousand-fold within weeks. This technological breakthrough led to the first cell-based tissue-engineered product, Epicel, which was marketed by Genzyme (Cambridge,
MA). Epicel consists of sheets of autologous (i.e., derived from the recipient) keratinocytes that are used to cover patients suffering from catastrophic cutaneous burn injuries who do not have enough viable skin remaining to be treated with traditional autografting techniques. The product does not have a dermis and is only a few cells thick; therefore, it is extremely fragile and is not commonly used (only approximately 60–70 patients per year on average). The U.S. Food and
Drug Administration regulates Epicel as a xenogeneic (i.e., derived from another species, in this case nonhuman) product (because it uses a feeder layer of mouse cells), the first of its kind.
Another early product was developed by mechanical engineer Ioannis Yannas at the
Massachusetts Institute of Technology (MIT) in collaboration with burn surgeon John F. Burke at the Boston Shriners Hospital for Children (4, 5) and their colleagues. It consists of a bovine type
I collagen and shark chondroitin 6-sulfate mixture that is cross-linked and turned into a porous matrix by controlled freeze-drying. A silicone sheet attached to one side functions as a temporary epidermis-like barrier. Commercialized under the name Dermal Regeneration Template by
Integra Life Sciences (Plainsboro, NJ), this product is used to cover severe burn wounds where the damage extends deep into the dermis. Under these circumstances, the wound bed may not support a skin graft, or the absence of dermis may lead to extensive contraction and scarring of the healed wound. The matrix is biodegradable and presumably dissolves as the host’s cells—primarily fibroblasts, endothelial cells, and neural cells—migrate into it and deposit their own extracellular matrix (ECM). Ultimately, the matrix disappears and is entirely replaced with a neodermis made of the patient’s own cells and matrix, thus promoting dermal regeneration while inhibiting wound contraction and leading to better function and appearance of the healed wound. At that point, the silicone film is removed, and the wound is covered with a skin graft. Interestingly, the product contains no living cells, and its main purpose is to guide and stimulate the body’s repair and regenerative processes.
Also early on, Eugene Bell at MIT and colleagues (6) developed a composite skin product reconstituting both dermis and epidermis. The dermis is first made by seeding a collagen gel with dermal fibroblasts, which cause the gel to contract and form a neodermis. The keratinocytes are grown on top of the neodermis, initially submerged in culture medium, and then at some point in the manufacturing procedure exposed to the air-liquid interface to induce differentiation and formation of a keratinized layer. The entire process takes approximately 3 weeks and uses allogeneic (i.e., derived from donors of the same species) cells isolated from neonatal human foreskin, which provides the potential for off-the-shelf availability, but with the caveat that the allogeneic skin substitute can provide only temporary coverage, as the patient will eventually reject it. The current product based on this technology, Apligraf, marketed by Organogenesis
(Canton, MA), is used to stimulate the host’s wound healing response in recalcitrant venous leg ulcers and diabetic foot ulcers. Analogous skin constructs are also used for in vitro tests to measure transdermal transport and chemical corrosive properties.
During the 1990s, several of these and other tissue-engineered skin and subsequently cartilage products were successfully commercialized. These early successes fueled much enthusiasm, and many research laboratories embarked on applying tissue engineering to nearly every tissue in the body. Several new companies were spun off with great fanfare and the hope that, as some prominent spokespeople predicted just 15 years ago, tissue engineers would be making complex body parts by now (7). The strategy of simply combining cells and matrix worked for skin and cartilage because these tissues do not require extensive vascularization and other significant tissue processes. Furthermore, technologies to grow and differentiate keratinocytes made it possible to adequately source the needed cells for these products. This was not the case for other tissues. As www.annualreviews.org • Tissue Engineering: Progress and Challenges

405

ARI

12 May 2011

15:43

the same prominent spokespeople recently acknowledged (8), there remain significant hurdles to overcome, such as providing a functional vascular supply, controlling the complex arrangement of different cell types in a 3D tissue, and identifying qualitatively and quantitatively reliable cell sources to make those tissues.
In the early 2000s, the high-tech bubble burst, and weary investors stopped funding high-risk ventures including tissue-engineering companies, which led to a decline in the industry (9). A study conducted in 2004 found that activity in skin, cartilage, and other structural applications declined by more than 50% with a loss of 800 full-time employees (10). The decrease was partially offset by an increase in stem cell firms, which added more than 300 employees. Except for this transient resurgence fueled by the promise of stem cells, financing of startup activity since 2008 has been very limited. Although significant advances have occurred in some areas, such as bladder, cornea, and bronchial tubes, tissues such as blood vessels, heart, and liver—in spite of years of research efforts—are still far from offering clinically acceptable solutions.
During the maturation of tissue engineering over the past three decades, several technologies have been developed based on advances in molecular and cellular biology and micro- and nanosystems engineering. These technologies have been developed largely by basic scientists and engineers, who sometimes have a tendency to oversimplify the problem and do not always recognize the clinical issues. Nevertheless, some of these technologies have led to the development of molecular diagnostics, which as of 2002 comprised an industry market greater than $3 billion, growing at a rate of approximately 25% per year (11, 12). That nontherapeutic applications of tissue engineering are making strides may ultimately help support the development of new tissueengineered therapeutic products, which are much more difficult to produce than enthusiastic advocates originally thought. As we reiterate at the end of this review, ultimately tissue engineers must focus their energy on solving clinical problems to have a real impact.

Annu. Rev. Chem. Biomol. Eng. 2011.2:403-430. Downloaded from www.annualreviews.org
Access provided by Imperial College London on 01/18/16. For personal use only.

CH02CH19-Yarmush

BASIC PRINCIPLES OF TISSUE ENGINEERING
The ability to reconstitute tissue function with therapeutic products at a clinically meaningful scale has a wide spectrum of applications. The main targets are those tissues that are prone to injury, disease, and degeneration (Table 2). Corresponding organs that have been the targets of tissue-engineered equivalents are listed in Table 3.
Most tissue engineering utilizes living cells, and supplying enough cells is obviously a critically important issue. Cells are typically derived from (a) donor tissue, which is often in very limited supply, or (b) stem or progenitor cells. Stem cells possess two major properties that make them attractive for deriving large cell quantities: (a) their high proliferative capacity and (b) their pluripotency, or ability to differentiate into cells of multiple lineages. Ethical concerns about the use of human embryonic stem (ES) cells are a significant impediment for industrial adoption, but recent advances in the use of adult stem cells, induced pluripotent stem cells (iPS cells), and stem cells from placental and umbilical sources have in part allowed these other stem cell types to replace
ES cells as feasible sources.
A key need for effective tissue engineering is the cellular environment that allows the cells to function as they do in the native tissue. Often the environment mimics some critical aspects of the in vivo setting through proper control of the materials and mechanical setting as well as the chemical milieu. Cell scaffolds usually serve at least one of the following purposes:
1.
2.
3.
4.
406

cell attachment and perhaps migration; retention and presentation of biochemical factors; porous environment for adequate diffusion of nutrients, expressed products, and waste; and mechanical rigidity or flexibility.

Berthiaume

·

Maguire

·

Yarmush

CH02CH19-Yarmush

ARI

12 May 2011

15:43

Table 2 Incidence of injuries and diseases in the United States
Indications

Procedures or patients

Reference

2,000,000 total

(137)

144,000 total

(138)

2,500,000 total

(139)

259,000 total

(140)

Skin
Burns
Pressure sores
Venous stasis ulcers
Nervous system
Spinal cord injury

5,300,000 total

(141)

Eye surgery

5,500,000/year

(141)

Ear surgery

Annu. Rev. Chem. Biomol. Eng. 2011.2:403-430. Downloaded from www.annualreviews.org
Access provided by Imperial College London on 01/18/16. For personal use only.

Alzheimer’s disease

900,000/year

(141)

Musculoskeletal
Joint replacement (knee)

326,000/year

(142)

Joint replacement (hip)

165,000/year

(142)

Bone graft
Musculoskeletal (other)

500,000/year

(143)

6,300,000/year

(141)

26,800,000 total

(144)

1,500,000/year

(141)

Cardiovascular
Heart disease
Respiratory system surgeries
Liver
Liver cirrhosis
Liver cancer
Hepatitis C

400,000 total

(145)

16,260/year

(145)

3,200,000 total

(144)

24,000,000 total

(144)

Pancreas
Diabetes
Digestive system surgeries
Urinary system surgeries

11,000,000/year

(141)

2,500,000/year

(141)

Many of the synthetic biomaterials that have been used in tissue engineering, notably collagenbased materials and the polylactic, polyglycolic, and polycaprolactone family of polymers, were already well known in the medical community, having already been employed as bioresorbable sutures. These materials were attractive initially because they already had regulatory approval, but they were far from optimal for many tissue engineering purposes, particularly because the hydrolytic biodegradation process releases acid, which can be toxic to cells. Other synthetic materials have been engineered with customizable properties such as injectability, transparency, and optimal porosity and resorption rates. One such biomaterial is PuraMatrix (3DM, Cambridge,
MA), which consists of small (16 amino acids long) oligopeptide fragments that self-assemble into nanofibers on a scale similar to the in vivo ECM (13).
Natural scaffolds that use existing ECM materials are still extensively used, including proteinbased materials (e.g., collagen, fibrin) and polysaccharide-based materials (e.g., chitosan, alginate, glycosaminoglycans, hyaluronic acid) (14–16). Cross-linking agents (e.g., glutaraldehyde, watersoluble carbodiimide) may be used with these and other materials to reduce degradation rates.
Although biocompatibility with natural materials is obviously excellent, there remain issues with potential immunogenicity in some cases. Recently, there has been heightened interest in using decellularized tissue matrices obtained from processing discarded donor tissue. This approach to generating a tissue engineering scaffold has had some recent successes with the world’s first whole www.annualreviews.org • Tissue Engineering: Progress and Challenges

407

CH02CH19-Yarmush

ARI

12 May 2011

15:43

Table 3 Overview of tissue engineering thrusts
Challenges

Ref.

Skin

Barrier for the body

Matrix implanted to guide regeneration; implants with autologous or allogeneic cells

Lack of appendages, slow process for growing cells, slow vascularization (27, 146, 147)

Cornea

Transparent barrier for the eye

Matrix implants; extracellular matrix generated by cells cultured ex vivo

Maintain transparency and barrier properties of the matrix

(148–150)

Liver

Annu. Rev. Chem. Biomol. Eng. 2011.2:403-430. Downloaded from www.annualreviews.org
Access provided by Imperial College London on 01/18/16. For personal use only.

Tissue

Function

Approach

Detoxification, production of liver-specific proteins

Hepatocytes from xenogeneic, allogeneic or stem cell-derived sources, or immortalized hepatoma seeded in implantable matrices, extracorporeal bioreactor systems

Cell source, maintenance of hepatic function, high cell density, vascularization of implants (151–153)

Pancreas

Secrete insulin to maintain glucose homeostasis Free or encapsulated islet transplantation Choice of transplantation site, vascularization, cell source, immune rejection

(154–156)

Cartilage

Critical component of joints Matrix implanted to guide regeneration; implants with autologous or allogeneic cells

Slow process for growing cells, control of cell differentiation, host integration, long-term durability (157–160)

Heart

Provides blood circulation Materials, including decellularized organs, seeded with progenitor and stem cells differentiated into cardiomyocytes Tumorigenicity, control of cell differentiation, electrical integration (161–163)

Kidney

Regulates body fluid volume and pH, metabolite excretion

Stem cell–derived nephrons cultured ex vivo

Replicating glomerular selectivity while retaining high hydraulic permeability

(164–166)

Neurons and spinal cord

Send electrical stimuli to control bodily functions Materials shaped as tubes for axonal guidance and regeneration, sometimes used in combination with anti-inflammatory strategies; neural stem cells

Reconnecting proper axons, controlling proinflammatory environment, preventing scar tissue formation

(167–169)

tissue-engineered organ transplant of an engineered trachea (17) and rapid recent advancements with heart, liver, and lung tissues (18–21).
Another consideration with tissue-engineered constructs is the presence of exogenous chemical and mechanical stimuli such as soluble growth and differentiation factors as well as mechanical forces (e.g., cyclic mechanical loading, fluid shear). Among the chemical factors that frequently have been applied are bone morphogenetic proteins (BMPs), basic fibroblast growth factor
(bFGF or FGF-2), vascular endothelial growth factor (VEGF), and transforming growth factor-β
(TGF-β). Although these are chiefly soluble factors, they can be incorporated into the ECM during scaffold fabrication. In fact, one of the key nonstructural functions of the natural ECM in vivo is to bind, retain, and present growth factors to cells attached to the ECM. Controlled delivery schemes can also be used to increase the longevity of the original soluble factor load.
Applied techniques include encapsulation in small biodegradable particles, use of transfected cells to express and release the factors, and chemical conjugation to the scaffold material itself.
408

Berthiaume

·

Maguire

·

Yarmush

Annu. Rev. Chem. Biomol. Eng. 2011.2:403-430. Downloaded from www.annualreviews.org
Access provided by Imperial College London on 01/18/16. For personal use only.

CH02CH19-Yarmush

ARI

12 May 2011

15:43

With the cells, scaffolds, and environmental needs identified, one must consider adequate means for assembly. Early work in this area, which just randomly mixed all the components together before implantation, was fraught with major failures. Many a study was published claiming success, only to have succeeded in implanting cells that eventually died. One problem with this approach is mass transport limitations. Tissues engineered in this manner lacked a preexisting vascular network, thus making it difficult for implanted cells to obtain sufficient oxygen and nutrients to survive and/or function properly. Potential solutions have been offered in which scaffolds are engineered to promote rapid vascular ingrowth or vascular endothelial cells are introduced and allowed to form a vascular network prior to, or concomitant with, the seeding of tissue cells. With the advent of technologies such as soft lithography, robotic printing, and laser tweezers, some investigators have attempted to “print” tissues and even organs. Recent methods include an assembly that uses an ink-jet mechanism to print precise layers of cells in a matrix of thermoreversible gel (22, 23). For example, endothelial cells, which line blood vessels, have been printed in a set of stacked rings. When incubated, these fused into a tube. Through this controlled integration approach, it may be possible to generate an emergent vascular network.

REPRESENTATIVE EXAMPLES OF ENGINEERED TISSUES
In this section, we describe in greater detail three applications for which engineered tissues have been developed. The first example is skin, for which tissue-engineered products were first established and the field is most mature. Second, we look at cartilage, for which some products are available but with limited therapeutic use. Third, we discuss liver tissue engineering, an area that has seen much laboratory investigation, many animal studies, and even clinical trials, but no successful translation to the clinical arena yet.

Skin
All skin wounds that extend deep into the dermis and are more than 1 cm in diameter require specialized treatment, as they cannot close (i.e., regenerate the epithelial lining) on their own and may lead to extensive scarring that may result in joint mobility limitations and severe cosmetic deformities (24). The gold standard for serious cutaneous wounds remains autologous skin grafts, a technique that originated several thousand years ago (25). The limitation for autologous skin grafting is inadequate uninjured donor sites remaining to harvest skin graft material. Although it is possible to extend coverage by meshing the skin (a technique in which the skin graft is uniformly perforated and stretched to cover greater areas of the wound), the lack of dermis in the interstices of the stretched meshed skin graft as well as slow epithelialization from graft margins across interstices results in greater graft contraction and a pronounced crocodile skin appearance of the scar. In general, areas where injuries extend deep into the dermis may not support the skin graft, and/or severe scarring may occur owing to the lack of functional dermis.
Skin substitutes were originally developed to address some of these limitations. In particular, biodegradable matrix materials can emulate the dermis, and keratinocyte and fibroblast culture techniques have led to live cultured skin substitutes.
According to Shakespeare (26), the functions that tissue-engineered skin products can offer are: (a) protection—by establishing a mechanical barrier to microorganisms and vapor loss;
(b) procrastination—by providing some wound cover following early wound debridement until permanent wound closure can be achieved with serial skin grafts or cultured autologous cell applications, especially in extensive burns; (c) promotion—by delivering to the wound bed dermal matrix components, cytokines, and growth factors, which can promote and enhance natural www.annualreviews.org • Tissue Engineering: Progress and Challenges

409

CH02CH19-Yarmush

ARI

12 May 2011

15:43

Table 4 Representative skin substitutes approved by the U.S. Food and Drug Administration
Product

Indications

Comments

Burns and full-thickness injuries

Incorporates into patient’s skin, two-year shelf life

Apligraf

Venous and diabetic ulcers

Allogeneic, cryopreserved

Dermagraft

Diabetic ulcers, epidermolysis bullosa

Allogeneic, cryopreserved

Epicel

Deep partial- and full-thickness burns, congenital nevi

Incorporates into patient’s skin, variable take rate, made to order

Integra Dermal Regeneration
Template

Deep partial- and full-thickness burns

Incorporates into patient’s skin, moderate shelf life

OrCel

Annu. Rev. Chem. Biomol. Eng. 2011.2:403-430. Downloaded from www.annualreviews.org
Access provided by Imperial College London on 01/18/16. For personal use only.

Alloderm

Split-thickness donor sites, epidermolysis bullosa Allogeneic, nine-month shelf life cryopreserved

TransCyte

Deep partial- and full-thickness burns

1.5-year shelf life, cryopreserved

host wound healing responses; and (d ) provision—of new structures, such as dermal collagen or cultured cells, that are incorporated into the wound and persist during wound healing and/or thereafter. Although none of the existing products can fully replace damaged skin (27, 28), they have been used to treat extensive acute wounds (especially burns) as well as to promote healing of chronic nonhealing wounds such as diabetic ulcers and venous ulcers.
A representative listing of engineered skin substitutes that are available on the U.S. market is in Table 4. Figure 1 summarizes the current main approaches to skin tissue engineering. The simplest engineered skin substitutes, which are still in use today, consist of porous matrices that function as templates for dermal regeneration. The matrices are placed on the wound bed and allowed to integrate and vascularize. After sufficient revascularization of the matrix, these products must be covered with autografts (29). Integra Lifesciences’s Dermal Regeneration TemplateTM , which we described earlier, is primarily used for the treatment of deep burn wounds, which are prone to forming undesirable scars. The matrix degrades while the host’s cells invade and proliferate within it, thus promoting dermal regeneration while inhibiting wound contraction, leading to better function and appearance of the healed wound (30). Another skin substitute, Alloderm (LifeCell, Branchburg, NJ), is made from decellularized donor skin. Removing all the cells and keeping only matrix components prevents an allogeneic immunological response and also reduces the risk of disease transmission (31, 32). Alloderm is used for both wound repair and reconstructive surgery.
As with Integra, an autograft must be applied eventually to re-epithelialize the wound. Another tissue-engineered dermal analog consists of allogeneic neonatal dermal fibroblasts cultured in a polyglactin mesh. The cells produce ECM proteins as the mesh degrades, thus producing a matrix usable on the wound (31). This product, called Dermagraft (Advanced BioHealing, Westport,
CT), has been used to cover diabetic foot ulcers. Although Dermagraft is eventually rejected, it appears to help restore the dermis and promote keratinocyte migration to close the wound
(33).
−− − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − −→
−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−
Figure 1
Current approaches to skin tissue engineering. One approach consists of placing a biodegradable matrix in the wound to promote the regeneration of the skin dermis through a process of host cell migration and proliferation (a). Another approach focuses on regenerating the keratinocyte layer by putting on top of the wound cultured autologous keratinocytes or a temporary covering that contains extracellular matrix and growth factors that stimulate keratinocyte proliferation (b). These methods work best on partial-thickness wounds. Cultured autologous keratinocytes are used on full-thickness wounds as well, but the take is poor. The most comprehensive tissue-cultured skin incorporates both living dermis and epidermis, which are usually cultured from allogeneic sources (c).
410

Berthiaume

·

Maguire

·

Yarmush

CH02CH19-Yarmush

a

ARI

12 May 2011

15:43

Dermal regeneration matrix
Silicone barrier

Epidermis
Dermis

Hypodermis

Annu. Rev. Chem. Biomol. Eng. 2011.2:403-430. Downloaded from www.annualreviews.org
Access provided by Imperial College London on 01/18/16. For personal use only.

Porous biodegradable matrix
Full-thickness injury
• Porous biodegradable matrix with no cells
• Host fibroblasts and endothelial cells migrate into matrix
• Promotes regeneration of dermis
• Improves scar appearance

b

Epithelial regeneration

Silicone barrier

Epidermis
Dermis

Hypodermis

Partial-thickness injury

Porous biodegradable matrix
Partial-thickness injury

• Cultured autologous epithelial cells

c

• Temporary matrix/growth factor cover

• Permanent wound coverage

• Promotes host epithelial migration and proliferation

Full-thickness composite skin

Epidermis
Dermis

Hypodermis

Full-thickness injury
• Fibroblast-populated matrix covered with keratinocytes • Allogeneic cells used, so coverage is temporary
• Attempts to develop technology using autologous cells are underway

www.annualreviews.org • Tissue Engineering: Progress and Challenges

411

ARI

12 May 2011

15:43

The first cultured skin began with Epicel, a cultured autologous epidermis described earlier.
Some of the major limitations of this product are the absence of a dermis and its lack of off-the-shelf availability because a patient biopsy is used as starting material. To overcome these limitations, other tissue-engineered skin products were developed to include both epidermis and dermis. The first full-thickness engineered skin product is the earlier-described Apligraf, a bilayered construct using fibroblasts in a collagen gel and keratinocytes to create a dermis and epidermis, respectively. Other analogous products have since been developed, such as OrCel (FortiCell Bioscience,
Englewood Cliffs, NJ), which uses bovine type I collagen sponge as substrate. As is the case for any allogeneic tissue, Apligraf and Orcel ultimately are rejected (34). Similar products based on autologous cultured keratinocytes and fibroblasts may fulfill the role of true skin substitution; they are currently in research and development, and results of clinical trials seem encouraging (35).
At the same time, other products that are simpler and less expensive have also become available.
For example, TransCyte (Advanced BioHealing, Westport, CT) is a nylon mesh with a silicone membrane on one side and cultured foreskin human fibroblasts on the other side. The fibroblasts proliferate within the nylon mesh and deposit ECM as well as growth factors. The product is frozen and then thawed for application. Cells die in this process, but the ECM and growth factors remain essentially intact (36).
A relatively new approach involves distributing a minced micrograft over the wound area. This technique involves excision of a small area (∼2 cm2 ) of full-thickness skin from the patient, which is then minced. The resulting mixture, which contains both the dermal and epidermal components of skin, is combined with a hydrogel and applied to the wound. The distributed cells proliferate and participate in the wound healing process. This clever approach may provide a future alternative to traditional skin grafts, given its need for only a small donor area and its inherent simplicity when compared with full-fledged tissue-engineered products (37).
Even though tissue-engineered skin substitutes have been available for decades and arguably are the best-established tissue-engineered products, their practical role is limited to a specific niche within a complex approach to treating acute and chronic wounds. In most instances, they serve as temporary biologically active wound dressings until the patient’s own skin regenerates and can be used for serial autografting (27). This reality exists because the reported benefits of skin substitutes tend to be modest, and most experts in the field would agree that no existing product can claim to be a complete solution. In general, these tissue replacements only partially address specific functional requirements, and surgeons tend to use different products to achieve different purposes. One of the main limitations of engineered skin substitutes is slow revascularization and in some instances poor take (i.e., attachment to the wound bed). In fact, those systems that contain a dermal component, which is meant to help regenerate the dermis as well as provide a better surface for attachment of the epidermis, take a long time to vascularize and delay wound closure. Thus, the surgeon must balance the pros and cons of using a skin substitute to improve long-term scar appearance and function in the face of increased risk of infection owing to delayed vascularization and wound closure. However, new research is promising in this regard. For example, collagen has been used for some time in the design of skin substitutes and recently has been used to create a model of endothelialized, reconstructed dermis that promotes the spontaneous formation of a human capillary-like network (38).
Tissue-engineered skin also lacks several important structures and cell types, including sebaceous glands and sweat glands as well as melanocytes and dendritic or Langerhans cells (27, 28).
Recently two research teams described the use of bulge cells from hair follicles to regenerate skin appendages. In one case, freshly isolated bulge cells from adult mice, when combined with neonatal dermal cells, formed hair follicles after injection into immunodeficient mice (39). In the

Annu. Rev. Chem. Biomol. Eng. 2011.2:403-430. Downloaded from www.annualreviews.org
Access provided by Imperial College London on 01/18/16. For personal use only.

CH02CH19-Yarmush

412

Berthiaume

·

Maguire

·

Yarmush

CH02CH19-Yarmush

ARI

12 May 2011

15:43

Annu. Rev. Chem. Biomol. Eng. 2011.2:403-430. Downloaded from www.annualreviews.org
Access provided by Imperial College London on 01/18/16. For personal use only.

other case, a mixture of isolated neonatal dermal cells injected with epidermal aggregates (isolated hair follicle epithelial stem cells) into the dermis of nude mice led to hair morphogenesis and gave rise to cycling hair follicles within 8 to 12 days (40). These findings suggest that it may be possible to incorporate complex differentiated structures in a new generation of skin substitutes.
Genetically modified skin substitutes have also been developed. This area has been extensively explored with respect to correcting genetic defects (41, 42), augmenting the supply of deficient hormones (43), and enhancing the efficacy of a tissue-engineered product (44). The first two of these areas have received considerable attention but have not yet generated widely used therapeutic skin products. However, Stratatech (Madison, WI) is currently developing genetically engineered skin products to confer enhanced antibacterial, angiogenic, and antiproteolytic properties.

Cartilage
The demand for engineered and regenerative tissue approaches for cartilage has been growing in the face of the increasing prevalence of degenerative joint diseases (e.g., osteoarthritis) as the general population continues to age and become more overweight. The demand in young and healthy individuals is also high owing to the high incidence of sports injuries, given the limited spontaneous repair following articular cartilage injury (45). It is thought that the lack of vascularization of articular cartilage prevents the onset of an inflammatory response to tissue injury and resultant repair. The low cellularity and proliferative capacity of chondrocytes may also underlie an intrinsic inability to repair, leading to scar tissue of inferior mechanical properties and durability. As will be discussed below, it is noteworthy that most cartilage repair technologies work best when used early after injury and in young, healthy individuals. Evolution of the injury toward a chronic state may create an environment that is hostile to tissue repair and regeneration, and delayed treatment universally results in poorer outcomes. Furthermore, the use of matrix materials to supplement surgical methods and cell transplantation techniques is more widespread than the use of cultured cells. The development of tissue-engineered grafts that are made in the laboratory for eventual implantation is a more recent advance that has undergone very limited clinical testing.
Major issues pertaining to cartilage tissue engineering are depicted in Figure 2.
Current techniques to repair cartilage that are used to treat acute injuries generally fall into three categories: (a) marrow simulation–based techniques, (b) osteochondral transplantation techniques, and (c) cell-based repair techniques.
The most prevalent marrow-based technique is called microfracture; the damaged area is perforated below the subchondral plate, allowing blood to flow and clot in the microfractures.
The blood clot contains a relatively high proportion of marrow-derived mesenchymal stromal cells
(MSCs) with high chondrogenic differentiation potential (46), which subsequently produce a scar tissue more akin to fibrocartilage than true cartilage (47, 48). This technique is a first-line procedure in acute knee injuries for athletes younger than 40 years old. Favorable short-term outcomes have been observed, especially if treatment is applied early after injury; however, evidence suggests that the repair tissue undergoes significant deterioration after ∼2 years, and in general the outcome is highly variable. It has been suggested that the imperfect integration of the scar with surrounding healthy cartilage, together with inferior mechanical properties of the scar itself, may be responsible for the observed long-term deterioration. Further development of this approach—still in an active research stage—involves locally applying growth factors and anti-inflammatory agents. Another improvement of the technique uses biodegradable scaffold materials that are inserted into the microfracture. Clinical studies are ongoing and suggest that the scaffold improves cartilage repair volume, composition, and stability. www.annualreviews.org • Tissue Engineering: Progress and Challenges

413

CH02CH19-Yarmush

ARI

12 May 2011

15:43

Prevention of dedifferentiation
Prevention of erosion

Cell source
Ex vivo expansion

Macrophage
Stratification
IL-6
IL-6

Annu. Rev. Chem. Biomol. Eng. 2011.2:403-430. Downloaded from www.annualreviews.org
Access provided by Imperial College London on 01/18/16. For personal use only.

IL-6

Attachment to subchondral bone and nearby cartilage

IL-6
IL-6

Modulation of
M
inflammation

Figure 2
Critical issues in cartilage tissue engineering. The most advanced tissue-engineered cartilage constructs consist of matrices seeded with chondrocytes, although both have been used separately as well. These implants must be designed so that they firmly attach to the subchondral bone and nearby cartilage and have the appropriate type II–collagen density and orientation near the surface to withstand shear forces in the joint. In addition, stratification of the matrix is important to provide appropriate cushioning and an environment for chondrocyte survival and differentiation. Identification of the best source of cells and the ability to proliferate and differentiate these cells are also critical. Finally, one must take into account that the implant may be put in a hostile proinflammatory environment, and provisions to control the impact on implant performance must be taken. IL-6, interleukin-6.

Osteochondral transplantation techniques involve harvesting cartilage together with subchondral bone from nonweight-bearing regions of the joint and placing them in the weight-bearing area of the damage. Technical challenges are associated with this approach, mainly owing to mismatch in the surface shape (convexity) of the treated joint versus the donor tissue and fixation of the graft to the host tissue. This approach is used for mid-size defects (1–4 cm2 ) and has shown excellent results, although evidence suggests that preexisting joint degeneration decreases favorable outcomes. Acute donor site morbidity is also an issue, and there is some controversy as to whether even a small injury to the donor site could be detrimental to the nearby tissue and increase the risk of osteoarthritis in the long term (49, 50).
This technique is obviously limited by the rather short supply of autologous donor tissue available, and for this reason, surgeons are looking toward allografts and cell-free osteochondral graft substitutes. In the former case, the technique takes advantage of the fact that chondrocytes are immunoprivileged in their surrounding ECM (51). The data show excellent results for the first 5 years but significant loss of viability at 15 years (52). The latter approach obviates the need for tissue and makes possible off-the-shelf products that can be used whenever needed. These implants have been developed recently as substitute grafts for treatment of focal chondral and osteochondral defects and include a bone phase and a cartilage phase, each designed to physically and mechanically match the layers of the adjacent cartilage and subchondral bone. These implants are composites of polylactide-glycolide copolymers, calcium sulfate, polyglycolic acid fibers, and surfactant (TruFit, Smith & Nephew Endoscopy, San Antonio, TX). Injectable materials (BST
Cargel, BioSyntech, Montreal, QC, Canada) have also been described (53). Such implants are replaced with new tissue within 12 months; however, long-term data on performance are still
414

Berthiaume

·

Maguire

·

Yarmush

Annu. Rev. Chem. Biomol. Eng. 2011.2:403-430. Downloaded from www.annualreviews.org
Access provided by Imperial College London on 01/18/16. For personal use only.

CH02CH19-Yarmush

ARI

12 May 2011

15:43

lacking. Besides their use as graft substitutes, these bioresorbable implants can be used successfully to backfill donor sites in osteochondral autograft transfers.
Cell-based cartilage repair techniques were first reported in 1994 by Brittberg and colleagues
(54), who developed the first commercially available cell-based technology, now called Carticel (Genzyme, Cambridge, MA). This breakthrough technology successfully repaired articular cartilage lesions of the human knee by autologous chondrocyte transplantation. In this approach, autologous chondrocytes are harvested from a less weight-bearing area of the joint, extracted from the cartilage explant, and proliferated in culture before implantation on the defect is performed.
This therapeutic approach has been successfully used for full-thickness cartilage lesions in the knee with long-term durability of functional improvement exceeding 10 years (55). Long-term functional results were best in athletes with single lesions, age