Tissue Engineering in Orthopaedics

Abstract

• ➤It is important to carefully select the most appropriate combination of scaffold, signals, and cell types when designing tissue engineering approaches for an orthopaedic pathology.
• ➤Although clinical studies in which the tissue engineering paradigm has been applied in the treatment of orthopaedic diseases are limited in number, examining them can yield important lessons.
• ➤While there is a rapid rate of new discoveries in the basic sciences, substantial regulatory, economic, and clinical issues must be overcome with more consistency to translate a greater number of technologies from the laboratory to the operating room.

The field of orthopaedics has a long history of embracing new technologies. From total joint replacement to the use of recombinant human bone morphogenetic protein-2 (rhBMP-2), orthopaedic surgeons and biomedical engineers have worked together to create novel strategies to repair and/or replace damaged tissues. Tissue engineering, the science of generating living tissues, holds promise for treating human disease. Specific to the field of orthopaedics, tissue engineering strategies are being investigated for a number of challenging musculoskeletal pathologies, including fracture nonunion, osteonecrosis, and osteochondral defects.

While exciting developments are being investigated in the laboratory and in preclinical animal models^1,2, the majority of tissue engineering-based therapies may not be translated into the clinic^3-5. The primary goals of this review were (1) to provide an overview of tissue engineering as relevant to orthopaedic surgery; (2) to review how these principles are being applied to treat orthopaedic diseases today, with examples from current studies; and (3) to discuss future directions and challenges for continuing to translate tissue engineering strategies from the laboratory to the operating room.

Principles of Orthopaedic Tissue Engineering

Fundamentally, the tissue engineering paradigm consists of scaffolds, signals, and cells (Fig. 1). These 3 elements can be combined or used independently to attempt to generate tissues in a limitless number of arrangements. However, with an increasing complexity of design, there are greater challenges to translation⁶. For example, receiving regulatory approval for an acellular scaffold requires substantially less time and fewer resources than does a drug-eluting scaffold that has been pre-seeded with stem cells. In this section, these 3 fundamental elements of orthopaedic tissue engineering are reviewed.

Fig. 1.

The orthopaedic tissue engineering paradigm.

Scaffolds

Tissue engineering scaffolds are cytocompatible biomaterials that cells can adhere to and/or replace with extracellular matrix to produce native tissues. Scaffolds can be as simple as morselized autologous bone or as complex as injectable, thermally responsive synthetic hydrogels capable of mineralizing in situ⁷. On the basis of material composition, scaffolds can be divided into 3 basic classes: metals, ceramics, and polymers (Fig. 2). Scaffolds can be further divided by their source (naturally derived versus synthetically fabricated) and ability to degrade (nonresorbable versus resorbable). Although they are stable, nonresorbable scaffolds and delivery systems cannot be replaced by native tissues and may elicit a chronic foreign-body reaction detrimental to tissue healing. Naturally derived scaffolds (such as those made from collagen, chitosan, and hyaluronan) are generally all resorbable in situ and often already possess adhesion ligands for cellular attachment. However, naturally derived scaffolds typically have a narrow range of available physical properties, such as mechanical strength and degradation rate. Synthetic scaffolds can be tuned to have a wide variety of properties by altering synthesis components and parameters. Not all synthetic scaffolds are biodegradable, and cell adhesion motifs may need to be added in order to promote biocompatibility. Different scaffold materials can also be combined to create composite scaffolds that have novel properties not observed in either material used alone⁸.

Fig. 2.

Overview of the classes of scaffolds.

Many considerations are important in selecting a scaffold. For orthopaedics, mechanical properties and durability are paramount to a successful device. While a ceramic scaffold may have appropriate compressive strength in a femoral nonunion defect, its high stiffness and weak tensile properties would be inappropriate for use in a cartilage defect or repairing a rotator cuff. Another important factor to consider is the compatibility of the rate of scaffold resorption with the rate of native tissue replacement. If a scaffold resorbs too rapidly, it may not be able to support the growth of new tissues. If a scaffold resorbs too slowly, it may not fully integrate with surrounding tissues and may pose risks associated with chronic foreign bodies. The effect of the scaffold on native cells may also be important in choosing a material. Scaffolds may be osteoconductive (i.e., permit the growth of bone, such as the calcium sulfates⁹) or osteoinductive (i.e., actively promote bone growth in a defect that otherwise would not heal, such as demineralized bone matrix⁹). As scaffolds with carefully selected properties can recruit native stem cells and direct their differentiation into the proper tissue of choice, implantation of acellular scaffolds may become the gold standard in tissue engineering-based orthopaedic products¹⁰. Last, depending on the intended application, scaffolds may require gradient-based rather than isotropic properties, such as in the functional repair of osteochondral defects and tendon or ligament-to-bone insertions^11,12.

Signals

In the tissue engineering paradigm, signals are internally or externally derived environmental factors that can influence the regeneration of tissues. As in the case of scaffolds, these signals can be further broken down into subcategories including biological, chemical, mechanical, and electrical cues. In orthopaedics, the most commonly utilized biological signal is rhBMP-2, a potent osteogenic growth factor. Clinical products containing rhBMP-2 have been approved for specific applications by the U.S. Food and Drug Administration (FDA) (Table I)¹³. However, it appears that rhBMP-2 is associated with both a substantially greater risk of adverse events as well as a higher severity of adverse events than originally reported¹⁴, and there remain concerns about the use of growth factors in light of their association with malignancies such as osteosarcoma^15,16. Another common source of biological signals used in tissue engineering strategies is platelet rich plasma (PRP) and its different variants¹⁷. PRP is a combination of blood components that are isolated and concentrated (typically by centrifugation). While use of PRP is appealing because of the ease of autologous obtainment, its actual efficacy in the regeneration of musculoskeletal tissues currently remains uncertain¹⁸. In addition to exogenous delivery, biological factors can also be delivered by cells through genetic engineering techniques, although there is some controversy over safety for in situ approaches^19,20. Finally, co-delivery of combinations of biological cues, such as osteogenic and angiogenic growth factors^21,22, has resulted in enhanced tissue regeneration in animal models but has not been translated to clinical trials, perhaps because of concerns of carcinogenicity associated with the upregulation of angiogenic factors^23,24.

TABLE I.

Approved Indications for rhBMP-2-Based Devices

Application	Disease and/or Procedure	Device
Oral maxillofacial	Sinus augmentations	rhBMP-2 in a resorbable collagen sponge
Oral maxillofacial	Localized alveolar ridge augmentations for defects associated with extraction sockets	rhBMP-2 in a resorbable collagen sponge
Spine	Degenerative disc disease (at 1 level from L2 to S1)	rhBMP-2 in a resorbable collagen sponge within a metallic spinal fusion cage
Trauma	Acute open tibial fractures (stabilized by intramedullary nail fixation)	rhBMP-2 in a resorbable collagen sponge

Chemical cues may promote healing by influencing the inflammatory pathway (such as statins²⁵) or by directly acting on cells (such as site-specific bisphosphonate therapy²⁶). Locally delivered small molecules for bone tissue engineering is a growing area of interest in the regenerative medicine community²⁷. Antibiotics are another example of chemical cues often delivered in conjunction with tissue engineering strategies. In an infected orthopaedic defect, encapsulating antibiotics for local drug delivery into the scaffold may be required to stimulate healing²⁸.

Mechanical cues have long been used to stimulate bone formation, such as in distraction osteogenesis. Recent studies have demonstrated that passive mechanical signals provided by scaffolds (such as substrate stiffness, roughness, and porosity) can influence the differentiation of stem cells toward specific lineages^29,30. Electrical cues have been demonstrated to be important in generating functional skeletal muscle tissue as well as innervation of neotissues^31,32, but have not been explored as thoroughly relative to other cellular signals for orthopaedic tissue engineering applications.

Cells

In order to create living tissues, as well as integrate living engineered tissue with native host tissues, cells must be present. Cells can be recruited into an implanted scaffold by methods such as the release of chemokines³³, attachment of cell ligands to the scaffold³⁴, or osteoconduction or osteoinduction, or scaffolds containing cells can be implanted into a defect. Unlike in other tissue engineering fields, there is little controversy regarding stem cell type in orthopaedic tissue engineering. By far, the most commonly utilized cell type is the mesenchymal stem cell (MSC) (Table II)³⁵. Depending on their environment, MSCs have the ability to differentiate into osteoblasts, chondroblasts, myoblasts, and tenocytes as well as other adult cells³⁶. Compared with other stem cell populations, such as neural progenitor cells, MSCs are relatively easy to harvest from an autologous host. However, the optimal source for MSCs is controversial at this time. The current gold standard is MSCs harvested from bone marrow aspirate; however, adipose-derived MSCs are gaining more traction in the field because of their increased availability, lower harvesting costs, and ease of expansion³⁷. Amniotic fluid-derived MSCs are another intriguing source that has recently been shown to be capable of osteogenesis and chondrogenesis in small animal models^38,39. Other sources of MSCs include skin⁴⁰, periosteum⁴¹, and umbilical cord blood⁴². However, ease of collection should not be the only consideration in MSC harvesting. MSCs from different sources have different potential for differentiation^36,43,44 and may affect the quality of tissue repair⁴⁵.

TABLE II.

Criteria for Defining a Cell Population as MSCs According to the International Society for Cellular Therapy^35*

Criteria	Definition
Culture conditions	MSCs must be expandable as well as plastic-adherent under normal culture conditions
Trilineage potential	MSCs must be able to differentiate into osteoblasts, chondrocytes, and adipocytes in vitro as demonstrated by histological staining
Positive surface markers	MSCs must possess CD73, CD90, and CD105 surface markers (≥95% positive)
Negative surface markers	MSCs must lack CD14 or CD11b, CD19 or CD79α, CD34, CD45, and HLA-DR surface markers (<2% positive)

^*Cell populations not meeting these rigorous criteria but capable of differentiating into musculoskeletal cells should be referred to as “connective tissue progenitor” cells. HLA = human leukocyte antigen.

Initially, it was believed that the primary mechanism of action of implanted MSCs within an orthopaedic defect was structural. It was assumed that the implanted MSCs themselves would proliferate, differentiate, and generate the extracellular matrix required to repair the defect. However, recent studies have revealed the tremendous pleiotropic effects of MSCs. Beyond their ability to terminally differentiate into adult musculoskeletal cells, MSCs secrete a variety of cytokines and modulate inflammatory and immune response pathways⁴⁶. Because of these immunomodulatory effects, implantation of allogeneic MSCs carries minimal risk of rejection by the host and commercially available MSCs are being explored in a number of clinical trials for a multitude of autoimmune diseases^46-48. Despite all of these positive attributes, there is still concern over potential long-term side effects from MSC transplantation and the lack of comparability in demonstrating efficacy between clinical trials⁴⁹. Patients enrolled in clinical trials are carefully monitored; a recent report has demonstrated that no infection or tumor had developed in patients treated with MSCs for osteoarthritis at 11 years of follow-up⁵⁰.

The need to implant living cells into an orthopaedic defect for repair is situational. From a product development and regulatory standpoint, acellular strategies present advantages over cell-containing products¹⁰, which carry a potential risk of rejection or disease transmission, have heterogeneous cell populations, etc. For example, it is clear that rhBMP-2 combined with a collagen scaffold has ample regenerative capacity for spinal fusion without the addition of any exogenous MSCs⁵¹. However, for complex diseases such as osteonecrosis, the pleiotropic and immunomodulatory capacities of MSCs may be required for treatment⁵².

Orthopaedic Tissue Engineering Elements Applied in Recent Clinical Experiences

While it is outside the scope of this work to review all of the exciting developments in the field, a series of recent clinical studies (2011 to the present) have been chosen arbitrarily to highlight how the previously discussed elements of tissue engineering have been applied to treat orthopaedic diseases^53-63. Examples were chosen from the following areas: fracture nonunion, osteonecrosis, and chondral and osteochondral defects. For easy reference, these examples have been compiled in Table III. We recommend several excellent recent reviews on these topics by Panteli et al., Mont et al., and Nicolini et al. for a more complete discussion on disease pathophysiology^64-66.

TABLE III.

Recent Examples of Orthopaedic Tissue Engineering Strategies Applied in Human Disease (2011-Present) and Corresponding Elements of the Tissue Engineering Paradigm

Study	No. of Patients	Elements	Summary
Nonunion
Kuroda et al.53 (2014)	7	Scaffold and cells	Bone marrow-derived hematopoietic stem cells were delivered on atelocollagen in femoral and tibial nonunion defects
Calori et al.54 (2013)	52	Scaffold, signals, and cells	Retrospective review of patients treated at a single center; patients were divided into monotherapy (treated with scaffold, cells, or signals) or polytherapy (treated with all 3)
Desai et al.55 (2015)	49	Scaffold, signals, and cells	Bone marrow aspirate MSCs were delivered with demineralized bone matrix or rhBMP-2
Osteonecrosis
Hwang et al.56 (2011)	43	Cells	Autologous bone marrow-derived MSCs implanted alone
Aarvold et al.57 (2013)	4	Scaffold and cells	Autologous bone marrow-derived MSCs implanted with morselized allograft
Aoyama et al.58 (2014)	10	Scaffold and cells	Autologous bone marrow-derived MSCs implanted with β-tricalcium phosphate scaffold in combination with vascularized bone graft
Chondral and osteochondral defects
Jo et al.59 (2014)	9	Cells	Autologous adipose-derived MSCs injected into osteoarthritic knees at different dosages
Ha et al.60 (2015)	27	Signals and cells	Injection of allogeneic chondrocytes engineered to express TGF-β1
Delcogliano et al.61 (2014)	19	Scaffold	Acellular collagen-hydroxyapatite scaffold
Kon et al.62,63 (2011 and 2014)	27	Scaffold	Acellular gradient-based collagen-hydroxyapatite scaffold

Fracture Nonunion

While there is currently no standardized definition, fracture nonunion has been defined by the FDA and others as incomplete healing at 9 months after injury and the absence of healing progression over the following 3 consecutive months⁶⁴. Despite advances in surgical techniques, fracture nonunions continue to present clinical challenges⁶⁷. As disrupted vascularity is one of the major contributors to nonunion⁶⁸, strategies to enhance angiogenesis, including delivery of hematopoietic stem cells, are being explored. In one study, autologous bone marrow-derived hematopoietic stem cells were delivered on an atelocollagen scaffold into tibial and femoral nonunion defects in 7 patients⁵³. This combination of cells and scaffold resulted in fracture-healing in 5 (71%) of 7 patients at 12 weeks. While there was no control arm, the threshold of healing achieved in the historical outcome of the standard of care was 18% (2 of 11 patients).

In a retrospective review of 52 patients treated for forearm nonunion defects at a single center, patients were categorized as being treated with a single tissue engineering element (MSCs, a scaffold, or BMP, i.e., “monotherapy”) or with all 3 elements of the tissue engineering paradigm (“polytherapy”)⁵⁴. It is not clear if a standard concentration of MSCs was used among the patients. With a minimum follow-up time of 1 year, patients receiving all 3 elements had significantly improved radiographic healing, clinical success criteria, and rapidity of healing compared with patients treated with monotherapy. While that study lends support to the synergistic nature of combining cells, scaffolds, and signals to treat severe orthopaedic defects, a randomized prospective study would have provided stronger evidence by minimizing bias.

To better compare combinations of cells, scaffolds, and signals in treating nonunion defects, a nonrandomized retrospective-prospective cohort study with 46 patients compared patients treated with bone marrow-derived stem cells with a scaffold (demineralized bone matrix) or rhBMP-2⁵⁵. A third group of 3 patients was treated with both scaffold and signal but was underpowered for statistical analysis. In that study, a greater percentage of patients treated with cells and scaffolds demonstrated healing compared with those who had treatment with cells and rhBMP-2 (86.4% versus 70.8%; p = 0.033). The authors postulated that because the demineralized bone matrix scaffold is osteoconductive and can recruit osteoprogenitor cells, it may be more effective than a single growth factor. Interestingly, all 3 patients treated with polytherapy (cells, signals, and scaffold) in that study demonstrated healing.

Osteonecrosis

Given their necrotic nature, the lesions associated with osteonecrosis have poor innate regenerative capacity with few or no viable MSCs. In addition, it has been reported that, in corticosteroid-induced osteonecrosis, there is a global decrease in available MSCs⁵². In theory, MSC therapy may be promising, given the immunomodulatory properties of MSCs as well as their ability to secrete angiogenic growth factors and recruit local vasculature⁵². Therefore, the delivery of MSCs (alone or in a scaffold) has been the goal of at least 18 published reports, according to Mont et al., in 2015⁶⁵. Examples of clinical successes with MSC therapy alone were reported in recent meta-analyses, but it was noted that comparisons between studies were difficult because of variations in the numbers of MSCs delivered^52,65.

As implantation with scaffolds has been demonstrated to improve MSC viability in animal models⁶⁹, a logical application of the tissue engineering paradigm is to pair implanted MSCs with a scaffold to ensure consistent MSC dosage as well as to support proliferation. In a recent clinical study of 5 femoral heads in 4 patients, autologous bone marrow-derived MSCs were seeded onto morselized allogeneic bone as scaffold and were implanted in the defect after core decompression⁵⁷. Unfortunately, in the patient who received treatment bilaterally, disease progressed in both femoral heads, leading to bilateral total hip replacement (after 13 and 19 months, respectively). However, this presented an opportunity to harvest the bone to analyze the treatment with radiographic, histological, and mechanical testing. Microcomputed tomography revealed that the treated zone was greater in opacity than trabecular bone. The tissue in the treated zone was histologically and mechanically indistinguishable from the patient’s healthy trabecular bone. The other 3 patients had no more disease progression after treatment (22 to 44 months of follow-up). As the authors noted, “Further clinical trials are necessary, including comparison to concurrent therapies,” before the full efficacy of the treatment compared with the current state of the art can be determined. However, the study as it stands represents a unique opportunity in which repair by tissue engineering elements could be evaluated with precision, including by mechanical testing, in treated human tissues.

In a similar study in which cells and scaffolds were utilized in disease treatment, 10 patients with osteonecrosis were treated with vascularized bone grafts augmented with autologous bone marrow-derived MSCs seeded on a β-tricalcium phosphate scaffold⁵⁸. At 2 years of follow-up, 9 patients had completed the protocol and 7 had no further disease progression. (One patient had been excluded because hip surgery had been performed bilaterally before the end of the follow-up period.) In the 2 patients who had disease progression, the authors postulated that “an imbalance between bone resorption and formation” may have contributed to the cystic lesions observed in their femoral heads 1 year after surgery. Given the degradation rate of β-tricalcium phosphate (weeks to months) and the low innate regenerative potential of the femoral head in osteonecrosis, a mismatch in the rate of scaffold degradation and native tissue regeneration may have resulted in collapse of the lesion. While a prospective randomized study is required for conclusive evidence on treatment superiority, these studies in sum demonstrate how selection of scaffold properties may impact the clinical outcome.

Chondral and Osteochondral Defects

Perhaps because of the tremendous clinical need for new therapies, there have been many recent studies involving tissue engineering strategies for the treatment of cartilage-based defects. Cell monotherapy for the treatment of cartilage defects is more common than in other orthopaedic pathologies. In fact, autologous expanded chondrocytes are part of a product regulated by the FDA and approved for the repair of symptomatic cartilage defects (Carticel; Vericel)⁷⁰. Despite the popularity of these monotherapies, the optimal amount of MSCs required for efficacy is unclear. In a recent study, autologous adipose-derived MSCs in 3 different doses were injected without scaffold or exogenous signals into osteoarthritic knees in 18 patients⁵⁹. While no patient experienced treatment-related adverse events, the effects of dose are more difficult to ascertain as the low and medium dosage groups had only 3 patients each. However, patients in the high dosage group demonstrated significantly improved clinical, radiographic, and arthroscopic measurements compared with baseline, and these improvements were not reflected in the low and medium dosage groups.

Increasing in complexity, allogeneic chondrocytes that have been genetically modified to express transforming growth factor-beta 1 (TGF-β1) by retroviral modification ex vivo were injected at 2 different concentrations in 27 patients with osteoarthritis⁶⁰. While the 2 groups did not show significant differences in healing compared with one another, both groups demonstrated significantly improved clinical scores (including reduced pain and stiffness and increased physical function) compared with baseline values. Given the public perception of genetic engineering²⁰, perhaps the most important findings of the study were the absence of serious adverse events and no evidence of global changes in TGF-β1 expression in patients or the appearance of the retroviral vector DNA in patient samples. This recent work demonstrates that cells can be programmed to deliver signals themselves to mitigate disease in human patients, rather than requiring exogenous delivery of the signals through non-living vehicles such as collagen sponges or microparticles.

Acellular scaffolds for the repair of cartilage defects have also been investigated in the past 5 years^61-63,71. For osteochondral defects, acellular composite scaffolds consisting of both bone-like and cartilage-like layers have recently been investigated in human patients. In one of these studies, a commercially available acellular scaffold consisting of a top layer of equine-derived type-I collagen and a bottom layer consisting of magnesium-enriched hydroxyapatite was used to treat osteochondral defects in 19 patients with 2 years of follow-up⁶¹. Patients exhibited a significantly improved subjective score at the 1-year follow-up; the score improved further, but not significantly, at the 2-year follow-up. While treatment was satisfactory in these patients, as a nonrandomized study with no control group, it is difficult to comment on efficacy. Another similar composite scaffold, constructed using collagen and hydroxyapatite nanoparticles in a gradient-based fashion, was implemented in 27 patients^62,63. Again, patients showed significant improvement compared with baseline values at both 2 and 5 years following treatment, which is historically not the case with osteochondral defects. While this improvement could be due to the gradient-based nature of the acellular scaffold, without a nongradient-based control, it is difficult to draw any definitive conclusions regarding the effects of acellular gradient-based scaffolds on the treatment of osteochondral defects in human disease.

Future Directions and Challenges

Although the future for orthopaedic tissue engineering is bright, there is much work to be done. While the small number of clinical studies reviewed in the present work may not necessarily be a representative sample of the field as a whole, they demonstrate that small patient numbers, lack of randomization, and absence of control groups consisting of current clinical treatment standards can make it difficult to determine the efficacy of the various elements of the tissue engineering paradigm in orthopaedics. In addition, a dearth of standardization in clinical protocols makes it difficult to compare results across existing studies. Even in cases of excellent clinical results, regulatory hurdles and the need for large amounts of capital for product translation contribute to the considerable barriers in guiding more tissue engineering-based technologies into the operating room^3,4.

Our understanding of the interactions between materials and biological tissues continues to grow. New manufacturing techniques, such as the advent of high-resolution bioprinting^72,73, are allowing for the rapid creation of personalized devices to an extent that was not previously possible. Infections, which often plagued implantation of foreign materials like scaffolds, are being mitigated by advances in anti-infective strategies such as biofilm-repulsing surfaces and antibiotic-delivering materials⁷⁴. Synthetic biology and genetic engineering techniques are being utilized to reprogram cells to optimize healing⁶⁰. Ultimately, products such as INFUSE (Medtronic) and Carticel have demonstrated that the field of orthopaedics is willing to adopt tissue engineering strategies into clinical practice. Hopefully, product development will catch up with these breakthroughs in fundamental knowledge as tissue engineers and clinicians alike better understand how to tackle the hurdles surrounding translation.

In conclusion, the goals of this work were to introduce the basic tenets of orthopaedic tissue engineering and to use current examples in the field to discuss how these elements can aid in treating human disease. While the field is advancing at an exponential rate and new discoveries are constantly being made, these fundamental principles will remain relevant. When designing a tissue engineering-based approach to an orthopaedic pathology, one must consider the appropriate scaffold, cell type, and source, and whether the inclusion of exogenous signals is necessary. In the past, the benchmark for successful treatment of conditions such as osteoarthritis was lack of disease progression. With the new therapies inspired by tissue engineering, the new benchmark may be the absence of disease.