Can Bone Tissue Engineering Contribute to Therapy Concepts after Resection of Musculoskeletal Sarcoma?

Resection of musculoskeletal sarcoma can result in large bone defects where regeneration is needed in a quantity far beyond the normal potential of self-healing. In many cases, these defects exhibit a limited intrinsic regenerative potential due to an adjuvant therapeutic regimen, seroma, or infection. Therefore, reconstruction of these defects is still one of the most demanding procedures in orthopaedic surgery. The constraints of common treatment strategies have triggered a need for new therapeutic concepts to design and engineer unparalleled structural and functioning bone grafts. To satisfy the need for long-term repair and good clinical outcome, a paradigm shift is needed from methods to replace tissues with inert medical devices to more biological approaches that focus on the repair and reconstruction of tissue structure and function. It is within this context that the field of bone tissue engineering can offer solutions to be implemented into surgical therapy concepts after resection of bone and soft tissue sarcoma. In this paper we will discuss the implementation of tissue engineering concepts into the clinical field of orthopaedic oncology.


Introduction
Bone has to carry major loads. To ful�l this task it is created as a composite material, which comprises primarily of collagen, noncollageneous proteins, and hydroxyapatite. Its complex structure contains a wealth of mechanically relevant details [1]. Bone is a composite in several senses that is, being a porous material, a polymer-ceramic mixture, a lamellar material, and a �bre-matrix material. Its mechanical properties will therefore depend on each of these aspects of composition and structure. In general, bone displays a high intrinsic regenerative capacity following trauma or disease. erefore, the majority of fractures heal spontaneously by a recapitulation of the pathway of normal fetal skeletogenesis, Sarcoma including endochondral and intramembraneous ossi�cation [4]. �e�nements in surgical techniques, implant design and postoperative care have signi�cantly improved treatment outcomes of complex fractures and defects as caused by highenergy trauma, disease, developmental deformity, and revision surgery. However, there are conditions in which bone regeneration is compromised or in which bone regeneration is required in a large quantity.
A situation of the latter entity is the resection of malignant bone and so tissue sarcoma. is can result in large defects where regeneration is needed in a quantity far beyond the normal potential of self-healing [5]. Furthermore, an adjuvant therapeutic regimen or local factors such as postoperative seroma or infection can account for a limited intrinsic regenerative potential [6]. erefore, reconstruction of these defects is still one of the most demanding procedures in orthopaedic surgery.

Common Treatment Strategies for Bone Defects
Treatment protocols of bone and so tissue sarcoma are based not only on the tumour biology and location, but also on the patient's needs and age [7]. e primary surgical goal should be to obtain adequate surgical margins in order to ensure local tumor control [8][9][10]. With the introduction of multimodal therapeutic concepts and improved reconstruction techniques, limb salvage procedures have largely replaced ablative surgery [11][12][13][14], but only few of them can restore the original anatomical and functional conditions. For the reconstruction of skeletal defects aer tumor resection, both biological techniques (e.g., autogras, allogras, or rotationplasty) and prostheses are used. ese procedures, mainly bone graing and metallic implants, are well established and the comparative advantages and disadvantages have been discussed at length in the literature [15][16][17]. Problems of autologous bone graing can be donor site morbidity and limitation of the gra mass. For gra harvesting, additional personnel and time are needed. e use of allogras or xenogras carries the risk of immunomediated rejection, transmission of infectious diseases, or gra sequestration. In addition, the acquisition costs of allo-or xenogras are rather high. Gra devitalisation and consecutive absorption processes can lead to decreased mechanical stability. Failures usually result from incomplete transplant integration, particularly in critical sized defects. Due to the dense nature of cortical allogras, revascularisation and cellular invasion is impeded. is limited ability for revascularization and remodelling is believed to be responsible for the high complication rate associated with allogras [18]. Other biologic approaches used for the reconstruction of bone defects include distraction osteogenesis, segment transport, or the Masquelet technique, but all of these methods are technically demanding and they require lengthy treatment protocols, which can be highly inconvenient for patients [19,20]. e limitations of these conventional biological reconstruction techniques are exacerbated in cancer patients, who are oen elderly, have localised or systemic osteoporosis and suffer from impaired wound healing as a consequence of an adjuvant therapeutic regimen. e high tensile strength and fatigue resistance of metal would make it suitable for load-bearing applications, but the large mismatch in Young's modulus between metal and bone can lead to peri-implant bone resorption, a phenomenon known as stress shielding [21]. Furthermore, tumor endoprostheses exhibit a higher complication rate than standard implants with infection or aseptic loosening as the most common failure modes [22]. ese constraints have triggered a need for new therapeutic concepts to design and engineer unparalleled structural and functioning bone gras to replace current treatment options. To satisfy the need for long-term repair and good clinical outcome, a paradigm shi is needed from methods to replace tissues with inert medical devices to more biological approaches that focus on the repair and reconstruction of tissue structure and function [23]. It is within this context that the �eld of bone tissue engineering can offer solutions to be implemented into surgical therapy concepts aer resection of bone and so tissue sarcoma. While this has already led to a variety of novel therapeutic concepts particularly in the �eld of craniofacial surgery [3,24], only few called smart biomaterials have found their way into clinical application in the �eld of orthopaedic surgery.
In the following passages, we will discuss the implementation of tissue engineering concepts into treatment strategies of bone defects caused by musculoskeletal sarcoma. From a material science and especially clinical point of view, the future prospects and possible application spectrum will be outlined.

Tissue Engineering Constructs (TECs)
e �eld of tissue engineering is embodied in the collective vision of its early pioneers Langer and Vacanti, whose diverse yet symbiotic research approaches as an engineer and surgeon led to the commencement of this interdisciplinary �eld. eir seminal 1993 paper remains one of the most in�uential and cited works in the �eld [25]. e application of the principles of biology and engineering to the development of functional substitutes for damaged tissue has seen laboratories worldwide forging impressive multidisciplinary teams to focus on restoring, maintaining, or improving the function of a wide range of human tissues. While progress has been made to deliver bench to bedside solutions, the rate at which tissue engineering has seen innovations translated to the clinic has been slower than originally expected and the urgency for tissue-engineered products which achieve these ideals remains high [26][27][28].
e fundamental concept underlying tissue engineering is to combine a scaffold with living cells and/or biologically active molecules to form a "tissue engineering construct" (TEC) which promotes the repair and/or regeneration of tissues [29,30]. A suitable scaffold should (i) possess a porous interconnected pore network (pores and pore interconnections should be at least 400 m to allow vascularisation) with surface properties optimised for the attachment, migration, proliferation, and differentiation of cell types  of interest (depending on the targeted tissue) and enable �ow transport of nutrients and metabolic waste, (ii) be biocompatible, and (iii) be biodegradable with a controllable rate to complement cell/tissue growth and maturation [23,31]. e design of these scaffolds also needs to consider physicochemical properties and morphology. External size and shape of the construct are of importance, particularly if the construct is customised for an individual patient. e work by groups focussing on scaffold design and fabrication utilising additive manufacturing technologies has advanced the tissue engineering �eld tremendously over the past few years [32]. e ability to create scaffolds in a layer-bylayer manner enables a computer-aided design to be directly translated from a clinical scan (i.e., a patient CT or MRI scan) to produce customised and/or patient-speci�c scaffolds to �t any anatomical defect site [33,34].

Regeneration and Remodelling of TECs
Aer scaffold implantation, continuous cell and tissue remodelling is essential to achieve and maintain stable biomechanical conditions, vascularization, and integration within the host site [2]. Importantly, TECs should stimulate and support both the onset and the continuance of bone ingrowth as well as subsequent remodelling and maturation by providing optimal stiffness and external and internal geometrical shapes. Scaffolds must provide sufficient initial mechanical strength and stiffness to substitute for the loss of mechanical function of the diseased, damaged, or missing tissue. Furthermore TECs must degrade at a rate which is compatible with new tissue ingrowth and maturation [35]. It is essential to understand and control this scaffold degradation process for successful tissue formation, remodelling and maturation at the defect site. In the early days of tissue engineering, it was believed that scaffolds should degrade and vanish as the tissue is growing [36]. ough, tissue ingrowth and maturation differ temporally from tissue to tissue and, furthermore, tissue ingrowth does not equate to tissue maturation and remodelling. In other words, a defect �lled with an immature tissue should not be considered as "regenerated. " Hence, many scaffold-based strategies have failed in the past as scaffold degradation was more rapid than tissue remodelling and/or maturation [37]. Our concept of using a slow degrading composite scaffold fabricated with pores and pore interconnections with a size larger than 400 m is illustrated in Figure 1.

Translating Bone Tissue Engineering Concepts into the Clinical Field of Orthopaedic Oncology
Bone defects aer resection of musculoskeletal tumours represent a considerable surgical challenge, are associated with high socioeconomic costs and highly in�uence patients� quality of life. ese problems may be approached from the perspective of the nature of the gra material with which the surgeon works. e mission of our interdisciplinary group is to coordinate efforts between researchers and clinicians in the area of bone tissue engineering and the translation of tissue engineering platforms into orthopaedic oncology. Laboratories in Singapore, Germany, and Australia have spent the last decade in close collaboration translating a concept of bone tissue engineering based on slowly biodegradable composite scaffolds comprising medical grade poly (epsiloncaprolactone) (mPCL) and calcium phosphates from bench to bedside [38][39][40][41][42][43][44][45]. Aer a large series of in vitro experiments, we consequently performed small animal studies using mouse, rat, and rabbit models which demonstrated the ability of composite scaffolds in combination with BMPs or cells to promote bone regeneration within ectopic sites or bone defects [35]. Another key project of our group has been the development of a large animal model for bone regeneration research. We recently have established and fully characterised a critically sized defect model in sheep tibiae to evaluate different tissue-engineering-based treatment strategies [46,47].
In the following section, we will describe a part of the rationale and road map of how our multidisciplinary research team is approaching the �rst steps to translate bone tissue engineering concepts into orthopaedic oncology.
Our clinical partners have used custom-made metal prostheses for the treatment of pelvic defects aer sarcoma resection for more than 20 years [48,49]. Since 1988, the general production process of the prosthesis has only changed in details but has developed according to the technological advances available. In the �rst step a 1 : 1 pelvic model is made using data acquired via high-resolution computed tomography. e model is cut out from a block of polyurethane by a �ve axial C�C-milling machine. In the next step the surgeon uses this model to de�ne the levels of osteotomy with special regard to the later surgical margins. According to the planned osteotomy planes and the acquired CT-data not only the custom-made prosthesis but also special osteotomy guides are constructed by the manufacturer to ensure accurate �tting of the prosthesis. e series reported from our institution showed encouraging results (Figure 2) [50].
Although these massive endoprostheses provide orthopaedic oncologists with many reconstructive options, failure rates are still high. Especially in younger patients a reconstructive method would be desirable that does not rely on the use of permanent metal implants but rather on bioactive materials enabling customised reconstruction and supporting natural healing processes. Using computeraided design (CAD) and fused deposition modelling (FDM) technologies, we are able to produce bioresorbable composite scaffolds fabricated from mPCL, either with or without reinforcement using up to 20 wt% -tricalcium phosphate (TCP) for bone tissue engineering applications at load-bearing sites ( Figure 3). is scaffold exhibits mechanical and structural properties comparable to cancellous bone and can be speci�cally adapted to the clinical needs with a fully interconnected pore network structure [45,51]. A detailed description of the fabrication protocol has been given elsewhere [34,39,44]. ese scaffolds are already in clinical use and are FDA approved for craniofacial applications [3] (Figure 4).
In principle, the fabrication process of our scaffolds as depicted in Figure 5 is similar to that described for the patient-speci�c and individually customised pelvic metal implants. Preoperatively, high-resolution CT data is processed via a 3D medical imaging soware (e.g., InVesalius 3.0) supporting the medical DICOM/PACS format to generate a virtual model of the pelvis. In the next step, the data set is converted into a Standard Tessellation Language format (STL) which is the standard format for rapid prototyping applications. Accordingly, a model is built from an acrylonitrile butadiene styrene (ABS) polymer using a 3D fabricator (FDM3000, Stratasys, Eden Prairie, USA) based on fused deposition modeling technology. e model facilitates the haptic perception and orientation both before and during surgery. In close collaboration with the orthopaedic surgeon, the levels of osteotomy are marked in both the virtual and the physical model. Considering the dimensions of the tumor it should be possible to achieve tumor-free resection margins. As it has been previously described for the customised metal implants [50], based on the virtual model, special osteotomy guides can be manufactured to facilitate the later resection and implantation process of the scaffold. As the dimensions of the prospective bony defect are exactly known, the dimensions of the scaffold can be virtually adjusted. Moreover, the form of the scaffold can be adapted to the clinical needs. In the �rst step, we mirror the healthy side of the pelvis to the affected one to achieve near-physiological conditions. Aerwards the scaffold is armed with �anges and an intramedullary peg to improve its primary stability. en, Skeinforge soware is employed to generate the printing toolpath, which is subsequently modi�ed to introduce porosity allowing tissue ingrowth. An in�ll density of 0.2 is chosen, corresponding to 80% porosity. Furthermore, the perimeter sections of the toolpath are removed using a custom algorithm to generate open pores to the exterior of the scaffold. According to this modi�ed toolpath the scaffold is then manufactured using again a 3D fabricator. We used a commercially available MakerBot Replicator with poly(D,Llactide (PDLLA) as biomaterial. PDLLA is a biodegradable thermoplastic polymer which has been successfully applied for �xation in maxillofacial reconstructions before [52,53] and has been made available for several additive manufacturing techniques such as fused deposition modeling (as in this application) and stereolithography [54]. During surgery, the �anges are �xed with resorbable tacks or screws. Additionally, the contact area between the scaffold and the host bone can be covered with �brin glue which can serve F 5: Schematic illustration of the scaffold manufacturing process. A 3D computer-aided designed (CAD) model of the patient's pelvis is fabricated according to data obtained by high-resolution CT ((a), (b)). Using this prototype, the surgeon indicates the osteotomy planes needed to achieve tumour-free resection margins, aer which the CAD model is virtually resected ((c), (d)). A scaffold model is then derived by mirroring the healthy side of the pelvis and ad�usting the si�e to �t into the defect ((e), (f)). �e scaffold can be armed with �anges or an intramedullary peg to enhance its primary stability ((g), (h)) and exhibits a porous internal architecture to allow for tissue ingrowth and regeneration (i). F 6: In orthopaedic oncology, vascularised �bula transfer is considered as one of the most suitable techniques for the reconstruction of critically sized defects of the tibia diaphysis due to the mechanical strength, the predictable vascular pedicle, and the hypertrophic potential of the �bula. Combining the autogra with a large bone allogra can enhance the biomechanical properties of the construct. However, the use of allogras can be associated with signi�cant drawbacks such as immunomediated rejection, gra sequestration or transmission of infectious diseases. In addition, the acquisition costs are rather high. A novel biological approach could be to combine an intramedullary �bular autogra with a customised tissue engineered bone construct. Aer tumor resection (a) a customised mPCL/TCP tube is placed around the vascularised �bula (b) to �ll the defect. Together with an internal �xation device, it ensures load distribution and primary stability. Secondary stability is achieved by osseointegration of both the �bula and the porous scaffold (c).
as a biomimetic template promoting migration of osteogenic cells. e presented therapy strategy combines the advantages of both CAD/CAM procedures and tissue engineering concepts. Technically, it is not restricted to the application in defects caused by pelvic tumors but can also be transferred to other bony defect sites ( Figure 6).

Outlook
ough, small bony defects such as cysts are relatively easy to handle in the routine clinical setting, the management of large defects in load-bearing bones presents a particular challenge in reconstructive surgery and particularly in orthopaedic oncology. In this opinion paper, tissue engineering has been suggested as an alternative strategy to regenerate bone in patients with musculoskeletal sarcoma. We have developed an integrated holistic approach for the reconstruction of bone defects caused by musculoskeletal tumours using patient-speci�c scaffolds with well-de�ned macro-and microarchitecture. ough promising case reports have been presented in the literature [3,55], large clinical studies, which can show the efficacy of this approach in the clinical setting, are still missing. To tackle major bone tissue engineering problems in orthopaedic oncology, researchers have to perform functional assessment of the biological and biomechanical parameters of the regenerated bone. Furthermore, to allow a comparison between different studies, animal models, �xation devices, surgical procedures, and methods of taking measurements need to be standardised to achieve an efficient accumulation of reliable data as a foundation for future developments.