Histological assessment is commonly employed to evaluate osteochondral tissue regeneration. The regenerated tissue, including the SBP, is compared with corresponding host tissue to further assess repair quality. In addition, microcomputed tomography (µ-CT) is an acknowledged ‘gold standard’ for SBP evaluation, with its global view of bone architecture providing a detailed assessment. However, criteria for satisfactory SBP regeneration have not been established. Typical SBPs share common features, such as a suitable height, a flat and smooth surface, good interface integration and dense texture (17); SBP regeneration is not considered successful unless all four features are present. Other literature has summarized the histological scores of SBP repair by scoring abnormal activity following the repair (18). The standard for scoring abnormal activity is that a SBP with a low score has good mobility and good mechanical properties. However, a low score for repaired SBP is not representative (18), because few studies have used this scoring system and it cannot quantify performance appropriately.

In contrast to the aforementioned four criteria for adequate SBP regeneration, the four typical indicators of a failed repair are abnormal height, uneven surface, poor integration and loose lacunae structure (Fig. 3). When three or more indicators are present, defects filled with fibrous tissue have been reported, along with either tissue resembling collapsed cartilage or scarcely any cartilage regeneration. When one or two indicators were found in the repaired SBP, regenerated hyaline cartilage with sufficient integration with surrounding host cartilage was observed (Fig. 3). while a plate with one or two indicators of failure was defined as a moderate repair. Among these four indicators of failed repair, abnormal height was the most frequent. Moreover, when abnormal height was apparent, it was often accompanied by two or three other indicators, which suggests that abnormal height has a clear influence on SBP repair and that SBP may be required to reach an appropriate height during regeneration. These four histological findings provide a path to understanding the underlying causes of inadequate osteochondral regeneration from the perspective of the SBP.

One of the most significant functions of seeding cells is to promote the regeneration of cartilage. Research has indicated that BMSCs have a distinct advantage over ADSCs in cartilage repair because of their greater capacity for cartilage differentiation (21,22). Compared with chondrocytes, BMSCs can produce ECM with higher mechanical strength (21,22). However, the aforementioned advantages do not consider the effect of the graft itself. When widely applied to osteochondral grafts, BMSCs occasionally result in the formation of fibrocartilage instead of hyaline cartilage (23). This unsatisfactory repair should not be attributed to the cell type since the same limited cell types are used in bilayer and trilayer grafts. The bilayer graft, which is usually composed of chondrogenic and osteogenic layers, resembles a normal osteochondral unit, with the top layer encouraging cartilage repair. BMSCs can be inserted into the cartilage layer to restore the ECM. Alternatively, chondrocytes can be implanted in the chondrogenic layer to improve chondrogenesis because of their outstanding ability to synthesize cartilage matrix (22). Similarly, BMSCs and chondrocytes are utilized in trilayer or multilayer grafts for cartilage regeneration. Under these circumstances, the neocartilage is always reconstructed with firm and compact cell lineage. BMSCs and chondrocytes have significant potential to promote cartilage regeneration in both bilayer and multilayer grafts.

Although bone regeneration appears easier than cartilage regeneration over an extended period, bone tissue reconstruction must be relatively rapid and provide early support for superficial chondral tissue regeneration (24,25). For this purpose, cells were incorporated into tissue-engineered grafts to adjust the speed of bone reconstruction. BMSCs have been widely used in this process because of their self-renewal and differentiation capacities (1). Other studies have reported the use of periosteum-derived progenitor cells, which can induce the formation of subchondral bone. These cells can also stimulate differentiation of mineralized bone tissue and bone marrow (Fig. 4) (19,26). Furthermore, osteoblasts in the underlying layer of bone can promote the growth, differentiation and infiltration of new bone (27) and are mainly responsible for the synthesis, secretion and mineralization of bone matrix (19). When cultured on osteochondral tissue grafts, the regenerated bone tissue stimulated by osteoblasts is firm with good mechanical properties.

No existing literatures discuss the function of a certain cell to promote the regeneration of SBP exclusively, as SBP is often regarded as part of subchondral bone. Previous studies viewed that cells which can regenerate bone certainly can repair SBP (28–30). However, the SBP can be satisfactorily rebuilt when specific cell types are incorporated into a well-integrated graft that contains growth factors. It has been found in previous studies that BMSCs embedded in bi-layer or tri-layer grafts can obtain flat and compact SBP regeneration (28–30), which no is doubt related to the polypotent and powerful differentiation potential of BMSCs (22). Others have shown that ADSCs can achieve good SBP regeneration after implantation in bi-layer grafts (28–30). Besides these pluripotent stem cells, chondrocytes may have the similar function to repair SBP (24). A flat and neat SBP structure well integrated with surrounding tissues can be detected. This is also closely related to the chondrocyte's function of secretion, synthesis and induction of calcified cartilage matrix (22). Previous research revealed that the regeneration of the SBP could not be independently stimulated by any one cell type, including stem cells or somatic cells. In general, SBP, cortical bone above trabecular bone, can usually achieve ideal repair results right after subchondral trabecular bone is well repaired (Table II).

Subsequent research revealed that the biocompatibility of the three-layer/multi-layer graft with the typical osteochondral unit structure gave it unique benefits in the repair of osteochondral lesions. Three-layer graft, distinguished from the two-layer graft, adds a transition layer between the chondrogenesis layer and the osteogenesis layer. It was expected that the SBP would undergo good regeneration when the cells were grown into the transition layer (30). Some stem cells with osteogenic potential, such as BMSCs and ADSCs, are commonly applied into the translational layer. For instance, these investigations all share the construction of the calcified cartilage layer (CCZ) and the successful SBP regeneration that results from the addition of BMSCs to the CCZ (31–33). On the other hand, the SBP typically heals poorly if these cells are denied access to the intermediate transition layer (30). These results can prove that osteogenesis is one of the necessary factors for the SBP. The trabecular bone structure can provide a stable biomechanical environment and mechanical support for the regeneration of the SBP. Also, translational layer of grafts provides enough space to accommodate neo SBP.

As a pivotal part of tissue engineering strategy, seeding cells can promote regeneration of the osteochondral unit, but cannot be effective without supportive scaffolds. The critical concept of the supportive scaffold design is to induce cell growth, proliferation and differentiation at the defect sites (34). As research rapidly progressed, monolayer scaffolds were the first to be designed and studied. Owing to their inherent limitations, monolayer grafts cannot regenerate the entire osteochondral structure; instead, their primary goal is to stimulate cartilage regeneration (35–37). The most recent tissue engineering approaches for the development of monolayer scaffolds include acellular cartilage matrix (ACM) and hydrogels, which are shown in Fig. 5 (38–41). Of these, ACM is the most frequently used material (38–41) because cartilage matrix clearly promotes cartilage regeneration and has demonstrated the ability to mimic various distinctive requirements of an ECM-like microenvironment. Hydrogels can be made from natural or synthetic polymers, depending on their unique characteristics and specific functions (36). They both offer conditions suitable for cell proliferation and differentiation. Hydrogels based on natural polymers including silk fibroin protein, sodium alginate, porous chitosan and hyaluronic acid have been extensively documented (41–45). Widely reported synthetic polymers include polycaprolactone (PCL), poly lactic-co-glycolic acid (PLGA) and polyurethane (43). These hydrogels are often a good choice for scaffold design due to good biocompatibility and biodegradability (46). In addition to these single-component hydrogel scaffolds, other scaffolds using combinations of two or more biomaterials, such as sodium alginate-gelatin, collagen-fibroin, hyaluronic acid-chitosan, have been used (23,29,47–49). These composite scaffolds are not only more beneficial to cartilage reconstruction but also markedly enhance the biomechanical properties of scaffolds (47). However, only a portion of cartilage tissue may be restored by monolayer grafts and subchondral bone is frequently disregarded (35,50). Although monolayer grafts do not appear to improve the repair of osteochondral lesions, their creation provides a foundation for additional study and the development of bi- and trilayer grafts.

Typically, monolayer grafts only partly repair the defect and cannot achieve overall regeneration of the osteochondral unit. Due to the distinct mechanical strengths and biological environments of the cartilage and bone layers, the design of bilayer scaffolds is much more complicated. The cartilage layer of bilayer scaffolds shares a common structure with monolayer scaffolds, that is, polymer and ACM hydrogels (18,47,51). Due to the robust mechanical properties, excellent biocompatibility and slow biodegradability, cartilage regeneration was evident (38) and the arrangement of newly formed chondrocytes was similar to natural cartilage (52,53). The design of the osteogenesis layer of the bilayer scaffold is also important. To provide a suitable microenvironment and ensure bone regeneration, hydrogels and some inorganic materials became the main source for the osteogenesis layer (54–56). Commonly used materials, including bioceramics and bioglass (54–56), have been improved over the past few decades, with excellent osteoconductive and inductive properties, stiff mechanical strength and extraordinary biodegradability, allowing these materials to demonstrate superiority for subchondral bone repair (54). The osteogenesis layer of scaffolds constructed with hydroxyapatite (HAp) (57) show good bioplasticity (58) and a pattern similar to trabecular bone structure can be shaped using three-dimensional (3D) printing and other technologies. Bilayer scaffolds tend to achieve better repair results for both the cartilage layer and the subchondral bone layer compared with monolayer scaffolds.

Among bilayer graft studies, only a few reported that the SBP regenerated with an uneven surface or anomalous tissue formation between the cartilage and subchondral bone layer (58–60). At present, however, no phase of bilayer scaffolds has been developed that can precisely repair the SBP. Nonetheless, it has been demonstrated that offering a somewhat stable 3D microenvironment allows the bilayer scaffold to accomplish SBP regeneration (58). Proper porosity is one of the key conditions for creating a microenvironment and is also important for scaffold design (61,62). Relevant studies have reported that bilayer scaffolds with specific porosities can enhance SBP regeneration (55,63–65). Using different porosities, the microenvironment can support the growth and multiplication of encapsulated seeding cells as well as osteogenic and chondrogenic differentiation and can also improve structural stability and enhance mineralization in the subchondral bone region (66,67). For example, a bilayer scaffold with an upper porosity of 200 µm and a lower porosity of 400 µm (66,67) provided clear evidence of improving SBP regeneration (68). Although it is not entirely convincing to rely solely on designing different porosity to achieve SBP regeneration (55), these designs provide insight into how best to repair the SBP. If trilayer or multilayer scaffolds achieve unsatisfactory reconstruction, porosity may be a potential target to improve the SBP regeneration.

Trilayer grafts are designed based on bilayer grafts. The biggest advantage of the trilayer design is that it addresses the structural flaws in bilayer grafts by incorporating more structure into the graft. A transitional layer between the cartilage and bone layers often plays a pivotal role, with its functions including isolation of components and provision of a physicochemical barrier, cross-linking network and mechanical support (69). These grafts have two types of components; one that is similar to the composition of cartilage and another similar to the composition of bone (31,34). For the cartilage-like phase, Col-II-based scaffolds have been the most used. A dense isolated layer produced by Col-II/PLGA, with a small enough pore size and porosity, prevents excessive downward cartilage growth (34) and at the same time inhibits bone hyperplasia and hypertrophy (70). Others synthesized the calcified cartilage layer with Col-II/HAp and Col-I/HAp as a transitional layer (71) that served as a vital physical barrier, separating nutrients and cells within their respective spaces, avoiding crosstalk and interference between the cartilage and subchondral bone layers and ensuring a distinction between the cartilage and bone environments (72). For the bone-like phase, bioceramic was the most common material. PCL-β-tricalcium phosphate is a prepared transitional layer that has been reported to match the local Young's modulus in the middle and also provides steady support with excellent mechanical properties (31). Additionally, this transitional layer may offer sufficient space to regenerate the SBP. Trilayer scaffolds reported in the literature thus distinguish themselves from bilayer scaffolds. Together with the creation of stable cartilage and trabecular bone structures, trilayer tissue-engineered osteochondral grafts yield good repair of the SBP.

Compared with trilayer grafts, multilayer grafts could provide a more sophisticated structure to promote SBP regeneration (70,73). Typically, multilayer grafts are multifunctional, with each layer capable of performing a distinct function. Of those reported, the most common type includes a top layer promoting cartilage regeneration, a second calcified cartilage layer, a third transitional layer and a bottom layer that mainly mimics the porous structure of trabecular bone (33). This complex architecture creates a biomimetic environment for the entire osteochondral tissue (74). The calcified cartilage layer not only balances the differential physical load between upper cartilage and lower subchondral bone (34), but also helps stabilize the overall mechanical properties of the scaffold. In addition to the isolation and barrier function aforementioned (34), the intermediate transitional layer can release growth factors to the upper and lower layers of the scaffold (75). Notably, in both small and large animals, SBP regeneration results were good with a flat and even surface achieved (30,33,76). In general, the closer the structure of the artificial scaffold is to normal osteochondral structure, the more improved the repair efficacy. Multilayer grafts may thus result in improved osteochondral unit regeneration compared with bilayer grafts.

Given the complex structure of multilayer scaffolds, fabrication techniques have limited their progress. 3D printing has become a standard technique for fabricating biomimetic scaffolds. To date, a variety of 3D printing methods such as fused deposition modeling (FDM), inkjet printing, light-assisted bioprinting (digital light processing), stereolithography and laser-based printing have been used to engineer different tissue repair scaffolds (67,68) (Fig. 5). To mimic the structural and mechanical characteristics of subchondral bone and improve the hydrophilicity of the bone scaffold, some previous researchers fabricated the core-sheath structure bone layer using FDM (30). Other researchers have produced the chondral and osteochondral phases with 3D bioplotting; these have been designed with a cross-linking network and high porosity (71,73). These structures had surprising biocompatibility and could be adjusted according to the lesion position. Moreover, with 3D printing technologies, the fabrication of patient-specific scaffolds that perfectly match the size and shape of the defect would come true. And targeting porosity, layers, integration would be designed by 3D-printing technology, all these provide the possibility for future customized strategies. Consequently, multilayer scaffolds are unquestionably of higher quality because of 3D printing. Moreover, additional techniques for multiphase scaffolds are anticipated in the future.

Gradient scaffolds have gained increasing popularity over the past decades. Gradient scaffolds involved two types: Compositional gradient scaffolds and structural gradient scaffolds (77). First, as with other kinds of stratified scaffolds, the materials applied in gradient component scaffold included collagen or extracellular matrix (77–79) and inorganic polymers such as PCL, PLGA and gelatin methacryloyl (GelMA) (80,81): With their suitable biodegradability and mechanical properties, scaffolds physicochemical stability could be sustained continuously (77). In contrast to stratified scaffolds, the primary characteristic of gradient scaffolds was the smooth transition between layers. Gradient scaffolds can prevent sudden component changes in distinct zones since the components in the osteochondral micro-environment vary gradually as depths increased (81). A gradient component scaffold was designed with PLGA/TCP and the upper cartilage layer was constructed of a microtubule-like structure. These microtubules were interconnected and parallel arranged in perpendicular plane (diameter, 84.2±20.7 µm). The bone layer had a highly interconnected porous structure with large pores (450.5±47.2 µm) (76). The whole osteochondral structure was repaired integrally and both cartilage layer and trabecular bone layer showed perfect reconstruction. Other studies have also proposed a biodegradable hydroxyapatite-collagen (HAp/Coll) gradient distribution scaffold with collagen matrix synthesized in four different weight ratios as follows: 0:100, 10:90, 30:70 and 50:50 to simulate the normal amount of cartilage and bone components (82). As well as cartilage and trabecular bone regeneration, regeneration of SBP could be also detected (82). The special benefits of the compositional gradient scaffold, structural stability, a biocompatible interior environment and successive compositional alteration, were the main reason of all these satisfactory results (77,82).

Microsphere scaffolds, another common construction prototype as gradient scaffolds, organize a three-dimension and porous structure for cells proliferation and tissue formation. In addition to the inherent advantages such as biocompatible structure and smooth translation between layers, the microsphere scaffolds show unique features to repair osteochondral defect. First is the architectural and mechanical variations throughout their 3D structure. Chitosan/mesoporous silica nanoparticles was designed as microsphere scaffold by Yuan et al (83). The nanoparticles possess excellent bio-remodeling activity, different patterns can be arranged in the chondrogenesis zone and osteogenesis zone. With their 3D structure and porous formation, the differentiation of chondrocytes can be obviously promoted (83) (Fig. 5). By adjusting the arrangement of nanoparticles, the mechanical properties of the scaffolds can mimic the transition from soft cartilage tissue to the calcified cartilage and ultimately subchondral bone (84). Along with the feasibility of nanoparticles, osteochondral defects could gain regeneration. The second characteristic of microsphere scaffolds is their ability to create interfacial cohesiveness between several layers. PCL/HA was used to construct a composite microsphere scaffold by Gu et al (85). Due to their tiny spherical 3D structure, these microspheres have the potential to improve scaffold transitions while also strengthening the interface integration (77). Regeneration of cartilage and bone tissue was observed following implantation of these microsphere scaffolds (84,85). With their unlimited and promising prospect, the design of gradient grafts will be markedly improved. After long-term observation and evaluation, the satisfactory result of osteochondral repair would be achieved with the gradient grafts application.