Tissue engineering approaches for the treatment of meniscal defects have demonstrated a promising means of restoring native meniscus properties, especially injuries to the inner avascular region. In particular, treatment of meniscus injuries within the inner avascular region could utilize this technique. However, to evaluate translational approaches, models mimicking the degenerative situation are required. The present investigation sought to evaluate cell-based tissue engineering approaches for treatment of the avascular region of the meniscus within a rabbit model showing signs of early osteoarthritis. The present study demonstrated regenerative treatment of avascular meniscal defects in this situation for two different autologous cell sources, meniscal cells and MSCs that were seeded in a hyaluronan collagen composite matrix. Three months of in-vivo treatment with either cell type enabled defect filling with differentiated meniscus-like tissue that was completely integrated with the surrounding native tissue.

Early osteoarthritic changes of the knee are a very demanding situation, especially for regenerative treatment strategies [8]. Many of these degenerative changes might initiate an inflammatory status and secretion of catabolic factors that lead to development of late-stage osteoarthritis [10]. In the context of early osteoarthritis, significant correlations between early osteoarthritic changes in the submeniscal tibial plateau cartilage and meniscal degeneration have recently been detected [29]. Additionally, there appears to be a correlation between meniscal extrusion and cartilage damage in the peripheral region of the tibial plateau, underlining the fact that the submeniscal region is vulnerable to early osteoarthritis [30], and thus leads to structural and mechanical alterations of the meniscus leading to further implications within the joint [31, 32]. Although the onset of joint degeneration represents a very demanding situation for regenerative treatment, these facts emphasize the need to restore the meniscus to prevent knee collapse. Resections of both medial menisci in the animal model in this study leads to early degenerative changes in all rabbit knees after 3 months, with cartilage defects and formation of osteophytes. The average OARSI grading was 3.1, indicating an early OA situation [27] (Fig. 1).

Meniscal cells are derived from the tissue at the defect site and are a potential cell source for treatment of meniscal injuries, although the meniscal self-healing capacity is limited [1, 33]. Using a tissue engineering approach helps to overcome these limitations, with a reduced requirement for intrinsic meniscal regeneration. The application of autologous meniscal cells seeded on a hyaluronan collagen-based scaffold induced complete defect filling with differentiated tissue. Webber [34] showed that culture and differentiation of human meniscal cells was possible. However, monolayer expansion of the cells has been shown to result in dedifferentiation and thus requires three-dimensional culture to restore phenotype [35]. Furthermore, regional differences regarding the chondrogenic potential exist, as meniscal cells derived from the outer vascularized zone show a higher chondrogenic capacity than meniscal cells from the inner avascular part [16]. Thus, meniscal cells alone are not wholly responsible for the reduced intrinsic self-healing capacity. Hennerbichler et al. [15] showed that reinserted meniscal plugs in the outer and inner zones of the meniscus reintegrated into the surrounding meniscal tissue in vitro, with stable connecting fibers between the meniscal cells. Explants from the avascular inner zone and vascular outer zone of the meniscus exhibit similar healing potential and repair strength in vitro. In the present investigation, we have shown the repair capacity of the meniscus cells in an in-vivo situation. The reason for the enhanced repair in these previous investigations may be the existence of progenitor populations within the meniscus, particularly from the outer meniscus [1, 17].

A substantial disadvantage of autologous meniscal cells as a source for cell-based treatment is their limited availability and the resultant donor-side morbidity. In the in-vivo study, the resection of the medial menisci of both knee joints was necessary to obtain a sufficient number of cells for a cell-based treatment of a small 2-mm circular punch defect. Three months postoperatively, all knee joints began to show degenerative changes in the medial compartment with chondral lesions, softening of the surrounding cartilage, and formation of osteophytes. This is not surprising since the resection of the medial meniscus serves as a model for inducing the development of osteoarthritis in animal studies [36]. In clinical practice, the only possible option to obtain autologous meniscal cells would be to harvest meniscal debris or tissue derived from nonrepairable tears. Baker et al. described that cells derived from surgical debris are a potent cell source for engineered meniscus constructs in vitro [37]. However, their results are dependent upon two observations. Their use of a biodegradable nanofibrous scaffold contributed to the increasing content of proteoglycan and collagen II over the culture period of 70 days. Furthermore, some of the donors came from knee arthroplasty patients, and thus there were resections with large amounts of meniscus substance. Nevertheless, there was significant data variation relating to these observations and the continual passaging of meniscal cells increases the risk of cell dedifferentiation to obtain sufficient cell numbers for further culture and analysis.

In the present study, an in-vitro evaluation of human meniscal tissue from non-refixable tears was used to assess their potential in a clinical setting. However, the human meniscal cell pellets cultured in chondrogenic medium revealed moderate gene expression and no deposition of collagen II after 21 days (Figs. 4, 5 and 6). As previously discussed, monolayer expansion to achieve sufficient cell numbers for aggregate formation increases dedifferentiation and may have led to poor outcomes in vitro despite the presence of TGF-β. These results are also contrary to our in-vivo data or similar studies using meniscus tissue, and thus indicate that this may not be an appropriate cell source for meniscus repair in a clinical setting.

An alternative cell source for meniscus repair are MSCs, and these have been shown to induce meniscus regeneration [1113, 28]. MSCs have been detected in vivo following meniscal lesions in the knee synovial fluid [21], whilst a progenitor population has been identified within the avascular region of the injured meniscus that had high migratory potential towards the lesion [38]. However, our study focused on the MSCs derived from the bone marrow, which show the potential to differentiate into bone, adipose tissue, and cartilage in this setting [13]. These cells demonstrated meniscus-like repair in the punch defect after 3 months in vivo. This confirms the results of previously published studies on different meniscus defect types that all showed that untreated injuries showed no healing, with a “non-union” of the lesion comparable to the human situation [1113, 28]. Furthermore, in these studies, the application of a MSC-based treatment revealed significantly superior results compared to the use of a cell-free scaffold. Therefore, the described animal model can be considered as appropriate for the evaluation of the regenerative potential of different meniscus treatment options. As the historical controls with untreated defects and treatment with cell-free scaffolds showed only a reduced quality of regeneration with nonintegrated fibrous tissue, different cell-based repair strategies should be tested in this study. Meniscal cells and MSCs differentiated and integrated meniscus-like repair tissue, with scoring results between 12 and 16 points after 3 months. In comparison, the historical controls of cell-free treatment showed scoring results of only 8 points, indicating an improvement in meniscus regeneration using a cell-based repair strategy. In addition, the demanding early osteoarthritis situation within the model provides further evidence of the suitability of MSCs within a clinical context.

Previous studies have reported the positive effects of MSCs on meniscus regeneration both in vitro and in vivo and from different sources, including adipose and synovium [18, 39, 40]. Gonzalez-Fernandez et al. compared the regeneration potential induced by bone marrow-derived MSCs to adipose tissue-derived stem cells in an equine model and found no differences between the two cell sources [19]. As previously stated, the feasibility within a degenerative situation was not assessed in their equine model. The reasons for the successful repair of the meniscus using MSCs may be related to two distinct mechanisms. MSCs may have differentiated into meniscal cells due to the surrounding tissue matrix, and cell-cell communication or the secretion of trophic factors released by MSCs may have helped to heal the meniscus via pharmacological means or via recruitment of resident cell populations [41].

As MSCs from many sources show positive results regarding the enhancement of meniscal repair it seems that the question is not the origin of the progenitor cells but more their availability and applicability for clinical use. In other connective tissues, such as articular cartilage, preliminary clinical data show positive effects on regeneration with the application of MSCs. Sekiya et al. detected significant improvements after arthroscopic transplantation of synovial-derived stem cells in small (average 200 mm2) cartilage defects by magnetic resonance imaging (MRI) scoring, qualitative histology, and Lysholm score evaluation [42]. A further advantage of autologous MSCs is the potential for a single-step cell-based repair augmentation since they have a higher proliferation rate than meniscal cells and are less susceptible to dedifferentiation [43]. Our in-vitro results confirm the high chondrogenic potential of human MSCs and their qualification for augmentation of meniscal healing in a clinical setting.

However, limited data are available on the clinical use of MSCs for meniscus regeneration. Whitehouse et al. [44] conducted an open-label first-in-human study for repair of avascular meniscal lesions with autologous MSCs. Following isolation and expansion, MSCs from iliac crest bone marrow were seeded on collagen scaffolds. These cell-matrix constructs were placed and sutured into avascular meniscal tears of five patients. After 2 years, three patients were asymptomatic with no signs of a recurrent tear in the MRI control compared to two patients that required subsequent meniscectomy due to re-tear or nonhealing. In a controlled randomized trial, Vangsness et al. delivered allogenic MSCs 1 week after arthroscopic partial medial meniscectomy via intra-articular injection to the knee. A year after surgery, a significantly increased meniscal volume determined by quantitative MRI was detected in 24% of patients in the treatment group, while no patient in the control group showed an increasing amount of meniscus tissue. Additionally, stem cell injection revealed beneficial effects on pain management of these patients after partial meniscectomy [45].

A limitation of this study is the animal model and the comparison to a historical control. However, the study shows that regeneration of avascular meniscal defects is possible by a cell-based treatment even in a demanding and clinically very relevant situation such as early osteoarthritis. A strength of the study is the analysis of different human cell sources that may be used for an autologous cell-based meniscus treatment using meniscus cells or MSCs. In comparison to MSCs, human meniscal cells from non-refixable meniscal tears or meniscal debris demonstrated a very limited capacity for chondrogenic differentiation. Thus, MSCs appear to be the most promising cell source for an autologous cell-based meniscus treatment approach, including low donor site morbidity. However, there are disadvantages that limit their clinical applicability, particularly the high treatment costs, requirement for cell expansion prior to application, and regulatory burdens that currently inhibit their use in daily clinical practice [43]. These problems should be resolved to facilitate the cell-based treatment of meniscal defects with MSCs and therefore improve the clinical outcome of this common injury.