Keywords

1 Introduction

Over the last decade, there has been a great deal of interest in mesenchymal stem cells (MSCs) and their potential role to play in the treatment of osteoarthritis (OA). The burden of OA has seen an exponential increase, with the World Health Organization reporting 10% of men and 18% of women over the age of 60 years suffering from symptoms of OA [1]. This burden is expected to increase with the universal rising geriatric population [2,3,4]. The exact aetiology of OA remains uncertain, but literature has revealed age, obesity, trauma, genetics, infection and primary orthopaedic pathologies, together have a multifactorial role in contributing to the biochemical and biomechanical alterations in joint homeostasis to initiate or progress to OA [5, 6]. In the past, OA treatment strategies consisted of pain alleviation with drugs or interventions such as platelet-rich plasma [7], corticosteroid injections, viscosupplementation [8] and finally surgical interventions such as microfracture [9, 10], osteotomies [11] and finally arthroplasty [12, 13]. Recently, with further clarity on the pathophysiology of OA, research has shown numerous cytokines and free radicals having a significant role in increasing pro-inflammatory pathways leading to matrix degradation and onset of OA [14]. This has resulted in a keen interest in biological approaches for treatments using stem cells such as mesenchymal stem cells (MSCs), which have proven immunomodulatory and anti-inflammatory roles [15,16,17,18]. It has been suggested that some of these roles are fulfilled by paracrine signalling by employment of exosomes shed from MSCs, allowing for the regeneration and upregulation of endogenous chondrocytes [19, 20]. Recent studies also indicate that there may be subsets of MSCs that perform different functions [21]; MSC may, therefore, have multiple roles to play in cartilage repair. In the past, the most commonly used source of MSCs was bone marrow, but over time, literature has revealed that this source contributes inadequate cell numbers and is inferior when compared to other sources such as adipose and synovium [22,23,24,25]. Another source of stem cell is embryonic stem cells (ESCs), which are known to be superior in pluripotency but pose many ethical issues regarding their clinical and experimental use [26]. A more recent breakthrough cell source was discovered by Yamanaka et al. [27], where these authors were able to reprogramme mouse and human adult fibroblasts to become pluripotent cells, which exhibited embryonic stem cell morphology and growth properties [28]. These induced pluripotent stem (iPS) cells showed differentiation capacities into all three germ cell layers similar to ESCs, making them an additional potential cell source for all regenerative cell therapies. Table 15.1 summarizes the major differences between MSCs, ESCs and iPS cells, with modifications from a table compiled by Lin et al. [35].

Table 15.1 Comparing the major differences and similarities between MSCs, iPSCs and ESCs
Full size table

Clinical improvement with the use of stem cell therapies has been shown in a number of pre-clinical trials, but the objective outcome data have not been consistent, and this limitation remains a limitation for clinical trials [36,37,38,39]. Overall, with the use of MSC therapies being deemed safe in either autologous or allogeneic form, much research has been focused on identifying cell based therapies to retard OA progression and reverse the disease-associated catabolic pathways [40,41,42]. In this chapter, we discuss the current roles of stem cells and the current research on iPS cells in OA management.

2 Mesenchymal Stem Cells in OA

As mentioned above, various types of stem cells exist, and depending on their tissue of origin, they possess different advantages and disadvantages. MSCs may be harvested from bone marrow [43], synovium, adipose [44], dental pulp [45], umbilical cord blood [46], peripheral blood [43, 46], placenta [47], muscle [48], skin [49] and periosteum [50]. Figure 15.1 illustrates a few of the many sources of MSCs in the human body. MSCs are characterized by their fibroblast-like shape, adherence to plastic, tri-lineage differentiation capacity, and immunophenotypes [51]. At present, there remains no consensus on the ideal cell source for MSCs, and considerations include harvest cell number, donor site, differentiation capacity, and proliferative potential. MSCs have been applied for chondral repair as they have the potential to differentiate into chondral, adipose, and bone tissue [51, 52]. Due to their lack of human leukocyte antigen class II, these cells exhibit low immunogenicity making allogeneic cell use possible [53]. MSCs have been postulated to function by either acting as a precursor to chondrocytes or to mediate joint regeneration via enhanced secretion of trophic factors by endogenous cells [54], thus supporting the possibility of multiple subsets [21]. The paracrine effects of MSCs, mediated by the shedding of exosomes and secretion of bioactive molecules, have been identified to be a key feature allowing for immunomodulation and facilitated tissue regeneration [19, 54,55,56]. These trophic factors increase cellular migration and differentiation while regulating prostaglandin and inflammatory molecule production [57].

Fig. 15.1
figure 1

Illustrating a few of the various sources of mesenchymal stem cells in the human body

Full size image

MSC-based OA therapies have been investigated in both pre-clinical studies and clinical trials environments with cells from different tissue sources and various formulations. They are commonly isolated and expanded in culture before administration through direct intra-articular injection, or in combination with a tissue engineering strategy on a scaffold. One step injection protocols have been popular, with the earliest being simple bone marrow aspirate injections, although yields contained low MSC numbers [58]. Newer injections protocols have now been developed, such as using a stromal vascular fraction (SVF) [59,60,61] derived from adipose tissue and micro-fragmented adipose tissue [62]. These newer approaches utilizing adipose tissue have shown to yield higher number of MSCs [60, 63]. However, expansion of MSCs in vitro before administration as a treatment is not possible in many clinical settings due to regulatory restrictions imposed by government agencies.

3 Embryonic Stem Cells in OA

ESCs originate from the embryo, more specifically from the inner cell mass of the blastocyst depicted in Fig. 15.2. These cells are totipotent with the ability to differentiate into any cell type that would make up a fully developed human. ESCs are also highly proliferative and do not undergo differentiation like other cells. The fact that these cells are extremely young and possess the ability for self-renewal and differentiation into ectodermal, endodermal and mesodermal cells theoretically makes them the most superior stem cell source available for stem cell therapies. However, experiments with these cells have demonstrated teratoma formation, a finding that raises concerns about the use of ESCs in clinical cell therapies. The main challenges with ESCs have been related to ethical approvals and regulations, which do not allow for the harvest of the cells from the blastocyst. Therefore, despite the many positive reports on the potential of ESCs, the efficacy of such treatments and more importantly safety are yet to be determined.

Fig. 15.2
figure 2

Embryonic stem cells are harvested from the inner cell mass of the blastocyst

Full size image

4 Induced Pluripotent Stem Cells in OA

iPS cells were first generated using murine fibroblasts [27], soon after which they were created using human fibroblasts [28, 31]. In the murine models, the fibroblast cells were transduced using four factors; c-Myc, K1f4, Oct3/4, and Sox2, allowing for somatic cell reprogramming [27] illustrated in Fig. 15.3. These cells formed teratomas when transplanted into immunodeficient mice, where the production of hyaline cartilage was also noted [64]. iPS cells have been used in a variety of regenerative modalities with the hope of possible disease modification. Owing to their indefinite proliferative potential, ability to become any desired cell type, and abundance, iPS cells eliminate many shortcomings associated with the previously discussed stem cell therapies, making them an attractive option in current experimental trials [65]. Previous ethical concerns prevented the use of embryonic cells in such trials; however, iPS cells do not evoke such concerns. Another important advantage of iPS cells over bone marrow MSCs (BMMSCs) is that in vitro BMMSCs are primed towards endochondral ossification [66,67,68] resulting in the tissue produced being hypertrophic, expressing large amounts of collagen I and markers of calcifying cartilage [66, 69]. iPS cells could also allow for larger amounts of in vitro hyaline cartilage generation [68]. With respect to cartilage regeneration and OA, the use of iPS cells would allow for the generation of autologous cells with just the use of a skin fragment and initial dermal fibroblast cultures. After which, the cells can be induced to form iPS cells and then subsequently subjected to chondrogenic differentiation [70, 71]. Four induction methods have been studied for the conversion of iPS cells to chondrocytes [72, 73]. The first is with primary chondrocyte co-culture, where various secretory factors from the chondrocytes can stimulate the iPS cells towards chondrogenic differentiation. The second study method involves the use of growth factors for chondrocyte differentiation. A third method employing chondrogenic supplementation can be used similar to the method used for MSC chondrocyte differentiation. Finally, differentiation can be regulated by specific media changes mimicking that of normal developmental cell differentiation processes. This latter method appears to be the most successful method for the production of stable hyaline cartilage [74,75,76].

Fig. 15.3
figure 3

Somatic cells are reprogrammed using four factors; c-Myc, K1f4, Oct3/4 and Sox2, which result in the production of induced pluripotent cells

Full size image

iPS cell therapies do have some challenges to overcome, and making them available for cell-based therapies is a major hurdle at this point in time. Although any cell can be reprogrammed to form an iPS cell, all cells appear to undergo some element of genetic mutation during their lifetime. This risk could affect the reprogrammed cells negatively and enhance the potential for tumourigenicity [77]. Using embryonic cells from cord blood for iPS cell generation is, therefore, preferred in view of their low genetic modification rate [78, 79], but this option is not available in many cases. Cord blood banking has been a popular trend in recent times, but generating iPS cells from cord blood cells and subsequently banking them for a large population is an extensive and difficult task. Therefore, an allogeneic system may be more practical, and much research has been focused on making this a reality through iPS cell banks [80, 81]. Another concern is the unlimited proliferative potential of iPS cells, while being one of their major initial advantages, upon implantation, they theoretically could proliferate indeterminately and result in tumours. In view of such concerns, iPS cells cannot be implanted without prior differentiation, and exclusion of all iPS non-differentiated cells must be confirmed prior to the initiation of the therapy [68]. Numerous strategies to decrease such risks have been proposed, including the inclusion of suicide genes to destroy the cells in the event of an adverse effect [82]. There is no doubt regarding the value and potential of these universal cell sources to eliminate the shortcomings of previous cell-based therapies, especially with regard to the field of cartilage regeneration.

Clinical trials have not begun for cartilage repair treatments with iPS cells, but the ongoing pre-clinical research appears positive, along with the development of iPS cell banks. Pre-clinical studies aim to identify the safety of iPS cell treatments, including the ideal sources of these cells and optimal culture conditions. Yamashita et al. [74] performed a study with cultured human iPS cells in a chondrogenic medium containing specific growth factors resulting in chondrocytes, which were then transplanted into immunodeficient mice and mini-pigs. They found the presence of bone morphogenic protein 2 (BMP2), transforming growth factor b1 (TGF-b1), and GDF5 to be essential for chondrogenic differentiation of human iPS cells. They reported that transplantation of the cells into mice and mini-pigs showed good integration with the surrounding native cartilage. This is a positive finding, as when mature chondrocytes have been transplanted they do not exhibit such good integration. iPS cells being in very early phases of differentiation can mimic the normal developmental pathway allowing for better chondral maturation and, therefore, improved integration with the mature native cartilage tissue. These authors also did not report any teratomas or tumour formation in the in vivo studies, a major concern with the use of iPS cells. Various other studies have also used growth factors for chondrogenic differentiation of iPS cells and reported encouraging results [83, 84]. Other protocols to stimulate chondrogenic differentiation have involved MSC-like populations [70, 85], chondrocyte co-cultures [86], and embryoid body formation [87]. The progenitor cell for iPS cells has also been a topic of study, and neural crest cells were thought to be a good candidate given their ability to differentiate into osteochondral tissues. iPS cells derived from neural crest cells have been studied and shown to have good chondrogenic differentiation capacity under in vitro conditions; however, the cells did not achieve adequate defect filling when implanted to in vivo chondral defect sites. Again, there was no teratoma formation or tumour growth detected [88]. Other progenitor cell sources from which iPS cells have been derived and studied for differentiation into chondrocytes include umbilical cord blood [89], peripheral blood, [90, 91] and dermal fibroblasts [92]. Research is still experimenting with the ideal cell source and methods leading to chondrogenesis concerning iPS cells. For some, peripheral blood and umbilical cord blood have become preferred sources of iPS cells because of easy harvest and effective reprogramming [93]. Optimal methods for induction of chondrogenesis are under investigation, with each protocol having its own advantages. Suchorska et al. recently suggested the most direct, fast, and cost-effective methods to be monolayer cultures with growth factors or a medium conditioned with human chondrocytes [92]. These pre-clinical studies should lead to movement in the direction of further in vivo studies, and in time, clinical trials once they have achieved more efficient cell reprogramming and chondrogenesis protocols.

5 Review of Clinical Trials Using MSCs in OA

Several clinical trials have been reported using varying numbers of MSCs from different sources. These studies also report on different delivery modalities ranging from intra-articular injections to tissue engineered approaches. These variable approaches are discussed in the sub-sections that follow.

5.1 Intra-articular Injections

A systematic review conducted by Chahla et al. [94] investigated the use MSCs in the treatment of OA and concluded that they were unable to perform a meta-analysis due to the high heterogenicity between trials. Their review included 6 studies, of which 3 focused on MSC therapies in OA across 124 knees. Of the three studies reporting on OA, two utilized autologous adipose derived mesenchymal stem cells (ADMSCs) and one BMMSCs, which were expanded to passage 3. They noted overall positive clinical improvement in the selected studies and reported the therapies to be safe, but could not rule out a placebo effect. They concluded that literature quality is poor owing to lack of blinded trials, cell population definition, standardization, and quantitative metrics to define cell populations.

Kim et al. [95] analysed five randomized control trials (RCTs) (level II), where four trials employed BMMSCs and one ADMSCs. Their cumulative pain score assessment revealed significant improvement in clinical outcome scores [96,97,98,99]; however, the MRI evaluations from three of the selected studies showed no evidence for improvement [96,97,98]. They too concluded that the optimal cell concentration needed to be determined, along with better standardized trials and that, currently, despite the encouraging results, MSC injections in OA should be investigational based on the available literature.

A larger systematic review was performed by Ha et al. [100], where 17 level I–III studies were included. Their mean follow-up was to 28 months, and cell study sources included bone marrow, adipose, SVF, and umbilical cord blood. Of the 17 studies, all but 2 reported clinical improvement. Only seven studies compared the experimental arm to a control group, where four reported significantly better results in the MSC treated group [97,98,99, 101]. Eleven of 17 studies reported MRI evaluation, of which only two reported no change in cartilage status [96, 102]. The last two assessed outcomes were second look arthroscopy and histology. Of six studies, one reported no improvement at arthroscopy, [102] and out of four studies, one demonstrated osteoarthritic chondrocytes [102]. Their principal finding was similar to other reports in that they concluded there is limited evidence for the use of MSCs in knee osteoarthritis. Although several studies reported clinical benefit, the RCTs reported controversial results.

Jevotovsky et al. [103] performed a review to evaluate MSC use in OA, in relation to study quality and procedural specifics. Their conclusion was similar to the other discussed reviews in that MSC therapies alleviated symptoms of OA, but due to inconsistencies in study methodology, MSC preparations and protocol design, it is difficult to draw definite conclusions regarding the therapeutic benefits of MSC treatments. Most reviews regarding intra-articular therapies have reported MSC injections to be safe overall; however, a few adverse effects such as synovitis [96], pain and swelling have been reported, but such reactions were also found in study control groups, indicating that they could be associated with any injection [101]. The literature also remains inconclusive regarding the optimal MSC cell count in the intervention, as well as the number of doses, with some studies reporting higher cell number and multiple doses being more beneficial [39, 99, 104, 105].

5.2 Tissue Engineering Approaches

Tissue engineering utilizing cell-based strategies has aimed to take things further than simple injections, by programming the stem cells to differentiate towards specific target tissues [106, 107]. Studies have employed specific growth factors and scaffolds made of various biomaterials, all to provide the cells with an effective microenvironment to promote differentiation into chondral tissue [108]. Most clinical studies in this area have used MSCs in combination with a scaffold or an adjunct technique such as autologous chondrocyte implantation or microfracture. The most popular tissue source for clinical MSC tissue engineering treatments has been bone marrow, usually in the form of an autologous bone marrow aspirate concentrate. However, several of the other above-mentioned sources have also been used. MSCs have been combined as an adjunct to existing techniques such as augmented autologous matrix-induced chondrogenesis [109], as well as to microfracture [110], to improve the outcomes of already utilized techniques. MSCs have also been combined with scaffolds such as a collagen matrix [111,112,113,114], polyglycolic acid [115], polylactic acid, [116] and hyaluronan [117]. These studies have mostly reported clinical improvement and reasonable chondral defect fill; however, the quality of the repair tissue has been at best hyaline-like cartilage, which is still imperfect. MSCs appear to improve tissue quality and outcomes, but further research is required to generate repair tissue that is actual tissue regeneration. Isolation and quality control of MSCs remains the major challenge as, currently, the resultant cell populations are very heterogeneous with regard to proliferation, lineage differentiation, and molecular response patterns. This can lead to variable results in terms of chondrogenic differentiation efficiency [118]. Recently, a scaffold-free tissue engineering technique has been introduced using synovial MSCs in a high-density monolayer culture, which results in the formation of a three-dimensional tissue engineered construct (TEC) [119]. TEC implantation has shown favourable pre-clinical results demonstrating hyaline cartilage repair, which has both biological and mechanical properties similar to that of native cartilage [120]. With the excellent pre-clinical data, a clinical study was conducted using TEC in five patients with knee chondral defects. At 24-month follow up, patients had significantly improved clinical outcome scores, second-look arthroscopy demonstrated complete defect fill, and histology of a repair tissue biopsy showed the presence of hyaline cartilage [121]. The same group is currently performing a randomized control trial. Table 15.2 summarizes the results with each MSC tissue source and the resultant clinical outcomes.

Table 15.2 Summary of the differentiation capacities of bone marrow, adipose and synovium tissue and the clinical results for each MSC source
Full size table

6 Conclusion

Currently, available literature on MSC therapies in osteoarthritis is voluminous, and despite this, it is difficult to deduce precise inferences regarding the effects of MSC therapies for OA treatment and chondral regeneration. The heterogeneity and inferior quality of clinical trials have instigated misperceptions and unregulated non-standardized use of what may be a valuable clinical solution for OA. The iPS cell has in pre-clinical studies shown immense potential and superiority over MSC treatments but has also exhibited possible tumourigenic risks. Without extensive pre-clinical studies and steps to mitigate such risks, as well as ascertain the detailed behaviour of these cells, clinical trials should be delayed. It is hoped that with the introduction of MSC therapy definitions, and the development of superior isolation and quality control protocols, better standardized clinical trials and indications will be published allowing for higher quality analysis of level I data. At present, stem cell therapies for OA should be investigational, and clinicians using them should be encouraged to collect outcome data in the form of high-quality RCTs defining their cell source and specifics of preparation so as to contribute to the standardization of protocols and evaluation of optimized procedures.