Cartilage defect is a risk factor for osteoarthritis, and current therapies aim to relieve symptoms or prevent further degenerative changes to the articular cartilage. Cartilage damage is a key feature in degenerative diseases such as osteoarthritis (OA), with 26.6% of those aged 45 or above having a diagnosis of OA, which is projected to increase in the near future 
. The World Health Organisation suggested that 25% of OA patients are unable to perform major day-to-day activities and 80% suffer from movement limitations 
. Furthermore, knee and hip OA patients are at a higher risk of suffering from depression 
and 1.27 times more likely to have suicidal thoughts 
. With such widespread global impact and causing a significant socioeconomic burden for the patient, it is of great interest to scientists to find a solution for this.
Articular cartilage is avascular and devoid of lymphatics and nerves, limiting regenerative capabilities. Since the time of Hippocrates, many have astutely observed that once damaged, cartilage does not heal 
. Hence, current treatments for early to moderate OA are mostly palliative such as weight reduction, more exercise, walking aids and thermal modalities 
. Pharmacological treatment includes the administration of analgesics, non-steroidal, anti-inflammatory drugs and chondroprotective agents such as glucosamine sulphate, hyaluronic acid, chondroitin, growth factors and, hormones 
. These management options are insufficient in those with severe OA. Since there has not been any approved Disease-Modifying Osteoarthritis Drugs (DMOAD), these patients require surgical interventions such as arthroscopic lavage and debridement to remove inflammatory mediators and loose cartilaginous debris 
, osteotomy and knee arthroplasty to retain function, all of which are costly, fraught with perioperative and post-operative complications, and sometimes lead to unsatisfactory outcomes 
Research into cell-based therapy for the treatment of focal cartilage defects started at the end of the last century, with several of these products already on the market in Europe and the United States 
. Compared with microfracture, autologous chondrocyte implantation (ACI) resulted in better structural regeneration with minimal subchondral bone alterations, hence producing greater improvement in clinical outcomes 
. However, ACI is a two-stage process that is costly, associated with long rehabilitation, and has variable surgical risk factors affecting success rate. Arthroscopically-assisted osteochondral autologous transplantation surgery (OATS) has been successful, with a cohort of nine patients achieving an average post-operative AOFAS Ankle-Hindfoot score of 80.2 
. Mosaicplasty, which involves harvesting cylindrical osteochondral grafts of different sizes for transplant into focal cartilage lesions, has achieved good preliminary results and patient satisfaction 
. However, these procedures rely on harvesting healthy tissue from the same or other joints. Studies have reported donor-site morbidity from the knee after an OATS was performed to treat talar osteochondral defects, with patients having knee pain and patellar instability 
. A systematic review and meta-analysis reported a 7.8% donor-site morbidity rate from the knee after OATS for capitellar osteochondritis dissecans 
. Furthermore, the technique involves chondrocyte dedifferentiation 
and, ultimately, fails to reproduce the hyaline characteristics of original articular cartilage 
Mesenchymal stem cells (MSCs) are multipotent adult stromal cells that have the potential for self-renewal and multi-lineage differentiation. They can be isolated from various tissues such as bone marrow, adipose tissue, umbilical cord, dental pulp 
and even from discarded fragments during surgery such as the infrapatellar fat pad and synovial membrane 
. Depending on their tissue source, their methylation status determines their differentiation potentials. For example, the epigenetic memory of bone or adipose tissue-derived MSCs (AMSCs) favoured differentiation along an osteoblastic or adipocytic lineage 
. MSCs are also immunomodulatory and can act in a paracrine fashion to aid tissue repair and promote a host of cellular responses, ranging from survival to proliferation and migration 
. Linero et al. showed that AMSCs mediate bone regeneration mainly by releasing paracrine factors 
. MSC paracrine signalling can facilitate chemotaxis of macrophages and endothelial lineage cells 
, and MSCs themselves secrete mitogens that stimulate the proliferation of keratinocytes and endothelial cells 
. MSCs have, hence, successfully been used to treat chronic wounds and restart stalled healing processes 
. Despite endogenous MSCs in synovial fluid being able to proliferate following tissue injury, numbers available for cartilage repair are few 
MSCs avoid the ethical dilemma surrounding the use of embryonic stem cells (ESCs). However, they can influence tumour development and lead to cancer drug resistance 
, for example, by increased VEGF expression, leading to angiogenesis in pancreatic carcinoma 
. Furthermore, when integrated into tumour-associated stroma, they can cause cancer cells to increase their metastatic potency 
. MSCs can become immunogenic in certain situations, for example, after IFNγ stimulation, umbilical cord-derived MSCs increase MHC class I and develop MHC class II expression 
Given the increasing interest in MSCs for treating focal cartilage defects, many clinical studies have been performed to assess their efficacy in terms of functional and patient-reported outcomes, especially in relation to the site of MSCs extraction. While the conventional site of harvest of MSCs is the bone marrow, AMSC is an increasingly popular choice due to ease of harvest 
, purported higher proliferative capacity 
and anti-inflammatory properties 
. Aside from being administered as a culture-expanded population of MSCs, AMSCs can also be administered in the form of Stromal Vascular Fraction (SVF), a mixture of heterogeneous cells containing a fraction of AMSCs as well as pericytes, endothelial cells, smooth muscle cells, fibroblasts, etc. 
. Although SVF contains a fraction of AMSCs, since they express different cell markers and possesses different properties, they are not strictly defined as MSCs 
2. Adipose Tissue-Derived Mesenchymal Stem Cells and Cartilage Defects
MSC-based therapies for cartilage defects induce hyaline-like cartilage regeneration and, therefore, have the potential to improve clinical outcomes in patients with cartilage defects. This disease-modifying approach is vastly different to conventional palliative therapies.
2.1. Optimal Source of Stem Cell
AMSCs are an increasingly attractive source for MSCs. A lot of progress on MSC-based cell therapy was achieved using bone marrow-derived MSCs (BMSCs), which was the dominant source of MSCs 
, however, the recent literature has expanded to including MSCs derived from almost all human tissue, including pluripotent stem cells. Among numerous options, adipose tissue has emerged as a dependable and rich source of MSCs, with regards to increased quantity, higher yield, lack of ethical issues and ease of harvest with minimally invasive procedures such as liposuction 
. In vivo studies show that the differentiation potential of AMSCs is less attenuated by age when compared with BMSCs 
, and AMSCs have better immunosuppressive function 
. AMSCs also have higher proliferation potential according to growth curve, cell cycle and telomerase activity analyses 
, although other studies have suggested that the differentiation potential of both types of MSCs are comparable 
. Overall, while the exact properties of AMSCs compared with other MSCs have yet to be ascertained, AMSC-based therapies may be preferable to other MSC sources due to ease of harvest and having fewer ethical hurdles to overcome, which would aid its expansion in the healthcare market.
The ideal source for AMSCs is still a subject under research, however, IFPFs are an attractive source compared with adipose tissue since IFPF is usually obtained during the resection of inflamed tissue in knee arthroscopy 
. IFPF-MSCs also display similar surface markers compared with other cells around the knee joint, and may reduce immunologic rejection 
. Ding et al. suggested that IFPF-derived AMSCs may have higher proliferative potential than AMSCs derived from abdominal fat in vitro, although in vivo studies are needed to support this finding 
An in vivo study compared culture-expanded AMSCs with SVF regarding its use for osteoarthritis in sheep and reported better imaging, macroscopic and immunohistochemistry outcomes when AMSCs were used 
. However, the culture expansion of MSCs resulted in lower migration and homing ability 
, and, therefore, culture-expanded AMSCs may require surgery to expose the site of lesion. Furthermore, SVFs that are minimally manipulated may be favoured for economic or regulatory reasons 
. More comparative studies would be required to define the superior treatment for cartilage defects.
Zhao et al. evaluated the use of allogeneic stem cells as opposed to autologous stem cells 
. Allogeneic stem cells were obtained from three healthy donors and preclinical toxicity and chronic tumorigenicity were evaluated in vivo with no adverse events reported. Allogeneic MSC transplants have been previously demonstrated in animals as well as in human clinical trials, again with no adverse events reported 
. This may be due to the low immunogenicity of MSCs as well as the immune-privileged character of cartilage tissue.
2.2. Augmenting the Function of AMSCs
Despite the benefits of AMSCs, the outcomes of advanced clinical trials have sometimes fallen short of expectations 
. This could be due to the vast dimensions across which MSC heterogeneity is present, such as among donors, tissue sources and even cell subpopulations with the same origin 
, and studies showing that MSCs disappeared after 24 h post-infusion 
. It is, hence, prudent to ensure that studies utilising AMSCs adhere to well-defined international guidelines such as the International Federation for Adipose Therapeutics and Science (IFATS) 
while pursuing novel avenues to artificially boost AMSC potency to overcome the shortcomings of naïve AMSCs.
AMSCs can be pre-conditioned in vitro, which prepares it for the harsh microenvironment of the host, enhances its migration to its site of action, or enhances its biological properties. Given that synovial inflammation plays a key role in the pathogenies of osteoarthritis, and is associated with cartilage destruction and pain, enhancing the immunomodulatory function of AMSCs has gained a lot of attention 
. Pre-treating AMSCs with pro-inflammatory stimuli increases their immunosuppressive and anti-inflammatory potential by reducing NF-κB activity and promoting macrophage differentiation into the M2 anti-inflammatory phenotype 
. Studies have shown that AMSCs express toll-like receptors (TLRs), which can be exploited via specific TLF-agonist engagement to affect their migration, secretion of immune modulating factors and induce a change in cell fate 
. Furthermore, IFNγ-stimulated MSCs can enhance chondrogenesis 
. However, the degree to which differentiation is affected depends on the MSC source, despite ligation of the same TLR, is under investigation 
, and future, in-depth comparative studies are needed to determine the best MSC source for chondrogenic differentiation in this scenario.
AMSCs are also amenable to genetic modification and may be adequate vehicles for gene delivery 
. Targeted overexpression of microRNAs have produced favourable clinical outcomes 
. miR-302 transfection increased proliferation and inhibited oxidant-induced cell death in AMSCs, and can be used to enhance the therapeutic efficacy of AMSCs in vivo 
. However, studies have shown that genetic manipulation can affect differentiation potential, although results are conflicting and the effect on chondrogenic differentiation is unknown 
. As described above, the immunomodulatory potential of AMSCs to promote cartilage repair is attractive, and genetic engineering has the potential to improve immunomodulation. CTLA4Ig-overexpressing AMSCs have been shown in a mouse model to protect against cartilage destruction and ameliorated severe rheumatoid arthritis 
. Nevertheless, there is a risk of tumorigenicity and immunogenicity 
, and studies that examine the long-term clinical outcomes of using genetically modified AMSCs at a local and systemic level have yet to be published.
The niche microenvironment is crucial for stem cell integrity, fate and behaviour, and the regulation between an active and quiescent state 
. Culture conditions must be optimised and can increase the biological potency of AMSCs. Conventional practice involves growing MSCs on a two-dimensional system as monolayers. This is artificial and lacks the key cell–cell and cell–extracellular matrix contact present in vivo. Three-dimensional cultured MSCs have shown superior expansive and differentiation potential, such as undergoing large-scale in vitro chondrogenic differentiation and enhanced in vivo cartilage formation in an animal model 
. MSC aggregation into spheroids can also augment their immunomodulatory ability, as shown in a mouse model of cartilage damage 
. Nevertheless, with the increased complexity of three-dimensional culture, it is necessary to regulate the time duration of the culture and size of spheroids formed 
. Furthermore, studies have reported that AMSCs cultured under hypoxic (2% oxygen) conditions enhanced early chondrogenic differentiation, while decreasing osteogenesis, thus, favoured chondrocyte formation 
Finally, AMSC therapy can be combined with concomitant application after other bioactive molecules, or performed with other conventional procedures. Two studies injected AMSCs or SVF with PRP 
. In vitro research has suggested that PRP optimizes MSC-based therapy by stimulating MSC proliferation, migration and immune modulation without affecting differentiation potential 
. PRP may also optimize MSC-based therapy for cartilage regeneration by enhancing chondrogenic differentiation 
. However these studies did not investigate the effect of PRP alone, making it hard to determine the actual role PRP has on chondrogenesis when administered together with AMSCs. Koh et al. compared AMSCs concurrent with microfracture therapy with just microfracture therapy 
and reported improved radiologic and KOOS pain and function scores when AMSCs were administered together with microfracture. Nevertheless, only large cartilage defects (≥3 cm2
) were investigated, and future research is needed to investigate the effects of AMSC and microfracture co-therapy for lesions of different sizes.
2.3. Methods of AMSC Administration
Methods of AMSC administration varied greatly, though most studies favoured an intra-articular injection of AMSCs or SVF. Three studies investigated the use of implantation of AMSCs, none of which reported any adverse events 
. Kim et al. compared implantation versus injection of AMSCs in their study and reported a significant difference 
, suggesting that in the implantation group, ICRS scores and clinical scores such as IKDC and Tegner scores were higher than those in the injection group. The article suggests that the cartilage regenerated by implantation has greater durability than that of injection. This may also be due to better cell retention and survival at the lesion site in implantation compared with injection.
Some studies also compared the use of AMSC therapy and other pre-existing therapies for focal cartilage defects. Hong et al. compared SVF with the injection of hyaluronic acid 
, while Lu et al. compared the AMSC injection against the injection of hyaluronic acid 
. Both studies reported significant improvement in the AMSC or SVF group when compared with the control group with only the injection of hyaluronic acid. However, no study compared the efficacy of AMSC with SVF. Studies investigating cellular therapies for knee osteoarthritis suggested that iatrogenic complication may be higher in SVF groups, with more frequent knee effusion (SVF 8%, AMSC 2%) 
and minor complications related to the fat harvest site (SVF 34%, AMSC 5%). However, clinical outcomes such as pain VAS improved earlier and to a greater degree in the AMSC group than VAS group.
2.4. Improved Clinical Outcomes
Studies concluded that AMSC administration correlates with an improvement in pain and functional outcomes. Pooled analyses of WOMAC scores showed a statistically significant improvement in all studies across all follow-up times, regardless of the dose of SVF or AMSC used. This suggests that disease modification is long term, rather than simply acting as short-term analgesics, hence avoiding the need for the repeated administration of MSC therapy. However, a meta-analysis that directly compares SVF and AMSC treatments therapies could not be performed due to the low number of studies; furthermore, heterogeneity in the studies precluded any subgroup analysis of AMSC or SVF therapy.
Despite this suggesting a relationship between the number of AMSCs administered and therapeutic effect, Pers et al. found that only the group that received a low AMSC dose experienced significant improvements in pain levels and function compared to baseline 
. This could be explained by the fact that significant synovial inflammation was present in the low dose cohort, and, as stated previously, MSCs can be primed by an inflammatory environment to exert their immunomodulatory effects 
. This could lead to lower costs and the increased speed of AMSC harvesting, increasing their appeal in clinical practice. Due to the low number of studies investigating the outcomes of AMSC doses, subgroup analyses were not performed. Further research is needed on the dose-dependent relationship between AMSC therapy and long-term clinical outcomes, and, perhaps, studies that investigate the difference between one low dose and multiple low dose AMSC therapy, rather than giving one large bolus dose.
Joint pain and loss of functionality are the major symptoms of osteoarthritis. In the present study, VAS scores and other pain-related outcomes have been shown to improve after AMSC administration. There is an improvement in KOOS scores as well, which suggests an improvement in both symptoms and functions and also in the patient’s quality of life. Improvement in such clinical parameters suggests the therapeutic potential of using AMSCs over conventional therapies for focal cartilage defects. The therapeutic potential of MSCs on osteoarthritis have been previously described 
Most studies did not stratify their patients based on the severity of osteoarthritis. Given that an inflammatory environment encourages AMSCs to bring out their immunomodulatory effects, they may be most effective during end stage osteoarthritis, given that inflammation levels are highest as the disease progresses. The current literature remains divided, with Nguyen et al. suggesting better efficacy in patients with less severe osteoarthritis 
, while Tran et al. reported greatly reduced WOMAC scores 24 months post-treatment in patients with Kellgren–Lawrence (K–L) grade 3 than K–L grade 2 
Additionally, studies could have stratified patients based on age or BMI. Studies have shown that ageing negatively impacted AMSC proliferation and reduced its chondrogenic differentiation ability in favour of adipogenic differentiation 
. Obesity has been shown to be a key risk factor for osteoarthritis, however, the biological role of adipose-derived inflammation on MSC efficacy have yet to be investigated 
2.5. Hyaline-Like Cartilage Regeneration
The underlying mechanism of AMSCs amelioration of cartilage defects lie in the potential of AMSCs to generate hyaline-like cartilage.
Previous in vitro studies have exhibited the chondrogenic differentiation ability of AMSCs with successful differentiation indicated by the immunohistological staining of type II collagen or expression of glycosaminoglycan in the regenerated tissue similar to hyaline cartilage 
. Animal studies have also corroborated with in vitro results and evaluated the safety of the administration of MSCs 
. Bone marrow-derived MSCs and synovium-derived MSCs have been shown to induce hyaline-like cartilage regeneration confirmed by biochemical results such as improved GAG content and histological results related to type II collagen expression and integration. Early human AMSC studies, such as Pak et al., showed probable cartilage regeneration in the knee joint after AMSC administration 
. In our included studies, histological and imaging outcomes as well as biomarker analysis shows the presence of hyaline-like cartilage regeneration. Koh et al. performed histological staining and showed that participants receiving AMSC administration in concurrence with microfracture exhibited a higher degree of staining for safranin O and type II collagen than patients who received microfracture alone, suggesting the presence of hyaline cartilage regeneration after AMSC administration 
. For imaging analysis, the MOCART tool assesses regenerated cartilage compared to its similarity in terms of morphology and signal intensity with the surrounding native hyaline cartilage. The improved MOCART score post AMSC administration therefore suggests the presence of hyaline-like properties of the regenerated cartilage. Aside from MOCART, Freitag et al. utilized additional T2-weighted mapping to evaluate the quality of the regenerated cartilage and showed progressive cartilage maturation over time 
. These results point towards the ability of AMSCs to generate hyaline-like cartilage in humans.
Nevertheless, with an average follow-up time of 18.3 months, the long-term effects of AMSC therapy on cartilage repair are unknown. Jo et al. reported that cartilage degeneration occurred after two years post-treatment, perhaps due to desensitisation of the knee joint to AMSC therapy, or simply the need for another AMSC injection 
. Park et al. provided more optimistic results, with all cartilage regeneration parameters remaining stable over a seven year follow-up period 
. There is a need in the literature for more studies with long follow-up times to determine the long-lasting effects of AMSC therapy.