Background
Osteoarthritis (OA) is a common and debilitating chronic degenerative disease of large joints, especially the hip and knee, characterized by a loss of articular cartilage, subchondral sclerosis, and marginal osteophyte formation. Worldwide, approximately 9.6% of men and 18% of women aged ≥60 years have symptomatic osteoarthritis [
1]. Current treatment in early-stage OA includes weight reduction, quadriceps strengthening exercises, non-steroidal anti-inflammatory drugs, intra-articular (IA) glucocorticoid injections, viscosupplements, and bracing [
2‐
4]. Total joint arthroplasty is the mainstay treatment for end-stage OA of the knee joint, which is often associated with serious and life-threatening complications including increase risk of infection [
5].
Currently, cell therapy- and tissue engineering-based approaches are being used to address the issue of repair of damaged articular cartilage. This includes autologous cultured chondrocytes and mesenchymal stromal cells (MSCs) obtained from various tissues that are used for transplantation into the cartilage lesion. Autologous chondrocyte implantation has inherent disadvantages such as a two-stage surgical procedure (harvesting healthy cartilage and transplanting culture-expanded chondrocytes from that sample) that may cause further cartilage damage and degeneration [
6,
7], and chondrocyte dedifferentiation during culture that might result in fibrocartilage rather than hyaline cartilage formation [
6,
8]. Thus, autologous or allogeneic MSCs are rapidly emerging as an investigational product for cartilage repair [
9‐
11]. The anti-inflammatory and immunomodulatory properties of MSCs suggest that these cells can reduce inflammation and pain reduction in the knee. Concurrently, MSCs may initiate the repair process of the damaged cartilage by differentiating into chondrocytes, as well as by inducing proliferation and maturation of the remaining healthy chondrocytes or by inducing differentiation of chondroprogenitors [
12]. A whole host of growth factors, biological modulators, and extracellular matrix proteins produced by MSCs may play a pivotal role in enhancing neocartilage formation [
12].
Several preclinical studies and clinical trials have been conducted using MSCs which have reported the safety and therapeutic effect of its administration in patients with OA, although the majority of these studies have been conducted as single-dose, single-arm pilot studies [
13‐
15]. Hence, there is a need for randomized, double-blind, controlled clinical trials. We have carried out in vitro studies to show the differentiation efficiency of adult human bone marrow-derived, cultured, pooled, allogeneic mesenchymal stromal cells (Stempeucel®) into the chondrogenic lineage and the expression of chondrocyte-specific markers. In order to determine if Stempeucel® is efficacious in a preclinical model, we have administered these cells intra-articularly into the knee joints of rats with mono-iodoacetate (MIA)-induced OA. After completion of these studies, a phase 2 dose-finding clinical study was initiated to evaluate the safety (primary endpoint), potential efficacy, and appropriate dose (secondary endpoints) of IA administration of Stempeucel® in patients with OA of the knee joint.
Discussion
The propensity of MSCs to differentiate into chondrocytes in vitro [
29] and their ability to repair articular cartilage has been shown in various preclinical models of OA [
30‐
32]. In several studies, MSCs were prepared and injected with sodium hyaluronan to increase the engraftment and chondrogenic activity [
30,
33]. In the present study, the efficacy of Stempeucel® was evaluated in a well-validated animal model of OA that was induced by MIA injection into the knee joints. Both low and high doses of Stempeucel® + HA treatment showed significant improvement in the pain threshold from week 2 onwards when compared to animals treated only with HA; treatment with only HA provided a short-term benefit on pain reduction, which corroborates with an earlier publication [
34]. We did not observe a significant difference between the two Stempeucel® treatment groups of animals (low and high dose) on pain reduction. However, it is important to note that the pain reduction in the high-dose animals continued to improve until the end of the study (12 weeks). Although the exact mechanism of action of MSCs on pain reduction is not known, anti-inflammatory activity has been attributed to this effect. To date, some studies have demonstrated the role of MSCs on OA-induced pain behavior [
35‐
37]. Van Buul et al. reported improvement of weight-bearing joints of the affected limb after intra-articular application of both rat and human BMMSCs in MIA-induced OA rats [
37]. However, unlike the results presented in this study, the authors did not observe cartilage regeneration. Furthermore, in several animal studies, it has been shown that the increased levels of pro-inflammatory cytokines might have contributed to pain increase. Intra-articularly administered MSCs probably play an important role in attenuating the inflammation-induced pain by secreting a wide range of anti-inflammatory cytokines and analgesic peptides [
38], and Stempeucel® might have also contributed to pain reduction through a similar mechanism.
We also demonstrated that the pooled BMMSC population are efficient in differentiating into chondrocytes in vitro, and secrete a significant amount of sGAG (Fig.
2c). When these cells were administered intra-articularly into OA-affected joints, we observed a progressive increase in proteoglycan staining. The improvement in cartilage repair was observed both macroscopically and microscopically. The sGAG intensity data revealed that the total proteoglycan content was significantly higher in both the cell + HA treated groups compared to animals treated only with HA. One of the short comings of the preclinical results is that we did not determine the therapeutic effect of BMMSCs without HA. However, based on the published data it appears that administration of MSCs in combination with HA provided better therapeutic benefit than either HA or MSC treatment alone in an experimental animal model of OA [
30]. The concomitant reduction in MIA-induced pain followed by an increase in cartilage regeneration observed in this study suggests that human bioactive factors synthesized by BMMSCs may be responsible for both the reduction in inflammation and promotion of endogenous cartilage regeneration via a paracrine mechanism [
12].
This clinical study met its predefined endpoint of safety of intra-articular administration of Stempeucel® in osteoarthritis of the knee joint. Adverse events were predominantly local pain and swelling, particularly seen in patients randomized to the higher dose groups (75 M and 150 M) and they resolved completely upon symptomatic treatment. There was no evidence of ectopic tissue or tumor formation locally at 1-year follow-up. Hematological, biochemical, and serological parameters were comparable in both the cell and placebo arm in all groups of patients. Limited joint space, higher dose, and volume of injection (6 ml) may be the reason for increased joint swelling and pain seen in cohort 2 (75 M and 150 M). Furthermore, it can be assumed that a proportion of the cells injected into the joint space have not survived and this phenomenon was more pronounced with higher cell doses. Probably, such non-viable cells produce an inflammatory reaction causing pain and swelling, as reported earlier [
39]. The frequency of these complications was similar to a report from another study using culture-expanded bone marrow-derived MSCs [
40]. In another study using allogeneic non-HLA matched BMMSCs in two different doses (50 and 150 million cells) which were pre-mixed with hyaluronic acid (5 ml) and administered in partial medial meniscectomy patients [
10], the adverse events were similar to those seen in our study, with the most frequently reported AE by system organ class being musculoskeletal and connective tissue disorders [
10]; however, the adverse events did not differ between the two doses tested. Recently, Vega et al. have conducted a study using IA injection of allogeneic BMMSCs (40 million cells suspended in 8 ml of Ringer-Lactate) in OA of the knee joint [
11]. Post-implantation pain was observed in 53% to 60% of patients in both the experimental and control groups. The pain responded to analgesics and improved within 1 to 6 days. Hence, pain and local swelling may be the most common post-injection complication in patients after IA injection of MSCs which responds within a few days of symptomatic treatment.
One of the most important factors influencing the clinical outcome of a study is to determine the optimal treatment dose. In this study, patients in the low-dose group (25 million cells) showed improved outcomes in the pain measurement scores, whereas those in the higher dose groups did not. The VAS and WOMAC composite index scores decreased by 64% and 64.4% in the 25-million-cell arm as compared to 36% and 49.3% in the active controls with HA, respectively, at 12 months follow-up. In a proof of concept study, three doses of autologous adipose tissue-derived MSCs (AD-MSCs) were used: 10 million, 50 million, and 100 million cells. The WOMAC score improved at 6 months follow-up in the high-dose group [
14]. In another study using allogeneic BMMSCs at a dose of 40 million cells, improvement in pain, disability, quality of life, and cartilage quality by MRI was noted in the cell-treated group [
11]. Several reasons are hypothesized for this effect in the low-dose group of patients as observed in this study. Firstly, a dose of 25 M cells may be optimum with the volume of hyaluronic acid (2 ml) used in the study as a supporting matrix. Secondly, the 25-million-cell dose maybe optimal for the limited IA space in the knee joint. Thirdly, doses higher than 25 million might cause cell aggregation due to a high cell concentration or insufficient space in the knee joint and subsequently cause cell death. Fourthly, the 25-million-cell dose may be lying in the upper range of the efficacy dose since numerous studies reports that doses in the range of 10 to 25 million BMSCs may be efficacious in OA of the knee joint [
15,
41‐
45]. Finally, higher doses of MSCs may activate the MSCs to function as an M1-type cell with a pro-inflammatory response [
46], whereas the 25 M dose may be the optimal concentration of cells which gives rise to an M2-type MSC with an anti-inflammatory/immunosuppressive response. Hence, which cell dose will lead to the best outcome cannot be determined until a series of dose-finding studies are carried out.
Various studies are ongoing to determine the optimal tissue source of MSCs for therapeutic repair of the cartilage tissue. The combination of MSCs with scaffolds, growth factors, platelet-rich plasma (PRP), and genetic modification have also been studied. It is not clear which source of stem cells, or a combination product, will be the best for the disease condition. Studies have shown that adipose tissue-derived stem cells are both safe and efficacious [
13,
14,
47‐
49], whereas other studies have shown that bone marrow-derived cells are equally efficacious [
10,
11,
50‐
52]. A current focus for knee cartilage repair is to use scaffolds that provide a three-dimensional environment for guiding and supporting the cells for cartilage repair. An advantage for using a scaffold is containment of the implanted cells on the lesion, and these biomaterials may act as barriers for fibroblast invasion of the graft [
53,
54]. Koh et al. have used PRP as a scaffold as it acts as an MSC accelerator for clinical chondrogenesis, is non-immunogeneic and bioabsorbable, and can be easily prepared preoperatively [
13]. In another study, fibrin glue has been used as a scaffold in MSC implantation to induce improved cell survival, proliferation, gene expression, differentiation, and matrix synthesis leading to repair of the cartilage lesion [
55]. Cartistem® (MEDIPOST Co. Ltd., South Korea) is a combination product of human umbilical cord blood-derived mesenchymal stem cells and hyaluronic acid [
56]. This acts as a biodegradable matrix in MSC implantation as it facilitates the migration and adherence of cells to the damaged cartilage, leading to better healing of the damaged lesion. Hence, more studies are required to pinpoint the best source of stem cells and the scaffold to be used to demonstrate both safety and efficacy.
The method of delivery of cells—either by direct intra-articular injection or by open arthroscopy injection—into the joint cavity is also important and may be one of the factors for deriving efficacy. In one of the initial studies, Wakitani et al. transplanted cells of bone marrow embedded in collagen gel into the articular cartilage defect at the time of high tibial osteotomy [
43]. Cartistem®, a combination product approved by the Korean FDA, has been applied to the damaged area through arthroscopy after conducting a microfracture [
57]. These open surgical methods have their disadvantages such as pain, longer hospital stay, and higher cost. Minimally invasive techniques such as intra-articular injection have been adopted by different groups [
14,
15,
41,
45,
50]. IA injection is patient-friendly in terms of being less invasive, with reduced hospital stay, and are likely to reach a larger patient population as it can be performed in peripheral hospitals. Ultrasound guidance of knee injections could be a better option to more precisely deliver the cells intra-articularly. Berkoff et al. have reported that ultrasound guidance of knee injections resulted in better IA accuracy of needle placement than anatomical guidance (95.8% versus 77.8%;
P < 0.001) [
58]. This enhanced injection accuracy achieved with ultrasound needle guidance directly improves patient-related clinical outcomes. However, in developing countries, ultrasound-guided intra-articular injection may be a challenge due to limited access to the instrument.
The present study, though it has shown good subjective improvement in pain and functional scores, did not demonstrate improvement in cartilage signal and morphology by MRI. We have used the WORMS scoring system, which is a semiquantitative MRI system for evaluating structural change in knee OA. WORMS scoring has been extensively studied for the prevalence and severity of cartilage loss, bone marrow lesions, and meniscal damage [
59,
60], and has seldom been studied for cartilage regeneration. Koh et al. studied the effect of adipose-derived MSCs with PRP in OA of the knee joint and found that WORMS score significantly improved from 60.0 points to 48.3 points and cartilage subscore improved from 28.3 points to 21.7 points at 24 months follow-up (
P < 0.001) as compared to baseline [
13]. However, in our study, the cartilage subscore did not demonstrate any significant worsening or improvement of the cartilage in any of the subgroups. The reason for the WORMS score differences between these two studies could be due to several reasons: the type of MSCs used are different, better complementarity between adipose-derived MSC and PRP, or the length of follow-up time (24 months vs. 12 months) after cell administration. The limited number of patients used in this study for MRI analysis might have contributed as well. Regardless of these differences, Stempeucel® administration in the preclinical model clearly suggested the ability of these cells in combination with HA to trigger adequate proteoglycan synthesis for cartilage repair. Future clinical trials of Stempeucel® in OA patients should consider using guided delivery of cells in and around the lesion site or by arthroscopy, followed by MRI measurements using delayed gadolinium-enhanced MRI of cartilage (dGEMRIC) or T2 mapping to perform compositional (sGAG) analysis of the cartilage before and after cell administration. One of the limitations of this study was unblinding of the trial after 6 months follow-up, particularly given that the subjective measurements of clinical data (VAS, WOMAC, and ICOAP) were the secondary endpoints.
Acknowledgments
This study was funded by Stempeutics Research Pvt. Ltd.