Safety, tolerability, clinical, and joint structural outcomes of a single intra-articular injection of allogeneic mesenchymal precursor cells in patients following anterior cruciate ligament reconstruction: a controlled double-blind randomised trial

Osteoarthritis (OA) is a widespread debilitating chronic condition that causes joint pain, functional disability, and impaired quality of life. It is a disease with a multifactorial aetiology, with obesity, ageing, occupational, hormonal, and genetic factors considered to be important contributors to its onset and progression [1]. Joint injury is also a well-established risk factor for the development of OA [2]. For approximately 12% of the OA population the condition is considered to arise secondary to a traumatic injury [3], leading to the classification of this group of patients as presenting with post-traumatic OA. Post-traumatic OA affects 5.6 million individuals in the United States with an annual direct medical cost of approximately $3 billion [3]. Anterior cruciate ligament (ACL) rupture is a relatively common injury, particularly during sporting activities. Individuals who sustained an ACL rupture account for an estimated 25% of the overall knee OA population [4, 5]. Approximately 80% of patients with traumatic rupture of the ACL show evidence of radiographic OA within 12–14 years following the injury, despite surgical repair [6, 7]. Moreover, there is no evidence that ACL reconstruction improves knee joint structural outcomes and the onset of OA [8].

Currently used modalities for the treatment of post-traumatic OA address the symptoms that arise subsequent to the initiating injury but, apart from their analgesic and anti-inflammatory activities [9, 10], offer limited value in restoring the joint structural changes induced by the initial traumatic event or diminishing the progression to symptomatic OA. Adult mesenchymal stem cells (MSC) are an abundant source of self-renewing, multipotent, undifferentiated cells that can be readily isolated from bone marrow, adipose tissue, muscle, or synovium and then readily culture expanded without undergoing differentiation [11]. The ability of these cells to differentiate into bone, cartilage, adipose, tendon, and other cells of the mesenchymal lineage under appropriate stimuli suggests the potential for the regeneration and repair of injured joint tissues in the clinic [11, 12]. Moreover, MSC exhibit trophic, anti-inflammatory, and immunomodulatory activities that have been shown to attenuate the mediators of joint tissue destruction as well as promoting their repair [11‐14]. As a consequence, MSC have been investigated for their therapeutic potential for the treatment of established OA and other arthropathies [13, 15‐19]. Positive outcomes from these studies have been reported in terms of safety and the modulation of joint tissue breakdown as well as improving joint pain and function, but for the most part the MSC used were autologous, were derived from a variety of sources, and often administered as multiple doses [13, 15‐19]. We are aware of only one randomised double-blind, placebo-controlled clinical trial that investigated the safety and efficacy of allogeneic MSC for the management of post-traumatic OA [19]. In this 24-month trial a single intra-articular injection of allogeneic bone marrow-derived MSC, suspended in hyaluronan (HA), was administered within 7 to 10 days of surgery to 55 patients who has been subjected to partial meniscectomy [19]. Although only 23% of the patients who received the lower intra-articular dose of 50 million MSC + HA showed greater than 15% increase in meniscus volume post-treatment, none of the HA-injected controls achieved this level of meniscal regeneration [19]. Moreover, over the course of the study, no safety concerns were reported, and significant improvements in OA symptoms and decreased progression of magnetic resonance imaging (MRI) indices of disease progression were observed relative to the HA-alone injected group [19].

Stromal tissue within the perivascular niche of the bone marrow contains a heterogeneous population of MSC that may be isolated by ficoll gradient centrifugation, followed by adherence in plastic culture flasks [20]. MSC preparations derived from adult human bone marrow by this means contain low abundance of colony-forming units-fibroblast (CFU-F) and are contaminated with mature stromal cells, non-mesenchymal cell types, and a small population of immature mesenchymal precursor cells (MPCs) [20, 21]. However, MPCs can be separated from the MSC and other stromal elements by magnetic-activated cell sorting using specific monoclonal antibodies that bind to antigens expressed on the surface of the MPCs [22, 23]. Examples of these antibodies include: STRO-1, STRO-3, STRO-4 (HSP-90b), VCAM-1 (CD106), and CD146 [23‐26]. The MPCs isolated using these antibodies for immuno-selection lack the phenotypic characteristics of mature stromal elements and include the majority of the CFU-F and plasticity present in the plastic-adherent bone marrow MSC population [23, 25, 27]. Previous studies utilising immune-selected STRO-1⁺ or STRO-3⁺ MPCs demonstrated that they exhibited superior clonogenicity, proliferative efficiency, and maintained their immunophenotype following culture expansion compared to plastic-adherent MSC isolated from the same bone marrow aspirates [27‐29].

Our previous preclinical studies using ovine models of post-traumatic OA [30] and collagen-induced arthritis [31, 32] have demonstrated the safety and efficacy of allogeneic STRO-3⁺-selected MPCs in preserving synovial joint articular cartilage, downregulating pro-inflammatory cytokine activities, and preserving cartilage and endothelial cell function. These encouraging preclinical findings using allogeneic STRO-3⁺ MPCs provided the rationale for our hypothesis that these cells were safe and might provide long-term clinical and joint structural benefits when administered to patients after ACL reconstruction. Accordingly, the primary aims of the present study were to evaluate the safety and tolerability over 24 months of a single intra-articular injection of 75 million allogeneic MPCs suspended in HA compared to injection of HA alone when administered to patients post-ACL rupture and surgical reconstruction. A secondary objective was to explore the efficacy of these agents in improving clinical outcomes, joint cartilage integrity, and subchondral bone remodelling over 24 months using radiographs and MRI.

This study is the first to evaluate the therapeutic effects of a single intra-articular injection of allogeneic immuno-selected MPCs on the long-term (2 years) outcomes of knee joint symptoms, cartilage, and subchondral bone changes post-ACL reconstruction. The study showed that intra-articular administration of allogeneic MPCs was safe and well tolerated over the 24-month study period following a single injection. Moreover, using validated clinical instruments under blinded conditions, there was evidence for improvements in pain, function, and quality of daily life, as well as favourable effects on knee joint structural changes over 24 months, supporting further investigation of MPCs as potential disease-modifying agents in post-traumatic knee OA.

In terms of both safety and efficacy our findings were consistent with previous studies using culture-expanded autologous bone marrow-derived MSC administered via the intra-articular route for the management of OA and a variety of other joint problems [16, 19, 40‐44]. Over our 2-year study no cell-related serious adverse events or abnormalities in laboratory parameters or clinical signs were observed.

Although the patient group enrolled in our study were young adults with a recent ACL reconstruction who exhibited no radiographic signs of OA, they reported persistent joint pain at study entry with the MPC + HA group exhibiting lower (worse) KOOS pain, symptom, and ADL scores than the HA group at baseline. Over the course of the study both MPC + HA- and HA-injected groups showed symptomatic improvement; however, the MPC + HA-treated group exhibited greater changes from baseline.

Although the primary aim of the study was to examine the safety and tolerability of the test substances, it also demonstrated favourable effects of MPCs on knee structural outcomes relative to the HA-alone group. We found that tibial bone expansion was halted in the MPC + HA arm (0.5 ± 2.4%), while there was a significant bone expansion in the HA alone group (4.0 ± 2.3%) over 6 months, with a non-statistically significant trend observed at 12 and 24 months. Although we did not observe a statistically significant effect of MPCs on knee cartilage, there was a trend for slowing in cartilage volume loss in the MPC + HA group (0.7 ± 5.9%) compared to the HA-only group (–4.0 ± 3.9%) over 6 months. This was supported by the finding of reduced joint space narrowing in the MPC + HA group compared to the HA-only group. Recent studies have shown that the earliest changes at the knee post-trauma are seen at the subchondral bone [45]. Hunter and colleagues showed that both ACL injury and ACL reconstruction were associated with significant flattening of articulating bone curvature over 5 years [45]. We showed that there was tibial bone expansion at 2 and 4 years post-meniscectomy which predated changes in cartilage volume [46]. Subchondral bone and cartilage are intimately integrated, since the avascular cartilage relies on the integrity of vascularized subchondral bone to remain functional. A greater tibial plateau bone area has been shown to be associated with the classical radiographic hallmarks of OA (osteophyte and joint space narrowing) [47] and cartilage defects [48]. Taken together, these preliminary results suggest that the administration of MPCs in post-ACL reconstruction could slow the rate of disease progression, thereby delaying the onset of post-traumatic OA in later years.

The mechanisms of action responsible for the beneficial effect of MPCs on patient-reported clinical outcomes and radiograph- and MRI-derived structural outcomes are presently unresolved. However, MPCs are known to exhibit anti-inflammatory and immunomodulatory activities [11‐14] which could suppress the injury-induced production of inflammatory mediators within joint tissues that are responsible for the clinical symptoms and breakdown of joint connective tissues [2]. The MPCs have surface receptors for interleukin (IL)-6, tumour necrosis factor (TNF), IL-1, and IL-17, among others, that bind these cytokines when induced at sites of tissue injury and inflammation. The receptor engagement results in secretion by the MPCs of prostaglandin E2 and indoleamine 2,3-dioxygenase which can polarise monocytes to an anti-inflammatory M2 phenotype and T helper 17 cells to FoxP3 T regulatory cells [14, 49]. This pathway results in increased IL-10 levels and reduction in the very inflammatory cytokines that initiated the response [30, 31, 50]. Moreover, co-culture of human synovial and cartilage explants exposed to conditioned media from MSC pre-treated with TNF-alpha showed reduced expression of IL-beta and matrix metalloproteinases and upregulation of the cytokine signalling suppressor 1 gene while, in cartilage, upregulation of IL-1 receptor antagonist and downregulation of the proteoglycan degrading proteinase ADAMTS-5 was observed [51]. MPCs injected intravenously to sheep with collagen-induced joint inflammation was demonstrated to decrease cartilage erosions, synovial stromal cell activation, angiogenesis, and plasma levels of activin A and IL-17A, confirming that systemic administration of these cells can modulate both local joint and systemic inflammation [32].

The present study has limitations. It is a pilot phase Ib/IIa randomised controlled trial with the primary aim being to determine the safety and tolerability of a single intra-articular injection of allogeneic MPCs. The small sample size has limited the power of our study to show significant results for some structural and clinical outcomes such as cartilage volume loss and ADL score. However, it has demonstrated consistent findings between clinical symptoms and knee structural changes. In terms of joint structural effects, we have observed a more pronounced effect of MPCs on subchondral bone remodelling accompanied by a preservative effect on articular cartilage which follows a biologically plausible sequence of effect, particularly after ACL reconstruction. Although there was a higher rate of loss to follow-up (17.6% at 6 months, 29.4% at 12 months, and 41.2% at 24 months), there were no significant differences between the completers and non-completers in terms of baseline characteristics of age, gender, body mass index, tibial cartilage volume, tibial bone area, or joint space width. The strengths of our study included the objective assessment of changes in cartilage volume and bone area over 24 months at four time points which were measured from MRI using validated methods. Both MRI outcome measures have been shown to play a role in the pathogenesis of knee OA [48, 52]. Furthermore, each measurement was independently performed by a trained observer who was blinded to the participant characteristics, the group allocation of the participants, and the sequence of MRIs, with high intra-observer reproducibility.

Intra-articular administration of a single injection of allogeneic immuno-selected MPCs following ACL reconstruction was shown to be safe and well tolerated in this 24-month study. There was also preliminary evidence to support the efficacy of MPCs in improving both symptoms and structural outcomes. These findings suggest that MPCs may modulate some of the pathological processes responsible for the onset and progression of post-traumatic OA and warrant further investigation of their potential as a disease-modifying agent for the treatment of early joint injuries.