Protecting the regenerative environment: selecting the optimal delivery vehicle for cartilage repair—a narrative review

Focal cartilage defects are common in youth and older adults, cause significant morbidity and constitute a major risk factor for developing osteoarthritis (OA). OA is the most common musculoskeletal (MSK) disease worldwide, resulting in pain, stiffness, loss of function, and is currently irreversible. Research into the optimal regenerative approach and methods in the setting of either focal cartilage defects and/or OA holds to the ideal of resolving both diseases. The two fundamentals required for cartilage regenerative treatment are 1) the biological element contributing to the regeneration (e.g., direct application of stem cells, or of an exogenous secretome), and 2) the vehicle by which the biological element is suspended and delivered. The vehicle provides support to the regenerative process by providing a protective environment, a structure that allows cell adherence and migration, and a source of growth and regenerative factors that can activate and sustain regeneration. Models of cartilage diseases include osteochondral defect (OCD) (which usually involve one focal lesion), or OA (which involves a more diffuse articular cartilage loss). Given the differing nature of these models, the optimal regenerative strategy to treat different cartilage diseases may not be universal. This could potentially impact the translatability of a successful approach in one condition to that of the other. An analogy would be the repair of a pothole (OCD) versus repaving the entire road (OA). In this narrative review, we explore the existing literature evaluating cartilage regeneration approaches for OCD and OA in animal then in human studies and the vehicles used for each of these two conditions. We then highlight strengths and challenges faced by the different approaches presented and discuss what might constitute the optimal cartilage regenerative delivery vehicle for clinical cartilage regeneration.


Introduction
Focal cartilage defects are common and cause significant morbidity, including pain, swelling, mechanical symptoms (such as a feeling of joint instability), and functional disability (Mandelbaum et al., 1998;Liu et al., 2022).Over a quarter of asymptomatic middle-aged individuals with no family history of cartilage loss were found to have some degree of focal cartilage defect; those with a family history of knee osteoarthritis (OA) demonstrated a cartilage defect prevalence of >50% (Ding et al., 2007;Everhart et al., 2019).In those with symptomatic knee OA, cartilage defects have a prevalence of >80% (Ding et al., 2007;Everhart et al., 2019).Correspondingly, cartilage defects represent an underrecognized, yet major risk factor for the development of OA (Mandelbaum et al., 1998;Arøen et al., 2004;Cicuttini et al., 2005;Ding et al., 2007;Guermazi et al., 2017;Everhart et al., 2019;Liu et al., 2022).While loss of the articular cartilage is often described as the hallmark of OA, OA is a total joint disease also affecting bone, synovium, meniscus and other intra-and extra-articular tissues such as ligament and muscle (Barr et al., 2015;Martel-Pelletier et al., 2016;Cunha et al., 2019;Hunter and Bierma-Zeinstra, 2019).Pain, loss of function, joint stiffness and reduced quality of life are welldescribed sequelae of OA (Barr et al., 2015;Martel-Pelletier et al., 2016;Hunter and Bierma-Zeinstra, 2019;Campbell and McGonagle, 2020).Additionally, while cartilage defects are a risk factor for OA (Martel-Pelletier et al., 2016;Hunter and Bierma-Zeinstra, 2019), OA can occur independently of a pre-established focal cartilage defect (Ding et al., 2007;Everhart et al., 2019).In addition, detection of focal cartilage defects requires advanced imaging (e.g., MRI), and may not be appreciated until more advanced stages.As such, patients will often present for medical treatment at more advanced stages of cartilage degeneration and joint dysfunction, with OA already established (Finley et al., 2018).The optimal method of regenerating cartilage for both focal cartilage defects and OA is actively researched (Clinicaltrials.gov, 2023) with the hope of discovering a treatment that could resolve both pathologies.
Stem cells have been delivered to cartilage lesions (both focal cartilage defects and OA) in the clinical setting before basic or translational evidence supported the regenerative efficacy of the various approaches (Marks et al., 2017;Canada, 2019;ISfSC, 2021).Stem cell delivery was rapidly extended to clinical practice, often at high cost to the patients (Marks et al., 2017;Canada, 2019;ISfSC, 2021).These two factors led national healthcare governing bodies to step in and limit this practice (Marks et al., 2017;Canada, 2019;ISfSC, 2021).The upside was an expansion of the basic science and translational research in this area to build a more solid foundation upon which to eventually base clinical regenerative treatment.Two major components of this regenerative treatment include 1) the biological element contributing to the regeneration (e.g., stem cells, extravesicular vesicles), and 2) the vehicle by which these biological elements are delivered (e.g., organic fluids, engineered scaffolds) (Campbell et al., 2022).Proposed advantages of using stem cells or EVs as the biological element have been reviewed previously (Kim et al., 2020;Taghiyar et al., 2021;Campbell et al., 2022) and include a direct ability to repair damaged cartilage (for stem cells), immunomodulation capacity to reduce inflammation within the OA joint, as well as paracrine effects that recruit other endogenous regenerative cells to the site of injury.While these potential advantages have been described for stem cells and EVs, there is currently no consensus regarding which method is more advantageous.The vehicle should be considered equally important to the biological element as it supports the regenerative process.A protective 3-dimensional (3D) environment may preserve stem cells' ability to differentiate; a structure may allow cell adherence and migration; a source of growth and regenerative factors may activate and sustain regeneration (Kim et al., 2019).
Various animal models of cartilage injury have been described, such as: focal osteochondral defects (OCD; a localized osteochondral defect surgically imposed through the articular surface) and various OA models such as the destabilization of the medial meniscus (DMM) model, the anterior cruciate ligament transection (ACLT) model, and chemical models for generalized OA-like cartilage loss; these models are described in detail by Cully (2015) (Culley et al., 2015) and Cope (2019) (Cope et al., 2019).OCD versus OA models cause different sizes and distributions of cartilage injuries, and potentially differing endogenous intraarticular responses.Consequently, the model may impact the optimal regenerative delivery vehicle selected to the specific injury or condition in need of repair when moving from translational to clinical therapy (Liu et al., 2022).
In this narrative review, we evaluated the existing literature on cartilage regeneration both focal OCDs and OA in both animal models and clinical trials.Since the optimal stem cell population (Campbell et al., 2022), or use of their secretome (Kim et al., 2020;Taghiyar et al., 2021), for cartilage regeneration has been reviewed elsewhere, we focus here on the vehicles used in these studies and consider both the context in which they were used, as well as their general applicability to focal OCDs and OA.We conclude by looking towards the application of scaffold technology in clinical trials.

Animal models of focal osteochondral defects
The last decade saw advancements in vehicle engineering to treat focal OCDs in animal models (Lam et al., 2014;Shimomura et al., 2014;Li et al., 2015a;Dashtdar et al., 2015;Ha et al., 2015;Goodrich et al., 2016;Park et al., 2017;Ruan et al., 2017;Li et al., 2018;Lu et al., 2018;Muttigi et al., 2018;Ye et al., 2018;Jia et al., 2019;Jin et al., 2019;Kondo et al., 2019;Park et al., 2019;Pascual-Garrido et al., 2019;Wang et al., 2019;Xu et al., 2019;Cho et al., 2020;Wu et al., 2020;You et al., 2020;Deng et al., 2021;Heirani-Tabasi et al., 2021;Murata et al., 2022;Zheng et al., 2022) (Table 1).In the rat, intraarticular needle injection delivered the vehicle, such as hyaluronic acid (HA) hydrogels (Li et al., 2018;Xu et al., 2019).While direct placement of the vehicle into the OCD, a more precise delivery of the vehicle is limited by the technical challenges of placing a pre-formed scaffold in a small lesion (~1.5-2 mm in width and <1 mm in depth in rats, smaller in mice).Two studies reported improved histologicbased cartilage scoring outcomes after implanting a gel or cell pellet scaffold into the OCD of the rat knee (Xu et al., 2019;Zheng et al., 2022).The most common model however remains the rabbit knee OCD (Table 1); providing a relatively large joint for surgical scaffold implantation handled with the naked eye and relatively lower cost (versus dog or horse, for example,).In the mid-2010s, vehicles used in rabbit OCD models included bilayered hydrogels that crosslinked at body temperature (Lam et al., 2014), synthetic hydroxyapatite (Shimomura et al., 2014) and Polyvinyl alcoholchitosan composite hydrogel (Dashtdar et al., 2015), implanted during open knee surgery.Nanofiber scaffolds, composed of ultra-fine biodegradable polymeric fibers that simulate the natural extracellular matrix (ECM), have also emerged in animal OCD studies (Kazemnejad et al., 2017;Ahmadian et al., 2023).Nanofiber scaffolds may be constructed using different techniques, including electrospinning, self-assembly, phase separation, and drawing (Ahmadian et al., 2023).Natural nanofiber hydrogel scaffolds (scaffolds with high water content and elastic properties with cross-linked, multiporous networks (Li et al., 2019)) include alginate, gelatin, agarose, HA, fibrin, and collagen, while synthetic scaffolds often contain polycaprolactone, polyethylene glycol, polyurethane, poly (pdioxanone), and/or poly (lactic acid) (Kazemnejad et al., 2017;Ahmadian et al., 2023).For the interested reader, reviews describing the advantages of nanofiber scaffold technology are available (Coburn et al., 2012;Kazemnejad et al., 2017;Ahmadian et al., 2023).Since the evolution of nanofiber technology, more sophisticated vehicles have been bioprinted that incorporated distinct osteogenic and chondrogenic regions or zones matching the anatomy of the predefined OCD (Yilmaz et al., 2020).For example, 2D-nanopattern chondroitin sulfate-coated scaffolds featured an inverse pattern of nanograting (250-nm line, 250-nm space, and 150-nm height) or nanohole (225-nm diameter, 400-nm pitch, and 300-nm height) patterning with cells delivered as a bilayered construct (Wu et al., 2020).In doing so, it was hypothesized that stem cells differentiated on specific nanotopographic patterns based on a zonally stratified cartilage construct would possess topographically induced mechanical memory, and could thus retain an induced chondrogenic phenotype (Wu et al., 2020).Correspondingly, these constructs included layer-defined pre-differentiation of incorporated mesenchymal stem cells (MSCs) to enhance their osteogenic and/or chondrogenic abilities, depending on where they were implanted (subchondral bone or articular surface) (Wu et al., 2020).In addition to stem cells, various growth factors have also been incorporated into these scaffolds to enhance anabolic cellular activity.For example, a biphasic scaffold utilizing gelatin methacryloyl (GM) in one phase and parathyroid hormone (PTH) covalently immobilized on silk fibroin (SF) with methacrylic anhydride (MA) in a second resulted in a corresponding biomechanical gradient due to the differing physical properties of the constituents (Deng et al., 2021).Benefits of potential growth factor use, such as insulin-like growth factor, fibroblast growth factors, bone morphogenic proteins and transforming growth factor beta, for cartilage regeneration have recently been reviewed (Hernigou et al., 2022;Trippel, 2023); however a growth factor that robustly augments articular cartilage healing remains elusive (Trippel, 2023).
Another interesting approach to OCD repair was injecting a mimetic hydrogel composed of polyethylene glycol, chondroitin sulfate, a matrix metalloprotein-2 degradable peptide crosslinker, and arginyl-glycyl-aspartic acid (Pascual-Garrido et al., 2019).This scaffold allowed cells encapsulation during the hydrogel formation process, with the final combination being formed in situ within the articular chondral defect, then cross-linked to maintain shape (Pascual-Garrido et al., 2019).In this study, the authors included the matrix metalloprotein-2 degradable peptide cross-linking photoinitiator (CVPLSLYSGC) that polymerized under blue light.After the defect was generated, a red light in the operating room allowed defect visualization and polymer delivery.Polymer solution was then injected onto the chondral defect, then photopolymerization was performed in situ by use of a 405-nm blue light in order to establish the final placing of the customized regenerative material (Pascual-Garrido et al., 2019).The majority of small animal OCD studies reported positive structural outcomes following stem cells delivery.
Larger animals studies of OCD have included the use of HA hydrogel (Ha et al., 2015) and hanging drop MSC aggregates (Kondo et al., 2019) in minipigs, as well as platelet-enhanced fibrin (Goodrich et al., 2016) and 3D-based microspheroids (Murata et al., 2022) in horses.These proof-of-concept studies (Table 1) often reported positive structural outcomes following stem cell application.
Overall, scaffolds used over the last decade evaluating the treatment of OCDs in animal models have been roughly evenly divided between engineered hydrogels (Lam et al., 2014;Ha et al., 2015;Park et al., 2017;Li et al., 2018;Lu et al., 2018;Jia et al., 2019;Pascual-Garrido et al., 2019;Wang et al., 2019;Xu et al., 2019;Cho et al., 2020), engineered nanofiber products (Shimomura et al., 2014;Li et al., 2015a;Dashtdar et al., 2015;Ruan et al., 2017;Jin et al., 2019;Kondo et al., 2019;Wu et al., 2020;Deng et al., 2021;Murata et al., 2022), and biological products (Goodrich et al., 2016;Muttigi et al., 2018;Ye et al., 2018;Park et al., 2019;You et al., 2020;Heirani-Tabasi et al., 2021;Zheng et al., 2022) (Figure 1).For details regarding structural outcomes following treatment delivery in both small and large animal OCD models, readers are directed to recent reviews by Jiang (Jiang et al., 2021) and Liu (Liu et al., 2022).In a narrative review, Jiang outlined that the current biomaterials and scaffolds used for the delivery of stem cells still have barriers to overcome, including the biocompatibility and degradation effects of polymer materials, as well as a need for better understanding regarding the effect of the degrading scaffold on stem cell activity and functionality in the body (Jiang et al., 2021).Though the majority of studies included in the review demonstrated improvement in structural outcomes, screening for the most regenerative subgroups of stem cells and preventing premature stem cell differentiation were also recommended (Jiang et al., 2021).Liu performed a narrative review summarizing MSC therapies for articular cartilage regeneration in large animals (Liu et al., 2022).Liu concluded that, given the biomechanical/anatomical differences and potent healing capacity of small animals, preclinical studies should utilize large animal models to adequately develop translatable therapeutics to humans, and that there was no clear consensus or standard regarding critical aspects of MSC therapy (e.g., cell concentration at implantation, need for pre-differentiation) for large animal cartilage regeneration (Liu et al., 2022).
It is critical to note that neither small nor large OA animal models used the more advanced tissue engineering stem cell delivery methods (Figure 1).This is likely owing to the diffuse nature of articular cartilage loss in the OA joint, making it difficult to engineer or "print" a scaffold that could be applied the non-uniform and diffusely compromised articular surface.
As with the OA clinical trials, the vast majority of the studies reported that the interventions led to positive clinical outcomes.Many of these studies reported improved cartilage structural outcomes assessed by MRI.Summary of these outcomes for clinical OCD studies are presented by Jiang (Jiang et al., 2021) and Lv (Lv et al., 2023).In a narrative review, Lv concluded that while MSCs have therapeutic value, a stronger evidence base is needed until they can be clinically applied (Lv et al., 2023).It was also noted that injecting MSCs has significant limitations, including cell death at injection and significant leakage at the injection site and that overcoming these limitations via biomaterial-assisted cell cartilage repair may be the optimal direction for OA treatment (Lv et al., 2023).Like small and large animal models of OA, randomized controlled trials evaluating stem use for the treatment of OA in humans (Vega et al., 2015;Gupta et al., 2016;Lamo-Espinosa et al., 2016;Goncars et al., 2017;Shapiro et al., 2017;Emadedin et al., 2018;Freitag et al., 2019;Lee et al., 2019;Lu et al., 2019;Matas et al., 2019;Anz et al., 2020;Bastos et al., 2020;Garza et al., 2020;Lamo-Espinosa et al., 2020;Chen et al., 2021) included a broad spectrum of OA severity (KL grades 1-4 included) and have largely made use of fluid-based vehicles such as saline (Goncars et al., 2017;Emadedin et al., 2018;Freitag et al., 2019;Lee et al., 2019;Matas et al., 2019), Ringer's lactate (Vega et al., 2015;Garza et al., 2020), proprietary electrolyte solutions (Gupta et al., 2016;Lu et al., 2019;Chen et al., 2021), HA (Lamo-Espinosa et al., 2016), or PRP (Anz et al., 2020;Bastos et al., 2020;Lamo-Espinosa et al., 2020) (Table 5).Stem cell delivery was often performed by intra-articular injection, with or without imaging guidance.These studies generally reported no major adverse events with the treatment provided.A majority of trials reported improved clinical outcomes, though no study reported improved structural outcomes.Clinical outcomes reported by these trials were summarized by Jiang (Jiang et al., 2021) and Lv (Lv et al., 2023) in the reviews noted above.

Lessons learned from the current literature
Emerging technologies, including 3D printing of biological scaffolds, allow the standardized computer-assisted creation of delivery vehicles with precise control over composition, pore size, stiffness and geometry, reviewed in detail by Kim (2019) (Kim et al., 2019) and Lafuente-Merchan (2022) (Lafuente-Merchan et al., 2022).These differing 3D environments influence migration, cellcell integrin interactions, signaling and phenotype of the cells encapsulated within the scaffold (Kim et al., 2019;Lafuente-Merchan et al., 2022).These regenerative cells, as well as growth factors important for regenerative cell viability, proliferation and differentiation can be embedded in the scaffold, adapted specifically to the tissue requiring repair (Kim et al., 2019;Campbell, 2020).Solubility of the scaffold can be adjusted to emphasize anabolic paracrine functionality of incorporated cells as well as growth factor release rate (Murata et al., 2022).Overall, it is anticipated that these combined advantages will provide superior cartilage regeneration capacity for injured cartilage (Mandrycky et al., 2016;Kim et al., 2019;Lafuente-Merchan et al., 2022).Ultimately, the ideal regenerative vehicle could treat all forms of cartilage injury, including OCDs and OA.The development of such vehicles was unfortunately delayed by the clinical use of stem cells for cartilage regeneration prior to establishing a strong evidence base (Marks et al., 2017;Canada, 2019;ISfSC, 2021).Formal regulation of stem cell application in the clinical setting created considerable movement forward towards this goal of applying tissue engineering technology to cartilage regeneration.Based on the available literature, advanced multi-layered scaffolds with molecular and cellular composition gradients optimized for the tissues to be repaired are applied to animal models of OCD (Wu et al., 2020;Deng et al., 2021;Murata et al., 2022).Implantation of nanopatterned differentiated MSCs as stratified bilayered hydrogel constructs improved the repair quality of cartilage defects, as indicated by histological scoring, mechanical properties, and polarized microscopy analysis in rabbit (Wu et al., 2020).A GM + SF-PTH/GM + SF-MA osteochondral biphasic scaffold promoted the regeneration of OCDs in rabbit and maintained hyaline cartilage phenotype, as evaluated by histologic staining and immunohistochemistry (Deng et al., 2021).Synovial fluid-derived MSCs contained within 3D spheroids reduced the sizes of OCDs in an equine model as seen on MRI and the histology-based ICRS scale (Murata et al., 2022).This technology has been piloted in clinical studies of OCD (Abdelhamid et al., 2023), with positive structural results (Jiang et al., 2021;Lv et al., 2023).Given the large number of individuals affected by OCD, the research advances represent an important step forward (Mandelbaum et al., 1998;Ding et al., 2007;Everhart et al., 2019;Liu et al., 2022).Presently, however applying these technologies to OA, the most common musculoskeletal disease worldwide, faces more challenges (Barr et al., 2015;Martel-Pelletier et al., 2016;Hunter and Bierma-Zeinstra, 2019;Campbell and McGonagle, 2020).One of the challenges presented by the OA joint is the heterogeneous quality of diffuse articular cartilage damage.Studies treating OCD have the advantage of "filling" a lesion of pre-defined dimensions (in the case of animal models), or that is limited in its dimensions (in the case of clinical OCDs).In OA, the articular damage is often less delineated, making it difficult to pre-fabricate a scaffold that will treat the entire articular surface.An analogy would be the repair of a pothole (OCD) versus repaving the entire road (OA).Other challenges in OA regenerative treatment includes the non-chondrogenic contributors to the OA joint (e.g., subchondral bone changes, synovial hypertrophy, altered forces across the joint due to abnormal joint alignment, etc.) that must also be considered (McGonagle et al., 2017).

Future directions
Scaffold engineering advances for OCD are progressing rapidly with ever-more sophisticated methods of delivery.For those with OA, current strategies to optimize the delivery vehicle have included the use of existing biological fluids such as PRP, or synthesized biological material such as HA.PRP has been cited to reduce the occurrence of inflammation, improve angiogenesis, and promote the proliferation and differentiation of chondrocytes by secreting cytokines, chemokines, and growth factors, as well as exosomes with similar effects, so as to promote the healing of bone and cartilage injuries (Liang et al., 2022).In clinical studies, patients with knee OA treated with PRP showed improvement in physical function and pain, suggesting that intra-articular injection of PRP may be a potential therapeutic strategy for relieving knee pain (Liang  , 2022;Riewruja et al., 2022).HA is a naturally-occurring, nontoxic, biodegradable biopolymer essential for bone growth and chondrocyte differentiation, and has been reported to stimulate stem cell chondrogenesis as well as cartilage-specific ECM production by stem cells (Matsumoto et al., 2009;Yoon et al., 2011;Toh et al., 2012;Davoudi et al., 2018;Park et al., 2020).HA may also provide a protective milieu for the cells in the damaged joint surface environment and ensure cells remain localized to cartilage injury (Li et al., 2015b;Campbell, 2020).HA is used clinically to treat OA via intra-articular application (Chavda et al., 2022).For OA, vehicles such as PRP and HA may represent a compromise between form and function, with custom and commercial versions of each being developed towards regenerative optimization (Gato-Calvo et al., 2019;Chavda et al., 2022).Ideally, tissue engineering advances for cartilage repair will 1 day be applicable to the OA joint.Possible solutions for matching OA cartilage loss may include joint surface modeling using CT or MRI that would allow the printing of a cartilage scaffold to be overlain on the damaged articular surface.Such an approach is currently being evaluated for the custom manufacturing of total knee arthroplasty components (Beckmann et al., 2020;Moret et al., 2021), but might also be applicable to cartilage regeneration.Applying a fluidic vehicle that will polymerize over the articular surface at body temperature or with another catalyzing stimulus, such as specific wavelengths of light, could secure a regenerating agent in place while permitting the supportive benefits of the scaffold.Concurrent treatment, such as joint distraction or alignment correction (McGonagle et al., 2017;Jansen and Mastbergen, 2022) may be of benefit when applying cartilage regeneration treatments to the OA joint in order to reduce the potential for damage to the treatment components during the regeneration process.

Conclusion
Regenerative stem cell delivery for cartilage repair has advanced considerably over the last decade with the most impressive advances incorporating both structural and biological features designed to enhance cartilage repair.At present these advances are being directed towards focal OCDs in animal models, foreshadowing competition with popular biological products such as HA and PRP.While the application of such technologies to OCD in the clinical setting is likely just around the corner, whether their application can be adapted to the more diffusely damaged articular surface of the common and debilitating OA joint is yet to be seen, and whether they prove superior to biological agents such as HA or PRP in the OA setting is yet to be demonstrated.

TABLE 1
Small and large animal osteochondral defect studies and associated injection vehicle.

TABLE 1 (
Continued) Small and large animal osteochondral defect studies and associated injection vehicle.

TABLE 2
Small animal models of OA and associated injection vehicle.
(Continued on following page)Frontiers in Bioengineering and Biotechnology frontiersin.org (Table 4) often included individuals with isolated cartilage lesions of International Cartilage Repair Society (ICRS)

TABLE 2 (
Continued) Small animal models of OA and associated injection vehicle.

TABLE 3
Large animal models of OA and associated injection vehicle.

TABLE 4
Clinical studies using stem cells for isolated chondral lesions and associated injection vehicle.

TABLE 5
Human randomized controlled trials using stem cells for knee OA and associated injection vehicle.