Tendon healing in the era of regenerative medicine: literature review

Abstract

Background and Objective

Tendon injuries are common musculoskeletal conditions that affect both athletic and working populations. Although surgical intervention remains the mainstay of treatment for large tendon injuries, conventional approaches often result in suboptimal healing and functional outcomes. Recent evidence has shown that stem cell therapy may play a role in the management of these injuries. This review comprehensively examines the current literature on stem cell applications in tendon regeneration by analyzing both preclinical and clinical evidence. Specifically, we evaluated various stem cell populations, their characteristics, delivery mechanisms, and repair processes. Additionally, we addressed the limitations of stem cell-based therapies while highlighting emerging trends and future research directions in this rapidly evolving field.

Methods

PubMed was searched for articles published in August 2025. Boolean operators “Tendon” OR “Tendon repair” OR “Tendon regeneration” OR “Tendon injury” AND “Scaffold” AND “Secretomes” OR “Exosomes” OR “Stem Cells” were used to search for articles. Inclusion criteria included studies within the last 10 years, performed on humans or animals, written in English, as well as articles considered clinical trials, meta-analyses, randomized controlled trials, reviews, or systematic reviews. We identified approximately 1,800 studies, which were screened for relevance to our topic. Additional reference screenings and targeted searches were performed to identify other relevant studies.

Key Content and Findings

With advancements in regenerative medicine and material science, new solutions, such as the integration of stem cells and growth factors with specialized scaffolds, offer innovative solutions for tendon regeneration. Various stem cell populations, including mesenchymal stem cells (MSCs), tendon-derived stem cells (TDSCs), and perinatal stem cells (PSCs), have demonstrated potential for assisting in tendon repair. However, significant challenges persist regarding ethical considerations, safety protocols, and treatment standardization.

Conclusions

Stem cell therapy represents a promising frontier in tendon healing, with growing preclinical and clinical support for its regenerative efficacy.

Introduction

Tendons play a vital role in the musculoskeletal system by transferring stress between the muscles, bones, and stabilizing joints (1). Unfortunately, these tissues are susceptible to injury, resulting in an estimated 15 million injuries annually, with over 33,000 subsequent reconstructions performed in the United States alone (2,3). Common tendon injuries include damage to the rotator cuff, finger flexor tendons, patellar tendons, and Achilles tendons (4). Common causes of acute tendon injuries include trauma, overuse, and age-related degeneration, with injuries commonly seen in both athletic and workplace settings (1,4). While some tendon injuries can be managed non-operatively, more severe injuries may require surgical intervention, including specialized suture techniques, tendon-to-bone fixation, graft augmentation (with autografts, allografts, or xenografts), or the use of prostheses (5). However, these current techniques lead to the formation of fibrotic scars rather than regenerative healing, resulting in decreased or limited function (6). Current treatment options also have other notable drawbacks, including the risk of infection, graft rejection, donor site morbidity, poor graft integration, and a high rate of recurrent tears/ruptures (2,5,7). Recognizing these drawbacks, there has been a recent push to develop alternative and complementary treatment strategies for tendon injuries (2,5). This study aimed to review the current literature on various stem cells used for tendon regeneration and evaluate their effectiveness in preclinical and clinical studies. Additionally, this study discusses the challenges and limitations of stem cell-based therapies as well as emerging trends and opportunities for future research. We present this article in accordance with the Narrative Review reporting checklist (available at https://atm.amegroups.com/article/view/10.21037/atm-25-124/rc).

Methods

We performed a PubMed search in August 2025 targeting studies looking at the intersection of (I) tendon injuries; (II) stem cell therapies; and (III) innovative, stem cell adjacent therapies used to promote better healing of tendinopathies. We used various Boolean Operators such as “Tendon” OR “Tendon repair” OR “Tendon regeneration” OR “Tendon injury” AND “Scaffold” AND “Secretomes” OR “Exosomes” OR “Stem Cells” to find relevant studies. The search was limited to studies published in the past 10 years, written in English, and involving humans and animals. Additionally, we limited the search to clinical trials, meta-analyses, randomized controlled trials, reviews, and systematic reviews. This resulted in nearly 1,800 studies that were screened for relevance by assessing their abstracts and titles. Finally, reference screening and targeted searches using both PubMed and Google Scholar were performed to identify additional sources for review (Table 1).

Table 1. The search strategy summary.

Items	Specification
Date of search	August 12, 2025
Database and other source searched	PubMed, Google Scholar
Search terms used	“Tendon” OR “Tendon repair” OR “Tendon regeneration” OR “Tendon injury” AND “Scaffold” AND “Secretomes” OR “Exosomes” OR “Stem Cells” to find relevant studies. Filters: past 10 years, English language, human or animals, clinical trials, meta-analyses, randomized controlled trials, reviews, and systematic reviews
Timeframe	2015–2025
Inclusion criteria	Human and animal studies; English language
Selection process	Collaborative selection was performed by our study team based on perceived relevance to our topic

Tendon basics

Tendon anatomy and structure

Tendons are composed of dense extracellular matrix (ECM) and cells (8). Collagen, specifically type I collagen, is the most abundant component of the ECM, comprising roughly 70–80% of the tendon’s dry mass (4). Other less abundant collagen types include types III, V, IX, X, XI, and XII, which have unique purposes (4). Non-collagenous components of the ECM primarily consist of proteoglycans, such as aggrecan and decorin (DCN), and glycoproteins, such as tenascin-C, fibronectin, and elastin, all of which aid in the structure and function (4). The cellular components of tendons include tendon-derived stem cells (TDSCs) and tenocytes, which comprise up to 90–95% of the cellular elements (9). Other cells within the tendon include chondrocytes, synovial cells, and vascular cells (e.g., endothelial and smooth muscle cells) (9). Each tendon is a multilevel structure composed of units surrounded by an epitendon (4). The tendon units themselves are composed of (from largest structural level to smallest) fascicles/fiber bundles, fibers, fibrils, and collagen molecules (4). The tendon blood supply is limited compared to other tissues and comes from three primary sources: the intrinsic systems at the myotendinous junction (MTJ) and osteotendinous junction (OTJ), as well as from the extrinsic system through the paratenon or synovial sheaths (9). Blood vessels from the muscle pass through the MTJ and supply the proximal third of the tendon, while blood supply from the OTJ is sparse and requires communication with the extrinsic system (9). Tendon blood flow decreases with age and increases during exercise but remains limited compared to that of the surrounding tissues (9).

Tendons are connective tissues that attach muscles to bones (10). Tendons transfer tensile stress from the muscle to bone, allowing joint motion and increasing joint stability (4). They can also limit muscle damage by acting as buffers against external forces (9). The densely packed parallel collagen is key to the tendon’s strength as it transfers this stress (4). ECM components, such as tenascin-C, fibronectin, and elastin, allow tendons to stretch and return to their natural length and shape (4). The stress-carrying capacity of tendons depends on the strain rate, with a greater capacity at lower strain rates (4). This is reflected in the stress-strain curve with four regions (Figure 1). In the first region, also known as the toe region, the tendon strain is less than 2% and represents the “stretching out” of the tendon fibrils (4). In the second region, the tendon strain is less than 4% and is depicted by a linear pattern (4). This is the upper limit of the physiological strain in which collagen fibrils are oriented in the direction of the strain (4). The slope of this region is representative of tendon stiffness (4). The third region represents a strain greater than 4%, which results in microscopic tearing of tendon fibers (4). The fourth region represents a strain between 8–10%, which leads to macroscopic tearing of the tendon fibers, ultimately resulting in tendon rupture (4). It is also important to note that different tendons have different strain capacities, owing to their varied structures and functions (4).

Figure 1.

Tendon stress strain curve. The four zones of tendon strain. In zone 3 and 4 microscopic and macroscopic tendon injuries are observed respectively (4). Figure adapted from Robi et al. (2013). The Physiology of Sports Injuries and Repair Processes (DOI: 10.5772/54234) (11). [Created in BioRender. Ogunsola, A. (2025) https://BioRender.com/l10y285].

Natural tendon repair processes

Acute tendon injury triggers a structured healing process consisting of inflammation, proliferation, and remodeling (12). During inflammation, a clot forms to stop bleeding and releases cytokines such as transforming growth factor β (TGF-β), insulin-like growth factor I (IGF-I), and platelet-derived growth factor (PDGF), which promote inflammation, recruit immune cells, and stimulate fibroblast and tenocyte activity (12,13). Neutrophils and macrophages clear debris, while TGF-β and IGF-I regulate collagen and ECM production (12,13). About two days post-injury, the proliferative phase begins, marked by increased ECM formation, fibroblast recruitment, and angiogenesis driven by basic fibroblast growth factor (bFGF) and vascular endothelial growth factor (VEGF) (12,13). Type III collagen, water, and glycosaminoglycan levels rise as macrophages continue to support repair (9,12). The remodeling phase begins between two weeks and two months after injury, with declining cellularity and vascularity, increased type I collagen, and alignment of collagen fibers along stress lines (9,12,13). Over time, the scar tissue becomes less metabolically active and structurally inferior to native tendon, leading to reduced mechanical strength, risk of reinjury, and potential adhesion formation (4-6,12,13).

Treatment challenges and the potential of stem cell-based therapies

Even with the current surgical treatment options, tendon tear injuries still face an extended recovery process. Tendon structure and function remain limited compared to uninjured tendons 1 year after surgical repair (12). There are also added complications that can appear from surgery. Suturing methods can lead to adhesion formation, repair of site gaps, and chondrolysis (14). Tissue graft methods can precipitate donor site morbidity, tissue rejection, disease transmission, and zoonotic transmission depending on the graft type (14). These methods all lead to scar tissues similar to those seen in native regeneration, which are at an elevated risk of re-tear/rupture compared to native healthy tissues (13).

Stem cells are naturally occurring cells that self-replicate and differentiate into other cell types (15). Stem cell-based regeneration for tendon injuries has been studied since 1993 (13). The current theory of using stem cells to treat tendon injury is that doing so would result in a shift towards more cellular regeneration of the injured tendon rather than scar formation by producing various growth factors and cytokines, as well as stem cell differentiation into tenocytes (16). Over the years, several studies have reported promising results in vitro and in vivo using both animal and human models (12,16).

Evaluation of tendon repair and functional outcome

A thorough understanding of how tendon healing outcomes are evaluated is essential for determining the effectiveness of cell-based therapies. Furthermore, the methods used to assess study outcomes may affect the future applicability of stem cell research. Among the studies assessed in this review, a vast array of different mechanisms was used to evaluate functional outcomes. Several studies have performed biomechanical tests, including load-bearing exercises, gait analyses, tensile strength testing, range of motion testing, and flexibility testing. In addition, multiple studies have performed histological analysis of the tendons to analyze the growth of collagen fibers and the surrounding environment. They often assess the organization and parallelism of collagen fibers, vascularity of the area, cellularity, ECM formation, and inflammation within the tendon. Although these aspects do not directly indicate strength, the microenvironment can help predict the health and functional capacity of a tendon. Imaging modalities, including micro-computed tomography (CT) and magnetic resonance imaging (MRI), have also been used to assess functional outcomes. Tendon morphology, injury patterns, and healing can be visualized using MRI. Finally, when performing clinical trials, one of the most effective ways to assess patient functionality is to provide self-reported functional assessments. These assessments, often referred to as patient-reported outcome measures (PROMs), are administered during clinic visits. One advantage of these metrics is that they can be observed over time, beginning before the intervention and continuing for as long as desired. This helps providers to understand the current status of their patients and paint a picture of the effectiveness of the intervention.

Stem cell sources for tendon repair

Stem cells are primitive cells that can self-renew and have not yet reached terminal differentiation into their final cell type (17-19). Several types of Stem Cells are included under the umbrella of “Stem Cells”, all of which have various differentiation capabilities. This hierarchy of capability to differentiate into different cell lineages is deemed a given cell’s “potency” (20). The least differentiated and, thus, most “potent” cells are totipotent stem cells (Figure 2). They are named based on their ability to differentiate into cells in the human body. Examples of totipotent cells are zygotes and early blastomeres, which exist in the first three days of oocyte fertilization (19). After four days, the zygote becomes a sphere known as the morula, forming the blastocyst on day five. Blastocysts are composed of an inner cell mass and trophectoderm (21). The inner cell mass comprises many pluripotent cells called embryonic stem cells (ESCs) that give rise to the entire developing embryo. The trophectoderm is an extra-embryonic tissue composed of epithelial cells that gives rise to the placenta (21). The last type of pluripotent stem cells is called induced pluripotent stem cells (iPSCs). These unique cells were first engineered in 2006 by Takahashi and Yamanaka (22).

Figure 2.

Tendon stem cell origin. Stem cells can be harvested from different places along a cell’s path to maturity. The stage at which a cell is harvested determines the stem cell’s potency. [Created in BioRender. Ogunsola, A. (2025) https://BioRender.com/s40p990]. MSC, mesenchymal stem cell.

As development and cell differentiation progress, a postnatal cell known as the adult stem cell is eventually formed. These cells are present throughout the entire duration of a person’s life and are responsible for regeneration and repair in the human body (23). These cells are either multipotent, meaning that they can differentiate into many different cell types within their given cell lineage, or unipotent, meaning that they can proliferate into only a specific cell type (24). Although less powerful, adult stem cells have been widely used in scientific experimentation and in numerous clinical trials (20). Some of the most common types of adult stem cells studied in the context of tendon healing include mesenchymal stem cells (MSCs) and TDSCs. After terminal differentiation, these cells become functional and contribute to tissue maintenance. The functional cells in tendons are called tenocytes, which are a type of fibroblast cell embedded in a collagen-rich ECM (25). The tendon ECM and tenocyte array are further layered and cross-linked into what is grossly observed as a tendon (25,26). Upon injury, tenocytes and activated tendon fibroblasts are primarily responsible for tendon repair through collagen synthesis and matrix remodeling, whereas TDSCs/MSCs play supportive roles through both direct differentiation and paracrine signaling to facilitate the healing process (25,26).

Adult stem cells

As mentioned above, adult stem cells can be multipotent or unipotent with the ability to regenerate and repair specific tissues in the human body. They are advantageous because they are relatively easy to harvest, can be harvested autologously, decrease the risk of autoimmune rejection, pose minimal ethical dilemmas, and are generally considered less teratogenic than other stem cells (17,27).

Mesenchymal stem cells

Mesenchymal stem cells, also known as mesenchymal stromal cells, are multipotent adult stem cells that serve as widespread progenitors (28). They are identified by their ability to adhere to plastic, the presence of surface markers cluster of differentiation (CD)73, CD90, and CD105, the absence of hematopoietic stem cell markers such as CD45, CD34, CD14 or CD11b, CD79a or CD19, and human leukocyte antigen-DR isotype (HLA-DR), and their ability to differentiate in vitro (29,30). These cells can be harvested from various sites in the human body (31). The types of MSCs used for tendon healing therapies include bone marrow-derived MSCs (BM-MSC), adipose tissue-derived MSCs (AT-MSC), synovial membrane-derived MSCs (Sy-MSC), periodontal ligament-derived MSCs (PDL-MSC), and umbilical cord-derived MSCs (UC-MSC).

BM-MSCs

The bone marrow is one of the primary sites for MSC harvesting. BM-MSCs appear to be the most studied stem cells for their potential therapeutic use in tendon healing. Young et al. (1998) laid the foundational evidence for the therapeutic use of BM-MSCs in tendon repair (32). By culturing MSCs from rabbit bone marrow and embedding them into sutures, this study pioneered a practical approach for delivering MSCs directly to the site of injury. The experimental design included a controlled tendon injury model, in which BM-MSC-loaded sutures were implanted in half of the subjects and standard sutures were used as controls. Biomechanical tests at multiple postoperative time points showed that the BM-MSC-treated tendons consistently showed significantly higher load-bearing strength and tendon cross-sectional area than the control tendons, demonstrating the potential for improved healing and providing a foundation for future research on MSCs for tissue engineering and regenerative medicine (32).

Recent studies on the use of BM-MSCs in tendon and ligament repair have converged on a few central findings. BM-MSC-seeded scaffolds lead to increased collagen deposition, proteoglycan levels, collagen types I and II ratios, and improved tissue organization, which correlates with increased tensile strength and improved biomechanical outcomes (33,34). The addition of growth factors, regardless of the method of addition, direct injection, or incorporation into BM-MSCs scaffold treatment, further amplifies tendon healing effects (34). This synergy suggests that BM-MSCs combined with growth factors have superior healing effects compared with stand-alone treatments. Furthermore, Ramos et al. (2022) highlighted that bioactive modifications, such as the use of insulin-functionalized fibers, can support BM-MSC-mediated healing by promoting collagen types I and III organization and density (35). These findings suggest that insulin may play a role similar to insulin growth factor 1 (IGF-1) in enhancing tenogenesis and overall tissue quality (35).

BM-MSCs in conjunction with platelet-rich plasma (PRP) or bone marrow aspirate concentrate (BMAC) help control inflammation and reduce adhesion over time (36). Studies have also examined whether the cell itself is needed to achieve the desired healing or whether exosomes alone are sufficient. Li et al. (2022) created an in vivo anterior cruciate ligament (ACL) injury model in rats with three treatment groups, which included rats injected with either phosphate-buffered saline (PBS) (control), BM-MSC exosomes alone (BM-Exos), or BM-MSC exosomes with miR-23a-3p overexpression (BM-Exos mimic) (37). They found that, while both treatment groups showed increased tendon-to-bone healing and biomechanical strength compared to the controls, BM-Exos did so to a significantly greater degree (37).

This collection of studies seemed to show varied results. In some cases, BM-MSCs had minimal to no effect on tendon healing, while in others, an effect was seen. We suspect a large factor is the lack of standardization between studies. More information on these studies and additional studies that utilized BM-MSCs is presented and synthesized in Table 2.

Table 2. Outcome of BM-MSCs and exosomes for tendon healing.

Author	Model	Defect	Intervention	Outcome	Significance
Caniglia et al., 2012 (38)	Horses (n=6)	SDFT	Core lesions were created in the SDFTs of both forelimbs. After 4 weeks, one limb in each horse was injected with MSCs while the other injured limb was injected with Bone Marrow supernatant. Horses were euthanized 12 weeks after BM-MSC implantation, and the tendons were removed. Sections from each tendon at the maximum injury zone were evaluated using electron microscopy	BM-MSC therapy did not significantly increase matrix regeneration, as evidenced by no change in the diameter of collagen fibrils compared to untreated controls	This finding suggests that while BM-MSCs may have some benefits, they do not necessarily enhance the structural properties of tendons in terms of collagen fibril size
Renzi et al., 2013 (39)	Horses (n=33)	Tendonitis and desmitis	The horses were treated with PRP and autologous eBM-MSCs (n=21) while the other group received the control treatment of pin firing (n=12). Following this, both groups received the same rehabilitation regimen	One horse died (unrelated to treatment) and another retired from racing leaving 19 in the experimental group. Of these 13 of the 19 horses treated with eBM-MSC returned to racing while 6 were reinjured. In contrast, only 3 of the 12 control horses were able to return to sport. However, the difference in these clinical outcomes were not statistically significant (P=0.06)	While not statistically significant, the clinical results suggest that eBM-MSCs may enhance tendon defect healing in horse models, leading to greater recovery and return to activity
Daghan et al., 2023 (36)	Rabbit (n=160 fingers)	Transection of the flexor digitorum profundus tendon	The transections were sutured with a modified Kessler suture technique, and the rabbits were treated with either isotonic solution (control), PRP, BMAC, or micro-fragmented adipose tissue. At 3 and 8 weeks post-treatment, they examined the injury site’s morphology, histopathology, range of motion, and biomechanic	The BMAC group displayed the greatest tendinous adhesion thickness and inflammation at three weeks, but the lowest in both categories at eight weeks. Adhesions were further reduced with a maintained range of motion and biomechanical strength	BM-MSCs show potential for further use based on the result from this study.
Du et al., 2024 (34)	Rabbit (n=56)	5 cm anterolateral humeral head defect	Rabbits were assigned to one of four treatment groups. The control group received standard suture repair (CLT group), the first group received BM-MSCs seeded into the polycaprolactone scaffold (PB group), the second group received the same as the PB group with the addition of cerasomes loaded with growth factors onto the scaffold (PBC group), and the third group only received growth factors introduced via cerasome injection (IC group)	The PBC group showed increased levels of proteoglycan deposition, collagen volume fraction, and proportions of collagen I and II when compared to the CLT group	BM-MSCs are effective in promoting collagen growth, especially in tangent with cerasomes, and may have future use in regenerative medicine
Christogiannis et al., 2025 (40)	Rabbits (n=16)	Full thickness rotator cuff deficit	Following rotator cuff transection, rabbits underwent one of the following treatments intraoperatively: (I) no treatment (n=4); (II) treatment with scaffold alone (n=4); (III) treatment with BM-MSC-loaded scaffold (n=4); or (IV) treatment with rotator cuff MSC-loaded scaffold (n=4). Rabbits were euthanized 16 weeks postoperatively for histologic and IHC analyses	Group 1, which underwent treatment with scaffold alone, showed the highest maturing score on histological analysis compared to all 3 other groups. Similarly, group 1 showed the highest mean collagen I/collagen III ratio compared to the other groups, but the difference was not statistically significant	This study provides a contrary perspective to the utility of stem cells for tendon healing. They conclude that the scaffold alone helped rotator cuff repair compared to the MSC constructs
Salman Hamza et al., 2023 (33)	Dog (n=20)	1.5 cm Achilles defect	Following resection, the control group was given no treatment or repair (n=10), while the experimental group had a BM-MSC seeded small intestinal submucosa-extracellular matrix implant and a modified Kessler suture repair (n=10). At 4, 8, and 12 weeks postoperatively, they resected the tendon and performed biomechanical and histological analysis	They found that the experimental group showed increased tensile strength and histological qualities indicative of better tendon healing, including a greater density and organization of the tissue with high cellularity and blood vessels	This further emphasized the fact that BM-MSC therapy in conjunction with various scaffold technologies may serve to promote tendon healing on cellular and functional levels
Ramos et al., 2022 (35)	Rat (n=84)	Cut Achilles tendon	Insulin functionalized bioactive fiber repair was performed with or without BM-MSC seeding	They found that at 4 and 8 weeks postoperatively, BM-MSC seeded fiber repair showed significantly higher levels of collagen I and III compared to insulin-functionalized bioactive fiber repair alone. The BM-MSC treated group also exemplified greater collagen organization analyzed via second harmonic generation imaging	Ramos et al. posit that insulin itself acts similarly to insulin growth factor 1 and may serve to push BM-MSCs toward tenogenesis, improving tendon healing engineering
Li et al. 2022 (37)	Rat (n=90)	ACL resection	Bone tunnels were created through the distal femur and proximal tibia and a graft was sutured at each exit to repair the ACL. Three and 7 days following the reconstruction, rats were injected with either PBS (control), BM-Exos, or BM-Exos mimic	Both BM-Exos and BM-Exos mimic led to increased tendon-to-bone healing and biomechanical strength compared to controls. However, the BM-Exos mimic did so to a significantly greater degree, with less inflammation and better healing and attachment of the tendon	Both BM-Exos and BM-Exos mimic are effective in tendon regeneration, but miR-23a-3p gene overexpression inhibits interferon regulatory factor 1 and the NF-kB pathway, leading to a decrease of pro-inflammatory M1 macrophages and an increase of anti-inflammatory M2 macrophages, making the latter more effective
Zhang et al., 2024 (41)	Sprague-Dawley rats (n=45)	Rotator cuff tear	Rat rotator cuffs were incised and injected with hydrogel containing Y-BM-MSCs-exo or A-BM-MSCs-exo. After four weeks of treatment, rotator cuff tissue was harvested for further experimentation	Treatment involving Y- and A-BM-MSCs-exo displayed an improved failure load and stiffness of the TBI. The TBI also had new bone, cartilage, collagen fibers, and a large number of Sharpey-like fibers growing in a longitudinal direction with good alignment. BM-MSC treatment also promoted extracellular matrix remodeling, osteogenic differentiation, angiogenesis, and tendon cell gene expression. Notably, Y-BM-MSCs were superior to A-BM-MSCs in all aspects mentioned above	BM-MSC treatment promotes tenogenesis and has therapeutic potential in healing rotator cuff tears. Furthermore, this study reveals the influence of age in BM-MSCs, indicating that younger cells have greater regenerative capabilities
Lui, 2021 (42)	Systematic Review	Tendon injuries	Using BM-MSC EVs for tendon healing	All 5 of the studies analyzed showed increased tendon healing with the use of BM-MSC EVs, however there was a severe lack of standardization across studies	EV therapy may produce tenogenesis through paracrine signaling, which would provide many advantages to this therapy, including stability, low chance of immune rejection, no need to maintain cell viability, and no unexpected differentiation of cells

A-BM-MSCs-exo, aged bone marrow stem cell derived exosomes; ACL, anterior cruciate ligament; BM-Exos, BM-MSC exosome alone; BM-Exos mimic, BM-MSC exosome with a miR-23a-3p gene overexpression; BM-MSC, bone marrow mesenchymal stem cell; BMAC, bone marrow aspirate concentrate; eBM-MSC, equine BM-MSC; EV, extracellular vesicle; IHC, immunohistochemistry; PBS, phosphate-buffered saline; PRP, platelet-rich plasma; SDFT, superficial digital flexor tendon; TBI, tendon-bone interface; Y-BM-MSCs-exo, young bone marrow stem cell derived exosomes.

AT-derived MSCs

AT-MSCs are another commonly used form of adult stem cells in the context of tendon healing. They pose an advantage to BM-MSCs in that they are less invasive and painful to harvest and still seem to show promise in their regenerative capacity. AT-MSCs have also shown comparable results to BM-MSCs in terms of tendon healing. Nam et al. (2023) investigated tenogenic gene expression in MSCs from both the marrow and adipose tissues under stress (43). The bone marrow and adipose tissue retrieved from the recruited study subjects were processed and MSCs were isolated. The isolated MSCs surface markers were verified by flow cytometry and then seeded in a collagen-induced silicone chamber. Subsequently, the harvested cells were subjected to cyclical tensile loading for 24 and 48 hours at passage two. Morphology, orientation, proliferation rate, and protein and gene expression analyses after tensile loading at each selected time point showed that both MSC types exhibited similar structural changes, and the direction of cell alignment was perpendicular to the direction of tensile force. Both MSCs expressed collagen types I, III, and key tenogenic markers in the same loading environment (43). This suggests that AT-MSCs could be a viable alternative to BM-MSCs for tendon healing owing to their relative ease of harvest and comparable response to tensile forces.

Evidence also suggests that the use of additional growth factors in concordance with AT-MSCs may lead to better healing outcomes, as revealed by Fitzgerald et al. (2023) in their rat Achilles model (44). Tissue analysis revealed that AT-MSCs treated with growth differentiation factor (GDF)5 and PDGF significantly enhanced Achilles tendon healing in rats by improving fiber organization and blood vessel alignment. Additionally, gene expression analysis showed that AT-MSC treatment increased tendon cell development, proliferation, and connection while reducing oxidative stress and inflammation (44). The method by which AT-MSCs are delivered may also play a role in their effectiveness. Chen et al. (2024) found that AT-MSC injections delivered in spheroid form showed superior tendon healing than an injection of AT-MSCs in the form of single cell suspension and saline control injections (45).

A systematic review by Senesi et al. (2023) examined thirty-two studies that used AT-MSCs for tendon healing. Eighteen of the studies were in animal models, ten in humans, and four analyzed biomaterial applications combined with AT-MSCs in animal models (46). Fifteen animal studies reported favorable histological outcomes following AT-MSC treatment, demonstrating better regenerative environments characterized by reduced inflammatory markers and increased regenerative and angiogenic markers. All ten human studies showed an increase in functional outcomes and a decrease in pain following AT-MSC treatment. Tendon healing was further verified using MRI (46). Environmental factors can affect the healing properties of AT-MSCs. Guo et al. (2022) found that hypoxic conditions (5% O2) reduced hypoxia-inducible factor-1α (HIF-1α) in AT-MSCs and increased the levels of VEGF and other tenogenic markers [scleraxis (SCX), tenomodulin (TNMD), and DCN], suggesting enhanced tendon-like differentiation (47). The introduction of a HIF-1α blocker (2-MeOE2) inhibited these effects, suggesting that hypoxia may promote AT-MSC differentiation into tenocytes via the HIF-1α pathway, although it may also inhibit cell proliferation and migration (47).

In summary, the potential use of AT-MSCs for tendon healing is promising, with multiple studies supporting their utility, but there has been limited successful translation of these techniques into human trials. Additionally, various environmental factors such as hypoxia and stem cell configuration may play an important role in the success of these therapies. Information on these studies and additional studies that utilized AT-MSCs is presented and synthesized in Table 3.

Table 3. Outcome of AT-MSCs and exosomes for tendon healing.

Author	Model	Defect	Intervention	Outcome	Significance
Kryger et al., 2007 (48)	Rabbit (n=30)	Flexor profundus tendon defect	Four cell types (tenocytes, sheath fibroblasts, BM-MSCs, and AT-MSCs) were obtained from adult rabbits, along with flexor tendons that were then acellularized. The cells were cultured and then used to reseed the tendons which were then implanted into a flexor profundus tendon defect in vivo	AT-MSCs proliferated faster in culture. However, all cell types effectively repopulated acellularized tendons	All of the four cell types may be successfully used to engineer tendons
Manning et al., 2013 (49)	Dog (n=15)	Flexor digitorum profundus tendon transection	In vitro: PDGF-BB and AT-MSCs were integrated into a heparin/fibrin-based scaffold. This was layered with an electrospun nanofiber PLGA backbone, which provided structural integrity necessary for surgical handling and implantation	In vitro: the delivery system facilitated controlled release of growth factors and maintained cell viability	The novel scaffold-based delivery system shows promise for clinical applications and for the ability of AT-MSCs to mount a healing response
			In vivo: flexor tendon repair was performed using similar surgical repair to human flexor tendon repair and introducing cellular or acellular scaffolds or naive repair. At 3 and 9 days the dogs were euthanized and gene expression, histological, and DNA analyses were performed	In vivo: histologically, the groups with scaffolds showed increased polymorphonulear cells and monocytes, however adhesions were also observed. Fluorescent imaging showed the successful viability of AT-MSC implanted cells
Carvalho et al., 2013 (50)	Horses (n=8)	Lesions of the superficial digital flexor tendon	AT-MSCs were administered into the induced lesions. After 16 weeks, the tendons were biopsied and underwent histopathological, immunohistochemical, and gene expression analyses	AT-MSCs prevented lesion progression, led to better organization of collagen fibers, and decreased the amount of inflammatory infiltrates	This indicates that AT-MSC treatment has therapeutic potential for equine tendonitis, as it may induce tissue regeneration rather than merely repairing damaged tissue
Fitzgerald et al., 2023 (44)	Rat (n=34)	Lesions of the Achilles tendon	Rat Achilles tendons were cut and either left severed or repaired. Then, one of three MSC regimens (hydrogel, a hydrogel with AT-MSCs, or a hydrogel with AT-MSCs, cocultured with GDF5 and PDGF) were administered	In both the repaired and unrepaired tendons, the AT-MSCs + GDF5/PDGF displayed the greatest healing capacity, indicated by tight fiber organization and blood vessel formation. Additionally, AT-MSCs + GDF5/PDGF had the highest gene expression of SOX9, a protein-inducing gene, and proliferating cell nuclear antigen and Tenascin C, markers of cellular proliferation and cell-to-cell connections	These findings suggest that the GDF5/PDGF helps the AT-MSCs thrive in an injury environment deficient in cellularity and angiogenesis and promotes tendon repair. Relatively low oxidative stress and inflammation with high extracellular matrix turnover also indicate that AT-MSCs may be ready for implantation and healing upon treatment
Shen & Lane, 2023 (51)	Mice (n=32)	Lesions of the Achilles tendon	Extracellular vesicles of iEVs were introduced to Achilles tendon injuries in mice	The introduction of the iEVs was able to increase anti-inflammatory gene expression, decrease mononuclear cell activity, increase collagen, promote recovery of the tendon structure, develop less peritendinous scar, decrease incidence of postoperative tendon gap or rupture, and increase functional recovery time at the site of the injury	The environment created by AT-MSCs has healing capacities and may be beneficial in tendon regeneration
Senesi et al., 2023 (46)	Meta-analysis	N/A	This review examined 32 different studies utilizing AT-MSCs: 18 in animal models, 10 in humans, and 4 analyzing biomaterial applications in animal models	Fifteen of the animal studies found positive results with AT-MSCs when analyzed histologically, demonstrating better regeneration and less inflammation. All studies in humans displayed an increase in functional outcomes and a decrease in pain following treatment	AT-MSCs are an effective form of treatment and have the capacity for future, more advanced use in regenerative medicine
Guo et al., 2022 (47)	In vitro	Normoxic and hypoxic conditions	AT-MSCs were put under hypoxic stress (5% O2), and another group of AT-MSCs were put under normoxic conditions (20% O2). Later, both groups were also introduced to HIF-1α inhibitor (2-MeOE2)	The hypoxic group produced a marked decrease of HIF-1α compared to the normoxic group, and the hypoxic group displayed higher levels of VEGF and tenogenic markers SCX, TNMD, and DCN. The HIF-1α inhibitor impeded these effects	These findings indicate that hypoxia may be both inhibitory to AT-MSC proliferation and migration, and at the same time, induce differentiation of tenocytes via the HIF-1α pathway
Nam et al., 2023 (43)	In vitro	Cyclic tensile loading	Bone marrow and adipose tissue were retrieved from 10 patients, and then the BM-MSCs and AT-MSCs were isolated and subjected to cyclical tensile loading for 24–48 hours	Both types of MSCs aligned perpendicular to the tensile force, with no significant proliferation observed. Additionally, both groups expressed collagen and tenogenic genes	This indicates that AT-MSCs may be a substitutionary solution for BM-MSCs in the context of tendon healing due to the relative ease of harvest and similar response to tensile forces
Schmitz et al., 2025 (52)	Rabbit (n=32)	Full thickness 3 mm lesion in the midsubstance of right gastrocnemius tendon (a component of the common calcaneus tendon)	The rabbits received a single injection of either AT-MSCs mixed with RLS or saline (n=16) or RLS or saline alone (n=16). The rabbits were euthanized, and tendons were analyzed at 4 and 12 weeks via histology, IHC, and biomechanical testing	The AT-MSC treated lesions showed healing that had morphology of regenerated tendons including newly formed connective tissue, compared to controls which showed scar tissue. On biomechanical testing, the AT-MSC treated tendons showed greater percent relaxation indicative of increased viscoelasticity of the tendon	This study exemplified the way that AT-MSCs can enhance natural tendon regeneration and reduce scar tissue. In addition to improved connective tissue organization, AT-MSC use can also lead to better functioning tendons from a biomechanical standpoint. This has tremendous implications in healing from tendon injuries
Chen et al., 2024 (45)	Rabbit	Achilles tendon transection	Rabbits had their transection repaired using suture and were assigned to one of three injection treatments: saline injection, AT-MSC single-cell suspension injection, or AT-MSC spheroid injections. The rabbits were euthanized at 1 and 4 weeks for imaging, histologic, and biomechanical analyses of the tendons	On fluorescent imaging, the AT-MSC spheroid injection cohort showed significantly levels of fluorescence intensity at both 1 and 4 weeks compared to AT-MSC single-cell suspension injections. On biomechanical testing, the AT-MSC single-cell suspension group was able to withstand a significantly higher tensile force than the saline group, but the AT-MSC spheroid group was able to withstand a significantly higher tensile force than the AT-MSC single cell group. Histological analysis also demonstrated better fiber alignment in the AT-MSC spheroid group compared to the other groups. Finally, both AT-MSC treatments resulted in lower Col III levels, higher decorin and tenomodulin levels, and lower inflammatory markers IL-6 and IL-1β	This study showed that while AT-MSCs in single cell suspensions can have beneficial effects compared to saline control’s, AT-MSCs prepared in spheroid configurations may be even more effective for tendon healing. This study emphasized that it is important to consider both the type of cells that are administered, cell configuration (2D vs. 3D), and how those cells are administered to achieve maximal healing effects
Sam et al., 2025 (53)	Rat (n=14)	Achilles tendon transection	Four weeks after full length transection of their Achilles’ tendons, rats were treated with (I) TDSC secretome injection (n=4); (II) AT-MSC secretome injection (n=4); (III) TDSC + AT-MSC injection (n=3); or (IV) phosphate-buffered saline solution injection (n=3). Rats were euthanized two weeks following treatment and tendons were studied for PINP, PIIINP, and histopathologic analyses	Group 3 (TDSC + AT-MSC combined secretomes) showed increased levels of PINP. Otherwise, PIIINP levels and histopathological analyses showed no differences between any of the groups	This study proposes the benefit of using more than one stem cell lineage for treatment. It showed that using both TDSC and AT-MSCs secretomes in conjunction led to higher levels of PINP which is indicative of active type I collagen production

2D, two-dimensional; 3D, three-dimensional; AT-MSC, adipose tissue-derived MSC; BM-MSC, bone marrow-derived MSC; DCN, decorin; GDF5, growth differentiation factor 5; HIF-1α, hypoxia-inducible factor-1α; iEVs, inflammation-primed AT-MSCs; IHC, immunohistochemistry; IL, interleukin; MSC, mesenchymal stem cell; N/A, not applicable; PDGF, platelet-derived growth factor; PDGF-BB, platelet-derived growth factor subunit B; PIIINP, procollagen type III N-terminal propeptide; PINP, procollagen type I N-terminal propeptide; PLGA, polylactic glycolic acid; RLS, Ringer’s lactated solution; SCX, scleraxis; TDSC, tendon-derived stem cell; TNMD, tenomodulin; VEGF, vascular endothelial growth factor.

Sy-MSCs

Sy-MSCs are stem cells found in the synovial joints. First elucidated in 2001 and later characterized more descriptively in 2005, Sy-MSCs have been explored widely as potential therapeutic agents (54,55). However, only a few studies have examined Sy-MSCs in the context of tendon healing and have demonstrated their promising regenerative properties. The central theme of studies that explored Sy-MSCs as potential agents in tendon and ligament healing focused on their role in enhancing tendon-to-bone integration and collagen formation. Ju et al. (2008) and Noh et al. (2024) showed that Sy-MSCs accelerate early tendon healing and improve bone-tendon interface (BTI) strength in animal models (56,57). Noh et al.’s treatment included the incorporation of a scaffold-free fibrocartilage construct seeded with Sy-MSCs and treatment with growth factors [TGF-β and connective tissue growth factor (CTGF)] to further support tendon-to-bone attachment (57). Bami et al. (2020) also found that Sy-MSCs showed the highest expression of tenogenic proliferation and collagen marker (TNMD) compared to other MSCs, suggesting that Sy-MSCs may possess superior tenogenic potential (58).

However, these results have not been unanimously replicated. Khan et al. (2020) observed limited effects of Sy-MSCs in ovine deep digital flexor tendon injuries, regardless of euthanasia time (59). The findings in this section suggest that the effectiveness of Sy-MSCs may vary depending on the specific injury context and the tendon type. Overall, these studies imply that Sy-MSCs have the potential for tendon and ligament repair; however, their effectiveness and long-term impact are influenced by environmental and injury-specific factors. These results warrant further research on the efficacy of Sy-MSCs and their ability to regenerate the tendons. Information on these studies and additional studies that utilized Sy-MSCs is presented and synthesized in Table 4.

Table 4. Outcome of Sy-MSCs and exosomes for tendon healing.

Author	Model	Defect	Intervention	Outcome	Significance
Ju et al., 2008 (56)	Sprague-Dawley rats (n=18)	Patella tendon	ACL graft only (n=9); ACL graft + Sy-MSC (n=9)	At 1-, 2-, and 4-weeks post-op, Sy-MSC group showed more collagen deposition in the bone tunnels compared to the controls at 1 week. At 2 weeks, more oblique collagen fibers that looked like Sharpey’s fibers in the Sy-MSC group compared to the control group appeared. At 4 weeks post-op, the tendon implant had attached to the bone in both the control and Sy-MSC groups	This served as one of the earliest in vivo studies to use Sy-MSC for tendon healing showing clinical potential for this stem cell lineage
Khan et al., 2020 (59)	Sheep (n=26)	Deep digital flexor tendon	Iron nano-particle-labeled cell distribution was assessed at 1 (n=2) and 2 (n=2) weeks post-implantation; assess effects of unlabeled cells on tendon repair at 4 weeks (n=6), 12 weeks (n=8), and 24 weeks (n=8) post-implantation	Analysis of tendon was performed at 4, 12, and 24 weeks following Sy-MSC treatment. Macroscopic analysis showed that the lesions remained visible in all 22 animals regardless of the timing of their euthanasia, but that the Sy-MSC’s engrafted into the synovium	In contrast to other studies, this study provides an alternative finding that though the Sy-MSC engraft and survive within the synovium, they may not be effective for tendon repair
Noh et al., 2024 (57)	Sprague-Dawley rats (n=56)	ACL	In vitro: scaffold in TGF-β only, scaffold in CTGF only, and Scaffold in TGF-β + CTGF	In vitro: cells co-cultured with TGF-β + CTGF produced fibrocartilage-like scaffold structures. Also treating implant site with collagenase promotes better tendon to bone integration by the scaffold	This recent study shows the viability of using Sy-MSCs in scaffold constructs and indicates greater structural stability in a tendon construct following in vivo implantation in rats. This provides support that the same effects may be observable clinically
			In vivo: ACL reconstruction only and ACL + Sy-MSC	In vivo: at 2 and 4 weeks post-operative, Sy-MSC fibrocartilage scaffold ACL reconstructions were 88% stiffer 4 weeks postoperatively than the non-treated group
Bami et al., 2020 (58)	In vitro	N/A	Monodose rabbit MSCs with 10−10 mol/L insulin and cultured them for 2 weeks in DMEM-LG	After 2 weeks, qRT-PCR was performed and assessed for presence of TNMD. Sy-MSCs showed the highest level of TNMD, nearly doubling the expression compared to any other MSC	Given TNMD’s effect on tendon maturation and proliferation, and Sy-MSCs ability to produce TNMD more efficiently than other cell lineages, Sy-MSCs may be an effective enhancer of tendon healing.

ACL, anterior cruciate ligament; CTGF, connective tissue growth factor; DMEM-LG, Dulbecco’s Modified Eagle Medium-low glucose; MSC, mesenchymal stem cell; N/A, not applicable; qRT-PCR, quantitative reverse transcription polymerase chain reaction; Sy-MSC, synovial-derived MSC; TGF-β, transforming growth factor-beta; TNMD, tenomodulin.

PDL-MSCs

PDL-MSCs are stem cells harvested from the periodontal ligament (PDL) (60). They are of particular interest because of their self-renewal properties, potential as immunomodulatory agents, differentiation capacity, and accessibility, which make them ideal for tendon regeneration (61,62). To examine the suitability of these cells for ligament regeneration, Menicanin et al. (2014) implanted ovine PDL-MSCs into mice, which displayed the capacity to form mineralized tissue and fibrous ligament structures, even after removal (62). Additionally, Hsieh et al. (2016) and Chen et al. (2021) found that PDL-MSCs were effective in improving tendon strength and reducing ectopic fibrocartilage formation in rat Achilles tendons, especially when paired with aligned silk scaffolding (63,64). Moshaverinia et al. (2014) encapsulated PDL-MSCs in TGF-β3-loaded Arg-Gly-Asp (RGD)-coupled alginate microspheres and implanted them into mice, where they exhibited high mRNA expression of gene markers related to tendon regeneration and were more effective than other dental MSCs and BM-MSCs (65). Taken together, these findings highlight the potential of PDL-MSCs as a suitable source for tendon and other musculoskeletal regenerative therapies. That said, there are a limited number of studies with mostly low sample sizes and thus more studies may need to be performed and analyzed before concluding the utility of this cell type. Information on studies that utilized PDL-MSCs is presented and synthesized in Table 5.

Table 5. Outcome of PDL-MSCs and exosomes for tendon healing.

Author	Model	Defect	Intervention	Outcome	Significance
Menicanin et al., 2014 (62)	Sheep (n=7)	Bilateral standard periodontal defect	BrdU-labeled PDL-MSCs taken from Merino ewes were implanted into immunodeficient mice. They were able to form bone/cementum-like mineralized tissue. The PDL-MSCs were then retrieved from harvested primary transplants and implanted into secondary ectopic xenogeneic ovine transplants	Histomorphological analysis revealed that four out of six donor re-derived PDL-MSC populations displayed a potential to survive and form fibrous ligament structures and mineralized tissues associated with vasculature in vivo but at lower levels than primary PDL-MSCs	This exemplified the potential viability of PDL-MSCs for healing in a live model
Moshaverinia et al., 2014 (65)	Immunocompromised mice (n=8)	Subcutaneous implantation	A co-delivery system based on TGF-β3-loaded RGD-coupled alginate microspheres was used to encapsulate PDL-MSCs and GMSCs to examine their ability to differentiate in vitro and in vivo. Human GMSCs and PDL-MSCs were obtained from twenty healthy male patients aged 18–25 years. The encapsulated dental-derived MSCs were then implanted subcutaneously into immunocompromised mice	In vitro: after 4 weeks of differentiation in vitro, PDL-MSCs and GMSCs showed high mRNA expression levels for gene makers linked to tendon regeneration	These findings show that periodontal ligament and gingival tissues can be sources for stem cells in tendon engineering and that PDL-MSCs and GMSCs show potential in tendon regeneration
				In vivo: in the animal model, ectopic neo-tendon regeneration was seen in subcutaneously implanted MSC-alginate constructs. In quantitative PCR and in vivo histomorphometric analyses, PDL-MSCs had a remarkably greater capacity for tendon regeneration than GMSCs and positive control BM-MSCs
Hsieh et al., 2016 (63)	Athymic female rats (n=40)	Achilles tendon	Four study groups (with ten athymic female rats each) were used to compare the healing properties of ATs and PDL-MSCs: (I) NT, non-injured left Achilles tendon; (II) ED control group, injured right Achilles tendon without cell implantation; (III) PDL, each animal received a one-shot implantation PDL-MSC pellets within injured right Achilles tendon; (IV) hAT, each animal received a one-shot implantation hAT-cell pellets at the same site	Each group was assessed 16 weeks after surgery using various techniques, including histological and immunohistological stainings and examinations of cell morphometry. Results revealed that PDL cell-implanted tendons displayed reduced fibrocartilage formation, improved tensile strength, tendon matrix expression corresponding to the final healing stage, and better cell morphometry parameters when compared with the ED group, but did not show significant differences compared to the tendons treated with hAT-derived cells	This study shows that PDLs are suitable as a novel cell source for tendon regeneration
Chen et al., 2021 (64)	Sprague-Dawley rats (n=5)	Achilles tendon	Obtained PDL-MSCs from the third molars of patients aged 20–40 years and fabricated aligned and random silk scaffolds using the controlled directional freezing technique. The scaffolds were differentiated by surface structure, water contact angle, swelling ratio, degradation speed, and mechanical properties. Then, PDL-MSCs were seeded onto the silk scaffolds and transplanted into ruptured rat Achilles tendons. Scaffolds without cells were used as control groups	After 4 weeks, histological examination revealed that the PDL-MSCs seeded in the aligned scaffold exhibited more tendon-like tissue formation compared to the scaffold-only groups. The aligned scaffold also enabled the proliferation and ordered arrangement of PDL-MSCs more than the random scaffold	This data reveals that PDL-MSCs paired with aligned biomimetic silk scaffolding are suitable for tendon regeneration

ATs, Achilles tendon-derived cells; BrdU, bromo-deoxyuridine; ED, empty defect; GMSC, gingival MSC; hAT, human Achilles tendon; NT, native tendon; PDL-MSC, periodontal ligament-derived MSC; RGD, Arg-Gly-Asp; TGF-β3, transforming growth factor-beta 3.

TDSCs

Recently, TDSCs have emerged as a promising avenue for tendon repair and healing. TDSCs are MSCs that can differentiate into various cell types including tenocytes, chondrocytes, and osteocytes. This ability makes them essential components of tendon repair and regeneration. TDSCs can be isolated from tendon tissue, typically from the patellar, hamstring, or Achilles tendons (66). The cells typically undergo enzymatic digestion and are cultured in specific growth media that allow them to proliferate and maintain their stem cell properties (67). Harvey et al. (2019) identified TDSC populations using an in vivo mouse model (68). The aim of their study was to define adult tendon stem cells and explore the mechanism of tendon regeneration using a paratenon-sheathed patellar tendon model. They identified a population of tubulin polymerization-promoting protein family member 3-expressing (Tppp3+) cells as tendon stem cells. These cells generate new tenocytes that undergo self-renewal after injury (68). Additionally, a subset of these cells expressed PDGF receptor alpha (PDGFRα), which was found to be crucial for new tenocyte production, as inactivation of PDGFRα inhibited regeneration. Overall, Harvey’s team was able to trace a lineage of TDSCs and demonstrated that PDGFRα signaling is necessary for the lineage to produce new tenocytes (68).

TDSCs are effective agents for tendon repair, particularly when used in combination with other biological treatments and environmental conditions. Chen et al. (2012) demonstrated that TDSCs combined with PRP in mechanically loaded conditions improved collagen levels and tendon healing (69). This result highlights the synergistic effect of the TDSCs and mechanical loading (69). Similarly, Jiang et al. (2014) studied TDSCs co-cultured with decellularized fibroblast-derived matrix (dFM) both in vitro and in vivo (70). First, they showed that TDSCs proliferate in the presence of dFM in vitro. Using rats, they showed that in vivo introduction of TDSC + dFM led to more organized type 1 collagen, greater biomechanical strength, and an increase in type 1 collagen expression (70). Furthermore, Ni et al. (2012) supported previous findings by showing that TDSCs delivered with fibrin glue enhanced collagen alignment and tensile strength in a rat patellar tendon model (71). Lui et al. (2014) found that TDSCs effectively promote tendon repair without increasing the risk of ectopic bone formation, suggesting that their therapeutic effects might not be due to direct differentiation (72). Environmental factors significantly influence TDSC behavior, with studies showing that hypoxic conditions enhance TDSC proliferation and reduce undesired differentiation (73). In addition, injury activation increases the expression of tenogenic markers and pluripotency of TDSCs, supporting tendon repair (74). Overall, TDSCs exhibit strong potential for tendon regeneration, especially when optimized through mechanical loading, scaffolds, and controlled hypoxia. Information on these studies and additional studies that utilized TDSCs is presented and synthesized in Table 6.

Table 6. Outcome of tendon-derived stem cells and exosomes for tendon healing.

Author	Model	Defect	Intervention	Outcome	Significance
Chen et al., 2012 (69)	Sprague Dawley rats (n=96)	Achilles tendon	Tendons were injured and then subdivided into loaded and unloaded groups. Unloaded conditions were created through injection of botulinum toxin. Injured tendons were then randomly treated with PRP, TDSCs, PBS, or a combination of TDSCs and PRP. At 3 and 14 days post-treatment, the rats were euthanized and the tendon excised for mRNA and protein expression were measured through PCR and immunohistochemical analyses	Loaded tendons treated with any intervention had higher collagen levels compared to unloaded tendons, with the TDSC + PRP combination showing the highest collagen levels compared to PBS-treated loaded tendons	The results showed that TDSCs and PRP have synergistic effects on tendon healing and that loaded conditions improve healing
Ni et al., 2012 (71)	Sprague-Dawley rats (n=38)	1 mm defect in the center of the distal apex of the patellar tendon	The tendons were repaired via constructs consisting of fibrin glue only (n=19) or allogeneic TDSCs and fibrin glue (n=19). At 1, 2, and 4 weeks 3 rats in each group were euthanized and had their tendons harvested for fluorescent GFP signal imaging, histologic analysis, and polarization microscopy. The remaining 10 living rats from each group were euthanized at 4 weeks for tendon harvest and biomechanical testing	In both groups cellularity decreased from week 1 to 2 and an increase in extracellular matrices by week 4. The TDSC group showed more extracellular matrix formation compared to the control group. The TDSC group also showed better cellular alignment, collagen alignment, and collagen birefringence. Also, the TDSC treated group showed significantly greater biomechanical strength on both the Young’s modulus and ultimate stress test	This study showed that TDSCs may promote quicker healing in the early stages of tendon healing compared to fibrin glue. They may aid in both the microscopic/cellular healing of the tendon, and also the gross biomechanical function of injured tendons
Lui et al., 2014 (72)	Sprague-Dawley rats (n=60)	~1 mm defect in the center of the distal apex of the patellar tendon	Defects were repaired using fibrin constructs. The experimental group had constructs with GFP-TDSCs (n=30) while the control group had fibrin constructs alone (n=30). Fluorescence imaging, histology, immunohistochemistry, and image analysis were performed to analyze the tendon tissue following euthanasia at 1, 2, 4, 8, and 16 weeks	Photomicrograph analysis showed that fiber arrangement and cell alignment were seemingly enhanced in the TDSC group in early weeks, but overall healing by 16 weeks was comparable in the TDSC and control groups. The TDSC group did however show less vascularity at 16 weeks. Fluorescent imaging showed that the TDSCs labeled with GFP were not detected after week 4. Image analysis also showed generally decreased (CD3, CD68, and mast cell tryptase) or non-significantly increased (CD163) levels of immune cells in the TDSC group compared to the control group in the early weeks of healing	This study showed that TDSC-enhanced constructs aided in tendon healing and did increase the risk for side effects such as ectopic chondro-ossification. It also showed that the TDSC construct was minimally immunoreactive. This also suggested that their mechanism of action may not be direct differentiation
Jiang et al., 2014 (70)	Nude rats (n=8)	Dorsal subcutaneous pockets	Dorsal subcutaneous pockets were created surgically. Either TDSC or TDSC + dFM was inserted	After 3 weeks, a “tendon-like structure” was observed in the TDSC + dFM in contrast to no tendon formation in the control group. Similarly, the cells in the TDSC + dFM group were longitudinal compared to no organization in the TDSC group. Upon immunohistochemistry analysis of the tendon, type 1 collagen was seen in the TDSC + dFM implanted tissue whereas a weaker expression was seen in the control group	This exemplified that dFM may be a useful adjunct for TDSC therapy
		Middle 1/3rd patella tendon defect	Surgical defects were introduced to the middle 1/3rd of the rat patella. The tendon was sutured with TDSC + dFM tissue. The control group received an injection of TDSC only
Jiang et al., 2014 (same as above) (70)	Sprague-Dawley rats (n=40)	Window injury of the patella tendon	TDSC with no scaffold or TDSC + dFM scaffolds were inserted into the created lesion. At 3 and 8 weeks, tissue was harvested and analyzed histologically and biomechanically	The TDSC + dFM showed more longitudinal arrangement of collagen, produced more extracellular matrix, and showed higher expression of type 1 collagen compared to the TDSC only group. Similarly, only one tendon failed biomechanical testing due to rupture in the TDSC only group. Overall, the TDSC + dFM group endured higher stress than TDSC only group	Not only may dFM increase the tenogenesis of tendons, but it also may serve to increase functional capacity of tendons when stressed
Wu et al., 2025 (75)	Sprague-Dawley rats (n=120)	1×2 mm Achilles tendon defect	Rat TDSCs were subjected to hypoxia to release sEVs. sEVs were then injected into injured rat Achilles tendons in a hydrogel mixture	sEVs harvested from TDSCs subjected to hypoxia were found to express higher tendon healing protein levels, such as tenomodulin and collagen I. Injured tendons injected with hypoxic sEVs displayed greater strength during biomechanical testing compared to control groups, as well as reduced collagen III formation, which is known to cause scar tissue formation. They also had lower M1 macrophage activation and downregulated fibrosis markers such as α-SMA, COX2, and COLIII	Tendon derived stem cells and the extracellular vesicles they release have therapeutic potential and promote tendon healing in vivo. This study also demonstrates the benefits of exposing tendon derived stem cells to hypoxic conditions, as they enhance their regenerative properties
Lee et al., 2012 (73)	In-vitro	N/A	hTDSC isolated from human patellar tendons of 3 patients undergoing ACL reconstruction. The TDSCs were isolated from this tissue and processed. The cells were then exposed to either 2% or 20% O2 concentration	Low oxygen tension increased cell numbers by 25% and increased colony numbers but reduced the differentiation potential. However, the hypoxia-preconditioned hTDSCs could successfully differentiate once returned to 20% O2 tension. There was no effect on collagen production	Explored the effect of low oxygen tension on the function of hTDSCs
					A hypoxic environment is most suitable for hTDSC expansion in vitro and once expanded the cells can differentiate normally once exposed to normal oxygen tension. This research provides a technique useful for tendon tissue engineering
Yin et al., 2013 (76)	In-vitro	N/A	Human adult Achilles tendon samples were obtained following leg amputation surgery. hTDSCs were isolated from these samples. The collagen matrices were harvested from pig bones, tendons, and dermis	All matrices supported stem cell adhesion and proliferation. Tendon-derived matrix promoted tendinous phenotypes in hTDSCs while inhibiting osteogenesis. Bone-derived matrix induced osteogenic differentiation. The collagen matrix derived from the dermis had no apparent effect	Evaluated how the tissue origin of collagenous matrices used as stem-cell delivery systems affects the differentiation of tendon-derived stem cells
					Their findings underscored the potential of using a decellularized matrix for tendon engineering applications and displays increased efficacy gained by using a tendon-derived matrix in conjunction with TDSCs
Jiang et al., 2014 (70)	In-vitro	N/A	A dFM was derived from rats seeded with TDSCs	dFM as a substrate was demonstrated to support the growth and tenogenic differentiation of TDSCs	Explored the use of dFM as a substrate for tissue engineering and found that dFM combined with TDSCs may be useful in tendon repair and regeneration
Wang et al., 2019 (77)	In-vitro	N/A	Rat TDSCs were treated with interleukin-1β and added to conditioned medium with or without exosomes. Tendon matrix markers and tenogenesis markers were measured via immunostaining and western blot	The conditioned medium with exosomes decreased MMP-3 expression while increasing TIMP-3, creating a balance between the two that is thought to be ideal for repair. The exosome-containing medium also yielded increased expression of tenogenesis markers TNMD and Col-1a1	This further demonstrates the efficacy of exosomes in tendon repair while elucidating some of the potential mechanisms by which the exosomes contribute to healing
Song et al., 2022 (78)	In-vitro	N/A	Rat tenocytes were incubated with labeled TDSC exosomes and cultured. The tenocytes underwent several assays to test for exosomes’ effect on oxidative stress-protection, proliferation of tenocytes, and tenocyte migration	High concentration of exosomes demonstrated protection of tenocytes from oxidative stress. The Exos also increased type I collagen production, tenocyte migration, and expression of tendon-specific markers (SCX, Col-1a1 and DCN)	Tendon derived stem cell exosomes could have therapeutic potential for promoting tendon healing in vivo
Sam et al., 2025 (53)	See Table 2 for study details	See Table 2 for study details	See Table 2 for study details	See Table 2 for study details	See Table 2 for study details

ACL, anterior cruciate ligament; COX2, cyclooxygenase-2; DCN, decorin; dFM, decellularized fibroblast-derived matrix; GFP, green fluorescent protein; hTDSC, human TDSCs; MMP-3, matrix metalloproteinase-3; N/A, not applicable; PBS, phosphate-buffered saline; PCR, polymerase chain reaction; PRP, platelet-rich plasma; SCX, scleraxis; sEVs, small extracellular vesicles; TDSC, tendon-derived stem cell; TIMP-3, tissue inhibitor of metalloproteinase-3; TNMD, tenomodulin; α-SMA, alpha-smooth muscle actin.

ESCs

ESCs are pluripotent cells that can be distinguished by the presence of markers such as octamer-binding transcription factor 3/4 (OCT3/4), stage-specific embryonic antigens 3 and 4 (SSEA-3 and SSEA-4), tumor related antigen (TRA)-1-60, TRA-1-81, and alkaline phosphatase (79). Given their powerful regenerative capabilities, ESCs are considered an extremely attractive option for stem cell therapy. However, there are numerous limitations and concerns regarding their use, including ethical and religious concerns (17), governmental regulations (80), concern for teratoma formation (81-90), and concern for host immune rejection (91,92). Because of these concerns, ESCs are less frequently used as a source in modern stem cell research, but some studies have examined their therapeutic capabilities in tendon healing.

Yao et al. (2008) optimized suture coatings to enhance ESC adherence and suggested practical approaches for tendon repair applications (93). Watts et al. (2011) provided evidence that fetal-derived ESCs (fdESCs) promoted tendon regeneration in horses by enhancing collagen alignment without altering gene expression, highlighting their structural impact on tendon repair (94). Furthermore, Dale et al. (2018) demonstrated that ESCs cultured with bone morphogenetic proteins (BMP)12/13 developed tenocyte markers and tendon-like morphology (1). Several studies have also employed genetically modified ESCs and engineered scaffolds to develop tendon-like tissues, showing that ESC-derived MSCs can replicate tendon architecture and functionality in animal models under certain conditions, such as mechanical stimulation (3,95). These studies showed that dynamic loading improved cell alignment, collagen organization, and integration with host tissue in ESC-MSCs seeded onto collagen or silk scaffolds. These studies emphasize the feasibility of using ESCs for tendon regeneration and underscore the importance of combining stem cell sources with growth factors, scaffolds, and mechanical cues to optimize tendon-like tissue formation for potential clinical applications. Information on these studies and additional studies that utilized ESCs is presented and synthesized in Table 7.

Table 7. Outcome of ESCs and exosomes for tendon healing.

Author	Model	Defect	Methods	Results	Significance
Guest et al., 2010 (96)	Horses (n=8)	Mechanically induced damage in SDFT	Three lesions were made at 3 sites in one SDFT of each horse. Sites then received injections of either MSC or ESC. Contralateral limbs were used as the control group. Horses were euthanized at 10, 30, 60, or 90 days after injection of cells. The SDFTs were recovered and underwent immunohistochemical analysis and cell counting to determine the immune response and survival of each stem cell type, respectively	ESCs survived better than MSCs. Neither of them induced adverse immune reactions or tumor formation	Exploration of the efficacy of ESCs versus MSCs in repairing mechanically induced damage in equine SDFTs underscored the potential of MSCs in tendon repair and regeneration, which offers promising avenues of exploration for clinical application
Watts et al., 2011 (94)	Female thorough-bred horses (n=8)	Mid-metacarpal SDFT injury via a collagenase gel physical defect model	fdESC were injected into induced tendon injuries in the superficial digital flexor tendons of thoroughbred horses. After 8 weeks, results were analyzed using MRI, biomechanical assays, gene expression, and histology	Analysis revealed that cell survival was demonstrated in fdESC-treated tendons but not in controls. The fdESC-treated tendons showed improved tendon size, linear fiber pattern, tissue architecture, and decreased lesion size	Supports the idea that pluripotent stem cells like fdESCs can promote musculoskeletal regeneration rather than just repair
Cohen et al., 2010 (95)	Nude mice (n=17)	Full thickness Achilles tendon defect	CTP [which include fetal-derived CTPs (n=7) and hESCs (n=7)] were harvested, cultured, and seeded into a scaffold. Then following a 4 mm long section removed from the Achilles tendon, the CTP scaffold was included in the suture repair. Control groups had their tendon resected with no repair. After 8 weeks, the mice were euthanized, and analysis was performed	When analyzed, the mice treated with the grafts showed extension >90° on range of motion. They also were able to run significantly faster than control mice	This study provided an early tangible example of ESC seeded scaffolding for tendon repair
Chen et al., 2010 (3)	Nude mice (n=6)	Subcutaneous pocket	An incision was created on the right dorsum of the mice and a 5 cm hESC-MSC bioengineered tendon was inserted and sutured into the fascia in order to assess for stem cell viability. Controls included a left dorsum pocket that received a 2 cm tendon that was left un-sutured, thus did not receive the repetitive mechanical stimulation that the experimental group experienced	Immunofluorescent analysis indicated a decrease in the number of cells after 4 weeks, but still some amount of survival. Additionally, the sutured group, exposed to mechanical stimulation, showed greater tendon-like morphology compared to the control group. Similar results were seen upon histological analysis. The tendon that was stressed showed greater expression of collagen and showed better parallel alignment of fibers compared to the non-stressed tendon	This exemplified the viability of the hESC-MSC bioengineered tendon and how it may be able to be affected by mechanical stresses
Chen et al., 2010 (same as above) (3)	Sprague-Dawley rat (n=11)	6 mm Achilles tendon lesion	Surgically induced Achilles lesions were performed and then treated with either GFP-transfected hESC-MSC scaffolds or non-cell seeded scaffolds. At 2 and 4 weeks, animals were sacrificed and histology, gene expression, protein expression, and extracellular matrix secretion analyses were performed. At 4 weeks biomechanical testing was also performed	Cell viability was confirmed. Histological analysis showed better tendon histology scores and more cell alignment in the hESC-MSC scaffold group compared to the scaffold only group. The hESC-MSC group also expressed higher levels of collagen and showed greater strength on biomechanical testing	hESC-MSC scaffolds serve as a potential adjunct therapy for tendon defect healing
Barsby and Guest, 2013 (97)	In vitro	N/A	Three lines of ESCs and tenocytes isolated from healthy equine tendon tissue postmortem were cultured and passaged. For both ESCs and tenocytes, transcription TGF-β1, 2, or 3 was added and cells were harvested at 1, 3, 7, or 14 days, at which point RNA extraction and immunohistochemistry was performed	ESCs can differentiate into tenocytes when exposed to TGF-β signaling, TGF-β3 having the greatest effect. In addition, SCX, a gene required for the development of tendons and ligaments, is upregulated in response to TGF-β expression in the injured tendons	This study hoped to determine the role of TGF-β signaling in equine tendon repair and whether its signaling can drive tenocyte differentiation by equine embryo-derived stem cells. These findings indicate a potential method for enhancing tendon repair through targeted cellular therapies
Dale et al., 2018 (1)	In vitro	N/A	Tenocytes were isolated from Sprague-Dawley rats. These cells were then cultured with various ESC solutions. Tenogenesis was attempted via hESC alone, hESC + BMP 12 and BMP 13 or hESC + BMP 12/13 + dorsomorphin (an inhibitor of the Smad signaling cascade)	Histological analysis of the culture hESC + BMP12/13 co-culture (with the addition of ascorbic acid) showed a morphology resembling tendons. In addition to this the reverse transcription polymerase chain reaction showed increased expression of COL1A2, COL3A1, DCN, TNC, THBS4, and TNMD. In contrast the cells cultured with BMP12/13 + dorsomorphin showed decreases in COL3A1, DCN, and TNC gene expression	Similar to the results observed in MSCs, this study exemplifies that ESCs may be able to be modified to be more or less effective at creating collagen

BMP, bone morphogenic protein; COL1A2, collagen type I alpha 2 chain; COL3A1, collagen type III alpha 1 chain; CTP, connective tissue progenitor; DCN, decorin; ESC, embryonic stem cell; fdESC, fetal-derived ESC; GFP, green fluorescent protein; hESC, human ESC; MRI, magnetic resonance imaging; MSC, mesenchymal stem cell; N/A, not applicable; SCX, scleraxis; SDFT, superior dorsal flexor tendons; TGF, transforming growth factor; THBS4, thrombospondin 4; TNC, tenascin C; TNMD, tenomodulin.

iPSCs

iPSCs are an extremely promising cell line that was first introduced in 2006 (22,98). Takahashi et al. (2007) were able to create pluripotent stem cells from adult fibroblast cells via the introduction and culture of four factors: OCT3/4, Sex-determining Region Y-box transcription factor 2 (Sox2), Myc proto-oncogene protein (c-Myc), and Krüppel-like factor 4 (Klf4) (98). This provides a tremendous advantage over ESCs from an ethical standpoint, eliminating the necessity of using an embryo while still achieving pluripotency (99). However, the challenges of teratoma formation observed with ESCs have also been observed in iPSCs (100,101).

Several studies have highlighted the potential of iPSC-derived cells in tendon repair, focusing on their differentiation into tenocyte-like cells, biomechanical support, and paracrine signaling. Xu et al. (2013) demonstrated that iPSC-derived neural stem cells enhance tendon healing and improve biomechanical strength via tenogenic differentiation in a rat model (102). Komura et al. (2020) developed a protocol for tenogenic differentiation by mimicking normal embryonic tendon development in iPSCs from mouse models (10). Their results showed an improved fiber structure and arrangement, although not fully equivalent to that of the uninjured tendon (10). Kaneda et al. (2023) then used iPSC-derived MSC-seeded 3D-printed polycaprolactone scaffolds in rat Achilles injuries, finding that this approach enhanced gait, tendon biomechanics, and tissue architecture compared to controls (103). Studies using human-derived iPSCs have shown comparable results, suggesting their clinical applicability. Bavin et al. (2015) showed that iPSCs gradually but effectively increased tendon-related genes in a 2D environment, although they struggled with 3D differentiation compared to ESCs (104). These results highlight the need for optimized differentiation protocols. Nakajima et al. (2021) showed improved motor function, tensile strength, and Achilles function index scores by using human iPSC-derived tenocytes in a rat Achilles model (105). Their results suggested both engraftment and paracrine effects through the secretion of IGF- and TGFβ-related proteins (105). Together, these findings show the potential usefulness of iPSC in tendon regeneration, especially human-derived iPSCs, which continue to show clinical potential when used in combination with optimized differentiation and scaffold techniques.

Perinatal stem cells (PSCs)

PSCs are a type of cell derived from tissues associated with pregnancy and birth. These tissues include amniotic-derived stem cells (ADSCs) and umbilical cord stem cells (UCSCs).

ADSCs

ADSCs are pluripotent stem cells derived from the amniotic environment. These cells are often derived from either amniotic fluid (AF) or amniotic fluid membranes (AFM) and are collected from females during the birthing process. This makes them easy to obtain and has few ethical implications. Muttini et al. (2010) explored the role of epithelial cell-derived ADSCs in tendon repair using an ovine model (106). They found that ADSCs survived for 30 days post-treatment, began forming tendon-like structures, and increased the presence of repair cells in the treated lesions. Their findings suggest that ADSCs play a regenerative role in tendon repair (106). Mirzayan and Suh (2022) addressed concerns regarding the anti-inflammatory nature of AFM-derived mesenchymal stem cells (AFM-MSCs) and whether they might inhibit scarring necessary for healing at the BTI (107). Their study revealed that ADSCs did not interfere with the BTI healing process, addressing concerns about their role in scar formation and confirming their usefulness for tendon repair (107).

Recently, Wang et al. (2025) examined the use of AFM scaffolds in a rat rotator cuff model and found that AFM scaffolds, especially when interposed at the BTI, significantly improved bone and cartilage formation, collagen alignment, and biomechanical strength compared to suture-only repairs (108). Together, these findings demonstrate that ADSCs and AFM scaffolds can enhance tendon repair, suggesting that ADSCs and AFM scaffolds hold promise for improving tendon regeneration through careful scaffold application and placement. However, there are a limited number of studies looking at tendon healing with this cell lineage, and of the ones presented, all are in different models. Thus, there is a great need for further investigation into the utility of ADSCs for tendon healing.

UCSCs

UCSCs are derived from the umbilical cord tissue and blood and can be obtained with minimal invasiveness. These stem cells can undergo great proliferation and differentiation, forming a variety of cell types, while also demonstrating low immunogenicity (66,109). Studies on UCSCs have shown promising results for tendon repair across animal models, although they have also revealed the need for consistent methodologies and controls. Starting with a 2009–2012 study on Warmblood horses, Van Loon et al. (2014) treated superficial digital flexor tendon injuries using UCSC injections and found that 77% of horses returned to work successfully (110). However, the interpretation was limited by a single-horse control group that fully recovered with rehabilitation alone, underscoring the importance of an adequate control group for assessing UCSC efficacy (110). Park et al. (2015) assessed the impact of UCSCs on a surgically induced subscapularis tendon tear model in rabbits (111). They observed that the UCSC-treated group exhibited smaller tear sizes, more organized fibroblastic bundles, and collagen regeneration than the hyaluronic acid- or saline-treated groups (111). The motion analysis also showed superior functional outcomes in the UCSC-treated group, including farther walking distance, faster walking time, and greater mean speed compared to saline controls (111). These findings highlight UCSC’s regenerative effect on tendon tissue.

Rak Kwon et al. (2020) examined UCSC’s effect in supraspinatus tendon injury model in rabbits, utilizing four groups: UCSC-only, polymer scaffold, polymer scaffold with UCSCs, and a saline control (112). They found that UCSC treatment, particularly when used with a scaffold, resulted in a significantly smaller tear size, higher collagen levels, and improved motion metrics (walking distance and speed) compared to the saline control. In addition, the UCSC scaffold group produced superior results compared with UCSCs alone (112).

Stem cell delivery and incorporation strategies

Direct injection of stem cells

Direct injection into the injury site is the most straightforward method of stem cell delivery (13). This method of direct injection of MSCs and ESCs into superficial digital flexor tendons of horses was evaluated by Guest et al. (2010) (96). A higher number of ESCs remained at the injection site than MSCs. ESCs also demonstrated a greater ability than MSCs to migrate to other damaged areas within the tendon. This could be because MSCs provide paracrine support in tendon regeneration by manipulating the environment, whereas ESCs play a role in proliferation and differentiation (96). MacLean et al. (2012) demonstrated a “significant improvement” in tendon healing in rabbits that received a direct injection of MSCs into the patellar tendon compared to a collagen-only control group (113). The group that received MSCs demonstrated significant increases in stress, modulus, and strain energy density of 26%, 18%, and 33%, respectively” (113). Yang et al. (2013) also discussed a clinical trial demonstrating improved healing of patellar tendinopathy following the direct injection of dermal fibroblasts suspended in autologous plasma (12). Furthermore, they emphasized the potential of dermal fibroblasts, suggesting that they should be considered a viable alternative to tenocytes, given the donor site morbidity associated with harvesting autologous tenocytes (12).

Stem cell-seeded scaffolds

Scaffolds are engineered 3D structures used to promote regeneration and healing (114). Stem cells are usually seeded onto or cultured on a scaffold before its implantation (13). Scaffolds aid in the healing of tendon tears by providing mechanical support while new ECM is produced, preventing re-tear/rupture. Scaffolds can also improve healing by promoting cell proliferation, matrix production, and matrix organization into functional tissues (12). Vasiliadis and Katakalos (2020) described certain properties required for an ideal scaffold to achieve tendon restoration and function (115). The first property is biocompatibility, which allows the scaffold to integrate into the injury site after implantation. Suggestions on how to promote this integration include carefully selecting and modifying the materials used. A negligible immune response is also important, as increased inflammatory responses from the immune system might decrease the healing ability of the scaffold. Ensuring the controlled biodegradability of scaffold materials is also necessary for the adequate integration of the scaffold into the surrounding tissue. Mechanical integrity is also key, as the scaffold needs to maintain its structure and integration within a dynamic physiological environment (115).

There are many types of scaffold materials used, with each one having a different effect on cell behavior. Scaffold materials can be classified into three categories: biological/natural, synthetic, and composite. Biological scaffolds are advantageous because of their ability to integrate into the injury site while eliciting a weak immunological response. These scaffolds are mainly composed of type I collagen, with other types of collagen, elastin, and proteoglycans also being used (115). In a recent study, citrate-based scaffolds were demonstrated to improve bone and cartilage healing and thus might be beneficial for tendon regeneration (116). Synthetic scaffolds are composed of polymers, such as polystyrene, poly (L-lactic acid) (PLLA), poly(glycolic acid) (PGA), and poly (D,L-lactic-co-glycolic acid) (PLGA), with PGA being the most commonly used. Synthetic scaffolds are generally more immunoreactive than biological scaffolds; however, they are typically more versatile and have better mechanical properties. Composite scaffolds are composed of two or more materials and have the added advantage of modulating cellular function and replicating the native microarchitecture of the tissue (115,117-120). Among these materials, polymer-ceramic composites are becoming widely used because of their osteoinductive characteristics, good biocompatibility, and corrosion resistance conferred by the ceramic component (115,121,122). However, their use is limited because of their poor mechanical strength and low degradability (115). In general, scaffolds allow for tailored mechanical and biological properties specific to different tissues and injuries in tissue engineering applications.

Stem cell-derived exosomes, extracellular vesicles (EVs), and secretomes

EVs, exosomes, and secretomes are often used interchangeably (123), but differ in size (EVs: 50–1,000 nm, exosomes: 30–100 nm, secretome: includes all secreted factors), origin (EVs: plasma membrane and endosomal, exosomes: endosomal multivesicular bodies, secretome: encompasses both vesicular and soluble factors), and composition (EVs: membrane proteins and cytosolic cargo, exosomes: specific proteins, secretome: all secreted molecules including EVs) (124-126). Despite these differences, they are similar as mediators of cell-to-cell communication, regulators of biological processes, and contributors to tissue repair and regeneration. They may contain various components including nucleic acids, proteins, lipids, and metabolites, which can be used to promote regeneration and healing (127). Exosomes achieve their regenerative abilities through several methods, including transferring regenerative factors and genetic material, modulating immune responses, and stimulating resident stem cells. Owing to their acellular nature, exosomes mitigate some risks associated with traditional stem cell therapies, such as tumor formation. Previous studies using exosomes have shown promising results for the treatment of skin, bone, and cartilage injuries (128). Lui (2021) reviewed five studies on BM-MSC EVs and reported consistent improvements in tendon healing through paracrine signaling (42). Lui noted EV’s stability, low risk of immune rejection, and ease of application over direct stem cells (42). Supporting these findings, Chen et al. (2021) used BM-MSC EVs in a rat rotator cuff model and revealed enhanced BTI healing with predominantly anti-inflammatory M2 macrophages over inflammatory M1 macrophages (129).

Li et al. (2022) created a rat ACL model and used BM-MSC exosomes (BM-Exos) with and without miR-23a-3p gene overexpression (BM-Exos mimic) for treatment (37). Both types of BM-Exos improved tendon-to-bone healing and biomechanical strength, with the BM-Exos mimic showing heightened efficacy by reducing M1 macrophages and increasing M2 macrophage activity through inflammatory pathway inhibition (37). Similarly, Shen and Lane (2023) tested inflammatory-primed AT-MSC EVs (iEVs) in mice with Achilles tendon injuries and found that iEVs alone reduced inflammation, enhanced collagen deposition, minimized scarring, and improved structural and functional recovery (51). These findings highlight the potential of primed EV environments in driving regenerative outcomes.

In studies that specifically examined tendon stem cell-derived exosomes (TDSC-Exos), Wang et al. (2019) developed an Achilles tendinopathy model in rats to demonstrate that both TDSCs and TDSC-Exos intra-tendon injection help balance matrix remodeling by decreasing MMP-3 expression, increasing tissue inhibitor of metalloproteinase 3 (TIMP-3) levels, and elevating tenogenic markers such as TNMD and collagen type 1 alpha chain 1 (COL1A1) (77). While both treatments promoted healing, TDSCs Exos delivered similar benefits to TDSCs solely through paracrine factors, offering a cell-free therapeutic approach. This strategy resulted in improved tendon biomechanics, including the maximum load and ultimate stress (77). Song et al. (2022) developed a rat tendon injury model and explored treatment with TDSC-Exos loaded onto a photopolymerizable hyaluronic acid (pHA) scaffold (78). The scaffold facilitated early tendon healing, improved fiber alignment, and enhanced biomechanics. Transcriptomics identified miR-144-3p within TDSC-Exos as a promoter of tenocyte proliferation and migration, suggesting that specific miRNAs may amplify the regenerative effects of TDSC-Exos (78).

Preconditioning of stem cells

The excitement surrounding the use of stem cells has increased with the recent discovery of the genetic engineering capabilities of iPSCs. This adds to the advantage of future stem cell use, as it eliminates the risk of transplanted cell rejection and avoids ethical concerns of using ESCs (102). Concerning stem cell use in tendon injury, Leong et al. (2016) discussed a group of tendon cells with the marker CD146 that demonstrated proliferation and tenogenic differentiation via the focal adhesion kinase (FAK)/extracellular signal-regulated kinases 1 and 2 (ERK1/2) signaling pathway when administered in conjunction with CTGF (130). They concluded that their findings, which supported the concept of conditioning/promoting TDSCs, could prove beneficial to tendon regeneration efforts (130).

Mechanisms of stem cell-mediated tendon repair

Stem cell therapy has been shown to be a promising treatment for tendon injuries, largely due to its contribution to tissue regeneration rather than tissue repair, the latter of which is usually accomplished through the deposition of scar tissue. Their regenerative ability can be attributed to their potential to differentiate into tenocytes. Tenocytes are specialized cells that produce tendon ECM components, such as type I collagen. While normal healing involves scar tissue formation, which eventually contains type I collagen, it is less organized than the tissue created by tenocytes and, therefore, has properties that are different from those of healthy, non-injured tissue. Tenocyte-mediated healing results in tissues that are most similar to those of healthy tendons.

Differentiation into tenocytes and tendon-like formation

Once stem cells have migrated to the zone of injury, they ideally commit to a tenogenic lineage. This process typically occurs during the proliferation phase of tendon healing. This commitment involves cultivating a variety of signals. Yin et al. (2016) found that exposure of BM-MSCs to TGF-β1, followed by CTGF, was an effective protocol for inducing the commitment of stem cells and initiating their tenogenic differentiation in vitro (131). In addition, Dale et al. (2018) found that exposure to BMP also demonstrated the ability to induce tenogenic differentiation in human ESCs (1). In addition, transcription factors, such as SCX (70) and Mohawk, and signaling pathways, such as the Wnt/β-catenin pathway, are crucial for tenogenic differentiation. When activated, these genes and pathways regulate the expression of tendon-related genes, which promote tendon development, tenocyte differentiation, and maturation (13).

Not all stem cells migrating to injured tendon sites differentiate into functional tenocytes. A critical fate decision node determines whether progenitor cells will become functional tenocytes [expressing SCX, mohawk homeobox (MKX), and TNMD] or differentiate into scar-forming myofibroblasts [expressing alpha smooth muscle actin (α-SMA) and producing disorganized type III collagen] (132,133). The decision of the fate node depends on the interplay within upstream signals or influential factors, which are mechanical cues, growth factor signaling, and inflammatory environment (132,133) (Figure 3). Optimal mechanical loading (4–8% cyclic strain) combined with transient TGF-β and BMP12/14 exposure promotes SCX/MKX activation and tenogenic differentiation (133-135). In contrast, sustained high-dose TGF-β, mechanical overload, or a Periostin-rich ECM niche favors myofibroblast differentiation and fibrotic scar formation (132,133) (Table 8). The activation of SCX and MKX plays a crucial role in activating tendon-specific genes and suppressing alternative progenitor cell fate (133). It is important to note that SCX and MKX functions are not mutually exclusive because MKX cannot function in the absence of SCX, despite the fact that it supports tendon differentiation and collagen fibril formation better. Thus, the role of early growth response 1 (EGR1) in the induction of SCX and the production of type I collagen is essential for guiding progenitor cells for tenocytic differentiation (133). Understanding these molecular cues in tendon repair has important therapeutic implications. Successful regenerative strategies must engineer both mechanical and biological microenvironments to favor tenogenic differentiation over fibrotic outcomes. Options may involve the delivery of pro-tenogenic factors (BMP12/14), controlled mechanical rehabilitation, modification of inflammation, or targeting pro-fibrotic signals (sustained TGF-β, periostin deposition) to shift the balance towards true tissue regeneration.

Figure 3.

Fate decision between tenogenic and fibrotic lineage. Numerous genetic and environmental factors play a role in determining whether stem cells differentiate towards a tenogenic fate versus a fibrotic fate. (Created in BioRender. Ogunsola, A. (2025) https://BioRender.com/by6p49e). α-SMA, alpha-smooth muscle actin; AKT, protein kinase B; BMP, bone morphogenetic protein; COL3, collagen type III; ECM, extracellular matrix; ERK, extracellular signal-regulated kinase; EGR1, early growth response 1; MAPK, mitogen-activated protein kinase; MKX, mohawk homeobox; mTOR, mammalian target of rapamycin; NF-κB, nuclear factor kappa-light-chain-enhancer of activated B cells; NGF, nerve growth factor; TAZ, transcriptional co-activator with PDZ-binding motif; TGF-β, transforming growth factor beta; TNMD, tenomodulin; TrkA, tropomyosin receptor kinase A; YAP, yes-associated protein.

Table 8. Signaling pathways and cues governing stem/progenitor cell fate decisions in tendon repair.

Pathway/cues	Key study	Model	Net effect	Mechanism/notes
Neuroregulation (NGF-TrkA signaling)	Cherief et al., 2023 (136)	Mouse Achilles tendon injury model with sural neurectomy and TrkA agonist intervention	Tenogenic	Tendon cells produce NGF after injury, which binds TrKA on sensory neurons and promotes nerve sprouting. The nerve-tendon interaction activates TSPC proliferation and TGF-β signaling. TrkA agonist treatment enhances TNMD expression and organized collagen repair. Neurectomy impairs tenogenic differentiation and results in fibrosis
ECM niche-periostin (POSTN) matrix and PostnLin progenitors	Ackerman et al., 2024 (132)	Murine flexor tendon repair (PostnMCM lineage tracing ± DTA ablation), human tendon scar tissue	Biphasic: protective (early and transient), Fibrotic (persistent)	Progenitors: PostnLin cells emerged D7-10 and differentiated into transient αSMA+ myofibroblast
				Mechanism: the matrix recruits and sustains myofibroblasts via integrin β1/β3 and F-actin
				Biphasic: early (D5-14) promotes mechanical recovery; persistence causes fibrosis
				Limitation: Does not enhance Scx or tenogenic differentiation
H19/miR-29b/TGF-β1 regulatory loop	Lu et al., 2017 (137)	TGF-β induced human TDSC differentiation	Tenogenic (when H19 is high/miR-29b low)	LncRNA H19 acts as miR-29b-3p sponge that prevents miR-29b from degrading TGF-β1 mRNA. It creates a positive feedback loop that amplifies TGF-β1 signaling during early differentiation, promoting collagen synthesis and tenogenic marker expression
miR-29b/TGF-β1/Smad3 axis	Chen et al., 2014 (138)	Rat Achilles tendon repair with chitosan treatment	Tenogenic (when TGF-β is reduced)	Chitosan upregulates miR-29b, which targets TGF-β1 mRNA, reducing TGF-β1/p-Smad3 expression, and dampening fibroblast proliferation (G1 arrest via P21). This dampens excessive fibroblast proliferation and collagen deposition, improving gliding function and reducing adhesion. Thus high miR-29b or low TGF-β1 signaling favors functional healing over fibrotic scarring
TGF-β/Smad2/3 signaling	Li et al., 2021 (139)	In vitro rat TSC culture treated with exosomes derived from TSCs enriched in TGF-β	Mixed (early-tenogenic, late-fibrotic)	Early repair (tenogenic): promotes TSPC proliferation, migration, and COL I/III synthesis and induces differentiation into SCX+ tenocytes
				Persistent activation (fibrotic): results in myofibroblast differentiation (α-SMA+) and excess ECM deposition, causing fibrosis and adhesions
				Mechanosensitivity: mechanical strain enhances this pathway via Smad2/3 phosphorylation; timing and magnitude determine tenogenic versus fibrotic outcomes (140)
TGF-β/cytoskeletal-mechanotransduction	Wang et al., 2016 (141)	In vitro hDF cultured on microgroove membranes enforcing elongated morphology	Mixed (low TGF-β1 dose—tenogenic, High TGF-β1 dose—fibrotic)	Elongated cell morphology enhanced endogenous TGF-β1 and ROCK activity, promoting SCX, TNMD, and collagen expression without inducing α-SMA. At low doses, TGF-β1 reinforced this tenogenic program, whereas high-dose treatment led to α-SMA activation and myofibroblast differentiation. Cytoskeletal tension and TGF-β dosage jointly determine the tenogenic or fibrotic fate (141)
YAP/TAZ (hippo-mechanotransduction) pathway	Xu et al., 2022 (142)	Irisin treatment of rat TSPCs (in vitro)	Tenogenic	Irisin activates YAP/TAZ by inhibiting proteasomal degradation, driving nuclear translocation, and dose-dependent upregulation of tenogenic markers (SCX, TNMD, and COL1) in rat TSPCs, while YAP inhibition with verteporfin completely abolished this tenogenic differentiation. This identifies YAP/TAZ as a critical binary switch determining whether TSPCs differentiate into functional tenocytes or non-tenogenic cells that characterize tendinopathy pathogenesis. The mechanism provides a molecular explanation for why exercise-based therapy works in tendinopathy: mechanical loading stimulates irisin secretion from the muscle, which activates YAP/TAZ signaling to promote tenogenic fate commitment in resident stem cells
TGF-β3 and BMP-12 signaling	Perucca Orfei et al., 2019 (143)	In-vitro culture of human TCs, ASCs and BM-MSCs treated with TGF-β3, BMP-12, ascorbic acid, and other growth factors	Tenogenic	TGF-β3 was the main inducer of SCX but inhibited DCN, a late matrix-maturation marker, during early induction. In contrast, BMP-12, b-FGF, and AA promoted DCN expression and collagen fiber maturation. Under controlled sequential stimulation, early TGF-β3 exposure followed by matrix maturation cues (BMP-12, FGF, and AA) resulted in tenogenesis. Prolonged or unmodulated TGF-β3 could impair collagen maturation and shift toward fibrosis (143)
TGF-β maintenance signaling	Tan et al., 2020 (135)	Mouse Scx-CreER/lox Tgfbr2 conditional knockout, lineage tracing, scRNA-seq, AAV rescue	Tenogenic	TGF-β/Tgfbr2-Smad2/3 signaling maintains differentiated tenocyte identity in a cell-autonomous manner. Loss of signaling in mature tenocytes causes dedifferentiation to a progenitor-like state (Sca-1, CD44+, CD34+) with loss of SCX, MKX, and TNMD expression and activation of wound-healing pathways. AAV-mediated Tgfbr2 reactivation restores tenogenic markers, demonstrating continuous TGF-β signaling is required for tendon cell fate maintenance and homeostasis
Inflammatory environment-macrophage polarization	Ackerman et al., 2017 (144)	Murine flexor tendon repair in diet-induced obese/T2DM mice	Fibrotic	Obesity/T2DM causes delayed M1 activation and excessive, prolonged M2 polarization, resulting in exuberant ECM deposition. Tendons show increased type I/III collagen, persistent adhesions, impaired gliding and reduced mechanical strength. While M2 macrophages are necessary for normal tissue repair, chronic and excessive M2-mediated matrix deposition drives the healing response towards fibrosis rather than tenogenic regeneration
ERK (MAPK) pathway	Morita et al., 2019 (145)	Human model of primary tendon derived cells from patients with torn rotator cuff vs. healthy donors (in vitro)	Fibrotic	TGF-β induces ERK1/2 activation, which drives tendon cell proliferation and adhesion. In diseased human tendon-derived cells, proinflammatory factors such as IL-1β induce strong ERK1/2 signaling, leading to increased expression of TGFB1, SERPIN E1, and BMP2. Subsequently, this increased α-SMA and COL3A1 expression, which are markers of fibrosis. Healthy tendon cells only showed transient ERK activation and minimal fibrotic gene induction (139)
NF-κB Signaling	Best et al., 2020 (146)	Murine flexor tendon repair (IKKβ knockout) model	Fibrotic	Chronic activation (fibrotic): chronic NF-kB activation during tendon repair upregulates COL1A1/COL3A1 expression, maintains myofibroblast survival, sustains inflammation, and promotes fibrosis (139)
				Transient inhibition (therapeutic): transient or pharmacological NF-kB inhibition (147) reduces inflammatory cytokine levels and fibrous tissue formation, improving collagen organization (147)
				Complete suppression (pathological): however, conditional IKKβ deletion completely suppresses NF-kB signaling, leading to excessive apoptosis, myofibroblast loss, and disorganized periosteum-rich scar (146)
				Conclusion: moderate NF-kB activity is essential for tenocyte survival, and chronic activation or complete loss of NF-κB activity disrupts normal healing
AKT-mTOR (mechanotransduction and translational control)	Cong et al., 2018 (148)	Mouse MSCs, mTOR-KO mice	Tenogenic (when activated)	Morphogenic growth factors (TGF-β1 and IGF-1) and cyclic strain activate AKT-mTOR signaling to drive tenogenesis through the translational control of collagen I and ECM proteins, promoting SCX, TNMD, and COL1A1 expression. Critically, mTOR regulates cell fate at the post-transcriptional level and its inhibition or genetic ablation reduces collagen protein and fibril diameter without affecting tendon gene transcription. mTOR is downregulated in human tendinopathy and by statin treatment. AKT-activated MSCs generate superior tendon-like repair tissues with improved biomechanics in vivo
PI3K-Akt signaling (curcumin-Induced tenogenesis)	Zhang et al., 2024 (149)	Rat Achilles tendon injury model; rat TSPCs in vitro	Tenogenic	Curcumin activates the PI3K/Akt pathway in TSPCs, increasing p-Akt and driving the upregulation of SCX, TNMD, COL1A1, and TNC. This activation suppresses stemness markers (e.g., nucleostemin) and commits cells to a tenogenic fate. Inhibition of PI3K (LY294002) abolished all curcumin-induced effects, confirming that PI3K/Akt is a critical decision node for tenogenic differentiation
HIF-1α	Guo et al., 2022 (47)	AT-MSCs in vitro	Tenogenic	HIF-1α is a transcription factor that regulates mitochondrial metabolism and stress on the endoplasmic reticulum. In a hypoxic environment AT-MSCs showed marked reduction of HIF-1α expression and increased expression levels of tenogenic markers (SCX, TNMD, DCN) and VEGF. Paradoxically, introduction of a HIF-1α inhibitor decreased the expression of those tenogenic markers, indicating that the HIF-1α has complex, context-dependent effects on tendon cell differentiation

α-SMA, alpha-smooth muscle actin; β-FGF, basic fibroblast growth factor; AA, ascorbic acid; AAV, adeno-associated virus; AKT, protein kinase B; ASC, adipose-derived stem cell; AT-MSC, adipose tissue-derived MSC; BM-MSC, bone marrow-derived mesenchymal stem cells; BMP, bone morphogenetic protein; CD34, cluster of differentiation 34; CD44, cluster of differentiation 44; COL1/COL1A1, collagen type I alpha 1; COL3/COL3A1, collagen type III alpha 1; DCN, decorin; DTA, diphtheria toxin A; ECM, extracellular matrix; EGR1, early growth response 1; ERK, extracellular signal-regulated kinase; FGF, fibroblast growth factor; hDF, human dermal fibroblasts; HIF-1α, hypoxia-inducible factor 1-alpha; IGF-1, insulin-like growth factor 1; IKKβ, inhibitor of nuclear factor kappa-B kinase subunit beta; IL-1β, interleukin-1 beta; lncRNA, long non-coding RNA; M1, M1 macrophages (pro-inflammatory); M2, M2 macrophages (anti-inflammatory); MAPK, mitogen-activated protein kinase; miR-29b, microRNA-29b; MKX, mohawk homeobox; mRNA, messenger RNA; MSC, mesenchymal stem cells; mTOR, mammalian target of rapamycin; NF-κB, nuclear factor kappa-light-chain-enhancer of activated B cells; NGF, nerve growth factor; PI3K, phosphoinositide 3-kinase; POSTN, periostin; PostnLin, periostin-lineage cells; ROCK, Rho-associated protein kinase; Sca-1, stem cell antigen-1; scRNA-seq, single-cell RNA sequencing; SCX, scleraxis; Smad2/3, mothers against decapentaplegic homolog 2/3; T2DM, type 2 diabetes mellitus; TAZ, transcriptional co-activator with PDZ-binding motif; TC, tendon cells; TDSC, tendon-derived stem cells; TGF-β, transforming growth factor beta; Tgfbr2, transforming growth factor beta receptor 2; TNC, tenascin-C; TNMD, tenomodulin; TrkA, tropomyosin receptor kinase A; TSC, tendon stem cells; TSPC, tendon stem/progenitor cells; YAP, yes-associated protein.

Certain studies have used some of the approaches suggested above to determine whether some of the influential cues can be altered to favor tenocytic regeneration rather than fibrosis formation upstream or downstream. Ackerman et al. (2022) used lineage cell tracing and spatial transcriptomics profiling to classify SCX-lineage cells from murine flexor tendons into six clusters (C0 to C5), identifying the C3 subtypes as having a fibrotic trajectory due to the presence of profibrotic markers (150). Understanding these spatial patterns could assist in identifying treatment targets. Another study showed that the reduction in TGF-β effect by knocking down one of its downstream effectors, plasminogen activator inhibitor-1 (PAI-1), which is a downstream mediator in TGF-β-induced fibrotic healing, reduced the generation of fibrotic scars in injured flexor tendons without the loss of mechanical properties (151). Finally, several studies have shown that targeted inflammatory modification in tendon models with low dose systemic corticosteroid injection, as well as controlled joint loading during the early remodeling or proliferative phase of injury, can improve mechanical strength and reduce fibrosis (152-154). Dietrich-Zagonel et al.’s (2024) study emphasize that injecting corticosteroids could be potentially harmful if overdosed, given too early, or under high mechanical stress, as it suppresses SCX/TNMD and impair tendon-cell matrix formation (154). The use of corticosteroids for tendon injuries is controversial, and these studies, while insightful, do not help clinicians decide on the use of corticosteroids due to the heterogeneity in the route of administration, uncertainty of timing and appropriate dosing, and the presence of too many uncertain variables that need to be controlled. Therefore, more clinical studies and randomized controlled trials are needed to determine the usefulness of steroids in tendon healing. Currently, the authors would still err on the side of caution and avoid steroids for the treatment of tendon tear post-repair until more compelling studies are available.

Paracrine effects and modulation of the tendon microenvironment

This section focuses on the role of MSCs in modulating the tendon microenvironment during regeneration.

Angiogenesis

Angiogenesis, the generation of new blood vessels, is a key component of tendon tissue regeneration. Stem cells contribute significantly to this process through the release of key growth factors, such as VEGF, TGF-β, and bFGF. These growth factors then initiate and promote angiogenesis. Angiogenesis is necessary in the healing process because it provides a network that allows the delivery of extrinsic cells, nutrients, and other growth factors to the injured area, which are all necessary for proper repair. The timing of angiogenesis is also critical, as earlier angiogenesis has been shown to improve healing and repair, whereas delayed or insufficient vessel formation can impair healing or lead to degenerative processes (155).

MSCs have also been shown to play a regulatory role in angiogenesis via the VEGF pathway. MSCs secrete VEGF, which binds to its receptor, located on the endothelial cell membrane. This VEGF/VEGFR interaction activates the phosphoinositide 3-kinase (PI3K) and mitogen-activated protein kinase (MAPK) signaling pathways, which inhibit the phosphorylation of Yes-associated protein 1 (YAP1). This allows for the activation of YAP1, which induces angiogenesis by enhancing the expression of certain angiogenic genes and provides positive feedback by promoting further VEGF production. MSC-mediated induction of angiogenic signaling pathways facilitates the optimization of both the timing and intensity of the angiogenetic process (156).

Immunomodulation

MSCs play an important role in modulating the immune response at the site of tendon injury. They play a dual role in both pro- and anti-inflammatory processes. They accomplish this through interactions with several types of immune cells and by secreting proteins, such as macrophage inflammatory protein-1 (MIP-1) (157). Initially, MSCs promote an inflammatory environment by activating T-cells and enhancing lymphocyte recruitment. In their anti-inflammatory capacity, MSCs can influence the polarization of macrophages, a process in which macrophages differentiate into either M1 (inflammatory) or M2 (anti-inflammatory) types. M1-type macrophages predominate during the early stages of repair, whereas the M2 type accumulates over time and eventually takes over during the latter anti-inflammatory healing stages. Some studies have shown that AT-MSCs can induce polarization of M1 macrophages into M2 macrophages, thereby limiting inflammation through immunomodulation (157).

MSCs can also act as sensors for immune system responses to injury. Through Toll-like receptors (TLRs) on their surface, MSCs can become polarized in response to their environment, similar to macrophages. When the inflammatory response becomes overactive, MSCs are triggered to switch gears into an anti-inflammatory mode to temper the immune response. This can be accomplished through the induction of M2 macrophages and the release of anti-inflammatory cytokines.

ECM remodeling

MSCs contribute to ECM remodeling, which is essential for providing structural support and facilitating tissue regeneration. They secrete MMPs and other remodeling enzymes that degrade and reorganize the ECM, creating a scaffold for new tissue formation and allowing stem cells to invade the injured area through the membranes (158). In addition, their ability to differentiate into tenocytes means that they can produce tendon ECM components, such as type I collagen, elastin, and glycosaminoglycans. MSCs also secrete various growth factors and cytokines that regulate ECM synthesis and remodeling by other cells, such as fibroblasts and existing tenocytes. Chen et al. (2009) found that hESC-MSCs accelerate ECM creation and remodeling (6).

Recruitment and activation of endogenous progenitor cells

Following tendon injury, inflammatory processes attract immune cells to the site of injury. These immune cells, as well as the cells in the injured tissue, produce growth factors and cytokines that attract MSCs to the injured area via vascular transport. Common cytokines involved in this process include TGF-β, interleukin (IL)-1, IL-6, and monocyte chemoattractant protein-1 (159). Integrin expression is upregulated on the surface of endothelial cells, allowing MSCs to migrate through vessel walls to the site of injury (159). Damage also causes ECM remodeling. The process of this remodeling exposes binding sites and is accompanied by the release of growth factors, such as VEGF and PDGF, which further attract MSCs. In addition, interferon gamma (IFN-γ) and tumor necrosis factor alpha (TNF-α) secreted by immune cells at the site of injury further enhance MSC homing (160).

Clinical trials and translational considerations

Overview of clinical trials

Several clinical trials have investigated different approaches to tendon repair. Connell et al. (2009) conducted a clinical trial to evaluate the efficacy of collagen-producing cells derived from skin fibroblasts for the treatment of refractory lateral epicondylitis (161). Skin tenocyte-like cells yielded a substantial number of collagen-producing cells when cultured, thereby demonstrating the feasibility of their therapeutic use. In a prospective study involving 12 patients with refractory lateral epicondylitis, aged approximately 39.1 years, these cells were injected into tendon injury sites. Ultrasound assessments revealed tear reduction, new vessel formation, and tendon thickness. Patients experienced significant decreases in pain and functional disability, as evaluated by the patient-rated tennis elbow evaluation (PRTEE). Eleven patients reported satisfactory outcomes, with only one requiring surgery after treatment failure (161).

Usuelli et al. (2018) compared injections of the stromal vascular fraction (SVF) from adipose tissue versus PRP for non-insertional Achilles tendinopathy in a prospective study of 44 patients (162). Both treatments showed significant improvement, with SVF initially demonstrating superior outcomes in pain reduction, suggesting faster efficacy than PRP. At later time points, both groups showed comparable effectiveness, suggesting that SVF might offer quicker recovery to return to sports or daily activities (162).

Randelli et al. (2022) performed a randomized controlled trial that explored the safety and efficacy of autologous lipoaspirate tissue in arthroscopic rotator cuff repair (163). In this context, lipoaspirate tissue functioned as a source of AT-MSCs. Following single-row arthroscopic rotator cuff repair, the treatment group received intraoperative injections of autologous micro-fragmented adipose tissue, whereas the control group received no injections. After 6 months, the treatment group demonstrated superior results. However, further time points revealed no differences between the two groups. There were no significant differences in reinjury rates, complications, or adverse events between the treatment and control groups (163). This trial showed that these injections are safe and effective for short-term results after a single-row arthroscopic rotator cuff repair.

Clinical trials investigating tendon repair therapies have highlighted important safety and efficacy considerations. For example, studies assessing stem cell therapies for tendon disorders often face moderate to high risks of bias, particularly concerning outcome measurements, such as ultrasound and PROMs, as highlighted by systematic reviews such as Van Den Boom et al. (2020) (16). Non-randomized trials also raise concerns about participant selection bias, impacting the reliability of the findings (16). In addition, Connell et al. (2009) examine the use of collagen-producing cells from skin fibroblasts for treating refractory lateral epicondylitis, but uncertainties persist regarding the exact nature and mechanisms of these cells within the tendon tissue (161). This raises concerns about the efficacy of the treatment, as well as the safety of this procedure.

While studies specific to tendon injuries have not reported adverse effects of stem cell treatments so far (16), potential risks include bloodstream infections, joint infections, epidural abscesses, and particularly the formation of ectopic bone, tumors, and angio proliferative lesions, all of which have been observed in various stem cell therapeutic studies (164-166). Moreover, the process of obtaining tendon-derived stem progenitor cells may pose risks and complications to the donor tissue (164). These findings underscore the importance of rigorous safety monitoring and further research to clarify and mitigate the potential risks associated with stem cell therapies for tendon repair.

Challenges and future perspectives

Challenges and limitations of stem cell-based therapies

Stem cell therapy for tendon healing and regeneration is of great clinical interest. Although our knowledge of how various stem cells work and the cellular microenvironment has improved tremendously over the years, there remain challenges and limitations with stem cell-based therapies. Concerns surrounding the use of various stem cells for treatment are related to their source and potential for teratoma formation, particularly with the use of ESCs (20,123). Additionally, ESCs sourced from developing embryos have sparked ethical conflicts owing to the destruction of such embryos during the process (167). iPSCs have become an alternative to ESCs, but their potential use in human cloning and the risk of teratoma formation remain a concern (167).

Exploring ways to mitigate risks of teratogenicity, several studies have found that ADSCs have a low tumor-forming potential in vivo (168-170), however, there is some evidence that it can promote tumors under the right conditions (171,172). In fact, Meng et al.’s (2019) study challenged the known norm about the safety of ADSCs and showed that human ADSCs (hADSCs) promoted the proliferation of cancer cells but inhibited cancer cell motility in mice (172). Their study demonstrated that human umbilical stem cells may be a safer alternative to hADSCs because they inhibit cancer growth and motility in their experiments (172). Given these controversial reports on the safety profile of stem cell-based therapies, more evidence is needed.

Another safety concern, particular with the use of IPSCs, is the risk of viral insertional mutations owing to the use of the viral system to deliver reprogramming factors into somatic stem cells. This method of creating iPSCs not only elicits safety concerns but can also lead to unintended consequences due to the potential integration of residual viral DNA and unintended re-expression of reprogramming genes, which is another pathway to teratogenicity (167,173,174). Another unintended consequence of re-expression of reprogramming genes is genomic instability and challenges with the maturation of iPSC-derived cells, which can lead to reduced efficacy (173,174). The need to devise various ways to remove undifferentiated iPSCs from cellular products to reduce the risk of teratomas and genetic analysis at several stages of preparation to ensure that it is free from deleterious mutations makes the use of iPSCs challenging clinically (174).

Much remains to be done to ensure the safety of iPSCs. Non-integrating methods have been developed to generate iPSCs to minimize risks such as insertional mutagenesis. Stadtfeld et al. (2010) used non-integrating adenoviral vectors that provided transient expression of reprogramming factors (Oct 4, Sox2, Klf4, and c-Myc), ensuring that the resulting iPSCs were viral integration-free (175). A disadvantage of this approach is its lower reprogramming efficiency, which leads to dilution over time and loss of reprogramming factor expression. Recently, Conrad et al. (2024) used episomal plasmids encoding reprogramming factors to generate transgene-free porcine iPSCs from porcine fibroblasts (176). This new technique showed more stable expression of endogenous pluripotency factors without genome integration of the episomal plasmids.

The risk of immunogenicity seems to vary among the various stem cell categories. ESCs have low immunogenicity in their undifferentiated state (173); however, immunogenicity increases once they differentiate owing to an increase in the number of donor major histocompatibility class (MHC) proteins on their surface (168). Adult stem cells, such as MSCs, are considered less immunogenic when derived from the same individual (autologous) (173), but the risk persists with allogeneic forms. In contrast, ADSCs have low immunogenicity due to reduced expression of MHC class I antigen (HLA-ABC) and its co-stimulatory molecules (CD40, CD80, and CD86), and lack of MHC class II (HLA-DR) antigen and β2 microglobulin expression (168,177). In contrast, some studies have shown that human amniotic cells can stimulate both the innate and adaptive immune responses (177-179).

Recently, there has been a shift towards the use of stem cell secretomes/exosomes for treatment because their use eliminates the risk of immunogenicity and tumor formation because they are devoid of cellular components (123,180). Many studies have shown that they are as effective as stem cells because of their abundant growth factor and cytokine content (123,180). Compared to stem cell therapy, which requires stringent storage and transport requirements, exosomes are stable for transport and long-term storage (180). Moreover, the use of stem cell secretomes eliminates ethical concerns associated with stem cell use. One notable limitation of the use of secretomes is the lack of standardized purification and storage methods (180). As with other stem cell-based therapies, including secretomes, the lack of standardization and characterization of stem cell treatments makes their clinical application challenging. The need for stem cell characterization and standardization is increasing because it is crucial to ensure the reliability of stem cell therapies (168). While liquid chromatography/mass spectrometry, protein microarray, and cytokine array have been used to define stem cell populations, cytokine, and protein content, the protein content in various secretomes varies in concentration, and there have been inconsistent findings in the literature owing to the heterogeneity of stem cell populations (168,181). Moreover, stem cell sensitivity to their microenvironment affects their differentiation pattern, as well as protein and cytokine secretion, contributing to inconsistent findings (169,181).

Finally, the interpretations of tendon injury studies, as well as their measured outcomes, must be carefully considered. For example, caution must be exercised when interpreting studies that show reduced scar formation, as there may be a complex trade-off for maximal load. Epanomeritakis et al. (2024) noted this trade-off in a systematic review of nine flexor tendon studies that used adipose-derived stromal cells, where cellular therapy effectively reduced scar and adhesion formation but did not significantly improve maximum load compared to suture-only repair (182). This outcome may be due to the antifibrotic effect of treatment, which may initially compromise tensile strength and is probably dependent on early fibrotic scar formation (1-4). Variations in results were attributed to differences in cell delivery methods, with scaffold-seeded cells and intratendinous injections showing superior maximum load compared to droplet administration, although inconsistent control of cell number, retention, and follow-up periods complicates the interpretation (182).

Future research directions and emerging trends

The retention, survival, and integration of stem cells into damaged tissues is a major challenge in stem cell therapy. To enhance the stability and integration of stem cells in tissues, researchers have explored new delivery strategies such as genetic manipulation, biological or material incorporation, and pharmacological methods of improvement.

Genetic techniques such as somatic cell nuclear transfer (SCNT) can generate embryonic cells genetically identical to the donor from somatic cells (therapeutic cloning) (167,183). This process has already been proven in mice and non-human primates and could be useful for personalized treatments, including the regeneration of damaged tendons (167,183). Moreover, SCNT can address concerns regarding immunogenicity, offer the ability to study disease models and explore therapeutic approaches (183). SCNT has some challenges, such as the low efficiency of deriving ESCs, chromosomal abnormalities, and immune rejection due to the presence of mitochondrial DNA from the oocyte donor. Byrne et al. (2007) studied the derivation of ESCs from SCNT in rhesus macaques and solved the low-efficiency problem by reporting an increased blastocyst formation rate (16%) compared with previous studies (183). Furthermore, their study partially resolved the chromosomal abnormality problem by demonstrating that normal karyotypes could be achieved using SCNT-derived ESCs. However, the challenges of immune rejection due to the presence of mitochondrial DNA in the oocytes of donors are yet to be resolved, and their model is still at the proof-of-concept stage. The development of advanced gene-editing tools (CRISPR-Cas9) and various methods for reprogramming cells could also provide solutions to immune-related and genomic instability problems (173). There is still a long way to go to make genetically engineered stem cell therapy safe for clinical applications. Based on current evidence, the use of stem cell-derived secretomes/exosomes and autologous MSCs appears to be the most promising approach for promoting safer therapeutic outcomes in tendon repair.

Other emerging trends in tendon repair and regeneration have focused on the development of novel tissue engineering and biofabrication techniques to address the challenges associated with tendon injuries. The most crucial task is to develop scaffold materials that closely mimic the structural and functional properties of the native tendons. Freedman et al. (2022) reviewed recent developments toward integrated tendon therapies, emphasizing the importance of biomaterials that provide mechanical cues and controlled drug delivery to enhance tendon healing (184). They highlighted promising strategies such as injectable hydrogels and electrospun scaffolds loaded with growth factors or anti-inflammatory drugs, underscoring the potential for these systems to simultaneously provide structural support and sustained bioactive molecule release (184). Natural materials, such as collagen, fibrin, and silk fibroin, offer biocompatibility and support for cell adhesion, while synthetic materials, such as poly(glycolide-co-caprolactone) (PGCL) and PCL, provide mechanical strength (185). However, synthetic biomaterials must be enhanced to improve their bioactivities. Technologies such as electrospinning and 3D bioprinting have been used to fabricate scaffolds with highly aligned fibers, which are critical for restoring tendon function. These technologies are promising for the production of scaffolds that replicate the biological environment and mechanical properties of tendons (185). Additionally, integrating mechanical loading and strain environments using advanced bioreactors can enhance the functional regeneration of engineered tendons by increasing collagen type I production, collagen fibril alignment, cross-sectional area, cellularity, and tensile strength of tendons (185-187). However, Ning et al. (2023) noted that current bioreactor technologies are still unable to replicate the intricate ECM remodeling process observed in native tendon tissues (185). This limitation opens an avenue for the further development of bioreactor technology.

For large tendon defects in which end-to-end repair is impossible, surgeons usually employ grafts or tendon transfers for repair. Recent advancements have ensured that scaffolds with seeded stem cells can be incorporated into graft repair to enhance healing under appropriate conditions. However, Ning et al. (2023) highlighted key challenges in tendon scaffold design, including the need to align scaffold degradation rates with tissue regeneration and to improve interactions between seeded cells and biomaterials (185). Nevertheless, recent studies suggest that new approaches may help address some of these limitations. Jiang et al. (2022) designed a 3D-bioprinted multilayered scaffold seeded with spatially differentiated AT-MSCs to mimic the tendon-to-bone interface (188). Their scaffold promoted better collagen organization, fibrocartilage formation, and histological healing, though the improvement in biomechanical strength was limited (188). Aykora et al. (2025) characterized the use of artificial intelligence to guide bioink formulation and scaffold design (189). This AI-driven approach may improve scaffold precision and function, enabling the fabrication of tendon constructs that more closely replicate native tissue structure and mechanical properties (189). Additionally, Sarmiento and Little (2021) reviewed multiomics approaches (genomics, transcriptomics, proteomics, metabolomics) in tendon regeneration research. They demonstrated how integrating these techniques could help identify molecular differences associated with tendon healing, degeneration, and engineered constructs (190). Their review suggests that multiomics data could inform and improve scaffold design and lead to targeted, personalized regenerative therapies.

Given the role of macrophages in tendon healing, another promising area of research involves modulating macrophage polarization to create a precise early inflammatory environment via M1 macrophages and a late anti-inflammatory environment via M2 macrophages to support tissue repair and regeneration (185,191). Future directions in tendon repair will involve the combination of carefully selected scaffolds, growth factors, and cytokines from stem cells or exosomes, and gene therapy to promote tendon healing.

Conclusions

The use of stem cells to enhance tendon healing is an exciting prospect that has garnered considerable attention and research over the past few decades. However, further research is needed to determine the best method to refine the application of this therapy. In our review, promising results were observed across all different cell lineages explored without a predilection for one specific lineage. However, some stem cells, including AT-MSCs, are far easier to harvest and are associated with far fewer ethical concerns than alternative stem cell lineages. Similarly, when considering the administration of stem cell therapy, secretomes may be more advantageous than whole stem cells. They carry fewer risks, including less immune rejection and decreased tumorigenicity, yet still produce healing outcomes similar to those of whole stem cells.

Advances in stem cell and scaffold engineering for tendon repair are likely to have implications in clinical practice in the coming years. With the continued enhancement and improved safety of these therapies, they are likely to complement surgical intervention. Although surgical intervention is inevitable for large tendon defects, stem cell therapy combined with scaffolds, when needed, could serve as a complementary therapeutic option. This is advantageous, given that surgical repair often relies on scar tissue to enhance bone-to-tendon healing, whereas stem cell therapy has been shown to promote anatomical tendon microanatomy. Thus, stem cell therapy combined with surgical repair may lead to more organized collagen deposition, fewer post-injury adhesions, less scarring, increased biomechanical strength, and greater functional outcomes. This will decrease the patients’ pain and increase PROMs postoperatively. Stem cell therapy could also help the body to heal small tears that may not require surgical repair, particularly among at-risk and low-demand patient populations. This may mitigate some of the inherent risks of surgery, including anesthetic complications, deep vein thrombosis, and iatrogenic damage.

Overall, the ever-changing landscape of stem cell tendon therapy is rapidly advancing and full of potential and promise.

Supplementary

The article’s supplementary files as

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.