Stem cell therapy: a promising biological strategy for tendon–bone healing after anterior cruciate ligament reconstruction

Abstract

Tendon–bone healing after anterior cruciate ligament (ACL) reconstruction is a complex process, impacting significantly on patients' prognosis. Natural tendon–bone healing usually results in fibrous scar tissue, which is of inferior quality compared to native attachment. In addition, the early formed fibrous attachment after surgery is often not reliable to support functional rehabilitation, which may lead to graft failure or unsatisfied function of the knee joint. Thus, strategies to promote tendon–bone healing are crucial for prompt and satisfactory functional recovery. Recently, a variety of biological approaches, including active substances, gene transfer, tissue engineering and stem cells, have been proposed and applied to enhance tendon–bone healing. Among these, stem cell therapy has been shown to have promising prospects and draws increasing attention. From commonly investigated bone marrow‐derived mesenchymal stem cells (bMSCs) to emerging ACL‐derived CD34+ stem cells, multiple stem cell types have been proven to be effective in accelerating tendon–bone healing. This review describes the current understanding of tendon–bone healing and summarizes the current status of related stem cell therapy. Future limitations and perspectives are also discussed.

Introduction

ACL rupture is a common sports injury with a prevalence estimated to be 1 in 3000 in the United States. It may result in a spectrum of secondary symptoms, including knee laxity, movement dysfunction, meniscus and cartilage damage and even early post‐traumatic osteoarthritis, which exerts a heavy burden on the patients and impact heavily on the society 1, 2, 3, 4.

As for the poor intrinsic regenerative properties of ACL, conservative and surgical suture strategies often result in the failure of ACL 5, 6, 7. Currently, surgical ACL reconstruction using autologous graft is the primary approach to treat ACL tears, which exhibits superior clinical safety and graft biocompatibility compared with allogenic or artificial graft 8, 9, 10, 11, 12. The hamstring tendon and the middle part of the patella tendon with bone plugs attached are commonly used as the graft for ACL reconstruction. However, for the risk of pain and morbidity of the knee joint by using patella tendon autograft, hamstring tendon autograft is becoming more popular 13, 14, 15. With the improved knowledge of ACL healing characteristics and development of biological strategies, some novel ACL repair approaches like bio‐enhanced repair 16, 17 and the dynamic intraligamentary stabilization (DIS) system 18, 19, 20 have been introduced. These alternative approaches have been proved effective for ACL rupture through plenty of animal trials 21, 22, 23, 24. They might be potential techniques to treat ACL tears in the future but more clinical researches are necessary before.

Although the reconstruction surgery can provide initial stability of the knee joint, there are still masses of issues persist post‐operatively. The interface between the transplanted tendon and the bone tunnel usually heals with a fibrous scar tissue rather than a natural insertion in a few months, failing to restore the normal joint kinematics and kinetics 25, 26, 27, 28, 29, 30. As a consequence, rehabilitation programme has to be delayed to protect the fragile tendon–bone interface and the ultimate function of knee joint is often adversely restricted 31, 32, 33. Additionally, the changed mechanical structure may lead to cartilage wear and meniscus injury, eventually resulting in the early post‐traumatic osteoarthritis 34, 35, 36. Therefore, to restore the solid bio‐mechanical property of tendon–bone interface is critical for patients to return to normal work and sports activities.

Focusing on enhancing tendon–bone healing, plenty of biological strategies, including active substances like platelet‐rich plasma 37, 38, 39 and growth factors 40, 41, gene transfer 42, 43, 44, 45, tissue engineering 46, 47, 48, 49, 50 and stem cells therapy 51, 52, 53, 54, 55, 56, 57, 58 have been introduced and aggressively developed in recent years. Among these treatments, stem cells derived from multiple tissues, such as bMSCs 51, 52, 53, 54 and synovial MSCs (sMSCs) 55 have been demonstrated promising potential in promoting tendon–bone healing. Recently, CD34+ stem cells from ACL tissues have presented admirable performance on tendon–bone healing 56, 57, 58. The aim of this review was to describe the current understanding of tendon–bone healing and summarize current status of stem cells therapy on enhancing tendon–bone healing.

Biological characteristics of tendon–bone healing

The native tendon–bone attachment of ACL is a typical direct insertion that transmits complex mechanical loads from soft ligament to hard bone. This fibrocartilage structure consists of four distinct and transition histological zones: ligament, unmineralized fibrocartilage, mineralized fibrocartilage and bone, which allows a gradual change in stiffness and decreases stress concentration 59, 60, 61, 62.

The biological mechanism of tendon–bone healing process after ACL reconstruction is incompletely understood. Based on current animal trials, the chronological and morphological changes of tendon‐to‐bone interface have been illustrated 26, 63, 64, 65, 66, 67. Instead of regenerating the four zones of the native direct insertion, the graft heals with an interposed layer of fibrovascular scar tissue at the graft–tunnel interface. Initially, the fibrovascular tissue arranges in disorder. By several weeks after surgery, the collagen organizes and forms perpendicular fibres resembling the Sharpey's fibres of an indirect insertion. Eventually, the surrounding bone grows into the interface tissue and incorporates with the outer portion of the graft, enhancing graft attachment strength.

It has been demonstrated that the fibrous insertion shows inferior mechanical property compared with the native ACL insertion 25, 26, 27, 28, 29, 30, leading to suboptimal functional recovery, higher risk of reinjury and post‐traumatic osteoarthritis 31, 32, 33, 34, 35, 36. Thus, enhancing tendon–bone healing to restore the natural biomechanics is crucial for ensuring rapid return to pre‐injury activity level.

Stem cells therapy for tendon–bone healing: current status

Possessing the capacities of self‐renewing and multi‐differentiation, stem cells have been used in tissue regeneration 68, 69, 70. Recently, the therapeutic potential of stem cells for tendon–bone healing has been investigated in animal models (Table 1). From the widely studied bMSCs to the emerging CD34+ ACL‐derived stem cells, different kinds of stem cells have shown promise prospects in promoting tendon–bone healing.

Table 1.

Recent studies on stem cells therapy for promoting tendon–bone healing

Cells	Graft	Carrier	Animal Model	Histological evaluation	Biochemical evaluation	Other evaluations	Author	Year
bMSCs	Hallucis longus tendons	Fibrin glue	Rabbit	More perpendicular collagen fibres and cartilage‐like cells	None	None	Ouyang 51	2004
bMSCs	Hamstring tendon	Fibrin glue	Rabbit	A fibrocartilage interface	Higher failure loads and stiffness	None	Jit‐Kheng Lim 52	2004
PDGF‐B gene‐transfected bMSCs	Achilles tendon allograft	Fibrin glue	Rabbit	More hypervascularity and a structure similar to normal insertion	None	None	Feng Li 103	2007
bMSCs	Achilles tendon allograft	Fibrin glue	Rabbit	A fibrocartilage interface	Higher failure load	None	Michael Y. H. Soon 53	2007
sMSCs	Achilles tendon Autograft	Atelocollagen gel	Rat	More Sharpey's fibres	None	Implanted sMSCs differentiated into fibroblasts	Young‐Jin Ju 55	2008
Human adult ACL ruptured tissues‐derived CD34+ stem cells	Autologous flexor digitorum longus tendon	Intra‐articular injection	Rat	A fibrocartilage interface and enhanced angiogenesis and osteogenesis	Higher failure load	Implanted cells differentiated into endothelial cells and osteoblasts	Yutaka Mifune 56	2012
Autologous Ruptured Tissue	Autologous flexor digitorum longus tendon	Sutured to the tibial side of the graft	Dog	Mature bone ingrowth and enhanced angiogenesis and osteogenesis	Higher failure load	Implanted cells differentiated into endothelial cells and osteoblasts	Tomoyuki Matsumoto 57	2012
BMP‐2 gene‐transfected bMSCs	Gastrocnemius tendons	Fibrin glue	Rabbit	More cartilage‐like cells and fibrocartilage‐like tissue	Enhanced maximum load and stiffness	The BMP‐2 mRNA and protein levels were increased.	Yu Dong 104	2012
Human adult ACL ruptured tissues‐derived CD34+ cells	Flexor digitorum longus tendons	Cell sheet	Rat	More cells derived from the cell sheets incorporated within the bone tunnel site and grafted tendon	Higher failure load	Implanted cells differentiated into endothelial cells and osteoblasts	Yutaka Mifune 58	2013
bMSCs	Aotologous semitendinosus tendons	collagen sponge or a fibrin sealant	Rabbit	only one specimen of total seven showed a chondroid cell layer at both the anterior and posterior interfaces	None	None	Tomonoshin Kanazawa 54	2014
Tendon‐derived stem cells	Autologous flexor digitorum longus tendon	Cell sheet	Rat	Higher tunnel graft integrity and more Sharpey's fibres	No significant differences	None	Pauline Po Yee Lui 110	2014

Bone marrow‐derived mesenchymal stem cells

BMSCs are non‐hematopoietic stromal cells with multi‐potential to differentiate into osteoblasts, chondrocytes and adipocytes. Based on the capacities of proliferating in culture while retaining their growth and multi‐lineage potential 71, 72, 73, 74, 75, 76, bMSCs have been widely studied for enhancing tendon–bone healing in pre‐clinical investigations 51, 52, 53, 54.

To observe the effect of bMSCs on promoting tendon–bone healing, Ouyang et al. 51 created an intra‐articular tendon–bone healing rabbit model. They implanted bMSCs embedded within fibrin glue into the bone tunnel after reconstruction surgery with halluces longus tendon. Four weeks later, the tendon–bone interface showed more perpendicular collagen fibre formation and increased proliferation of cartilage‐like cells in the bMSCs‐treated group, which resembled a fibrocartilage attachment.

Similar positive results came from Lim et al. 52 and Michael et al. 53. With the intra‐articular ACL reconstruction rabbit model, they investigated the effect of bMSCs on osteointegration of an auto‐ and allograft in the bone tunnel respectively. A mature zone of fibrocartilage blending from bone to the graft was observed in the bMSCs‐enhanced group at week 8, while the controls showed mature scar tissue resembling Sharpey's fibres spanning the tendon–bone interface. Meanwhile, the bMSCs‐enhanced grafts had significantly higher failure load than the controls.

However, in the study of Tomonoshin Kanazawa et al. 54, only one specimen of total seven showed a chondroid cell layer at both the anterior and posterior interfaces after autologous bMSCs treatment. Even though, the potential effect of bMSCs could not be ruled out for the only one evidence. The result might be affected by the special reconstruction animal models without a tibial bone tunnel or the cultivation of cells.

Although the exact biological mechanism of transplanted bMSCs on tendon–bone healing is still largely unknown, these cells seem to contribute to a more natural and biomechanical insertion. One possibility is that the implanted bMSCs proliferate and differentiate into ligament fibroblasts and fibrocartilage cells, contributing to the accumulation of the fibrocartilage tissues. Another possibility is that the fibrocartilage interface originates from cells recruited locally, and injected bMSCs may secrete a variety of growth factors to stimulate the activation and recruitment of local fibroblast and fibrocartilage cells.

Synovial mesenchymal stem cells

SMSCs are widely investigated and applied in cell‐based strategy for tissue repair and regeneration. 77, 78, 79, 80, 81. It has been reported that sMSCs have a higher proliferation and differentiation potential than bMSCs 82, 83. Recently, sMSCs have been introduced in the enhancement of tendon–bone healing and proved pro‐healing efficacy.

Young‐Jin Ju et al. 55 implanted DIL‐labelled sMSCs into tendon–bone interface and confirmed the enhancement in the production of collagen fibres and formation of Sharpey's fibres between the tendon graft and bone tunnel 2 weeks after the intervention. At week 4, the interface tissue disappeared, and the implanted tendon appeared to attach to the bone directly without fibrocartilage formation. They suggested that the implanted sMSCs could directly differentiate into fibroblasts to produce collagen fibres for strong connection between tendon and bone. However, there was not a formation of fibrocartilage insertion. In conclusion, sMSCs may effectively improve tendon–bone healing by accelerating the formation of fibrous structure but not meeting the goal to regenerate a normal ACL insertion site.

Stem cells derived from ACL tissues

It has been demonstrated that stem cells can be isolated and identified from ACL. Although it is controversial about their capacities of in vitro proliferation and multi‐differentiation, they are effective for ligament repair and regeneration 84, 85, 86, 87, 88. A recent report from Matsumoto et al. 89 revealed that there were abundant vascular stem cells with a characteristic expression of CD34 in the injured ACL tissues. These cells displayed higher expansion and multiple lineages differentiation potential when compared to CD34− cell population.

To figure out the efficacy of ACL‐derived CD34+ cells on tendon–bone healing, Mifune et al. 56 injected adult human ACL‐derived CD34+ stem cells into the nude mice articular cavity after ACL reconstruction. More fibrocartilage cells as well as enhanced angiogenesis and osteogenesis were suggested in the CD34+ and non‐sorted cells groups compared with CD34− cells and PBS groups. The immunostaining analysis suggested that the implanted ACL‐derived CD34+ stem cells could directly differentiate into endothelial cells and osteoblasts at both the tendon–bone interface zone and perigraft site. Mechanical test showed the failure load after the CD34+ cells treatment exhibited almost the same strength as the uninjured ACL. Subsequently, this team advanced the study by using ACL‐derived CD 34+ cells based on tissue engineering therapy to enhance tendon–bone healing, which also exhibited promising healing potential 58.

Compared to other commonly applied stem cells, CD34+ stem cells promoted early remodelling of tendon–bone attachment and incorporation of the graft into the bone via enhancing angiogenesis and osteogenesis. It has demonstrated that the implanted CD34+ stem cells could directly differentiate into endothelial cells and osteoblasts in the above‐mentioned studies 56, 58, which might be associated with expression of certain genes within the special stem cells 90, 91, 92. In addition, it is also uncertain that the fibrocartilage cells come from the implanted stem cells or the recruitment of local cells. The real mechanisms need to be further investigated.

Cell sheet technique

Direct intra‐articular injection or embedding within fibrin glue is most commonly used methods to transfer stem cells into bone tunnel. Cell leakage is an unavoidable problem in these ways, which might lead to sub‐optimal efficacy 51, 52, 53, 56, 93. Thus, alternative approaches which can deliver stem cells to the tendon–bone interface more efficiently are required. Recently, cell sheet technique which has been used in various fields such as treating heart diseases 94, 95, 96 and promoting cartilage regeneration 97, 98 may serve as a new strategy for stem cell transplantation in promoting tendon–bone healing.

To clarify the effect of cell sheet on promoting tendon–bone healing, Mifune et al. 58 wrapped the ACL‐derived CD34+ cell sheet on the tendon graft before it was implanted into the bone tunnel in a rat ACL reconstruction model. A greater number of transplanted cells could be identified within the bone tunnel site as well as the grafted tendon in the cell sheet group, which ultimately contributed to superior tendon–bone healing compared to the intra‐articular injection group.

With the application of cell sheet technique, delivery of stem cells was enhanced at the tendon–bone interface as well as the graft, which accelerated the maturation of the graft while promoting tendon–bone healing. The mature graft improves the stability and biological properties of the reconstructed knee further 25, 99. Thus, cell sheet technique is rather a superior strategy to delivery stem cells into the reconstructed ACL compared to direct intra‐articular injection or fibrin glue technique.

Gene therapy based on stem cells

It has been proved that several active substances such as transforming growth factor‐β(TGF‐β), platelet‐derived growth factor (PDGF), bone morphogenetic protein‐2 (BMP‐2) have been demonstrated effective in promoting tendon–bone healing by enhancing osteogenesis or angiogenesis based on animal studies 37, 38, 39, 40, 41. However, the effects of these growth factors were dose‐ and time‐dependent 100, 101, 102. Thus, to achieve continuous and stable concentration of these factors at tendon–bone interface, the gene therapy based on stem cells has been introduced.

Feng Li et al. 103 treated the reconstructed tendon–bone interface with PDGF‐B transfected bMSCs, which effectively enhanced vascularity and collagen deposition. In another study from Yu Dong et al. 104, BMP‐2 gene‐transfected bMSCs were transplanted into bone tunnels, which promoted the proliferation of cartilage‐like cells and the formation of the fibrocartilage‐like tissues. The maximum load and stiffness were also improved.

Compared to direct application of growth factors, the genetic intervention based on stem cells is a superior alternative to promote tendon–bone healing. These transfected stem cells can secret relevant growth factors steadily. Moreover, the efficacy of growth factors is enhanced with the proliferation and differentiation of stem cells. Thus, gene therapy based on stem cells exerts a strong, positive, continuous and stable effect of growth factors.

However, there are several issues to be considered before clinical application of gene therapy 105, 106. One major limitation is that the expression of the transferred gene will decrease with time, which may lead to the loss of desired product after several weeks. In addition, the safety of gene therapy is not guaranteed because mutagenesis, development of malignancy and other side effects may appear.

Limitations of stem cells therapy and directions of future researches

In spite of the promising consequences of stem cells therapy, there are still quite some limitations. The biological mechanism of implanted bMSCs on tendon–bone healing is largely unknown 51, 52, 53, 54. It seems that sMSCs just contribute to accelerating a fibrous attachment formation rather than regenerating a reliable fibrocartilage insertion, which may be difficult to achieve the ideal effect 55. And the related investigations of sMSCs on tendon–bone healing are quite insufficient, which still need further investigations to prove their regenerative potential. The currently discovered CD34+ stem cells from ACL ruptured tissues have advanced our knowledge of the stem cells therapy on promoting tendon–bone healing. However, the biological mechanism affecting the directions of differentiation needs to be further clarified 56, 58. The short‐term effect of stem cells therapy has been proven relatively effective based on animal studies. However, the long‐term effect or data based on human study are still scarce.

As for clinical applications, the major goal of stem cells therapy for tendon–bone healing is to promote incorporation of the graft into the bone and formation of a reliable fibrocartilige insertion. However, there is no consensus on dose and frequency of the stem cells treatment to achieve an optimal effect 107. To get sufficient number of autologous stem cells, a two‐staged arthroscopic surgery is unavoidable, which often leads to extra suffering and cost. A more reliable, surgeon‐friendly and low‐cost strategy is required. In addition, it is suggested that the portion of stem cells in tissues is related to the age of patients. Thus, age is a crucial factor to be considered with the stem cells therapy 108, 109. Currently, most studies of stem cells therapy are based on small animal models, in which the tendon–bone healing process occurs at a faster rate than in humans. Further studies to test the effectiveness of this application in large animal models are necessary to confirm its clinical feasibility. Besides the reparation and regeneration efficacy of stem cells, the potential risks or adverse effects such as carcinogenesis should be adequately evaluated before they were applied in clinical.

Conclusion

Stem cells therapy has been enjoying popularity and drawing increasing attention in tendon‐to‐bone healing. This strategy helps to promote the healing process and contributes to a reliable fibrocartilage tendon–bone interface. However, these results are under limited conditions without clinical data. Future studies are required to solve multiple problems before clinical applications, including application conditions, technical and safety issues. The recently reported CD34+ stem cells isolated from ACL ruptured tissues have provided a more promising direction for future studies. Meanwhile, researches focused on the biological healing behaviours of the implanted stem cells for tendon‐to bone healing are also required.

Competing interests

The authors declare that they have no competing interests.

Author contributions

Zi‐chen Hao studied related articles and drafted the manuscript. Shan‐zheng Wang, Xue‐jun Zhang and Jun Lu participated in revising the manuscript. All authors read and approved the final manuscript.