Consensus on Regenerative Medicine in Orthopedics and Sports Medicine

INTRODUCTION

Regenerative medicine (RM), or biologic therapy, is a rapidly emerging field for which there is considerable excitement and potential. There are numerous therapies that promise to modulate cellular activity of injured or surgically repaired tissue to produce a higher fidelity, more restorative form of healing than results from the current standard of care.  Physicians and patients are often confused by the lack of clarity about the exact nature of these products, the mechanism by which they enhance healing, and in which injuries or clinical scenarios a particular biologic product can demonstrate benefit.

The primary objective of any regenerative therapy is to improve functional outcomes by enhancing the overall healing paradigm to produce higher quality repair at the injured region or surgical site. In the professional, collegiate, high school, or weekend athlete, RM products are used both as an intermediate intervention for inflammatory or non-catastrophic injuries or as adjuvant therapy to surgical reconstruction.  In less invasive interventions, RM is intended to manage inflammation and/or drive enhanced healing of musculoskeletal tissues.  In surgical indications, RM is expected to accelerate healing and decrease scar formation and adhesions, leading to better outcomes.

Orthopedic surgeons, professional team physicians, and sports medicine specialists widely use these products to support healing of extra-articular connective tissue pathology and for intra-articular conditions following meniscal repair and after ligamentous or cartilaginous reconstructive procedures. Significant interest is developing in the use of biologics in the non-operative management of acute and chronic soft tissue injuries and in the modulation of degenerative osteoarthritis.

In orthopedic sports medicine today, primary RM options include: platelet-rich plasma (PRP), bone marrow aspirate concentrate (BMAC), adipose-derived mesenchymal stem cells, and human amniotic membrane (AM)-derived tissues. While all fall under the umbrella of “Regenerative Medicine,” these therapies actually take very different approaches while seeking to accomplish similar objectives.

These therapies can be characterized by the site of origin of the regenerative cell and mode of action of the cell and/or matrix molecules upon injection. PRP is a concentrate of platelet-rich plasma protein derived from the patient’s whole blood and is intended to introduce a concentrated mixture of platelets and growth factors into the healing environment to support normal cellular activity.^1-4 Mesenchymal stem cells can be recovered from bone, adipose, or muscle. Bone and adipose sources are commercially available.

Autologous BMAC reintroduces a concentrated volume of the patient’s hematopoietic and osteogenic stem cells into the natural healing response.  BMAC and adipose-derived stem cells have the potential to differentiate into several mature cell types. This differentiation process is determined by the local environment and/or local milieu interacting with mesenchymal stem cells.^5-7 Theoretically, these cells can then aid in the repopulation and restoration of damaged tissue, promoting improved repair at the histologic level.^8-11

AM-deprived tissue has proven to be a source of anti-fibrotic and anti-inflammatory cytokines. It also has unique matrix proteins, which guide restorative healing while reducing inflammation, adhesion, and scarring as observed in the fetal environment.^12-17

Besides modes of action, RM modalities have other key differences, such as their preparation and clinical consistency.  The inherent donor biology (e.g., patient age and general health), harvesting, and preparation of autologous products (PRP, BMAC, and adipose tissue) introduce a multitude of variables that can affect the final “product” in an unpredictable manner.  AM tissues, on the other hand, have a well-defined composition; These tissues are procured only following healthy, full-term live births; and are processed using standardized, reproducible methods.  They also have a lengthy and well-documented track record of clinical success, having been commercially processed since 1997 to heal ocular injuries, for example.^18-20  More recently, AM tissue has become the advanced therapy of choice for the treatment of difficult-to-heal chronic wounds.^{21 -24}

As a whole, there are key questions that we as clinicians need to consider.  What are the therapeutic advantages and disadvantages of these therapies individually?  Can their success in connective tissue applications translate into different environments, such as intra-articular spaces?  How can we best determine which patients will potentially benefit?  Are autologous options truly benefiting the healing cascade if the donor is older, sick, or diseased? 

There are no clinical guidelines from professional medical societies or randomized clinical trials published to establish a particular biologic therapy as the standard of care. As a result, we convened a panel of national experts to review the medical literature, assess their own clinical experience, and develop recommendations for the use of these technologies. The panel members’ clinical experience with RM modalities spans thousands of patients over 15 years. The following statements and related comments are what arose from those discussions.  Our consensus recommendations are intended to provide orthopedic sports medicine physicians with best practices and guidance in the effective use of these therapies.

Clearly, more robust research on the clinical effectiveness of these therapies is needed.  Furthermore, an analysis of each therapy’s cost-effectiveness will continue to be a significant factor in influencing adoption of these new technologies. It is our intent that these discussions advance the current practice of regenerative medicine and this intersection of science and medicine to fulfill this field’s considerable promise.

CONSENSUS STATEMENTS/DISCUSSION

OPTIMAL RESTORATION OF FUNCTION: A CRUCIAL OUTCOME

QUESTION: When it comes to the broad topic of RM, how important is it that patients experience healing that supports optimal restoration of function, whether the intervention is conservative or surgical?

CONSENSUS STATEMENT: Regardless of the technical aspects of the procedure, restoration of function, speed of healing, and diminished pain are the important parameters of overall outcomes and patient satisfaction. In the panel’s experience, restorative medicine should optimize the environment for healing the injury, and, despite advances in surgical techniques, there are still opportunities to specifically limit inflammatory, scarring, and adhesive complications typically seen in damaged and/or surgically manipulated structures.  An optimized healing environment should subsequently improve overall outcomes of surgical or nonsurgical procedures.

DISCUSSION: At our current level of understanding and experience, RM appears to have its greatest sports medicine applicability in soft tissue repair.  RM can be beneficial in improving the healing process by introducing key biological drivers and optimizing the available cellular resources when treating an injury.  As such, it is important to consider that RM may benefit any individuals where the optimization of end-result functionality is critical, not only those seeking a return to athletic endeavors.

Regardless of the surgeon’s ability or the success of the surgical procedure, patients often experience undesirable side effects from the surgery itself.  Repaired connective tissues can scar very rapidly after surgery and lead to poor functional outcomes that correlate more to the patient’s individual biology than the quality of the surgical repair.  This is a frustrating phenomenon for surgeons.  How might we better influence the quality of our interventions by limiting the impact of fibrosis, particularly in the area of potential function?  Of the therapies considered in this discussion, AM is the only one that has specifically been demonstrated to impede scar and adhesion formation around soft tissue by modulating the inflammatory cascade, improving overall healing and providing for a better functional outcome.^25-27

The efficiency or speed of healing may also be increased with RM, and that accelerated healing should also correlate with decreased pain.  This additional advantage provides patients with a head start on key post-operative goals, such as weight bearing, aggressive rehabilitation, and reestablishing range of motion.  Early achievement of these goals aids in limiting muscle atrophy and rehabilitation-related injuries of proximal structures, potentially leading to more complete restoration of function.

An example of the use of RM in a non-operative procedure is the treatment of plantar fasciitis (PF). Healthy fascia has a certain degree of flexibility that allows it to stretch and bear load while providing support. In PF, the fascia is stiff and develops micro tears. To treat PF, the goal is to change how the fascia is functioning as a biologic structure, which is challenging. Injecting the fascia with cortisone, for example, may precipitate a rupture and subsequent biomechanical problems. RM may allow us to maintain the structure and function of the fascia, aiding the patient’s own physiology to hopefully undo the effects of this degenerative condition. RM can up-regulate the biological healing potential, increasing blood flow and signaling resident stem cells that have been kept dormant. 

UMBILICAL CORD TISSUE MODULATES INFLAMMATORY CASCADE

QUESTION: Where does umbilical cord tissue fall within the range of options available to support optimal healing and restoration of function?

CONSENSUS STATEMENT: Injectable particulate umbilical cord is a targeted RM therapy that follows a proven pathway,^28-30  in contrast to therapies such as PRP, BMAC, and adipose cell therapy. Amniotic tissue is a proven technology in ophthalmology, where it has become the gold standard of treatment in several indications and has demonstrated 20 years of success in an area where outcomes are highly sensitive to inflammation and scarring, which can have deleterious implications to the final clinical results.^31-33 Both the longevity of clinical use and its status as the standard of care in conditions such as Stevens-Johnson Syndrome make amniotic tissues unique agents in RM.³⁴   Other modalities such as PRP, BMAC, and adipose therapies have struggled to show conclusive and widely accepted evidence regarding efficacy in any specific indication. 

DISCUSSION: Umbilical cord and amniotic tissue are transplantable and immunologically privileged human tissue products that are a more complete biological cocktail than PRP and BMAC. At the local level, umbilical cord is the only one of these therapies that modulates the inflammatory cascade in a very controlled manner, creating a direct and effective healing response.^28-30  In contrast, when we introduce PRP, BMAC, or adipose cells to a site of injury, we are assuming that the patient is deficient in specific healing nutrients or cells, and that by delivering high concentrations of these components, the patient’s innate healing mechanisms will be enhanced to a higher functioning paradigm.

As a processed and regulated tissue, umbilical cord has proven its consistency and quality.^35,36    Autologous RM products are variable from their inception, as they can be affected by the overall health and quality of donor biology.  This is further complicated by variability in the users’ techniques of harvesting, processing, and transferring the therapies. These variabilities make it difficult to see how autologous PRP, BMAC, or adipose cell therapies might achieve consistent and reproducible clinical benefits.

Inflammation is the fundamental mechanism in any healing paradigm.  Inflammation is necessary to alert the body to injury, but its escalation is directly correlated with increased pain and disorganized fibrotic tissue repair.  It is well recognized that both PRP^37,38  and BMAC incite a significant inflammatory response, which often creates significant pain, particularly when an injection is performed in an acute injury setting. This is most likely due to the high concentration of white blood cells within the final preparation. By design, these RM products strive to elevate the local inflammatory reaction following an injury. In contrast, amniotic tissue, encompassing the umbilical cord and amniotic membrane, has demonstrated inherent anti-inflammatory characteristics. Transplanted cryopreserved amniotic membrane causes an induction of neutrophil, monocyte, and macrophage apoptosis.^39-41 This tissue also reduces infiltration by neutrophils, macrophages, and lymphocytes^42,43, and it has been shown to promote macrophage polarization.⁴⁴ These intricate anti-inflammatory effects have the potential to decrease acute discomfort and speed up recovery following injection. Interestingly, AM stromal tissue down-regulates TGF-beta in human fibroblasts, directly limiting scarring.^45,46 Finally, AM tissue has shown an anti-angiogenic effect in a rat model, a characteristic that has demonstrated clinical benefit in corneal surface reconstruction.^47-50

As mentioned above, the relative “age” of the biological components of these therapies is a significant factor in viability.  Umbilical cord, however, is less variable. Fetal tissue is inherently young tissue and may show a consistently greater capacity for altering repair when compared with technologies that, in many patients, utilize more mature adult biology.

Currently available stem cell technologies in RM are predicated around harvesting cells from areas known to hold mesenchymal stem cells of connective tissue, bone, and cartilage lineages and redeploying them into affected areas of injury.  Today, bone marrow and adipose tissue are the most readily available sources for harvesting.  Contemporary harvesting systems, however, are still relatively limited in their ability to isolate viable undifferentiated stem cells from fully differentiated cellular components.  These fully differentiated cells, such as leukocytes, are known to be prevalent in concentrated forms in these therapies and tend to complicate healing, ultimately negating some of the RM promise of BMAC and adipose tissue.  This area is extremely exciting, but we are admittedly still at the infancy of this science with much of the promotion based on anecdotal or intellectual justification, not empirical evidence.

Noting the potential as well as the limitations of PRP, BMAC, and adipose cell therapies relative to the clinical understanding established for amniotic tissues, this panel supports the primary consideration of amniotic tissues when choosing a regenerative therapy. Amniotic tissues are consistent in quality, simple to use, sterile, natural, and easy to obtain. They do not require a large time commitment on the provider’s part—no harvesting or preparation, for example. These products also have an extremely well-documented safety profile.

Based on the clinical experience of this panel, current potential indications for amniotic tissues, including fluid, matrix, and umbilical cord, are:

• To reduce the risk of adhesions in soft tissue repair (based on experience in Achilles tendon, peroneal tendon, and ACL surgery)
• To prevent capsular adhesions in total ankle arthroplasty
• To support the healing of challenging surgical wounds, such as in revision cases and in areas where there is not good blood flow (anterior ankle, Achilles tendon, etc.)
• To improve the healing environment around nerve injury or nerve decompression
• Non-operative treatment of partial tendon tears
• Non-operative treatment of plantar fasciitis and tennis elbow

Indications with promising early outcomes but requiring further investigation are:

• Osteoarthritis
• Tendinopathy
• Avascular necrosis
• Subchondral lesions
• Osteochondral lesions
• Ligamentous reconstruction/repair
• Meniscal repair
• Articular cartilage regeneration
• Fracture nonunion

AMNIOTIC PRODUCTS: A 20-YEAR CLINICAL HISTORY

QUESTION: Based on medical literature, clinical experience, and your own judgment, what are the fundamental characteristics that you look for when including amniotic products in your treatment?

CONSENSUS STATEMENT: Basic science on placental tissue in its diverse forms is substantial, but additional clinical evidence is needed to differentiate the efficacy of the various commercial products. Clinical experience and judgment currently guide the choice of placental tissue. Important factors include the method of processing placental tissue to maintain the biological integrity, consistency of the product’s content and quality, the product’s overall safety record, clinical experience in sensitive anatomy, and the company’s dedication to peer-reviewed data.

DISCUSSION: There is very little, if any, medical literature investigating the efficacy of one AM product versus another. This is an inherent weakness surrounding all biologics.  However, there is robust scientific evidence clearly demonstrating that processing methods greatly affect how the collagen matrix, available growth factors, and unique effector proteins are maintained within AM tissues. Specifically, cryopreservation has been shown to maintain the inherent biology to be equivalent to fresh tissue.³⁵   In contrast, processing methods such as dehydration have been shown to alter the biological potential by denaturing and depleting available biologically active factors. ³⁶ 

Although relatively new to orthopedics, amniotic membrane has been used clinically in ophthalmology for 20 years for the treatment of severe conditions including traumatic burns, Stevens-Johnson Syndrome, and dry eye.^18-20,35  The introduction of amniotic tissues to ophthalmology has contributed to great advances in the field and has become the gold standard of care in several ophthalmic indications.  Based on the excellent outcomes observed in ophthalmology, the adoption of placental tissues into orthopedics is very intuitive.  On the other hand, PRP, BMAC, and adipose tissue have yet to demonstrate conclusive clinical effectiveness in other areas of medicine from which we could draw a translational efficacy. 

The current body of literature investigating amniotic tissue use specific to orthopedics is young but is quickly evolving and will become much more robust in the next few years as adoption and utilization continue to grow.  In the studies that have been conducted on amniotic tissue in orthopedics, however, the product has performed as expected, demonstrating safety, reduced inflammation, and improved functional outcomes in the repair of soft tissue.^51-57

This panel recognizes that cost is always a fundamental consideration in RM, but the benefits must also be considered.  It is incumbent on the industry to analyze these technologies from a healthcare economics perspective in order to enable good decision-making when determining which patients can benefit appropriately.  We fully expect that the long-term benefits may far outweigh the initial cost required for these therapies.

AN EXPECTATION OF IMPROVED HEALING

QUESTION: What do you expect of therapies that use the term “regenerative medicine”?

 

CONENSUS STATEMENT: At this time, our use of available regenerative therapies might be better described as “augmented restorative healing.” We don’t presume or require that these products regenerate damaged tissues and joints or make them normal again. We can, however, intervene in the normal adult healing process by creating a more favorable healing environment instead of leaving it up to an individual’s predetermined healing cascade.  By improving the healing environment, our goal is to achieve the best functional outcomes from surgical and nonsurgical treatment, which is the top priority for the patient and the physician.

DISCUSSION: The goal of RM is to attempt to orchestrate and improve the adult healing process to achieve a higher quality repair of tendons, ligaments, and cartilage, which, in turn, will enhance functional outcomes. By appropriately modulating the inflammatory cascade during healing (mimicking fetal healing), these tissues should be repaired in a more normal physiologic state or architecture.  Ultimately, restorative therapies should allow surgeons to extend their therapeutic influence beyond the operating room by dictating the healing process after an incision is closed.  Regardless whether the patient is a professional athlete or a 50-year-old weekend warrior, RM may offer the opportunity to heal more quickly, with less pain, and with better functional results.  The pursuit of that promise is fundamental to advancing medicine and justifies the clinical experience as data are collected.

MORE CLINICAL EVIDENCE NEEDED

QUESTION: What are the expectations for clinical evidence that are appropriate for the continued adoption of placental tissue?

CONSENSUS STATEMENT: Our conclusions are based on the current state of basic science and clinical evidence, as well as our own clinical experience. The panel acknowledges that the published clinical evidence is limited in scope and in long-term follow-up.  We challenge companies developing regenerative or restorative therapies to conduct the clinical studies that will advance this field and demonstrate value for cost. This is especially true for therapies including PRP, BMAC, and adipose tissue, where definitive clinical evidence is lacking and experience seems to vary. Data supporting indications and patient segmentation would have particular utility, as would case series from a variety of practitioners.

DISCUSSION: While the general effectiveness of cryopreserved umbilical cord has been proven in multiple clinical studies, other RM therapies either lack scientific and clinical evidence demonstrating their effectiveness, or their efficacy claims are based on poorly designed studies.  Based on the available clinical evidence and experience, this panel has concluded that placental tissues may play an important role in the continuum of treatment for the indications outlined earlier.  We anticipate that this role may further expand as ongoing and future clinical studies are completed.

Current many clinicians are using RM products that fail to show compelling evidence for their effectiveness and safety.  We as clinicians have an obligation to recommend treatments based on clinical evidence, not those based solely on marketing claims.  We have an obligation to our patients, who often spend thousands of dollars on RM treatments, to vet the products in this emerging field and guide them to modalities that offer both biological and economic value. 

THE FACULTY

Robert Anderson, MD, is a member of the Green Bay Packers’ medical team and a partner at Bellin Health Titletown Sports Medicine and Orthopedics in Green Bay, Wisconsin.  Formerly the chief of foot and ankle surgery at OrthoCarolina/Carolinas Medical Center in Charlotte, North Carolina, Anderson retains a part-time practice there.  He is also the co-chair of the NFL Musculoskeletal Committee and a consultant to numerous professional and collegiate teams.

Gregory C. Berlet, MD, is an orthopedic surgeon with subspecialty fellowship training in sports medicine and foot and ankle surgery. His practice focus is on sports injuries and ankle arthritis. He is a consulting physician for the Columbus Blue Jackets NHL franchise and former hear of food and ankle at Ohio State University. Berlet has been treating patients with autogenous biologics for over 15 years. His group treats 4,000 patients with foot and ankle problems annually, and approximately half undergo BMAC augmentation.

David N. Garras, MD, is an orthopedic specialist who specializes in foot and ankle orthopaedic surgery, with an emphasis on total ankle replacement, work and sports injuries, ligament and tendon repairs/reconstruction, minimally invasive and arthroscopic techniques, deformity correction and reconstructive surgery, and foot and ankle fracture care. He also has an acute interest in Achilles ruptures, sports injuries, and foot and ankle arthritis. Garras has used PRP, BMAC, and stem cell therapies for the past five years.

Deryk Jones, MD, is the section head of Sports Medicine and Cartilage Restoration at the Ochsner Sports Medicine Institute and a professor of orthopaedic surgery at the University of Queensland. He is also team physician for the New Orleans Pelicans and numerous colleges and high school teams in the New Orleans metropolitan area and surround parishes. He was one of the founding members of the Cartilage Repair Centers of America and has a unique practice performing biologic reconstructions in the knee, shoulder, hip, and ankle. Jones has used viscosupplementation since 1998 to treat over 2,000 patients with knee, shoulder, and hip OA. He has used PRP since 2005 to treat a large number of patients for multiple musculoskeletal issues. He has used BMAC over the last two years in the OR to augment biologic procedures and to treat knee OA/chrondromalacia. Jones has used amniotic tissue therapy over the last two years to augment biologic procedures with more than 350 injections to data.

Anthony Miniaci, MD, FRCSC, is a staff physician with Cleveland Clinic’s Center for Sports Health. His specialty interests include shoulder and knee reconstruction, and anatomic, cartilage, and prosthetic restorative procedures. He was a full professor at the University of Toronto and is now a professor at Cleveland Clinic and Case Western University. He previously served as the team physician/orthopedic consultant to the London Tigers Baseball Club, Toronto Blue Jays, Cleveland Browns, and he is a consultant to the National Hockey League Players’ Association.

Murray J. Penner, MD, FRCSC, is an orthopedic surgeon and a clinical professor in the Department of Orthopaedics at the University of British Columbia and the head of the Regional Department of Orthopaedics, Vancouver Coastal Health Authority & Providence Health Care. He is a past president of the Canadian Orthopaedic Foot and Ankle Society. He specializes in adult orthopedic foot and ankle reconstruction and is a co-designer of three total ankle replacement systems in use worldwide.

References

1. van den Dolder J et al. Tissue Eng. 2006; 12(11):3067-3073.

2. Kakudo N, et al. Plast Reconstr Surg. 2008; 122(5):1352-1360.

3. Caceres M et al. J Periodontol. 2008; 79(4):714-720.

4. Hu ZM et al. Growth Factors. 2009; 27(5):280-288.

5. Dominici M et al. International Society for Cellular Therapy position statement. Cytotherapy (2006)   8(4):315–7.

6. Wagner Wet al. Exp Hematol (2005) 33(11):1402–16.

7. Delorme B, Charbord P. Methods Mol Med (2007) 140:67–81.

8. Gobbi A, et al.  Cartilage (2011) 2(3):286–99.

9. Gobbi A, et al. Knee Surg Sports Traumatol Arthrosc (2016).

10. Giannini S, et al. Injury (2010) 41(11):1196–203.

11. Krych AJ et al. Am J Sports Med (2016) 44(1):91–8.

12. Espana EM et al. Investigative Ophthalmology and Visual Science. (2004) 45 (9):2985-2991

13. He H et al. Journal of Biological Chemistry. (2009) 284 (30):20136-20146

14. Lee SB et al. Current Eye Research. (2000) 20 (4):325-334.

15. Tseng et al. Journal of Cellular Physiology. (1999) 179:325-335

16. Nicodemo, MC et al. Acta Cirurgica Brasileira. 32(2):125-139

17. Ellington JK et al. Surgical Technology International. (2014) 25:63-67

18. Rai P et al.  American Journal of Ophthalmology. (2005) 140(6):1148-1152

19. Lee SH et al. American Journal of Ophthalmology. (1997) 123(3):303-312

20. Azuara-Blanco A et al. British Journal of Ophthalmology. (1999) 83(4):399-402

21. Swan J. Surgical Technology International. (2014) 25:73-78

22. Raphael A. Journal of Wound Care. (2016) 25 (Suppl 7):S10-7   

23. Couture M.  Wounds. July 2016

24. Caputo W. Wound Repair and Regeneration. (2016) 24:885-893

25. S. Tseng et al. American Journal of Ophthalmology (1997) 124(6): 765–774.

26. H. Sheha et al. Journal of AAPOS (2009) 13(1): 99–101.

27. S. Tseng et al. Journal of Cellular Physiology (1999) 179: 325–335.

28. He H et al. Journal of Biological Chemistry. (2009) 24;284(30):20135-46

29. Zhang S et al. Journal of Biological Chemistry. (2012) 6;287(15):12433-44

30. He H et al. Journal of Biological Chemistry. (2013) 6;288(36):25792-803

31. A. Solomon et al. Ophthalmology (2003) 110(1): 93–100.

32. W. C. Park and S. C. G. Tseng. Investigative Ophthalmology and Visual Science. (2000) 41(10): 2906–2914.

33. S. C. G. Tseng, et al. American Journal of Ophthalmology. (1997) 124 (6): 765–774.

34. Hsu M et al. Cornea. (2012) 31 (12):1394-402

35. Tan EK et al. Journal of Biomaterials and Tissue Engineering. (2014) 4:379-388

36. Cooke M et al. Journal of Wound Care. (2014) 23 (10):465-74, 76

37. Zhou Y et al. Stem Cell Research Therapies. (2015) 6(1):173

38. Boswell SG et al. American Journal of Sports Medicine. (2014) 42(1):42-9

39. Park WC, Tseng SC. Invest Ophthalmol Vis Sci. (2000) 41:2906–14.

40. Wang MX et al. J Cataract Refract Surg. (2001) 27:310– 9.

41. Shimmura S et al. Cornea. (2001) 20:408–13.

42. Bauer D et al. Exp Eye Res. (2007) 85:335–45.

43. Heiligenhaus A et al. Invest Ophthalmol Vis Sci. (2001) 42:1969–74.

44. Bauer D et al. Invest Ophthalmol Vis Sci. (2012) 53:799–810.

45. Lee SB et al. Curr Eye Res. (2000) 20:325– 34

46. Tseng SC et al. J Cell Physiol. (1999) 179:325–35.

47. Tseng SC et al. J Ophthalmol. (1998) 2:26–34.

48. Tsubota K et al. N Eng J Med. (1999) 340:1697–703.

49. Tsai RJ et al. N Eng J Med. (2000) 343:86–93.

50. Jiang A et al. Cornea. (2006) 25:S36–S40.

51. Keeley R et al. Critical Reviews in Biomedical Engineering. (2014) 42(3-4):249-270.

52. Hanselman AE et al. Foot & Ankle International. (2015) 36(2):151-8

53. DeMill SL et al. Surgical Technology International. (2014) 25:257-261

54. Warner M et al. Surgical Technology International. (2014) 25:251-255

55. Li Z et al. Neural Regeneration Research. (2013) 8(36):3441-3448

56. Riccio M et al. Microsurgery. (2014) 34(8):616-622

57. Zelen CM et al. Foot & Ankle International. (2013) 34(10):1332-9