 |
 |
 |
|
| home | CME | help | feedback | subscriptions | archive |
|
© 2003 The Journal of Bone and Joint Surgery, Inc.
Biologic Aspects of Flexor Tendon Laceration and Repair
• Flexor tendon repairs in the hand are often complicated by the formation of peritendinous adhesions that result in loss of normal tendon gliding, digital stiffness, and functional disability. • Advances in suture techniques and rehabilitation protocols have improved the outcomes of tendon repair. • Recent progress in our understanding of the biology of tissue healing may lead to the enhancement of flexor tendon healing. • Manipulation of cytokine levels, introduction of genetic material into cells, and addition of pluripotential mesenchymal stem cells at the site of the repair are potential therapeutic strategies for the modulation of scar formation after flexor tendon injury. Restoration of normal hand function following flexor tendon laceration requires reestablishment not only of the continuity of the tendon fibers, but also of the gliding mechanism between the tendon and its surrounding structures. Like many other tissues, tendons heal by deposition of scar tissue at the site of injury. While the initial formation of scar tissue between the tendon ends provides physical continuity at the site of the disruption, proliferation of the scar tissue between the tendon and adjacent tissues is undesirable, indeed harmful, because these attachments impede tendon gliding, which is of critical importance for function. This problem becomes more pronounced in zone II of the flexor tendon system (located between the origin of the flexor sheath in the palm and the insertion of the superficialis tendon on the middle phalanx), where the tendons travel through a fibro-osseous canal along the palmar aspect of the digits. Thus, unlike the repair of other musculoskeletal tissues, which can fail because of too little healing, a repair of a lacerated flexor tendon can also fail because of too much healing: adhesions result in loss of motion, contracture formation, and functional disability. The majority of the research in the area of tendon repair has focused on mechanical aspects—for example, the improvement of suture repair techniques and the enhancement of postoperative rehabilitation protocols allowing early motion. These improved surgical methods and rehabilitation schemes have no doubt led to better clinical results. Indeed, before those innovations, attempted repair was so prone to failure that experts advocated delayed reconstruction, not primary repair. Because of these improvements in suture repair techniques and postoperative rehabilitation protocols, primary repair of flexor tendon lacerations has become the standard of care. Nonetheless, postoperative scarring and adhesion formation still remain disappointingly frequent complications. Even with the best technical repairs and the optimal rehabilitation protocols, functional restoration has not been reliably achieved and results have been highly unpredictable 1 . In the last two decades, our understanding of the molecular biology of the growth and repair of soft tissues has expanded dramatically. We now know, albeit incompletely, how genes are regulated, how these genes are expressed and proteins are synthesized, and how these proteins affect macroscopic and biomechanical changes in tissues. It is probable that the next advance in the repair of flexor tendons will result from the application of this basic-science knowledge. The purpose of this Current Concepts Review is to present the fundamentals of the problem of postoperative scarring following flexor tendon repair in the hand and to review the therapeutic approaches that have been used to address it. In addition, our current understanding of the biology of tendon healing at the molecular and genetic levels will be discussed. Implicit in this presentation is a point of view; namely, that the solution to this vexing clinical problem will encompass not only the finest suture repair techniques and postoperative rehabilitation protocols but also the optimal biologic modulation of the healing response.
The treatment of lacerations of the flexor tendons in the digital sheath presents one of the most challenging problems in orthopaedic surgery. While stability is critical for a successful tendon repair, mobility of the repaired tendon is also important, as motion of the repaired tendon decreases the formation of postoperative adhesions and increases the strength of the repair. Unfortunately, factors that promote stability impede mobility and vice versa. Immobilization can ensure the integrity of the repair but also lead to scarring, stiffness, and joint contractures. Conversely, postoperative motion can prevent the formation of adhesions and contractures but also place the repair site at risk for dehiscence. To date, research in this area has been directed at achieving the optimal balance between stability and mobility and at devising methods for increasing strength and motion without compromising one for the other. There are several factors that are unique to intrasynovial tendon injuries (in zone II) that make this an especially difficult problem. To begin, most injuries that lacerate the tendon also disrupt the nutritional systems that feed the tendon and sustain the repair effort. Specifically, violation of the digital sheath leads to leakage of the synovial fluid contained within it. This loss of synovial fluid may starve the tendon repair process, as nutrients are normally provided to the tendon primarily by means of diffusion through this fluid. Moreover, even if the aggregate loss of fluid is not great, a defect in the sheath can impair tendon nutrition, as it allows the pressure of the fluid in the sheath to dissipate. This loss of pressure can deprive the tendon of nutrients, as the delivery of these nutrients is normally driven by a process of imbibition, in which synovial fluid is forced into the interstices between tendon fascicles during flexion and extension of the digit 2,3 . In addition to leakage of synovial fluid, direct injury to the vincula and segmental branches from the digital arteries that provide direct blood supply to the tendon can also interfere with the nutritional supply to the tendon proper 1 ( Fig. 1 ). In summary, injury to the surrounding tissue can lead to decreased nutrition because of a poor blood supply, decreased synovial fluid, or loss of pressure to drive the nutrients into the tendon.
The second factor that leads to difficulty with the treatment of flexor tendon injuries is that all such injuries must heal in the setting of surgical trauma. Unlike the Achilles tendon, which can be reapproximated by plantar flexion of the foot (and no operative treatment), flexor tendons in the hand are not amenable to nonoperative treatment: the two tendon ends must be surgically apposed in order for healing to occur. Yet operative intervention can interfere with the biology of healing. It has been observed that adhesions form in proportion to the degree of crushing and manipulation of the tendon tissue during surgery 4 . The effects of surgical trauma are superimposed on those of the original injury and contribute to a biologic environment that is favorable for scarring. The third difficulty is that the anatomy of the flexor tendon apparatus renders it uniquely unforgiving of scarring. To carry out their mechanical function, flexor tendons glide through fibro-osseous tunnels formed by the phalanges and an overlying system of pulleys ( Fig. 2 ). Scarring often leads to formation of adhesions between the tendon and the surrounding tunnel, which interferes with the normal gliding mechanism of the tendons. This limitation of gliding restricts tendon excursion, which is critical for digital motion. In addition, the fibro-osseous tunnels are only slightly larger than the tendons themselves. For this reason, even slight bulkiness of the tendon proper due to scarring increases friction between the tendon and the sheath, further limiting tendon excursion and compromising function.
The last factor that causes difficulty with the treatment of flexor tendon injuries is that when these tendons are immobilized, the joints over which they pass are immobilized as well. Accordingly, if tendon excursion is limited, capsular motion is inhibited. This eventually leads to joint contracture and stiffness, in addition to the limitations of motion imposed by the scarred tendon.
The ideal method of flexor tendon repair either ensures adequate nutrition or somehow proceeds without it, allows a healing response precisely at the tendon ends but not between the tendon and its surroundings, creates a repair site with minimal bulk and low friction, and places enough force across the repair to promote motion and remodeling of the repair tissue without placing it at risk for failure. The traditional approach to the clinical problem of flexor tendon lacerations has been to create a repair strong enough to withstand relatively high loads, thereby lessening the need for protection of the repair site. In turn, this makes it possible to carry out a more aggressive program of postoperative motion, with the hope that this will lead to fewer adhesions and improved strength. There is a large volume of research data regarding methods of tendon suture repair, the effects of repair type on strength and shape, and the effects of motion on the repair tissue itself. On the basis of the large body of literature in this area, certain important principles of flexor tendon repair have been established 1 : 1. Core sutures (placed through the cut ends of the tendon at the laceration site) in combination with a peripheral epitendinous suture provide the most strength and prevent gapping at the repair site. 2. Stress (motion) at the repair site increases the amount of collagen deposited at the site of the injury and aids in the organization of the collagen deposited, resulting in a stronger repair but not in faster healing. 3. The strength of the tendon repair is proportional to the number of sutures and the caliber (size) of the suture crossing the repair site. 4. Nonabsorbable braided sutures (3-0 or 4-0) are optimal for use as core sutures. 5. Dorsal placement of the core sutures provides more mechanical strength than does volar placement of the sutures. However, because of the location of the nutrient vessels, dorsal sutures can also interfere with the blood supply to the tendon proper. 6. There should be equal tension across all suture strands. 7. Because the repairs of tendons that have been treated surgically and immobilized become weakest at approximately twenty-one days, rehabilitation must proceed with caution. In addition to the large body of literature regarding the method of tendon repair, a considerable amount of attention has been given to the development of postoperative rehabilitation protocols. The goals of postoperative rehabilitation techniques include preventing the development of peritendinous adhesions and joint contractures, protecting the integrity of the tendon repair, and improving the tensile strength of the repaired tendon. A multitude of rehabilitation protocols have been developed, with variations in the type of motion instituted (passive or active), type of splints used (static or dynamic), and timing of the progression of motion. In spite of this diversity, there are certain common important principles of postoperative therapy protocols that have become accepted and can be considered standard at this point in time 1 : 1. Wrist and metacarpophalangeal joints should be maintained in flexion at rest. 2. Distal and proximal interphalangeal joints should be maintained in extension at rest. 3. Therapy should involve frequent application of motion (active or passive). Application of these core principles has substantially improved the treatment of flexor tendon lacerations. Until relatively recently, primary repair of tendon injuries was discouraged and secondary tendon grafting was recommended because the results of the latter procedure seemed to be more reliable. As Bunnell noted in 1928, "one cannot join [lacerated flexor tendons] together by suture with success." 5 Although it is now known that these injuries can be treated successfully with primary repair, the results are not always optimal. Therefore, further research in this area is needed.
Research on the biology of tendon healing has provided a basic understanding of the processes by which tendons heal after injury. This reparative response has been characterized as having three sequential phases: inflammatory, fibroblastic, and remodeling ( Fig. 3 ) 6-8 . During the inflammatory phase, inflammatory cells from the surrounding tissues migrate to the injury site. These cells phagocytize necrotic tissue and clot 8 . During the fibroblastic phase, fibroblasts proliferate about the injury site and synthesize collagen and other components of the extracellular matrix. Finally, during the remodeling phase, newly produced collagen fibers become organized longitudinally along the axis of the tendon. Fibroblasts are the main cells in fibrotic healing reactions and are responsible for collagen deposition and the formation of scar.
Two mechanisms of tendon healing have been proposed. The first is the so-called extrinsic mechanism, whereby fibroblasts and inflammatory cells from the periphery invade the healing site to promote repair of the injured tendon. The second mechanism is the so-called intrinsic mechanism, whereby fibroblasts and inflammatory cells from within the tendon and epitenon invade the healing site 3,7,9 . It is likely that the healing response observed clinically is a combination of extrinsic and intrinsic mechanisms ( Fig. 4 ) 1 .
The differences between the extrinsic and intrinsic mechanisms extend beyond the site of origin of the healing cells. The extrinsic mechanism appears to be active earlier in the healing sequence, whereas the intrinsic mechanism is often delayed 10,11 . Some studies have demonstrated that the synovial sheath reacts with a greater proliferative and inflammatory response compared with the endotenon and tendon proper 12 . Other studies have shown the synovial fibroblasts to be more reactive to cytokines and to have a greater capacity for degradation of the extracellular matrix 11 . In addition, it is thought that predominance of the extrinsic mechanism of healing leads to increased collagen content at the injury site as well as to a reduction in the level of collagen organization and material properties of the reparative tendon tissue 13,14 . For these reasons, it is believed that the predominance of the extrinsic mechanism of healing leads to scar formation and adhesions between the tendon and surrounding peritendinous structures. Modulating the healing process to enhance the intrinsic pathway (and augment end-to-end tendon healing) while suppressing the extrinsic pathway (and diminishing adhesions of the tendon to surrounding tissues) could lead to improvements in the treatment of these injuries. While investigations into the biology of tendon healing have shed some light on the mechanisms by which tendon tissue heals, there has been relatively little progress toward the biologic enhancement of the healing process after injury and repair. As a result, modulation of the process by which tendons heal remains an appealing, but not currently practicable, concept.
It is well recognized that motion prevents scar formation 1,7,15-17 . There is not a specific amount of motion that is needed, only a relative amount—that is, enough to overcome the problems of scarring and adhesion formation. If one could decrease the amount of scarring associated with a given amount of motion, less motion would be required. This approach would eliminate the need for aggressive postoperative motion protocols that place the tendon repair site at risk for dehiscence. If the ideal scar inhibitor could be found, the patient could be treated with postoperative immobilization in a cast or splint until the tendon healed, allowing the repair to heal in a mechanically protected environment without concern about adhesions. Historically, prevention of adhesion formation has been attempted on two fronts. The first method has been to place a physical and mechanical barrier between the healing tendon and the surrounding tissues. The rationale for this approach is that limiting contact between the tendon and its sheath diminishes the amount of adhesions around the repaired tendon—that is, the tendon is allowed to heal to itself but not to the sheath and surrounding tissues. The various barrier materials that have been tried include silicone 18 , polyethylene membranes 19 , alumina sheaths 20 , polytetrafluoroethylene 21 , and chondroitin sulfate-coated polyhydroxyethyl methacrylate membranes 22 , among many others. In spite of the many different materials studied for this purpose, none are in routine clinical use at this time. In a similar approach, many authors have attempted to use chemical modulation to diminish the amount of scar formation after repair. The chemical agents that have been used in these efforts include local 23,24 and parenteral 25,26 corticosteroids, dimethyl sulfoxide 27 , beta-aminoproprionitrile 28 , hyaluronic acid 29,30 , and 5-fluorouracil 31,32 , among many others. The common principle of these methodologies is reduction of inflammation. In the case of the corticosteroids and hyaluronic acid, the goal is to diminish inflammation by inhibiting lymphocyte migration, proliferation, and chemotaxis as well as macrophage motility. Similarly, 5-fluorouracil, an antimetabolite, suppresses scar formation by inhibiting contraction of collagen lattice and proliferation of inflammatory cells 31,32 . Researchers using barrier and chemical techniques have reported some degree of adhesion reduction in laboratory and clinical studies of tendon repair 18,20,21,23,25,29,31,32 . In spite of these findings, these methods are not widely used in clinical practice. This suggests that none of them have been demonstrated to be effective in most clinical settings.
Despite the lack of reliable strategies to modulate the tendon repair process, there is reason to believe that a biologic approach may be feasible. Experimental evidence has shown that fetal tissue in the early to mid-gestational stage responds to injury in a fundamentally different manner than does adult tissue 33,34 . In general, fetal wound healing occurs at a faster rate and in the absence of scar formation. It is well known that fetal skin wound healing is characterized by scarless repair with restoration of normal dermal architecture and a markedly diminished and delayed inflammatory response 35,36 . In addition to skin, fetal articular cartilage, nerve, and bone have demonstrated the ability to heal without scarring in experimental models 37-41 . Furthermore, preliminary data obtained in a study of a fetal sheep tendon injury model in our laboratory has provided support for the idea that fetal tendon tissue undergoes scarless, regenerative healing ( Figs. 5 , 6 , 7-A, and 7-B ) 42 . These observations are highly preliminary and require further validation.
While the basis for the ability of fetal tissues to heal without scarring remains unknown, it is believed that the regenerative response of these tissues is a property of the tissues and not of the fetal environment 43 . It is also widely believed that the differential expression of growth factors and cytokines in fetal tissue in the early gestational period compared with that in the late gestational period and in adult tissue plays an important role in this distinct healing response 33,35,44 . These findings suggest that biologic modulation of tendon tissue repair potentially could lead to a regenerative healing process in an adult.
The cellular events of tendon healing are relatively well understood. Comprehending the healing process at the molecular and genetic levels is the next frontier in the investigation of flexor tendon healing following injury. A mechanistic insight into the process of healing and scar formation could lead to biologic modulation of the healing response to decrease the development of scar formation. While there is some understanding of these processes, much remains to be learned.
Molecular Mechanisms of Repair The role of cytokines in tendon healing has been investigated recently, and several cytokines have been implicated in the healing processes of flexor tendons after wounding. Cytokines such as transforming growth factor-beta (TGF-ß) 45-51 , insulin-like growth factor (IGF) 52-55 , platelet-derived growth factor (PDGF) 56-58 , basic fibroblast growth factor (b-FGF) 12,56,59-61 , and epidermal growth factor (EGF) 56,62 have been found to play a potential role in the healing process of tendons following injury. TGF-ß is a cytokine that has received much attention in this area 34,45,46,48,50,51 . TGF-ß is directly responsible for increased fibroblast and macrophage recruitment and proliferation, angiogenesis, upregulation of metalloproteinase inhibitor activity, downregulation of proteinase activity, and synthesis of collagen following injury; thus, it is associated with scar formation following injury or surgery 45 . In addition, it has been implicated in the pathogenesis of excessive scar formation 46 . There are three isoforms of TGF-ß: ß1, ß2, and ß3 34,51 . TGF-ß1 is the most prevalent isoform found in adults and is thought to be primarily responsible for its proinflammatory and scarring activities 34 . Addition of exogenous TGF-ß1 is also known to induce angiogenesis and deposition of collagen and granulation tissue 45 . Intrinsic tenocytes and extrinsic tendon sheath fibroblasts are known to produce TGF-ß1, and this production is activated in a tendon wound environment 45 . Furthermore, it has been shown that TGF-ß1 receptor synthesis is upregulated during tendon injury 50 . It has been suggested that perioperative modulation of TGF-ß1 levels (for example, with neutralizing antibodies to TGF-ß1) may help to limit flexor tendon adhesion following repair 46 . Interestingly, TGF-ß3 has been found to have anti-scarring properties in a cutaneous rat wound model, and it has been postulated to act as a negative regulator of scarring in an injury environment 51 . Another cytokine that has recently generated some interest in this area of investigation is b-FGF 56,59-61 . This cytokine is a 146-amino-acid polypeptide grouped in the heparin-binding growth factor family. The principal effects of b-FGF include induction of angiogenesis and fibroblast chemotaxis and proliferation. It has been found to be present in normal canine and rabbit synovial flexor tendons 56,61 . In addition, b-FGF mRNA has been found to be upregulated in tenocytes and tendon sheath fibroblasts following synovial tendon wounding in a rabbit model 61 . These findings suggest that b-FGF may play an important role in the repair of synovial tendons. IGF has also raised interest recently 52-55,62,63 . Two isomers, IGF-I and IGF-II, have been shown to stimulate proteoglycan, collagen, noncollagen protein, and DNA synthesis in a dose-dependent manner when applied to rabbit flexor tendon specimens in culture 52 . Other cytokines thought to be implicated in the healing process are listed in Table I .
While our understanding of the roles of various cytokines in tendon healing is in its infancy, their use to modulate the healing process is an appealing concept. It follows that the exogenous addition of a cytokine that limits excessive scar formation (such as TGF-ß3) or the suppression of a cytokine that promotes excessive scar formation (such as TGF-ß1) at the time of tendon repair could alter the healing process and result in a more favorable healing response. While the manipulation of cytokine concentrations at the site of a repair is an attractive notion, appropriate implementation of this concept may be complex. First, changing the concentration of one cytokine may affect other cytokine systems. Second, it is likely that these cytokines, like most other biologic systems, have a built-in level of redundancy; it seems unlikely that suppression of only one proinflammatory protein will decrease the degree of scarring at one particular site. Third, it is likely that the effect of these cytokines depends on specific concentrations of these proteins at the site of injury. Fourth, the role of each cytokine probably depends on timing and is effective only within a narrow window in the process. (For example, it is probable that the angiogenic cytokines [b-FGF] exert their effects early in the healing process, whereas the cytokines involved within extracellular matrix production [IGF and TGF-ß1] are more important later.) In spite of these limitations, it is possible that increased knowledge of cytokine biology may allow us to modulate tendon healing in the future.
Gene Therapy The introduction of these genetic materials into cells can be achieved with use of viral or nonviral techniques. Viruses used for this purpose include retrovirus, adenovirus, herpes simplex, and adeno-associated virus ( Fig. 8 ). Nonviral techniques include liposomes, "gene guns" (DNA-coated microprojectiles shot into cells), and injection of naked DNA plasmid into cells 65 . The advantages of viral vectors for genetic transfer include easier amplification of the genetic material and a greater efficiency of gene transfer than are possible with nonviral methods. The disadvantages include cellular toxicity, immunogenicity and the potential for genetic mutation of the host DNA increasing the risk of tumor formation 65 .
Several researchers have reported successful transfer of genetic material into tendons with use of viral vectors 57,66-69 . Nakamura et al. were able to transfect cells from an injured patellar tendon with the PDGF-B gene, using liposomes in a rat model 57 . The authors found enhanced expression of PDGF-B up to four weeks after transfection, leading to an initial promotion of angiogenesis and subsequent enhancement of collagen deposition in the wound. In an effort to diminish adhesion formation after injury in chicken tendons, Lou et al. used an adenoviral vector to induce overproduction of pp125 focal adhesion kinase (FAK), a protein that may play a role in the formation of adhesions after tendon injury and repair 67 . Those authors found that the group of animals that overexpressed adenovirus-mediated FAK protein showed substantial increases in adhesion formation, as detected histologically and as reflected by increased work of flexion. (Work of flexion is proportional to friction between the tendon and the tendon sheath, which is increased when peritendinous adhesions are present.) This finding suggests that suppression of pp125 FAK synthesis might decrease adhesion formation after injury and repair. In a similar study, researchers found that adenovirus-mediated in vitro BMP-12 gene transfer into chicken tendon cells increased type-I collagen synthesis 69 . In addition, the same gene transfer into a tendon laceration chicken model resulted in a twofold increase in tensile strength and stiffness of repaired tendons. Goomer et al. demonstrated high-efficiency in vivo gene transfer using nonviral techniques 66 . With a novel method of gene transfection using liposomes, this group introduced a marker gene into a canine model of intrasynovial flexor tendon injury during surgical repair. The authors found that the transfection efficiency rates (a measure of the efficacy of a process by which genetic material is transferred into cells) approached 100% in the superficial tissue layers. Moreover, transfected cells were found several layers below the exposed tissue surfaces six days following the procedure. The endothelial cells of blood vessels near the site of injury also exhibited a high level of gene expression. These findings indicate that nonviral in vivo gene transfer is an effective method of transfer of genetic material. Another method of transfer of genetic material that has potential applications in the biologic improvement of flexor tendon surgery is the transfer of antisense oligodeoxynucleotides (ODNs). Antisense ODNs are sequences of genetic material that are complementary to messenger RNA (mRNA). When transferred into a cell, the antisense ODN hybridizes to mRNA, preventing the translation of the message and effectively arresting synthesis of a specific protein. Nakamura et al. used decorin antisense gene therapy to improve the mechanical properties in a rabbit ligament injury model 70 . Decorin is a proteoglycan that is known to inhibit collagen fibrillogenesis after injury 71 . The authors found that in vivo treatment of injured ligaments with decorin antisense oligodeoxynucleotides led to improvement in the biomechanical properties of the healing ligaments. While this approach has not been used in tendon tissue, the potential applications of this technology in the treatment of flexor tendon injuries are appealing.
Cell Therapies To deliver the mesenchymal stem cells to the site of injury, investigators have placed the cultured cells in a degradable delivery system (typically a scaffold) onto which the cells can adhere. The delivery matrix and attached cells are then placed at the site of injury, where the cells can carry out their synthetic functions while adhering to a structurally competent matrix. This tissue engineering approach to improve the quality of the healing process has been attempted in an Achilles tendon model 75 ; the investigators found that implantation of a mesenchymal stem cell-collagen gel composite substantially improved the material properties of the injured tendon. To date, this approach has not been attempted to improve the healing properties of flexor tendons, but it remains an interesting avenue of investigation.
The ability to restore functional ability following treatment of flexor tendon injuries in the hand has been greatly improved by the development of better suture techniques and rehabilitation protocols. In spite of these improvements, however, the outcome of surgical repair is often less than desirable. Developments in our understanding of soft-tissue healing at the cellular, molecular, and genetic levels are likely to enable treating surgeons to modulate the healing response to improve the strength of the repair while at the same time diminishing the degree of adhesion formation following injury and operative repair. Note: The author acknowledges Joseph Bernstein, MD, for his invaluable help in the preparation of this manuscript.
  CiteULike  Connotea  Del.icio.us  Technorati What's this?
This article has been cited by other articles:
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Top
Introduction
The Clinical Problem 
| home | CME | help | feedback | subscriptions | archive | site index |
|
Stanford University Libraries' HighWire Press (TM) assists in the publication of JBJS Online. |