Despite the refinement of surgical technique and postoperative rehabilitation, the results of tendon surgery are disappointing for a substantial proportion of patients.
Clinical reviews of the results of isolated digital flexor tendon repairs indicate at least 20% of patients achieve only a fair result,2,6,7,35,37,39 with compromise of active joint motion and grip and pinch strength.13 The treatment for established tendon adhesions is tenolysis, although as an additional traumatic insult to the tendon and surrounding soft tissues, adhesions can also complicate this procedure. Prevention, rather than treatment, of adhesions after primary tendon surgery or tenolysis is preferable. The challenge of this clinical problem is to reconcile the competing needs of limiting scarring between the tendon and surrounding soft tissues and skin while allowing scarring in the tendon, overlying soft tissue, and skin to achieve wound healing.
Numerous strategies have been used experimentally to reduce formation of the collagenous bands of scar tissue which form adhesions. These approaches can be grouped into those limiting the wound-healing response to limit adhesion formation5,11,20,24,30 or those interposing a physical or chemical barrier between the tendon and the surrounding soft tissues.1,17,21,22,28,29,31,32,34 Although the experimental results of these approaches have shown success, few have been accepted in clinical practice. Inhibiting the wound-healing response by systemic treatments is nonspecific and does not resolve the dilemmas of allowing the tendon and overlying skin to heal independently and inhibiting the intervening adhesions. More recently, topical administration has been used to improve localization of various therapeutic agents, which is promising. The barrier techniques separate the tendon and adjacent soft tissues but can potentially provoke an important inflammatory response and interfere with the physiologic environment for wound and tendon healing. However, other than a few reports,29,40 clinical trials with inhibitors of tendon adhesions have not shown substantial benefit after tendon injury.15,16
Collagen prolyl 4-hydroxylase inhibitors (CPHIs) reduce the formation of extracellular collagen in vitro and in vivo.3,8-10,27 Inhibiting proline hydroxylation of the collagen peptide residues reduces the collagen triple helix structure stability and limits its secretion to the extra-cellular matrix.36
We questioned whether CPHI-I administered orally would have adequate bioavailability and favorable pharmacokinetics to allow for a therapeutic effect. We also questioned whether CPHI-I, as a selective inhibitor of collagen production, would inhibit the formation of adhesions after tendon injury. Finally, we wondered whether cutaneous wound healing is affected by CPHI-I-mediated collagen synthesis inhibition.
MATERIALS AND METHODS
The pharmacokinetics of CPHI-I were established after one dose of 50mg/kg body weight CPHI-I suspended in 1% carboxymethyl-cellulose (CMC) administered by oral gavage. Blood samples were taken before and 1, 2, 4, and 6 hours after administration by tail venipuncture. The samples were centrifuged to obtain plasma samples, which then were stored at −20°C. The CPHI-I is hydrolyzed rapidly and completely to an active metabolite in plasma. The CPHI-I metabolite was measured in the samples with high pressure liquid chromatography, and results were quantified by comparison to a standard curve generated with standard concentrations of CPHI-I metabolite.
The tendon adhesion, wound healing, and dermal collagen synthesis experiments were performed in 300 g (± 50 g) male Sprague-Dawley rats. The tendon adhesion model was restricted to an in-continuity tendon injury to exclude the potentially confounding effect of CPHI-I on tendon repair. The biomechanical parameter of the work of flexion was chosen to assess the effect magnitude of the adhesions formed. The tendon adhesion model has been modified from the rat model described by Bora et al.4 The effect of CPHI-I on cutaneous wound healing was assessed in three experiments: measurement of the rate of healing of an excised wound on the flank, ranking of collagen density in the healed excised wound, and a quantitative analysis of dermal collagen deposits in a subcutaneous tissue sponge buried in the contralateral flank. The effect of CPHI-I was assessed at an early time with animals sacrificed 2 weeks after injury, and a later time 6 weeks after injury. The durability of the CPHI-I effect was examined by varying the duration of treatment in these groups (Fig 1). The treatment group animals received 50 mg/kg body weight of CPHI-I in a 1 mL suspension of 1% CMC twice daily by oral gavage for the treatment period. The animals in the control groups received 1 mL of 1% CMC twice daily. Treatment started within 24 hours of surgery. Animal weights were measured preoperatively and then weekly during the treatment period. Seven animals were in each group based on a power calculation with a level of significance of p < 0.05 to detect a difference of 50%. The expected standard deviation of results was based on data from a pilot study of CPHI-I in a similar model (data not shown).
The experimental procedure was performed with the animals under general anesthesia, using tourniquet control, and using magnification. The flexor digitorum longus (FDL) tendon was exposed in the left hindfoot proximal to the flexor sheath and distally in the third digit over the proximal phalanx via an incision in the flexor sheath between pulleys. A 7.5-mm length of the intrasynovial portion of the FDL tendon was crushed with a standard artery forcep using a force measured to be 50 N, for 15 seconds. The FDL tendon was immobilized in the flexor sheath with a continuous running 5-0 polypropylene suture in a standard pattern along the digit from distal to proximal (Fig 2). The skin wound in the foot was closed with a Nylon suture. The back of the animal then was shaved and a 10-mm by 10-mm full-thickness skin wound, preserving the deep fascia, was created on the left flank to allow for serial measurement of wound closure and histologic analysis of dermal collagen synthesis. Dermal collagen synthesis was quantified by measurement of hydroxyproline deposits in a sponge buried in a subcutaneous pocket. The subcutaneous pocket was developed via an incision on the right flank remote from the excised wound described above. A 10 × 10 × 2-mm piece of polyethylene sponge (Graele Scientific, Melbourne, Australia) was placed in this pocket and the wound was closed. Animals were allowed to recover and mobilize without restriction. Postoperatively, the animals were monitored for gait abnormalities and wound-healing problems. The CPHI-I or CMC suspensions were gavaged according to protocol and the dimensions of the excised wound on the left flank were measured on alternate days by tracing the wound margins onto transparent acetate paper.
At the harvest procedure, the previously injured digit and the contralateral uninjured third digit were prepared for testing. The distal phalanx was transfixed with a suture used to form a loop. The FDL tendon to the third toe was isolated proximal to the flexor sheath through an incision limited to the proximal sole of the foot and secured with another transfixion suture. After the flexor tendon was stabilized, the Prolene suture inserted at the initial procedure was removed through the plantar incision without interfering with the digit. The skin of the second and fourth webs was incised to the level of the metatarsophalangeal joint to allow isolated flexion of the third digit. The foot was amputated through the ankle and mounted on a jig in a standard fashion. The scars from the excised wound on the left flank were excised en bloc and fixed in 10%-buffered formalin. The subcutaneous sponge in the right flank was removed and dissected free of its associated capsule. The sponges were stored at −20°C until all groups were completed.
The flexor tendon adhesion was assessed by measuring the force and excursion required to flex the digit with a mechanical testing machine (Instron, Melbourne, Australia). Each test was observed by two blinded observers (DMc, MK). Each foot was mounted on the Instron machine with the tip of the digit orientated inferiorly. A 10-g weight was attached to the digit tip to extend the digit to a standard neutral position. The proximal end of the FDL tendon was attached to the crosshead clamp of the machine, which then was elevated at 20 mm/minute, flexing the digit until the tip of the digit reached the plantar surface of the foot. At this point the test was terminated. The test was performed once for each digit. The force and displacement required to achieve this flexed position were recorded on a calibrated constant-speed chart recorder. The maximum force and the excursion of the clamp required to flex the digit were measured directly, and the work of flexion was calculated by curve integration. The work of flexion related to adhesions formed as a result of injury was the difference between the work of flexion in the injured digit and the uninjured digit in each animal.
The serial measurements of the excised wound on the animals' flanks were analyzed with image analysis software (Videopro, Fortitude Valley, Queensland, Australia) to calculate wound area. The biopsy specimens of the healed scars from these wounds were sectioned through the middle of the scar and stained in a standard manner with Masson's trichrome. The sections were ranked in order from most to least collagen in the scar by two blinded observers (EWT, EW). The data were separated into two groups, the 2-week and 6-week harvest groups. The ranks were recalculated for each group before analysis.
Dermal collagen synthesis was quantified by measuring hydroxyproline deposits in the subcutaneous sponge. The hydroxyproline was extracted from the sponge by hydrolysis in 2 mL of 6 mol/L hydrochloric acid at 110°C for 16 hours. An aliquot of the hydrolysate was buffered to neutral pH with 6 mol/L NaOH. This solution was diluted to 2 mL in the assay buffer (a one in 10 dilution of the stock buffer). One mL chloramine-T reagent was added, and the solution stood at room temperature for 20 minutes. One milliliter diamino-benzaldehyde reagent was added, and the solution was mixed, incubated at 60°C for 15 minutes, and cooled in tap water for 5 minutes. The absorbance of the solution at a wavelength of 540 nm was measured in a spectrophotometer. The absorbance was measured against a calibration curve to derive the hydroxyproline results.
The initial analysis of the biomechanical testing results was performed using a two-way ANOVA, which analyzed the effect of the injury, the effect of the treatment, and the interaction between the treatment and the injury. A one-way ANOVA was performed on the value of work attributed to adhesion with a post hoc Newman-Keuls test for multiple comparisons between the various treatment and control groups. The cutaneous wound healing data were analyzed according to the individual experiment. The serial wound healing measurements were analyzed with linear regression (GraphPad Prism v3.0a, San Diego, CA) and the rates of wound healing were compared. The histologic ranking of collagen density in the excised wound scars were compared using Monte Carlo method Kruskal-Wallis Test (StatXact v4.01, Saugus, MA). The hydroxyproline deposition in the subcutaneous sponges results were analyzed with a one-way ANOVA with a post hoc Newman-Keuls test for multiple comparisons.
After oral administration, CPHI-I was absorbed rapidly and reached substantial plasma concentrations. Maximum plasma levels of 6.5 to 16 mg/L were measured between 45 and 60 minutes after oral dosing with 50 mg/kg CPHI-I (Fig 3). The bioavailability, as calculated by comparison with the previously established pharmacokinetics of intravenous CPHI-I (results not shown), was estimated at 10% to 20% of the oral dose.
The remaining experimental procedures and CPHI-I treatment were well tolerated. No substantial gait abnormalities or wound complications were observed. One animal in the CPHI-I treatment group died prematurely. The cause was not ascertained at postmortem examination. Each treatment group showed failure to gain weight during the first week compared with the control groups. Thereafter, the treated animals gained weight at a similar rate as the control animals. One animal in the 2-week control group was excluded because of a testing procedure failure and one of the 6-week control group animals was excluded because no immobilizing suture was found in the tendon at exploration. The tendon adhesion model generated a substantial adhesion. The work of flexion in the injured digit was greater (p < 0.0001) than in the uninjured digit (0.0307 J versus 0.0079 J, respectively) in the 2-week control animals. At 6 weeks, the injured digit work of flexion also was greater (p < 0.0001) than in the uninjured digit (0.0331 J versus 0.0049 J). The mean additional work of flexion because of the adhesion was similar in the 2-week and the 6-week control groups.
Treatment with CPHI-I led to a reduction in the work of flexion because of the adhesion in the treatment groups compared with the control groups (Fig 4). The work of flexion because of adhesion was less (p < 0.05) in both groups treated with CPHI-I assessed at 2 weeks (0.0135 J for the 1-week treated group and 0.0112 J for the 2-week treated group versus 0.0228 J for the 2-week control group). The effect of treatment for 1 week was similar to treatment for 2 weeks in these groups assessed at 2 weeks. In the animals assessed at 6 weeks, the work of flexion because of adhesion was less (p < 0.05) in the 2-week and 6-week treatment groups compared with the control group (0.0117 J and 0.0161 J versus 0.0319 J respectively). The work of flexion because of adhesion was similar in the animals treated for 2 weeks and 6 weeks.
The cutaneous wounds healed without delay or complications in the treated animals, but dermal collagen synthesis was reduced in the treated animals. The incised wounds of the foot and back healed without complications. The serial area measurements of the excised wounds largely followed a linear relationship in the early phase of wound healing. Linear regression analysis performed on the excised wound area data from Days 1 to 7 showed the rates of wound healing of the treated groups were similar to rates of the control groups. All excised and incised wounds were healed by Day 14. The ranking of the collagen density in the histologic sections of the scars of the excised wounds were less (p < 0.05) in the animals treated for 1 week compared with the 2-week control group. The 2-week and 6-week treatment groups showed trends toward lower collagen density rankings than their respective control groups. The dermal collagen synthesis measured by hydroxyproline estimation was less (p < 0.01 and p < 0.05, respectively) in the 2-week and 6-week treatment groups than synthesis in the control groups (Fig 5).
Preventing tendon adhesion after surgery or injury requires manipulation of the wound-healing process by physical or pharmacologic means. Collagen is the predominant structural component of adhesions and therefore a specific target for adhesion prevention.12,18 To initially study the effect of collagen synthesis inhibition in adhesion prevention alone, without the effect of inhibiting the tendon laceration repair process, a crush injury tendon model was used instead of the more popular models of tendon repair adhesion formation. This experiment provides support for the use of CPHI-I as an adjunct to tenolysis. The effect of collagen synthesis inhibition on skin wound healing, however, remains relevant even with tenolysis and was studied in conjunction with the adhesion-prevention experiments. There were no wound complications, although there was evidence that dermal collagen production was affected.
We acknowledge the limitations of this study. The use of a tendon crush model rather than tenorrhaphy limits application of the findings because collagen synthesis inhibition is likely to compromise the strength of tendon repair. Additional study in a tendon repair model is required to answer this question. Treatment with CPHI-I was required only for the first week after injury to achieve a substantial benefit in reducing the effect of adhesions. The experimental data would be strengthened with longer-term followup of animals having this abbreviated treatment and with studies of shorter treatment periods to establish a minimum. The cutaneous wound-healing experiments used in our study are less stringent than tensile wound strength testing in an incised wound model. The discrepancy in the dermal collagen production assay results, where all CPHI-I treatment groups showed a substantial reduction in collagen deposits in the sponge experiment compared with the ranking of collagen density in the excised wound scar, and where the reduction in collagen density was seen only in the animals treated for the first week after injury, is likely to represent the relative insensitivity of the histologic ranking, rather than a true difference in these parameters. The remainder of the treated animal rankings showed a trend toward less collagen consistent with the dermal collagen production sponge assay.
The production inhibition of collagen has been one of numerous strategies used to prevent tendon adhesions. Specifically, cis-hydroxyproline,4,25 β-aminopropionitrile,4,19,26,33,38 αα dipyridyl,4 and D-Penacillamine4 have been shown in experimental models to inhibit the effect of tendon adhesions through their action on the collagen biosynthetic pathway. These agents, however, have been limited in their application because of their toxicity,4 their effect on other noncollagenous proteins (eg, elastin), and their effect on wound healing. The agent used in this study, CPHI-I, is a competitive antagonist of the co-substrate of prolyl 4-hydroxylase, 2-oxoglutarate. It is a member of a class of compounds shown to effectively inhibit collagen production in numerous tissues and to be well tolerated in in vivo animal studies.10 Although the compounds show activity against hypoxia-inducible factors (HIF)-prolyl hydroxylases, they are highly selective for collagen prolyl 4-hydroxylase. At the concentrations used, the compounds effectively inhibit collagen hydroxylation, but have only a minimal effect on the HIF system.
Treatment with CPHI-I reduced the work generated by the adhesion by approximately 50% in all treatment groups. This order of benefit is similar to those in other experimental series where work of flexion was used to measure the effect of peritendinous adhesions, albeit in different animal models with different adhesion-prevention strategies.23,30 Prolonged CPHI-I treatment did not produce a substantial benefit in reducing the work of flexion because of adhesions compared with a shorter early period of treatment. This finding is consistent with the biologic features of tendon injury. Collagen production begins within 3 days of injury in the epitenon and sheath regions,12,14 which are responsible for adhesions of the intrasynovial flexor tendon. After 2 weeks, collagen is produced predominantly at the repair site and the endotenon.12 This phase of collagen synthesis may be allowed to continue if treatment is restricted to the first week to 10 days. Although the biomechanical effect of the inhibitor is durable, it is likely the fibroblast populations of the flexor tendon system and other connective tissues are capable of resuming collagen synthesis, as Franklin et al10 observed recovery of proline hydroxylation and procollagen secretion within 18 hours of terminating treatment with CPHI-I in in vitro and in vivo models.
The use of a specific collagen synthesis inhibitor to limit the formation of collagenous adhesions after tendon trauma is effective. The treatment can be targeted to the critical period of adhesion formation and, by virtue of its limited biologic half-life, collagen synthesis can resume completing the normal wound healing process in the tendon. The effect of the orally administered treatment however, is not restricted to the site of tendon-adhesion formation, which has implications for cutaneous healing and the organism as a whole.
We thank A. Penington for assistance with the statistical analysis of the data, R. Romeo-Meeuw for assistance with the immunohistochemistry, and Dr. Elizabeth Williams for her role as a blinded observer in the assay of dermal collagen production.
1. Akasaka T, Nishida J, Araki S, Shimamura T, Amadio P, An K. Hyaluronic acid diminishes resistance to excursion after flexor tendon repair: an in vitro biomechanical study. J Biomech
. 2005;38: 503-507.
2. Baktir A, Turk CY, Kabak S, Sahin V, Kardas Y. Flexor tendon repair in zone 2 followed by early active mobilization. J Hand Surg Br
3. Bickel M, Baringhaus KH, Gerl M, Gunzler V, Kanta J, Schmidts L, Stapf M, Tschank G, Weidmann K, Werner U. Selective inhibition of hepatic collagen accumulation in experimental liver fibrosis in rats by a new prolyl-4-hydroxylase inhibitor. Hepatology
. 1998;28: 404-411.
4. Bora FW Jr, Lane JM, Prockop DJ. Inhibitors of collagen biosyn-thesis as a means of controlling scar formation in tendon injury. J Bone Joint Surg Am
5. Chang J, Thunder R, Most D, Longaker M, Lineaweaver W. Studies in flexor tendon wound healing: neutralizing antibody to TGF-b1 increases postoperative range of motion. Plast Reconstr Surg
6. Cullen KW, Tolhurst P, Lang D, Page RE. Flexor tendon repair in zone 2 followed by controlled active mobilisation. J Hand Surg Br
7. Elliot D, Moiemen NS, Flemming AF, Harris SB, Foster AJ. The rupture rate of acute flexor tendon repairs mobilized by the controlled active motion regimen. J Hand Surg Br
8. Franklin TJ, Hales NJ, Johnstone D, Morris WB, Cunliffe CJ, Millest AJ, Hill GB. Approaches to the design of anti-fibrotic drugs. Biochem Soc Trans
9. Franklin TJ, Hitchen M. Inhibition of collagen hydroxylation by 2,7,8-trihydroxyanthraquinone in embryonic-chick tendon cells. Biochem J
10. Franklin TJ, Morris WP, Edwards PN. Large Ms, Stephenson R. Inhibition of prolyl 4-hydroxylase in vitro and in vivo by members of a novel series of phenanthrolinones. Biochem J
. 2001;353: 333-338.
11. Frykman E, Jacobsson S, Widenfalk B. Fibrin sealant in prevention of flexor tendon adhesions: an experimental study in the rabbit. J Hand Surg Am
12. Garner WL, McDonald JA, Koo M, Kuhn C3rd, Weeks PM. Identification of the collagen-producing cells in healing flexor tendons. Plast Reconstr Surg
13. Gault DT. Reduction of grip strength, finger flexion pressure, finger pinch pressure and key pinch following flexor tendon repair. J Hand Surg Br
14. Gelberman RH, Amiel D, Harwood F. Genetic expression for type I procollagen in the early stages of flexor tendon healing. J Hand Surg Am
15. Golash A, Kay A, Warner J, Peck F, Watson J, Lees V. Efficacy of ADCON-T/N after primary flexor tendon in Zone II: a controlled clinical trial. J Hand Surg Br
16. Hagberg L. Exogenous hyaluronate as an adjunct in the prevention of adhesions after flexor tendon surgery: a controlled clinical trial. J Hand Surg Am
17. Hagberg L, Gerdin B. Sodium hyaluronate as an adjunct in adhesion prevention after flexor tendon surgery in rabbits. J Hand Surg Am
18. Hagberg L, Wik O, Gerdin B. Determination of biomechanical characteristics of restrictive adhesions and of functional impairment after flexor tendon surgery: a methodological study of rabbits. J Biomech
19. Herzog M, Lindsay W, McCain W. Effect of beta-aminopropionitrile of adhesions following digital flexor tendon repair in chickens. Surg Forum
20. Kapetanos G. The effect of a local corticosteroid on the healing and biomechanical properties of the partially injured tendon. Clin Orthop Relat Res
21. Karakurum G, Buyukbebeci O, Kalender M, Gulec A. Seprafilm interposition for preventing adhesion formation after tenolysis: an experimental study on the chicken flexor tendons. J Surg Res
22. Kobayashi M, Toguchida J, Oka M. Development of polyvinyl alcohol-hydrogel (PVA-H) shields with a high water content for tendon injury repair. J Hand Surg Br
23. Komurcu M, Akkus O, Basbozkurt M, Gur E, Akkas N. Reduction of restrictive adhesions by local aprotonin application and primary sheath repair in surgically traumatized flexor tendons of the rabbit. J Hand Surg Am
24. Kulick MI, Smith S, Hadler K. Oral ibuprofen: evaluation of its effect on peritendinous adhesions and the breaking strength of a tenorraphy. J Hand Surg Am
25. Lane JM, Bora FW Jr, Black J. cis-Hydroxyproline limits work necessary toflex a digit after tendon injury. Clin Orthop Relat Res
26. Madden JW, Peacock EE Jr. Effect of controlled lathyrism following repair of flexor tendons in man. Surg Forum
27. McCaffrey TA, Pomerantz KB, Sanborn TA, Spokojny AM, Du B, Park MH, Folk JE, Lamberg A, Kivirikko KI, Falcone DJ, Mehta SB, Hanauske-Abel HM. Specific inhibition of eIF-5A and collagen hydroxylation by a single agent. J Clin Invest
28. Menderes A, Mola F, Tayfur V, Vayvada H. Prevention of peritendinous adhesions following flexor tendon injury with seprafilm. Ann Plast Surg
29. Liew SH, Potokar T, Bantick GL, Morgan I, Ford C, Murison MS. The use of ADCON-T/N after repair of zone II flexor tendons. Chir Main
30. Moran SL, Ryan CK, Orlando GS, Pratt CE, Michalko KB. Effects of 5-flurouracil on flexor tendon repair. J Hand Surg Am
. 2000;25: 242-251.
31. Nishida J, Araki S, Akasaka T, Toba T, Shimamura T, Amadio PC, An KN. Effect of hyaluronic acid on the excursion resistance of tendon grafts: a biomechanical study in a canine model in vitro. J Bone Joint Surg Br
32. Ozgenel GY. The effects of a combination of hyaluronic and amniotic membrane on the formation of peritendinous adhesions after flexor tendon surgery in chickens. J Bone Joint Surg Br
. 2004;86: 301-307.
33. Peacock EE Jr, Madden JW. Some studies on the effects of beta-aminopropionitrile in patients with injured flexor tendons. Surgery
34. Peterson WW, Manske PR, Dunlap J, Horwitz DS, Kahn B. Effect of various methods of restoring flexor sheath integrity on the formation of adhesions after tendon injury. J Hand Surg Am
. 1990;15: 48-56.
35. Riaz M, Hill C, Khan K, Small JO. Long term outcome of early active mobilization following flexor tendon repair in zone 2. J Hand Surg Br
36. Rosenbloom J, Harsch M, Jimenez S. Hydroxyproline content determines denaturation temperature of chick tendon collagen. Arch Biochem Biophys
37. Small JO, Brennen MD, Colville J. Early active mobilisation following flexor tendon repair in zone 2. J Hand Surg Br
. 1989;14: 383-391.
38. Speer DP, Feldman S, Chvapil M. The control of peritendinous adhesions using topical beta-aminopropionitrile base. J Surg Res
39. Strickland JW. Results of flexor tendon surgery in zone II. Hand Clin
40. Wang SJ, Dai GF, Li X, Liu Q, Shi DP. The clinical study of adhesion prevention of sodium hyaluronate in flexer tendon surgery. Zhongguo Xiu Fu Chong Jian Wai Ke Za Zhi