- Email: [email protected]
Gliding resistance after repair of partially lacerated human flexor digitorum profundus tendon in vitro
Clinical Biomechanics 16 (2001) 696±701
www.elsevier.com/locate/clinbiomech
Gliding resistance after repair of partially lacerated human ¯exor digitorum profundus tendon in vitro Chunfeng Zhao, Peter C. Amadio *, Mark E. Zobitz, Toshimitsu Momose, Paulus Couvreur, Kai-Nan An Orthopaedic Biomechanical Laboratory, Division of Orthopedic Research, Mayo Clinic/Mayo Foundation, 200 First Street, S.W., Gligg 1-128, Rochester, MN 55905, USA Received 2 January 2001; accepted 5 June 2001
Abstract Objective. This study reports the gliding resistance between repaired, partially lacerated tendon and pulley in human cadaver digits, using several commonly employed repair techniques. Background. Suture techniques with multi-strands and locking loops have been recommended to reduce the risk of rupture of the repair tendon with early active motion. Such sutures may increase the gliding resistance, and the gliding resistance after tendon repair is also an important factor in¯uencing the rehabilitation. Method. 105 specimens of second, third, or fourth ®ngers from 36 adult human hands were tested for the gliding resistance between ¯exor digitorum profundus tendon and A2 pulley in the normal condition. After an 80% laceration, each tendon was repaired with one of the following suture techniques: (1) Kessler; (2) modi®ed Kessler; (3) Savage; (4) Lee; (5) Tsuge; and (6) Becker. All suture techniques were reinforced with a circumferential epitenon simple running suture. After tendon repair, the gliding resistance was remeasured. Results. The gliding resistance of the Becker repair was signi®cantly greater than each of the other four repairs P < 0:05. The resistance of the modi®ed Kessler repair was signi®cantly less than that of the Kessler, Savage, or Tsuge repairs. Conclusions. We conclude that the type of tendon repair can signi®cantly aect the gliding resistance between the tendon and pulley system after tendon repair. Relevance The design of the tendon repair, through its eect on friction, may have an adverse eect on the clinical results of tendon mobilization. Ó 2001 Published by Elsevier Science Ltd. Keywords: Flexor tendon; Frictional force; Tendon repair; Partial laceration
1. Introduction Since the 1970s, when early motion postoperative rehabilitation protocols were accepted, the outcome for ¯exor tendon injuries has improved [1±3]. Strong suture techniques are necessary to prevent gap formation and tendon rupture with early mobilization after tendon repair [4±6]. Resistance to tendon gliding is also very important, especially in zone 2, where two tendons reside within a ®bro-osseous pulley system. In the normal condition, the gliding resistance between tendon and
*
Corresponding author. E-mail address: [email protected] (P.C. Amadio).
0268-0033/01/$ - see front matter Ó 2001 Published by Elsevier Science Ltd. PII: S 0 2 6 8 - 0 0 3 3 ( 0 1 ) 0 0 0 5 6 - 0
pulley is very small due to the smooth surfaces and the presence of a lubrication mechanism [7,8]. However, following injury and repair, the rough tendon surface and bulk of the repair can produce higher gliding resistance, which may aect the rehabilitation, especially, for a low force therapy [9,10]. Although the work of ¯exion can used to evaluate ®nger function, this method cannot isolate the resistance within the tendon sheath, which is the only resistance directly aected by the tendon repair [11]. An experimental system has been developed which allows the measurement of the frictional interaction at the tendon±pulley interface [12]. Recently, this methodology was used to characterize the gliding resistance with several suture techniques on the partially lacerated tendon in a canine model [13].
C. Zhao et al. / Clinical Biomechanics 16 (2001) 696±701
697
The partial laceration model was used to permit perfect coaptation, so that the only variable would be the suture type. It was reported that the high strength suture techniques frequently present high gliding resistance. However, this ®nding has not been con®rmed in human ®ngers. The purpose of this study was to assess the gliding characteristics in a human model, using several commonly employed tendon repair techniques. In order to provide direct comparability with the canine data, a partial laceration model was again used. 2. Methods Thirty-six human hands were harvested from 24 cadavers which ranged from 49- to 93-years old (average 73 years). The index, middle and ring ®ngers of each hand were used for testing. A transverse incision of the ¯exor sheath was made just distal to the A2 pulley in order to mark the lateral surface of the ¯exor digitorum profundus tendon with the digit in full extension. The tendon was then pulled proximally to full proximal interphalangeal (PIP) and distal interphalangeal (DIP) joint ¯exion. In this position, the tendon was again marked through the previous incision. The distance between these two marks represented the tendon excursion. In each ®nger the proximal and middle phalanges, ¯exor digitorum profundus (FDP) and ¯exor digitorum super®cialis (FDS) tendons and their insertions, and A2 pulley were preserved. A 1.5 mm Kirschner wire was inserted through the phalanx parallel to the long axis of the bone and passed through the PIP joint to maintain it in an extended position. The gliding resistance between the FDP and A2 pulley, FDS and bone was measured with the following procedures. The proximal phalanx (with PIP joint, A2 pulley, FDS and FDP tendons) was mounted on a custom jig with the palmar side upward. To maintain tension in the FDS tendon a 2.0 N weight was attached to its proximal end. The measurement system consisted of one mechanical actuator with a linear potentiometer, two tensile load transducers, and a mechanical pulley (Fig. 1). Load transducers were connected to the proximal and distal ends of the FDP tendon. Based on experience in previous studies, a 4.9 N weight was attached to the distal end of the FDP tendon to simulate passive mobilization of the ®nger. The distal transducer (F1) was connected to the weight and the proximal load transducer (F2) was connected to the mechanical actuator. The actuator was positioned at the preselected angle, de®ned as the angle in degrees formed between the horizontal plane and the proximal cable extension. The pulley for the weight was positioned at angle b, de®ned as the angle in degrees formed between the horizontal plane and the distal cable extension. Based on the experience of our previous studies [14,15], a set
Fig. 1. The device for gliding resistance measurement between the FDP tendon and pulley system. 1 ± Linear potentiometer, 2 ± Load transducer, 3 ± FDS tendon, 4 ± FDP tendon, 5 ± Saline bath, 6 ± Specimen, 7 ± Weight.
arc of contact a 20; b 30 was able to provide adequate measurement of the gliding resistance. This set angle was adopted for all the measurements in this study. The tendon was pulled proximally by the actuator against the weight at a rate of 2.0 mm/s. This movement of the tendon toward the actuator was regarded as ¯exion. The actuator movement was then reversed, causing the tendon to be pulled distally by the weight. This movement of the tendon toward the weight was regarded as extension. The readings from the load transducers (F1, F2) and the corresponding excursion were recorded by a digital computer at a sampling rate of 10 Hz. Excursion was limited to the distance between the two tendon markers. All specimens were kept moist throughout testing by immersion in a saline bath, which was incorporated into the testing jig. After the gliding resistance test for the normal tendon, the FDP tendon was removed from the specimen and lacerated to 80% of its transverse section according to the method of Dobyns et al. [16]. The laceration was located 10 mm distal to the proximal tendon marker, in order to allow the repair site to travel the full length of the A2 pulley. Each tendon was then repaired with one of the following six randomly assigned suture techniques: (1) Kessler; (2) Modi®ed Kessler; (3) Tsuge; (4) Savage (modi®ed with four-strand); (5) Lee (modi®ed with two-strand, simple loops); and (6) Augmented Becker (MGH) (Fig. 2). All suture techniques were reinforced with a circumferential epitenon simple running suture based on the work of Coert et al. [9] which showed the repairs without a running peripheral suture would lock or catch on the pulley edge. A 4/0 coated polyester suture (Ticron) was used for all core sutures except the MGH, which used a 4/0 nylon suture. The epitenon running suture used 6/0 nylon suture in all cases. In order to investigate the gliding characteristics after tendon repair, all suture knots and loops were located within the two tendon markers so that they would
698
C. Zhao et al. / Clinical Biomechanics 16 (2001) 696±701
Dierences in the mean gliding resistances between the sutured tendon and the intact were analyzed using a two-sample t-test assuming unequal variances to detect signi®cant dierences. Results between normal and immediately after repair were evaluated using a paired ttest. In all cases a level of P < 0:05 was considered to be statistically signi®cant. 4. Results Fig. 2. Six suture techniques: (a) Kessler: two-strands with double knots outside tendon surface and four loops. (b) Modi®ed Kessler: two-strands with single knot within repair site and four loops. (c) Tsuge: loop suture with single knot outside tendon surface and locking loops. (d) Savage (modi®ed): four-strands with single knot within repair site and eight loops. (e) Lee (modi®ed): two-strands with double knots within repair site and eight loops. (f) Becker: six-strands with three knots outside tendon surface and 24 loops.
3. Data analysis
The mean intact tendon gliding resistance was 0.27 N (SD 0.14 N) with a range from 0.11 to 0.93 N. There was no signi®cant dierence either grouped by dierent digits or by subsequently assigned suture method P > 0:05. After tendon repair, the gliding resistance signi®cantly increased to 1.08 N (SD 0.44 N) with a range from 0.52 to 3.09 N P < 0:05 (Fig. 3). The mean and standard deviation of the gliding resistance for the six groups of repaired tendons are listed on Table 1. To compare dierences between the suture techniques, the data were normalized by the dierence between repaired and normal tendon. The gliding resistance of the MGH repair was signi®cantly higher than any of the other ®ve repairs (Tsuge, Savage, Kessler, Lee, and modi®ed Kessler repairs) P < 0:05. The
Data collection during tendon excursion began when the proximal marker on the tendon surface reached the distal edge of the A2 pulley and ended when the distal surface marker moved past the distal edge of the pulley. The frictional force could then be calculated from the absolute dierence between the two load transducers (F1 and F2). In the ¯exion direction the actuator provides the driving force and therefore F2 is greater than F1, while in the extension direction the 4.9 Newton dead weight drives the excursion and thus F1 is higher. Since F1 remains constant during the test, the gliding resistance may be calculated as half of the dierence of the extremes: 1=2 F 2flexion F 2extension .
Fig. 3. Intact versus sutured. The gliding resistance after tendon repair was signi®cantly greater than intact P < 0:05. N 18 in each group (N 15 in Tsuge epair).
pass beneath the A2 pulley during testing. Each repaired tendon was tested in the same bed in which it had been tested prior to laceration, with the tendon put back through the A2 pulley in its original orientation.
Table 1 The characteristics of six suture techniques and gliding resistance Suture method (n)
Resistance (N) (SD)
Strands
Loops
Knots and location
Suture material
Statisticsa (resistance)
Intact (105) Modi®ed Kessler (18) Lee (modi®ed) (18) Kessler (18) Savage (18) Tsuge (15)
0.27 0.81 0.87 1.07 1.10 1.12
(0.14) (0.25) (0.24) (0.40) (0.30) (0.37)
± 2 2 2 4 2
± 1, 2, 2, 1, 1,
± Polyester Polyester Polyester Polyester Polyester
± + ++ +++ ++++ ++++
Becker (18)
1.52 (0.62)
6
± 4 8 4 12 2 Locking loop 24
Nylon
++++++
SD: Standard deviation. a Equal or larger than ++ (signi®cant dierence).
Inside Inside Outside Inside Outside
3, Outside
C. Zhao et al. / Clinical Biomechanics 16 (2001) 696±701
Fig. 4. The dierence between intact and repair in the average gliding resistance. The dark ``-'' markers indicate the average for the 18 repairs in each group (15 repairs in Tsuge technique).
modi®ed Kessler repair had a signi®cantly lower resistance than the Tsuge, Savage, or Kessler repairs P < 0:05. There was no signi®cant dierence between MK and Lee, Lee and Kessler, and Kessler, Savage, and Tsuge repairs P > 0:05 (Fig. 4). 5. Discussion Eective postoperative therapy after ¯exor tendon repair in zone 2 will maintain tendon gliding within the sheath, which is the key to preventing adhesion formation [17,18]. Because the tendon±sheath interaction is not frictionless, it is clear that the gliding resistance must be overcome before the tendon can move within its sheath. In the normal state, this resistance is small because of the smoothness of the opposing surfaces and the lubricating eect of the synovial ¯uid on the ¯exor tendon sheath and joints [19,20]. After tendon repair, however, resistance to ®nger motion can be dramatically increased by edema, suture technique, or the eects of wound healing. Later on, adhesions can also reduce tendon gliding. We believe that a better understanding of the evolution of the gliding resistance after tendon repair will aid our knowledge of repair and rehabilitation of tendon injuries, and will guide future improvement in treatment of these complex injuries. In our study, the gliding resistance after tendon repair increased to 1.08 N, which is four times greater than the resistance of a normal tendon. Although this resistance is small compared to the suture strength itself [6,21], it is comparable to the force applied to the tendon during the rehabilitation. This may not aect a tendon moving actively after tendon repair, as high loading is applied to the tendon. However, this amount of gliding resistance may be very important with passive motion therapy, since only a small force is then applied to the tendon [22]. Schuind et al. [23] reported that the force applied to the tendon during active motion ranged from 0.98 to 8.4 N, averaging 18.6 N. However, during passive motion,
699
the force only ranged from 0.098 to 2.9 N, averaging 0.49 N. Thus, this increase of resistance may not contribute to gap formation or suture rupture, but it could easily limit tendon gliding during passive motion therapy [24], the most common form of therapy currently in use for ¯exor tendon rehabilitation. The six suture methods used in this study are clinically common techniques which have individual characteristics, including dierent number of suture strands and loops, dierent knot styles and locations, and different suture materials (Table 1). The modi®ed Kessler suture technique, which includes two strands, one knot inside of the tendon ends, and four simple loops, had the lowest gliding resistance among the six methods. Although the Kessler repair is very similar, except the knots are outside the tendon surface, its resistance was greater than that of the modi®ed Kessler repair. Therefore, we concluded that knots on the outside tendon surface increase the gliding resistance compared to knots buried between the repaired tendon ends. Correspondingly, although the Lee suture technique also has two strands with knots inside tendon ends, it has eight loops, double the number in the MK repair. The higher gliding resistance in the Lee suture suggests that more loops on the outside tendon surface increase the gliding resistance between tendon and pulley. Both the modi®ed Lee and Savage suture techniques have eight loops, with knots inside the cut tendon ends. These repairs dier in that the Savage is a 4-strand suture, twice as many as the Lee repair. The resistance of the Savage suture method was higher than that of the Lee repair. This suggests that the bulk of the extra strands has an eect on the gliding resistance. Similarly, the resistance of the Tsuge repair was also high in our testing group, and its knot was located outside of the tendon surface. Based on the above considerations, one could predict that the MGH repair would have the highest gliding resistance among the repairs studied here. It has six strands, two outside knots, and 24 loops on the tendon surface. Our results are consistent with this prediction. Previously we performed a similar study on gliding resistance following partial tendon repair in a canine model [13]. The similarity in results demonstrates that the gliding characteristics of human tendon are reliably mimicked using a canine model of ¯exor tendon repair. Our partial laceration model was chosen because it is clinically relevant and surgical management of such injuries remains controversial [16,25]. The partial laceration model also allowed us to more perfectly isolate the eect of suture repair, as it is easier to avoid malalignment and bunching during suturing, so that the gliding characteristics of dierent repairs can be precisely assessed and compared. This model is of course also directly applicable to the clinical situation of high percentage partial laceration, where many authors advocate repair [26,27]. In such circumstances, as repair
700
C. Zhao et al. / Clinical Biomechanics 16 (2001) 696±701
strength is less of an issue with partial than it is for complete lacerations [25] these data suggest that a low friction repair would be preferred. For tendon repairs in general, while multiple suture strands and loops increase suture strength, they also increase the gliding resistance. The ideal suture construct would have a combination of high strength and low gliding resistance. Unfortunately, for the sutures we studied a tradeo exists, as the stronger repairs (based on studies of these repairs done elsewhere) tended to have higher gliding resistance. This raises the question of which is more important for successful tendon healing: gliding resistance or suture strength. Low gliding resistance suture techniques should not be overemphasized, since gap formation and tendon rupture are serious complications of weaker repairs. Nevertheless, at some point we believe there is a diminishing return, where greater suture strength produces little added protection against rupture, while increasingly jeopardizing tendon gliding. One limitation of our study was that the gliding resistance we measured included only the FDP, A1 and A2 pulley, FDS, and bony interaction. Although it did not exactly represent the gliding resistance between the tendon and the entire sheath system, we believe that our measurement is clinically relevant, since the proximal pulley is the most important structure in the sheath [15,28]. Secondly, the gliding resistance we measured included both the surface friction and the bulk friction [29]. As this was an in vitro study, however, the bulk friction may be not precisely measured, as there was no edema in the repaired tendon or surrounding soft tissues. 6. Conclusions From our study, we conclude that ¯exor tendon gliding resistance is signi®cantly aected by suture technique. The modi®ed Kessler suture had the lowest resistance among the six suture techniques studied. Based on our analysis, we believe that suture techniques with a multi-strand core suture, knots located outside the tendon surface, and multiple or locking loops on the tendon surface may result in increased gliding resistance between the tendon and pulley system after tendon repair. This increased resistance is of a magnitude similar to that generated in tendon mobilized passively. Thus, we believe that the design of the tendon repair, through its eect on friction, may have an adverse eect on the clinical results of tendon mobilization in certain circumstances. Acknowledgements This study was funded by grant # AR 44391, awarded by the National Institutes of Health (NIAMS).
References [1] Verdan C. Half a century of ¯exor tendon surgery. J Bone Joint Surg Am 1972;54:472±91. [2] Strickland JW. Results of ¯exor tendon surgery in zone II. Hand Clin 1985;1:167±79. [3] Seiler 3rd JG, Chu CR, Amiel D, Woo SL, Gelberman RH. The Marshall R. Urist Young Investigator Award. Autogenous ¯exor tendon grafts. Biologic mechanisms for incorporation. Clin Orthop 1997;345:239±47. [4] Savage R, Risitano G. Flexor tendon repair using a ``six strand'' method of repair and early active mobilisation. J Hand Surg Br 1989;14:396±9. [5] Lee H. Double loop locking suture. J Hand Surg Am 1990;15:953±8. [6] Winters SC, Gelberman RH, Woo SL, Chan SS, Grewal R, Seiler 3rd JG. The eects of multiple-strand suture methods on the strength and excursion of repaired intrasynovial ¯exor tendons a biomechanical study in dogs. J Hand Surg Am 1998;23:97±104. [7] Uchiyama S, Amadio PC, Ishikawa J, An KN. Boundary lubrication between the tendon and the pulley in the ®nger. J Bone Joint Surg Am 1997;79:213±8. [8] Moro-oka T, Miura H, Higaki H, Arimura S, Mawatari T, Murakami T, Iwamoto Y. A new friction tester of the ¯exor tendon. J Biomech 1999;32:1131±4. [9] Coert JH, Uchiyama S, Amadio PC, Berglund LJ, An KN. Flexor tendon±pulley interaction after tendon repair. A biomechanical study. J Hand Surg Br 1995;20:573±7. [10] Lieber R, Silva M, Amiel D, Gelberman R. Wrist and digital joint motion produce unique ¯exor tendon force and excursion in the canine forelimb. J Biomech 1999;32:175±81. [11] Aoki M, Manske PR, Pruitt DL, Larson BJ. Work of ¯exion after tendon repair with various suture methods. A human cadaveric study. J Hand Surg Br 1995;20:310±3. [12] An KN, Berglund L, Uchiyama S, Coert JH. Measurement of friction between pulley and ¯exor tendon. Biomed Sci Instrum 1993;29:1±7. [13] Zhao CF, Amadio PC, An KN, Zobitz ME. Gliding characteristics of tendon repair in partially lacerated canine FDP tendons. Trans Orthop Res Soc 1999;24:120. [14] Uchiyama S, Coert JH, Berglund L, Amadio PC, An KN. Method for the measurement of friction between tendon and pulley. J Orthop Res 1995;13:83±9. [15] Zhao CF, Amadio PC, Berglund L, An KN. The A3 pulley. J Hand Surg Am 2000;25:270±6. [16] Dobyns RC, Cooney WC, Wood MB. Eect of partial lacerations on canine ¯exor tendons. Minn Med 1982;65:27±32. [17] Lister GD, Kleinert HE, Kutz JE, Atasoy E. Primary ¯exor tendon repair followed by immediate controlled mobilization. J Hand Surg Am 1977;2:441±51. [18] Silfverskiold KL, May EJ. Flexor tendon repair in zone II with a new suture technique and an early mobilization program combining passive and active ¯exion. J Hand Surg Am 1994;19:53±60 see comments. [19] Doyle J. Anatomy of the ®nger ¯exor tendon sheath and pulley system. J Hand Surg Am 1988;13:473±84. [20] Hagberg L, Heinegard D, Ohlsson K. The contents of macromoleculer solutes in ¯exor tendon sheath ¯uid and ther relation to synovial ¯uid: a quantitative analysis. J Hand Surg Br 1992;17:167±71. [21] McLarney E, Homan H, Wolfe SW. Biomechanical analysis of the cruciate four-strand ¯exor tendon repair. J Hand Surg Am 1999;24:295±301. [22] Lieber RL, Amiel D, Kaufman KR, Whitney J, Gelberman RH. Relationship between joint motion and ¯exor tendon force in the canine forelimb. J Hand Surg Am 1996;21:957±62.
C. Zhao et al. / Clinical Biomechanics 16 (2001) 696±701 [23] Schuind F, Garcia-Elias M, Cooney WP, An KN. Flexor tendon forces in vivo measurements. J Hand Surg Am 1992;17:291±8. [24] Zhao CF, Amadio PC, Momose T, Couvreur P, Zobitz ME, An KN. The eect of synergistic motion on FDP excursion after tendon repair in a canine model in vivo. Trans Orthop Res Soc 2000;25:46. [25] Hariharan JS, Diao E, Soejima O, Lotz JC. Partial lacerations of human digital ¯exor tendons: a biomechanical analysis. J Hand Surg Am 1997;22:1011±5 see comments. [26] McGeorge DD, Stilwell JH. Partial ¯exor tendon injuries: to repair or not. J Hand Surg Br 1992;17:176±7.
701
[27] McCarthy DM, Boardman ND, 3rd, Tramaglini DM, Sotereanos DG, Herndon JH. Clinical management of partially lacerated digital ¯exor tendons: a survey [corrected] of hand surgeons [published erratum appears in J Hand Surg Am 1995; 20(3):526]. J Hand Surg Am 1995;20:273±275. [28] Lin GT, Cooney WP, Amadio PC, An KN. Mechanical properties of human pulleys. J Hand Surg Br 1990;15:429±34. [29] Mow VC, Soslowsky LJ. Friction, lubrication, and wear of diarthrodial joints. In: Mow VC, Hayes WC, editors. Basic orthopaedic biomechanics. New York: Raven Press; 1991. p. 245± 92.
www.elsevier.com/locate/clinbiomech
Gliding resistance after repair of partially lacerated human ¯exor digitorum profundus tendon in vitro Chunfeng Zhao, Peter C. Amadio *, Mark E. Zobitz, Toshimitsu Momose, Paulus Couvreur, Kai-Nan An Orthopaedic Biomechanical Laboratory, Division of Orthopedic Research, Mayo Clinic/Mayo Foundation, 200 First Street, S.W., Gligg 1-128, Rochester, MN 55905, USA Received 2 January 2001; accepted 5 June 2001
Abstract Objective. This study reports the gliding resistance between repaired, partially lacerated tendon and pulley in human cadaver digits, using several commonly employed repair techniques. Background. Suture techniques with multi-strands and locking loops have been recommended to reduce the risk of rupture of the repair tendon with early active motion. Such sutures may increase the gliding resistance, and the gliding resistance after tendon repair is also an important factor in¯uencing the rehabilitation. Method. 105 specimens of second, third, or fourth ®ngers from 36 adult human hands were tested for the gliding resistance between ¯exor digitorum profundus tendon and A2 pulley in the normal condition. After an 80% laceration, each tendon was repaired with one of the following suture techniques: (1) Kessler; (2) modi®ed Kessler; (3) Savage; (4) Lee; (5) Tsuge; and (6) Becker. All suture techniques were reinforced with a circumferential epitenon simple running suture. After tendon repair, the gliding resistance was remeasured. Results. The gliding resistance of the Becker repair was signi®cantly greater than each of the other four repairs P < 0:05. The resistance of the modi®ed Kessler repair was signi®cantly less than that of the Kessler, Savage, or Tsuge repairs. Conclusions. We conclude that the type of tendon repair can signi®cantly aect the gliding resistance between the tendon and pulley system after tendon repair. Relevance The design of the tendon repair, through its eect on friction, may have an adverse eect on the clinical results of tendon mobilization. Ó 2001 Published by Elsevier Science Ltd. Keywords: Flexor tendon; Frictional force; Tendon repair; Partial laceration
1. Introduction Since the 1970s, when early motion postoperative rehabilitation protocols were accepted, the outcome for ¯exor tendon injuries has improved [1±3]. Strong suture techniques are necessary to prevent gap formation and tendon rupture with early mobilization after tendon repair [4±6]. Resistance to tendon gliding is also very important, especially in zone 2, where two tendons reside within a ®bro-osseous pulley system. In the normal condition, the gliding resistance between tendon and
*
Corresponding author. E-mail address: [email protected] (P.C. Amadio).
0268-0033/01/$ - see front matter Ó 2001 Published by Elsevier Science Ltd. PII: S 0 2 6 8 - 0 0 3 3 ( 0 1 ) 0 0 0 5 6 - 0
pulley is very small due to the smooth surfaces and the presence of a lubrication mechanism [7,8]. However, following injury and repair, the rough tendon surface and bulk of the repair can produce higher gliding resistance, which may aect the rehabilitation, especially, for a low force therapy [9,10]. Although the work of ¯exion can used to evaluate ®nger function, this method cannot isolate the resistance within the tendon sheath, which is the only resistance directly aected by the tendon repair [11]. An experimental system has been developed which allows the measurement of the frictional interaction at the tendon±pulley interface [12]. Recently, this methodology was used to characterize the gliding resistance with several suture techniques on the partially lacerated tendon in a canine model [13].
C. Zhao et al. / Clinical Biomechanics 16 (2001) 696±701
697
The partial laceration model was used to permit perfect coaptation, so that the only variable would be the suture type. It was reported that the high strength suture techniques frequently present high gliding resistance. However, this ®nding has not been con®rmed in human ®ngers. The purpose of this study was to assess the gliding characteristics in a human model, using several commonly employed tendon repair techniques. In order to provide direct comparability with the canine data, a partial laceration model was again used. 2. Methods Thirty-six human hands were harvested from 24 cadavers which ranged from 49- to 93-years old (average 73 years). The index, middle and ring ®ngers of each hand were used for testing. A transverse incision of the ¯exor sheath was made just distal to the A2 pulley in order to mark the lateral surface of the ¯exor digitorum profundus tendon with the digit in full extension. The tendon was then pulled proximally to full proximal interphalangeal (PIP) and distal interphalangeal (DIP) joint ¯exion. In this position, the tendon was again marked through the previous incision. The distance between these two marks represented the tendon excursion. In each ®nger the proximal and middle phalanges, ¯exor digitorum profundus (FDP) and ¯exor digitorum super®cialis (FDS) tendons and their insertions, and A2 pulley were preserved. A 1.5 mm Kirschner wire was inserted through the phalanx parallel to the long axis of the bone and passed through the PIP joint to maintain it in an extended position. The gliding resistance between the FDP and A2 pulley, FDS and bone was measured with the following procedures. The proximal phalanx (with PIP joint, A2 pulley, FDS and FDP tendons) was mounted on a custom jig with the palmar side upward. To maintain tension in the FDS tendon a 2.0 N weight was attached to its proximal end. The measurement system consisted of one mechanical actuator with a linear potentiometer, two tensile load transducers, and a mechanical pulley (Fig. 1). Load transducers were connected to the proximal and distal ends of the FDP tendon. Based on experience in previous studies, a 4.9 N weight was attached to the distal end of the FDP tendon to simulate passive mobilization of the ®nger. The distal transducer (F1) was connected to the weight and the proximal load transducer (F2) was connected to the mechanical actuator. The actuator was positioned at the preselected angle, de®ned as the angle in degrees formed between the horizontal plane and the proximal cable extension. The pulley for the weight was positioned at angle b, de®ned as the angle in degrees formed between the horizontal plane and the distal cable extension. Based on the experience of our previous studies [14,15], a set
Fig. 1. The device for gliding resistance measurement between the FDP tendon and pulley system. 1 ± Linear potentiometer, 2 ± Load transducer, 3 ± FDS tendon, 4 ± FDP tendon, 5 ± Saline bath, 6 ± Specimen, 7 ± Weight.
arc of contact a 20; b 30 was able to provide adequate measurement of the gliding resistance. This set angle was adopted for all the measurements in this study. The tendon was pulled proximally by the actuator against the weight at a rate of 2.0 mm/s. This movement of the tendon toward the actuator was regarded as ¯exion. The actuator movement was then reversed, causing the tendon to be pulled distally by the weight. This movement of the tendon toward the weight was regarded as extension. The readings from the load transducers (F1, F2) and the corresponding excursion were recorded by a digital computer at a sampling rate of 10 Hz. Excursion was limited to the distance between the two tendon markers. All specimens were kept moist throughout testing by immersion in a saline bath, which was incorporated into the testing jig. After the gliding resistance test for the normal tendon, the FDP tendon was removed from the specimen and lacerated to 80% of its transverse section according to the method of Dobyns et al. [16]. The laceration was located 10 mm distal to the proximal tendon marker, in order to allow the repair site to travel the full length of the A2 pulley. Each tendon was then repaired with one of the following six randomly assigned suture techniques: (1) Kessler; (2) Modi®ed Kessler; (3) Tsuge; (4) Savage (modi®ed with four-strand); (5) Lee (modi®ed with two-strand, simple loops); and (6) Augmented Becker (MGH) (Fig. 2). All suture techniques were reinforced with a circumferential epitenon simple running suture based on the work of Coert et al. [9] which showed the repairs without a running peripheral suture would lock or catch on the pulley edge. A 4/0 coated polyester suture (Ticron) was used for all core sutures except the MGH, which used a 4/0 nylon suture. The epitenon running suture used 6/0 nylon suture in all cases. In order to investigate the gliding characteristics after tendon repair, all suture knots and loops were located within the two tendon markers so that they would
698
C. Zhao et al. / Clinical Biomechanics 16 (2001) 696±701
Dierences in the mean gliding resistances between the sutured tendon and the intact were analyzed using a two-sample t-test assuming unequal variances to detect signi®cant dierences. Results between normal and immediately after repair were evaluated using a paired ttest. In all cases a level of P < 0:05 was considered to be statistically signi®cant. 4. Results Fig. 2. Six suture techniques: (a) Kessler: two-strands with double knots outside tendon surface and four loops. (b) Modi®ed Kessler: two-strands with single knot within repair site and four loops. (c) Tsuge: loop suture with single knot outside tendon surface and locking loops. (d) Savage (modi®ed): four-strands with single knot within repair site and eight loops. (e) Lee (modi®ed): two-strands with double knots within repair site and eight loops. (f) Becker: six-strands with three knots outside tendon surface and 24 loops.
3. Data analysis
The mean intact tendon gliding resistance was 0.27 N (SD 0.14 N) with a range from 0.11 to 0.93 N. There was no signi®cant dierence either grouped by dierent digits or by subsequently assigned suture method P > 0:05. After tendon repair, the gliding resistance signi®cantly increased to 1.08 N (SD 0.44 N) with a range from 0.52 to 3.09 N P < 0:05 (Fig. 3). The mean and standard deviation of the gliding resistance for the six groups of repaired tendons are listed on Table 1. To compare dierences between the suture techniques, the data were normalized by the dierence between repaired and normal tendon. The gliding resistance of the MGH repair was signi®cantly higher than any of the other ®ve repairs (Tsuge, Savage, Kessler, Lee, and modi®ed Kessler repairs) P < 0:05. The
Data collection during tendon excursion began when the proximal marker on the tendon surface reached the distal edge of the A2 pulley and ended when the distal surface marker moved past the distal edge of the pulley. The frictional force could then be calculated from the absolute dierence between the two load transducers (F1 and F2). In the ¯exion direction the actuator provides the driving force and therefore F2 is greater than F1, while in the extension direction the 4.9 Newton dead weight drives the excursion and thus F1 is higher. Since F1 remains constant during the test, the gliding resistance may be calculated as half of the dierence of the extremes: 1=2 F 2flexion F 2extension .
Fig. 3. Intact versus sutured. The gliding resistance after tendon repair was signi®cantly greater than intact P < 0:05. N 18 in each group (N 15 in Tsuge epair).
pass beneath the A2 pulley during testing. Each repaired tendon was tested in the same bed in which it had been tested prior to laceration, with the tendon put back through the A2 pulley in its original orientation.
Table 1 The characteristics of six suture techniques and gliding resistance Suture method (n)
Resistance (N) (SD)
Strands
Loops
Knots and location
Suture material
Statisticsa (resistance)
Intact (105) Modi®ed Kessler (18) Lee (modi®ed) (18) Kessler (18) Savage (18) Tsuge (15)
0.27 0.81 0.87 1.07 1.10 1.12
(0.14) (0.25) (0.24) (0.40) (0.30) (0.37)
± 2 2 2 4 2
± 1, 2, 2, 1, 1,
± Polyester Polyester Polyester Polyester Polyester
± + ++ +++ ++++ ++++
Becker (18)
1.52 (0.62)
6
± 4 8 4 12 2 Locking loop 24
Nylon
++++++
SD: Standard deviation. a Equal or larger than ++ (signi®cant dierence).
Inside Inside Outside Inside Outside
3, Outside
C. Zhao et al. / Clinical Biomechanics 16 (2001) 696±701
Fig. 4. The dierence between intact and repair in the average gliding resistance. The dark ``-'' markers indicate the average for the 18 repairs in each group (15 repairs in Tsuge technique).
modi®ed Kessler repair had a signi®cantly lower resistance than the Tsuge, Savage, or Kessler repairs P < 0:05. There was no signi®cant dierence between MK and Lee, Lee and Kessler, and Kessler, Savage, and Tsuge repairs P > 0:05 (Fig. 4). 5. Discussion Eective postoperative therapy after ¯exor tendon repair in zone 2 will maintain tendon gliding within the sheath, which is the key to preventing adhesion formation [17,18]. Because the tendon±sheath interaction is not frictionless, it is clear that the gliding resistance must be overcome before the tendon can move within its sheath. In the normal state, this resistance is small because of the smoothness of the opposing surfaces and the lubricating eect of the synovial ¯uid on the ¯exor tendon sheath and joints [19,20]. After tendon repair, however, resistance to ®nger motion can be dramatically increased by edema, suture technique, or the eects of wound healing. Later on, adhesions can also reduce tendon gliding. We believe that a better understanding of the evolution of the gliding resistance after tendon repair will aid our knowledge of repair and rehabilitation of tendon injuries, and will guide future improvement in treatment of these complex injuries. In our study, the gliding resistance after tendon repair increased to 1.08 N, which is four times greater than the resistance of a normal tendon. Although this resistance is small compared to the suture strength itself [6,21], it is comparable to the force applied to the tendon during the rehabilitation. This may not aect a tendon moving actively after tendon repair, as high loading is applied to the tendon. However, this amount of gliding resistance may be very important with passive motion therapy, since only a small force is then applied to the tendon [22]. Schuind et al. [23] reported that the force applied to the tendon during active motion ranged from 0.98 to 8.4 N, averaging 18.6 N. However, during passive motion,
699
the force only ranged from 0.098 to 2.9 N, averaging 0.49 N. Thus, this increase of resistance may not contribute to gap formation or suture rupture, but it could easily limit tendon gliding during passive motion therapy [24], the most common form of therapy currently in use for ¯exor tendon rehabilitation. The six suture methods used in this study are clinically common techniques which have individual characteristics, including dierent number of suture strands and loops, dierent knot styles and locations, and different suture materials (Table 1). The modi®ed Kessler suture technique, which includes two strands, one knot inside of the tendon ends, and four simple loops, had the lowest gliding resistance among the six methods. Although the Kessler repair is very similar, except the knots are outside the tendon surface, its resistance was greater than that of the modi®ed Kessler repair. Therefore, we concluded that knots on the outside tendon surface increase the gliding resistance compared to knots buried between the repaired tendon ends. Correspondingly, although the Lee suture technique also has two strands with knots inside tendon ends, it has eight loops, double the number in the MK repair. The higher gliding resistance in the Lee suture suggests that more loops on the outside tendon surface increase the gliding resistance between tendon and pulley. Both the modi®ed Lee and Savage suture techniques have eight loops, with knots inside the cut tendon ends. These repairs dier in that the Savage is a 4-strand suture, twice as many as the Lee repair. The resistance of the Savage suture method was higher than that of the Lee repair. This suggests that the bulk of the extra strands has an eect on the gliding resistance. Similarly, the resistance of the Tsuge repair was also high in our testing group, and its knot was located outside of the tendon surface. Based on the above considerations, one could predict that the MGH repair would have the highest gliding resistance among the repairs studied here. It has six strands, two outside knots, and 24 loops on the tendon surface. Our results are consistent with this prediction. Previously we performed a similar study on gliding resistance following partial tendon repair in a canine model [13]. The similarity in results demonstrates that the gliding characteristics of human tendon are reliably mimicked using a canine model of ¯exor tendon repair. Our partial laceration model was chosen because it is clinically relevant and surgical management of such injuries remains controversial [16,25]. The partial laceration model also allowed us to more perfectly isolate the eect of suture repair, as it is easier to avoid malalignment and bunching during suturing, so that the gliding characteristics of dierent repairs can be precisely assessed and compared. This model is of course also directly applicable to the clinical situation of high percentage partial laceration, where many authors advocate repair [26,27]. In such circumstances, as repair
700
C. Zhao et al. / Clinical Biomechanics 16 (2001) 696±701
strength is less of an issue with partial than it is for complete lacerations [25] these data suggest that a low friction repair would be preferred. For tendon repairs in general, while multiple suture strands and loops increase suture strength, they also increase the gliding resistance. The ideal suture construct would have a combination of high strength and low gliding resistance. Unfortunately, for the sutures we studied a tradeo exists, as the stronger repairs (based on studies of these repairs done elsewhere) tended to have higher gliding resistance. This raises the question of which is more important for successful tendon healing: gliding resistance or suture strength. Low gliding resistance suture techniques should not be overemphasized, since gap formation and tendon rupture are serious complications of weaker repairs. Nevertheless, at some point we believe there is a diminishing return, where greater suture strength produces little added protection against rupture, while increasingly jeopardizing tendon gliding. One limitation of our study was that the gliding resistance we measured included only the FDP, A1 and A2 pulley, FDS, and bony interaction. Although it did not exactly represent the gliding resistance between the tendon and the entire sheath system, we believe that our measurement is clinically relevant, since the proximal pulley is the most important structure in the sheath [15,28]. Secondly, the gliding resistance we measured included both the surface friction and the bulk friction [29]. As this was an in vitro study, however, the bulk friction may be not precisely measured, as there was no edema in the repaired tendon or surrounding soft tissues. 6. Conclusions From our study, we conclude that ¯exor tendon gliding resistance is signi®cantly aected by suture technique. The modi®ed Kessler suture had the lowest resistance among the six suture techniques studied. Based on our analysis, we believe that suture techniques with a multi-strand core suture, knots located outside the tendon surface, and multiple or locking loops on the tendon surface may result in increased gliding resistance between the tendon and pulley system after tendon repair. This increased resistance is of a magnitude similar to that generated in tendon mobilized passively. Thus, we believe that the design of the tendon repair, through its eect on friction, may have an adverse eect on the clinical results of tendon mobilization in certain circumstances. Acknowledgements This study was funded by grant # AR 44391, awarded by the National Institutes of Health (NIAMS).
References [1] Verdan C. Half a century of ¯exor tendon surgery. J Bone Joint Surg Am 1972;54:472±91. [2] Strickland JW. Results of ¯exor tendon surgery in zone II. Hand Clin 1985;1:167±79. [3] Seiler 3rd JG, Chu CR, Amiel D, Woo SL, Gelberman RH. The Marshall R. Urist Young Investigator Award. Autogenous ¯exor tendon grafts. Biologic mechanisms for incorporation. Clin Orthop 1997;345:239±47. [4] Savage R, Risitano G. Flexor tendon repair using a ``six strand'' method of repair and early active mobilisation. J Hand Surg Br 1989;14:396±9. [5] Lee H. Double loop locking suture. J Hand Surg Am 1990;15:953±8. [6] Winters SC, Gelberman RH, Woo SL, Chan SS, Grewal R, Seiler 3rd JG. The eects of multiple-strand suture methods on the strength and excursion of repaired intrasynovial ¯exor tendons a biomechanical study in dogs. J Hand Surg Am 1998;23:97±104. [7] Uchiyama S, Amadio PC, Ishikawa J, An KN. Boundary lubrication between the tendon and the pulley in the ®nger. J Bone Joint Surg Am 1997;79:213±8. [8] Moro-oka T, Miura H, Higaki H, Arimura S, Mawatari T, Murakami T, Iwamoto Y. A new friction tester of the ¯exor tendon. J Biomech 1999;32:1131±4. [9] Coert JH, Uchiyama S, Amadio PC, Berglund LJ, An KN. Flexor tendon±pulley interaction after tendon repair. A biomechanical study. J Hand Surg Br 1995;20:573±7. [10] Lieber R, Silva M, Amiel D, Gelberman R. Wrist and digital joint motion produce unique ¯exor tendon force and excursion in the canine forelimb. J Biomech 1999;32:175±81. [11] Aoki M, Manske PR, Pruitt DL, Larson BJ. Work of ¯exion after tendon repair with various suture methods. A human cadaveric study. J Hand Surg Br 1995;20:310±3. [12] An KN, Berglund L, Uchiyama S, Coert JH. Measurement of friction between pulley and ¯exor tendon. Biomed Sci Instrum 1993;29:1±7. [13] Zhao CF, Amadio PC, An KN, Zobitz ME. Gliding characteristics of tendon repair in partially lacerated canine FDP tendons. Trans Orthop Res Soc 1999;24:120. [14] Uchiyama S, Coert JH, Berglund L, Amadio PC, An KN. Method for the measurement of friction between tendon and pulley. J Orthop Res 1995;13:83±9. [15] Zhao CF, Amadio PC, Berglund L, An KN. The A3 pulley. J Hand Surg Am 2000;25:270±6. [16] Dobyns RC, Cooney WC, Wood MB. Eect of partial lacerations on canine ¯exor tendons. Minn Med 1982;65:27±32. [17] Lister GD, Kleinert HE, Kutz JE, Atasoy E. Primary ¯exor tendon repair followed by immediate controlled mobilization. J Hand Surg Am 1977;2:441±51. [18] Silfverskiold KL, May EJ. Flexor tendon repair in zone II with a new suture technique and an early mobilization program combining passive and active ¯exion. J Hand Surg Am 1994;19:53±60 see comments. [19] Doyle J. Anatomy of the ®nger ¯exor tendon sheath and pulley system. J Hand Surg Am 1988;13:473±84. [20] Hagberg L, Heinegard D, Ohlsson K. The contents of macromoleculer solutes in ¯exor tendon sheath ¯uid and ther relation to synovial ¯uid: a quantitative analysis. J Hand Surg Br 1992;17:167±71. [21] McLarney E, Homan H, Wolfe SW. Biomechanical analysis of the cruciate four-strand ¯exor tendon repair. J Hand Surg Am 1999;24:295±301. [22] Lieber RL, Amiel D, Kaufman KR, Whitney J, Gelberman RH. Relationship between joint motion and ¯exor tendon force in the canine forelimb. J Hand Surg Am 1996;21:957±62.
C. Zhao et al. / Clinical Biomechanics 16 (2001) 696±701 [23] Schuind F, Garcia-Elias M, Cooney WP, An KN. Flexor tendon forces in vivo measurements. J Hand Surg Am 1992;17:291±8. [24] Zhao CF, Amadio PC, Momose T, Couvreur P, Zobitz ME, An KN. The eect of synergistic motion on FDP excursion after tendon repair in a canine model in vivo. Trans Orthop Res Soc 2000;25:46. [25] Hariharan JS, Diao E, Soejima O, Lotz JC. Partial lacerations of human digital ¯exor tendons: a biomechanical analysis. J Hand Surg Am 1997;22:1011±5 see comments. [26] McGeorge DD, Stilwell JH. Partial ¯exor tendon injuries: to repair or not. J Hand Surg Br 1992;17:176±7.
701
[27] McCarthy DM, Boardman ND, 3rd, Tramaglini DM, Sotereanos DG, Herndon JH. Clinical management of partially lacerated digital ¯exor tendons: a survey [corrected] of hand surgeons [published erratum appears in J Hand Surg Am 1995; 20(3):526]. J Hand Surg Am 1995;20:273±275. [28] Lin GT, Cooney WP, Amadio PC, An KN. Mechanical properties of human pulleys. J Hand Surg Br 1990;15:429±34. [29] Mow VC, Soslowsky LJ. Friction, lubrication, and wear of diarthrodial joints. In: Mow VC, Hayes WC, editors. Basic orthopaedic biomechanics. New York: Raven Press; 1991. p. 245± 92.