The effects of multiple-strand suture methods on the strength and excusion of repaired intrasynovial flexor tendons: A biomechanical study in dogs

The Effects of Multiple-Strand Suture Methods on the Strength and Excursion of Repaired Intrasynovial Flexor Tendons: A Biomechanical Study in Dogs Steven C. Winters, MD, Richard H. Gelberman, MD, St. Louis, MO, Savio L-Y. Woo, PhD, Serena S. Chan, MS, Rupinder Grewal, MD, Pittsburgh, PA, John Gray SeUer III, MD, St. Louis, MO This study was designed to determine the effects of in vivo multistrand, multigrasp suture techniques on the strength and gliding of repaired intrasynovial tendons when controlled passivemotion rehabilitation was used. Twenty-four adult mongrel dogs were divided into 4 groups and their medial and lateral forepaw flexor tendons were transsected and sutured by either the Savage, the Tajima, the Kessler, or the recently developed 8-strand suture method. The tendon excursion, joint rotation, and tensile properties of the repaired tendons were evaluated biomechanically at 3 and 6 weeks after surgery. It was found that neither time nor suture method significantly effected proximal and distal inteq3halangeal joint rotation or tendon excursion when the 4 techniques were compared to each other. Normalized load value (experimental/control) was significantly affected by both the suture method and the amount of time after surgery, however. The Savage and 8-strand repair methods had significantly greater strength than did the Tajima method at each time interval (p < .05 for each comparison). In addition, the 8-strand method had significantly greater normalized load values than did the Savage method at each time interval (p < .05 for each comparison). Normalized stiffness (experimental/control) for the 8-strand repair method was significantly greater than that for the Tajima and Savage methods at 3 and 6 weeks after surgery (p < .05). In addition, the normalized stiffness values for the 6-week groups was significantly greater than those for the 3-week groups (p < .05). It was concluded that the method of tendon suture was a significant variable insofar as the regaining of tendon strength was concerned and that the newer low-profile 8-strand repair method significantly expands the safety zone for the application of increased in vivo load during the early stages of rehabilitation. (J Hand Surg 1998;23A: 97-104. Copyright 9 1998 by the American Society for Surgery of the Hand.)

Recent clinical studies have determined that increased levels of proximal musculotendinous load From the Department of Orthopaedic Surgery, Washington University School of Medicine, St. Louis, MO, and the Musculoskeletal Research Center, Department of Orthopaedic Surgery, Universit3~ of Pittsburgh School of Medicine, Pittsburgh, PA. Supported by grant AR33097 tu the National Institutes of Health. Received for publication Aug. 5, 1996; accepted in revised form Sept. 18, 1997. No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. Reprint requests: Steven C. Winters, MD, Department of Orthopaedic Surgery, Washington University School of Medicine, One Barnes Hospital Plaza, suite 11300, St. Louis, MO 63110. Copyright 9 1998 by the American Society for Surgery of the Hand. 0363-5023/98/23A01-001753.00/0

and greater magnitudes of tendon repair-site excursion may enhance the healing response of repaired intrasynovial flexor tendons. 1-4 Realizing that the greater the force applied to the repair site during rehabilitation, the greater the risk of gap formation and tendon rupture, investigators have modified their methods for suturing tendons in an effort to increase the repair site's strength for the application of early controlled passive and active mobilization. 5-11 The resulting innovative tendon suture methods have featured a greater number of intratendinous grasping points and an increased number of suture strands extending across the repair site, coupled with modifications of the epitenon circumferential suture. The Journal of Hand Surgery 97

98 Winters et al. / Tendons Stronger With 8-Strand Suture A comprehensive study of the relationship between the strength and excursion of tendons repaired with multistrand, multigrasp suture methods and repair-site deformation resulting from early mobilization has not been carried out in a clinically relevant in vivo animal model. The primary goal of this project was to evaluate these tendon repair techniques with regard to the strength and gliding that they provide within the digital sheath on the basis of the method by which the suture was accomplished in a clinically applicable early passive motion model. Specifically, the ultimate strength, stiffness, linear excursion, and angular joint rotation was measured in canine flexor tendons after transection and suture by either the Savage, 1~ the Tajima, 12 the Kessler, 13 or the recently developed 8-strand suture method.

J Figure 2. Schematic drawing of the Tajima suture method.

Materials and Methods Twenty-four adult dogs, each weighing between 20 and 30 kg, were used as experimental animals. All operations and postoperative care were conducted within a vivarium. The animals were anesthetized with an initial intravenous dose of sodium Thiamylol, 0.5 mL/kg, which was supplemented by intermittent injections of atropine (0.5 mL) and Acepromazine (0.2 mL tim). The animals were intubated and anesthesia maintained with 1% halothane anesthesia. Midlateral incisions were used at the second and fifth toes to expose the digital sheath. The sheath was entered between the annular pulleys proximal and distal to the proximal interphalangeal (PIP) joint, and the flexor digitorum profundus (FDP) tendon was transversely incised to simulate a lacerated tendon. Each forepaw underwent random placement of a repair technique in the second and fifth toes so that by

completion, there were 12 repairs for each of 4 repair methods (n = 6 per group per repair interval). Two 2-strand methods were used: (1) the Kessler method, which is a traditional grasping suture technique with 2 external knots (Fig. 1), and (2) the Tajima method, a modification of the Kessler technique that calls for the use of 2 sutures with placement of the suture knots within the repair site (Fig. 2). Two multistrand techniques were used: (1) the Savage method, which emphasizes multiple strands as well as grasps and was modified from 6 to 4 strands to accommodate the smaller size of the canine as compared to the human flexor tendon (Fig. 3), and (2) an 8-strand technique, which maximized tendon strand numbers while minimizing tendon grasps by using a double-stranded needle to create continuous nonlocking volar and dorsal rectangles across the repair site (Fig. 4). The Kessler, Tajima, and Savage repairs were carried out with 4 - 0 braided Dacron suture. Owing to the unavailability of double-stranded braided Dacron suture, the 8-strand repair was carried out with a 4 - 0 double-

Figure 1. Schematic drawing of the Kessler suture

Figure 3. Schematic drawing of the modified Savage

method.

technique.

The Journal of Hand Surgery / Vol. 23A No. 1 January 1998

A

B

E

F

99

Figure 4. Schematic drawing of 8-strand repair method. (See text for explanation of steps.)

stranded polyfilament caprolactan suture ( 4 - 0 S'upramid, S. Jackson, Alexandria, VA). For the 8-strand repair, the needle was inserted into the tendon at the repair site and was extended through its posterolateral quadrant, exiting 1 cm from the cut tendon edge (Fig. 4A). The process progressed in a counterclockwise direction: the needle was inserted next just distal to its previous exit point to anchor the tendon transversely. The next insertion point completed the first posterolateral rect-

angle by paralleling the first suture pass along the tendon edge (Fig. 4B). The procedure was carried out in the same manner in the opposite tendon stump, completing a dorsal rectangle. The needle was then advanced to the palmar half of the tendon and the previous steps were duplicated, with the needle ultimately exiting the repair site opposite and palmar to the initial entry site (Fig. 4C, D). As none of the grasps were locking, tension was placed on the double-stranded suture to allow apposition of the tendon

100

Winters et al. / Tendons Stronger With 8-Strand Suture

ends before a 4-throw surgeon's knot was placed within the repair site (Fig. 4E). The suture knots lay within the repair site with the low-profile Tajima, Savage, and 8-strand repair methods, whereas they lay outside the repaired tendon with the Kessler method. For all repair groups, including the 8-strand method, a 6 - 0 nylon epitendinous running suture was inserted to invaginate the free ends of the tendons. An example of a completed 8-strand repair is shown in Figure 4F. After surgery, the animals were placed in polyurethane shoulder spica casts with their elbows at 90 ~ and their wrists at 45 ~. Early controlled passive mobilization was initiated on the day after surgery. The wrists and digits were passively flexed and extended to the limits of a dorsal extension bloc for 10 minutes daily. Twelve animals (24 tendons) were killed at 3 weeks and 12 animals were killed at 6 weeks (24 tendons). At the time of death, the digits were disarticulated at the metacarpophalangeal (MP) joints and the FDP tendons were transsected 2-3 cm proximal to the MP joints for biomechanical evaluation. In addition, the contralateral digits were similarly removed to serve as control digits and were subjected to the same testing protocol as the experimental digits. The metacarpal bone was removed first, and 2 1.6-mm stainless-steel pins were inserted into 2 holes drilled perpendicularly to the proximal phalanx to secure the specimen to the testing device. Verhoeff' s stain ~4 was used to create markers for the evaluation of gliding function. Two markers were drawn on the middle phalanx, 1 marker on the distal phalanx, and a fourth marker on the nail at a point along the imaginary extension of the longitudinal axis of the distal phalanx. To measure the gliding function of the tendon, we used a device that was designed to evaluate the angular rotation of the distal interphalangeal (DIP) joint, PIP joint, and excursion of the FDP tendon ~5 (Fig. 5). The markers on the digits were first digitized with the digit fully extended and then again after the application of a 1.5-N load to the FDP tendon, which placed the digit in a flexed position. The data from the 2 rotational variable differential transducers, together with the positions of the markers, permitted centers of rotation and angular displacement of the DIP and PIP joints to be calculated. 16 For linear excursion of the tendon, we used a linear variable differential transducer to measure the distance the tendon clamp translated after application of the load. These procedures were repeated 3 times, with the values obtained in the last 2 trials averaged to produce 1 result per specimen.

D DT Canir PIF FDP

Figure 5. The device designed to measure the angular rotations of both the distal and proximal interphalangeal (DIP and PIP) joints and linear excursion of the tendon. FDP, flexor digitorum profundus; LVDT, linear variable differential transducer; RVDT, rotational variable differential transducer.

After the gliding function of the tendons had been measured, the proximal and middle phalanges of each specimen were further dissected away, leaving the tendon and its attachment to the distal phalanx. This bone-tendon complex was subjected to uniaxial tensile testing to disruption. 17 The initial length for each tendon (i.e., the clamp-to-clamp distance) was set at 45 mm. The specimen was immersed in a bath of 0.9% saline at 37~ for 15 minutes to reach temperature equilibrium. A 0.5-N preload was applied and the tendon was loaded in tension at a crosshead speed of 20 mm/min until tendon disruption. A load-elongation curve, representing the structural properties of the bone-tendon complex, was recorded and the stiffness and ultimate load were determined from the curve. All experimental tendons in each of the 4 suture method groups were tested, as were the corresponding contralateral digits, which were used as controls. Data on DIP and PIP rotation, excursion, stiffness, and ultimate load of lacerated FDP tendons were

The Journal of Hand Surgery / Vol. 23A No. 1 January 1998

obtained. The difference between the experimental and control values was evaluated using a paired t-test with a significance level of values less than .05. To minimize the interanimal size differences, data from the experimental tendons were normalized by dividing by the respective control values. A 2-way analysis of variance was then used to evaluate the effects of time and suture methods. Significance was further evaluated with a Fisher's protected least-squared difference post-hoc test, with the level of significance set at values less than .05.

Results Each of the experimental animals tolerated the surgery and all postoperative rehabilitation protocols. There were no infections or cast failures. At the time of tendon harvesting, there was evidence of mild tendon gapping for tendons repaired with either the Tajima, the Savage, or the 8-strand repair methods. The 3-week specimens demonstrated a 2.5-mm gap of 1 Tajima repair, a 1-mm gap of 1 8-strand repair, and a 2-mm gap of another 8-strand repair. The 6-week specimens showed gapping of 1 mm for 1 Savage repair and 2 mm for 1 Tajima repair. In the 3-week Kessler group, however, there was 1 ruptured tendon and 1 tendon that was untestable (2 of 6 tendons failed). Further, 2 tendon ruptures occurred in the 6-week Kessler group and 1 additional tendon in that group was attenuated (3 of 6 tendons failed). Data in the Kessler group were therefore insufficient for statistical comparison with other suture methods for tendon repair stiffness (Fig. 6) and ultimate load (Fig. 7). The averages of the successful Kessler repairs were included in Table 1 for reference with respect to the other groups.

A

60

u} 50 (,5 0~ C

.~- 4o

[]

Tajima

[]

Savage

[]

Eight-strand Repair

9 20 .N_ ale

E lo

0 Z

0

i

3 Weeks

6 Weeks

1

Figure 6. Experimental/control (E/C) graph for normalized stiffness at 3 and 6 weeks for the Tajima, Savage, and 8-strand repair methods (mean -+ SD). Statistical significance is described in the text.

101

30 [ ] Tajima

~ =E

20

[]

Savage

[]

Eight-strand Repair

10

Z 3 Weeks

6 W~ eks

Figure 7. Experimental/control (E/C) graph for normalized load at 3 and 6 weeks for the Tajima, Savage, and 8-strand repair methods (mean --- SD). Statistical significance is described in the text.

Table 1 shows the experimental values as well as an average control value of the gliding function and tensile properties for the canine specimens. For gliding function, DIP rotation, PIP rotation, and excursion of the FDP were significantly less than the control values for the Savage group at 3 weeks. In addition, PIP rotation for the Tajima group was significantly less than the control value at 3 weeks. At 6 weeks, tendon excursion was significantly less than the control value for the 8-strand repair group. For intergroup comparison, the normalized (experimental/control [E/C]) data revealed no significant effects as a result of time passage or suture method for DIP rotation, PIP rotation, and FDP excursion. When we examined the structural properties, stiffness of all experimental values was significantly less than the control values for all groups at both time periods. When we compared the effects of the Tajima, Savage, and 8-strand repair groups on normalized structural properties, the stiffness of the 8-strand repair group was significantly greater than that of the Tajima and Savage groups (Fig. 6) at 3 and 6 weeks. In addition, the normalized (E/C) values for stiffness for the 6-week groups were significantly greater than those for the 3-week groups. Normalized stiffness also demonstrated a significant interaction between repair method and time interval. Ultimate load was significantly less than control for all suture methods at both 3 and 6 weeks (Table 1). The normalized values for ultimate load of the Savage and 8-strand repair groups were found to be significantly greater than values for the ultimate load of the Tajima group at each time interval (Fig. 7). In addition, the 8-strand repair group's ultimate load was significantly greater than that of the Savage

102

W i n t e r s et al. / T e n d o n s S t r o n g e r W i t h 8-Strand Suture

+1

+1

+1

+1

+1

group at 3 weeks (52.6 N) and 6 weeks (70.9 N). Overall, the 6-week values for normalized load were significantly greater than those obtained at 3 weeks.

Discussion

+1

+1

+1

+1

+1

ee~

oo

~

~

~1

+1

+1

+1

+1

+1

t'-: +1

~ +1

e-i +1

e.i +1

+1

oo

+1

+1

+1

+1

+1

t<

~

oo

~6

t'q

+1

+1

+[

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

~ +1

u4 +1

+1

,.~

~

t'q

+ 1 +~0 1

g

=

9~

9

M'~ +1

,,

~'o

m +1

~ ~

~ ' ~

~

.

~ =

Disappointed with the results seen with the application of the low repair-site loads brought about with traditional methods of protected passive motion, some authors have advocated techniques for increasing the in vivo forces applied after tendon repair. 7'1~ In support of an accelerated rehabilitation regimen, Kubota et al. found that the cumulative effects of motion and tension on the healing response of injured FDP tendons in chickens resulted in significantly greater tendon repair strength than occurred when either of these 2 parameters were applied alone. 4 In a clinical study, Small et al. reported 77% good to excellent results in a series of 98 patients treated with controlled active motion rehabilitation. Their reported rupture rate of approximately 10%, however, caused them to emphasize the necessity for identifying safe parameters for tendon gliding used during rehabilitation. 19 In an effort to define the in vivo forces seen clinically with passive and active motion and with pinch and grasp activities, Shuind et al. applied S-shaped buckle transducers to the flexor tendons of patients undergoing surgery for carpal tunnel syndrome.2~ The authors found a large variation in in vivo force, from a 0.1-kg (0.98-N) force with passive digital motion to a 3.5-kg (34.3-N) force with active digital flexion, which increased to 12-kg (117.6-N) force with tip pinch. The authors recommended that these data be used in designing rehabilitation programs with passive and active mobilization after surgery. There has been considerable concern in the past about the degree to which early motion leads to repair-site gap formation, inadequate healing, and ultimate tendon rupture, leading some to suggest that the interval of immobilization intervals be inc r e a s e d . 21-23 Recent s t u d i e s 8'9'24-26 have suggested that carefully applied early protected motion can produce a consistent range of gliding without leading to excessive repair site deformation, however. The data provided by our study indicate that while there was no statistically significant difference between the tested techniques with regard to tendon excursion and joint rotation in vivo, the newer multistrand flexor tendon repairs were significantly stronger than were the traditional 2-strand repairs, with the 8-strand technique having statistically significant greater strength than all of the techniques at both 3 weeks (52.6 N) and 6 weeks (70.9 N).

The Journal of Hand Surgery / Vol. 23A No. 1 January 1998

In respect to the "safe zones" defined by Shuind et al., all repairs exceeded the 0.1-kg (0.98-N) force required for passive digital motion, with the exception of the Kessler technique, for which a high failure rate was noted (33%). Only the 8-strand repair exceeded the 3.4-kg (34.3-N) force required for active digital motion. On the basis of the promising results noted in this experimental study, a clinical trial using the 8-strand technique with controlled early active motion has been initiated. It is interesting to consider the features provided by newer repair techniques that may have been responsible for the achievement of greater ultimate load and linear slope values than those provided by the 2-strand repairs at all time intervals in this experiment. An increased number of contact points and/or an increased number of strands extending across the repair site could have contributed to these results. The Kessler and Tajima techniques used 2 intratendinous grasping points and had 2 strands extending across the repair site. They differed primarily in suture knot placement. The increased number of tendon disruptions at 3 and 6 weeks for the Kessler repair's extrinsic knots suggests that this may be a factor in its poor performance when compared to the lower-profile Tajima repair. The modified Savage technique employed 12 grasping points and had 4 strands extending across the repair site. The 8-strand repair was designed to be easy to use and have a low-profile configuration (i.e., knot buried in the repair site) while employing a maximum number of strands extending across the repair site. Since there are no formal intratendinous locking grasps, its strength is derived from an effective anchoring double-box configuration with 8 suture strands and a double-suture intratendinous knot extending across the repair site. The necessity of performing the 8-strand repairs with double-stranded polyfilament caprolactam introduced an additional variable for comparison with the remaining 3 techniques, which used braided Dacron. In a previous experiment, however, investigators found that 4 - 0 Supramid had similar biomechanical properties to 4 - 0 braided Dacron in in )ivo canine flexor tendon repair models with no statistically significant differences demonstrated between the 2 materials' strength with regard to tendon rupture. While we conclude from the findings of this experiment that the improved strength characteristics of the 8-strand repair were due to the fact that an increased number of strands extended across the repair site, the suture material that we used for this

103

repair might have had a positive effect on the values that were obtained. Attempts to optimize the postoperative regimen for flexor tendon repair have been essentially empiric. Both the time and graduation of the exercise regimen have lacked clear, conceptual guidelines. Recent clinical and experimental studies have demonstrated that higher levels of proximal musculotendinous load and greater magnitudes of tendon repair-site excursion significantly improve the healing of repaired intrasynovial flexor tendons. The observation made in this experiment, that the low-profile 8-strand suture method significantly improves the in vivo strength characteristics of repaired tendons obtained at 3 and 6 weeks, supports the concept that higher levels of in vivo load should be well tolerated during the early stages of tendon repair rehabilitation. We gratefully acknowledge the technical assistance of Dawn M. Tramaglini, MS.

References 1. Hitchcock TF, Light TR, Bunch WK et al. The effect of immediate constrained digital motion on the strength of flexor tendon repairs in chickens. J Hand Surg 1987;12A:590-595. 2. Strickland JW, Glogovac SV. Digital function following flexor tendon repair in zone II: a comparison of immobilization and controlled passive motion techniques. J Hand Surg 1980;5:537-543. 3. Feehan LM, Beauchene JG. Early tensile properties of healing chicken flexor tendons: early controlled passive motion versus postoperative immobilization. J Hand Surg t990; 15A:63-68. 4. Kubota H, Manske PR, Aoki M, Pruitt DL, Larson BJ. Effect of motion and tension on injured flexor tendons in chickens. J Hand Surg 1996;21A:456-464. 5. Becket H, Davidoff M. Eliminating the gap in flexor tendon surgery. Hand 1977;9:306-311. 6. Hagberg L, Selvik G. Tendon excursion and dehiscence during early controlled mobilization after flexor tendon repair in zone II: an x-ray stereophotogrammetric analysis. J Hand Surg 1991;16A:669-680. 7. Lee H. Double loop locking suture: a t~hnique of tendon repair for early active mobilization.Part I: evolution of technique and experimental study. J Hand Surg 1990;15A:945-952. 8. Lin GT, An KN, Amadio PC, Cooney WP. Studies of running suture for flexor tendon repair in dogs. J Hand Surg 1988;13A:553-558. 9. Mashadi ZB, Amis AA. Strength of the suture in the epitenon and within the tendon fibers: development of stronger peripheral suture technique. J Hand Surg 1992; 17B: 172-175. 10. Savage R, In vitro studies of a new method of flexor tendon repair. J Hand Surg 1985;10B:135-141. 11. Wade PJ, Muir IF, Hutcheon LL. Primary flexor tendon repair: the mechanical limitations of the modified Kessler technique. J Hand Surg 1986;11B:71-76.

104

Winters et al. / Tendons Stronger With 8-Strand Suture

12. Tajima T. History, current status, and aspects of hand surgery in Japan. Clin Orthop 1984;184:41-49. 13. Kessler I, Nissim F. Primary repair without immobilization of flexor tendon division within the tendon sheath: an experimental and clinical study. Acta Orthop Scand 1969; 40:587-601. 14. Luna LG. Pathology. In: Manual of histological staining methods of the Armed Forces Institutes of Pathology. New York: McGraw-Hill, 1960:72-100. 15. Noguchi M, Seiler JG HI, Gelberman RG, Sofranko RA, Woo SL-Y. In vitro biomechanical analysis of suture methods for flexor tendon repair. J Orthop Res 1993;11:603-611. 16. Spiegelman JJ, Woo SL-Y. A rigid-body method for finding centers of rotation and angular displacement of planar joint motion. J Biomech 1987;20:715-721. 17. Woo SL-Y, Gelberman RH, Cobb NG, Amiel D, Lothringer K, Akeson WH. The importance of controlled passive mobilization on flexor tendon healing: a biomechanical study. Acta Orthop Scand 1981;16:615-622. 18. Silverskiold KL, May EJ. Gap formation after flexor tendon repair in zone II. Scand J Plast Reconstr Surg 1993; 27:263-268.

19. Small JO, Brennan MD, Colville J. Early active mobilization following flexor tendon repair in zone 2. J Hand Surg 1989;14B:383-391. 20. Schuind F, Garcia-Elias M, Cooney WP, An K-N. Flexor tendon forces: in vivo measurements. J Hand Surg 1992; 17A:291-298. 21. Ketchum LD, Martin NL, Kappel DA. Experimental evaluation of factors affecting the strength of tendon repairs. Plast Reconstr Surg 1977;59:708-719. 22. Lindsay WK, Thompson HG, Walker FG. Digital flexor tendons: an experimental study. Br J Plast Surg 1960; 13:1-9. 23. Seradge H. Elongation of the repair configuration following flexor tendon repair. J Hand Surg 1983;8:182-185. 24. Silverskiold KL, Andersson CH. Two new methods of tendon repair: an in vitro evaluation of tensile strength and gap formation. J Hand Surg 1993;18A:58-65. 25. Becker H, Davidoff M. Eliminating the gap in flexor tendon surgery. Hand 1977;9:306-311. 26. Urbaniak JR, Cahill J, Mortenson R. Tendon suturing methods: analysis of tensile strengths. In: AAOS symposium on tendon surgery in the hand. St. Louis: CV Mosby, 1975:70- 80.