- Email: [email protected]
Repair of Zone II Flexor Digitorum Profundus Lacerations Using Varying Suture Sizes: A Comparative Biomechanical Study
Repair of Zone II Flexor Digitorum Profundus Lacerations Using Varying Suture Sizes: A Comparative Biomechanical Study George Alavanja, MD, Chesterton, IN, Elizabeth Dailey, BS, Daniel P. Mass, MD, Chicago, IL
Purpose: To compare the maximum tensile load, change in work of flexion, and gapping at the repair site after zone II flexor digitorum profundus tendon repairs using 2-0, 3-0, and 4-0 braided polyester 4-strand locked cruciate repair technique in fresh-frozen cadaveric hands with standard 6-0 suture epitenon repairs, to determine which suture size is the best for a core repair. Methods: A randomized study was designed using 41 tendons from 15 fresh-frozen cadaveric hands. We included only the flexor digitorum profundus tendons from the index, middle, and ring fingers to minimize variation between digits. Core suture size was randomized for each finger. A sharp laceration through the flexor digitorum profundus in zone II was made and a 4-strand locked cruciate braided polyester core stitch was performed along with a locked epitenon stitch. Cyclic loading was performed for 1,000 cycles. For each tendon the mean work of flexion (before/after zone II repair) and maximum tensile load were measured using a custom-designed tensiometer, as was gapping before maximum tensile loading. Results: Mean gaps after 1,000 load-unload cycles to 3.9 N of pulp pinch did not approach the clinically significant limit of 3 mm in each group. By using a regression model, we found that the prerepair and postrepair comparisons for mean work of flexion to a 3.9-N pulp pinch showed the greatest change in work of flexion for 2-0 braided polyester. Statistical significance was found between 2-0 braided polyester and 3-0 or 4-0 braided polyester; however, the work of flexion between the 3-0 and 4-0 polyester was not clinically significant. The highest maximum tensile load was obtained with suture size 2-0 braided polyester. The maximal tensile load statistically showed 2-0 braided polyester to be stronger than 4-0 braided polyester but we found no statistically significant difference between 3-0 and 2-0 braided polyester or between 3-0 and 4-0 braided polyester. Conclusions: In this cadaveric study we found that increasing locking cruciate suture caliber from 4-0 to 2-0 increased maximum tensile strength but also caused increased work of flexion. Gapping was not affected by suture caliber. There was no significant difference in strength or mean change in work of flexion between 3-0 or 4-0 braided polyester sutures. (J Hand Surg 2005;30A:448-454. Copyright © 2005 by the American Society for Surgery of the Hand.) Key words: Flexor tendon, tendon repair, 4-strand repair, in situ testing, zone II injury.
From the Lakeshore Bone and Joint Institute, Valparaiso Orthopedic Clinic, Inc., Chesterton, IN; and the Section of Orthopaedic Surgery and Rehabilitation Medicine, Department of Surgery, University of Chicago Hospitals, Chicago, IL. Received for publication March 19, 2004; accepted in revised form February 10, 2005. 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. Corresponding author: Daniel P. Mass, MD, Section of Orthopaedic Surgery and Rehabilitation Medicine, Department of Surgery, University of Chicago Hospitals, 5841 S Maryland Ave, MC 3079, Chicago, IL 60637; e-mail: [email protected]. Copyright © 2005 by the American Society for Surgery of the Hand 0363-5023/05/30A03-0004$30.00/0 doi:10.1016/j.jhsa.2005.02.008
448
The Journal of Hand Surgery
Alavanja, Dailey, and Mass / Four-Strand Flexor Tendon Repair
Repair of flexor tendons has undergone a notable evolution over the past 50 years. Early primary repair of flexor tendons in no man’s land has replaced tendon grafting as the standard of care.1 Rehabilitation after repair of flexor tendon injuries also has evolved from delayed motion to early passive motion and now early active motion.2–5 Some have advocated using large-caliber sutures6 – 8 (2-0 and 3-0) for their repair strength but some of these studies were performed in nonanatomic (linear) models, which do not account for the friction of the flexor pulley system. Comparing strength and the work of flexion (WoF) of various suture sizes in an anatomic model9 –13 would determine the optimum suture size for balance between strength and gliding resistance at the repair sites within zone II. Based on the work of Choueka et al10 and Angeles et al14 4-0 braided polyester core repairs with either the modified Becker or the locked cruciate repair are strong enough for a potential active postoperative motion rehabilitation program. Others have advocated larger core sutures for this rehabilitation program. We hypothesized that by studying the locked cruciate core repair shown by Angeles et al14 to be the best core repair and by testing the same repair with different core sutures— 4-0, 3-0, and 2-0 —we could determine the best suture size based on a combination of repair strength, gapping, and WoF. We randomly repaired 41 flexor digitorum profundus (FDP) tendons with 2-0, 3-0, and 4-0 braided polyester core sutures with a 6-0 locked epitenon suture (Dermalon, Excel Ethicon, Inc, Somerville, NJ) and compared their maximum strength, WoF, and gap after 1,000 cycles of pulling to 3.9 N.
Materials and Methods Fifteen fresh-frozen cadaveric hands (41 tendons) were amputated approximately 3 cm proximal to the radiocarpal joint. Four tendons were not included in the study because of failure during cycling. Two 4-0 and one 2-0 repair ruptured and one 3-0 repair pulled out of the tendons during cycling. Each hand was allowed to thaw overnight or in a warm water bath while in a plastic bag. We included only the FDP tendons from the index, middle, and ring fingers to minimize variation between digits. By using a Bruner incision we inspected the flexor tendons for abnormalities between the A2 and A4 pulleys and closed the incision by using a 4-0 suture (Monosof, United States Surgical, Norwalk, CT). The flexor digitorum superficialis and FDP tendons for the index, middle,
449
and ring fingers were tagged proximal to the carpal canal with a Krakow-type stitch using 0 sutures (Ethibond; Excel Ethicon, Inc., Somerville, NJ) after dividing the interconnecting tissues between the tendons to allow for independent tendon gliding. We placed a 1.1-mm (.045 in) K-wire transversely through the base of the metacarpals, avoiding damage to the flexor and extensor tendons to allow the hand to be immobilized on the sliding platform (Fig. 1). Based on random assignments for each finger we placed variously sized 4-strand locked cruciate core sutures (Fig. 2) using braided polyester with a 6-0 locked epitenon stitch (Dermalon). We performed the repair under the A-2 pulley by approaching the tendon between the A-2 and A-4 pulleys while attempting to preserve the A-3 pulley when possible. The skin was closed with a 4-0 suture (Monosof) after completing the repairs. There were 14 tendons repaired with 2-0, 14 with 3-0, and 13 with 4-0 sutures.
Tendon Loading The hand was mounted onto a specially designed tensiometer that simulated anatomic digital motion (Fig. 1). To monitor internal tendon loads the proximally tagged portion of the flexor tendon being tested was attached by a nonslip clamp to a force transducer (Lucas Schaevitz, Pennsauken, NJ). To measure displacement of the tendon a linear variable differential transducer (Lucas Schaevitz) was attached to the motorized platform. To measure the pulp pinch we used a digital pinch meter (Greenleaf Medical Systems, Palo Alto, CA) attached to an adjustable bracket that was supported 1 cm above the distal palmar crease. When the fingertip of the digit being tested rested on the pinch meter, the position approximated the digital joint angles in the activehold component of the minimal active muscle tension technique of Evans and Thompson3 for active motion immediately after flexor tendon repair. The corresponding flexor digitorum superficialis tendon had been identified and loaded with a 20-g weight, which allowed free gliding between the flexor tendons. A 200-g weight was attached to size 0 sutures (Ethibond) that were attached to the extensors to simulate the physiologic resistance of the extensor tendons, which in turn allowed the digit to return to full extension at the end of each cycle. By using custom software we programmed the tensiometer for a single pull (eg, to failure or a predetermined pinch force) or cyclic loading. Before any testing run the tendon was preloaded until a tendon force load of 12 N was
450
The Journal of Hand Surgery / Vol. 30A No. 3 May 2005
Figure 1. (A) Hand mounted on a dynamic testing apparatus. (B) Cadaveric hand mounted on a dynamic testing apparatus.
generated by producing a force of 3.9 N at the pinch meter. This is greater than the force of 9 N required for active motion against mild resistance.15 The force was detected by the transducer to take up the slack and this was set as the 0 point for force and displacement measurements on the computer. For prerepair WoF measurements we programmed the tensiometer to apply tension on the FDP tendon at 10 cm/min from full extension to flexion until a force of 3.9 N was measured at the pinch meter. The WoF needed to generate the 3.9-N pulp pinch was calculated and recorded by a computer. Three trials were performed and then an average was computed for each tendon. Once the prerepair WoF measurements were completed the incisions were reopened and the flexor tendon was exposed to allow for laceration and repair
between the A2 and A4 pulleys using variously sized 4-strand locked cruciate braided polyester sutures with a running locked 6-0 epitenon repair (Dermalon). As the tendon was loaded to a pinch force of 3.9 N the WoF was measured and the average for 3 trials was calculated. To determine the effect of core suture placement on gliding function we computed the change in WoF (WoF after repair – WoF before repair) for each tendon.
Gapping at Repair Sites After the WoF measurements were completed we programmed our loading apparatus to run 4 sets of 250 cycles for each tendon to evaluate gap resistance. A preprogrammed 3.9-N force generated at the pinch meter would allow the digit to return to the starting
Alavanja, Dailey, and Mass / Four-Strand Flexor Tendon Repair
451
Results Change in WoF
Figure 2. Locked cruciate repair.
position of full extension to complete 1 cycle. At the end of each 250-cycle set the finger was placed in full extension, the skin incision was reopened, and a 12-N preload tension was maintained across the FDP tendon during measurements.15 A single observer measured the gap between the tendon ends by using an electronic caliper. Three measurements (2 at the lateral aspect, 1 at the midpoint of the laceration’s transverse diameter) were taken and the average between them was obtained. The mean ⫹ SD gap measurements at the end of each 250-cycle set were calculated.
Ultimate Tensile Strength at Failure The tensiometer was set up for a single pull to failure resulting in suture rupture or pullout at a rate of 10 cm/min. To determine the ultimate tensile strength of a repair the maximum tension force at failure was recorded and stored in a computer. The mean ⫹ SD was computed for each suture technique.
Data Collection and Analysis The raw analog data from the force transducer, linear variable differential transducer, and pinch meter were sampled continuously at 100 Hz by an analogto-digital converter (ADC488/16; IO Tech, Cleveland, OH). An interface unit (MacSCSI 488; IO Tech) allowed us to store digital data on a computer (Apple Power Macintosh 8600, Apple Inc, Cupertino, CA) for analysis using custom software. A power analysis was performed with preliminary data, which showed that there would be 85% power (2sided ␣ ⫽ 0.05) with the current number of tendons. Analysis of variance was used to compare the change in WoF, maximum tensile load, and maximum pinch force among the 3 suture sizes. Then a regression model was fit to the data to compare suture sizes (reference group was suture size 2-0) while adjusting for different fingers and taking into account the correlation between fingers on the same hand. The interaction between suture size and finger also was included in all models.
The average baseline WoF with the 4-0 core suture fingers was 1.86 ⫾ 0.066, with the 3-0 core suture fingers it was 2.04 ⫾ 0.060, and with the 2-0 core suture fingers it was 1.83 ⫾ 0.078. The baseline WoF was not statistically different among the 3 groups (p ⫽ .88). There was a significant difference in the mean change in WoF between the 3 suture techniques (p ⫽ .005). A 2-0 suture repair with a mean change in WoF of 0.51 ⫾ 0.27 created significantly more WoF than a 3-0 suture with a mean change of 0.34 ⫾ 0.16 (p ⫽ .025) and than a 4-0 suture with a mean change of 0.31 ⫾ 0.11 (p ⫽ .007), but a 3-0 suture did not create significantly more WoF than a 4-0 suture (p ⫽ .70). Therefore, the 2-0 suture had a significantly greater change in WoF or gliding resistance than the 4-0 and 3-0 but there was no statistical difference between the 3-0 and 4-0 sutures (Fig. 3).
Maximum Tensile Load The highest maximum tensile load was obtained with 2-0 braided polyester sutures. Its mean maximum tensile load was 80 ⫾ 12 N whereas 3-0 and 4-0 braided polyester sutures had maximum tensile loads of 74 ⫾ 18 N and 66 ⫾ 12 N, respectively. The 3-0 polyester suture is on average 12% stronger than the 4-0 suture and the 2-0 polyester suture is on average 21% stronger than the 4-0 suture and 8% stronger than 3-0 suture. Results indicate that there was a significant difference in the maximum tensile load among the 3 suture sizes (p ⫽ .024). The regression model indicated that 2-0 braided polyester sutures were significantly different from 4-0 braided polyester sutures (p ⫽ .004) but that the difference was not significant between suture sizes 2-0 and 3-0 (p ⫽ .32) or between sizes 3-0 and 4-0 (p ⫽ .15). Therefore maximal tensile load statistically showed 2-0 braided polyester sutures to be stronger than 4-0 braided
Figure 3. Mean change of WoF. Error bars indicate SD.
452
The Journal of Hand Surgery / Vol. 30A No. 3 May 2005
Figure 4. Maximum tensile load in newtons. Error bars indicate SD.
polyester sutures but no statistically significant differences between 3-0 and 2-0 braided polyester sutures or between 3-0 and 4-0 braided polyester sutures were found (Fig. 4).
Gapping Gapping at the repairs was measured as described earlier. For 2-0 braided polyester sutures the average gapping measured was 0.5 ⫾ 0.6 mm and for 3-0 braided polyester sutures it was 0.4 ⫾ 0.8 mm. For the 4-0 sutures the gapping average was 0.3 ⫾ 0.5 mm. None of the repairs approached the clinically significant limit of 3 mm.16
Discussion The evolution of flexor tendon repair has progressed dramatically over the past 50 years.1 The pendulum has shifted to early primary repair to allow early passive- or even active-motion rehabilitation protocols that can improve the tendon nutrition, healing, and remodeling at the repair site.17–21 At the same time early motion can increase gapping at the repair site, which in turn increases the adhesion rate and need for tenolysis,16 resulting in a poor outcome. Because early active rehabilitation protocols were thought to lead to tendon repair rupture larger-caliber core sutures6 – 8 or more than 4-strand sutures22–25 have been developed and advocated. In the report by Choueka et al10 a literature review showed that the most strength lost was at day 5 during flexor tendon repairs and was 60%; therefore the strength of the repair only needed to be 3 times the strength needed to move the tendon without resistance (3 ⫻ 19 N), not 5 times as stated by Savage and Risitano.5 Strickland1 reviewed the forces needed for flexion and stated that for passive motion the tendon saw 2 to 4 N, for mild resistance up to 10 N, and for moderate resistance up to 17 N. It is only with strong composite grasp that significant force (70
N) is needed. Combining the review by Choueka et al10 and Strickland’s1 results we conclude that the required tendon force to flex a finger is between 30 and 51 N if one assumes that postoperative swelling changes active unresisted resistance to a moderate resistance. Taras et al7 studied the strength of progressively increasing suture sizes in the core repair and concluded that the strength of the repair is related directly to the size of the suture, with 2-0 being the strongest. The 2 problems with this work are (1) that the repair technique used was a 2-strand core repair, which cannot reach a strength of 30 to 50 N and (2) that they tested the strength to failure by linear distraction, which does not allow for the measurement of friction or the WoF, which could change with the increasing volume of the repair as the core suture gets larger. The article by Barrie et al8 showed the increasing strength of the 3-0 over the 4-0 suture with the cruciate repair tested in linear loading but this article did not address the 2-0 size as was recommended by Taras et al7 nor did it test the WoF. Hatanaka and Manske26 showed that the size of the core suture affected the ultimate strength and the gap strength in a linear distraction model. They found that larger core sutures were stronger and that locking repairs were stronger than grasping repairs. They did not evaluate the WoF, which could eliminate the larger core sutures, because they used a linear model for testing. In another article Aoki et al13 looked at WoF for multiple repair techniques, detailing the importance of looking at WoF for the best repair. A linear model takes the tendon from its surrounding anatomic structures (pulleys) and stresses it linearly. An anatomic model takes into account the confining aspects of the pulley system that restrict the amount of material placed at a repair site, which in turn affects gliding resistance. This model allows for more physiologic loading and testing, including looking at WoF.9 –13 Therefore we chose to use a curvilinear anatomic model to assess the locked cruciate core suture27 using braided polyester with a running locked epitenon stitch (Dermalon) because of its ease of placement, less interference (1 knot), and more than adequate strength needed for early active rehabilitation protocols.14 Each tendon in this study was cycled 1,000 times to a pinch force of 3.9 N (0.4 kg), far beyond what is needed to allow an early active rehabilitation program.15 In the previous study by Angeles et al14 not only did 2 different 4-0 braided polyester repair techniques yield more than 60 N of force but the locked cruciate was easier to
Alavanja, Dailey, and Mass / Four-Strand Flexor Tendon Repair
perform than the modified Becker and the WoF was significantly lower. They recommended this technique for flexor tendon repairs using early, unresisted, protected active-motion protocols. We performed this study to answer the question asked by Taras et al7: if 4-0 is good, should we be using 3-0 or even 2-0 locked cruciate core sutures to protect our repair better? In our study we were able to show that the strength advantage relates to 2-0 braided polyester over 4-0 braided polyester sutures, as all the other researchers have shown. This suture size could protect an aggressive postoperative early, unresisted, protected activemotion treatment protocol while avoiding rupture. Its higher WoF or gliding resistance, however, could preclude its use in the clinical setting. The 3-0 and 4-0 braided polyester suture10,14 strength and gliding function were not statistically different, with 3-0 being slightly favored in strength and 4-0 in WoF. They both were adequate to allow early protective active rehabilitation protocols. We felt obligated to study gapping because as the suture gets smaller there may be more stretch or elasticity in the repair; this is best tested with cycling. Gapping was found to be minimal among the 3 suture sizes after the cyclic loads, which supports gapping as a measure of suture configuration and not caliber. We were not measuring different suture configurations, however, so the gapping may show that the smaller caliber suture does not cut out of the tendon or stretch more, which also would be important to know. The weakness of this study is that it was in vitro and no healing took place, so that the conclusion of the strength of the needed repairs is a guess that is consistent with the literature. We also do not know how to factor in the increased resistance from edema after surgery. The 9 N needed for active motion against mild resistance is for normal tendons but is not necessarily the force needed to move repaired tendons or tendons with edema. We took this into consideration by using previous animal experiments1,10 and by requiring at least 3 times the repair strength for moderate resistance for a successful repair. Other weaknesses included dividing the interconnecting tissue in the carpal canal in individualizing each FDP tendon; this could cause an underestimate of the proximal resistance and increase the potential pull by the muscles. We conclude that 3-0 or 4-0 braided polyester locked cruciate core suture repairs of FDP tendons with a 6-0 locked epitenon repair
453
(Dermalon) should be strong enough for protected active postoperative rehabilitation. To confirm this proposal an in vivo study in the clinical setting is needed. The authors thank Kristen Kasza, MS, for her insights and assistance with the statistical analysis.
References 1. Strickland JW. Development of flexor tendon surgery: twenty-five years of progress. J Hand Surg 2000;25A:214 –235. 2. Cullen KW, Tolhurst P, Land D, Page RE. Flexor tendon repair in zone 2 followed by controlled active mobilisation. J Hand Surg 1989;14B:392–395. 3. Evans RB, Thompson DE. The application of force to the healing tendon. J Hand Ther 1993;6:266 –284. 4. Gratton P. Early active mobilization after flexor tendon repairs. J Hand Ther 1993;6:285–289. 5. Savage R, Risitano G. Flexor tendon repair using a “six strand” method of repair and early active mobilisation. J Hand Surg 1989;14B:396 –399. 6. Hatanaka H, Kojima T, Mizoguchi T, Ueshin Y. Aggressive active mobilization following zone II flexor tendon repair using a two-strand heavy-gauge locking loop technique. J Orthop Sci 2002;7:457– 461. 7. Taras JS, Raphael JS, Marczyk SC, Bauerle WB. Evaluation of suture caliber in flexor tendon repair. J Hand Surg 2001; 26A:1100 –1104. 8. Barrie KA, Tomak SL, Cholewicki J, Merrell GA, Wolfe SW. Effect of suture locking and suture caliber on fatigue strength of flexor tendon repairs. J Hand Surg 2001;26A: 340 –346. 9. Greenwald DP, Randolph MA, Hong H-Z, May JW. Augmented Becker versus modified Kessler tenorrhaphy in monkeys: dynamic mechanical analysis. J Hand Surg 1995; 20A:267–272. 10. Choueka J, Heminger H, Mass DP. Cyclical testing of zone II flexor tendon repairs. J Hand Surg 2000;25A:1127–1134. 11. Thurman RT, Trumble TE, Hanel DP, Tencer AF, Kiser PK. Two-, four-, and six-strand zone II flexor tendon repairs: an in situ biomechanical comparison using a cadaver model. J Hand Surg 1998;23A:261–265. 12. Komanduri M, Phillips CS, Mass DP. Tensile strength of flexor tendon repairs in a dynamic cadaver model. J Hand Surg 1996;21A:605– 611. 13. Aoki M, Manske PR, Pruitt DL, Larson BJ. Work of flexion after tendon repair with various suture methods. A human cadaveric study. J Hand Surg 1995;20B:310 –313. 14. Angeles JG, Heminger H, Mass DP. Comparative biomechanical performances of 4-strand core suture repairs for zone II flexor tendon lacerations. J Hand Surg 2002;27A: 508 –517. 15. Urbaniak JR, Cahill JD, Mortenson RA. Tendon suturing methods: analysis of tensile strengths. In: Hunter JM, Scheneider LHY, eds. Symposium on tendon surgery. St. Louis, MO: Mosby, 1975:70 – 80. 16. Seradge H. Elongation of the repair configuration following flexor tendon repair. J Hand Surg 1983;8:182–185. 17. Hitchcock TF, Light TR, Bunch WH, Knight GW, Sartori MJ, Patwardhan AG, et al. The effect of immediate con-
454
18.
19. 20.
21.
22.
23.
The Journal of Hand Surgery / Vol. 30A No. 3 May 2005 strained digital motion on the strength of flexor tendon repairs in chickens. J Hand Surg 1987;12A:590 –595. Gelberman RH, Woo SL, Lothringer K, Akeson WH, Amiel D. Effects of early intermittent passive mobilization on healing canine flexor tendons. J Hand Surg 1982;7:170 –175. Gelberman RH, Manske PR. Factors influencing flexor tendon adhesions. Hand Clin 1985;1:35– 42. Gelberman RH, Botte MJ, Spiegelman JJ, Akeson WH. The excursion and deformation of repaired flexor tendons treated with protected early motion. J Hand Surg 1986;11A:106 –110. Aoki M, Kubota H, Pruitt DL, Manske PR. Biomechanical and histologic characteristics of canine flexor tendon repair using early postoperative mobilization. J Hand Surg 1997; 22A:107–114. Xie RG, Zhang S, Tang JB, Chen F. Biomechanical studies of 3 different 6-strand flexor tendon repair techniques. J Hand Surg 2002;27A:621– 627. Sanders DW, Milne AD, Johnson JA, Dunning CE, Richards
24.
25.
26.
27.
RS, King GJW. The effect of flexor tendon repair bulk on tendon gliding during simulated active motion: an in vitro comparison of two-strand and six-strand techniques. J Hand Surg 2001;26A:833– 840. Barrie KA, Wolfe SW, Shean C, Shenbagamurthi D, Slade JF 3rd, Panjabi MM. A biomechanical comparison of multistrand flexor tendon repairs using an in situ testing model. J Hand Surg 2000;25A:499 –506. Barrie KA, Tomak SL, Cholewicki J, Wolfe SW. The role of multiple strands and locking sutures on gap formation of flexor tendon repairs during cyclical loading. J Hand Surg 2000;25A:714 –720. Hatanaka H, Manske PR. Effect of suture size on locking and grasping flexor tendon repair techniques. Clin Orthop 2000,375:267–274. McLarney E, Hoffman H, Wolfe SW. Biomechanical analysis of the cruciate four-strand flexor tendon repair. J Hand Surg 1999;24A:295–301.
Purpose: To compare the maximum tensile load, change in work of flexion, and gapping at the repair site after zone II flexor digitorum profundus tendon repairs using 2-0, 3-0, and 4-0 braided polyester 4-strand locked cruciate repair technique in fresh-frozen cadaveric hands with standard 6-0 suture epitenon repairs, to determine which suture size is the best for a core repair. Methods: A randomized study was designed using 41 tendons from 15 fresh-frozen cadaveric hands. We included only the flexor digitorum profundus tendons from the index, middle, and ring fingers to minimize variation between digits. Core suture size was randomized for each finger. A sharp laceration through the flexor digitorum profundus in zone II was made and a 4-strand locked cruciate braided polyester core stitch was performed along with a locked epitenon stitch. Cyclic loading was performed for 1,000 cycles. For each tendon the mean work of flexion (before/after zone II repair) and maximum tensile load were measured using a custom-designed tensiometer, as was gapping before maximum tensile loading. Results: Mean gaps after 1,000 load-unload cycles to 3.9 N of pulp pinch did not approach the clinically significant limit of 3 mm in each group. By using a regression model, we found that the prerepair and postrepair comparisons for mean work of flexion to a 3.9-N pulp pinch showed the greatest change in work of flexion for 2-0 braided polyester. Statistical significance was found between 2-0 braided polyester and 3-0 or 4-0 braided polyester; however, the work of flexion between the 3-0 and 4-0 polyester was not clinically significant. The highest maximum tensile load was obtained with suture size 2-0 braided polyester. The maximal tensile load statistically showed 2-0 braided polyester to be stronger than 4-0 braided polyester but we found no statistically significant difference between 3-0 and 2-0 braided polyester or between 3-0 and 4-0 braided polyester. Conclusions: In this cadaveric study we found that increasing locking cruciate suture caliber from 4-0 to 2-0 increased maximum tensile strength but also caused increased work of flexion. Gapping was not affected by suture caliber. There was no significant difference in strength or mean change in work of flexion between 3-0 or 4-0 braided polyester sutures. (J Hand Surg 2005;30A:448-454. Copyright © 2005 by the American Society for Surgery of the Hand.) Key words: Flexor tendon, tendon repair, 4-strand repair, in situ testing, zone II injury.
From the Lakeshore Bone and Joint Institute, Valparaiso Orthopedic Clinic, Inc., Chesterton, IN; and the Section of Orthopaedic Surgery and Rehabilitation Medicine, Department of Surgery, University of Chicago Hospitals, Chicago, IL. Received for publication March 19, 2004; accepted in revised form February 10, 2005. 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. Corresponding author: Daniel P. Mass, MD, Section of Orthopaedic Surgery and Rehabilitation Medicine, Department of Surgery, University of Chicago Hospitals, 5841 S Maryland Ave, MC 3079, Chicago, IL 60637; e-mail: [email protected]. Copyright © 2005 by the American Society for Surgery of the Hand 0363-5023/05/30A03-0004$30.00/0 doi:10.1016/j.jhsa.2005.02.008
448
The Journal of Hand Surgery
Alavanja, Dailey, and Mass / Four-Strand Flexor Tendon Repair
Repair of flexor tendons has undergone a notable evolution over the past 50 years. Early primary repair of flexor tendons in no man’s land has replaced tendon grafting as the standard of care.1 Rehabilitation after repair of flexor tendon injuries also has evolved from delayed motion to early passive motion and now early active motion.2–5 Some have advocated using large-caliber sutures6 – 8 (2-0 and 3-0) for their repair strength but some of these studies were performed in nonanatomic (linear) models, which do not account for the friction of the flexor pulley system. Comparing strength and the work of flexion (WoF) of various suture sizes in an anatomic model9 –13 would determine the optimum suture size for balance between strength and gliding resistance at the repair sites within zone II. Based on the work of Choueka et al10 and Angeles et al14 4-0 braided polyester core repairs with either the modified Becker or the locked cruciate repair are strong enough for a potential active postoperative motion rehabilitation program. Others have advocated larger core sutures for this rehabilitation program. We hypothesized that by studying the locked cruciate core repair shown by Angeles et al14 to be the best core repair and by testing the same repair with different core sutures— 4-0, 3-0, and 2-0 —we could determine the best suture size based on a combination of repair strength, gapping, and WoF. We randomly repaired 41 flexor digitorum profundus (FDP) tendons with 2-0, 3-0, and 4-0 braided polyester core sutures with a 6-0 locked epitenon suture (Dermalon, Excel Ethicon, Inc, Somerville, NJ) and compared their maximum strength, WoF, and gap after 1,000 cycles of pulling to 3.9 N.
Materials and Methods Fifteen fresh-frozen cadaveric hands (41 tendons) were amputated approximately 3 cm proximal to the radiocarpal joint. Four tendons were not included in the study because of failure during cycling. Two 4-0 and one 2-0 repair ruptured and one 3-0 repair pulled out of the tendons during cycling. Each hand was allowed to thaw overnight or in a warm water bath while in a plastic bag. We included only the FDP tendons from the index, middle, and ring fingers to minimize variation between digits. By using a Bruner incision we inspected the flexor tendons for abnormalities between the A2 and A4 pulleys and closed the incision by using a 4-0 suture (Monosof, United States Surgical, Norwalk, CT). The flexor digitorum superficialis and FDP tendons for the index, middle,
449
and ring fingers were tagged proximal to the carpal canal with a Krakow-type stitch using 0 sutures (Ethibond; Excel Ethicon, Inc., Somerville, NJ) after dividing the interconnecting tissues between the tendons to allow for independent tendon gliding. We placed a 1.1-mm (.045 in) K-wire transversely through the base of the metacarpals, avoiding damage to the flexor and extensor tendons to allow the hand to be immobilized on the sliding platform (Fig. 1). Based on random assignments for each finger we placed variously sized 4-strand locked cruciate core sutures (Fig. 2) using braided polyester with a 6-0 locked epitenon stitch (Dermalon). We performed the repair under the A-2 pulley by approaching the tendon between the A-2 and A-4 pulleys while attempting to preserve the A-3 pulley when possible. The skin was closed with a 4-0 suture (Monosof) after completing the repairs. There were 14 tendons repaired with 2-0, 14 with 3-0, and 13 with 4-0 sutures.
Tendon Loading The hand was mounted onto a specially designed tensiometer that simulated anatomic digital motion (Fig. 1). To monitor internal tendon loads the proximally tagged portion of the flexor tendon being tested was attached by a nonslip clamp to a force transducer (Lucas Schaevitz, Pennsauken, NJ). To measure displacement of the tendon a linear variable differential transducer (Lucas Schaevitz) was attached to the motorized platform. To measure the pulp pinch we used a digital pinch meter (Greenleaf Medical Systems, Palo Alto, CA) attached to an adjustable bracket that was supported 1 cm above the distal palmar crease. When the fingertip of the digit being tested rested on the pinch meter, the position approximated the digital joint angles in the activehold component of the minimal active muscle tension technique of Evans and Thompson3 for active motion immediately after flexor tendon repair. The corresponding flexor digitorum superficialis tendon had been identified and loaded with a 20-g weight, which allowed free gliding between the flexor tendons. A 200-g weight was attached to size 0 sutures (Ethibond) that were attached to the extensors to simulate the physiologic resistance of the extensor tendons, which in turn allowed the digit to return to full extension at the end of each cycle. By using custom software we programmed the tensiometer for a single pull (eg, to failure or a predetermined pinch force) or cyclic loading. Before any testing run the tendon was preloaded until a tendon force load of 12 N was
450
The Journal of Hand Surgery / Vol. 30A No. 3 May 2005
Figure 1. (A) Hand mounted on a dynamic testing apparatus. (B) Cadaveric hand mounted on a dynamic testing apparatus.
generated by producing a force of 3.9 N at the pinch meter. This is greater than the force of 9 N required for active motion against mild resistance.15 The force was detected by the transducer to take up the slack and this was set as the 0 point for force and displacement measurements on the computer. For prerepair WoF measurements we programmed the tensiometer to apply tension on the FDP tendon at 10 cm/min from full extension to flexion until a force of 3.9 N was measured at the pinch meter. The WoF needed to generate the 3.9-N pulp pinch was calculated and recorded by a computer. Three trials were performed and then an average was computed for each tendon. Once the prerepair WoF measurements were completed the incisions were reopened and the flexor tendon was exposed to allow for laceration and repair
between the A2 and A4 pulleys using variously sized 4-strand locked cruciate braided polyester sutures with a running locked 6-0 epitenon repair (Dermalon). As the tendon was loaded to a pinch force of 3.9 N the WoF was measured and the average for 3 trials was calculated. To determine the effect of core suture placement on gliding function we computed the change in WoF (WoF after repair – WoF before repair) for each tendon.
Gapping at Repair Sites After the WoF measurements were completed we programmed our loading apparatus to run 4 sets of 250 cycles for each tendon to evaluate gap resistance. A preprogrammed 3.9-N force generated at the pinch meter would allow the digit to return to the starting
Alavanja, Dailey, and Mass / Four-Strand Flexor Tendon Repair
451
Results Change in WoF
Figure 2. Locked cruciate repair.
position of full extension to complete 1 cycle. At the end of each 250-cycle set the finger was placed in full extension, the skin incision was reopened, and a 12-N preload tension was maintained across the FDP tendon during measurements.15 A single observer measured the gap between the tendon ends by using an electronic caliper. Three measurements (2 at the lateral aspect, 1 at the midpoint of the laceration’s transverse diameter) were taken and the average between them was obtained. The mean ⫹ SD gap measurements at the end of each 250-cycle set were calculated.
Ultimate Tensile Strength at Failure The tensiometer was set up for a single pull to failure resulting in suture rupture or pullout at a rate of 10 cm/min. To determine the ultimate tensile strength of a repair the maximum tension force at failure was recorded and stored in a computer. The mean ⫹ SD was computed for each suture technique.
Data Collection and Analysis The raw analog data from the force transducer, linear variable differential transducer, and pinch meter were sampled continuously at 100 Hz by an analogto-digital converter (ADC488/16; IO Tech, Cleveland, OH). An interface unit (MacSCSI 488; IO Tech) allowed us to store digital data on a computer (Apple Power Macintosh 8600, Apple Inc, Cupertino, CA) for analysis using custom software. A power analysis was performed with preliminary data, which showed that there would be 85% power (2sided ␣ ⫽ 0.05) with the current number of tendons. Analysis of variance was used to compare the change in WoF, maximum tensile load, and maximum pinch force among the 3 suture sizes. Then a regression model was fit to the data to compare suture sizes (reference group was suture size 2-0) while adjusting for different fingers and taking into account the correlation between fingers on the same hand. The interaction between suture size and finger also was included in all models.
The average baseline WoF with the 4-0 core suture fingers was 1.86 ⫾ 0.066, with the 3-0 core suture fingers it was 2.04 ⫾ 0.060, and with the 2-0 core suture fingers it was 1.83 ⫾ 0.078. The baseline WoF was not statistically different among the 3 groups (p ⫽ .88). There was a significant difference in the mean change in WoF between the 3 suture techniques (p ⫽ .005). A 2-0 suture repair with a mean change in WoF of 0.51 ⫾ 0.27 created significantly more WoF than a 3-0 suture with a mean change of 0.34 ⫾ 0.16 (p ⫽ .025) and than a 4-0 suture with a mean change of 0.31 ⫾ 0.11 (p ⫽ .007), but a 3-0 suture did not create significantly more WoF than a 4-0 suture (p ⫽ .70). Therefore, the 2-0 suture had a significantly greater change in WoF or gliding resistance than the 4-0 and 3-0 but there was no statistical difference between the 3-0 and 4-0 sutures (Fig. 3).
Maximum Tensile Load The highest maximum tensile load was obtained with 2-0 braided polyester sutures. Its mean maximum tensile load was 80 ⫾ 12 N whereas 3-0 and 4-0 braided polyester sutures had maximum tensile loads of 74 ⫾ 18 N and 66 ⫾ 12 N, respectively. The 3-0 polyester suture is on average 12% stronger than the 4-0 suture and the 2-0 polyester suture is on average 21% stronger than the 4-0 suture and 8% stronger than 3-0 suture. Results indicate that there was a significant difference in the maximum tensile load among the 3 suture sizes (p ⫽ .024). The regression model indicated that 2-0 braided polyester sutures were significantly different from 4-0 braided polyester sutures (p ⫽ .004) but that the difference was not significant between suture sizes 2-0 and 3-0 (p ⫽ .32) or between sizes 3-0 and 4-0 (p ⫽ .15). Therefore maximal tensile load statistically showed 2-0 braided polyester sutures to be stronger than 4-0 braided
Figure 3. Mean change of WoF. Error bars indicate SD.
452
The Journal of Hand Surgery / Vol. 30A No. 3 May 2005
Figure 4. Maximum tensile load in newtons. Error bars indicate SD.
polyester sutures but no statistically significant differences between 3-0 and 2-0 braided polyester sutures or between 3-0 and 4-0 braided polyester sutures were found (Fig. 4).
Gapping Gapping at the repairs was measured as described earlier. For 2-0 braided polyester sutures the average gapping measured was 0.5 ⫾ 0.6 mm and for 3-0 braided polyester sutures it was 0.4 ⫾ 0.8 mm. For the 4-0 sutures the gapping average was 0.3 ⫾ 0.5 mm. None of the repairs approached the clinically significant limit of 3 mm.16
Discussion The evolution of flexor tendon repair has progressed dramatically over the past 50 years.1 The pendulum has shifted to early primary repair to allow early passive- or even active-motion rehabilitation protocols that can improve the tendon nutrition, healing, and remodeling at the repair site.17–21 At the same time early motion can increase gapping at the repair site, which in turn increases the adhesion rate and need for tenolysis,16 resulting in a poor outcome. Because early active rehabilitation protocols were thought to lead to tendon repair rupture larger-caliber core sutures6 – 8 or more than 4-strand sutures22–25 have been developed and advocated. In the report by Choueka et al10 a literature review showed that the most strength lost was at day 5 during flexor tendon repairs and was 60%; therefore the strength of the repair only needed to be 3 times the strength needed to move the tendon without resistance (3 ⫻ 19 N), not 5 times as stated by Savage and Risitano.5 Strickland1 reviewed the forces needed for flexion and stated that for passive motion the tendon saw 2 to 4 N, for mild resistance up to 10 N, and for moderate resistance up to 17 N. It is only with strong composite grasp that significant force (70
N) is needed. Combining the review by Choueka et al10 and Strickland’s1 results we conclude that the required tendon force to flex a finger is between 30 and 51 N if one assumes that postoperative swelling changes active unresisted resistance to a moderate resistance. Taras et al7 studied the strength of progressively increasing suture sizes in the core repair and concluded that the strength of the repair is related directly to the size of the suture, with 2-0 being the strongest. The 2 problems with this work are (1) that the repair technique used was a 2-strand core repair, which cannot reach a strength of 30 to 50 N and (2) that they tested the strength to failure by linear distraction, which does not allow for the measurement of friction or the WoF, which could change with the increasing volume of the repair as the core suture gets larger. The article by Barrie et al8 showed the increasing strength of the 3-0 over the 4-0 suture with the cruciate repair tested in linear loading but this article did not address the 2-0 size as was recommended by Taras et al7 nor did it test the WoF. Hatanaka and Manske26 showed that the size of the core suture affected the ultimate strength and the gap strength in a linear distraction model. They found that larger core sutures were stronger and that locking repairs were stronger than grasping repairs. They did not evaluate the WoF, which could eliminate the larger core sutures, because they used a linear model for testing. In another article Aoki et al13 looked at WoF for multiple repair techniques, detailing the importance of looking at WoF for the best repair. A linear model takes the tendon from its surrounding anatomic structures (pulleys) and stresses it linearly. An anatomic model takes into account the confining aspects of the pulley system that restrict the amount of material placed at a repair site, which in turn affects gliding resistance. This model allows for more physiologic loading and testing, including looking at WoF.9 –13 Therefore we chose to use a curvilinear anatomic model to assess the locked cruciate core suture27 using braided polyester with a running locked epitenon stitch (Dermalon) because of its ease of placement, less interference (1 knot), and more than adequate strength needed for early active rehabilitation protocols.14 Each tendon in this study was cycled 1,000 times to a pinch force of 3.9 N (0.4 kg), far beyond what is needed to allow an early active rehabilitation program.15 In the previous study by Angeles et al14 not only did 2 different 4-0 braided polyester repair techniques yield more than 60 N of force but the locked cruciate was easier to
Alavanja, Dailey, and Mass / Four-Strand Flexor Tendon Repair
perform than the modified Becker and the WoF was significantly lower. They recommended this technique for flexor tendon repairs using early, unresisted, protected active-motion protocols. We performed this study to answer the question asked by Taras et al7: if 4-0 is good, should we be using 3-0 or even 2-0 locked cruciate core sutures to protect our repair better? In our study we were able to show that the strength advantage relates to 2-0 braided polyester over 4-0 braided polyester sutures, as all the other researchers have shown. This suture size could protect an aggressive postoperative early, unresisted, protected activemotion treatment protocol while avoiding rupture. Its higher WoF or gliding resistance, however, could preclude its use in the clinical setting. The 3-0 and 4-0 braided polyester suture10,14 strength and gliding function were not statistically different, with 3-0 being slightly favored in strength and 4-0 in WoF. They both were adequate to allow early protective active rehabilitation protocols. We felt obligated to study gapping because as the suture gets smaller there may be more stretch or elasticity in the repair; this is best tested with cycling. Gapping was found to be minimal among the 3 suture sizes after the cyclic loads, which supports gapping as a measure of suture configuration and not caliber. We were not measuring different suture configurations, however, so the gapping may show that the smaller caliber suture does not cut out of the tendon or stretch more, which also would be important to know. The weakness of this study is that it was in vitro and no healing took place, so that the conclusion of the strength of the needed repairs is a guess that is consistent with the literature. We also do not know how to factor in the increased resistance from edema after surgery. The 9 N needed for active motion against mild resistance is for normal tendons but is not necessarily the force needed to move repaired tendons or tendons with edema. We took this into consideration by using previous animal experiments1,10 and by requiring at least 3 times the repair strength for moderate resistance for a successful repair. Other weaknesses included dividing the interconnecting tissue in the carpal canal in individualizing each FDP tendon; this could cause an underestimate of the proximal resistance and increase the potential pull by the muscles. We conclude that 3-0 or 4-0 braided polyester locked cruciate core suture repairs of FDP tendons with a 6-0 locked epitenon repair
453
(Dermalon) should be strong enough for protected active postoperative rehabilitation. To confirm this proposal an in vivo study in the clinical setting is needed. The authors thank Kristen Kasza, MS, for her insights and assistance with the statistical analysis.
References 1. Strickland JW. Development of flexor tendon surgery: twenty-five years of progress. J Hand Surg 2000;25A:214 –235. 2. Cullen KW, Tolhurst P, Land D, Page RE. Flexor tendon repair in zone 2 followed by controlled active mobilisation. J Hand Surg 1989;14B:392–395. 3. Evans RB, Thompson DE. The application of force to the healing tendon. J Hand Ther 1993;6:266 –284. 4. Gratton P. Early active mobilization after flexor tendon repairs. J Hand Ther 1993;6:285–289. 5. Savage R, Risitano G. Flexor tendon repair using a “six strand” method of repair and early active mobilisation. J Hand Surg 1989;14B:396 –399. 6. Hatanaka H, Kojima T, Mizoguchi T, Ueshin Y. Aggressive active mobilization following zone II flexor tendon repair using a two-strand heavy-gauge locking loop technique. J Orthop Sci 2002;7:457– 461. 7. Taras JS, Raphael JS, Marczyk SC, Bauerle WB. Evaluation of suture caliber in flexor tendon repair. J Hand Surg 2001; 26A:1100 –1104. 8. Barrie KA, Tomak SL, Cholewicki J, Merrell GA, Wolfe SW. Effect of suture locking and suture caliber on fatigue strength of flexor tendon repairs. J Hand Surg 2001;26A: 340 –346. 9. Greenwald DP, Randolph MA, Hong H-Z, May JW. Augmented Becker versus modified Kessler tenorrhaphy in monkeys: dynamic mechanical analysis. J Hand Surg 1995; 20A:267–272. 10. Choueka J, Heminger H, Mass DP. Cyclical testing of zone II flexor tendon repairs. J Hand Surg 2000;25A:1127–1134. 11. Thurman RT, Trumble TE, Hanel DP, Tencer AF, Kiser PK. Two-, four-, and six-strand zone II flexor tendon repairs: an in situ biomechanical comparison using a cadaver model. J Hand Surg 1998;23A:261–265. 12. Komanduri M, Phillips CS, Mass DP. Tensile strength of flexor tendon repairs in a dynamic cadaver model. J Hand Surg 1996;21A:605– 611. 13. Aoki M, Manske PR, Pruitt DL, Larson BJ. Work of flexion after tendon repair with various suture methods. A human cadaveric study. J Hand Surg 1995;20B:310 –313. 14. Angeles JG, Heminger H, Mass DP. Comparative biomechanical performances of 4-strand core suture repairs for zone II flexor tendon lacerations. J Hand Surg 2002;27A: 508 –517. 15. Urbaniak JR, Cahill JD, Mortenson RA. Tendon suturing methods: analysis of tensile strengths. In: Hunter JM, Scheneider LHY, eds. Symposium on tendon surgery. St. Louis, MO: Mosby, 1975:70 – 80. 16. Seradge H. Elongation of the repair configuration following flexor tendon repair. J Hand Surg 1983;8:182–185. 17. Hitchcock TF, Light TR, Bunch WH, Knight GW, Sartori MJ, Patwardhan AG, et al. The effect of immediate con-
454
18.
19. 20.
21.
22.
23.
The Journal of Hand Surgery / Vol. 30A No. 3 May 2005 strained digital motion on the strength of flexor tendon repairs in chickens. J Hand Surg 1987;12A:590 –595. Gelberman RH, Woo SL, Lothringer K, Akeson WH, Amiel D. Effects of early intermittent passive mobilization on healing canine flexor tendons. J Hand Surg 1982;7:170 –175. Gelberman RH, Manske PR. Factors influencing flexor tendon adhesions. Hand Clin 1985;1:35– 42. Gelberman RH, Botte MJ, Spiegelman JJ, Akeson WH. The excursion and deformation of repaired flexor tendons treated with protected early motion. J Hand Surg 1986;11A:106 –110. Aoki M, Kubota H, Pruitt DL, Manske PR. Biomechanical and histologic characteristics of canine flexor tendon repair using early postoperative mobilization. J Hand Surg 1997; 22A:107–114. Xie RG, Zhang S, Tang JB, Chen F. Biomechanical studies of 3 different 6-strand flexor tendon repair techniques. J Hand Surg 2002;27A:621– 627. Sanders DW, Milne AD, Johnson JA, Dunning CE, Richards
24.
25.
26.
27.
RS, King GJW. The effect of flexor tendon repair bulk on tendon gliding during simulated active motion: an in vitro comparison of two-strand and six-strand techniques. J Hand Surg 2001;26A:833– 840. Barrie KA, Wolfe SW, Shean C, Shenbagamurthi D, Slade JF 3rd, Panjabi MM. A biomechanical comparison of multistrand flexor tendon repairs using an in situ testing model. J Hand Surg 2000;25A:499 –506. Barrie KA, Tomak SL, Cholewicki J, Wolfe SW. The role of multiple strands and locking sutures on gap formation of flexor tendon repairs during cyclical loading. J Hand Surg 2000;25A:714 –720. Hatanaka H, Manske PR. Effect of suture size on locking and grasping flexor tendon repair techniques. Clin Orthop 2000,375:267–274. McLarney E, Hoffman H, Wolfe SW. Biomechanical analysis of the cruciate four-strand flexor tendon repair. J Hand Surg 1999;24A:295–301.