*Department of Trauma, Hand, Plastic and Reconstructive SurgeryWuerzburg, Germany.

Received for publication April 8, 2015; accepted in revised form M

No benefits in any form have been received or will be receiveindirectly to the subject of this article.

Corresponding author: Martin C. Jordan, MD, Department of TraumReconstructive Surgery, University Hospital, Oberduerrbacher Str. 6Germany; e-mail: Jordan_M@ukw.de.

0363-5023/15/---0001$36.00/0http://dx.doi.org/10.1016/j.jhsa.2015.05.032

SCIENTIFIC ARTICLE

Biomechanical Analysis of the Modified Kessler,

Lahey, Adelaide, and Becker Sutures

for Flexor Tendon Repair

Martin C. Jordan, MD,* Vanessa Schmitt,* Hendrik Jansen, MD,* Rainer H. Meffert, MD,*Stefanie Hoelscher-Doht, MD*

Purpose To compare the biomechanical properties of the modified Kessler, Lahey, Adelaide,and Becker repairs, which are marked by either a locking-loop or a cross-lock configuration.

Methods Ninety-six lacerated porcine flexor tendons were repaired using the respective coresuture and an epitendinous repair. Biomechanical testing was conducted under static andcyclic loads. Parameters of interest were 2-mm gap formation force, displacement duringdifferent loads, stiffness, maximum force, and mode of failure.

Results The meaningful gap formation occurred in all 4 repairs at similar tension loads withoutany significant differences. Maximum force was highest in the Becker repair with aconsiderable difference compared with the modified Kessler and Lahey sutures. The Adelaiderepair showed the highest stiffness. Overall, the displacement during cyclic loading demon-strated similar results with an exception between the Lahey and the Adelaide repairs at 10N load. Failure by suture pull-out occurred in 42% in the modified Kessler, in 38% in theLahey, and in 4% in the Adelaide repairs. The Becker repair failed only by suture rupture.

Conclusions The results of our study suggest that the difference between the 4-strand repairswith a cross-lock or a locking-loop configuration is minor in regard to gap formation. A strongepitendinous suture and the application of core suture pretension might prevent differences ingapping. However, the modified Kessler and Lahey repairs had an inferior maximum tensilestrength and were prone to early failure caused by the narrow locking loops with their limitedlocking power.

Clinical relevance We suggest that surgeons should use pre-tension in repaired tendons to improvegap resistance and should avoid narrow locking loop anchoring to the tendon. (J Hand Surg Am.2015;-(-):-e-. Copyright � 2015 by the American Society for Surgery of the Hand. Allrights reserved.)Key words 4-strand repair, hand, reconstruction, suture, tendon.

, University Hospital,

ay 29, 2015.

d related directly or

a, Hand, Plastic and, 97080 Wuerzburg,

PURPOSEMany different kinds of suture techniques have beenintroduced to pursue the aim of early active mobiliza-tion.1,2 The postoperative treatment is of great im-portance in order to avoid restricting adhesions andto gain full excursion at the end of therapy.3,4 To allowsuch a beneficial mobilization, a highetensile strengthrepair is required. In recent years, suture techniquescontaining multiple core strands and different lockingconfigurations have been described to achieve a reliable

� 2015 ASSH r Published by Elsevier, Inc. All rights reserved. r 1

FIGURE 1: Illustration of the different 4-strand repairs. A, The modified Kessler suture above and a porcine flexor tendon repaired in thesame technique below.7 B, Lahey suture.8 C, Adelaide suture.9 D, Becker suture.12

2 MODIFIED SUTURES FOR FLEXOR TENDON REPAIR

repair strength.5 Despite an increasing knowledgeabout the biomechanical behavior of such sutures, thereis no consensus about the ideal technique,1,6 and sur-geons encounter a growing number of different repairs.The purpose of this biomechanical study was to analyzeand compare the modified Kessler,7 Lahey,8 Adelaide,9

and Becker repairs10e12 regarding their primary tensilestrength. These techniques mainly vary in their lockingconfiguration, and we hypothesized that there wouldbe a significant difference in gap formation force, sti-ffness, and displacement or maximum force betweenthe repairs.

MATERIAL AND METHODSSpecimens

For this study, 96 fresh-frozen porcine flexor digitorumprofundus tendons were used. Porcine flexor tendonshave similar biomechanical properties to human flexortendons and are frequently used for biomechanicalstudies.13 Tendons from the forelimb were dissectedbetween the A2 and the A4 pulleys.14 All tendons weremeasured to ensure equal sample size. Tendons withdeviating diameter or defects were excluded. Harvestedtendons were stored inside saline-soaked gauzes anddeep-frozen at e20�C. Before testing, tendons werethawed at room temperature for 12 hours and a scalpelwas used to carefully create a defect in the middle ofeach tendon. Throughout testing, tendons were keptmoist using saline spray to avoid desiccation.

Repair and material

Tendons were randomly assigned to 1 of 4 groups with24 specimens per group. Group 1 tendons were repairedwith a4-strandmodifiedKessler suture (Fig.1), inwhich

J Hand Surg Am. r V

the knot is buried inside the tendon.7,15 The knot lies inthe transverse component of the suture, and the suture isanchoredwith 8 locking loops. Tendons of group 2wererepairedwith a 4-strandLahey repair.8 TheLahey sutureis a cruciate repair with 8 locking loops. Tendons ofgroup 3 were repaired with a 4-strand Adelaide suture(also known as cruciate cross-stitch locked repair orlockedcruciate repair).9TheAdelaidesuture is acruciaterepair with 4 cross-locks. Tendons of group 4 wererepaired with a 4-strand Becker suture (also known asMGH repair).12 Twelve cross-locks can be found in themodified Becker suture, which are either exposed orembedded (Fig. 2). A core suture tendon purchase of0.7 cm and a 10% shortening was used for all repairs toensure the best tensile strength.16e18 Three consecutivethrowswereperformedfor the coreandperipheral sutureknots. Anchor configuration was either a locking-loop(modified Kessler and Lahey) with a size of 1 mm or across-lock (Adelaide and Becker) with the correspond-ing size of 2 mm (Fig. 2).14,19,20 Each repair was com-binedwith a peripheral suture (Fig. 3). The epitendinoussuture was a running locking suture with 2-mm tendonpurchase, slight tension, which crossed the repair zonebetween 8 and 9 times. For all core sutures, a 3-0 poly-dioxanone (PDS, Ethicon, Somerville, NJ) was used,and for all peripheral sutures, a 5-0 polydioxanone.Polydioxanone is a monofilament synthetic absorbablesuture material. The tendons were repaired by a trainedorthopedic surgeon (MCJ) experienced in tendon repairtechniques.

Biomechanical testing

Tests were conducted using a mechanical testing ma-chine (Z020; Zwick/Roell GmbH, Ulm, Germany) and

ol. -, - 2015

FIGURE 3: Epitendinous repair. A, Peripheral running, simple-locking suture was added to each core suture. B, Porcine flexortendon repaired with a core and epitendinous suture.

FIGURE 2: Different anchor techniques. A, Grasping loopwithout locking configuration. B, Locking-loop configurationwith Pennington lock.29 C, Cross-lock configuration (exposed).D, Cross-lock configuration (embedded).10,11,21

MODIFIED SUTURES FOR FLEXOR TENDON REPAIR 3

the test Xpert II software (Version 3.6, Zwick/Roell).Pretesting up to 300 N achieved sufficient gripping ofthe tendon ends without slipping. Uniaxial testing wasperformed using a 20-kN load cell and 2 stainless steelclamps. The distance between the 2 clamps was stan-dardized with a gap of 3 cm. Tendon length of 1.5 cmwas clamped at each side. Two different test settingswere applied, a static (n ¼ 12) and a cyclic (n ¼ 12)testing. The static test had 3-N preload and an ad-vancement rate of 20 mm/min and was an axial load tofailure test. The2-mmgap formation force, stiffness, andmaximum force were measured under static conditions(Fig. 4). Stiffness (N/mm) was calculated as the slope ofthe linear section of the load-displacement curve fromthe initial loading portion of the test.

The cyclic loading protocol consisted of 3 levels.First, each repair was loaded from 0 to 10 N at 20mm/min for 400 cycles. Afterwards, another 400 cyclesfrom 0 to 20 N were conducted. If the repair withstoodthis level without failure, another 400 cycles from 0 to

J Hand Surg Am. r V

30 N were applied. Load and displacement were con-tinuously recorded to generate a load-displacementcurve. The length increase of the tendon was recor-ded from the load-displacement graph at the final (fourhundredth) cycle of each load level.

Statistics

Investigated parameters were 2-mm gap formationforce (N), stiffness (N/mm), maximum force (N),displacement (mm) at 10, 20, and 30 N and modeof failure (pull-out vs rupture). A power analysis wasperformed using a power of 80% that proved the samplesize to be adequate. Results are presented asmean valueincluding SD. The Shapiro-Wilk test was performedto analyze the distribution. Analysis of variance withTukey post hoc test and the Kruskal-Wallis-test wereused for comparison of the means. A P value of lessthan .05 was considered to be statistically significant.

RESULTS2-mm Gap formation force

There was no difference in 2-mm gap formation forceamong the 4 repairs (Fig. 5).

Maximum force

The mean maximum force for the Becker repair wassignificantly higher than for the Lahey and modifiedKessler repairs. The difference between the modifiedKessler, the Lahey, and the Adelaide repair did notreach a significant level (Fig. 5).

Stiffness

The Adelaide repair showed a statistically higher sti-ffness in comparison with all other suture techniques.

ol. -, - 2015

FIGURE 5: Results for the different repairs. A, Tension force toproduce a 2-mm gap at the repair site and maximum force beforefailure. B, The Adelaide suture exhibits a significantly (P < .05)elevated stiffness. C, There is no notable displacement duringcyclic testing between the repairs except for the Lahey and theAdelaide suture at a tension level of 10 N (P < .01).FIGURE 4: Biomechanical testing. A, Load-displacement graph

generated during static testing. A 2-mm gap generally occurredbefore rupture of the epitendinous suture. B, Cyclic testing withdifferent load levels.

4 MODIFIED SUTURES FOR FLEXOR TENDON REPAIR

There was no difference between the remaining repairs(Fig. 5).

Displacement

There was a significant difference in displacement at10 N between the Lahey and the Adelaide repairs. Noother significant difference appeared during cyclicloading under 10, 20, and 30 N.

Mode of failure

In the modified Kessler repair group, 10 out of 24tendons failed by suture pull-out, and in the Laheygroup, 9 out of 24 failed in this manner. In the Ade-laide group, 1 out of 24 failed by suture pull-out. TheBecker repairs all failed by suture rupture. The ruptureoccurred between the tendon ends.

J Hand Surg Am. r V

DISCUSSIONThe results of our study indicate that the locking-loopand cross-lock configurations affected the biomechan-ical behavior of the different repairs. More specifically,there were similarities in gap formation and displace-ment and differences in stiffness, maximum force, andmode of failure.

The various locking configurations may have in-dividual mechanical properties in end-to-end flexortendon repair.1,21e23 Some of the previous studiesclaimed a superior repair strength for the cross-lockconfiguration;22,23 however, in our study, the gapformation appeared in all tested repairs at a similartension force without a measurable difference betweenthe locking-loop and the cross-lock configurations.This result is an extension to the finding of Xie andTang21 who demonstrated similar locking power forthe cross-lock and circle-lock component.

ol. -, - 2015

MODIFIED SUTURES FOR FLEXOR TENDON REPAIR 5

The applied pre-tension or the core suture mightexplain why there was no difference in gap formationin our study. Wu and Tang18 reported that using suturetension with a 10% shortening of the tendon areaencompassed by the core suture increased the resistanceto gap formation of the repair. It is possible that theapplied pre-tension prevented potential differences ingap formation that would otherwise have been quanti-fiable, even if this effect is controversial.22 Further-more, we used a 3-0 core suture and a 5-0 epitendinoussuture, which are both stronger than the 4-0 and 6-0sutures used by many surgeons.1,5,24,25 Throughout ourtesting, the 2-mm gap consistently occurred beforefailure of the epitendinous suture, which emphasizesthe important role of this additional suture in order toavoid gap formation. Overall, our data suggest that thepreviously described superiority of the cross-lockcomponent could be reversed by a strong epi-tendinous suture as has been suggested by Croog et al23

and by the application of core suture pre-tension. Thisfinding is supported not only by the 2-mm gap for-mation in our results but also by the similar displace-ment that represents the stability under repetitive load.

Despite the fact that our studied locking configu-rations did not differ in their resistance to gap for-mation, the repairs including locking loops weremarked by a higher rate of suture pull-out and lessmaximum tensile strength. Hatanaka and Manske26

showed that there is a relationship between the sizeof the locking-loop and the repair strength. Increasingthe tendon area encompassed by the locking loopcauses a proportional increase in maximum tensilestrength.26 In our study, especially the Lahey suturehad low repair strength that might be explained by thenarrow locking loops, which impair proper anchorageto the tendon. The small locking loops with crossdiameter of 1 mm in the Lahey and modified Kesslerrepair were further narrowed by the applied pre-tension of the core suture. Peltz et al27 demonstratedthat a loss of the loop configuration was caused byaxial tension under radiographic control. This mightaggravate anchorage of the suture to the tendon.Whileusing such small locking loops, it is further difficult toascertain the exact location of the transverse andlongitudinal suture, thus it might rather be a graspingloop instead of a locking loop with less holding ca-pacity. Therefore, the Lahey repair, including itsnarrow locking loops, seems to be the less favorablerepair of our tested techniques. The same problemapplies to the modified Kessler suture that has alsosmall locking loops and likewise low repair strength.Because the modified Kessler and Lahey repairs areboth locking loop and have a small lock size, we are

J Hand Surg Am. r V

unsure whether 1 or 2 of these factors lead to theinferior mechanical performance. According to theresult of Xie et al,19 a locking size of 1 mm producedsignificantly lower locking strength than a repairshowing a locking size of 2 or 3 mm. Therefore, thesmall lock size can be considered a factor that leads tolower strength in these 2 repair methods in our study.Nevertheless, we cannot state whether the lockingloop is the cause of lower strength because the repairsincluding locking loops tested here both had a smalldiameter, and small diameter has been proven to leadto lower strength.

Beside the locking diameter, we found that thenumber of anchor points can also raise the maximumstrength as seen in the Becker repair with its 12 crosslocks. Still, the maximum tensile strength is a lessimportant parameter because it appears far beyondthe gapping. The Adelaide repair stands out by onlya small amount of displacement, high stiffness, andalmost no suture pull-out. Therefore, we found it to be areliable suture technique. However, a clear recomme-ndation cannot be made because all repairs demon-strated sufficient repair strength.28 Furthermore, thisstudy is a biomechanical analysis in a nonhumanmodel, and it can be questioned if our results show aclinical difference. Factors like friction, bulk, or adhe-sions were not addressed. The number of repetitiveloads during our cyclic testing is limited and does notpresent the full course of mobilization until healing.

ACKNOWLEDGMENTThe authors thank the IZKF (Interdisciplinary Centerfor Clinical Research, University Clinics of Wuerz-burg) for supporting our biomechanical studies andAlice K. Jordan for writing assistance.

REFERENCES

1. Wu YF, Tang JB. Recent developments in flexor tendon repair tech-niques and factors influencing strength of the tendon repair. J HandSurg Eur Vol. 2014;39(1):6e19.

2. Strickland JW. The scientific basis for advances in flexor tendonsurgery. J Hand Ther. 2005;18(2):94e110; quiz 111.

3. Gelberman RH, Manske PR. Factors influencing flexor tendon ad-hesions. Hand Clin. 1985;1(1):35e42.

4. Aoki M, Kubota H, Pruitt DL, Manske PR. Biomechanical and his-tologic characteristics of canine flexor tendon repair using earlypostoperative mobilization. J Hand Surg Am. 1997;22(1):107e114.

5. Tang JB, Chang J, Elliot D, Lalonde DH, Sandow M, Vogelin E.IFSSH Flexor Tendon Committee report 2014: from the IFSSHFlexor Tendon Committee (Chairman: Jin Bo Tang). J Hand SurgEur Vol. 2014;39(1):107e115.

6. Strickland JW. Development of flexor tendon surgery: twenty-fiveyears of progress. J Hand Surg Am. 2000;25(2):214e235.

7. Zechner W, Buck-Gramcko D, Lohmann H, Goth D, Stock W.Improvement of suture technic in flexor tendon injuries. Clinical andexperimental study [in German]. Handchir Mikrochir Plast Chir.1985;17(1):8e13.

ol. -, - 2015

6 MODIFIED SUTURES FOR FLEXOR TENDON REPAIR

8. Manchio JV, Shashikant MP, Shrivastava A, Weinzweig J,Vernadakis AJ. Evaluation of a new 4-strand flexor tendon repair in acadaveric porcine model. J Hand Surg Am. 2009;34(1):102e107.

9. Sandow MJ, McMahon M. Active mobilisation following singlecross grasp four-strand flexor tenorrhaphy (Adelaide repair). J HandSurg Eur Vol. 2011;36(6):467e475.

10. Becker H, Davidoff M. Eliminating the gap in flexor tendon surgery.A new method of suture. Hand (N Y). 1977;9(3):306e311.

11. Becker H, Orak F, Duponselle E. Early active motion following abeveled technique of flexor tendon repair: report on fifty cases.J Hand Surg Am. 1979;4(5):454e460.

12. Greenwald DP, Hong HZ, May JW Jr. Mechanical analysis of tendonsuture techniques. J Hand Surg Am. 1994;19(4):641e647.

13. Mao WF, Wu YF, Zhou YL, Tang JB. A study of the anatomy andrepair strengths of porcine flexor and extensor tendons: are theyappropriate experimental models? J Hand Surg Eur Vol. 2011;36(8):663e669.

14. Peltz TS, Haddad R, Scougall PJ, Nicklin S, Gianoutsos MP,Walsh WR. Influence of locking stitch size in a four-strand cross-locked cruciate flexor tendon repair. J Hand Surg Am. 2011;36(3):450e455.

15. Sebastin SJ, Ho A, Karjalainen T, Chung KC. History and evolutionof the Kessler repair. J Hand Surg Am. 2013;38(3):552e561.

16. Lee SK, Goldstein RY, Zingman A, Terranova C, Nasser P,Hausman MR. The effects of core suture purchase on the biome-chanical characteristics of a multistrand locking flexor tendon repair:a cadaveric study. J Hand Surg Am. 2010;35(7):1165e1171.

17. Tang JB, Tan J, Xu Y. Lengthening and locking: two ways toimprove repair strength of obliquely lacerated tendons. J Hand SurgAm. 2003;28(5):832e837.

18. Wu YF, Tang JB. Effects of tension across the tendon repair site ontendon gap and ultimate strength. J Hand Surg Am. 2012;37(5):906e912.

J Hand Surg Am. r V

19. Xie RG, Xue HG, Gu JH, Tan J, Tang JB. Effects of locking area onstrength of 2- and 4-strand locking tendon repairs. J Hand Surg Am.2005;30(3):455e460.

20. Dona E, Gianoutsos MP, Walsh WR. Optimizing biomechanicalperformance of the 4-strand cruciate flexor tendon repair. J HandSurg Am. 2004;29(4):571e580.

21. Xie RG, Tang JB. Investigation of locking configurations for tendonrepair. J Hand Surg Am. 2005;30(3):461e465.

22. Wu YF, Cao Y, Zhou YL, Tang JB. Biomechanical comparisons offour-strand tendon repairs with double-stranded sutures: effects ofdifferent locks and suture geometry. J Hand Surg Eur Vol. 2011;36(1):34e39.

23. Croog A, Goldstein R, Nasser P, Lee SK. Comparative biomechanicperformances of locked cruciate four-strand flexor tendon repairs inan ex vivo porcine model. J Hand Surg Am. 2007;32(2):225e232.

24. Alavanja G, Dailey E, Mass DP. Repair of zone II flexor digitorumprofundus lacerations using varying suture sizes: a comparativebiomechanical study. J Hand Surg Am. 2005;30(3):448e454.

25. Taras JS, Raphael JS, Marczyk SC, Bauerle WB. Evaluation of suturecaliber in flexor tendon repair. J Hand Surg Am. 2001;26(6):1100e1104.

26. Hatanaka H, Manske PR. Effect of suture size on locking andgrasping flexor tendon repair techniques. Clin Orthop Relat Res.2000;375:267e274.

27. Peltz TS, Haddad R, Walsh WR. The change in three-dimensionalgeometry of the Kessler flexor tendon repair under tension: a qualita-tive assessment using radiographs. J Hand Surg Eur Vol. 2010;35(8):676e677.

28. Urbaniak JR, Cahill JD, Mortenson RA. Tendon suturing methods:analysis of tensile strengths. AAOS Symposium on Tendon Surgery ofthe Hand. St. Louis: CV Mosby; 1975:70e80.

29. Pennington DG. The locking loop tendon suture. Plast ReconstrSurg. 1979;63(5):648e652.

ol. -, - 2015