Cyclic testing of meniscal sutures

Cyclic Testing of Meniscal Sutures Romain Seil, M.D., Stefan Rupp, M.D., and Dieter M. Kohn, M.D.

Summary: Suturing the meniscus has become a standard procedure for repairable tears. Studies investigating the outcome of meniscal sutures report a considerable rate of failures. Regarding the indications, which have been extended to the avascular zones, and regarding some accelerated rehabilitation protocols, the need for further in vitro investigations has become obvious. The aim of this study was to compare different meniscal suture types (vertical and horizontal mattress sutures) and materials (absorbable monofilament PDS 2-0, and nonabsorbable braided Ethibond 2-0 [Ethicon, Somerville, NJ]) under standard and cyclic loading conditions. Testing was performed on medial porcine menisci. In group A, specimens were tested to failure at a cross-head speed of 50 mm/minute. In group B, cyclic testing (100 cycles) was performed first within different load intervals (5 to 20 N and 5 to 40 N). Finally, the specimens were loaded until failure. In both groups, the failure loads were recorded and the failure modes were analyzed. In group A, there was no difference between suture type or suture material, with a mean failure load of 60 N. The failure modes were significantly different for vertical (100% suture failure) and horizontal sutures (50% suture failure) (P ⬍ .0001). In group B, 13% of the sutures failed under cyclic loading (7 with 40-N load, 1 with 20-N load). The gap of the sutured tear that appeared within the first load cycles was broader in horizontal sutures (P ⬍ .001). During the first cycles, the thread cut through the meniscus tissue and disappeared from the surface (partial tissue failure). There was no difference according to suture material. The ultimate failure loads after cyclic loading did not differ from the values of group A. These results show that meniscal sutures may fail under repetitive loading conditions and that a gap appears between the meniscal margins within the first loading cycles irrespective of the suture type and suture material used. The appearance of the gap and suture disappearance on the meniscal surface because of partial tissue failures (which were more pronounced in the horizontal sutures) confirmed the superior resistance of meniscal tissue to vertical sutures. Key Words: Meniscus—Meniscal repair—Suture technique—Suture material—Cyclic testing.

M

eniscus sutures have become the treatment of choice for repairable meniscal tears. There are many suture techniques available and successful results have been reported by numerous authors, either by arthroscopic1-13 or open techniques.14 It is known that the success of a meniscal suture depends on adequate stability, coaptation of the tear site, and the biological healing capacity of the meniscus.15 Whereas the primary stability of meniscus sutures has been evaluated in several biomechanical studFrom the Department of Orthopaedic Surgery, the University of Saarland Medical School, Homburg/Saar, Germany. Address correspondence and reprint requests to Romain Seil, M.D., Department of Orthopaedic Surgery, University of Saarland Medical School, D-66421 Homburg/Saar, Germany. r 2000 by the Arthroscopy Association of North America 0749-8063/00/1605-2153$3.00/0 doi:10.1053/jars.2000.4379

ies,16-20 there are no studies available investigating the fixation strength of meniscal sutures under cyclic loading. Early range of motion and accelerated postoperative rehabilitation2,3,5,10 underline the importance of the stability of meniscus sutures under repetitive loading conditions. Cyclic loading seems to be the best way to simulate these conditions in vitro.21 This in vitro study compared 2 different suture materials and techniques under standard and under cyclic loading conditions using medial porcine menisci.

MATERIALS AND METHODS Thirty-four medial menisci were harvested from adult pigs. They were separated from their capsular and bony attachments and immediately frozen at ⫺20°C until the time of testing. Anatomy and function

Arthroscopy: The Journal of Arthroscopic and Related Surgery, Vol 16, No 5 (July-August), 2000: pp 505–510

505

506

R. SEIL ET AL.

of porcine menisci appear to be similar to human menisci.22,23 The knees from which the menisci were taken did not show any sign of joint degeneration. One hour before testing, the menisci were placed in a 0.9% saline solution at room temperature. The menisci were cut vertically with a scalpel 3 mm from the periphery (Fig 1). According to previous studies,16,20 a complete peripheral meniscus tear was created to test only the repair site and to prevent any load transmission from meniscus tissue other than at the repair site. After repair of the meniscus, 2 steel wires were placed on either side of the repair site on the peripheral rim and 2 more on the inner portion of the meniscus (Fig 1). Wires were used instead of holding sutures to diminish, as much as possible, bias of the results under cyclic loading due to the elasticity of holding sutures. The wires were fixed in the clamps of a universal testing machine (Zwick type 1474, Ulm, Germany). At testing, a vertical load was applied to both pairs of wires to distract the meniscal parts at the repair site. The tests were conducted at a speed of 50 mm/minute. The loading experiments were performed at room temperature. The specimens were kept moist continuously. The suture techniques investigated were vertical mattress and horizontal mattress sutures.15,17 Absorbable monofilament suture (PDS 2.0) and nonabsorb-

FIGURE 1. Meniscus specimen armed with wires on either side of the repair site. Repair was performed with a horizontal mattress suture using PDS 2-0. At the beginning of cyclic loading a gap appeared between the 2 parts of the meniscus (arrow).

able braided suture (Ethibond 2.0; Ethicon, Somerville, NJ) was used. Only 1 suture was tested at a time. The distance between suture arms was 3 mm. The sutures were placed 3 to 4 mm from the peripheral edge of the tear. They were inserted from the articular, femoral side of the meniscus. A knot was tied on the capsular side. In every case, we performed hand-tied square knots with 5 throws. Both suture arms were marked behind the knot so as to evaluate knot slippage during cyclic loading. Each meniscus was tested 3 times with the same suture technique. A previous study20 showed that there was no difference in failure strengths whether a specimen was used once or 3 times for testing. At the first test, the suture was placed in the midportion of the meniscus. For the following tests, the suture was placed either 3 mm more anterior or 3 mm more posterior. The study was conducted in 2 parts. In the first part (group A), standard testing was performed by single maximum loading. The primary fixation strength of the suture types and materials were evaluated to allow a better comparison with the results obtained after cyclic testing. Forty meniscus sutures (10 vertical and 10 horizontal mattress sutures with PDS and Ethibond each) were tested to failure. After applying a preload of 5 N to the specimen, they were loaded to failure at a strain rate of 50 mm/minute. In the second part of the study (group B), cyclic loading and testing to failure were performed. In group B-1, 100 load cycles with a load from 5 to 20 N were applied to 20 vertical mattress sutures (10 Ethibond and 10 PDS sutures). In group B-2, the specimens were tested with 100 load cycles between 5 and 40 N (20 vertical mattress sutures with 10 PDS and 10 Ethibond sutures [group B-2-1] and 20 horizontal mattress sutures with 10 PDS and 10 Ethibond sutures [group B-2-2]). The methodology for cyclic testing included a preloading of 5 N and a repetitive vertical loading up to 20 and 40 N, respectively. The strain rate was 50 mm/minute in each group. At the end of cyclic testing, the gap between the 2 parts of the meniscus was measured. During cyclic loading, the visible part of the suture at the femoral meniscus surface tended to disappear because of a partial tissue failure. In these cases, the suture was not torn out from the meniscus. At the opposite, in complete tissue failures, the suture was torn out of the meniscus. To quantify partial tissue failures, the length of the visible part of the suture was measured after cyclic loading. Knowing that the initial distance separating both arms of the suture was 3 mm, 3 categories could be established: less than 50% disappearance of the initial 3 mm, more than 50% disappearance, and

CYCLIC TESTING OF MENISCAL SUTURES complete disappearance of the suture material from the meniscal surface. In a second step, single vertical tensile loading was performed at a strain rate of 50 mm/minute until failure. In all groups, the load displacement curves were recorded and the ultimate failure loads were measured. The failure mode was analyzed visually. For statistical analysis, ␹-square and Wilcoxon tests were performed.

507

TABLE 2. Failure Modes of Vertical and Horizontal Mattress Sutures After Maximum Loading Suture Type and Material

n

I

II

III

IV

Horizontal (PDS 2-0) Horizontal (Ethibond 2-0) Vertical (PDS 2-0) Vertical (Ethibond 2-0)

10 10 10 10

4 4 10 10

2 0 0 0

0 3 0 0

4 3 0 0

Horizontal (PDS 2-0) Horizontal (Ethibond 2-0) Vertical (PDS 2-0) Vertical (Ethibond 2-0)

9 8 16 19

8 6 16 17

0 0 0 0

0 0 0 2

1 2 0 0

A

B

RESULTS Group A The mean ultimate failure strengths of the specimen tested with 1 single load to failure are listed in Table 1. There was neither a statistical difference between horizontal and vertical sutures nor between the 2 suture materials. Four types of failure modes were observed: suture breakage close to the knot, midsubstance suture breakage, knot loosening, and complete tissue failure. The differences in the failure modes between horizontal and vertical sutures were statistically significant (P ⬍ .0001) (Table 2). Group B Cyclic Loading Meniscal Gap: Visual analysis of cyclic loading showed the immediate appearance of a gap between the 2 meniscal margins from the beginning of the first cycle. The only exception was noted in a vertical PDS mattress suture in group B-2-1 (loading from 5 to 20 N). The progression of the gap was documented on the load-displacement curves. About 50% of the elongation of the meniscus-suture-system appeared during the 10 first load cycles (Fig 2). The mean gap was 3 mm. It varied significantly according to the applied load (group B-1 v group B-2-1, P ⬍ .001; group B-1 v group B-2-2, P ⬍ .0001 ) and suture type (group B-2-1

NOTE. Failure mode: I, suture breakage near to the knot; II, midsubstance suture breakage; III, suture loosening; and IV, tissue failure. No tissue failures were observed with vertical sutures. A: Single maximum loading. B: Maximum loading after cyclic loading (The failure modes of the 8 sutures which failed during cyclic loading are not represented).

v group B-2-2, P ⫽ .005), but not with the suture material (Table 3). Within the first cycles, the suture tended to disappear from the meniscal surface because of partial tissue failure. There were significant differences between vertical and horizontal sutures (Fig 3) (P ⬍ .0001). Partial tissue failures were also more pronounced with PDS sutures compared with Ethibond sutures. However, this difference was not statistically significant (P ⬎ .05). Knot slippage could be documented in 4 of 60 cases (group B-1, Ethibond 1 case of 1 mm slipping of 1 suture arm; group B-2, Ethibond 2 cases of 1 mm slipping; PDS 1 ⫻ 1 mm). Failure Rate: The failure rate during cyclic loading was 13.3% (8 cases). Seven failures appeared in groups B-2-1 and B-2-2 (loading between 5 and 40 N). Of a total of 40 cyclically tested vertical sutures, 5 failed (12.5%). Three horizontal sutures failed (15%). The failure modes were in 7 cases suture breakage near

TABLE 1. Ultimate Failure Strengths of Meniscus Mattress Sutures Suture Material

Vertical Sutures (N)

n

SD (N)

Range (N)

Horizontal Sutures (N)

n

SD (N)

Range (N)

PDS Ethibond

63 59

10 10

6.8 8.1

48-72 48-72

58 58

10 10

14.6 9.5

32-74 40-70

PDS Ethibond

54 58

9 10

14.5 11.5

20-70 28-70

— —

— —

— —

— —

PDS Ethibond

61 64

8 8

11 10

44-72 54-76

64 57

9 8

12.7 10.6

34-78 34-70

A

B

C

NOTE. No statistically significant differences could be determined. A: Single maximum loading. B: Maximum loading after cyclic testing (previous cyclic loads of 5 to 20 N). C: Maximum loading after cyclic testing (previous cyclic loads of 5 to 40 N).

508

R. SEIL ET AL.

FIGURE 2. Elongation of the meniscus-suture-system as documented on the load displacement curves (in percent). Nearly 50% of the elongation of the meniscus-suture system occur during the 10 first load cycles.

to the knot and in 1 case a complete tissue failure (horizontal PDS suture). Testing to Failure After Cyclic Loading The values of the ultimate failure strength after cyclic loading are presented in Table 1. There was no statistically significant difference between suture types and materials. As with failure modes after single testing to failure without previous cyclic loading, complete tissue failures were only noted with horizontal sutures. The vast majority of failures consisted of suture breakage (Table 2). Suture loosening in both groups A and B appeared only with Ethibond sutures (5 cases, 8.7% of Ethibond sutures). DISCUSSION Previous clinical studies have had failure rates of meniscal repairs reaching up to 50% with isolated repairs in anterior cruciate ligament stable knees.6 The reasons for these failures are still unclear. On the other hand, the indications for repairs have been extended to

FIGURE 3. Disappearance of the suture material from the meniscal surface as estimated by visual analysis after cyclic loading (partial tissue failure). There was a statistically significant difference between horizontal and vertical sutures in partial tissue failures (P ⬍ .0001). In 33% of the horizontal sutures, the suture material was not visible at the central meniscus surface at the end of cyclic loading.

tears in the avascular zone of the meniscus with a reduced healing potential. Accelerated rehabilitation protocols, inducing repetitive loading of the repair site, become more and more popular. Regarding these changes in meniscal repair indications and accelerated rehabilitation, Post et al.15 stated that a ‘‘thorough understanding of the effects of suture types and techniques on the load to failure of meniscal sutures is important.’’ Previous in vitro studies of meniscal sutures only investigated the ultimate failure load. The purpose of this study was to analyze meniscal sutures under repetitive loading conditions. To improve the validation of the data gained by cyclic loading, preliminary testing of ultimate failure strengths was performed. Single maximum loading showed no differences in failure strengths between

TABLE 3. Meniscal Gap After Cyclic Loading Group

Applied Load

n

Suture Type

Gap (mm)

SD (mm)

Range (mm)

B-1 B-2-1 B-2-2

5-20 N 5-40 N 5-40 N

19 17 16

Vertical mattress Vertical mattress Horizontal mattress

2.3 3.4 4.3

1 1 0.9

0-4.1 1.8-4.8 3.1-6.1

NOTE. Group 1: loading from 5 to 20 N. Group 2: loading from 5 to 40 N. The difference between each group was statistically significant.

CYCLIC TESTING OF MENISCAL SUTURES horizontal and vertical sutures and no differences between the 2 suture materials. This seems to contradict previous findings showing higher failure strengths of vertical sutures compared with horizontal sutures.19 However, comparison with previous data is difficult because the testing conditions differed from study to study. Studies using human cadaveric menisci19,20 may introduce potential variability of the failure loads because the strength of meniscal tissue may vary with specimens of different ages. Dervin et al.18 used incomplete peripheral longitudinal tears and vertical loop sutures with Ethibond 2-0 suture material. Despite using human cadaveric menisci, the failure strength of vertical loop sutures was comparable to the present data. Post et al.15 found significant differences between vertical and horizontal mattress sutures using Ethibond 2-0 in porcine menisci. Their data for Ethibond 2-0 horizontal mattress sutures were comparable to our findings (59.7 v 58 N). However, they were higher for vertical mattress sutures (89.3 v 63 N). This could be explained by the fact that they left the natural meniscotibial attachments intact ‘‘which might have absorbed some of the energy measured to produce failure of the repair’’ and which resulted in higher standard deviations. There are no data available in the literature concerning PDS 2-0 suture, although its clinical use is common.24-26 Testing of PDS 0 (116 N) and PDS 1 (146 N) vertical mattress sutures15 revealed higher loads to failure compared with our data for PDS 2-0 (63 N). To the contrary, the failure strength for horizontal mattress sutures using PDS showed only minimal increase of ultimate failure load (58 N for PDS 2-0, 66.1 N for PDS 0, and 73.8 for PDS 1) with increasing strength of the suture material. These data tend to confirm the findings of Post et al. that the choice of suture material is more important in the vertical mattress technique than in the horizontal mattress technique. The failure modes of vertical sutures showed only suture breakage. Tissue failures could only be observed with horizontal sutures. This confirms previous findings.15,19,20 However, failures seemed to occur less often than in other studies. An important increase of complete tissue failures can be observed comparing our data for PDS 2-0 (40% tissue failures) with those of PDS 0 and PDS 1 (90% tissue failures). Regarding these differences and the only minimal increase of ultimate failure strengths with stronger suture material, one could assume that the capacity of porcine meniscal tissue to resist to horizontal sutures is approximately 60 to 70 N, regardless of the suture material used.

509

There are no previous data available for cyclic testing of meniscal sutures. For this reason, the force levels at which the sutures were tested had to be defined. Koukoubis et al.27 showed that meniscus tears that had healed after suturing in dogs had an average load to failure of 46 N. Based on these findings, we considered that an ideal suture should resist to a maximum force of about 40 N until healing occurred, and we chose this value as an upper limit. However, the forces acting on the meniscal repair site in vivo are still unknown. Raunest and Derra28 analyzed the biomechanics of the repaired meniscus by studying load transmission and radial strain. They found that vertical sutures provided better stability than horizontal sutures, but they did not investigate the failure strengths of these sutures. They also observed the appearance of a gap of the tear site in the anterior horn with knee flexion and a simultaneous compression of the posterior horn. These findings are in contradiction with arthroscopic findings where an opening and closing of a posterior horn tear can be observed with flexion-extension movements. In a pilot study on 4 cadaver knees, Kirsch et al.29 investigated the forces applied to posterior horn sutures of the medial meniscus. They observed astonishingly small peak forces never exceeding 10 N. Even if this has to be confirmed by further studies, the forces acting on meniscus sutures in vivo might be lower than expected. However, these experiments were performed without weight bearing and could only partially imitate the in vivo conditions. Based on these findings, we performed the cyclic testing with a second group of specimens with an upper load of 20 N. Cyclic loading resulted in a 13% failure rate, mostly in the group with a maximum load of 40 N. But even in the group with lower maximum loads, failure of 1 horizontal suture could be observed. In the remaining sutures, the appearance of a gap could be observed between both parts of the meniscus during the first cycles. This gap was broader in the higher load group and with horizontal sutures. The last finding confirmed the fact that horizontal sutures provide lesser stability because less collagenous fibers are caught by the suture.19,30 Another finding confirming this theory was the significantly higher rate of partial tissue failures with horizontal sutures in the higher load group. Beyond these anatomic reasons, the viscoelastic properties of meniscal tissue2 and of the suture material might have been responsible for these findings. Analyzing the latter was not purpose of our study. Despite the findings of Mishra et al.,21 who found a 2- to 4-fold elongation of absorbable monofilament sutures com-

510

R. SEIL ET AL.

pared with nonabsorbable braided sutures, we did not find a difference between PDS and Ethibond sutures. A third reason for the elongation of the suture-meniscus system was knot slippage, which could be found in 6.6% of the cases. Slippage never exceeded 1 mm. This could have been more important if the knots had been tied arthroscopically. Especially for the Duncan and overhead loops, knot slippage could be observed in up to 30%.21 Loading to failure after cyclic loading showed similar failure strengths compared with single maximal loading. Regarding the failure modes, vertical sutures failed most often because of suture breakage. Horizontal sutures failed most often because of complete tissue failure. CONCLUSIONS Cyclic testing appeared not to change the failure strength of meniscal sutures. The ultimate failure strength of vertical sutures was similar to that of horizontal sutures. Cyclic testing showed that meniscal sutures may fail under repetitive loading and revealed the appearance of a gap of the suture site within the first cycles, irrespective of the suture type and material used. Even if the ultimate failure strength did not change after 100 load cycles, the appearance of the gap and suture disappearance on the meniscal surface resulting from partial tissue failures, which were both more pronounced in horizontal sutures, confirmed the superior resistance of meniscal tissue to vertical sutures. Although it is difficult to assess the clinical consequences of these findings, they indicate that a gap will appear in the repaired meniscus during repetitive loading that could lead to healing problems under conditions as yet unknown. REFERENCES 1. Barber FA, Stone RG. Meniscal repair, an arthroscopic technique. J Bone Joint Surg Br 1985;67:39-41. 2. Barber FA: Accelerated rehabilitation for meniscus repairs. Arthroscopy 1994;10:206-210. 3. Barber FA, Click SD. Meniscus repair rehabilitation with concurrent anterior cruciate reconstruction. Arthroscopy 1997; 13:433-437. 4. Barrett GR, Richardson K, Ruff CG, Jones A. The effect of suture type on meniscus repair. A clinical analysis. Am J Knee Surg 1997;10:2-9. 5. Buseck MS, Noyes FR. Arthroscopic evaluation of meniscal repairs after anterior cruciate ligament reconstruction and immediate motion. Am J Sports Med 1991;19:489-494. 6. Cannon WD, Vittori JM. The incidence of healing in arthroscopic meniscal repairs in anterior cruciate ligament reconstructed knees versus stable knees. Am J Sports Med 1992;20: 176-181.

7. Henning CE, Lynch MA, Yearout KM, Vequist SW, Stallbaumer RJ, Decker KA. Arthroscopic meniscal repair using an exogenous fibrin clot. Clin Orthop 1990;252:64-72. 8. Jakob RP, Sta¨ubli HU, Zuber K, Esser M. The arthroscopic meniscal repair. Techniques and clinical experience. Am J Sports Med 1988;16:137-142. 9. Lindenfeld TN. Arthroscopically aided meniscus repair. Am J Sports Med 1988;16:315-320. 10. Mariani PP, Santori N, Adriani E, Mastantuono M. Accelerated rehabilitation after arthroscopic meniscal repair: A clinical and magnetic resonance imaging investigation. Arthroscopy 1996; 12:680-686. 11. Miller DB. Arthroscopic meniscus repair. Am J Sports Med 1988;16:315-320. 12. Rosenberg TD, Scott SM, Coward DB, et al. Arthroscopic meniscal repair evaluated with repeat arthroscopy. Arthroscopy 1986;2:14-20. 13. Stone RG, Frewin PR, Gonzales S. Long-term assessment of arthroscopic meniscus repair: A two to six year follow-up study. Arthroscopy 1990;6:73-78. 14. DeHaven KE, Lohrer WA, Lovelock JE. Long-term results of open meniscal repair. Am J Sports Med 1995;23:524-530. 15. Post WR, Akers SR, Kish V. Load to failure of common meniscal repair techniques: Effects of suture technique and suture material. Arthroscopy 1997;13:731-736. 16. Albrecht-Olsen P, Lind T, Kristensen G, Falkenberg B. Failure strength of a new meniscus arrow repair technique: Biomechanical comparison with horizontal suture. Arthroscopy 1997;13: 183-187. 17. Asik M, Sener N, Akpinar S, Durmaz H, Go¨ksan A. Strength of different meniscus suturing techniques. Knee Surg Sports Traumatol Arthrosc 1997;5:80-83. 18. Dervin GF, Downing KJ, Keene GC, McBride DG. Failure strengths of suture versus biodegradable arrow for meniscal repair: an in vitro study. Arthroscopy 1997;13:296-300. 19. Kohn DM, Siebert W. Meniscus suture techniques: A comparative biomechanical cadaver study. Arthroscopy 1989;5:324327. 20. Rimmer MG, Nawana NS, Keene GC, Pearcy MJ. Failure strengths of different meniscal suturing techniques. Arthroscopy 1995;11:146-150. 21. Mishra DK, Cannon DW, Lucas DJ, Belzer JP. Elongation of arthroscopically tied knots. Am J Sports Med 1997;25:113-117. 22. Jaspers P, Lange A, Huiskes R, van Rens T. The mechanical function of the meniscus experiments on cadaveric pig kneejoints. Acta Orthop Belg 1980;46:663-668. 23. Nakano T, Aherne FX. Morphology and water and lipid contents of stifle menisci of growing swine. Can J Vet Res 1992;56:165-167. 24. Busenkell GL, Lee CS. Arthroscopic meniscal repair: A posterior cannulated technique. Arthroscopy 1992;8:247-253. 25. Esser RD. Arthroscopic meniscus repair: The easy way. Arthroscopy 1993;9:231-233. 26. Schulte KR, Fu FH. Meniscal repair using the inside-to-outside technique. Clin Sports Med 1996;15:455-467. 27. Koukoubis TD, Glisson RR, Feagin JA, Seaber AV, Schenkman D, Korompilias AV, Stahl DL. Meniscal fixation with an absorbable staple. An experimental study in dogs. Knee Surg Sports Traumatol Arthrosc 1997;5:22-30. 28. Raunest J, Derra E. Experimentelle Ergebnisse zur biomechanischen Belastbarkeit der Meniskusnaht im Bereich der Zone II. Unfallchirurg 1990;93:197-201. 29. Kirsch L, Kohn D, Glowik A. Forces in medial and lateral meniscus sutures during knee extension—An in vitro study. J Biomech 1999;31:104 (suppl 1). 30. Bullough PG, Munuera L, Murphy J, Weinstein AM. The strength of the menisci of the knee as it relates to their fine structure. J Bone Joint Surg Br 1979;52:564-570.