- Email: [email protected]
The Effects of Freezing on the Tensile Properties of Repaired Porcine Flexor Tendon
SCIENTIFIC ARTICLE
The Effects of Freezing on the Tensile Properties of Repaired Porcine Flexor Tendon Kieran M. Hirpara, MB, BCh, Paul J. Sullivan, MB, BCh, Michael E. O’Sullivan, MCh Purpose When conducting complex testing of tendon repairs, it is essential that the samples are adequately preserved to prevent degradation. Freezing of samples is the most convenient method of preservation; however, there is no evidence in the literature to prove that freezing tendon before or after repair is acceptable. We aimed to prove that freezing tendons does not significantly alter the results of linear load-to-failure testing of tendon repairs. Methods After a power study, 150 tendons were harvested from porcine forelimbs and randomized into 5 groups of 30 tendons. After division, tendons were repaired using a Pennington modified core technique with a Silfverskiöld peripheral cross-stitch. Tendons in group 1 were divided, repaired, and tested within 3 hours postmortem. Tendons in group 2 were refrigerated at 4°C for 24 hours prior to repair and testing. Tendons in group 3 were frozen at ⫺25°C for 3 months prior to repair and testing. Tendons in group 4 were frozen at ⫺25°C for 6 months prior to repair and testing. Tendons in group 5 were frozen at ⫺25°C for 6 months, repaired, refrozen for 1 month, and then tested. All repairs were linear load tested to ascertain the ultimate strength and force to produce 3-mm gap in the repair. Results Analysis of variance analysis of the results did not demonstrate any significant differences between groups. Conclusions Freezing tendons both before and after suture repair is an acceptable method of preservation when investigating the force to produce 3-mm gap and ultimate strength of tendon repairs. ( J Hand Surg 2008;33A:353 – 358. Copyright © 2008 by the American Society for Surgery of the Hand.) Key words Freezing, repair, tendon, tensile, porcine.
T
HE INVESTIGATION OF TENDON REPAIRS is becoming more complex and the time taken to run experimental protocols is increasing, making the scheduling of tendon harvest and testing increasingly challenging. When performing ex vivo tensile testing of tendon repair techniques, it is essential to ensure that samples are adequately preserved. To limit the number of variables between and during biomechanical testing of tendon repairs, it may be more efficient to preserve harvested and repaired tendons with subsequent testing in batches. The most common method of preserving tendon after harvest is freezing; however, with larger numbers, refreezing of samples after repair may also be required.1 Bhatia et al2
From The Department of Trauma and Orthopaedic Surgery and The Department of Plastic and Reconstructive Surgery, Galway Regional Hospitals, Galway, Ireland. Received for publication January 4, 2007; accepted in revised form December 17, 2007. Supported by Creganna Medical Devices and Galway. Corresponding author: Kieran M. Hirpara, MB, BCh, The Department of Trauma and Orthopaedic Surgery, Galway Regional Hospitals, Galway, Ireland; e-mail: [email protected]. 0363-5023/08/33A03-0011$34.00/0 doi:10.1016/j.jhsa.2007.12.011
suggested that freezing had no effect on tendon repairs but made no mention of whether this freezing was before or after repair and provided no results to support the claim. Despite the widespread use of freezing as part of experimental protocols, there is no evidence to substantiate its use when testing repaired tendons, and researchers wishing to validate their experimental protocols must extrapolate the effects of freezing from studies using intact tendons. Previous authors have shown that freezing tendons causes a subtle decrease in the elastic modulus3 or the ultimate strength,4 factors that only manifest themselves over the full extent of a stress-strain curve. As the stress-strain curve of a tendon repair only occupies a small portion of the curve for an intact tendon, these properties should not be evident. The response of the repaired tendon to freezing may be a function of the suture’s interaction with the tendon rather than the viscoelastic properties of the tendon itself. In this context, the results of previous studies may not support fully the freezing of tendons, and clarity is lacking on whether it is acceptable to freeze ex vivo tendons before or after repair or indeed both. The porcine tendon model is used extensively in the literature as it is readily available and similar to human flexor digitorum profundus tendon in zone II.5,6 The deep flexor tendon5–12 of the porcine forefoot is most commonly used, though some authors advocate use of extensor tendons.13–15 We aimed to investigate the effects of freezing and
© ASSH 䉬 Published by Elsevier, Inc. 䉬 353
354
TENSILE PROPERTIES OF FROZEN TENDON REPAIRS
refreezing of repaired porcine deep flexor tendon on the force to produce 3-mm gap, as described by Gelberman et al,16 and the load to failure (ultimate strength) of the repair. MATERIALS AND METHODS One hundred fifty profundus tendons were harvested from the central two rays of adult porcine forelimbs within 2 hours of slaughter. After harvest, the tendons were randomized into 5 groups, and each group of tendons was subjected to a different preservation protocol. During the experiments, tendons were kept wrapped in 0.9% saline– soaked swabs. The groups were as follows: ● ●
●
●
●
Group 1: Thirty tendons were harvested, divided, repaired, and tested within 3 hours of slaughter. Group 2: Thirty tendons were harvested and wrapped in an 0.9% saline–soaked swab. They were then left refrigerated for 24 hours at 4°C. The tendons were then transected, repaired, and tested within 2 hours. Group 3: Thirty tendons were harvested and wrapped in an 0.9% saline–soaked swab prior to freezing for 3 months and then thawed. The tendons were then transected, repaired, and tested within 2 hours. Group 4: Thirty tendons were harvested and wrapped in an 0.9% saline–soaked swab prior to freezing for 6 months and then thawed. The tendons were then transected, repaired, and tested within 2 hours. Group 5: Thirty tendons were harvested and wrapped in an 0.9% saline–soaked swab prior to freezing for 6 months. They were then thawed, repaired, and refrozen for 1 further month. The repaired tendons were then thawed and tested within 2 hours.
To freeze the samples, 10 tendons were wrapped in surgical gauze and placed into a polyethylene bag with a small volume of 0.9% saline solution, which was then frozen in a domestic freezer at ⫺25°C. Thawing involved warming the frozen tendons in a 37°C saline bath for 15 minutes, as previously described by Giannini et al.4 Whenever samples were not being repaired or tested, they were stored wrapped in a saline-soaked gauze. Tendons were divided using a scalpel at the level of the metatarsophalangeal joint, a level consistent with the literature6,7,11 and roughly equivalent to a zone II laceration.5,6 Prior to division, transverse lines were marked on the tendon surface using a piece of thread dipped in black ink (using the apparatus shown in Fig. 1), 2 further marks were placed 5 mm on either side of this line, and another 2 marks were placed 10 mm from this line as measured with a vernier caliper (Mitutoyo Corporation, Kawasaki, Kanagawa, Japan). The central mark was used to guide the division of the tendon, with the adjacent marks being used to guide suturing. The core repairs were performed as described by Pennington17 (Fig. 2) using 4-0 braided polyester with locking loops 10 mm from the cut end of the tendon (using the 10-mm ink mark as a guide). Peripheral repair was performed using the type B peripheral cross-stitch (Fig. 3) as
FIGURE 1: Device used for marking of tendons prior to repair.
FIGURE 2: Pennington modification of the Kessler 2-strand core repair.
FIGURE 3: Silfverskiöld type B peripheral cross-stitch.
described by Silfverskiöld and Andersson18 using 6-0 monofilament nylon with bites 5 mm from the repair site (using the 5-mm ink mark as a guide). All repairs were performed under 2.5⫻ loupe magnification by a single surgeon. Tensile testing was performed in a Zwick Tensiometer (Zwick GmbH & Co. KG, Ulm, Germany) with a 2.5 kN load cell and an interclamp distance of 60 mm. The tendons were secured in pneumatic clamps with coarse sandpaper grippers and then subjected to a 1 N preload followed by linear loading to failure at a rate of 10 mm/min. All tensile tests were recorded at 25 frames per second using a digital video camera linked to the tensiometer. The testing software (Zwick TestXPert 11.02) stored the
JHS 䉬 Vol A, March
355
TENSILE PROPERTIES OF FROZEN TENDON REPAIRS
TABLE 1: Results of Tensile Testing*
End Point Force to produce 3-mm gap Ultimate strength
Fresh (Group 1)
Refrigerated (Group 2)
Frozen 3 Months (Group 3)
Frozen 6 Months (Group 4)
Refrozen (Group 5)
39.0 (8.2) 60.0 (7.8)
36.1 (6.7) 61.5 (6.4)
33.8 (9.5) 60.6 (6.5)
36.7 (5.8) 61.191 (6.6)
38.6 (9.0) 60.2 (7.0)
*Means of repair strengths in newtons (standard deviation in parentheses).
TABLE 2: Results of Shapiro-Wilk W Testing for Normality of Data* Force to Produce 3-mm Gap
Ultimate Strength
Group
W Statistic
p Value
W Statistic
p Value
Fresh Refrigerated Frozen 3 months Frozen 6 months Refrozen
.945 .979 .962 .951 .945
.125 .784 .341 .180 .123
.985 .967 .946 .935 .961
.932 .469 .130 .066 .338
*A significant p value (⬍.05) means that data are unlikely to be normally distributed. The W statistic is the result of the Shapiro-Wilk W test. A small value of W indicates a departure from normality, and the significance of this result is given by the associated p value.
resultant video frames synchronized to time points on the force versus displacement curve, allowing subsequent assessment of the gap at various loads. A linear calibration tool was placed adjacent to the tendon in the field of view of the video camera, and a mirror was placed behind the tendon to allow recording of the back of the tendon. This methodology allowed us to assess the repair site gap on a frame by frame basis and determine the applied force required to produce a 3-mm gap. The ultimate strength of the repair was defined as the maximum force sustained by the repair as measured from the force versus displacement curve. Statistical Methods Prior to analysis, a power analysis was performed using the GⴱPower 3.0.4 program (Heinrich-Heine University, Du¨sseldorf, Germany).19 The power analysis was calculated using the “compromise” method as described by Faul et al.19 Analysis was performed using a medium effect size (f) of 0.25 as defined by Cohen20 and an ␣/ ratio of 1, showing that the power of our study was 0.842. All data were tested for normality using the Shapiro-Wilk W test prior to parametric analysis using 1-way analysis of variance (ANOVA). All statistics were performed using the StatsDirect 2.6.5 statistical package (StatsDirect Ltd., Altrincham, Cheshire, UK). RESULTS All core and peripheral repairs failed by suture breakage at or near the knot. The results for force to produce 3-mm gap
and ultimate strength are summarized in Table 1, and the results of normality testing are given in Table 2. Graphical representations of the results are shown in Figures 4 and 5. As all data were shown to be from a normal distribution, an ANOVA analysis was performed, and there was no significant difference between groups for the force to produce 3-mm gap (p ⫽ .091) or ultimate strength (p ⫽ .921). DISCUSSION The repair of digital flexor tendons has been extensively studied in the literature in an effort to increase the strength of repair. These studies are primarily a response to the rupture rate of tendon repairs subjected to early mobilization.21 An important element in research into tendon repairs is the analysis of cadaveric specimens subjected to repair in a linear load to failure model. Since the work of Smith,22 it has been recognized that degradation of fresh samples occurs quite rapidly without appropriate preservation. Smith22 found that, when studying rabbit cruciate ligaments, the samples become less extensible and react elastically to greater loads within 1 hour of death. He therefore advised that all tendon testing should be performed within 30 minutes of death. Matthews et al3 performed a more thorough investigation using rat tendon, finding that the length of tendon samples decreased in the first 3 hours postmortem and attributed this change to the absorption of the Ringer’s solution used to keep the samples moist. Matthews et al could not find any evidence that this
JHS 䉬 Vol A, March
356
TENSILE PROPERTIES OF FROZEN TENDON REPAIRS
FIGURE 4: Box-and-whiskers plot of force to produce 3-mm gap. The whiskers represent the range of the data, the boxes represent the standard deviation, and the diamond symbol (}) represents the mean.
FIGURE 5: Box-and-whiskers plot of ultimate strength. The whiskers represent the range of the data, the boxes represent the standard deviation, and the diamond symbol (}) represents the mean.
change in length was related to any alteration of stress-strain characteristics and therefore could not support the conclusions of Smith22; however, they advised against pooling results from fresh and frozen specimens.
Studies that are more recent have tried to analyze the structural changes of connective tissues subjected to freezing. Giannini et al4 studied the effects of ⫺80°C freezing on human posterior tibialis tendon. On transmission electron
JHS 䉬 Vol A, March
TENSILE PROPERTIES OF FROZEN TENDON REPAIRS
micrography, Giannini et al found an increase in the mean diameter of collagen fibrils and on the fibril nonoccupation mean ratio, with an associated decrease in the mean number of fibrils per unit area. This implies that frozen samples have less densely packed collagen fibrils that are larger in diameter. On light microscopy, the bundles of fibrils were less tightly packed and were wavier, a finding that was shared by Tsuchida et al23 when studying frozen rabbit patella tendon. Giannini et al concluded that when the tendons relaxed, there was elastic recoil of the collagen fibrils, which became larger in diameter. They also believed that freezing caused formation of ice crystals within the tendon substance leading to an increased nonoccupied space ratio and a less densely packed collagen structure. These conclusions are similar to those of several other authors in the literature.24 –26 Giannini et al4 also found that the principal biomechanical effects of freezing were a simultaneous decrease in ultimate load (due to splitting and fragmentation of collagen fibrils as seen on transmission electron microscopy) and ultimate tensile stress (due to the decrease in collagen fibril density). Despite these changes, they demonstrated that there was no effect on the Young’s modulus of the tendon as these properties decreased equally. The ultimate strength of intact tendons is far greater than that of tendon repairs, and so the alterations in the viscoelastic properties of intact tendons subjected to freezing should not be evident on the stress-strain curves of repaired samples. The theoretical concern is that the fragmentation of collagen fibrils may lead to a decreased grasping ability of the sutures and therefore to a reduction in the force to produce 3-mm gap and ultimate strength. In this study, a comparison was made between frozen tendons and fresh samples. Unlike previous studies,1– 4 a biomechanical analysis of divided and repaired tendons was performed, confirming that there was no significant difference in force to produce 3-mm gap or ultimate strength for tendons that were fresh, refrigerated for 24 hours, frozen for 3 months, or frozen for 6 months. Included in this work are tendons that had been frozen, thawed, repaired, and then refrozen. This group represents an increasingly likely situation where samples may require an extra cycle of freezing.27 There were no significant effects on force to produce 3-mm gap or ultimate strength when refreezing repaired tendons. As these are the recognized outcome measures for tendon repair models, it can be concluded that it is acceptable to harvest tendons and preserve them as a “bank” of samples for later use. These tendons may then be thawed and subjected to repair as appropriate. In order to maximize the power of the study and its relevance to previous research, certain limitations had to be made. Only the effects of freezing on tendons subjected to repair using both a core and peripheral stitch (as opposed to studying the effect of these components individually) were analyzed. This can be justified, as the addition of a
357
peripheral suture to a core repair is considered essential in routine clinical practice. By limiting the study in this way, the group size was increased while decreasing the overall number of groups, thereby increasing the study’s power. The Pennington 2-strand core technique was selected as most studies in the literature comparing tendon repairs use this technique as a control.6,12,13,22,28 –32 Clinically, the 4strand repairs are considered to be the standard of care33,34; however the Pennington-modified Kessler is an ideal suture for this study as it allows comparison of results with data previously reported in the literature. Another limitation of this work is its use of the porcine flexor tendon model. Although the literature makes extensive use of porcine tendon and claims it is similar to human tendon, there may have been a different outcome if the experiments were repeated with human samples. As such, our results can only be applied to future experimentation on the porcine model. It is also possible (though unlikely) that the storage of tendons in a swab soaked in 0.9% saline may have affected the structure of the tendon, resulting in changes common to all of the groups that would mask any changes caused by freezing. These results prove that there are no effects on the force to produce 3-mm gap or ultimate strength when tendons are frozen or refrozen. Theoretically, this means that data from frozen and fresh samples can be pooled; however, good experimental practice requires that any such variables should be controlled. We therefore advise that freezing of tendons is acceptable when performing linear load to failure testing of tendon repairs. We would also advise that a preservation protocol should be defined prior to tendon harvesting and adhered to throughout any experimentation. REFERENCES 1.Moon DK, Woo SLY, Takakura Y, Gabriel MT, Abramowitch SD. The effects of refreezing on the viscoelastic and tensile properties of ligaments. J Biomech 2006;39:1153–1157. 2.Bhatia D, Tanner KE, Bonfield W, Citron ND. Factors affecting the strength of flexor tendon repair. J Hand Surg 1992;17B:550 –552. 3.Matthews LS, Ellis D. Viscoelastic properties of cat tendon: effects of time after death and preservation by freezing. J Biomech 1968;1–2:65–71. 4.Giannini S, Buda R, Di Caprio F, Agati P, Bigi A, De Pasquale V, Ruggeri A. Effects of freezing on the biomechanical and structural properties of human posterior tibial tendons. Int Orthop 2007. 5.Zatiti S, Mazzer N, Barbieri C. Mechanical strengths of tendon sutures: an in vitro comparative study of six techniques. J Hand Surg 1998;23B:228 –233. 6.Wang B, Xie RG, Tang JB. Biomechanical analysis of a modification of Tang method of tendon repair. J Hand Surg 2003;28B:347–350. 7.Tan J, Wang B, Tan B, Xu Y, Tang JB. Changes in tendon strength after partial cut and effects of running peripheral sutures. J Hand Surg 2003;28B:478 – 482.
JHS 䉬 Vol A, March
358
TENSILE PROPERTIES OF FROZEN TENDON REPAIRS
8.Mishra V, Kuiper JH, Kelly CP. Influence of core suture material and peripheral repair technique on the strength of Kessler flexor tendon repair. J Hand Surg 2003;28B:357–362. 9.Tang JB, Zhang Y, Cao Y, Xie RG. Core suture purchase affects strength of tendon repairs. J Hand Surg 2005;30A: 1262–1266. 10.Smith AM, Evans DM. Biomechanical assessment of a new type of flexor tendon repair. J Hand Surg 2001:26B:217– 219. 11.Cao Y, Xie RG, Tang JB. Dorsal-enhanced sutures iimprove tension resistance of tendon repair. J Hand Surg 2002;27B:161–164. 12.Robertson GA, Al-Qattan MM. A biomechanical analysis of a new interlock suture technique for flexor tendon repair. J Hand Surg 1992;17B:92–93. 13.Savage R. In vitro studies of a new method of flexor tendon repair. J Hand Surg 1985;10B:135–141. 14.Viinikainen A, Goransson H, Huovinen K, Kellomaki M, Rokkanen P. A comparative analysis of the biomechanical behaviour of five flexor tendon core sutures. J Hand Surg 2004;29B:536 –543. 15.Haddad RJ, Kester MA, McCluskey GM, Brunet ME, Cook SD. Comparative mechanical analysis of a looped-suture tendon repair. J Hand Surg 1988;13A:709 –713. 16.Gelberman RH, Boyer MI, Brodt MD, Winters SC, Silva MJ. The effect of gap formation at the repair site on the strength and excursion of intrasynovial flexor tendons. An experimental study on the early stages of tendon-healing in dogs. J Bone Joint Surg 1999;81A:975–982. 17.Pennington DG. The locking loop tendon suture. Plast Reconstr Surg 1979;63:648 – 652. 18.Silfverskiöld KL, Andersson CH. Two new methods of tendon repair: an in vitro evaluation of tensile strength and gap formation. J Hand Surg 1993;18A:58 – 65. 19.Faul F, Erdfelder E, Lang AG, Buchner A. GⴱPower 3: a flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behav Res Methods 2007;39:175–191. 20.Cohen J. Statistical power for the behavioral sciences. 2nd ed. Hillsdale, NJ: Lawrence Earlbaum Associates, 1988. 21.Elliot D, Moiemen NS, Flemming AF, Harris SB, Foster AJ. The rupture rate of acute flexor tendon repairs mobilized by the controlled active motion regimen. J Hand Surg 1994; 19B:607– 612.
22.Smith JW. The elastic properties of the anterior cruciate ligament of the rabbit. J Anat 1954;88:369 –380. 23.Tsuchida T, Yasuda K, Kaneda K, Hayashi K, Yamamoto N, Miyakawa K, Tanaka K. Effects of in situ freezing and stress-shielding on the ultrastructure of rabbit patellar tendon. J Orthop Res 1997;15:904 –910. 24.Graf BK, Fujisaki K, Vanderby R, Vailas AC. The effect of in situ freezing on rabbit patellar tendon (a histologic, biochemical and biomechanical analysis). Am J Sports Med 1992;20:401– 405. 25.Jackson DW, Grood ES, Cohn BT, Arnoczky SP, Simon TM, Cummings JF. The effects of in situ freezing on the anterior cruciate ligament. An experimental study in goats. J Bone Joint Surg 1991;73A:201–213. 26.Ohno K, Yasuda K, Yamamoto N, Kaneda K, Hayashi K. Effects of complete stress-shielding on the mechanical properties and histology of in situ frozen patellar tendon. J Orthop Res 1993;11:592– 602. 27.Shoemaker SC, Markolf KL. In vitro rotator knee stability. Ligamentous and muscular contributions. J Bone Joint Surg 1982;64A:208 –216. 28.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. 29.Barrie KA, Wolfe SW, Shean C, Shenbagamurthi D, Slade JF, Panjabi MM. A biomechanical comparison of multistrand flexor tendon repairs using an in situ testing model. J Hand Surg 2000;25A:499 –506. 30.Shaieb M, Singer D. Tensile strengths of various suture techniques. J Hand Surg 1997;22B:764 –767. 31.McLarney E, Hoffman H, Wolfe SW. Biomechanical analysis of the cruciate four-strand flexor tendon repair. J Hand Surg 1999;24A:295–301. 32.Stein T, Ali A, Hamman J, Mass DP. A randomized biomechanical study of zone II human flexor tendon repairs analyzed in a linear model. J Hand Surg 1998;23A:1043– 1045. 33.Strickland J. Flexor tendon injuries: I. Foundations of treatment. J Am Acad Orthop Surg 1995;3:44 –54. 34.Amadio P, An K, Ejeskär A, Guimberteau JC, Harris S, Savage R, et al. IFSSH flexor tendon committee report. J Hand Surg 2005;30B:100 –116.
JHS 䉬 Vol A, March
The Effects of Freezing on the Tensile Properties of Repaired Porcine Flexor Tendon Kieran M. Hirpara, MB, BCh, Paul J. Sullivan, MB, BCh, Michael E. O’Sullivan, MCh Purpose When conducting complex testing of tendon repairs, it is essential that the samples are adequately preserved to prevent degradation. Freezing of samples is the most convenient method of preservation; however, there is no evidence in the literature to prove that freezing tendon before or after repair is acceptable. We aimed to prove that freezing tendons does not significantly alter the results of linear load-to-failure testing of tendon repairs. Methods After a power study, 150 tendons were harvested from porcine forelimbs and randomized into 5 groups of 30 tendons. After division, tendons were repaired using a Pennington modified core technique with a Silfverskiöld peripheral cross-stitch. Tendons in group 1 were divided, repaired, and tested within 3 hours postmortem. Tendons in group 2 were refrigerated at 4°C for 24 hours prior to repair and testing. Tendons in group 3 were frozen at ⫺25°C for 3 months prior to repair and testing. Tendons in group 4 were frozen at ⫺25°C for 6 months prior to repair and testing. Tendons in group 5 were frozen at ⫺25°C for 6 months, repaired, refrozen for 1 month, and then tested. All repairs were linear load tested to ascertain the ultimate strength and force to produce 3-mm gap in the repair. Results Analysis of variance analysis of the results did not demonstrate any significant differences between groups. Conclusions Freezing tendons both before and after suture repair is an acceptable method of preservation when investigating the force to produce 3-mm gap and ultimate strength of tendon repairs. ( J Hand Surg 2008;33A:353 – 358. Copyright © 2008 by the American Society for Surgery of the Hand.) Key words Freezing, repair, tendon, tensile, porcine.
T
HE INVESTIGATION OF TENDON REPAIRS is becoming more complex and the time taken to run experimental protocols is increasing, making the scheduling of tendon harvest and testing increasingly challenging. When performing ex vivo tensile testing of tendon repair techniques, it is essential to ensure that samples are adequately preserved. To limit the number of variables between and during biomechanical testing of tendon repairs, it may be more efficient to preserve harvested and repaired tendons with subsequent testing in batches. The most common method of preserving tendon after harvest is freezing; however, with larger numbers, refreezing of samples after repair may also be required.1 Bhatia et al2
From The Department of Trauma and Orthopaedic Surgery and The Department of Plastic and Reconstructive Surgery, Galway Regional Hospitals, Galway, Ireland. Received for publication January 4, 2007; accepted in revised form December 17, 2007. Supported by Creganna Medical Devices and Galway. Corresponding author: Kieran M. Hirpara, MB, BCh, The Department of Trauma and Orthopaedic Surgery, Galway Regional Hospitals, Galway, Ireland; e-mail: [email protected]. 0363-5023/08/33A03-0011$34.00/0 doi:10.1016/j.jhsa.2007.12.011
suggested that freezing had no effect on tendon repairs but made no mention of whether this freezing was before or after repair and provided no results to support the claim. Despite the widespread use of freezing as part of experimental protocols, there is no evidence to substantiate its use when testing repaired tendons, and researchers wishing to validate their experimental protocols must extrapolate the effects of freezing from studies using intact tendons. Previous authors have shown that freezing tendons causes a subtle decrease in the elastic modulus3 or the ultimate strength,4 factors that only manifest themselves over the full extent of a stress-strain curve. As the stress-strain curve of a tendon repair only occupies a small portion of the curve for an intact tendon, these properties should not be evident. The response of the repaired tendon to freezing may be a function of the suture’s interaction with the tendon rather than the viscoelastic properties of the tendon itself. In this context, the results of previous studies may not support fully the freezing of tendons, and clarity is lacking on whether it is acceptable to freeze ex vivo tendons before or after repair or indeed both. The porcine tendon model is used extensively in the literature as it is readily available and similar to human flexor digitorum profundus tendon in zone II.5,6 The deep flexor tendon5–12 of the porcine forefoot is most commonly used, though some authors advocate use of extensor tendons.13–15 We aimed to investigate the effects of freezing and
© ASSH 䉬 Published by Elsevier, Inc. 䉬 353
354
TENSILE PROPERTIES OF FROZEN TENDON REPAIRS
refreezing of repaired porcine deep flexor tendon on the force to produce 3-mm gap, as described by Gelberman et al,16 and the load to failure (ultimate strength) of the repair. MATERIALS AND METHODS One hundred fifty profundus tendons were harvested from the central two rays of adult porcine forelimbs within 2 hours of slaughter. After harvest, the tendons were randomized into 5 groups, and each group of tendons was subjected to a different preservation protocol. During the experiments, tendons were kept wrapped in 0.9% saline– soaked swabs. The groups were as follows: ● ●
●
●
●
Group 1: Thirty tendons were harvested, divided, repaired, and tested within 3 hours of slaughter. Group 2: Thirty tendons were harvested and wrapped in an 0.9% saline–soaked swab. They were then left refrigerated for 24 hours at 4°C. The tendons were then transected, repaired, and tested within 2 hours. Group 3: Thirty tendons were harvested and wrapped in an 0.9% saline–soaked swab prior to freezing for 3 months and then thawed. The tendons were then transected, repaired, and tested within 2 hours. Group 4: Thirty tendons were harvested and wrapped in an 0.9% saline–soaked swab prior to freezing for 6 months and then thawed. The tendons were then transected, repaired, and tested within 2 hours. Group 5: Thirty tendons were harvested and wrapped in an 0.9% saline–soaked swab prior to freezing for 6 months. They were then thawed, repaired, and refrozen for 1 further month. The repaired tendons were then thawed and tested within 2 hours.
To freeze the samples, 10 tendons were wrapped in surgical gauze and placed into a polyethylene bag with a small volume of 0.9% saline solution, which was then frozen in a domestic freezer at ⫺25°C. Thawing involved warming the frozen tendons in a 37°C saline bath for 15 minutes, as previously described by Giannini et al.4 Whenever samples were not being repaired or tested, they were stored wrapped in a saline-soaked gauze. Tendons were divided using a scalpel at the level of the metatarsophalangeal joint, a level consistent with the literature6,7,11 and roughly equivalent to a zone II laceration.5,6 Prior to division, transverse lines were marked on the tendon surface using a piece of thread dipped in black ink (using the apparatus shown in Fig. 1), 2 further marks were placed 5 mm on either side of this line, and another 2 marks were placed 10 mm from this line as measured with a vernier caliper (Mitutoyo Corporation, Kawasaki, Kanagawa, Japan). The central mark was used to guide the division of the tendon, with the adjacent marks being used to guide suturing. The core repairs were performed as described by Pennington17 (Fig. 2) using 4-0 braided polyester with locking loops 10 mm from the cut end of the tendon (using the 10-mm ink mark as a guide). Peripheral repair was performed using the type B peripheral cross-stitch (Fig. 3) as
FIGURE 1: Device used for marking of tendons prior to repair.
FIGURE 2: Pennington modification of the Kessler 2-strand core repair.
FIGURE 3: Silfverskiöld type B peripheral cross-stitch.
described by Silfverskiöld and Andersson18 using 6-0 monofilament nylon with bites 5 mm from the repair site (using the 5-mm ink mark as a guide). All repairs were performed under 2.5⫻ loupe magnification by a single surgeon. Tensile testing was performed in a Zwick Tensiometer (Zwick GmbH & Co. KG, Ulm, Germany) with a 2.5 kN load cell and an interclamp distance of 60 mm. The tendons were secured in pneumatic clamps with coarse sandpaper grippers and then subjected to a 1 N preload followed by linear loading to failure at a rate of 10 mm/min. All tensile tests were recorded at 25 frames per second using a digital video camera linked to the tensiometer. The testing software (Zwick TestXPert 11.02) stored the
JHS 䉬 Vol A, March
355
TENSILE PROPERTIES OF FROZEN TENDON REPAIRS
TABLE 1: Results of Tensile Testing*
End Point Force to produce 3-mm gap Ultimate strength
Fresh (Group 1)
Refrigerated (Group 2)
Frozen 3 Months (Group 3)
Frozen 6 Months (Group 4)
Refrozen (Group 5)
39.0 (8.2) 60.0 (7.8)
36.1 (6.7) 61.5 (6.4)
33.8 (9.5) 60.6 (6.5)
36.7 (5.8) 61.191 (6.6)
38.6 (9.0) 60.2 (7.0)
*Means of repair strengths in newtons (standard deviation in parentheses).
TABLE 2: Results of Shapiro-Wilk W Testing for Normality of Data* Force to Produce 3-mm Gap
Ultimate Strength
Group
W Statistic
p Value
W Statistic
p Value
Fresh Refrigerated Frozen 3 months Frozen 6 months Refrozen
.945 .979 .962 .951 .945
.125 .784 .341 .180 .123
.985 .967 .946 .935 .961
.932 .469 .130 .066 .338
*A significant p value (⬍.05) means that data are unlikely to be normally distributed. The W statistic is the result of the Shapiro-Wilk W test. A small value of W indicates a departure from normality, and the significance of this result is given by the associated p value.
resultant video frames synchronized to time points on the force versus displacement curve, allowing subsequent assessment of the gap at various loads. A linear calibration tool was placed adjacent to the tendon in the field of view of the video camera, and a mirror was placed behind the tendon to allow recording of the back of the tendon. This methodology allowed us to assess the repair site gap on a frame by frame basis and determine the applied force required to produce a 3-mm gap. The ultimate strength of the repair was defined as the maximum force sustained by the repair as measured from the force versus displacement curve. Statistical Methods Prior to analysis, a power analysis was performed using the GⴱPower 3.0.4 program (Heinrich-Heine University, Du¨sseldorf, Germany).19 The power analysis was calculated using the “compromise” method as described by Faul et al.19 Analysis was performed using a medium effect size (f) of 0.25 as defined by Cohen20 and an ␣/ ratio of 1, showing that the power of our study was 0.842. All data were tested for normality using the Shapiro-Wilk W test prior to parametric analysis using 1-way analysis of variance (ANOVA). All statistics were performed using the StatsDirect 2.6.5 statistical package (StatsDirect Ltd., Altrincham, Cheshire, UK). RESULTS All core and peripheral repairs failed by suture breakage at or near the knot. The results for force to produce 3-mm gap
and ultimate strength are summarized in Table 1, and the results of normality testing are given in Table 2. Graphical representations of the results are shown in Figures 4 and 5. As all data were shown to be from a normal distribution, an ANOVA analysis was performed, and there was no significant difference between groups for the force to produce 3-mm gap (p ⫽ .091) or ultimate strength (p ⫽ .921). DISCUSSION The repair of digital flexor tendons has been extensively studied in the literature in an effort to increase the strength of repair. These studies are primarily a response to the rupture rate of tendon repairs subjected to early mobilization.21 An important element in research into tendon repairs is the analysis of cadaveric specimens subjected to repair in a linear load to failure model. Since the work of Smith,22 it has been recognized that degradation of fresh samples occurs quite rapidly without appropriate preservation. Smith22 found that, when studying rabbit cruciate ligaments, the samples become less extensible and react elastically to greater loads within 1 hour of death. He therefore advised that all tendon testing should be performed within 30 minutes of death. Matthews et al3 performed a more thorough investigation using rat tendon, finding that the length of tendon samples decreased in the first 3 hours postmortem and attributed this change to the absorption of the Ringer’s solution used to keep the samples moist. Matthews et al could not find any evidence that this
JHS 䉬 Vol A, March
356
TENSILE PROPERTIES OF FROZEN TENDON REPAIRS
FIGURE 4: Box-and-whiskers plot of force to produce 3-mm gap. The whiskers represent the range of the data, the boxes represent the standard deviation, and the diamond symbol (}) represents the mean.
FIGURE 5: Box-and-whiskers plot of ultimate strength. The whiskers represent the range of the data, the boxes represent the standard deviation, and the diamond symbol (}) represents the mean.
change in length was related to any alteration of stress-strain characteristics and therefore could not support the conclusions of Smith22; however, they advised against pooling results from fresh and frozen specimens.
Studies that are more recent have tried to analyze the structural changes of connective tissues subjected to freezing. Giannini et al4 studied the effects of ⫺80°C freezing on human posterior tibialis tendon. On transmission electron
JHS 䉬 Vol A, March
TENSILE PROPERTIES OF FROZEN TENDON REPAIRS
micrography, Giannini et al found an increase in the mean diameter of collagen fibrils and on the fibril nonoccupation mean ratio, with an associated decrease in the mean number of fibrils per unit area. This implies that frozen samples have less densely packed collagen fibrils that are larger in diameter. On light microscopy, the bundles of fibrils were less tightly packed and were wavier, a finding that was shared by Tsuchida et al23 when studying frozen rabbit patella tendon. Giannini et al concluded that when the tendons relaxed, there was elastic recoil of the collagen fibrils, which became larger in diameter. They also believed that freezing caused formation of ice crystals within the tendon substance leading to an increased nonoccupied space ratio and a less densely packed collagen structure. These conclusions are similar to those of several other authors in the literature.24 –26 Giannini et al4 also found that the principal biomechanical effects of freezing were a simultaneous decrease in ultimate load (due to splitting and fragmentation of collagen fibrils as seen on transmission electron microscopy) and ultimate tensile stress (due to the decrease in collagen fibril density). Despite these changes, they demonstrated that there was no effect on the Young’s modulus of the tendon as these properties decreased equally. The ultimate strength of intact tendons is far greater than that of tendon repairs, and so the alterations in the viscoelastic properties of intact tendons subjected to freezing should not be evident on the stress-strain curves of repaired samples. The theoretical concern is that the fragmentation of collagen fibrils may lead to a decreased grasping ability of the sutures and therefore to a reduction in the force to produce 3-mm gap and ultimate strength. In this study, a comparison was made between frozen tendons and fresh samples. Unlike previous studies,1– 4 a biomechanical analysis of divided and repaired tendons was performed, confirming that there was no significant difference in force to produce 3-mm gap or ultimate strength for tendons that were fresh, refrigerated for 24 hours, frozen for 3 months, or frozen for 6 months. Included in this work are tendons that had been frozen, thawed, repaired, and then refrozen. This group represents an increasingly likely situation where samples may require an extra cycle of freezing.27 There were no significant effects on force to produce 3-mm gap or ultimate strength when refreezing repaired tendons. As these are the recognized outcome measures for tendon repair models, it can be concluded that it is acceptable to harvest tendons and preserve them as a “bank” of samples for later use. These tendons may then be thawed and subjected to repair as appropriate. In order to maximize the power of the study and its relevance to previous research, certain limitations had to be made. Only the effects of freezing on tendons subjected to repair using both a core and peripheral stitch (as opposed to studying the effect of these components individually) were analyzed. This can be justified, as the addition of a
357
peripheral suture to a core repair is considered essential in routine clinical practice. By limiting the study in this way, the group size was increased while decreasing the overall number of groups, thereby increasing the study’s power. The Pennington 2-strand core technique was selected as most studies in the literature comparing tendon repairs use this technique as a control.6,12,13,22,28 –32 Clinically, the 4strand repairs are considered to be the standard of care33,34; however the Pennington-modified Kessler is an ideal suture for this study as it allows comparison of results with data previously reported in the literature. Another limitation of this work is its use of the porcine flexor tendon model. Although the literature makes extensive use of porcine tendon and claims it is similar to human tendon, there may have been a different outcome if the experiments were repeated with human samples. As such, our results can only be applied to future experimentation on the porcine model. It is also possible (though unlikely) that the storage of tendons in a swab soaked in 0.9% saline may have affected the structure of the tendon, resulting in changes common to all of the groups that would mask any changes caused by freezing. These results prove that there are no effects on the force to produce 3-mm gap or ultimate strength when tendons are frozen or refrozen. Theoretically, this means that data from frozen and fresh samples can be pooled; however, good experimental practice requires that any such variables should be controlled. We therefore advise that freezing of tendons is acceptable when performing linear load to failure testing of tendon repairs. We would also advise that a preservation protocol should be defined prior to tendon harvesting and adhered to throughout any experimentation. REFERENCES 1.Moon DK, Woo SLY, Takakura Y, Gabriel MT, Abramowitch SD. The effects of refreezing on the viscoelastic and tensile properties of ligaments. J Biomech 2006;39:1153–1157. 2.Bhatia D, Tanner KE, Bonfield W, Citron ND. Factors affecting the strength of flexor tendon repair. J Hand Surg 1992;17B:550 –552. 3.Matthews LS, Ellis D. Viscoelastic properties of cat tendon: effects of time after death and preservation by freezing. J Biomech 1968;1–2:65–71. 4.Giannini S, Buda R, Di Caprio F, Agati P, Bigi A, De Pasquale V, Ruggeri A. Effects of freezing on the biomechanical and structural properties of human posterior tibial tendons. Int Orthop 2007. 5.Zatiti S, Mazzer N, Barbieri C. Mechanical strengths of tendon sutures: an in vitro comparative study of six techniques. J Hand Surg 1998;23B:228 –233. 6.Wang B, Xie RG, Tang JB. Biomechanical analysis of a modification of Tang method of tendon repair. J Hand Surg 2003;28B:347–350. 7.Tan J, Wang B, Tan B, Xu Y, Tang JB. Changes in tendon strength after partial cut and effects of running peripheral sutures. J Hand Surg 2003;28B:478 – 482.
JHS 䉬 Vol A, March
358
TENSILE PROPERTIES OF FROZEN TENDON REPAIRS
8.Mishra V, Kuiper JH, Kelly CP. Influence of core suture material and peripheral repair technique on the strength of Kessler flexor tendon repair. J Hand Surg 2003;28B:357–362. 9.Tang JB, Zhang Y, Cao Y, Xie RG. Core suture purchase affects strength of tendon repairs. J Hand Surg 2005;30A: 1262–1266. 10.Smith AM, Evans DM. Biomechanical assessment of a new type of flexor tendon repair. J Hand Surg 2001:26B:217– 219. 11.Cao Y, Xie RG, Tang JB. Dorsal-enhanced sutures iimprove tension resistance of tendon repair. J Hand Surg 2002;27B:161–164. 12.Robertson GA, Al-Qattan MM. A biomechanical analysis of a new interlock suture technique for flexor tendon repair. J Hand Surg 1992;17B:92–93. 13.Savage R. In vitro studies of a new method of flexor tendon repair. J Hand Surg 1985;10B:135–141. 14.Viinikainen A, Goransson H, Huovinen K, Kellomaki M, Rokkanen P. A comparative analysis of the biomechanical behaviour of five flexor tendon core sutures. J Hand Surg 2004;29B:536 –543. 15.Haddad RJ, Kester MA, McCluskey GM, Brunet ME, Cook SD. Comparative mechanical analysis of a looped-suture tendon repair. J Hand Surg 1988;13A:709 –713. 16.Gelberman RH, Boyer MI, Brodt MD, Winters SC, Silva MJ. The effect of gap formation at the repair site on the strength and excursion of intrasynovial flexor tendons. An experimental study on the early stages of tendon-healing in dogs. J Bone Joint Surg 1999;81A:975–982. 17.Pennington DG. The locking loop tendon suture. Plast Reconstr Surg 1979;63:648 – 652. 18.Silfverskiöld KL, Andersson CH. Two new methods of tendon repair: an in vitro evaluation of tensile strength and gap formation. J Hand Surg 1993;18A:58 – 65. 19.Faul F, Erdfelder E, Lang AG, Buchner A. GⴱPower 3: a flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behav Res Methods 2007;39:175–191. 20.Cohen J. Statistical power for the behavioral sciences. 2nd ed. Hillsdale, NJ: Lawrence Earlbaum Associates, 1988. 21.Elliot D, Moiemen NS, Flemming AF, Harris SB, Foster AJ. The rupture rate of acute flexor tendon repairs mobilized by the controlled active motion regimen. J Hand Surg 1994; 19B:607– 612.
22.Smith JW. The elastic properties of the anterior cruciate ligament of the rabbit. J Anat 1954;88:369 –380. 23.Tsuchida T, Yasuda K, Kaneda K, Hayashi K, Yamamoto N, Miyakawa K, Tanaka K. Effects of in situ freezing and stress-shielding on the ultrastructure of rabbit patellar tendon. J Orthop Res 1997;15:904 –910. 24.Graf BK, Fujisaki K, Vanderby R, Vailas AC. The effect of in situ freezing on rabbit patellar tendon (a histologic, biochemical and biomechanical analysis). Am J Sports Med 1992;20:401– 405. 25.Jackson DW, Grood ES, Cohn BT, Arnoczky SP, Simon TM, Cummings JF. The effects of in situ freezing on the anterior cruciate ligament. An experimental study in goats. J Bone Joint Surg 1991;73A:201–213. 26.Ohno K, Yasuda K, Yamamoto N, Kaneda K, Hayashi K. Effects of complete stress-shielding on the mechanical properties and histology of in situ frozen patellar tendon. J Orthop Res 1993;11:592– 602. 27.Shoemaker SC, Markolf KL. In vitro rotator knee stability. Ligamentous and muscular contributions. J Bone Joint Surg 1982;64A:208 –216. 28.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. 29.Barrie KA, Wolfe SW, Shean C, Shenbagamurthi D, Slade JF, Panjabi MM. A biomechanical comparison of multistrand flexor tendon repairs using an in situ testing model. J Hand Surg 2000;25A:499 –506. 30.Shaieb M, Singer D. Tensile strengths of various suture techniques. J Hand Surg 1997;22B:764 –767. 31.McLarney E, Hoffman H, Wolfe SW. Biomechanical analysis of the cruciate four-strand flexor tendon repair. J Hand Surg 1999;24A:295–301. 32.Stein T, Ali A, Hamman J, Mass DP. A randomized biomechanical study of zone II human flexor tendon repairs analyzed in a linear model. J Hand Surg 1998;23A:1043– 1045. 33.Strickland J. Flexor tendon injuries: I. Foundations of treatment. J Am Acad Orthop Surg 1995;3:44 –54. 34.Amadio P, An K, Ejeskär A, Guimberteau JC, Harris S, Savage R, et al. IFSSH flexor tendon committee report. J Hand Surg 2005;30B:100 –116.
JHS 䉬 Vol A, March