- Email: [email protected]
Electrothermal shrinkage reduces laxity but alters creep behavior in a lapine ligament model
ORIGINAL ARTICLES Electrothermal shrinkage reduces laxity but alters creep behavior in a lapine ligament model Andrew L. Wallace, MBBS, PhD, FRACS(Orth), Robert M. Hollinshead, MD, FRCSC, and Cyril B. Frank, MD, FRCSC, London, United Kingdom, and Calgary, Alberta, Canada
Thermal denaturation of collagen in ligament tissue has the potential to enhance arthroscopic shoulder stabilization. Previous studies have shown that laser energy produces significant capsular shortening without alteration of viscoelastic properties, but little information is available on the effects of radio frequency electrothermal energy. We assessed the acute effects of radio frequency shrinkage with use of the lapine medial collateral ligament model, in which the tibial insertion was shifted proximally to produce abnormal laxity. Thermal treatment resulted in restoration of laxity from 3.33 ± 0.25 mm to 0.66 ± 0.31 mm, which was not significantly different from medial collateral ligaments replaced anatomically (0.50 ± 0.34 mm). When tested at 4.1 megapascals, cyclic and static creep strains were increased twofold to threefold in thermally-treated ligaments (P < .01), and partial failure occurred in 2 of 8 cases. We conclude that radio frequency electrothermal shrinkage is effective at reducing laxity but significantly alters viscoelastic properties, posing a risk of recurrent stretching-out at “physiological” loads. (J Shoulder Elbow Surg 2001;10:1-6.)
From McCaig Centre for Joint Injury and Arthritis Research, University of Calgary, Alberta, Canada. Nominated for the Charles S. Neer Award in Basic Science, American Shoulder and Elbow Surgeons Meeting, Anaheim, Calif, February 1999. Reprint requests: Andrew L. Wallace, MBBS, PhD, FRACS(Orth), Senior Lecturer/Honorary Consultant Orthopaedic Surgeon, Imperial College School of Medicine, Charing Cross Hospital, Fulham Palace Road, London W6 8RF, United Kingdom (Email: [email protected]). Copyright © 2001 by Journal of Shoulder and Elbow Surgery Board of Trustees. 1058-2746/2001/$35.00 + 0 32/1/112023 doi:10.1067/mse.2001.112023
INTRODUCTION Instability of the shoulder remains a common clinical presentation in young people, particularly as a consequence of athletic activity. Increasingly, the expectation in those requiring surgical management of the unstable shoulder is for minimal perioperative morbidity, expedient recovery, no recurrence of symptoms, and rehabilitation facilitating a rapid return to full activity without loss of motion. Recent technical advances in arthroscopic stabilization have brought this ideal within reach, but published results to date have not proven superior to those reported for open reconstruction.12 Difficulty in reducing excess capsular laxity at the time of surgery and recurrence of laxity with time have been cited as the major reasons for failure.11,16 Thermal shrinkage of collagen is a novel development that has considerable potential for dealing with this problem. Tissue heating, achieved either by laser light irradiation or application of radio frequency electric current, causes shrinkage by denaturation of the collagen triple helix at a critical temperature, usually 60°C to 65°C for mammalian tissue.22 In vitro studies have demonstrated that shortening of up to 50% can occur,4 though this was accompanied by significant reductions in high-load structural properties. Immediately after heating, tensile strength of isolated specimens was reduced by 30%, whereas tensile stiffness was reduced by up to 90%.4,21 Few studies have attempted to document lowload mechanical behavior, though it appears that ligament stresses during normal functional activity in vivo are typically less than 10% of ultimate tensile stress.6 Creep refers to the time-dependent increase in deformation seen in certain materials under constant cyclic or static loads, and this phenomenon may play a role in the development of recurrent laxity seen after ligament repair or autograft reconstruction.7,8 Unfortunately, at present there are no established animal models with sufficient acquired ligament laxity for adequate evaluation of new therapies for enhancing stabilization. For this
1
2
Wallace, Hollinshead, and Frank
J Shoulder Elbow Surg January/February 2001
Table I Morphologic characteristics Group 1 (n = 8) Anatomic Length* (mm) Area (mm2) Width (mm) Thickness (mm) Mass (mg) Water content (%)
23.53 4.90 4.60 1.54 71.83 62.46
(±1.36) (±0.57) (±0.47) (±0.19) (±7.43) (±3.51)
Group 2 (n = 8) Shifted 23.39 4.89 4.20 1.68 68.28 65.09
(±1.24)† (±0.52) (±0.50) (±0.16) (±9.92) (±2.91)
Group 3 (n = 8) Thermal 20.50 5.20 4.56 1.67 76.21 65.58
(±1.20)† (±0.80) (±0.44) (±0.21) (±10.46) (±2.44)
*Length measured at ligament zero (0.1 N). †P < .001.
experiment, we hypothesized that radio frequency electrothermal shrinkage would restore “anatomic” tension in a ligament with abnormal laxity without compromising low-load creep behavior. Given its discrete anatomy, similar overall geometry to the human inferior glenohumeral ligament, and the large database of existing information on its material and structural properties, the medial collateral ligament (MCL) of the rabbit knee was selected as the appropriate initial model.
MATERIALS AND METHODS Twelve skeletally mature female New Zealand White rabbits, approximately 12 months of age and weighing 4 to 5 kg, were used in this study. Animals were purchased from a single supplier (Riemens Fur Ranches, St Agatha, Ontario, Canada) and were allowed to acclimatize for a week before surgery. After sedation with 0.18 mL (10 mg/mL) acepromazine maleate, general anesthesia was induced with a mixture of 2% halothane and oxygen by inhalation. Through bilateral medial longitudinal incisions, the MCLs were exposed and allocated randomly to 3 treatment groups. In group 1 (n = 8) the tibial insertion of the MCL was isolated on a bone block, elevated, and replaced in anatomic position. The bone block was secured to the underlying tibia with a 2-mm screw. In group 2 (n = 8), the tibial insertional bone block was shifted 5 mm proximally toward the jointline before fixation to create abnormal laxity of the MCL. In group 3 (n = 8), the same procedure as used for group 2 was performed, followed by application of radio frequency electric energy (ORATEC Interventions Inc, Calif) to the midsubstance of the MCL until the probe tip temperature sensor reached 65°C. Heating was carried out while the MCL was submerged in a local bath of 0.9% sodium chloride solution, and the probe tip remained stationary while in contact with the ligament surface. A preliminary histologic study demonstrated that this method produced a zone of thermal effect that extended across 50% of the width of the MCL. Immediately after treatment, the animals were euthanatized and the hindlimbs amputated and stored at –20°C for later analysis. All surgical procedures were approved by the Faculty of Medicine Animal Care and Ethics Committee. The hindlimbs were thawed overnight in double plastic bags at room temperature before testing. The knee joints were dissected free of muscle and soft tissues, including joint capsule, leaving the collateral and cruciate ligaments
and menisci intact. The femur and tibia were mounted in pots with methylmethacrylate resin, and the whole preparation was mounted at 70° flexion in a servohydraulic materials testing machine (MTS Corporation, Minneapolis, Minn). The long axis of the MCL was centralized with the principal axis for tensile testing. The testing protocol consisted of (1) determination of whole-joint laxity, (2) determination of laxity of the isolated MCL, (3) determination of cross-sectional area and length of the MCL, (4) “cyclic” and “static” creep testing, and (5) recovery. Whole-joint laxity was calculated from the second of two cycles of displacement between –5 N compression and +2 N tension. The preparation was then held at 0.1 N and all remaining structures divided, leaving only the MCL intact. Laxity of the isolated MCL was calculated from the displacement during cycling between –0.1 N compression and +0.1 N tension. The MCL was then held at +0.1 N (“ligament zero”), and length was measured from femoral to tibial insertion with digital calipers (Mitutoyo Corporation, Kawasaki-Shi, Japan). The MCL was loaded to +5 N tension, and the cross-sectional area was measured with an electronic caliper, as previously described.15 To prevent errors caused by dehydration, the preparation was enclosed in a humidity chamber and heated to 37°C ± 1°C and 99% humidity. Ligament zero was then redefined by further cycling to control for any error induced by heating of the testing apparatus. Cyclic creep was measured during 30 cycles of load to 4.1 megapascals (equivalent to 5% of ultimate tensile stress of the normal MCL). At the peak of the 30th cycle, the load was held constant at 4.1 megapascals for 20 minutes and static creep deformation recorded. The load was then reduced to approximately +0.1 N for an additional 20 minutes and the test concluded. The entire MCL was harvested and divided into two longitudinal strips. From one strip, water content was calculated from the difference between wet weight and stable dry weight after lyophilization for approximately 96 hours. The remaining strip was fixed in Karnovsky’s fixative and osmium tetroxide and stained with lead citrate and uranyl acetate. Images of the transversely sectioned MCL midsubstance were obtained at ×30,000 magnification with transmission electron microscopy. A 1-way analysis of variance was performed across the 3 groups. When a difference in a single parameter was found between 2 groups, the means were compared with the unpaired Student t test, assuming unequal variance. After the t tests, P values were subjected to the Bonferroni
J Shoulder Elbow Surg Volume 10, Number 1
Wallace, Hollinshead, and Frank
3
Figure 1 Laxity of the lapine knee joint with intact menisci, collateral, and cruciate ligaments determined by cycling from –5 N compression to +2 N tension. No significant differences were found between anatomic placement of the medial collateral ligament (0.24 ± 0.10 mm) and the other groups.
Figure 2 Laxity of the isolated medial collateral ligament determined by cycling from –0.1 N compression to +0.1 N tension. Shifting resulted in marked increase in laxity. Thermal treatment resulted in restoration of laxity to levels not significantly different from anatomic placement.
correction for multiple comparisons. Significance was attained when P < .05, and data were expressed as mean ± 1 SD.
RESULTS No significant differences existed between the 3 groups in ligament width or thickness (Table I). A trend was seen toward higher mean values for ligament mass, cross-sectional area, and water content in group 3 (thermal) when compared with the other groups but these did not reach statistical significance. Shifting the tibial insertional bone block toward the jointline had no net effect on whole joint laxity (Figure 1) but resulted in a highly significant increase in laxity
of the isolated MCL (Figure 2). The discrepancy between the extent of shift created at surgery (5 mm) and the laxity measured during mechanical testing (3.33 ± 0.25 mm) probably existed because the axis of shift was slightly divergent from the long axis of the MCL by several degrees. However, this effect was equal across all 3 groups. Thermal treatment resulted in restoration of anatomic laxity (0.66 ± 0.31 mm), and this was associated with a corresponding significant reduction in MCL length when measured at ligament zero, defined as the point at which the ligament was just beginning to carry load (Table I). Not surprisingly, the magnitude of cyclic and static creep between groups 1 and 2 was similar (Table II).
4
Wallace, Hollinshead, and Frank
J Shoulder Elbow Surg January/February 2001
Table II Creep behavior Group 1 (n = 8) Anatomic Low-load stiffness (N/mm) Cyclic strain (%) Cyclic creep strain (%) Static strain (%) Static creep strain (%)
25.85 3.88 0.57 4.96 0.95
(±6.85) (±0.99) (±0.17) (±0.97) (±0.18)
Group 2 (n = 8) Shifted* 28.86 3.53 0.83 4.77 1.10
(±5.74) (±0.52) (±0.20) (±0.80) (±0.37)
Group 3 (n = 8) Thermal* 7.87 11.10 3.25 13.70 2.04
(±2.60) (±2.70) (±1.18) (±3.27) (±0.41)
Cyclic (Static) strain, Total deformation (including creep) / MCL length at ligament zero; cyclic (static) creep strain, cyclic (static) creep deformation / MCL length at ligament zero; MCL, medial collateral ligament. *Significant differences at P < .001.
No differences existed between these two groups in creep strain, defined as the magnitude of cyclic or static creep expressed as a proportion of the MCL length at ligament zero. However, during static creep testing, evidence of partial failure was seen in 2 of 8 specimens in group 3 (thermal), and these were excluded from subsequent analysis (Figure 2). With the remaining 6 specimens, threefold increases were found in cyclic creep strain (3.25% versus 0.83%, P < .001) and twofold increases in static creep strain (2.03% versus 0.95%, P < .001), when compared with group 2. In view of these findings, low-load stiffness was calculated from the ratio of the load to the deformation recorded at the peak of the first cycle of cyclic creep testing. This revealed a reduction in low-load stiffness of approximately 75% in the thermally-treated ligaments when compared with the other groups (Table II). Qualitative observation of transmission electron microscopy images revealed substantial changes in cross-sectional architecture of the ligament, particularly with respect to the normal bimodal population of large and small diameter fibrils.1 In the thermally treated specimens, the majority of collagen fibrils were grossly enlarged, deformed, and intermingled with adjacent fibrils. Electron density was reduced, and the margins of the fibrils were indistinct.
DISCUSSION Biologic tissues such as ligaments exhibit timedependent viscoelastic properties, that is, mechanical behavior that depends on the rate of loading. These properties are also related to the state of hydration of the tissue.3 Generally, most studies of soft connective tissues have reported viscoelasticity in terms of stress relaxation (reduction of tissue load under cyclic or static elongation) rather than creep.2,9,10 Hayashi et al4 examined stress relaxation of rabbit patellar tendon tissue immediately after thermal shrinkage with a holmium–YAG laser and found no change with increasing energy application, although stiffness was markedly reduced. However, it appears that creep and stress relaxation may not be simply inversely related. Recent studies have demonstrated that in the lapine
MCL in vitro, creep behavior cannot be accurately predicted from measured stress relaxation17 and therefore may reflect different events occurring at the fibrillar or molecular level in the tissue. Proposed mechanisms include straightening of the crimp or waviness of the collagen fibrils,19 realignment of fibrils within the matrix, direct stretching of or sliding between the individual matrix molecules, or other interactions at the matrix-fibrillar interface.13 During normal joint function, ligaments are likely subjected to repetitive loads rather than finite displacements, and therefore assessment of creep behavior is probably more relevant. Creep has been measured in non-pretensioned bone-ligament-bone preparations for anterior cruciate ligament reconstruction7 and has been implicated clinically as a mechanism in cruciate graft healing that contributes to recurrent laxity.8 In experimental studies at various stages after surgical division, healing ligaments were found to be more susceptible to creep at low physiologic stresses than intact, noninjured ligaments.18 In this study, we found that low-load stiffness and creep of the lapine MCL were very significantly altered after radio frequency electrothermal shrinkage, and that 25% of thermally-treated specimens failed during the low-load creep test. We speculate that the increased deformation may reflect relative straightening or unraveling of the denatured collagen chains, whereas the reduction in strength may be caused by loss of the heat-labile intramolecular bonds stabilizing the crystalline structure of the triple helix of type I collagen. One limitation of this study is that the mechanical testing was performed immediately after surgery; it therefore takes no account of the biologic response of the tissue to thermal treatment that occurred with time after surgery. However, the clinical relevance of these findings is that mobilization of joints stabilized with thermal treatment may be contraindicated in the early postoperative phase, given that there is an increased risk of elongation and possible failure of ligament structures. This tendency may be amplified by the subsequent biologic response during the healing and remod-
Wallace, Hollinshead, and Frank
J Shoulder Elbow Surg Volume 10, Number 1
5
Figure 3 Examples of creep curves obtained from an “anatomic” ligament (bottom) and a thermally treated liga-
ment (top) after 20 minutes of static load at 4.1 megapascals stress. This discontinuity evident in the thermal curve was interpreted as partial ligament failure, and this specimen was excluded from quantitative analysis.
eling process, as suggested by others.14 Further studies that use this model are underway in our laboratory to address the specific question of the safe limits for mobilization, but until this information is known, rehabilitation should probably be undertaken with caution. Application of laser energy to ligament tissue results in heating caused by absorption of light by water. On the other hand, electric energy can create heat by rapid changes of polarity in a radio frequency current, causing molecular friction.22 Recently, a comparative study found radio frequency electrothermal shrinkage to be as effective as laser shrinkage in reducing glenohumeral translations in a cadaveric model in response to low loads.20 In our model, electrothermal shrinkage also reduced the surgically induced laxity to anatomic levels, but at the cost of increased susceptibility to creep. Ultrastructural evidence indicated that the ligament matrix underwent substantial modification in response to heat, confirming the results of previous studies that used laser as the energy source.5 In summary, these results refute previous reports suggesting that viscoelastic properties are not altered by thermal treatment. In the long term, if ligament shortening is maintained during the remodeling process, the final result may well be comparable with, or superior to, conventional open methods of capsulorrhaphy. Clearly, further experimental studies will be required to determine whether the thermally-treated matrix acts as
a scaffold for rapid fibroblast repopulation of the ligament tissue, or whether the matrix must be removed and replaced de novo by extrinsic scar tissue, with concomitant delay in restoration of normal viscoelastic behavior. REFERENCES 1. Frank C, McDonald D, Bray D, Bray R, Rangayyan R, Chimich D, et al. Collagen fibril diameters in the healing adult rabbit medial collateral ligament. Connect Tissue Res 1992;27:251-63. 2. Graf BK, Vanderby R, Ulm MJ, Rogalski RP, Thielke RJ. Effect of preconditioning on the viscoelastic response of primate patellar tendon. Arthroscopy 1994;10:90-6. 3. Haut TL, Haut RC. The state of tissue hydration determines the strain-rate-sensitive stiffness of human patellar tendon. J Biomech 1997;30:79-81. 4. Hayashi K, Markel MD, Thabit G, Bogdanske JJ, Thielke RJ. The effect of nonablative laser energy on joint capsular properties: an in vitro mechanical study using a rabbit model. Am J Sports Med 1995;23:482-7. 5. Hayashi K, Thabit G, Bogdanske JJ, Mascio LN, Markel MD. The effect of nonablative laser energy on the ultrastructure of joint capsular collagen. Arthroscopy 1996;12:474-81. 6. Holden JP, Grood ES, Korvick DL, Cummings JF, Butler DL, Bylski-Austrow DI. In vivo forces in the anterior cruciate ligament: direct measurements during walking and trotting in a quadruped. J Biomech 1994;27:517-26. 7. Howard ME, Cawley PW, Losse GM, Johnston RB. Bone-patellar tendon-bone grafts for anterior cruciate ligament reconstruction: the effects of graft pretensioning. Arthroscopy 1996;12:28792.
6
Wallace, Hollinshead, and Frank
J Shoulder Elbow Surg January/February 2001
8. Jenkins WL, Munns SW, Jayaraman G, Wertzberger KL, Neely K. A measurement of anterior tibial displacement in the closed and open kinetic chain. J Orthop Sports Phys Ther 1997;25: 49-5. 9. King GJW, Edwards P, Brant RF, Shrive NG, Frank CB. Intraoperative graft tensioning alters viscoelastic but not failure behaviours of rabbit medial collateral ligament autografts. J Orthop Res 1995;13:915-22. 10. Lam TC, Frank CB, Shrive NG. Changes in the cyclic and static relaxations of the rabbit medial collateral ligament complex during maturation. J Biomech 1993;26:9-17. 11. Manta JP, Organ S, Nirschl RP, Pettrone FA. Arthroscopic transglenoid suture capsulolabral repair: five year followup. Am J Sports Med 1997;25:614-18. 12. Pagnani MJ, Warren RF, Altchek DW, Wickiewicz TL, Anderson AF. Arthroscopic shoulder stabilization using transglenoid sutures: a four-year minimum followup. Am J Sports Med 1996;24:459-67. 13. Purslow PP, Wess TJ, Hukins DWL. Collagen orientation and molecular spacing during creep and stress-relaxation in soft connective tissues. J Exp Biol 1998;201:135-42. 14. Schaefer SL, Ciarelli MJ, Arnoczky SP, Ross HE. Tissue shrinkage with the holmium-yttrium aluminium garnet laser: a postoperative assessment of tissue length, stiffness and structure. Am J Sports Med 1997;25:841-8. 15. Shrive NG, Lam TC, Damson E, Frank CB. A new method of
16. 17.
18. 19.
20.
21. 22.
measuring the cross-sectional area of connective tissue structures. J Biomech Eng 1988;110:104-9. Speer KP, Warren RF, Pagnani M, Warner JJP. An arthroscopic technique for anterior stabilization of the shoulder with a bioabsorbable tack. J Bone Joint Surg Am 1996;78:1801-7. Thornton GM, Olinyk A, Frank CB, Shrive NG. Ligament creep cannot be predicted from stress relaxation at low stress: a biomechanical study of the rabbit medial collateral ligament. J Orthop Res 1997;15:652-6. Thornton GM, Frank CB, Shrive NG, Leask GP. Ligament scar creeps more than normal ligament in early healing. Trans Orth Res Soc 1997;22:73. Thornton GM, Sutherland CA, Barclay LD, Leask GP, Marchuk LL, Frank CB, et al. Creep is resisted by fibre recruitment in ligament: evidence from altered crimp patterns. Trans Orth Res Soc 1998;23:45. Tibone JE, McMahon PJ, Shrader TA, Black AD, Sandusky MD, Lee TQ. Glenohumeral translation after thermal anterior capsulorrhaphy: a comparison of application with laser and radiofrequency. Trans Orth Res Soc 1998;23:735. Vangsness CT, Mitchell W, Nimni M, Erlich M, Saadat V, Schmotzer H. Collagen shortening: an experimental approach with heat. Clin Orthop 1997;337:267-71. Wallace AL, Hollinshead RM, Frank CB. The scientific basis of thermal capsular shrinkage. J Shoulder Elbow Surg 2000; 9:354-60.
THE 8TH ANNUAL CONGRESS OF THE GERMAN SOCIETY OF SHOULDER AND ELBOW SURGERY JUNE 15 AND 16, 2001 MUNICH, GERMANY Topics: • • • •
Rotator cuff tear: Basic science, reconstruction, rehabilitation Biceps and biceps tendon: Degeneration and trauma Traumatology of the elbow Free topics
Also: Symposium on shoulder arthroplasty For more information, contact INTERPLAN, Albert-Rosshaupter-Str. 65, 81369 München, Germany. Telephone: (089) 54 82 34-15; Fax: (089) 54 82 34-43; E-mail: [email protected]; Web site: www.dvse2001.de.
Thermal denaturation of collagen in ligament tissue has the potential to enhance arthroscopic shoulder stabilization. Previous studies have shown that laser energy produces significant capsular shortening without alteration of viscoelastic properties, but little information is available on the effects of radio frequency electrothermal energy. We assessed the acute effects of radio frequency shrinkage with use of the lapine medial collateral ligament model, in which the tibial insertion was shifted proximally to produce abnormal laxity. Thermal treatment resulted in restoration of laxity from 3.33 ± 0.25 mm to 0.66 ± 0.31 mm, which was not significantly different from medial collateral ligaments replaced anatomically (0.50 ± 0.34 mm). When tested at 4.1 megapascals, cyclic and static creep strains were increased twofold to threefold in thermally-treated ligaments (P < .01), and partial failure occurred in 2 of 8 cases. We conclude that radio frequency electrothermal shrinkage is effective at reducing laxity but significantly alters viscoelastic properties, posing a risk of recurrent stretching-out at “physiological” loads. (J Shoulder Elbow Surg 2001;10:1-6.)
From McCaig Centre for Joint Injury and Arthritis Research, University of Calgary, Alberta, Canada. Nominated for the Charles S. Neer Award in Basic Science, American Shoulder and Elbow Surgeons Meeting, Anaheim, Calif, February 1999. Reprint requests: Andrew L. Wallace, MBBS, PhD, FRACS(Orth), Senior Lecturer/Honorary Consultant Orthopaedic Surgeon, Imperial College School of Medicine, Charing Cross Hospital, Fulham Palace Road, London W6 8RF, United Kingdom (Email: [email protected]). Copyright © 2001 by Journal of Shoulder and Elbow Surgery Board of Trustees. 1058-2746/2001/$35.00 + 0 32/1/112023 doi:10.1067/mse.2001.112023
INTRODUCTION Instability of the shoulder remains a common clinical presentation in young people, particularly as a consequence of athletic activity. Increasingly, the expectation in those requiring surgical management of the unstable shoulder is for minimal perioperative morbidity, expedient recovery, no recurrence of symptoms, and rehabilitation facilitating a rapid return to full activity without loss of motion. Recent technical advances in arthroscopic stabilization have brought this ideal within reach, but published results to date have not proven superior to those reported for open reconstruction.12 Difficulty in reducing excess capsular laxity at the time of surgery and recurrence of laxity with time have been cited as the major reasons for failure.11,16 Thermal shrinkage of collagen is a novel development that has considerable potential for dealing with this problem. Tissue heating, achieved either by laser light irradiation or application of radio frequency electric current, causes shrinkage by denaturation of the collagen triple helix at a critical temperature, usually 60°C to 65°C for mammalian tissue.22 In vitro studies have demonstrated that shortening of up to 50% can occur,4 though this was accompanied by significant reductions in high-load structural properties. Immediately after heating, tensile strength of isolated specimens was reduced by 30%, whereas tensile stiffness was reduced by up to 90%.4,21 Few studies have attempted to document lowload mechanical behavior, though it appears that ligament stresses during normal functional activity in vivo are typically less than 10% of ultimate tensile stress.6 Creep refers to the time-dependent increase in deformation seen in certain materials under constant cyclic or static loads, and this phenomenon may play a role in the development of recurrent laxity seen after ligament repair or autograft reconstruction.7,8 Unfortunately, at present there are no established animal models with sufficient acquired ligament laxity for adequate evaluation of new therapies for enhancing stabilization. For this
1
2
Wallace, Hollinshead, and Frank
J Shoulder Elbow Surg January/February 2001
Table I Morphologic characteristics Group 1 (n = 8) Anatomic Length* (mm) Area (mm2) Width (mm) Thickness (mm) Mass (mg) Water content (%)
23.53 4.90 4.60 1.54 71.83 62.46
(±1.36) (±0.57) (±0.47) (±0.19) (±7.43) (±3.51)
Group 2 (n = 8) Shifted 23.39 4.89 4.20 1.68 68.28 65.09
(±1.24)† (±0.52) (±0.50) (±0.16) (±9.92) (±2.91)
Group 3 (n = 8) Thermal 20.50 5.20 4.56 1.67 76.21 65.58
(±1.20)† (±0.80) (±0.44) (±0.21) (±10.46) (±2.44)
*Length measured at ligament zero (0.1 N). †P < .001.
experiment, we hypothesized that radio frequency electrothermal shrinkage would restore “anatomic” tension in a ligament with abnormal laxity without compromising low-load creep behavior. Given its discrete anatomy, similar overall geometry to the human inferior glenohumeral ligament, and the large database of existing information on its material and structural properties, the medial collateral ligament (MCL) of the rabbit knee was selected as the appropriate initial model.
MATERIALS AND METHODS Twelve skeletally mature female New Zealand White rabbits, approximately 12 months of age and weighing 4 to 5 kg, were used in this study. Animals were purchased from a single supplier (Riemens Fur Ranches, St Agatha, Ontario, Canada) and were allowed to acclimatize for a week before surgery. After sedation with 0.18 mL (10 mg/mL) acepromazine maleate, general anesthesia was induced with a mixture of 2% halothane and oxygen by inhalation. Through bilateral medial longitudinal incisions, the MCLs were exposed and allocated randomly to 3 treatment groups. In group 1 (n = 8) the tibial insertion of the MCL was isolated on a bone block, elevated, and replaced in anatomic position. The bone block was secured to the underlying tibia with a 2-mm screw. In group 2 (n = 8), the tibial insertional bone block was shifted 5 mm proximally toward the jointline before fixation to create abnormal laxity of the MCL. In group 3 (n = 8), the same procedure as used for group 2 was performed, followed by application of radio frequency electric energy (ORATEC Interventions Inc, Calif) to the midsubstance of the MCL until the probe tip temperature sensor reached 65°C. Heating was carried out while the MCL was submerged in a local bath of 0.9% sodium chloride solution, and the probe tip remained stationary while in contact with the ligament surface. A preliminary histologic study demonstrated that this method produced a zone of thermal effect that extended across 50% of the width of the MCL. Immediately after treatment, the animals were euthanatized and the hindlimbs amputated and stored at –20°C for later analysis. All surgical procedures were approved by the Faculty of Medicine Animal Care and Ethics Committee. The hindlimbs were thawed overnight in double plastic bags at room temperature before testing. The knee joints were dissected free of muscle and soft tissues, including joint capsule, leaving the collateral and cruciate ligaments
and menisci intact. The femur and tibia were mounted in pots with methylmethacrylate resin, and the whole preparation was mounted at 70° flexion in a servohydraulic materials testing machine (MTS Corporation, Minneapolis, Minn). The long axis of the MCL was centralized with the principal axis for tensile testing. The testing protocol consisted of (1) determination of whole-joint laxity, (2) determination of laxity of the isolated MCL, (3) determination of cross-sectional area and length of the MCL, (4) “cyclic” and “static” creep testing, and (5) recovery. Whole-joint laxity was calculated from the second of two cycles of displacement between –5 N compression and +2 N tension. The preparation was then held at 0.1 N and all remaining structures divided, leaving only the MCL intact. Laxity of the isolated MCL was calculated from the displacement during cycling between –0.1 N compression and +0.1 N tension. The MCL was then held at +0.1 N (“ligament zero”), and length was measured from femoral to tibial insertion with digital calipers (Mitutoyo Corporation, Kawasaki-Shi, Japan). The MCL was loaded to +5 N tension, and the cross-sectional area was measured with an electronic caliper, as previously described.15 To prevent errors caused by dehydration, the preparation was enclosed in a humidity chamber and heated to 37°C ± 1°C and 99% humidity. Ligament zero was then redefined by further cycling to control for any error induced by heating of the testing apparatus. Cyclic creep was measured during 30 cycles of load to 4.1 megapascals (equivalent to 5% of ultimate tensile stress of the normal MCL). At the peak of the 30th cycle, the load was held constant at 4.1 megapascals for 20 minutes and static creep deformation recorded. The load was then reduced to approximately +0.1 N for an additional 20 minutes and the test concluded. The entire MCL was harvested and divided into two longitudinal strips. From one strip, water content was calculated from the difference between wet weight and stable dry weight after lyophilization for approximately 96 hours. The remaining strip was fixed in Karnovsky’s fixative and osmium tetroxide and stained with lead citrate and uranyl acetate. Images of the transversely sectioned MCL midsubstance were obtained at ×30,000 magnification with transmission electron microscopy. A 1-way analysis of variance was performed across the 3 groups. When a difference in a single parameter was found between 2 groups, the means were compared with the unpaired Student t test, assuming unequal variance. After the t tests, P values were subjected to the Bonferroni
J Shoulder Elbow Surg Volume 10, Number 1
Wallace, Hollinshead, and Frank
3
Figure 1 Laxity of the lapine knee joint with intact menisci, collateral, and cruciate ligaments determined by cycling from –5 N compression to +2 N tension. No significant differences were found between anatomic placement of the medial collateral ligament (0.24 ± 0.10 mm) and the other groups.
Figure 2 Laxity of the isolated medial collateral ligament determined by cycling from –0.1 N compression to +0.1 N tension. Shifting resulted in marked increase in laxity. Thermal treatment resulted in restoration of laxity to levels not significantly different from anatomic placement.
correction for multiple comparisons. Significance was attained when P < .05, and data were expressed as mean ± 1 SD.
RESULTS No significant differences existed between the 3 groups in ligament width or thickness (Table I). A trend was seen toward higher mean values for ligament mass, cross-sectional area, and water content in group 3 (thermal) when compared with the other groups but these did not reach statistical significance. Shifting the tibial insertional bone block toward the jointline had no net effect on whole joint laxity (Figure 1) but resulted in a highly significant increase in laxity
of the isolated MCL (Figure 2). The discrepancy between the extent of shift created at surgery (5 mm) and the laxity measured during mechanical testing (3.33 ± 0.25 mm) probably existed because the axis of shift was slightly divergent from the long axis of the MCL by several degrees. However, this effect was equal across all 3 groups. Thermal treatment resulted in restoration of anatomic laxity (0.66 ± 0.31 mm), and this was associated with a corresponding significant reduction in MCL length when measured at ligament zero, defined as the point at which the ligament was just beginning to carry load (Table I). Not surprisingly, the magnitude of cyclic and static creep between groups 1 and 2 was similar (Table II).
4
Wallace, Hollinshead, and Frank
J Shoulder Elbow Surg January/February 2001
Table II Creep behavior Group 1 (n = 8) Anatomic Low-load stiffness (N/mm) Cyclic strain (%) Cyclic creep strain (%) Static strain (%) Static creep strain (%)
25.85 3.88 0.57 4.96 0.95
(±6.85) (±0.99) (±0.17) (±0.97) (±0.18)
Group 2 (n = 8) Shifted* 28.86 3.53 0.83 4.77 1.10
(±5.74) (±0.52) (±0.20) (±0.80) (±0.37)
Group 3 (n = 8) Thermal* 7.87 11.10 3.25 13.70 2.04
(±2.60) (±2.70) (±1.18) (±3.27) (±0.41)
Cyclic (Static) strain, Total deformation (including creep) / MCL length at ligament zero; cyclic (static) creep strain, cyclic (static) creep deformation / MCL length at ligament zero; MCL, medial collateral ligament. *Significant differences at P < .001.
No differences existed between these two groups in creep strain, defined as the magnitude of cyclic or static creep expressed as a proportion of the MCL length at ligament zero. However, during static creep testing, evidence of partial failure was seen in 2 of 8 specimens in group 3 (thermal), and these were excluded from subsequent analysis (Figure 2). With the remaining 6 specimens, threefold increases were found in cyclic creep strain (3.25% versus 0.83%, P < .001) and twofold increases in static creep strain (2.03% versus 0.95%, P < .001), when compared with group 2. In view of these findings, low-load stiffness was calculated from the ratio of the load to the deformation recorded at the peak of the first cycle of cyclic creep testing. This revealed a reduction in low-load stiffness of approximately 75% in the thermally-treated ligaments when compared with the other groups (Table II). Qualitative observation of transmission electron microscopy images revealed substantial changes in cross-sectional architecture of the ligament, particularly with respect to the normal bimodal population of large and small diameter fibrils.1 In the thermally treated specimens, the majority of collagen fibrils were grossly enlarged, deformed, and intermingled with adjacent fibrils. Electron density was reduced, and the margins of the fibrils were indistinct.
DISCUSSION Biologic tissues such as ligaments exhibit timedependent viscoelastic properties, that is, mechanical behavior that depends on the rate of loading. These properties are also related to the state of hydration of the tissue.3 Generally, most studies of soft connective tissues have reported viscoelasticity in terms of stress relaxation (reduction of tissue load under cyclic or static elongation) rather than creep.2,9,10 Hayashi et al4 examined stress relaxation of rabbit patellar tendon tissue immediately after thermal shrinkage with a holmium–YAG laser and found no change with increasing energy application, although stiffness was markedly reduced. However, it appears that creep and stress relaxation may not be simply inversely related. Recent studies have demonstrated that in the lapine
MCL in vitro, creep behavior cannot be accurately predicted from measured stress relaxation17 and therefore may reflect different events occurring at the fibrillar or molecular level in the tissue. Proposed mechanisms include straightening of the crimp or waviness of the collagen fibrils,19 realignment of fibrils within the matrix, direct stretching of or sliding between the individual matrix molecules, or other interactions at the matrix-fibrillar interface.13 During normal joint function, ligaments are likely subjected to repetitive loads rather than finite displacements, and therefore assessment of creep behavior is probably more relevant. Creep has been measured in non-pretensioned bone-ligament-bone preparations for anterior cruciate ligament reconstruction7 and has been implicated clinically as a mechanism in cruciate graft healing that contributes to recurrent laxity.8 In experimental studies at various stages after surgical division, healing ligaments were found to be more susceptible to creep at low physiologic stresses than intact, noninjured ligaments.18 In this study, we found that low-load stiffness and creep of the lapine MCL were very significantly altered after radio frequency electrothermal shrinkage, and that 25% of thermally-treated specimens failed during the low-load creep test. We speculate that the increased deformation may reflect relative straightening or unraveling of the denatured collagen chains, whereas the reduction in strength may be caused by loss of the heat-labile intramolecular bonds stabilizing the crystalline structure of the triple helix of type I collagen. One limitation of this study is that the mechanical testing was performed immediately after surgery; it therefore takes no account of the biologic response of the tissue to thermal treatment that occurred with time after surgery. However, the clinical relevance of these findings is that mobilization of joints stabilized with thermal treatment may be contraindicated in the early postoperative phase, given that there is an increased risk of elongation and possible failure of ligament structures. This tendency may be amplified by the subsequent biologic response during the healing and remod-
Wallace, Hollinshead, and Frank
J Shoulder Elbow Surg Volume 10, Number 1
5
Figure 3 Examples of creep curves obtained from an “anatomic” ligament (bottom) and a thermally treated liga-
ment (top) after 20 minutes of static load at 4.1 megapascals stress. This discontinuity evident in the thermal curve was interpreted as partial ligament failure, and this specimen was excluded from quantitative analysis.
eling process, as suggested by others.14 Further studies that use this model are underway in our laboratory to address the specific question of the safe limits for mobilization, but until this information is known, rehabilitation should probably be undertaken with caution. Application of laser energy to ligament tissue results in heating caused by absorption of light by water. On the other hand, electric energy can create heat by rapid changes of polarity in a radio frequency current, causing molecular friction.22 Recently, a comparative study found radio frequency electrothermal shrinkage to be as effective as laser shrinkage in reducing glenohumeral translations in a cadaveric model in response to low loads.20 In our model, electrothermal shrinkage also reduced the surgically induced laxity to anatomic levels, but at the cost of increased susceptibility to creep. Ultrastructural evidence indicated that the ligament matrix underwent substantial modification in response to heat, confirming the results of previous studies that used laser as the energy source.5 In summary, these results refute previous reports suggesting that viscoelastic properties are not altered by thermal treatment. In the long term, if ligament shortening is maintained during the remodeling process, the final result may well be comparable with, or superior to, conventional open methods of capsulorrhaphy. Clearly, further experimental studies will be required to determine whether the thermally-treated matrix acts as
a scaffold for rapid fibroblast repopulation of the ligament tissue, or whether the matrix must be removed and replaced de novo by extrinsic scar tissue, with concomitant delay in restoration of normal viscoelastic behavior. REFERENCES 1. Frank C, McDonald D, Bray D, Bray R, Rangayyan R, Chimich D, et al. Collagen fibril diameters in the healing adult rabbit medial collateral ligament. Connect Tissue Res 1992;27:251-63. 2. Graf BK, Vanderby R, Ulm MJ, Rogalski RP, Thielke RJ. Effect of preconditioning on the viscoelastic response of primate patellar tendon. Arthroscopy 1994;10:90-6. 3. Haut TL, Haut RC. The state of tissue hydration determines the strain-rate-sensitive stiffness of human patellar tendon. J Biomech 1997;30:79-81. 4. Hayashi K, Markel MD, Thabit G, Bogdanske JJ, Thielke RJ. The effect of nonablative laser energy on joint capsular properties: an in vitro mechanical study using a rabbit model. Am J Sports Med 1995;23:482-7. 5. Hayashi K, Thabit G, Bogdanske JJ, Mascio LN, Markel MD. The effect of nonablative laser energy on the ultrastructure of joint capsular collagen. Arthroscopy 1996;12:474-81. 6. Holden JP, Grood ES, Korvick DL, Cummings JF, Butler DL, Bylski-Austrow DI. In vivo forces in the anterior cruciate ligament: direct measurements during walking and trotting in a quadruped. J Biomech 1994;27:517-26. 7. Howard ME, Cawley PW, Losse GM, Johnston RB. Bone-patellar tendon-bone grafts for anterior cruciate ligament reconstruction: the effects of graft pretensioning. Arthroscopy 1996;12:28792.
6
Wallace, Hollinshead, and Frank
J Shoulder Elbow Surg January/February 2001
8. Jenkins WL, Munns SW, Jayaraman G, Wertzberger KL, Neely K. A measurement of anterior tibial displacement in the closed and open kinetic chain. J Orthop Sports Phys Ther 1997;25: 49-5. 9. King GJW, Edwards P, Brant RF, Shrive NG, Frank CB. Intraoperative graft tensioning alters viscoelastic but not failure behaviours of rabbit medial collateral ligament autografts. J Orthop Res 1995;13:915-22. 10. Lam TC, Frank CB, Shrive NG. Changes in the cyclic and static relaxations of the rabbit medial collateral ligament complex during maturation. J Biomech 1993;26:9-17. 11. Manta JP, Organ S, Nirschl RP, Pettrone FA. Arthroscopic transglenoid suture capsulolabral repair: five year followup. Am J Sports Med 1997;25:614-18. 12. Pagnani MJ, Warren RF, Altchek DW, Wickiewicz TL, Anderson AF. Arthroscopic shoulder stabilization using transglenoid sutures: a four-year minimum followup. Am J Sports Med 1996;24:459-67. 13. Purslow PP, Wess TJ, Hukins DWL. Collagen orientation and molecular spacing during creep and stress-relaxation in soft connective tissues. J Exp Biol 1998;201:135-42. 14. Schaefer SL, Ciarelli MJ, Arnoczky SP, Ross HE. Tissue shrinkage with the holmium-yttrium aluminium garnet laser: a postoperative assessment of tissue length, stiffness and structure. Am J Sports Med 1997;25:841-8. 15. Shrive NG, Lam TC, Damson E, Frank CB. A new method of
16. 17.
18. 19.
20.
21. 22.
measuring the cross-sectional area of connective tissue structures. J Biomech Eng 1988;110:104-9. Speer KP, Warren RF, Pagnani M, Warner JJP. An arthroscopic technique for anterior stabilization of the shoulder with a bioabsorbable tack. J Bone Joint Surg Am 1996;78:1801-7. Thornton GM, Olinyk A, Frank CB, Shrive NG. Ligament creep cannot be predicted from stress relaxation at low stress: a biomechanical study of the rabbit medial collateral ligament. J Orthop Res 1997;15:652-6. Thornton GM, Frank CB, Shrive NG, Leask GP. Ligament scar creeps more than normal ligament in early healing. Trans Orth Res Soc 1997;22:73. Thornton GM, Sutherland CA, Barclay LD, Leask GP, Marchuk LL, Frank CB, et al. Creep is resisted by fibre recruitment in ligament: evidence from altered crimp patterns. Trans Orth Res Soc 1998;23:45. Tibone JE, McMahon PJ, Shrader TA, Black AD, Sandusky MD, Lee TQ. Glenohumeral translation after thermal anterior capsulorrhaphy: a comparison of application with laser and radiofrequency. Trans Orth Res Soc 1998;23:735. Vangsness CT, Mitchell W, Nimni M, Erlich M, Saadat V, Schmotzer H. Collagen shortening: an experimental approach with heat. Clin Orthop 1997;337:267-71. Wallace AL, Hollinshead RM, Frank CB. The scientific basis of thermal capsular shrinkage. J Shoulder Elbow Surg 2000; 9:354-60.
THE 8TH ANNUAL CONGRESS OF THE GERMAN SOCIETY OF SHOULDER AND ELBOW SURGERY JUNE 15 AND 16, 2001 MUNICH, GERMANY Topics: • • • •
Rotator cuff tear: Basic science, reconstruction, rehabilitation Biceps and biceps tendon: Degeneration and trauma Traumatology of the elbow Free topics
Also: Symposium on shoulder arthroplasty For more information, contact INTERPLAN, Albert-Rosshaupter-Str. 65, 81369 München, Germany. Telephone: (089) 54 82 34-15; Fax: (089) 54 82 34-43; E-mail: [email protected]; Web site: www.dvse2001.de.