- Email: [email protected]
Loosening of Transtibial Pullout Meniscal Root Repairs due to Simulated Rehabilitation Is Unrecoverable: A Biomechanical Study
Loosening of Transtibial Pullout Meniscal Root Repairs due to Simulated Rehabilitation Is Unrecoverable: A Biomechanical Study Brett D. Steineman, Ph.D., Robert F. LaPrade, M.D., Ph.D., and Tammy L. Haut Donahue, Ph.D.
Purpose: To determine whether meniscal root repairs recover from displacement due to rehabilitative loading. Methods: Transtibial pullout repairs of the posteromedial meniscal root were performed in 16 cadaveric ovine knees. Single- and double-tunnel repairs using the 2esimple suture technique were cyclically loaded in tension to 10,000 cycles, allowed to rest, and loaded in tension again. Paired differences in displacement with rest were recorded to evaluate recoverability. Displacement of repairs at cycles of interest was recorded, and the response of repairs to 10,000 cycles was assessed. Results: All outcomes were not significantly different between the single- and double-tunnel techniques; therefore, the results were pooled. The difference in displacement between the first cycle and the first cycle after rest was 1.59 0.69 mm. Repair displacement did not reach an equilibrium within 10,000 cycles and instead resulted in a steady increase in displacement of 0.05 0.02 mm per additional 1,000 cycles. Sutures macroscopically began to cut out of the meniscus in both single- and double-tunnel repairs. Conclusions: This study showed that significant, unrecoverable loosening from rehabilitative loading occurred in single- and double-tunnel meniscal root repairs. Root repairs also gradually displaced with continued loading instead of reaching an equilibrium displacement after 10,000 cycles. This progressive, unrecoverable loosening needs to be studied further to better understand the resultant impact on knee mechanics. In addition, the quality and quantity of meniscal root repair healing at the time of rehabilitation should be studied to determine how susceptible patients are to repair loosening. Clinical Relevance: Rehabilitative loading caused unrecoverable and progressive loosening of root repairs, showing the importance of healing before loading. Investigations on the effects of loosening on mechanics and the quality of repair healing at weight bearing are necessary to better understand the clinical implications.
M
eniscal root tears may occur acutely, chronically, or as a result of iatrogenic injury.1 The posterior meniscal insertions are more susceptible to tears, especially when loaded when the knee is in deep
From the Department of Biomechanics, Hospital for Special Surgery (B.D.S.), New York, New York; The Steadman Clinic (R.F.L.), Vail, Colorado; Steadman Philippon Research Institute (R.F.L.), Vail, Colorado; and Department of Biomedical Engineering, University of Massachusetts (T.L.H.D.), Amherst, Massachusetts, U.S.A. The authors report the following potential conflicts of interest or sources of funding: R.F.L. is a consultant for and receives royalties from Arthrex, Össur, and Smith & Nephew. Full ICMJE author disclosure forms are available for this article online, as supplementary material. Received July 6, 2018; accepted November 16, 2018. Address correspondence to Tammy L. Haut Donahue, Ph.D., Department of Biomedical Engineering, University of Massachusetts, 130 Natural Resources Rd, Amherst, MA 01003, U.S.A. E-mail: [email protected] Ó 2018 by the Arthroscopy Association of North America 0749-8063/18813/$36.00 https://doi.org/10.1016/j.arthro.2018.11.041
flexion.2 Untreated meniscal root tears are becoming increasingly recognized with progressive increases in meniscal extrusion and articular cartilage degeneration detectable within a year.3 Therefore, surgical treatment is essential to reduce the risk of osteoarthritis development after meniscal root tears. Repairs of meniscal root tears are becoming more desirable than meniscectomy because they significantly improve clinical and radiologic outcomes while reducing cost.4 Previous ex vivo biomechanical experiments have reported that anatomic, transtibial pullout meniscal root repairs nearly restore intact contact mechanics immediately after repair.5-7 Despite this, postoperative follow-up studies have shown that meniscal extrusion was only restored in 56% of patients and progression of osteoarthritis was not always prevented.8,9 A potential reason for limited success of repairs may be due to significant displacement in meniscal root repairs shown to occur in biomechanical studies that simulated rehabilitative loading.10,11 Although these studies showed
Arthroscopy: The Journal of Arthroscopic and Related Surgery, Vol
-,
No
-
(Month), 2019: pp 1-8
1
2
B. D. STEINEMAN ET AL.
that displacement accumulates with loading of repairs, they were unable to distinguish whether this displacement was due to viscoelastic creep of the meniscal root or due to permanent, unrecoverable loosening of the repair. Displacement due to tissue viscoelasticity would be recoverable and result in small changes in the intraspecimen repeatability of tests with appropriate rest.12 If these repairs are unable to recover to the initial state of repair, then nearly restorative repairs may loosen into a less restorative position in vivo, resulting in decreased load transmission through the menisci and increased load transfer through the articular cartilage. In previous second-look arthroscopy studies of meniscal root repairs, little to no healing has been shown to occur in some patients with earliest follow-up at 6 months postoperatively.13,14 Failed healing may be defined as no continuity, an easily raised meniscus on probing, and no evidence of meniscal healing at the repair site.13 Another study defined little healing as no true continuity of the meniscus with some connecting scar tissue and the meniscus being easily raised on probing.13 Previous studies evaluating repair displacement have loaded repairs for only 1,000 cycles because they assumed some healing occurs before partial weight bearing in rehabilitation.10,11 Therefore, 1,000 loading cycles does not represent certain cases of meniscal root repairs that may occur in vivo. Although prescribed weight bearing varies between patients, partial weight bearing typically begins around 6 weeks.15 Because some cases of partial or no healing have been observed and the progress of healing near rehabilitative loading is unknown, the extent of repair displacement needs to be examined further. In addition, although previous studies showed that single- and double-tunnel repairs mechanically respond similarly,11 both repairs were also performed and compared to assess for differences in recoverability of displacement. The purpose of this study was to determine whether meniscal root repairs recover from displacement due to rehabilitative loading. It was hypothesized that a significant amount of displacement due to rehabilitative loading would be caused by unrecoverable loosening of repairs. It was also hypothesized that single- and double-tunnel repairs would not be significantly different in repair displacement or in recovery from this displacement.
Methods Specimen Preparation Sixteen cadaveric knees from adult sheep that had been euthanized for unrelated purposes were collected from the Colorado State University Veterinary Teaching Hospital; therefore, these specimens were exempt from Institutional Animal Care and Use Committee approval. The animals cared for and evaluated by veterinarians at the Colorado State University Veterinary Teaching
Hospital that were used in this study were approximately 3 years old and weighed between 65 and 90 kg at the time of euthanasia. Knees were dissected free of all extra-articular skin, muscle, and soft tissue while fresh without previous freezing. The anterior and posterior cruciate ligaments, medial and fibular collateral ligaments, and lateral meniscal insertions were sharply transected to expose the tibial plateau with only the medial menisci intact. The tibiae were then cut at the mid diaphysis to remove the distal portion and potted in a 2-part urethane casting resin (Smooth Cast 321; Smooth-on, Easton, PA). Repair Technique The knees were randomly distributed between the single- and double-tunnel transtibial pullout repair groups, with 8 specimens per group. All repairs were performed and assessed postoperatively by a boardcertified sports medicine orthopaedic surgeon (R.F.L.) experienced with meniscal root repairs. Initially, the posteromedial meniscal root insertion was cut with a scalpel at the interface with the tibial plateau, and the insertion area was marked with a surgical pen to identify anatomic placement. For the single-tunnel group, a 4.5-mm-diameter transtibial tunnel was created using a guide pin and reamer (Smith & Nephew, Andover, MA) in the middle of the posteromedial meniscal root attachment. Two No. 2 nonabsorbable sutures (Ultrabraid; Smith & Nephew) were passed between 5 and 7 mm medial within the meniscal root.16 The 2esimple suture technique was performed for all repairs because this technique has shown the lowest technical difficulty and resists displacement from cyclic loading better than other techniques assessed.17 In addition, this technique is the one that most current published studies have used (Fig 1).6,18 The sutures were placed into a looped nitinol wire, pulled through the tunnel, manually tensioned to reduce the menisci to the native position, and firmly tied to a 4-mm 12mm surgical fixation button (EndoButton; Smith & Nephew) over the anteromedial tibia using a surgeon’s knot with 5 suture throws followed by 5 half-hitches on alternating posts over the fixation button. For the double-tunnel group, the first transtibial tunnel was drilled in the same manner as in the singletunnel group, with a diameter of 3 mm. A second transtibial tunnel was then drilled using a calibrated offset guide to position the second tunnel parallel and 3 mm posterior to the initial tunnel and to ensure no tunnel convergence. In addition, the passing cannulas were left in place until the sutures were pulled down the tunnels. The sutures in the anterior portion of the meniscal root were then shuttled through the anatomic transtibial bone tunnel, whereas the sutures in the posterior portion of the meniscal root were pulled through the posterior transtibial bone tunnel. The
UNRECOVERABLE MENISCAL ROOT REPAIR LOOSENING
Fig 1. Transtibial pullout repair of posteromedial meniscal root tear performed with 2esimple suture technique. (reproduced with permission from Padalecki et al.6)
sutures were both tied to a 4-mm 12-mm surgical fixation button over the anteromedial surface of the tibia using the same knotting technique as in the singletunnel technique. Once repairs were complete, the limbs were frozen at e20 C until individually tested because careful freezing of specimens has been shown to have little effect on the biomechanical properties of biological soft tissues.19 Biomechanical Testing The limbs were individually thawed for experimental testing. The anteromedial meniscal insertion was transected with a scalpel so that the meniscus was only attached to the tibia posteriorly by the meniscal root repair. The anterior portion of the meniscus was then rigidly clamped 1 cm from the sutures in line with the circumferential collagen fibers and marked with a surgical pen at the boundary of the clamp to confirm no slippage occurred during the experiment. The clamp was fastened to the actuator of a servohydraulic material testing machine (Bionic model 370.02; MTS, Eden Prairie, MN) equipped with a 500N load cell (1500ASK-100; Interface, Scottsdale, AZ) with a maximum error specification of 0.25 N. The potted diaphysis was fastened to a custom fixture allowing the tibia to be positioned with the transtibial tunnel (or tunnels) in the axial direction of the actuator (Fig 2). Both single- and double-tunnel repairs were subjected to the same cyclic loading procedure. All tests
3
were performed at room temperature, and the specimens were kept continuously moist with roomtemperature phosphate-buffered saline solution through the duration of the experiment. The repaired attachments were first tensioned for 10 cycles from 1 N to 10 N at 0.1 Hz to precondition and minimize creep within the repair. After preconditioning, the menisci were cyclically tensioned for 10,000 cycles from 10 to 30 N at 0.5 Hz. This loading protocol has been used by previous studies to simulate tension that the posteromedial meniscal root may experience under neutral rotation, a range of motion from 30 to 60 , and 500 N of tibiofemoral load, representing the range of motion and partial weight bearing seen in a typical postoperative rehabilitation regimen after meniscal root repair.10,11,20 Repairs were then returned to their original, unloaded position and allowed to recover for 30 minutes. Good intra-specimen repeatability for tension testing of ligaments has been shown with at least 30 minutes of rest; thus, this period was presumed to allow the meniscal root sufficient time for reasonable recovery.12 After rest, the repairs underwent another loading session of preconditioning for 10 cycles from 1 N to 10 N at 0.1 Hz and were then tensioned for an additional 1,000 cycles from 10 N to 30 N at 0.5 Hz to assess repair recoverability after rest. Repair displacement was measured at the peak displacement for each cycle as tension reached 30 N. The displacement of single- and double-tunnel repairs at 1, 100, 500, 1,000, 5,000, and 10,000 cycles was recorded for comparison between groups. Displacement at the testing machine actuator was measured, as in previous studies evaluating meniscal displacement.10,11 The difference in displacement between the first loading cycle and the first loading cycle after rest was calculated to evaluate whether repairs recovered to their original position or whether the simulated rehabilitative loading caused unrecoverable loosening within repairs. Recoverability was also evaluated by determining whether repairs returned to the displacement of the final loading cycle after rest and additional loading. The difference in displacement between cycle 10,000 and cycle 1,000 after the rest period was calculated to assess this recovery. Statistical Analysis By use of previous literature results, a sample size calculation was performed for this experiment. The standard deviation for the sample size calculation was assumed to be 0.6 mm, which was taken as an approximate value from a previous study.11 On the basis of an assumption of a significance level (a) of .05 and requirement of 80% power, group sample sizes of 8 were sufficiently powered to detect a 0.9-mm difference for comparisons between the single- and doubletunnel repairs.
4
B. D. STEINEMAN ET AL.
resultant P .05 and indicates that changes are unlikely to have occurred owing to random chance. The effect size with the bias-corrected Hedges g and its 95% confidence interval (CI) was calculated to interpret the practical importance of paired comparisons.21 The Hedges g effect size was used instead of the Cohen d effect size to account for the small sample size, giving a more conservative interpretation of the effect size.21 The Hedges g may be interpreted similarly to the Cohen d, with g equal to 0.2 for a small effect, g equal to 0.5 for a moderate effect, and g of 0.8 or greater for a large effect. A large effect size would suggest a large, important difference is present but still needs to be investigated within context to determine the clinical significance.
Results Repair displacements at cycles of interest were not significantly different between the single- and doubletunnel techniques (P .3 for all comparisons, Tables 1 and 2). Because there were no significant differences between the 2 groups, repair displacements for the singleand double-tunnel groups were pooled to increase the sample size of the reported repair displacements (Tables 1 and 2). Recoverability outcomes between the single- and double-tunnel repairs were also not significantly different; therefore, these outcomes were reported for the single-tunnel group, for the double-tunnel group, and for the groups pooled. The difference in displacement between the first loading cycle and the first loading cycle after rest was 1.59 0.69 mm for the pooled repair groups and was significantly different from 0 mm (P < .001, Fig 3A). The Hedges g effect size was calculated to be 2.1 with a 95% CI of 1.2 to 3.1. The measured difference in displacement between the first cycle and the first cycle after rest was 1.65 0.51 mm and 1.54 0.86 mm for single- and double-tunnel repairs, respectively (P ¼ .38). When single- and double-tunnel repair data were pooled, the difference in displacement between cycle 10,000 and cycle 1,000 after rest was 0.04 0.04 mm and was significantly different from 0 mm (P ¼ .004) (Fig 3B). The Hedges g effect size was calculated as 0.04 with a 95% CI of 0.01 to 0.07. The difference in displacement between cycle 10,000 and cycle 1,000 after rest was 0.05 0.05 mm and 0.04 0.03 mm for single- and double-tunnel repairs, respectively (P ¼ .29).
Fig 2. Experimental testing setup. The potted limbs were secured to allow the tibiae to be positioned with the transtibial tunnel (or tunnels) in the axial direction of the actuator. The repaired menisci were clamped with the circumferential fibers in line with the actuator.
The results were initially tested for normality using the Shapiro-Wilk test. Nonparametric statistical analyses were performed for non-normal distributions of specimen results instead of parametric analyses. MannWhitney U tests were performed to compare the singleand double-tunnel repairs for repair displacement at cycles of interest and for recoverability outcomes. A paired-samples t test was performed for comparison of the difference in displacement between the first loading cycles and the first loading cycles after rest to determine whether loosening occurred and was significantly different from 0 mm. The Wilcoxon signed rank test was performed for comparison of repair displacement at cycle 10,000 versus displacement at cycle 1,000 after rest to determine whether the repairs returned to their peak displacement within 1,000 additional cycles after rest. Statistical significance was defined as a
Table 1. Cyclic Displacement of Single- and Double-Tunnel Techniques and With Data Pooled Displacement, mm Group Single Double Pooled
1 Cycle
100 Cycles
500 Cycles
1,000 Cycles
5,000 Cycles
10,000 Cycles
2.20 0.33 2.25 0.57 2.22 0.45
2.98 0.47 2.86 0.70 2.92 0.58
3.36 0.61 3.30 0.81 3.33 0.69
3.73 0.62 3.52 0.88 3.63 0.74
4.33 0.77 4.11 1.09 4.22 0.92
4.55 0.83 4.38 1.11 4.47 1.00
NOTE. Data are reported as mean standard deviation. No significant differences were noted between the single- and double-tunnel groups.
UNRECOVERABLE MENISCAL ROOT REPAIR LOOSENING Table 2. Cyclic Displacement of Single- and Double-Tunnel Techniques After Rest and With Data Pooled Displacement, mm Group Single Double Pooled
1 Cycle After Rest
1,000 Cycles After Rest
3.86 0.77 3.78 1.11 3.82 0.92
4.51 0.85 4.39 1.25 4.45 1.04
NOTE. Data are reported as mean standard deviation. No significant differences were noted between the single- and double-tunnel groups.
Repairs did not appear to reach an equilibrium displacement within 10,000 cycles. Instead, the repair displacement gradually increased with additional cycles. The cycle peaks were fit to a linear regression model to identify the point within 10,000 cycles at which creep began to increase linearly with a rootmean-square error of 0.01 or less (Fig 3C). The median at which steady-state creep began was cycle 4,531 (interquartile range, cycle 3,752 to cycle 5,253). The steady-state creep rate was calculated as an increase of 0.05 0.02 mm per additional 1,000 cycles once the repair reached the point of linear progression. The suture-meniscus interfaces of repairs were inspected macroscopically after the testing procedure. The holes created by the suture needles when initially
5
passed through the meniscal root were visibly elongated from the sutures beginning to pull through the meniscal root in both single- and double-tunnel repairs (Fig 4). This elongation was permanent damage caused by sutures gradually cutting through the meniscal root during the loading procedure. None of the specimens exhibited failure of fixation during the experiment.
Discussion In this study, the major finding was that transtibial pullout meniscal root repairs showed a significant amount of unrecoverable loosening due to simulated rehabilitative loading. Although this loosening cannot be individually attributed to the meniscus-suture interface, sutures, or bone-button interface, the important finding of this study was that this overall loosening of repairs was unrecoverable with rest. The large effect size calculated from this measurement indicates that 10,000 cycles of simulated rehabilitation and rest had a substantial impact on root repair loosening. Despite the statistical significance and large effect size, the clinical relevance of the 1.59 mm of unrecoverable loosening and the resultant change regarding knee mechanics is unknown. A previous experimental study showed that an intact anteromedial meniscal insertion loosened medially by only 3 mm significantly affected the transmission of tibiofemoral loads through
Fig 3. Example of sample cyclic test data (black) and displacement (green) of first cycle and first cycle after rest (A), displacement (green) of cycle 10,000 and cycle 1,000 after rest (B), and steady-state creep rate (green) of repairs (C). The dotted yellow lines indicate intervals of 2,500 cycles.
6
B. D. STEINEMAN ET AL.
Fig 4. Macroscopic elongation of the suture holes in the meniscal root from the original position of the sutures (black arrow) to the final position (white arrow) was present in both single- and double-tunnel repairs after the loading procedure.
the meniscus.22 Direct comparison with our study is difficult because the 1.59 mm of loosening occurred from the suture repair whereas Stärke et al.22 assessed mechanics with 3 mm of loosening with the native meniscal insertion intact. The strength of the intact, native meniscal insertions has been shown to not be restored with repairs.23 Therefore, a weaker fixation of the meniscus and unrecoverable loosening of 1.59 mm in the repair may result in similarly poor load transmission. Further investigation of repair loosening and the effect on knee mechanics is warranted. Another finding of this study was that displacement in repairs continued to gradually increase with additional loading instead of reaching an equilibrium. An average increase in displacement of 0.25 mm occurred over the final 5,000 cycles of the experiment. Although this average increase in repair displacement with additional loading may not appear like much, this result may have implications for patients postoperatively. Because no equilibrium displacement was reached within 10,000 cycles before rest and second-look arthroscopy studies have shown that no healing is a possible outcome at 6 months postoperatively,13 the results of this study suggest that loosening may continue to progress in some patients with insufficient healing. Our study only assessed these repairs up to 11,000 total loading cycles; however, 11,000 steps or loading cycles would be a low estimation for patients participating in partial or full weight bearing up to 6 months postoperatively when little to no healing has been shown. Therefore, repair loosening may be significantly greater in patients with early, insufficient healing than the 1.59 mm determined experimentally in our study. This study may help to explain why some patients have no reduction or have an increase in meniscal extrusion at postoperative follow-up and
advocates for the importance of early healing progress.8,9 The displacement of repairs at cycle 10,000 and displacement at cycle 1,000 after rest were significantly different; however, the Hedges g effect size was low. The low effect size indicates that resting of the root repair contributed little to the recovery of repair displacement. Thus, patients with meniscal root repairs that loosen from rehabilitative loading may not recover the intended position of the repaired meniscus with rest. Our study only assessed repairs with 30 minutes of rest before another loading session. It is possible that 30 minutes was not sufficient for repairs to recover; however, previous studies have shown minimal change in the intra-specimen repeatability of viscoelastic soft tissues in tension with 30 minutes of rest.12 In addition, visible damage to sutures pulling through the meniscal root suggests that rest time is not the main concern. Instead, suture cutout with loading causes concern for repair loosening and should be addressed in future studies. Compared with previous repair displacement studies, this study supports the conclusion that single- and double-tunnel repairs mechanically respond similarly in tension. A couple of studies have assessed the 2esimple suture technique with either single- or double-tunnel repairs.10,11 Both of these studies have only assessed repairs up to 1,000 cycles, but they reported comparable displacement values with respect to our study. On the basis of reported means and standard deviations, there were no significant differences between displacement of either technique at 500 or 1,000 cycles. Conversely, the repair displacement at the earlier cycles of 1 and 100 was significantly different. The difference between displacement in this study and that in other repair displacement studies at early cycles may be explained by the loading rate used to transition between preconditioning and full loading cycles. The strain rate dependence of meniscal tissue could cause different repair displacements at initial cycles when transitioned using the 0.1 Hz of preconditioning or the 0.5 Hz of the full loading cycles. Because the displacement values in this study with respect to other studies are similar for the higher cycles, this likely had minimal effect on the results of the initial loading cycles of repairs. The results of this study may also be beneficial for refining the current root repair surgical technique. In this study, the 2esimple suture method with traditional sutures was used and progressive, unrecoverable loosening of repairs occurred. During comparisons of different repair techniques at 1,000 cycles, the 2esimple suture method showed low displacement with respect to other techniques.23 Other studies have shown that a 3esimple suture technique, simple cinch, or different configurations of the modified Mason-Allen technique are potential
UNRECOVERABLE MENISCAL ROOT REPAIR LOOSENING
candidates for reducing repair displacement.23-26 Although some of these suture techniques are perceived as more technically challenging, they may help to prevent unrecoverable loosening of repairs and the gradual damage that occurs with further loading shown with the 2esimple suture technique. Further analysis of suture techniques that have shown less repair displacement, such as the 3esimple suture or modified Mason-Allen technique, should be studied further to assess their ability to resist repair loosening. In addition, a previous study compared traditional sutures with suture tape and found no significant differences within 1,000 cycles; however, it is unknown whether the suture tape helps to prevent loosening and suture cutout with further loading.27 Wider sutures through the meniscal root would distribute the tensile load to a greater area at the suture-meniscus interface and decrease the shear stress. Therefore, the use of wider sutures, such as suture tape, may also need to be investigated further with respect to high-strength or standard sutures for their ability to resist loosening. Other characteristics of sutures such as stiffness and yield strength may also have an influence on repair loosening.28 Until there is a better understanding of how other suture techniques and suture types respond to extended loading and rest, we recommend conservative rehabilitation with all techniques to promote healing that should help support the repair. This study is clinically relevant because the results show there is a potential for progressive, unrecoverable loosening of meniscal root repairs until sufficient healing occurs. Thus, it is essential that guided rehabilitation protocols be followed in the first few weeks postoperatively to ensure that healing of the root repair occurs before loosening of the repair sutures due to repetitive knee motion. In addition, noneweight bearing during the initial 6 weeks means that patients should not be loading their repairs to allow the repairs time to heal; however, any noncompliance may put repairs at greater risk of insufficient healing and repair healing. Because early healing of meniscal root repairs is not well understood, it is unknown whether the amount of healing present at early rehabilitation will be able to mitigate the amount of loosening that occurs from repetitive loading. Therefore, further evaluation of healing progress and quality should be performed around 6 weeks, when partial weight bearing typically occurs. Studies evaluating healing progression and quality would help to determine the contribution of scar tissue that may increase the stability of repairs and determine whether augmentation of healing or conservative rehabilitation is necessary.23 Until healing progress is better understood, a slower incorporation of partial and full weight bearing may be beneficial for patients. Other studies have confirmed that meniscal root repairs, independent of suture technique, are not able to restore the full strength of the native meniscal
7
root. This fact emphasizes the importance of healing to the success of meniscal root repairs. Thus, slower incorporation of weight bearing may be necessary to ensure early healing occurs to support the repair. Limitations Little scar tissue development or no healing at all and an easily raised meniscus with probing have only been shown to occur in a minority of cases reported13,14; however, assessment of the quality of healing regarding mechanics and response of repairs without healing is warranted to better address patients susceptible to insufficient healing. Another potentially useful assessment of repairs would investigate intermittent loading in which rest occurs between shorter loading sessions. This would simulate a more physiological loading scenario because patients will rarely load their repairs 10,000 times during rehabilitation without rest. In addition, although previous studies have shown that biological tissues from frozen cadaveric specimens behave similarly to fresh samples, the recovery response of these tissues is understudied and could have affected the results of this experiment. Another potential limitation of this study was that specimens were tested at room temperature instead of body temperature. Room temperature has been shown to be an acceptable testing condition for nonabsorbable fixation methods with menisci,29 but the extent of our study’s duration at room temperature may have influenced the results. This study was also limited by the use of ovine menisci instead of human menisci, which may have affected the results. Despite this, ovine menisci are structurally and mechanically similar to human menisci, so the differences resulting from this are anticipated to be minimal.30
Conclusions This study showed that significant, unrecoverable loosening from rehabilitative loading occurred in singleand double-tunnel meniscal root repairs. Root repairs also gradually displaced with continued loading instead of reaching an equilibrium displacement after 10,000 cycles. This progressive, unrecoverable loosening needs to be studied further to better understand the resultant impact on knee mechanics. In addition, the quality and quantity of meniscal root repair healing at the time of rehabilitation should be studied to determine how susceptible patients are to repair loosening.
References 1. Bonasia DE, Pietro P, D’Amelio A, Cottino U, Rossi R. Meniscal root tear repair: Why, when and how? Orthop Rev 2015;7:5792. 2. Papalia R, Vasta S, Franceschi F, D’Adamio S, Maffulli N, Denaro V. Meniscal root tears: From basic science to ultimate surgery. Br Med Bull 2013;106:91-115.
8
B. D. STEINEMAN ET AL.
3. Krych AJ, Johnson NR, Mohan R, et al. Arthritis progression on serial MRIs following diagnosis of medial meniscal posterior horn root tear. J Knee Surg 2018;31:698-704. 4. Faucett SC, Geisler BP, Chahla J, et al. Meniscus root repair vs meniscectomy or nonoperative management to prevent knee osteoarthritis after medial meniscus root tears: Clinical and economic effectiveness [published online March 8, 2018]. Am J Sports Med, https://doi.org/10.1177/0363546518755754. 5. LaPrade CM, Jansson KS, Dornan G, Smith SD, Wijdicks CA, LaPrade RF. Altered tibiofemoral contact mechanics due to lateral meniscus posterior horn root avulsions and radial tears can be restored with in situ pullout suture repairs. J Bone Joint Surg Am 2014;96:471-479. 6. Padalecki JR, Jansson KS, Smith SD, et al. Biomechanical consequences of a complete radial tear adjacent to the medial meniscus posterior root attachment site: In situ pull-out repair restores derangement of joint mechanics. Am J Sports Med 2014;42:699-707. 7. LaPrade CM, Foad A, Smith SD, et al. Biomechanical consequences of a nonanatomic posterior medial meniscal root repair. Am J Sports Med 2015;43:912-920. 8. Feucht MJ, Kühle J, Bode G, et al. Arthroscopic transtibial pullout repair for posterior medial meniscus root tears: A systematic review of clinical, radiographic, and secondlook arthroscopic results. Arthroscopy 2015;31:1808-1816. 9. Chung KS, Ha JK, Ra HJ, Kim JG. A meta-analysis of clinical and radiographic outcomes of posterior horn medial meniscus root repairs. Knee Surg Sports Traumatol Arthrosc 2016;24:1455-1468. 10. Cerminara AJ, LaPrade CM, Smith SD, Ellman MB, Wijdicks CA, LaPrade RF. Biomechanical evaluation of a transtibial pull-out meniscal root repair: Challenging the bungee effect. Am J Sports Med 2014;42:2988-2995. 11. LaPrade CM, LaPrade MD, Turnbull TL, Wijdicks CA, LaPrade RF. Biomechanical evaluation of the transtibial pull-out technique for posterior medial meniscal root repairs using 1 and 2 transtibial bone tunnels. Am J Sports Med 2015;43:899-904. 12. Öhman C, Baleani M, Viceconti M. Repeatability of experimental procedures to determine mechanical behavior of ligaments. Acta Bioeng Biomech 2009;11:19-23. 13. Cho JH, Song JG. Second-look arthroscopic assessment and clinical results of modified pull-out suture for posterior root tear of the medial meniscus. Knee Surg Relat Res 2014;26:106-113. 14. Lee SS, Ahn JH, Kim JH, Kyung BS, Wang JH. Evaluation of healing after medial meniscal root repair using secondlook arthroscopy, clinical, and radiological criteria. Am J Sports Med 2018;46:2661-2668. 15. Bhatia S, LaPrade CM, Ellman MB, LaPrade RF. Meniscal root tears: Significance, diagnosis, and treatment. Am J Sports Med 2014;42:3016-3030. 16. Kim YM, Joo YB, Noh CK, Park IY. The optimal suture site for the repair of posterior horn root tears: Biomechanical evaluation of pullout strength in porcine menisci. Knee Surg Relat Res 2016;28:147-152. 17. LaPrade RF, LaPrade CM, Ellman MB, Turnbull TL, Cerminara AJ, Wijdicks CA. Cyclic displacement after
18.
19.
20.
21. 22.
23.
24.
25.
26.
27.
28.
29.
30.
meniscal root repair fixation: A human biomechanical evaluation. Am J Sports Med 2015;43:892-898. LaPrade RF, Matheny M, Moulton SG, James EW, Dean CS. Posterior meniscal root repairs: Outcomes of an anatomic transtibial pull-out technique. Am J Sports Med 2017;45:884-891. 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. Stärke C, Kopf S, Lippisch R, Lohmann CH, Becker R. Tensile forces on repaired medial meniscal root tears. Arthroscopy 2013;29:205-212. Hedges LV, Olkin I. Statistical methods for meta-analysis. New York: Academic Press, 1985. Stärke C, Kopf S, Gröbel KH, Becker R. The effect of a nonanatomic repair of the meniscal horn attachment on meniscal tension: A biomechanical study. Arthroscopy 2010;26:358-365. Feucht MJ, Grande E, Brunhuber J, Burgkart R, Imhoff AB, Braun S. Biomechanical evaluation of different suture techniques for arthroscopic transtibial pull-out repair of posterior medial meniscus root tears. Am J Sports Med 2013;41:2784-2790. Camarda L, Pitarresi G, Lauria M, Fazzari F, D’Arienzo M. Three single loops enhance the biomechanical behavior of the transtibial pull-out technique for posterior meniscal root repair. Arch Orthop Trauma Surg 2017;137: 1301-1306. Fujii M, Furumatsu T, Xue H, et al. Tensile strength of the pullout repair technique for the medial meniscus posterior root tear: A porcine study. Int Orthop 2017;41: 2113-2118. Krych AJ, Johnson NR, Wu IT, Smith PA, Stuart MJ. A simple cinch is superior to a locking loop for meniscus root repair: A human biomechanical comparison of suture constructs in a transtibial pull-out model. Knee Surg Sports Traumatol Arthrosc 2018;26: 2239-2244. Perez-Blanca A, Prado Nóvoa M, Lombardo Torre M, Espejo-Reina A, Ezquerro Juanco F, Espejo-Baena A. The role of suture cutout in the failure of meniscal root repair during the early post-operative period: A biomechanical study. Int Orthop 2018;42:811-818. Feucht MJ, Grande E, Brunhuber J, et al. Biomechanical evaluation of different suture materials for arthroscopic transtibial pull-out repair of posterior meniscus root tears. Knee Surg Sports Traumatol Arthrosc 2015;23: 132-139. Brucker PU, Favre P, Puskas GJ, von Campe A, Koch PP, Meyer DC. Influence of test temperature on biomechanical properties of all-inside meniscal repair devices and inside-out meniscus suturesdEvaluation of an isolated distraction loading, worst-case scenario. Clin Biomech 2011;26:749-753. Sandmann GH, Adamczyk C, Garcia EG, et al. Biomechanical comparison of menisci from different species and artificial constructs. BMC Musculoskelet Disord 2013;14:324.
Purpose: To determine whether meniscal root repairs recover from displacement due to rehabilitative loading. Methods: Transtibial pullout repairs of the posteromedial meniscal root were performed in 16 cadaveric ovine knees. Single- and double-tunnel repairs using the 2esimple suture technique were cyclically loaded in tension to 10,000 cycles, allowed to rest, and loaded in tension again. Paired differences in displacement with rest were recorded to evaluate recoverability. Displacement of repairs at cycles of interest was recorded, and the response of repairs to 10,000 cycles was assessed. Results: All outcomes were not significantly different between the single- and double-tunnel techniques; therefore, the results were pooled. The difference in displacement between the first cycle and the first cycle after rest was 1.59 0.69 mm. Repair displacement did not reach an equilibrium within 10,000 cycles and instead resulted in a steady increase in displacement of 0.05 0.02 mm per additional 1,000 cycles. Sutures macroscopically began to cut out of the meniscus in both single- and double-tunnel repairs. Conclusions: This study showed that significant, unrecoverable loosening from rehabilitative loading occurred in single- and double-tunnel meniscal root repairs. Root repairs also gradually displaced with continued loading instead of reaching an equilibrium displacement after 10,000 cycles. This progressive, unrecoverable loosening needs to be studied further to better understand the resultant impact on knee mechanics. In addition, the quality and quantity of meniscal root repair healing at the time of rehabilitation should be studied to determine how susceptible patients are to repair loosening. Clinical Relevance: Rehabilitative loading caused unrecoverable and progressive loosening of root repairs, showing the importance of healing before loading. Investigations on the effects of loosening on mechanics and the quality of repair healing at weight bearing are necessary to better understand the clinical implications.
M
eniscal root tears may occur acutely, chronically, or as a result of iatrogenic injury.1 The posterior meniscal insertions are more susceptible to tears, especially when loaded when the knee is in deep
From the Department of Biomechanics, Hospital for Special Surgery (B.D.S.), New York, New York; The Steadman Clinic (R.F.L.), Vail, Colorado; Steadman Philippon Research Institute (R.F.L.), Vail, Colorado; and Department of Biomedical Engineering, University of Massachusetts (T.L.H.D.), Amherst, Massachusetts, U.S.A. The authors report the following potential conflicts of interest or sources of funding: R.F.L. is a consultant for and receives royalties from Arthrex, Össur, and Smith & Nephew. Full ICMJE author disclosure forms are available for this article online, as supplementary material. Received July 6, 2018; accepted November 16, 2018. Address correspondence to Tammy L. Haut Donahue, Ph.D., Department of Biomedical Engineering, University of Massachusetts, 130 Natural Resources Rd, Amherst, MA 01003, U.S.A. E-mail: [email protected] Ó 2018 by the Arthroscopy Association of North America 0749-8063/18813/$36.00 https://doi.org/10.1016/j.arthro.2018.11.041
flexion.2 Untreated meniscal root tears are becoming increasingly recognized with progressive increases in meniscal extrusion and articular cartilage degeneration detectable within a year.3 Therefore, surgical treatment is essential to reduce the risk of osteoarthritis development after meniscal root tears. Repairs of meniscal root tears are becoming more desirable than meniscectomy because they significantly improve clinical and radiologic outcomes while reducing cost.4 Previous ex vivo biomechanical experiments have reported that anatomic, transtibial pullout meniscal root repairs nearly restore intact contact mechanics immediately after repair.5-7 Despite this, postoperative follow-up studies have shown that meniscal extrusion was only restored in 56% of patients and progression of osteoarthritis was not always prevented.8,9 A potential reason for limited success of repairs may be due to significant displacement in meniscal root repairs shown to occur in biomechanical studies that simulated rehabilitative loading.10,11 Although these studies showed
Arthroscopy: The Journal of Arthroscopic and Related Surgery, Vol
-,
No
-
(Month), 2019: pp 1-8
1
2
B. D. STEINEMAN ET AL.
that displacement accumulates with loading of repairs, they were unable to distinguish whether this displacement was due to viscoelastic creep of the meniscal root or due to permanent, unrecoverable loosening of the repair. Displacement due to tissue viscoelasticity would be recoverable and result in small changes in the intraspecimen repeatability of tests with appropriate rest.12 If these repairs are unable to recover to the initial state of repair, then nearly restorative repairs may loosen into a less restorative position in vivo, resulting in decreased load transmission through the menisci and increased load transfer through the articular cartilage. In previous second-look arthroscopy studies of meniscal root repairs, little to no healing has been shown to occur in some patients with earliest follow-up at 6 months postoperatively.13,14 Failed healing may be defined as no continuity, an easily raised meniscus on probing, and no evidence of meniscal healing at the repair site.13 Another study defined little healing as no true continuity of the meniscus with some connecting scar tissue and the meniscus being easily raised on probing.13 Previous studies evaluating repair displacement have loaded repairs for only 1,000 cycles because they assumed some healing occurs before partial weight bearing in rehabilitation.10,11 Therefore, 1,000 loading cycles does not represent certain cases of meniscal root repairs that may occur in vivo. Although prescribed weight bearing varies between patients, partial weight bearing typically begins around 6 weeks.15 Because some cases of partial or no healing have been observed and the progress of healing near rehabilitative loading is unknown, the extent of repair displacement needs to be examined further. In addition, although previous studies showed that single- and double-tunnel repairs mechanically respond similarly,11 both repairs were also performed and compared to assess for differences in recoverability of displacement. The purpose of this study was to determine whether meniscal root repairs recover from displacement due to rehabilitative loading. It was hypothesized that a significant amount of displacement due to rehabilitative loading would be caused by unrecoverable loosening of repairs. It was also hypothesized that single- and double-tunnel repairs would not be significantly different in repair displacement or in recovery from this displacement.
Methods Specimen Preparation Sixteen cadaveric knees from adult sheep that had been euthanized for unrelated purposes were collected from the Colorado State University Veterinary Teaching Hospital; therefore, these specimens were exempt from Institutional Animal Care and Use Committee approval. The animals cared for and evaluated by veterinarians at the Colorado State University Veterinary Teaching
Hospital that were used in this study were approximately 3 years old and weighed between 65 and 90 kg at the time of euthanasia. Knees were dissected free of all extra-articular skin, muscle, and soft tissue while fresh without previous freezing. The anterior and posterior cruciate ligaments, medial and fibular collateral ligaments, and lateral meniscal insertions were sharply transected to expose the tibial plateau with only the medial menisci intact. The tibiae were then cut at the mid diaphysis to remove the distal portion and potted in a 2-part urethane casting resin (Smooth Cast 321; Smooth-on, Easton, PA). Repair Technique The knees were randomly distributed between the single- and double-tunnel transtibial pullout repair groups, with 8 specimens per group. All repairs were performed and assessed postoperatively by a boardcertified sports medicine orthopaedic surgeon (R.F.L.) experienced with meniscal root repairs. Initially, the posteromedial meniscal root insertion was cut with a scalpel at the interface with the tibial plateau, and the insertion area was marked with a surgical pen to identify anatomic placement. For the single-tunnel group, a 4.5-mm-diameter transtibial tunnel was created using a guide pin and reamer (Smith & Nephew, Andover, MA) in the middle of the posteromedial meniscal root attachment. Two No. 2 nonabsorbable sutures (Ultrabraid; Smith & Nephew) were passed between 5 and 7 mm medial within the meniscal root.16 The 2esimple suture technique was performed for all repairs because this technique has shown the lowest technical difficulty and resists displacement from cyclic loading better than other techniques assessed.17 In addition, this technique is the one that most current published studies have used (Fig 1).6,18 The sutures were placed into a looped nitinol wire, pulled through the tunnel, manually tensioned to reduce the menisci to the native position, and firmly tied to a 4-mm 12mm surgical fixation button (EndoButton; Smith & Nephew) over the anteromedial tibia using a surgeon’s knot with 5 suture throws followed by 5 half-hitches on alternating posts over the fixation button. For the double-tunnel group, the first transtibial tunnel was drilled in the same manner as in the singletunnel group, with a diameter of 3 mm. A second transtibial tunnel was then drilled using a calibrated offset guide to position the second tunnel parallel and 3 mm posterior to the initial tunnel and to ensure no tunnel convergence. In addition, the passing cannulas were left in place until the sutures were pulled down the tunnels. The sutures in the anterior portion of the meniscal root were then shuttled through the anatomic transtibial bone tunnel, whereas the sutures in the posterior portion of the meniscal root were pulled through the posterior transtibial bone tunnel. The
UNRECOVERABLE MENISCAL ROOT REPAIR LOOSENING
Fig 1. Transtibial pullout repair of posteromedial meniscal root tear performed with 2esimple suture technique. (reproduced with permission from Padalecki et al.6)
sutures were both tied to a 4-mm 12-mm surgical fixation button over the anteromedial surface of the tibia using the same knotting technique as in the singletunnel technique. Once repairs were complete, the limbs were frozen at e20 C until individually tested because careful freezing of specimens has been shown to have little effect on the biomechanical properties of biological soft tissues.19 Biomechanical Testing The limbs were individually thawed for experimental testing. The anteromedial meniscal insertion was transected with a scalpel so that the meniscus was only attached to the tibia posteriorly by the meniscal root repair. The anterior portion of the meniscus was then rigidly clamped 1 cm from the sutures in line with the circumferential collagen fibers and marked with a surgical pen at the boundary of the clamp to confirm no slippage occurred during the experiment. The clamp was fastened to the actuator of a servohydraulic material testing machine (Bionic model 370.02; MTS, Eden Prairie, MN) equipped with a 500N load cell (1500ASK-100; Interface, Scottsdale, AZ) with a maximum error specification of 0.25 N. The potted diaphysis was fastened to a custom fixture allowing the tibia to be positioned with the transtibial tunnel (or tunnels) in the axial direction of the actuator (Fig 2). Both single- and double-tunnel repairs were subjected to the same cyclic loading procedure. All tests
3
were performed at room temperature, and the specimens were kept continuously moist with roomtemperature phosphate-buffered saline solution through the duration of the experiment. The repaired attachments were first tensioned for 10 cycles from 1 N to 10 N at 0.1 Hz to precondition and minimize creep within the repair. After preconditioning, the menisci were cyclically tensioned for 10,000 cycles from 10 to 30 N at 0.5 Hz. This loading protocol has been used by previous studies to simulate tension that the posteromedial meniscal root may experience under neutral rotation, a range of motion from 30 to 60 , and 500 N of tibiofemoral load, representing the range of motion and partial weight bearing seen in a typical postoperative rehabilitation regimen after meniscal root repair.10,11,20 Repairs were then returned to their original, unloaded position and allowed to recover for 30 minutes. Good intra-specimen repeatability for tension testing of ligaments has been shown with at least 30 minutes of rest; thus, this period was presumed to allow the meniscal root sufficient time for reasonable recovery.12 After rest, the repairs underwent another loading session of preconditioning for 10 cycles from 1 N to 10 N at 0.1 Hz and were then tensioned for an additional 1,000 cycles from 10 N to 30 N at 0.5 Hz to assess repair recoverability after rest. Repair displacement was measured at the peak displacement for each cycle as tension reached 30 N. The displacement of single- and double-tunnel repairs at 1, 100, 500, 1,000, 5,000, and 10,000 cycles was recorded for comparison between groups. Displacement at the testing machine actuator was measured, as in previous studies evaluating meniscal displacement.10,11 The difference in displacement between the first loading cycle and the first loading cycle after rest was calculated to evaluate whether repairs recovered to their original position or whether the simulated rehabilitative loading caused unrecoverable loosening within repairs. Recoverability was also evaluated by determining whether repairs returned to the displacement of the final loading cycle after rest and additional loading. The difference in displacement between cycle 10,000 and cycle 1,000 after the rest period was calculated to assess this recovery. Statistical Analysis By use of previous literature results, a sample size calculation was performed for this experiment. The standard deviation for the sample size calculation was assumed to be 0.6 mm, which was taken as an approximate value from a previous study.11 On the basis of an assumption of a significance level (a) of .05 and requirement of 80% power, group sample sizes of 8 were sufficiently powered to detect a 0.9-mm difference for comparisons between the single- and doubletunnel repairs.
4
B. D. STEINEMAN ET AL.
resultant P .05 and indicates that changes are unlikely to have occurred owing to random chance. The effect size with the bias-corrected Hedges g and its 95% confidence interval (CI) was calculated to interpret the practical importance of paired comparisons.21 The Hedges g effect size was used instead of the Cohen d effect size to account for the small sample size, giving a more conservative interpretation of the effect size.21 The Hedges g may be interpreted similarly to the Cohen d, with g equal to 0.2 for a small effect, g equal to 0.5 for a moderate effect, and g of 0.8 or greater for a large effect. A large effect size would suggest a large, important difference is present but still needs to be investigated within context to determine the clinical significance.
Results Repair displacements at cycles of interest were not significantly different between the single- and doubletunnel techniques (P .3 for all comparisons, Tables 1 and 2). Because there were no significant differences between the 2 groups, repair displacements for the singleand double-tunnel groups were pooled to increase the sample size of the reported repair displacements (Tables 1 and 2). Recoverability outcomes between the single- and double-tunnel repairs were also not significantly different; therefore, these outcomes were reported for the single-tunnel group, for the double-tunnel group, and for the groups pooled. The difference in displacement between the first loading cycle and the first loading cycle after rest was 1.59 0.69 mm for the pooled repair groups and was significantly different from 0 mm (P < .001, Fig 3A). The Hedges g effect size was calculated to be 2.1 with a 95% CI of 1.2 to 3.1. The measured difference in displacement between the first cycle and the first cycle after rest was 1.65 0.51 mm and 1.54 0.86 mm for single- and double-tunnel repairs, respectively (P ¼ .38). When single- and double-tunnel repair data were pooled, the difference in displacement between cycle 10,000 and cycle 1,000 after rest was 0.04 0.04 mm and was significantly different from 0 mm (P ¼ .004) (Fig 3B). The Hedges g effect size was calculated as 0.04 with a 95% CI of 0.01 to 0.07. The difference in displacement between cycle 10,000 and cycle 1,000 after rest was 0.05 0.05 mm and 0.04 0.03 mm for single- and double-tunnel repairs, respectively (P ¼ .29).
Fig 2. Experimental testing setup. The potted limbs were secured to allow the tibiae to be positioned with the transtibial tunnel (or tunnels) in the axial direction of the actuator. The repaired menisci were clamped with the circumferential fibers in line with the actuator.
The results were initially tested for normality using the Shapiro-Wilk test. Nonparametric statistical analyses were performed for non-normal distributions of specimen results instead of parametric analyses. MannWhitney U tests were performed to compare the singleand double-tunnel repairs for repair displacement at cycles of interest and for recoverability outcomes. A paired-samples t test was performed for comparison of the difference in displacement between the first loading cycles and the first loading cycles after rest to determine whether loosening occurred and was significantly different from 0 mm. The Wilcoxon signed rank test was performed for comparison of repair displacement at cycle 10,000 versus displacement at cycle 1,000 after rest to determine whether the repairs returned to their peak displacement within 1,000 additional cycles after rest. Statistical significance was defined as a
Table 1. Cyclic Displacement of Single- and Double-Tunnel Techniques and With Data Pooled Displacement, mm Group Single Double Pooled
1 Cycle
100 Cycles
500 Cycles
1,000 Cycles
5,000 Cycles
10,000 Cycles
2.20 0.33 2.25 0.57 2.22 0.45
2.98 0.47 2.86 0.70 2.92 0.58
3.36 0.61 3.30 0.81 3.33 0.69
3.73 0.62 3.52 0.88 3.63 0.74
4.33 0.77 4.11 1.09 4.22 0.92
4.55 0.83 4.38 1.11 4.47 1.00
NOTE. Data are reported as mean standard deviation. No significant differences were noted between the single- and double-tunnel groups.
UNRECOVERABLE MENISCAL ROOT REPAIR LOOSENING Table 2. Cyclic Displacement of Single- and Double-Tunnel Techniques After Rest and With Data Pooled Displacement, mm Group Single Double Pooled
1 Cycle After Rest
1,000 Cycles After Rest
3.86 0.77 3.78 1.11 3.82 0.92
4.51 0.85 4.39 1.25 4.45 1.04
NOTE. Data are reported as mean standard deviation. No significant differences were noted between the single- and double-tunnel groups.
Repairs did not appear to reach an equilibrium displacement within 10,000 cycles. Instead, the repair displacement gradually increased with additional cycles. The cycle peaks were fit to a linear regression model to identify the point within 10,000 cycles at which creep began to increase linearly with a rootmean-square error of 0.01 or less (Fig 3C). The median at which steady-state creep began was cycle 4,531 (interquartile range, cycle 3,752 to cycle 5,253). The steady-state creep rate was calculated as an increase of 0.05 0.02 mm per additional 1,000 cycles once the repair reached the point of linear progression. The suture-meniscus interfaces of repairs were inspected macroscopically after the testing procedure. The holes created by the suture needles when initially
5
passed through the meniscal root were visibly elongated from the sutures beginning to pull through the meniscal root in both single- and double-tunnel repairs (Fig 4). This elongation was permanent damage caused by sutures gradually cutting through the meniscal root during the loading procedure. None of the specimens exhibited failure of fixation during the experiment.
Discussion In this study, the major finding was that transtibial pullout meniscal root repairs showed a significant amount of unrecoverable loosening due to simulated rehabilitative loading. Although this loosening cannot be individually attributed to the meniscus-suture interface, sutures, or bone-button interface, the important finding of this study was that this overall loosening of repairs was unrecoverable with rest. The large effect size calculated from this measurement indicates that 10,000 cycles of simulated rehabilitation and rest had a substantial impact on root repair loosening. Despite the statistical significance and large effect size, the clinical relevance of the 1.59 mm of unrecoverable loosening and the resultant change regarding knee mechanics is unknown. A previous experimental study showed that an intact anteromedial meniscal insertion loosened medially by only 3 mm significantly affected the transmission of tibiofemoral loads through
Fig 3. Example of sample cyclic test data (black) and displacement (green) of first cycle and first cycle after rest (A), displacement (green) of cycle 10,000 and cycle 1,000 after rest (B), and steady-state creep rate (green) of repairs (C). The dotted yellow lines indicate intervals of 2,500 cycles.
6
B. D. STEINEMAN ET AL.
Fig 4. Macroscopic elongation of the suture holes in the meniscal root from the original position of the sutures (black arrow) to the final position (white arrow) was present in both single- and double-tunnel repairs after the loading procedure.
the meniscus.22 Direct comparison with our study is difficult because the 1.59 mm of loosening occurred from the suture repair whereas Stärke et al.22 assessed mechanics with 3 mm of loosening with the native meniscal insertion intact. The strength of the intact, native meniscal insertions has been shown to not be restored with repairs.23 Therefore, a weaker fixation of the meniscus and unrecoverable loosening of 1.59 mm in the repair may result in similarly poor load transmission. Further investigation of repair loosening and the effect on knee mechanics is warranted. Another finding of this study was that displacement in repairs continued to gradually increase with additional loading instead of reaching an equilibrium. An average increase in displacement of 0.25 mm occurred over the final 5,000 cycles of the experiment. Although this average increase in repair displacement with additional loading may not appear like much, this result may have implications for patients postoperatively. Because no equilibrium displacement was reached within 10,000 cycles before rest and second-look arthroscopy studies have shown that no healing is a possible outcome at 6 months postoperatively,13 the results of this study suggest that loosening may continue to progress in some patients with insufficient healing. Our study only assessed these repairs up to 11,000 total loading cycles; however, 11,000 steps or loading cycles would be a low estimation for patients participating in partial or full weight bearing up to 6 months postoperatively when little to no healing has been shown. Therefore, repair loosening may be significantly greater in patients with early, insufficient healing than the 1.59 mm determined experimentally in our study. This study may help to explain why some patients have no reduction or have an increase in meniscal extrusion at postoperative follow-up and
advocates for the importance of early healing progress.8,9 The displacement of repairs at cycle 10,000 and displacement at cycle 1,000 after rest were significantly different; however, the Hedges g effect size was low. The low effect size indicates that resting of the root repair contributed little to the recovery of repair displacement. Thus, patients with meniscal root repairs that loosen from rehabilitative loading may not recover the intended position of the repaired meniscus with rest. Our study only assessed repairs with 30 minutes of rest before another loading session. It is possible that 30 minutes was not sufficient for repairs to recover; however, previous studies have shown minimal change in the intra-specimen repeatability of viscoelastic soft tissues in tension with 30 minutes of rest.12 In addition, visible damage to sutures pulling through the meniscal root suggests that rest time is not the main concern. Instead, suture cutout with loading causes concern for repair loosening and should be addressed in future studies. Compared with previous repair displacement studies, this study supports the conclusion that single- and double-tunnel repairs mechanically respond similarly in tension. A couple of studies have assessed the 2esimple suture technique with either single- or double-tunnel repairs.10,11 Both of these studies have only assessed repairs up to 1,000 cycles, but they reported comparable displacement values with respect to our study. On the basis of reported means and standard deviations, there were no significant differences between displacement of either technique at 500 or 1,000 cycles. Conversely, the repair displacement at the earlier cycles of 1 and 100 was significantly different. The difference between displacement in this study and that in other repair displacement studies at early cycles may be explained by the loading rate used to transition between preconditioning and full loading cycles. The strain rate dependence of meniscal tissue could cause different repair displacements at initial cycles when transitioned using the 0.1 Hz of preconditioning or the 0.5 Hz of the full loading cycles. Because the displacement values in this study with respect to other studies are similar for the higher cycles, this likely had minimal effect on the results of the initial loading cycles of repairs. The results of this study may also be beneficial for refining the current root repair surgical technique. In this study, the 2esimple suture method with traditional sutures was used and progressive, unrecoverable loosening of repairs occurred. During comparisons of different repair techniques at 1,000 cycles, the 2esimple suture method showed low displacement with respect to other techniques.23 Other studies have shown that a 3esimple suture technique, simple cinch, or different configurations of the modified Mason-Allen technique are potential
UNRECOVERABLE MENISCAL ROOT REPAIR LOOSENING
candidates for reducing repair displacement.23-26 Although some of these suture techniques are perceived as more technically challenging, they may help to prevent unrecoverable loosening of repairs and the gradual damage that occurs with further loading shown with the 2esimple suture technique. Further analysis of suture techniques that have shown less repair displacement, such as the 3esimple suture or modified Mason-Allen technique, should be studied further to assess their ability to resist repair loosening. In addition, a previous study compared traditional sutures with suture tape and found no significant differences within 1,000 cycles; however, it is unknown whether the suture tape helps to prevent loosening and suture cutout with further loading.27 Wider sutures through the meniscal root would distribute the tensile load to a greater area at the suture-meniscus interface and decrease the shear stress. Therefore, the use of wider sutures, such as suture tape, may also need to be investigated further with respect to high-strength or standard sutures for their ability to resist loosening. Other characteristics of sutures such as stiffness and yield strength may also have an influence on repair loosening.28 Until there is a better understanding of how other suture techniques and suture types respond to extended loading and rest, we recommend conservative rehabilitation with all techniques to promote healing that should help support the repair. This study is clinically relevant because the results show there is a potential for progressive, unrecoverable loosening of meniscal root repairs until sufficient healing occurs. Thus, it is essential that guided rehabilitation protocols be followed in the first few weeks postoperatively to ensure that healing of the root repair occurs before loosening of the repair sutures due to repetitive knee motion. In addition, noneweight bearing during the initial 6 weeks means that patients should not be loading their repairs to allow the repairs time to heal; however, any noncompliance may put repairs at greater risk of insufficient healing and repair healing. Because early healing of meniscal root repairs is not well understood, it is unknown whether the amount of healing present at early rehabilitation will be able to mitigate the amount of loosening that occurs from repetitive loading. Therefore, further evaluation of healing progress and quality should be performed around 6 weeks, when partial weight bearing typically occurs. Studies evaluating healing progression and quality would help to determine the contribution of scar tissue that may increase the stability of repairs and determine whether augmentation of healing or conservative rehabilitation is necessary.23 Until healing progress is better understood, a slower incorporation of partial and full weight bearing may be beneficial for patients. Other studies have confirmed that meniscal root repairs, independent of suture technique, are not able to restore the full strength of the native meniscal
7
root. This fact emphasizes the importance of healing to the success of meniscal root repairs. Thus, slower incorporation of weight bearing may be necessary to ensure early healing occurs to support the repair. Limitations Little scar tissue development or no healing at all and an easily raised meniscus with probing have only been shown to occur in a minority of cases reported13,14; however, assessment of the quality of healing regarding mechanics and response of repairs without healing is warranted to better address patients susceptible to insufficient healing. Another potentially useful assessment of repairs would investigate intermittent loading in which rest occurs between shorter loading sessions. This would simulate a more physiological loading scenario because patients will rarely load their repairs 10,000 times during rehabilitation without rest. In addition, although previous studies have shown that biological tissues from frozen cadaveric specimens behave similarly to fresh samples, the recovery response of these tissues is understudied and could have affected the results of this experiment. Another potential limitation of this study was that specimens were tested at room temperature instead of body temperature. Room temperature has been shown to be an acceptable testing condition for nonabsorbable fixation methods with menisci,29 but the extent of our study’s duration at room temperature may have influenced the results. This study was also limited by the use of ovine menisci instead of human menisci, which may have affected the results. Despite this, ovine menisci are structurally and mechanically similar to human menisci, so the differences resulting from this are anticipated to be minimal.30
Conclusions This study showed that significant, unrecoverable loosening from rehabilitative loading occurred in singleand double-tunnel meniscal root repairs. Root repairs also gradually displaced with continued loading instead of reaching an equilibrium displacement after 10,000 cycles. This progressive, unrecoverable loosening needs to be studied further to better understand the resultant impact on knee mechanics. In addition, the quality and quantity of meniscal root repair healing at the time of rehabilitation should be studied to determine how susceptible patients are to repair loosening.
References 1. Bonasia DE, Pietro P, D’Amelio A, Cottino U, Rossi R. Meniscal root tear repair: Why, when and how? Orthop Rev 2015;7:5792. 2. Papalia R, Vasta S, Franceschi F, D’Adamio S, Maffulli N, Denaro V. Meniscal root tears: From basic science to ultimate surgery. Br Med Bull 2013;106:91-115.
8
B. D. STEINEMAN ET AL.
3. Krych AJ, Johnson NR, Mohan R, et al. Arthritis progression on serial MRIs following diagnosis of medial meniscal posterior horn root tear. J Knee Surg 2018;31:698-704. 4. Faucett SC, Geisler BP, Chahla J, et al. Meniscus root repair vs meniscectomy or nonoperative management to prevent knee osteoarthritis after medial meniscus root tears: Clinical and economic effectiveness [published online March 8, 2018]. Am J Sports Med, https://doi.org/10.1177/0363546518755754. 5. LaPrade CM, Jansson KS, Dornan G, Smith SD, Wijdicks CA, LaPrade RF. Altered tibiofemoral contact mechanics due to lateral meniscus posterior horn root avulsions and radial tears can be restored with in situ pullout suture repairs. J Bone Joint Surg Am 2014;96:471-479. 6. Padalecki JR, Jansson KS, Smith SD, et al. Biomechanical consequences of a complete radial tear adjacent to the medial meniscus posterior root attachment site: In situ pull-out repair restores derangement of joint mechanics. Am J Sports Med 2014;42:699-707. 7. LaPrade CM, Foad A, Smith SD, et al. Biomechanical consequences of a nonanatomic posterior medial meniscal root repair. Am J Sports Med 2015;43:912-920. 8. Feucht MJ, Kühle J, Bode G, et al. Arthroscopic transtibial pullout repair for posterior medial meniscus root tears: A systematic review of clinical, radiographic, and secondlook arthroscopic results. Arthroscopy 2015;31:1808-1816. 9. Chung KS, Ha JK, Ra HJ, Kim JG. A meta-analysis of clinical and radiographic outcomes of posterior horn medial meniscus root repairs. Knee Surg Sports Traumatol Arthrosc 2016;24:1455-1468. 10. Cerminara AJ, LaPrade CM, Smith SD, Ellman MB, Wijdicks CA, LaPrade RF. Biomechanical evaluation of a transtibial pull-out meniscal root repair: Challenging the bungee effect. Am J Sports Med 2014;42:2988-2995. 11. LaPrade CM, LaPrade MD, Turnbull TL, Wijdicks CA, LaPrade RF. Biomechanical evaluation of the transtibial pull-out technique for posterior medial meniscal root repairs using 1 and 2 transtibial bone tunnels. Am J Sports Med 2015;43:899-904. 12. Öhman C, Baleani M, Viceconti M. Repeatability of experimental procedures to determine mechanical behavior of ligaments. Acta Bioeng Biomech 2009;11:19-23. 13. Cho JH, Song JG. Second-look arthroscopic assessment and clinical results of modified pull-out suture for posterior root tear of the medial meniscus. Knee Surg Relat Res 2014;26:106-113. 14. Lee SS, Ahn JH, Kim JH, Kyung BS, Wang JH. Evaluation of healing after medial meniscal root repair using secondlook arthroscopy, clinical, and radiological criteria. Am J Sports Med 2018;46:2661-2668. 15. Bhatia S, LaPrade CM, Ellman MB, LaPrade RF. Meniscal root tears: Significance, diagnosis, and treatment. Am J Sports Med 2014;42:3016-3030. 16. Kim YM, Joo YB, Noh CK, Park IY. The optimal suture site for the repair of posterior horn root tears: Biomechanical evaluation of pullout strength in porcine menisci. Knee Surg Relat Res 2016;28:147-152. 17. LaPrade RF, LaPrade CM, Ellman MB, Turnbull TL, Cerminara AJ, Wijdicks CA. Cyclic displacement after
18.
19.
20.
21. 22.
23.
24.
25.
26.
27.
28.
29.
30.
meniscal root repair fixation: A human biomechanical evaluation. Am J Sports Med 2015;43:892-898. LaPrade RF, Matheny M, Moulton SG, James EW, Dean CS. Posterior meniscal root repairs: Outcomes of an anatomic transtibial pull-out technique. Am J Sports Med 2017;45:884-891. 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. Stärke C, Kopf S, Lippisch R, Lohmann CH, Becker R. Tensile forces on repaired medial meniscal root tears. Arthroscopy 2013;29:205-212. Hedges LV, Olkin I. Statistical methods for meta-analysis. New York: Academic Press, 1985. Stärke C, Kopf S, Gröbel KH, Becker R. The effect of a nonanatomic repair of the meniscal horn attachment on meniscal tension: A biomechanical study. Arthroscopy 2010;26:358-365. Feucht MJ, Grande E, Brunhuber J, Burgkart R, Imhoff AB, Braun S. Biomechanical evaluation of different suture techniques for arthroscopic transtibial pull-out repair of posterior medial meniscus root tears. Am J Sports Med 2013;41:2784-2790. Camarda L, Pitarresi G, Lauria M, Fazzari F, D’Arienzo M. Three single loops enhance the biomechanical behavior of the transtibial pull-out technique for posterior meniscal root repair. Arch Orthop Trauma Surg 2017;137: 1301-1306. Fujii M, Furumatsu T, Xue H, et al. Tensile strength of the pullout repair technique for the medial meniscus posterior root tear: A porcine study. Int Orthop 2017;41: 2113-2118. Krych AJ, Johnson NR, Wu IT, Smith PA, Stuart MJ. A simple cinch is superior to a locking loop for meniscus root repair: A human biomechanical comparison of suture constructs in a transtibial pull-out model. Knee Surg Sports Traumatol Arthrosc 2018;26: 2239-2244. Perez-Blanca A, Prado Nóvoa M, Lombardo Torre M, Espejo-Reina A, Ezquerro Juanco F, Espejo-Baena A. The role of suture cutout in the failure of meniscal root repair during the early post-operative period: A biomechanical study. Int Orthop 2018;42:811-818. Feucht MJ, Grande E, Brunhuber J, et al. Biomechanical evaluation of different suture materials for arthroscopic transtibial pull-out repair of posterior meniscus root tears. Knee Surg Sports Traumatol Arthrosc 2015;23: 132-139. Brucker PU, Favre P, Puskas GJ, von Campe A, Koch PP, Meyer DC. Influence of test temperature on biomechanical properties of all-inside meniscal repair devices and inside-out meniscus suturesdEvaluation of an isolated distraction loading, worst-case scenario. Clin Biomech 2011;26:749-753. Sandmann GH, Adamczyk C, Garcia EG, et al. Biomechanical comparison of menisci from different species and artificial constructs. BMC Musculoskelet Disord 2013;14:324.