Abstract
Background:
Meniscal deficiency has been reported to increase contact pressures in the affected tibiofemoral joint, possibly leading to degenerative changes. Current surgical options include meniscal allograft transplantation and insertion of segmental meniscal scaffolds. Little is known about segmental meniscal allograft transplantation.
Purpose:
To evaluate the effectiveness of segmental medial meniscal allograft transplantation in the setting of partial medial meniscectomy in restoring native knee loading characteristics.
Study Design:
Controlled laboratory study.
Methods:
Ten fresh-frozen human cadaveric knees underwent central midbody medial meniscectomy and subsequent segmental medial meniscal allograft transplantation. Knees were loaded in a dynamic tensile testing machine to 1000 N for 20 seconds at 0°, 30°, 60°, and 90° of flexion. Four conditions were tested: (1) intact medial meniscus, (2) deficient medial meniscus, (3) segmental medial meniscal transplant fixed with 7 meniscocapsular sutures, and (4) segmental medial meniscal transplant fixed with 7 meniscocapsular sutures and 1 suture fixed through 2 bone tunnels. Submeniscal medial and lateral pressure-mapping sensors assessed mean contact pressure, peak contact pressure, mean contact area, and pressure mapping. Two-factor random-intercepts linear mixed effects models compared pressure and contact area measurements among experimental conditions.
Results:
The meniscal-deficient state demonstrated a significantly higher mean contact pressure than all other testing conditions (mean difference, ≥0.35 MPa; P < .001 for all comparisons) and a significantly smaller total contact area as compared with all other testing conditions (mean difference, ≤140 mm2; P < .001 for all comparisons). There were no significant differences in mean contact pressure or total contact area among the intact, transplant, or transplant-with-tunnel groups or in any outcome measure across all comparisons in the lateral compartment. No significant differences existed in center of pressure and relative pressure distribution across testing conditions.
Conclusion:
Segmental medial meniscal allograft transplantation restored the medial compartment mean contact pressure and mean contact area to values measured in the intact medial compartment.
Clinical Relevance:
Segmental medial meniscal transplantation may provide an alternative to full meniscal transplantation by addressing only the deficient portion of the meniscus with transplanted tissue. Additional work is required to validate long-term fixation strength and biologic integration.
Intact menisci are essential for normal knee kinematics and load-bearing characteristics.1,7,18,23,25,28,29,39 Meniscal tears and subsequent meniscal tissue excision have been shown to dramatically increase articular cartilage contact stresses in the tibiofemoral joint leading to degenerative changes.13,25,29 Numerous studies have documented the poor clinical outcomes that occur subsequent to partial and complete meniscectomy.7,11,13,20,23,34
While partial meniscectomy in the setting of an extensive or complex meniscal tear has been shown to provide short-term symptomatic relief, the long-term effect on the articular cartilage in the affected tibiofemoral joint remains a major concern.7,11,13,20,23,34 To prevent knee arthritis after partial or complete meniscal excision, attention has been turned to restoring normal knee kinematics and biomechanical forces through meniscal repair and transplantation. § Although the biomechanical characteristics of knees with meniscal allograft transplantation have been reported to resemble native knee loading characteristics, in some patients, significant amounts of native meniscal tissue must be excised to perform a complete meniscal transplantation.19,24,25,30,35,38,45 In certain patients, only the midbody of the meniscus is compromised or irreparable, while the anterior and posterior horns remain largely intact. In this circumstance, standard meniscal allograft transplantation requires removal of a considerable amount of healthy native meniscal tissue. Implantation of a segmental meniscal allograft in this setting has the potential to more closely replicate native knee biomechanics and biology while preserving as much native tissue as possible.19,31
Nyland et al 31 first evaluated partial meniscal allograft transplantation in a porcine knee cadaveric model as an alternative to complete meniscal allograft transplantation and meniscal scaffolds. They found that segmental meniscal allograft transplantation in porcine knees restored native knee contact pressures in the tibiofemoral joint. However, to the best of our knowledge, this technique had not yet been tested in human cadaveric knees.
The purpose of this study was to evaluate the biomechanical effects of partial meniscal transplantation on restoring normal knee loading characteristics in a cadaver model. It was hypothesized that segmental midbody meniscal transplantation using modern meniscal repair techniques would restore normal meniscal load-bearing characteristics in a human cadaveric model.
Methods
Specimen Preparation
Ten nonpaired male fresh-frozen cadaveric human knees with an average age of 46.9 years (range, 38-55) were procured. All test specimens were prescreened with diagnostic arthroscopy. Specimens were excluded if there was chondromalacia greater than grade II in the medial or lateral compartment, meniscal deficiency, evidence of previous meniscal surgery, ligamentous laxity, or gross malalignment or deformity. The specimens used in this study were donated to a tissue bank for medical research purposes and then purchased by our institution. Institutional review board approval was not required, as the use of cadaveric specimens is exempt at our institution.
All specimens meeting criteria were dissected, with all skin and subcutaneous tissue removed, including all posterior musculature other than the popliteal muscle belly. The extensor mechanism was also removed for access to the knee joint. All ligamentous attachments of the knee joint were identified and kept intact. The femur, tibia, and fibula were cut 14 cm proximal and distal to the knee joint line, respectively. The distal tibia and fibula were then fixed in a neutral anatomic position and potted in polymethyl methacrylate (Fricke Dental International).
Surgical Technique
A medial femoral condyle osteotomy was performed using a sagittal saw on each specimen to facilitate access to the medial compartment per a previously established protocol (Figure 1). 28 A midbody (middle 33%) partial medial meniscectomy was performed on each specimen using a No. 11 blade. The size of the meniscectomy was determined by measuring the circumference of the meniscus, dividing this by 3, and excising the central third portion. A size-matched meniscal allograft (donated by JRF Ortho) was used to fashion a segmental meniscal graft matching the meniscal defect. The segmental meniscal allograft was secured in the anatomic position with 7 meniscal suture tapes (Arthrex Inc) using horizontal and vertical mattresses via an inside-out technique. For the testing state including bone tunnel fixation, an additional central horizontal mattress suture tape was added, with each limb passed through 1 of 2 parallel bone tunnels (2.8-mm diameter), which were drilled from the anterior medial tibial metaphysis proximally to the central portion of the meniscal defect. The suture tape was then tied over a 4-hole 10-mm titanium cortical button (Smith & Nephew) against the anterior tibia (Figure 2). Two posterior submeniscal arthrotomies measuring 20 to 25 mm were made, 1 medial and 1 lateral, to facilitate passage of the pressure-mapping sensors (model 4000; Tekscan Inc). Care was taken to avoid cutting the popliteal, collateral, and cruciate ligaments and to ensure flat positioning of the pressure-mapping sensors against the tibial articular cartilage. The medial femoral condyle osteotomy was then secured in the anatomic position using a medial-to-lateral bolt and washer, a custom-contoured stainless steel plate, and 4 screws (Figure 3).
Right knee medial femoral condyle osteotomy. A medial femoral condyle osteotomy was performed on each specimen to facilitate access to the medial compartment per a previously established protocol. 28 A segmental meniscal allograft has been transplanted to the medial meniscal midbody.
Left knee medial meniscal allograft transplantation. A size-matched medial meniscal allograft was used to fashion a segmental meniscal graft matching the meniscal defect. The segmental meniscal allograft was secured in the anatomic position with 7 meniscal suture tapes using horizontal and vertical mattresses via an inside-out technique. This image includes an additional central horizontal mattress suture tape with each limb passed through 1 of 2 parallel bone tunnels and tied over a titanium suture button against the anterior tibia (button not visualized).
(A) Left cadaveric knee held at full extension and loaded in a dynamic tensile testing machine and (B) left knee fixed at 30° of flexion in a computer model of the apparatus. Each specimen’s medial femoral osteotomy was fixed in the anatomic position with a medial-to-lateral bolt and washer, a custom-contoured stainless steel plate, and 4 screws. A transepicondylar “loading rod” (10-mm diameter) was placed medial to lateral and acted as the load-bearing axis during testing. An additional “flexion rod” (8-mm diameter) was passed medial to lateral through the proximal femur and allowed for changes to the knee flexion angle from 0° to 90° of flexion in 30° increments. The potted distal tibia was rigidly fixed to a custom pivoting base that allowed for freedom of motion in the transverse plane and for adjustment of the tibial orientation to standardize varus and valgus positioning.
Mechanical Testing Protocol
All specimens were biomechanically evaluated using a dynamic tensile testing machine (ElectroPuls E10000; Instron). The machine was digitally recalibrated before each specimen was mounted. The device is annually serviced by the manufacturer and certificated to be within 0.1% of indicated force values and 0.1% of indicated position values. The potted portions of the distal tibia and fibula were oriented vertically and rigidly secured into a custom pivoting base that allowed for freedom of motion in the transverse plane on the base of the testing bed; this base also allowed for adjustment of the tibial orientation to standardize varus and valgus positions of the knee before testing. The potted portion of the femur was rigidly fixed to a custom fixture mounted on the end of the actuator using a transepicondylar rod (10-mm diameter), which acted as the load-bearing axis throughout testing. A subsequent 8-mm rod was passed through the proximal femur and secured at various positions on the custom fixture to facilitate adjustment in knee flexion angles of 0°, 30°, 60°, and 90° (Figure 3).
Before each specimen was tested, pressure-mapping sensors (sensitivity, 0.04 MPa; model 4000, K-Scan [Tekscan Inc]) were inserted as previously described on the medial and lateral sides of the knee. The medial and lateral sensor areas were each 33 mm in the anterior-posterior direction and 27.9 mm in the medial-lateral direction and were separated into grids of 26 × 22 individual sensing units (“sensels”), which resulted in a spatial resolution of 62 sensels/cm2. Each sensor was calibrated before testing using the sensor program’s 2-point calibration curve of 400 and 1000 N of applied force from a rubber end effector attached to the actuator of the dynamic tensile testing machine; this ensured accurate measurements within the range of forces being exerted on the knee. The knee joint was fixed to 45° of flexion and loaded to 500 N to dynamically visualize the placement of the submeniscal pressure sensors in real time, which were adjusted to visually capture the largest portion of each meniscus. The anterior and posterior ends of the sensors were then sutured and tethered to screws anchored on the anterior and posterior portions of the tibia to ensure consistent placement of the sensors during all trials and across all flexion angles (Figure 3). While the joint still bore 500 N of compression, the varus and valgus alignment of the tibia was adjusted such that the center of pressure was centered between the medial and lateral menisci as observed by the pressure-mapping sensors. Further observation of the center of pressure under load at the extreme ranges of 0° and 90° of flexion was used to adjust the center of pressure more medially or laterally to keep it generally centered throughout the entire range of flexion. The varus and valgus alignment was locked in this position for all states and flexion angles for each specimen.
This study included the following states, all tested at 0°, 30°, 60°, and 90° of flexion: intact medial meniscus, segmental medial meniscal loss (central 33% of meniscus excised), segmental medial meniscal transplant with all suture fixation, and segmental medial meniscal transplant with suture and bone tunnel fixation. One trial was collected for each flexion angle in every state. The order of flexion angles tested was randomized for each testing state. Each specimen underwent testing with an intact meniscus as the first state and segmental loss as the final state, with the 2 transplant states tested in a random order as determined a priori to equally distribute the number of knees in each testing sequence.
During each trial, the dynamic tensile testing machine began in 20 N of compression, increased to a load of 1000 N over 10 seconds, held the applied load for 20 seconds, and decreased to 20 N over 10 seconds. Force and position data were measured from the load cell and actuator at a rate of 1000 Hz, and the pressure on each meniscus was recorded throughout each trial via the pressure-mapping sensors at a rate of 10 Hz. All data were processed using a custom-written script (MATLAB R2019b; MathWorks, Inc) to determine the peak contact pressure, mean contact pressure, and total contact area for the medial and lateral menisci at the instant of maximum loading as observed from the pressure-mapping sensors.
Statistical Methods
To match the repeated measures experimental design, 2-factor random-intercepts linear mixed effects models were used to compare pressure and contact area measurements among the 4 experimental meniscal conditions. Data from all tested flexion angles were included in each linear mixed effects model, and the likelihood ratio test was used to assess the interaction effect between flexion angle and state. When the likelihood ratio test was not statistically significant, a reduced main effects model was reported. The Tukey method was used to make all pairwise comparisons among meniscal states, while holding the flexion angle constant. Residual diagnostics were inspected to ensure model fit and that assumptions were met. An alpha level of .05 was set to interpret statistical significance. Statistical power for this experimental design and the fixed feasible sample size were considered a priori. Based on an assumed alpha level of .05, 2-tailed testing, and parametric dependent-group comparisons of means, 10 specimens are sufficient to detect an effect size of d = 1.0 with 80% statistical power. Statistical software R (Version 4.0.0; R Core Team) was used for all plots and analyses with the packages nlme, emmeans, and ggplot2.
Results
When the medial meniscus alone (where transplant occurred) was examined, the meniscal loss state demonstrated significantly higher mean contact pressure than the other states: intact, transplant, and transplant with bone tunnel (P < .001 for all comparisons). The meniscal loss state also had a significantly smaller total contact area as compared with the other states (P < .001 for all comparisons). There was no significant difference in peak contact pressure in the meniscal loss state as compared with intact (P = .586), transplant (P = .716), and transplant with tunnel (P = .916). There were no significant differences in mean contact pressure, peak contact pressure, or total contact area among the intact, transplant, or transplant-with-tunnel groups. When the lateral meniscus alone was examined, there were no significant differences in any outcome measure across all comparison states (Table 1).
Results From 6 Two-Factor Linear Mixed-Effects Models a
White cells contain the Tukey-adjusted P values for the post hoc comparison between the states indicated, assuming that the flexion angle is held constant. Shaded cells contain the pairwise group effect (column minus row) in the units of measurement and 95% CIs in parentheses.
The locations of the center of pressure for each state and flexion angle were analyzed with respect to medial-lateral and anterior-posterior positioning. State was not a significant independent contributor to the medial-lateral center of pressure (P = .756) or anterior-posterior center of pressure (P = .780). Flexion angle was a significant independent contributor to the anterior-posterior position of the center of pressure (P < .001), resulting in a more anterior position as flexion angle increased; all flexion angle comparisons were significantly different (all P < .01). Flexion angle was a significant independent contributor to the medial-lateral position of the center of pressure (P = .012), resulting in a more lateral position in the 30° flexion angle as compared with 60° (P = .045) and 90° (P = .006); all other comparisons were insignificant (Table 2).
Contact Pressure, Peak Contact Pressure, and Total Contact Area Stratified by Flexion Angle and Meniscal State a
Values are presented as mean ± SD.
Relative pressure distribution on the medial meniscus was analyzed with respect to state and flexion angle. State was not a significant independent contributor to the relative pressure distribution in the knee (P = .743). However, flexion angle was a significant independent contributor to the relative pressure distribution (P = .012), resulting in a lower percentage of total pressure on the medial meniscus at 30° of flexion as compared with 60° (P = .041) and 90° (P = .007); all other comparisons were insignificant.
Discussion
The most important finding of this study is that segmental medial meniscal allograft transplantation with and without bone tunnel fixation restored medial compartment mean contact pressures and mean contact area to values measured in the intact medial compartment. Segmental medial meniscal allograft transplantation demonstrated no significant difference in loading characteristics in the lateral compartment when compared with an intact medial compartment.
As our understanding of normal meniscal function has increased, so too have our efforts to preserve the meniscus. Similar to other biomechanical studies, the current study has shown that segmental meniscal loss from a partial medial meniscectomy creates altered tibiofemoral compartment pressures and loading characteristics.1,3,7,11,23,25,26,30 Given these findings, several authors have gone to great lengths to study methods of meniscal preservation, such as meniscal repair,4,13,28,29,38 meniscal allograft transplantation,19,24,25,30,31,35,40,45 and meniscal scaffold transplantation. ‖
Most meniscal treatment algorithms support meniscal repair whenever meniscal tissue is amenable. 21 However, focal meniscal loss is not an uncommon result of failed meniscal repair or in tear patterns when repair is not possible. In these cases, meniscal allograft transplantation is a viable option, although current meniscal allograft transplantation techniques can be challenging, with complete excision of the native meniscus; in some cases, this involves excision of healthy native meniscal tissue, including the native meniscus roots, adjacent to the focal meniscal deficiency.
Several studies have reported that various intra-articular components such as menisci are sensate, capable of generating neurosensory signals that reach spinal, cerebellar, and higher central nervous system levels.2,8,22,32,46 It is believed that these neurosensory signals result in conscious perception and are important for normal knee joint function and maintenance of tissue homeostasis. 10 One concern with performing a total meniscectomy for meniscal allograft transplantation is that it effectively removes all meniscal mechanoreceptors in that compartment, including Ruffini endings, Pacinian corpuscles, and Golgi tendon organs, which aid in proprioception and joint homeostasis.2,8,46
Given these findings, attention has been turned to meniscal preservation through segmental meniscal scaffolds. Meniscal scaffolds generally come in 2 varieties: a collagen-based implant5,14,15,36,37,41,42 and a polyurethane-based scaffold.9,12,17,40,44 The 2 available implants are the Collagen Meniscal Implant (CMI; Ivy Sports Medicine) and Actifit polyurethane scaffold (Orteq Ltd). CMI consists of type 1 collagen from bovine Achilles tendons. The use of CMI is designed as a regeneration template into which the body’s own fibrocartilaginous tissue may grow and be resorbed within 2 years of implantation.27,37 The Actifit scaffold degrades slowly over a 5-year period and consists of porous polycaprolactone and urethane segments; it is thought to be replaced by fibrocartilaginous tissue. 35
Rates of success have been variable with both scaffold implants, with failure rates ranging from 0% to 38% at 4 years. 17 In a recent systematic review, Houck et al 17 reported that the mean treatment failure rate was higher in Actifit polyurethane scaffolds (9.9%) than in CMI (6.7%) at 4 years. Long-term results have been less predictable. The authors noted that meniscal scaffolds may become nonfunctional owing to fragmentation, shrinkage, and extrusion, which may result in failure to increase articular cartilage coverage, reduce peak pressure, and achieve a balanced load distribution. 17
Nyland et al 31 first evaluated segmental meniscal allograft transplantation in a porcine knee cadaveric model, looking for an alternative to meniscal scaffolds. They found that segmental meniscal allograft transplantation in porcine knees restored native knee contact pressures in the tibiofemoral joint. Strauss et al 43 demonstrated early healing of segmental meniscal transplants to adjacent native meniscal tissue in an ovine model. However, segmental meniscal transplantation had not yet been tested in human cadaveric knees.
The present study is the first human biomechanical assessment of a segmental meniscal allograft transplantation. Although further study is needed, the current findings raise the possibility of using a transplanted segmental meniscal allograft to treat patients who have focal meniscal deficiency. This would allow preservation of adjacent, healthy meniscal tissue and maintain its proprioceptive capacity. However, in contrast to complete meniscal allograft transplantation, an added challenge of segmental meniscal grafting is that healing must occur not just at the capsular periphery but also at the anterior and posterior meniscal tissue remnants. Additional in vivo research is needed to determine whether the segmental meniscal allograft heals to the remnant meniscal tissue. The prospect of being able to offer patients another option to restore meniscus-mediated chondroprotection is promising.
Attempting to assess the stability of transplant fixation constructs, we tested an all–soft tissue repair with 2-0 meniscal tapes and a construct with bone tunnel fixation. The bone tunnel fixation transplant state included the addition of 2 parallel bone tunnels from the anterior tibial metaphysis to the central defect. An additional tape suture was placed in a horizontal mattress through the central meniscal allograft and passed through the tunnels. The sutures were tensioned and tied over a suture button against the anterior tibial cortex. There were no significant differences observed in loading characteristics between these 2 transplant constructs. However, this study tested axial loading alone at several flexion angles. Had we tested meniscal extrusion and other more physiologic loading characteristics, we suspect that a significant difference may have been observed, favoring the construct with bone tunnel fixation. The bone tunnel may also prove to be helpful in an in vivo model to assist in healing, similar to how meniscal repairs have been reported to have higher healing rates in the context of bone tunnels for anterior cruciate ligament reconstruction. 6
Interestingly, medial compartment peak pressure showed no significant differences across all tested states. Although this finding suggests that segmental medial allograft transplant had no significant difference as compared with the intact meniscal state, the intact and meniscal-deficient state also showed no significant difference in peak contact pressure. While surprising, 1 possible explanation was saturation of the pressure sensors at peak contact pressure. We favored higher sensor sensitivity, which resulted in a lower maximum observable pressure, or “saturation pressure.” This likely prevented our testing software from capturing the full net difference in peak pressure between states. Other studies have validated that meniscal deficiency or meniscal tears result in increased tibiofemoral peak contact pressures.33,44
Limitations
This cadaveric model was limited to all male knees given the need for larger medial and lateral compartments to fit the pressure sensors under the menisci. Therefore, findings cannot be fully generalized among male and female specimens. Furthermore, medial femoral condyle osteotomies were required to fully access the medial compartment, and although the condyles were fixed stably in the anatomic position, the osteotomy fixation may have loosened with each successive state. To mitigate this, transplant states were randomized and continually observed for movement or loosening. Still, the sequential testing of the same knee may have resulted in a more degraded or lax specimen with each successive testing state. The testing jig, given the degrees of freedom, required manual adjustment and calibration. This may have introduced some unintended variability among specimens. The number of specimens tested in this study afforded 80% power to detect an effect size of Cohen d = 1.0. Thus, between-group differences more subtle than d = 1.0 cannot be ruled out by non–statistically significant comparisons in this study. Last, this was a time-zero study that did not account for biological healing or potential loosening of the suture that may occur in vivo. Additional in vivo studies may be beneficial in understanding the segmental meniscal allograft’s ability to heal and withstand more physiologic stress.
Despite these limitations, this study is a first step in determining if segmental medial meniscal allograft transplantation is a viable alternative to complete meniscal allograft transplantation and placement of meniscal scaffolds.
Conclusion
Segmental medial meniscal allograft transplantation with and without bone tunnel fixation restored the medial compartment mean contact pressure and mean contact area to values measured in the intact medial compartment. Segmental medial meniscal transplantation may provide an alternative to full meniscal transplantation and meniscal scaffold insertion by addressing only the deficient portion of the meniscus with transplanted tissue. However, additional work is required to validate long-term fixation strength and biologic integration.
Footnotes
Submitted August 12, 2020; accepted April 21, 2021.
One or more of the authors has declared the following potential conflict of interest or source of funding: Arthrex provided partial funding via grant RDGT2003 and surgical product donation. JRF Ortho donated meniscal allografts for this study. D.B.H. has received a grant from DJO. J.W.A. has received a grant from DJO and hospitality payments from Smith + Nephew. A.F.V. has received consulting fees from Stryker; research support from Arthrex; and hospitality payments from Arthrex, Vericel, and Steris Corp. M.T.P. has received research support and royalties from Arthrex; royalties from Arthrosurface; honoraria from Flexion Therapeutics; and hospitality payments from Arthrex, Joint Restoration Foundation, and SLACK Inc. AOSSM checks author disclosures against the Open Payments Database (OPD). AOSSM has not conducted an independent investigation on the OPD and disclaims any liability or responsibility relating thereto.
