An anatomically shaped medial meniscus prosthesis is able to partially restore the contact mechanics of the meniscectomized knee joint – Journal of Experimental Orthopaedics

This study consisted of two experiments in different test setups: In an artificial knee joint, the load sharing capabilities of the second-generation meniscus prosthesis (Fig. 2) were compared to the first generation (Fig. 1). The effects of water uptake of the meniscus prosthesis and environmental temperature were investigated as well. In the second test, axial loads were applied to cadaveric knee joints under physiological knee conditions. Contact mechanics on the medial and lateral tibial plateau were evaluated with pressure-sensitive sensors (Tekscan, South Boston, MA, USA, Fig. 3), in three situations (Fig. 4): With both intact native menisci (Fig. 5), after total medial meniscectomy and with the second-generation artificial medial meniscus prosthesis implanted.

Tekscan 4011 pressure-sensitive sensor

Schematic representation of the different situations tested in the cadaveric knee joints. a Native menisci. b Total medial meniscectomy. c Second-generation medial meniscus prosthesis. The arrows indicate the anterior (A), posterior (P), lateral (L) and medial (M) sides of the tibial plateau

Native medial meniscus, excised from a cadaveric knee joint

Meniscus prostheses

The rigid first generation of the medial meniscus prosthesis consists of a stiff core of Bionate® 75D polycarbonate urethane (PCU, DSM Biomedical, Berkeley, CA, USA), which extends into two meniscal horns of the same material. This reinforcing core is covered by a soft, cartilage-contacting outer layer of Bionate® II 80A PCU (Fig. 1). The horns of the meniscus prosthesis are fixed to the tibial plateau by using two dedicated fixation screws, made of a titanium alloy (Ti-6Al-7Nb).

For the design of the second-generation meniscus prosthesis (Fig. 2), two changes were implemented: Firstly, the stiff reinforcing core was removed, resulting in a single-component, soft and flexible meniscus body made of Bionate® II 80A PCU. The attachment horns of the prosthesis remained the same and are made of the stronger and stiffer Bionate® 75D PCU. Secondly, the titanium fixation screws were replaced by a polyethylene terephthalate (PET) anchoring tape. This tape is currently used clinically for fixation of ACL grafts and is secured on the anterior side of the tibia with polyether ether ketone (PEEK) anchoring screws.

All evaluated meniscus prostheses were manufactured by injection moulding of PCU, followed by EtO (ethylene oxide) sterilization. Since the approximately 1% water uptake is expected to strongly affect the mechanical properties of both materials, the wet-tested prostheses were pre-soaked for at least three weeks prior to testing [7]. Only average-sized (size 3) meniscus prostheses were used during this study, to match the selected cadaveric specimens.

Test setup—Artificial knee joint

As no test setup for assessing the load sharing between the meniscus prosthesis and the direct tibiofemoral contact has been described, an artificial setup was developed (Fig. 6). The medial tibial plateau of an average-sized knee joint was segmented from a CT-scan and machined from ultra-high molecular weight polyethylene (UHMWPE). The bottom side of the plateau was hollow, to reproduce the deformation that occurs in the tibial and femoral cartilage of the human knee joint [8]. The different meniscus prostheses could be placed on the tibial plateau by use of two pre-drilled fixation holes, suitable for both the screw and tape fixation techniques. The femoral part of the test setup consisted of an average-sized femur component of a total knee replacement, made of a cobalt chrome alloy (CoCr). Axial loads could be applied to the femur condyle by a hydraulic testing rig (MTS Systems, Eden Prairie, MN, USA). The test setup (Fig. 7) was placed inside a temperature-controlled water bath. The measurements were performed both dry at room temperature and in a water bath of 37 ± 2 °C.

Second-generation meniscus prosthesis in the artificial knee joint

First-generation meniscus prosthesis and sensor in the artificial knee joint

Test setup—Cadaveric knee joints

The femoral and tibial shafts of all specimens were shortened to approximately 12 and 10 cm respectively, to avoid any unnecessary effects of bending. Soft tissue was removed from the ends of the bones to enable potting of the bones with polymethyl methacrylate (PMMA) bone cement in stainless steel cups, which could be mounted in the test setup (Fig. 8). The setup was positioned in a water bath with a temperature of 37° ± 2 °C, to try to mimic the physiological conditions inside the knee joint [9]. It was decided to perform all tests in full extension, for better reproducibility of the alignment of the joints during testing and to maintain knee stability. The tibial cup was mounted in a dedicated support, using a steel rod oriented in the anteroposterior direction to make varus/valgus rotation of the knee joint possible. The axis of rotation was located approximately 5 mm medially of the centre, to allow for the unequal physiological load distribution between the medial and lateral compartment [9, 20]. Anteroposterior and mediolateral translations were not possible in this setup. Axial loads could be applied to the femoral bone cup by the MTS hydraulic testing rig.

Schematic representation of the cadaveric knee test setup: 1 Axial load applied by the testing rig. 2 Femoral bone cup, restricted in all directions except for proximodistal translation. 3 Tibial bone cup and 4 Steel rod of ø8mm, allowing small anteroposterior translations and 14° of varus and valgus rotation. 5 Support for the steel rod, restricted in all directions

Specimens

After obtaining ethical approval from the Anatomy Department, fresh frozen human cadaveric knee joints were obtained from the Radboud University Medical Centre (Nijmegen, The Netherlands). Anteroposterior and mediolateral radiographs were used to select one right and five left average-sized knee joints, to match the size of the pressure-sensitive sensors and the available meniscus prostheses. Specimens were excluded if they showed significant signs of osteoarthritis, such as osteophytes and joint space narrowing. The absence of osteoarthritis was confirmed during the surgical procedure.

Cadaveric surgery

All surgical activities were performed by an experienced orthopaedic knee surgeon (TvT). Firstly, all excessive skin, fat and muscle tissue was removed from the cadaveric knee joints. To improve visibility and to prevent folding of the pressure-sensitive sensor during insertion, the patella and the anterior and posterior parts of the joint capsule were removed. Care was taken to leave the collateral and cruciate ligaments intact, to maintain stability and physiological alignment of the knee joint. The meniscal horn attachments and the meniscocapsular and meniscotibial ligaments were left intact, while the periphery of the native medial and lateral menisci was detached from the tibial plateau. In this way, the pressure-sensitive sensor could be inserted underneath the menisci without folding. Detachment of the circumferential fixation is not expected to affect kinematics of the native meniscus [28], which is therefore expected to maintain its functional performance.

After the native condition was tested, the medial meniscus was completely removed by cutting both horn attachments and the attachment to the remainder of the joint capsule. After insertion of the pressure-sensitive sensor, the meniscectomy measurements were performed.

Subsequently, the second-generation medial meniscus prosthesis was implanted. Firstly, both anchoring tapes were led through the fixation holes of the prosthesis horns. A bone tunnel was drilled from the posterior horn attachment of the native meniscus to the anteromedial aspect of the tibia. By pulling the anchoring tape, the prosthesis was inserted into the knee joint, and the posterior horn was positioned above the tunnel. The optimal location of the anterior drill hole was determined with the knee in extension, thus preventing impingement or extension deficit. After drilling the anterior bone tunnel to the anterolateral aspect of the tibia, both anchoring tapes were tensioned and fixed in the tunnel by a PEEK anchoring screw. The pressure-sensitive sensor was positioned under the meniscus prosthesis before performing the final measurements, as shown in Fig. 9.

Second-generation meniscus prosthesis and sensor in a cadaveric knee joint

Loading protocol – Cadaveric knee joints

Before each series of measurements in the cadaveric knee joints, three pre-conditioning cycles of 1000 N were applied to allow the knee joint to find its natural alignment. Subsequently, the measurements were performed at the end of 120 s of constant axial loading of 500 and 1000 N respectively. Although 1000 N is below physiological loads, the load was not increased any further, to prevent breakage of the cadaveric bones, which has occurred in pilot experiments. In between load cycles, the joint was always unloaded for at least 30 s.

Loading protocol – Artificial knee joint

Approximately 70% of the physiological load is transferred through the medial joint compartment [20, 32]. Therefore, axial loads of 350 and 700 N were applied in the artificial knee joint, which only consist of a medial tibial plateau, to allow comparison with the 500 and 1000 N applied in the cadaveric tests.

Pressure measurements

For this study, piezoelectric pressure mapping sensors of type 4011 (Fig. 3) were used for the measurements in both the artificial and the cadaveric knee joints. Prior to use, the sensors were pre-conditioned and calibrated by applying five different pressures between 0 and 7 MPa. In between the measurements of the different meniscal conditions, additional calibration measurements were performed to evaluate and correct for any possible loss of sensor sensitivity.

From the obtained pressure maps, the peak pressure, mean pressure and contact area on the medial tibial plateau were determined. Furthermore, the load sharing ratio between the medial meniscus or medial meniscus prosthesis and the direct tibiofemoral cartilage contact, i.e. the percentage of the medial compartment load that is transferred through the meniscus, was estimated from the pressure maps. Finally, the pressures on the lateral tibial plateau of the cadaveric knee joints were determined, to assess a potential shift in the load distribution from one joint compartment to another.

Statistical analysis

Based on a previous study [27] and preliminary results from pilot testing, a sample size calculation was performed to determine the number of cadaveric specimens required. Using a power of 80% and a significance level of 0.05, six knees are required to detect contact pressure differences of 1.0 MPa, assuming a standard deviation of 0.5 MPa.

Linear mixed models were used to study the effect of meniscal condition (i.e. native meniscus, meniscectomy and prosthesis) of the medial knee compartment on contact mechanics. Medial peak pressure, mean pressure and contact area were analysed separately. In addition, the percentage of the load transferred through the meniscus or prosthesis and directly through the cartilage was analysed. Finally, the lateral mean pressure was analysed. The models included specimen as a random factor and all other variables that applied to the outcome measure under analysis (i.e. meniscal condition and axial load) as fixed factors. A random intercept was included to account for the specimen’s individual response to each experimental condition. Interaction terms between the fixed factors were also evaluated. Pairwise comparisons between the different levels of the fixed variables were performed by Tukey’s tests that were Bonferroni-corrected to account for multiple comparisons. 95% confidence intervals were determined and P-values below 0.05 were considered statistically significant. Statistical analyses were performed using R (version 4.1.2; R Foundation for Statistical Computing, Vienna, Austria).