Biomechanical analysis of medial tibial cartilage by 7T magnetic resonance imaging and digital volume correlation: a preliminary study of variations caused meniscus ramp lesions

Abstract

Background

See More

The objective study was to compare medial tibial stress variations in three different situations: (1) on native knees, (2) after performing medial meniscus ramp lesions then (3) after their arthroscopic treatments with all-inside suture or total menisectomy.

Methods

We performed a biomechanical cadaveric study on specimens with a normal knee joint line alignment and without meniscal and/or cartilaginous injuries. The comparison of medial tibial stress variations was measured by a digital volume correlation (DVC) method using 7 Tesla MRI images at different loadings as close as possible to physiological conditions (1500 N).

Results

We observed that medial tibial cartilage deformation and mechanical response, and thus the thickness of the cartilage, are restored in the case of all-inside arthroscopic suture whereas they increase in the case of menisectomy.

Conclusion

Medial menisectomy increases medial tibial cartilage displacement fields in physiological compression, whereas all-inside arthroscopic suture allows to get closer to the initial conditions before medial meniscus ramp lesion. This biomechanical study suggests that ramp lesions are potentially at risk of osteoarthritic evolution and that its repair would make it possible to reduce the risks associated with increased stress on the medial tibial cartilage.

Background

In 1988, Strobel described an atypical subtype of meniscal injury associated with anterior cruciate ligament (ACL) rupture involving the peripheral attachment of the posterior horn of the medial meniscus [1]. Anatomically, this lesion corresponds to a meniscocapsular junction and/or a meniscotibial ligament rupture which could be responsible for an increase knee anterior tibial translation, an internal and external rotation, and a pivot shift in case of ACL-deficient knees [2,3,4,5,6,7]. The description and epidemiology of these lesions were performed by Thaunat et al. [8]. Type 4 ramp lesions are common and correspond to a complete tear with a menisco capsular ligament and a meniscotibial tear. While some authors suggest that a ramp lesion constitutes a stable tear pattern located in a highly vascular zone with a favorable biological environment to heal, especially in the setting of acute ACLR rupture [9], the available literature discussing the biomechanical consequences on cartilage of ramp lesions remains limited. It is not clear whether these lesions affect joint kinematics and loading in the medial compartment [10, 11]. A recent review of MRI studies of knee joint under mechanical loading showed that a majority of studies compared only healthy knees with osteoarthritic (OA) knees under load [12]. Unfortunately, too few studies on femorotibial stresses after meniscal injury have been conducted: they were either numerical by finite element analysis or experimental on animal models. Experimental approach on human specimen represents a challenge to be able to measure mechanical fields in the soft structures during a loading. Digital Volume Correlation (DVC) seems to be the well adapted 3D mechanical field measurement method. Initially developed to characterize mechanical response in bone tissues [13], it is currently used for various studies [14, 15]. To analyze human soft tissues, few studies have been performed, DVC has already been coupled with MRI to measure displacement components in intervertebral discs [16].

The objective of the present cadaveric biomechanical study was to compare medial tibial stress variations by DVC method using 7 Tesla MRI images at different loadings in three different cases: (1) on native knees, (2) after performing medial meniscus ramp lesions then 3) after their arthroscopic treatments with all-inside suture or total menisectomy. The hypothesis of this study is that subtype 4 ramp lesion increase displacement fields of the medial tibial cartilage and that arthroscopic suture could restore a native medial compartment knee stress.

Methods

Specimens

This is a biomechanical cadaveric study. Two cadaveric knees specimens with a normal knee joint line alignment (HKA angle of 178.9° and 177.8° measured on whole segment CT radiographic acquisitions) and without meniscal and/or cartilaginous injuries on 7 Tesla MRI (7 T MRI, MAGNETOM Terra.X, Siemens Healthineers©)) examination were collected after obtaining the approval of the Ministry of Education and Research (No. DC-2008-137): a 63-year-old man (73 kg, BMI: 26.8) and an 81-year-old woman (79 kg, BMI: 23.6) with no osteoarticular history. The cartilage status was first assessed using high-resolution 7T MRI. Its integrity was then definitively confirmed under arthroscopic control by palpation of all femorotibial compartments prior to the creation of the RAMP lesion.

The knees were then dissected without opening the joint and the capsule and peripheral ligaments were preserved. Polyurethane rigid fixations have been molded on each bone end to facilitate attachment to the MRI-compatible loading bench (Fig. 1).

Fig. 1

Cadaveric knee with its proximal and distal fixations during arthroscopy in the anatomy lab (A). RMI-compatible loading bench allowing progressive axial compression ranging from 0 N to 1500 N (B, C)

Experimental procedure

An initial series of 7 T MRI imaging was performed on the native knees at progressive loads ranging from 0 N to 1500 N, equivalent to more than twice the body weight load. A second series of images was taken on these same knees with the same loads after grade 4 medial meniscus ramp lesions [8] had been made under arthroscopy using a posteromedial instrumental approach. The last image sequences were then taken after all-inside anterior arthroscopic suture on the left knee and after total menisectomy on the right one (Figs. 2, 3). A specific loading appartus was designed and developed to perform these experimentations in the 7 T MRI device (Fig. 1B, C). It was exclusively made with polymer materials (non-ferromagnetic materials) and able to impose a compression loading up to 2000 N. The imposed load ability was preliminary calibrated and controlled by a specific homemade hydraulic sensor during experiments. The loading apparatus was tested and validated on a synthetic specimen before to be used on knees.

Fig. 2

Arthroscopic posterior view of intact knee 1 and its meniscocapsular junction and meniscotibial ligament (A). Subtype 4 medial meniscus ramp lesion created by a posteromedial approach (B). Ramp lesion sutures through the standard anterior approach using an all-inside meniscal repair device (C) [35]

Fig. 3

Arthroscopic posteromedial view of intact knee 2 (A). Subtype 4 medial meniscus ramp lesion created by a posteromedial approach (B). Anterior arthroscopic view of a medial menisectomy performed with the shaver (C, D)

All the steps of experiments on a specimen were performed successively in 24 h to limit any degradation. Specimens were maintained under hydration, except during MRI acquisitions (10 min for on volume acquisition), to preserve a native mechanical behavior.

Digital volume correlation (DVC)

Digital volume correlation (DVC) is a measurement field method used to determine the three components of displacement and spatial variations of a material or structure from volume images [13,14,15,16,17,18]. In the initial image, a sub-volume of voxels (D) is defined at each voxel. Each sub-volume is then searched by measuring the degree of similarity in the initial image. For this purpose, at the reference state, a correlation sub-volume is represented by the value of the voxels denoted $\:f\left(\overrightarrow{X}\right)$, with $\:\overrightarrow{X}\:$the initial position vector. The position of the searched sub-volume in the deformed state is denoted $\:\overrightarrow{x}$ and the grey levels $\:g\left(\overrightarrow{x}\right).$ The level of similarity of a sub-volume between of the initial and the deformed states is evaluated by a correlation coefficient [19] based on the optimization of a functional equation $\:f\left(\overrightarrow{X}\right)-g\left(\overrightarrow{\varphi\:}\right(\overrightarrow{X})$), where $\:\overrightarrow{\varphi\:}\:$ is the material transformation between the deformation states. This non-contact method allows the measurement of volume displacements in the structure from 1 μm to several tens of millimeters [20, 21].

Overlay of the constrained and unconstrained MRI images was performed by semiautomated tibial registration of the image sequences with 3D Slicer software (Version 4.11, Kitware, France). Manual segmentation of the medial tibial cartilage was performed on the first MRI to extract the area associated to cartilage. The DVC procedure was restricted to this cartilage area in order to reduce computation times and resources.

The displacement fields were analyzed in the three dimensions, and observation of the displacements in “ $\:\overrightarrow{Z}$” made it possible to analyze the sinking of cartilage after axial compression, with the knee positioned in full extension.

Results

Figures 4 and 5 show the displacement fields of the medial tibial cartilage along the Z axis with an axial progressive compression load of 0 N, 750 N and 1500 N. We observe that these displacement fields, and thus the thickness of the cartilage, are restored in the case of all-inside arthroscopic suture (Fig. 4), whereas they increase in the case of menisectomy (Fig. 5). Displacement fields are measured in each voxel of the cartilage structures which provide the kinematics of the structures under compressive loading. Measurement displacement values according to z are negative corresponding the considered load.

Fig. 4

Medial tibial cartilage displacement fields on a right cadaveric knee with an axial compression load of 0 N, 750 N and 1500 N. The first tests were performed on healthy meniscus (A), then after subtype 4 medial meniscus ramp lesion (B) and finally after all-inside arthroscopic suture (C)

Fig. 5

Medial tibial cartilage displacement fields on a left cadaveric knee with an axial compression load of 0 N, 750 N and 1500 N. The first tests were performed on healthy meniscus (A), then after subtype 4 medial meniscus ramp lesion (B) and finally after arthroscopic menisectomy (C)

Quantitatively, mean displacements, initially expressed in voxels (1 voxel = 0.35 mm, ± standard deviation), were measured on a superficial medial anteroposterior strip of the medial tibial cartilage and are presented in the Table 1. Displacement observations can be resume as following:

Table 1 Medial tibial cartilage sinking measurement (Z axis) between native knees, injured knees and after arthroscopic treatment obtained by digital volume correlation (DVC)

• 1. Normal loaded knees at 1500 N load: mean displacements were − 1.316 mm (± 0.132 mm) on the first specimen and − 1.419 mm (± 0.248 mm) on the second one.

• 2. Load of 1500 N on the knees after creation of a RAMP tear: mean displacements were − 2.404 mm (± 0.288 mm) on specimen 1 and − 2.404 mm (± 0 .288 mm) on specimen 2.

• 3. Load of 1500 N on the knee after a RAMP repair: mean displacement was − 1.172 mm (± 0.281 mm) on specimen 1.

• 4. Load of 1500 N on the knee after total internal meniscectomy: mean displacement was − 2.092 mm (± 0.792 mm) on specimen 2.

Discussion

The most important finding of this study is that medial menisectomy increase medial tibial cartilage displacement fields in compression, whereas all-inside arthroscopic suture allows to get closer to the initial conditions before medial meniscus ramp lesion for specimens with a normal knee joint line alignment. This study is the first quantitative work of mechanical fields on the knee cartilaginous under loading. Ramp lesions increases stress in cartilaginous and its surgical repair provide a behavior close to the physiological one. However, ramp lesion seems to affect less displacement fields than a meniscectomy and would a less arthrogenic potential.

Concerning the ramp lesion, these increases in displacement fields are probably due to the fact that untreated tears cause meniscal extrusion that slightly increased cartilage–cartilage contact and almost eliminated meniscus–cartilage contact [22, 23].

However, some authors continue to say that it is not clear whether these lesions affect joint kinematics and loading in the medial compartment of the knee similar [11].

With a freedom robotic testing sytem, Naendrup et al. [24] found that there was no influence of a 25 mm ramp lesion on knee kinematics, in situ forces in the native ACL, bony contact forces in the medial compartment, and bony contact forces in the lateral compartment. But unlike our loading force simulating bipodal support (1500 N), their loading conditions were a 90 N anterior load, a 5 N.m of external-rotation torque, a 134 N anterior load plus a 200 N compression load, 4 N.m of external-rotation torque plus a 200 N compression load, and 4 N.m of internal-rotation torque plus a 200 N compression load. Although our study did not allow for the analysis of stress variations in flexion and rotation, it did allow for the first analysis of the displacement fields of the medial cartilage under physiological conditions.

Regarding the treatment, some authors suggest that nonsurgical management of ramp lesions may be reasonable [25,26,27] because their repair failed to restore internal rotation and external rotation at higher knee flexion angles [3]. However, our results show that the stresses on the medial cartilage in case of medial meniscus ramp lesion are close to the results after menisectomy. As an example, Song et al. [28] studied sheep knee cartilage deformation for joints undergoing partial meniscus removal using T1-weighted gradient echo imaging. Following meniscectomy, they reported a significant decrease in the contact area and a significant increase in maximum cartilage deformation. They observed that meniscectomy resulted in a 60% decrease in the contact area (P = 0.001) and a 13% increase in maximum cartilage deformation (P = 0.01), and therefore at the origin of an osteoarthritic degeneration. These observations are in corroboration with our results on human specimens with a high deformation under loading in case of meniscectomy and a significant diminution after RAMP repair. In another biomechanical study, Nicolas et al. [29] confirmed that meniscus-capsular junction’s mechanical properties are similar to other knee ligaments and may play a role in knee stability. The findings provide insights into the the behavior of the meniscus-capsular junction could have clinical implications for diagnosing and surgical treatment of meniscocapsular lesions. Secondly, a cadaveric porcine study confirmed that untreated ramp lesions showed histological deterioration [30].

About the other studies, only magnetic resonance imaging (MRI)-based knee investigations are usually performed on joints at rest or in a non-weight-bearing condition that does not mimic the actual physiological condition of the joint and does not allow to evaluate the cartilage thickness variations under load [12, 31]. Contrary to our analysis by DVC, MRI measures in knee loading are categorized as tissue deformation and morphological changes directly from magnitude images, or tissue deformation from MRI phase information, and/or quantitative MRI [12, 32]. Unlike these MRI techniques, the accuracy of the DVC measurement was verified and validated before experimentation.

The main limitation of this study was its cadaveric character on only two specimens. These cadaveric knees initially from fresh and non-formalin fixed specimens were cryopreserved after the primary dissection phase. Before each experiment, the thawing protocol consisted in placing the cadaveric segments at room temperature for 48 h in order to optimize the elasticity/stiffness relationship and to get closer to the physiological conditions found in living patients. This work is a preliminary study on two specimens allowing the validation of the proof of the concept. The preparation and the availability of 7 T MRI device coupled with specific loading setup were time consuming which leaded difficult to perform experiments on numerous specimens. In future works, it would be interesting to envisage experiments on a larger group with a higher age variation to study mechanical effects of other parameters and other injury cases.

It is also important to emphasize that loading was only possible in extension, since no compatible MRI coil currently exists that allows maintaining a knee in flexion, with or without loading, during image acquisition. This represents another important limitation of our study.

Secondly, we have chosen an all-inside technique for the meniscal suture. Some authors have demonstrated that both all-inside and inside-out repair approaches to medial meniscal lesions can provide similar results [33, 34]. But no experimental study has yet demonstrated the superiority of one technique over another.

Conclusion

RAMP lesion increases medial tibial cartilage displacement fields. However, medial menisectomy would increase them more in physiological compression, whereas all-inside arthroscopic suture allows to get closer to the initial conditions before medial meniscus ramp lesion. This biomechanical study suggests that ramp lesions are potentially at risk of osteoarthritic evolution and that its repair would make it possible to reduce the risks associated with increased stress on the medial tibial cartilage.