Posterior Tibial Slope and Medial Tibial Depth Measured by MRI in Patients with Posterior Cruciate Ligament Injury: A Case-Control Study

Objectives To 1) gure out the difference in tibial morphologies between patients with posterior cruciate ligament (PCL) injury and those without, 2) identify any tibial anatomical variations contributing to PCL rupture. Methods 45 patients with PCL rupture from 2015 to 2020 in our hospital, and 45 patients matched by age and sex with intact PCL were enrolled in our study. The intraclass correlation coecient (ICC) was used to assess the reliability of each parameter. The independent t-test was conducted to identify the differences in tibial morphometric characteristics between the PCL-injured and PCL-intact individuals, including the posterior tibial slope (PTS), meniscal slope (MS), medial tibial depth (MTD). A binary logistic regression model was established to evaluate the roles of those anatomic parameters of interest play in PCL injuries. Results The interobserver reliability of each parameter showed excellent agreement. Signicant differences in the medial (P= .023) and lateral (P= .009) PTS were found between the PCL-injured group (3.68 ± 2.70 and 4.55 ± 3.19, respectively) and the controls (5.00 ± 2.73 and 6.39 ± 3.29, respectively). And the MTD was 1.98 ± 0.64 mm in the PCL-injured group and 2.37 ± 0.55 mm in the control group (P=0.007). Binary logistic regression analysis showed that smaller lateral PTS and MTD were directly associated with PCL injury, with an OR of 1.17 and OR of 3.14, respectively. The medial PTS was independent to PCL injures. Conclusion In patients with PCL injury, lateral PTS, medial PTS, and MTD were found smaller compared to PCL intact controls. Decreased lateral PTS and MTD were underlying anatomic risk factors for PCL injury, with each unit reduction, the risks of PCL injury were 1.2 and 3.1-fold higher, respectively.


Introduction
The main function of the posterior cruciate ligament (PCL) is to restrict the excessive posterior movement of the tibia relative to the femur, maintaining the posterior stability of knee joint 1 . PCL injury always combined with other injuries, such as medial collateral ligaments, anterior cruciate ligament (ACL), or chondral injuries, especially in the lateral tibiofemoral compartment 2 , which would cause instability, accelerating degeneration of the knee joint. It was accepted that femorotibial biomechanics were in uenced by the morphology of the tibial plateau and femoral condyles 3 . Biomechanically, excessive posterior tibial slope (PTS) was considered as a predisposing risk factor for ACL noncontact injury, because of its contribution to the increase of tensity for native ACL, particularly in axial compressive forces condition. In contrast, the stress on PCL would increase with smaller PTS.
As an adjunctive examination that can effectively detect cruciate ligament injury, measurements on MRI were valued by clinicians and researchers. For non-contact injuries of the ACL, recent research has identi ed the risk factors, such as the excessive bony posterior tibial slope (PTS), the shallow medial tibial plateau [4][5][6] , and the femoral inter-condylar fossa stenosis 7 8 . Biomechanical studies have shown that excessive bony PTS can increase the tension on ACL, but the soft tissue (cartilage and meniscus) has a compensatory effect on the abnormality of bony structures. Moreover, shallow medial tibial plateau depression (MTD) reduced the limitation of relative motion of the femur over the tibia during closed chain movements 9 10 . Similarly, anatomic abnormalities at the proximal tibia also have an impact on the PCL, the lesser PTS, the more tension the PCL suffered 11 .
Fully understanding the impact of the proximal tibial morphology on PCL is important to guide the treatment and postoperative rehabilitation exercise of patients. However, such a study is lacking in clinical practice. Bony PTS in PCL injured knees measured on lateral radiographs was decreased compared with controls, and Gi n et al. 12 determined that increased PTS in PCL ruptured knees shifted the resting position of the tibia relative to the femur anteriorly. On the other hand, little information is available regarding tibial plateau anatomy measuring on magnetic resonance imaging (MRI) in patients with PCL rupture 13 14 .
Due to the insu cient knowledge in the literature, the principal aim of this study was to gure out the difference in tibial morphologies between patients with PCL injury and those without, and to identify any variations contributing to PCL rupture.

Study population
Approval from the hospital ethics committee was obtained before this study started (IRB Number: 2021-038). 71 patients with PCL injury, identi ed in the Medical Record Registration System of our hospital from January 2015 to December 2020, were eligible for inclusion of this study. The inclusion criteria were subjects with an MRI scan in the picture archiving and communication system (PACS); the PCL rupture was identi ed by arthroscopy. Subjects with previous osteotomy or fracture, meniscal repair, or partial meniscectomy, which would in uence the measurement of PTS and/or meniscal slope (MS); subjects with PCL avulsion fractures or ACL injuries; subjects with discoid meniscus were excluded from this study. Among the 71 patients, 16 patients with ACL injury, ve patients with PCL avulsion fractures, three patients with severe menisci injury (Bucket Handle Tear) that can affect the accuracy of the MS, one patient with previous ACL reconstruction, one patient with lateral discoid meniscus were excluded from this study.
As a result, 45 patients with PCL rupture were included in the study. Among them, some concomitant injuries were presented and included partial meniscus rupture, cartilage injury, bone marrow edema, but those could not in uence the accuracy of the measurements. The PCL-intact control group, matched by age and sex was built to include 45 individuals who came to the radiology department of our hospital for knee-MRI-scanning only with anterior knee pain and without a history of knee injury. And the MRI of each individual was retrospectively reviewed by two experienced orthopedists to ensure the participants' eligibility for the study.
The informed consent requirement was waived by the hospital ethics committee.

MRI Technique
The MRI examinations were performed within one week prior to surgical planning in patients with PCL injuries. All examinations were performed with the same 1.5T MRI scanner (Siemens Magnetom Essenza, Germany).
The coronal and sagittal planes were scanned with the T1-weighted turbo spin-echo (TSE) sequence and proton density (PD) TSE with the fat-suppressed (FS) sequence, and the axial plane was scanned with the PD-TSE-FS. The layer thickness was set to 3 mm, the slice gap was 0.5 mm, the eld of view (FOV) was 160 mm, and the matrix size was 512×512.

Measurements
Anatomic parameters were obtained on MRI by two observers in a blinded and randomized fashion through PACS workstation to determine inter-observer reliability. Both observers were trained in consensus in the measuring methods. The values measured by the 2 observers were averaged for the statistical analysis.
The literature suggested that separate assessment of the PTS was not reliably possible on lateral radiographs, the reliable and reproducible methods conducted on MRI or computed tomography (CT) images were recommended to measure the PTS 6 . It varies in the literature regarding the de nition of the lateral tibial axis on MRI 15-17 , we conducted the method described by Hudek et al. 17 , which was reported to be the most repeatable and reliable method to measure sagittal tibial slopes on MRI 15 a T1-weighted mid-sagittal cut was selected with the appearance of PCL-tibial insertion point, the tibial axis was de ned as the line through both center of circles drawing on the sagittal image, and tibial axis was used to measure the medial and lateral PTS (

Statistical analysis
The average values of the variables measured by both observers were used for all analyses. The Independent t-test was conducted to compare the parameters between the two subgroups, and binary logistic regression analysis was used to determine the probability of PCL injury in an individual based on the measured covariates. All analysis was conducted by a coauthor independently via the SPSS software (Version 22.0; IBM Corp), and the P value of < 0.05 was considered statistic signi cant. The interclass correlation coe cient (ICC) was calculated to assess the reliability of each parameter, with a value of more than 0.75 indicating excellent agreement. G*Power 3.1 (Heinrich-Heine-Universitat Dusseldorf, Dusseldorf, Germany) was used to perform power analysis. Regarding the difference in the lateral PTS between the two groups, for a large effect size (0.56), the results indicated that at least 40 patients were needed in this study (alpha, 0.05; power, 0.8).

Results
The study, as a result, included 45 patients with PCL injuries, with a mean ± standard deviation (SD) age of 43.4 ± 11.8 years, and 45 age and sex matched individuals (age of 42.7 ± 10.2) with intact PCL. The interval time between injury and MRI examination in PCL injured group was 4.2 ± 2.5 days. The descriptive data of both groups is presented in table 1.
The interobserver reliability of each parameter both in the PCL injured group and control group showed excellent agreement, the ICC of each measurement was more than 0.75 (Table2). The results of the independent t-test for comparison of the variables between the two groups were presented in Table3.
The MTD was 1.98 ± 0.64 mm in the PCL-injured group and 2.37 ± 0.55 mm in the control group (P=0.007). Both medial MS and lateral MS were smaller in the PCL-injured group than those in the control group, but the differences did not show statistical signi cance (p= n.s.).
Showing by the results of Binary logistic regression analysis (

Discussion
The most inspiring ndings of the present study were that both PTS and MTD were smaller in patients with PCL injuries than those in controls, but the MS between the two groups was not statistically different. And except the medial PTS, the lateral PTS and MTD were considered underlying anatomic risk factors for PCL injury.
As for anatomic parameters in patients with PCL injury, recent literature 13 proposed that PTS measured on lateral radiographs was decreased as compared with PCL-intact controls, but to the best of our knowledge, the medial and lateral PTS measured on MRI were unclear. As depicted by Bernhardson et al. 19 , PCL graft forces increased as a small tibial slope presented after PCL reconstruction. In addition, Gwinner et al. 20 suggested that PTS could in uence the stability of the knees, especially in patients with attened PTS. In accordance with previous researches, the medial and lateral PTS were found smaller in patients with PCL-injures when compared to controls in this study. The difference in MS between subgroups in our study did not show statistical signi cance, indicating that soft tissue had some compensatory effect on bony PTS. Soft tissues (meniscus or cartilage) contributed to formulating a restriction of posterior direction motion on the knee joints.
3/71 (4.2%) patients in this study were accompanied by severe menisci injury, which was an important manifestation of the loss of soft tissue restriction after PCL injury. It was unclear whether the front of the tibial plateau was thicker or the rear was thinner, further researches should focus on this issue.
On the other hand, there was a controversy in the literature whether the MS in patients with ACL injury was higher compared with controls. [21][22][23] The concavity of medial tibial plateau was considered as a kind of stabilization of knee joints, and it was found decreased in patients with ACL injury 5 . In this study, the difference in MS between subgroups did not show statistical signi cance, and MTD was signi cantly decreased in patients with PCL injuries compared with controls, which was in accordance with previous studies related to the ACL injury 16 21 22 .
We also identi ed that decreased lateral PTS and MTD were risk factors for PCL injury. It suggested that in patients with PCL injury, bony structures (PTS and MTD) played an important role in limiting the forward and backward movement of the tibia related to the femur in patients with PCL injuries. Shelburne et al. 9 modeled cruciate force and found that increase in PTS decreased the native PCL force, noting an increase in PCL force when the slope was decreased. Patients with a normal tibial slope or increased tibial slope may require larger force at the time of injury to overcome the protective effect of posterior tibial slope on PCL injuries. Patients with decreased PTS could suffer from increased stress on the PCL, making it more prone to PCL injury with lesser violence.
Medial PTS was signi cantly decreased in patients with PCL injury compared with PCL intact controls, but it was found not a risk factor for PCL injury. The medial tibial plateau mainly affects the stability of the knee joint in the anteroposterior direction, while the lateral tibial plateau is mainly related to the stability of the knee joint rotation. This may indicate that PCL injury is closely related to the knee joint rotation instability, which needs to be con rmed by further studies. On the other hand, shallower MTD could contribute to PCL injury. The backward movement of the tibia related to the femur would be greater in patients with a shallow medial tibial plateau.
The result of this study may to some extend be helpful in rehabilitation after PCL reconstruction: it should be more conservative in terms of weight-bearing after PCLR in patients with shallow MTD and/or small lateral PTS angle, due to the tibia is more prone to backward movement relative to the femur. If the radical functional rehabilitation is carried out, the risk of the grafts loosening and rupture may be increased. In contrast, whether radical functional rehabilitation is appropriate in patients with larger MTD and/or shaft lateral PTS. The cutoff values of lateral and medial PTS and MTD were unclear.
We acknowledged some limitations to this study. Injury mechanisms (contact and non-contact) and concomitant injuries (medial or bular collateral ligaments) are not included in the analysis, which may in uence the importance of the tibial slope and depth. Some femoral or tibial anatomic parameters are not included in this study, such as lateral femoral condyle ratio 24 , the shape of femoral notch 25 , lateral tibial plateau height. Further research, including femoral morphology is warranted. Notwithstanding these de ciencies, the study has addressed the gap in understanding tibial morphology in patients with PCL injury, and identi es the underlying risk factors for PCL injury. Future researches on anatomic parameters of femoral condyle are required in patients with PCL ruptures.

Strengths and limitations of this study
It is the rst study measured medial and lateral posterior tibial slope angle on MRI in patients with PCL injury, including meniscus slope and medial tibial plateau depth.
We con rmed that soft tissue (meniscus, cartilage) can compensate for tibial plateau slope in patients with PCL injury.
Identify possible anatomic risk factors for PCL injury.
The parameters included in the study were not very comprehensive, and the study did not include the femoral morphological parameters.
Most patients with PCL injury are caused by great violence, however, the injury mechanism of the PCLruptured was not fully investigated in this study.

Conclusion
In patients with PCL injury, lateral PTS, medial PTS, and MTD were found smaller compared to PCL intact controls. Decreased lateral PTS and MTD were underlying anatomic risk factors for PCL injury, with each unit reduction, the risks of PCL injury were 1.2 and 3.1-fold higher, respectively.      Medial tibial depth. Line i: the line tangent to the medial plateau, which is drawn to the proximal cortex border of the tibial plateau. Line j gets through the lowest point of the medial concavity and parallel to line i. The distance between the two lines is de ned as medial tibial depth (MTD).