*3.1. Femoral Rollback*

KA TKA restored the mid-flexion medial and lateral rollback and tibiofemoral axial rotation to levels closer to those of the native knee than MA TKA. The medial and lateral rollback of KA TKA and native knees was similar over the entire mid-flexion range (Figures 4A and 5A). The medial and lateral rollback after MA TKA were significantly lower compared with native knees at both > 40◦ (Figure 4B) and >20◦ (Figure 5B) of flexion. In addition, tibiofemoral axial rotation during flexion after KA TKA was similar to that of the native knee (Figure 6A), while that of MA TKA differed from the native knee in the mid-flexion range (Figure 6B). Remarkably, the femur moved forward during flexion after MA TKA over the entire mid-flexion range, except at 20◦ of flexion (Figures 4B, 5B and 6B).

**Figure 4.** Medial femoral rollback after KA TKA (**A**) and MA TKA (**B**). Rollback following KA TKA and of the native knees were similar at all flexion angles. Meanwhile, rollback following MA TKA was reduced at flexion angles > 40◦ . Error bars denote standard deviations. Significant differences (*p* < 0.05) are marked with asterisks.

**Figure 5.** Lateral femoral rollback after KA TKA (**A**) and MA TKA (**B**). Rollback after KA TKA compared to native knee rollback over the entire mid-flexion range, but in the case of MA TKA, the rollback was significantly smaller, except at a flexion angle of 20◦ . A paradoxical forward movement was observed after MA TKA. Error bars denote standard deviations. Significant differences (*p* < 0.05) are marked with asterisks.

**Figure 6.** Tibiofemoral axial rotations after KA TKA (**A**) and MA TKA (**B**). During flexion, KA TKA and native knee rotations were similar, while rotation following MA TKA was quite different from that of the native knee. A paradoxical forward movement was observed after MA TKA. Significant differences (*p* < 0.05) are marked with asterisks.

### *3.2. MCL Strain*

KA TKA was better for restoring the MCL strain to that of the native knee over the entire mid-flexion range than MA TKA. The mean strain measurements following KA TKA and those of the native knee were alike over all ranges (Figure 7A). The MCL strain after MA TKA was two-fold greater than that of the native knee at flexion angles > 20◦ (Figure 7B).

**Figure 7.** MCL strain after KA TKA (**A**) and MA TKA (**B**). The mean strains following KA TKA and that of the native knee were consistently similar, while the mean strain after MA TKA was about two-fold higher at >30◦ of flexion. Error bars denote standard deviations. Significant differences (*p* < 0.05) are marked with asterisks.

#### **4. Discussion**

Despite advancements in both technology and surgical techniques of MA TKA, patient dissatisfaction with post-TKA pain relief and overall outcomes remains high; a substantial proportion of patients report knee abnormalities [1,2,5,30,31]. KA TKA seeks to restore the anatomy of each individual patient; the kinematic and clinical outcomes are better than those of MA TKA [14–16,18]. Joint line elevation after MA TKA is associated with a risk of mid-flexion instability; theoretically, KA TKA restoration of native joint line height and obliquity makes mid-flexion biomechanics more natural than MA TKA [20,21,25,32]. However, it remains unclear whether restoration of the joint line configuration affects post-TKA mid-flexion kinematics and laxity [12,18,21,24–26]. Therefore, this matched pair cadaveric study tested which TKA alignment concept, KA or MA, would reproduce more native mid-flexion rollback and MCL strain.

The present study's results endorse that KA TKA provides better physiological kinematics over the mid-flexion range than MA TKA, as hypothesized. We found that after KA TKA, medial and lateral femoral rollback, and axial rotation, were consistently similar to those of the native knee, whereas for MA TKA this was the case only at ≤40◦ of flexion. In addition, a paradoxical femoral forward movement during flexion was observed after MA TKA. Our findings agree with those of recent cadaveric studies, which suggest that compared to MA TKA, knee kinematics after KA TKA were more alike to those of native knees [14,15,26]. Our results, and those of previous studies, reveal why patients who undergo KA TKA often report superior mid-flexion functional performance compared with those who undergo MA TKA.

Our findings also support the hypothesis that the joint line following KA TKA is better for restoring natural MCL strain during mid-flexion than the perpendicular joint line after MA TKA. KA TKA resulted in MCL strain that was consistently comparable to that of native knees, whereas, in the case of MA TKA, it was twice as high. Our findings agree with those of a recent cadaveric study: KA TKA was better for restoring the magnitude and distribution of MCL strain to natural levels than MA TKA [18]. Although it is challenging to directly liken our results with those of prior studies that assessed MCL strain via linear, two-dimensional measurements of length changes under valgus stress, our findings support previous studies that report that restoration of the pre-arthritic joint line provided a more physiological MCL strain than traditional MA TKA [12,14,15,26,33]. Furthermore, our results, when taken into consideration with the widely acknowledged existence of a nociceptor in the MCL [34], indicate that patients may experience less pain and more native knee sensations during mid-flexion after KA TKA, as opposed to MA TKA.

The results of the present study propose that restoration of natural MCL strain via KA TKA may explain the more physiological knee kinematics evident after KA TKA compared to MA TKA. In this study, KA TKA reproduces more physiological medial pivot motion, while MA TKA results in paradoxical anterior motion during mid-flexion. Given that the MCL serves as the fundamental restraint in ACL-deficient, prosthetic knees, our findings suggest that MCL strain may be strongly associated with restoration of knee kinematics. Interestingly, a recent cadaveric study reported that restoration of joint line obliquity was not associated with mid-flexion coronal plane laxity, if the medial joint line height was restored [21]. However, although we restored the medial joint line height in all knees, our study found a significant difference between KA and MA TKA knees in the MCL strain. Still, the correlation between these findings is limited, since the previous study assessed soft tissue laxity by the length changes of MCL after the valgus load, and information on knee kinematics was not presented. Although the cause of the inconsistencies is unclear, one plausible explanation is that, although valgus stress increases the MCL strain, the MCL length may not change if the stress is lower than the threshold required for such change.

Due to the use of a cadaveric model, this study has some limitations. First, specimen preparation and the squatting loads used may not have been entirely natural. Second, as the tissue quality around the knee joint is associated with the severity of knee OA, the experimental results could be affected by the OA status of the cadaver. However, most knees in this study lacked advanced osteoarthritis necessitating TKA, so caution is required when extrapolating our findings to clinical practice. Third, this study featured PS prosthesis, which requires consideration of the implant feature prior to any broad generalizations, as implant design has been shown to be strongly correlated with knee kinematics [26,35]. Generally, when performing KA TKA, a cruciate-retaining (CR) prosthesis is recommended. However, a recent study found that native knees, and knees in which CR and PS prostheses were placed during KA TKA, had similar kinematics and soft tissue laxity [14]. Fourth, because the thresholds for pain and mechanical failure in the human knee are unknown,

assessing the clinical significance of MCL strain was difficult. Fifth, it is possible that the study was underpowered and subject type-II error with respect to detecting all relevant outcomes. Sixth, the group assignment for the knees were not kept blinded to all investigators, which could have raised ascertainment bias. Finally, biofeedback, which is natural in knee function and enhanced by the innovative TKA surgical technique, could have introduced nonlinearity in the feedback and improved the signal-to-noise ratio in the loop [36]. Thus, the video extensometer that we used does not process data with 100% accuracy; changes in illumination may affect image processing, because illumination affects the properties of biomaterials. Nevertheless, we used a matched-pair design to minimize confounders, and this is the first study to simultaneously measure femoral rollback and MCL strain. These results provide valuable insight on the differences regarding mid-flexion kinematics and patterns of MCL strain between KA TKA and MA TKA.

#### **5. Conclusions**

We investigated whether restoration of the pre-arthritic joint line following TKA would affect post-TKA biomechanics. Restoration of the height and obliquity of the prearthritic joint line following KA TKA reproduces more natural rollback and MCL strain than alteration of the joint line following MA TKA over the entire mid-flexion range. Future studies focused on the development of both the motion analysis system that assesses the knee kinematics of patients in real clinical practice and the algorithm that recommends the optimal implant position restoring native knee kinematics are required.

**Author Contributions:** Conceptualization, I.J.K., D.-S.K. and D.L.; methodology, D.-S.K. and M.S.K.; software, D.-S.K. and M.S.K.; validation, D.-S.K. and D.L.; formal analysis, D.-S.K.; investigation, I.J.K. and Y.I.; resources, D.-S.K., D.L. and I.J.K.; data curation, D.-S.K., N.C. and M.S.K.; writing—original draft preparation, Y.D.K. and D.-S.K.; writing—review and editing, I.J.K. and N.C.; visualization, D.-S.K. and I.J.K.; supervision, I.J.K. and Y.I.; project administration, I.J.K., D.-S.K. and D.L.; funding acquisition, I.J.K., D.-S.K. and D.L. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (No. 2019 R1F1A 1057842 and No. 2017M3A9E9073545) and the Industry-Academy Cooperation Program through the Industry Academic Cooperation Foundation of the Catholic University of Korea (5-2020-D0721-00001). The funding source did not play any scientific role in performing this study.

**Institutional Review Board Statement:** This cadaveric study was conducted in compliance with the Act on Dissection and Preservation of Corpses of the Republic of Korea (act number: 14885) and approved by the Institutional Cadaver Research Committee of College of Medicine, The Catholic University of Korea (code number: R19-A018).

**Informed Consent Statement:** Written informed consent for use of the cadavers and consent for use of future research on the related materials were provided by all donors or authorized representatives.

**Data Availability Statement:** All data are presented in the article. Instrumental readings are available upon request from the corresponding author.

**Acknowledgments:** We thank Smith & Nephew (Seoul, Korea) for providing all surgical instruments. In particular, we thank Su Gu Chai B.S. and Seung Jun Lee P.R.S. of Smith & Nephew and Ki Joon Yoo B.A., Jae Young Sung B.S. and Hyunggu Han B.A. of Daon HealthCare (Seoul, Korea) for their assistance. We also thank the cadaver donors and their families.

**Conflicts of Interest:** The authors declare that they have no conflict of interest.

#### **References**

