The most significant finding in the present study is that both hinge position and coronal correction angle cause tibial rotation to some extent during uniplanar MOWHTO. Sagittal hinge position changes can influence tibial rotation, while axial hinge position changes cannot. Further, linear correlations were observed between the coronal correction angle and tibial rotation at varying degrees, depending on different hinge positions.
The direction and extent of tibial rotation during MOWHTO have been controversial, but only a few studies have reported on such issues. In clinical settings, Jang et al.[9] observed a wide range of DTR after biplane MOWHTO, from 9.5° externally to 20.9° internally. Li et al.[5] reported DTR with a range of − 9.6° to + 2.8° after uniplane MOWHTO in 41 knees. In a study involving 60 knees, Sasaki et al.[16]found that 32% had internal rotations, 67% had external rotations, and 1% had no rotation. In Jing’s study[23], a total of 106 knee joints were included, being 60% cases internal rotation of distal tibia, 38% external rotation, and 2% no rotation. As a new direction, Hinterwimmer et al. [24] measured the axial rotation of the tibia using two K-wires fixed to the tibia distal and proximal to the osteotomy during MOWHTO by using a sterile goniometer involving 50 patients. The average rotation of the distal tibia was 4.4°±2.8° internally. A study by Kendoff et al.[11] demonstrated that in 13 cadavers, all osteotomies fixated with opening width of 12.5mm using a tapered Puddu plate, the distal tibia was externally rotated in 10 of the 13 cadavers and the mean rotation increased by 2°± 6°, ranging from a maximal internal rotation of 9°to a maximal external rotation of 12°. A study by Jacobi et al. [10] demonstrated that in 11 cadaver limbs, the average rotation change of distal tibia was 1.5°±2.9° internally with a range of 4.5° external rotation to 15° internal rotation. By means of virtual osteotomy with 3D tibial models, we found external, unchanged and internal rotation may occur and more cases tended to internal rotation as mentioned above.
The related influencing factors of tibial rotation have been investigated by several researchers, but an uncontested result has yet to be provided. Clinically, Kim et al. [8] found that the degree of DTR was not related to the amount of coronal correction after MOWHTO. Kim et al.[15] reported that the correction angle was the only predictor of DTR, with a larger correction angle observed in the group with distal tibial rotation > 3° compared to the group with rotation ≤ 3°. Lee et al.[17] found a positive correlation between the hinge axis angle and internal rotation of the distal tibia when the hinge axis was positioned more posterolaterally, while the gap ratio was negatively correlated. Kuwashima et al.[25]suggested that the postoperative tibial rotation was related to the degree of coronal correction. In Jing’s[23] study, DTR change was significantly associated with angles between standard and actual hinge in the sagittal planes, followed by opening width and flange angle. In a cadaveric study conducted by Jacobi et al.[10], advanced OA with concomitant flexion contracture was found to be associated with a significantly decreased tibial rotation change. In a 3D surgical simulation study by Chang et al.[14], a large angle of wedge inclination, which was the incline angle of the bone saw relative to the plateau plane, was found to be an indicator of horizontal rotation. In such study, the results showed that the distracted plateau rotated medially or anteriorly according to different angle of wedge inclination. The aim of Chang’s study was partially similar to that of the present study, but obviously, the deficiency of his study was that the sample size was too small, and only a proximal tibia of a healthy subject was reconstructed and printed as a 3D model. Meanwhile, through further research with virtual osteotomy involving 30 cases, the direction and extent of tibial rotation was found to be related to the hinge position and coronal correction angle, as previously mentioned in our study.
In the clinical practice of MOWHTO, preoperative planning for coronal correction angle is mainly dependent on standing hip-to-ankle AP radiography. Obviously, the hinge position and inclination in the 3D space cannot be considered in two-dimensional imaging, as described in our study and several previous studies. The hinge position has an effect on both PTS and tibial rotation [6, 12, 14, 26]. In a 3D simulation uniplanar MOWHTO study, Teng et al. [12] found that axial hinge position changes significantly influenced PTS in MOWHTO, but sagittal hinge position changes had no effect on changes in PTS. Notably, in the present study, sagittal hinge position changes could more obviously influence tibial rotation in uniplanar MOWHTO compared with the axial hinge position. Combined with clinical practice, both the axial and sagittal positions of the hinge must be carefully considered at the same time during MOWHTO to avoid unintended PTS and DTR.
Though an osteotomy line parallel to the medial joint line in sagittal plane has been recommended in previous studies[7, 27, 28], such method cannot be performed easily. Lee et al.[28] reported there was only 12.9% of cases in which a preoperative plan was performed parallel to the medial joint line, with most osteotomy lines being anteriorly inclined. Similarly, Chung et al. [7] noted that 72.5% of cases showed anteriorly inclined osteotomy, while the remaining cases showed posteriorly inclined osteotomy. In order to achieve an osteotomy line parallel to the medial joint line, Lee et al. [28] recommended a technique in which two syringe needles are used to identify the medial joint line and the pes anserine tendon is completely detached from its insertion.
There are some limitations in the present study. First, there was a relatively small sample size with only 30 models. Second, the uniplanar osteotomy performed is different from biplanar osteotomy which is more popular in clinical practice, because an ascending osteotomy plane would have an effect on tibial rotation in the biplanar osteotomy. Third, the present study is a virtual surgical study conducted using software, and the soft tissue around osteotomy site, the position of plate or bone density were not considered. In addition, clinical outcomes such as the influence on articular cartilage and function were not included. However, to the best of our knowledge, there has been no 3D simulation MOWHTO study focusing on tibial rotation yet, and further investigations with a larger sample size and clinical researches are needed.