Knee ankle joint line angle: a critical value at 10 (cid:0) for decompensated knee joint line obliquity in medial opening wedge osteotomy.

Background: Medial opening wedge high tibial osteotomy (MOWHTO) changes the knee joint inclination in the coronal plane, which can be compensated by the ankle joint. Once there is a decompensated knee joint obliquity, it can induce excessive shear force on the articular cartilage. This study aimed to investigate the capacity of the compensation by analyzing the correlation of the knee-ankle joint line angle (KAJA) and the knee joint line obliquity (KJLO). Methods: The correlations between postoperative KJLO and body mass index (BMI), correction amount, KAJA, mechanical lateral distal femoral angle (mLDFA), preoperative medial proximal tibia angle (MPTA), ankle joint line obliquity (AJLO), KJLO and mechanical hip-knee-ankle angle (mHKA) were analyzed using Pearson correlation coecient. The contribution of signicant factors was further analyzed using multiple linear regression. The KJLO between ≦ 5 (cid:0) , 5 (cid:0) -10 (cid:0) And ≧ 10 (cid:0) KAJA groups were compared using Kruskal-Wallis test. Results: Postoperative KAJA and preoperative KJLO moderately correlated to postoperative KJLO. Preoperative MPTA, mHKA, AJLO weakly correlated to postoperative KJLO. After multiple linear regression, only postoperative KAJA, preoperative MPTA and mHKA still showed signicant contribution, while preoperative KAJA made the greatest contribution. The KJLO was substantial higher in the ≧ 10 (cid:0) KAJA group with a high rate (68%) of high-degree KJLO. Conclusions: 10 (cid:0) postoperative KAJA is a critical value for decompensated KJLO. The results suggest us carefully assess the KAJA intraoperatively. Double osteotomy should be considered if ideal alignment cannot be achieved when the KAJA reach 10 (cid:0) .


Background
Medial opening wedge high tibial osteotomy (MOWHTO) is a commonly performed procedure for medial osteoarthritis of the knee [1][2][3][4]. In order to relieve the stress at medial compartment, MOWHTO shifts the mechanical axis of the lower extremity laterally. However, this procedure inevitably changes the joint inclination in the coronal plane since the medial tibia is elevated [5][6][7][8][9]. In cases of severe varus deformity of the knee joint due to proximal tibial deformity or combined deformity of both the distal femur and proximal tibia, single-level osteotomy can result in high-degree joint obliquity because a large amount of correction is required [5][6][7], which can induce excessive shear force on the articular cartilage [10].
Increased shear force may subsequently cause chondrocyte death [11,12]. Therefore, in the circumstances, double osteotomy is indicated to ensure a physiologically oriented joint line [6,13].
There have been several studies discussing preoperative parameters associated with greater postoperative knee joint line obliquity [9,14]. Preoperative mechanical hip-knee-ankle angle (mHKA), medial proximal tibial angle (MPTA) and joint line convergence angle (JCLA) are well-established predictors of postoperative abnormal joint obliquity. The reason why these parameters can predict abnormal obliquity was explained by the necessity of a large correction in previous studies. However, in some cases, we observed that after MOWHTO, even a relatively small correction can still result in highdegree joint obliquity. Therefore, there must be some other factors which was not found yet.
The postoperative knee joint obliquity can be compensated by the ankle joint [8,[15][16][17][18]. The capacity of compensation was suggested to be determined by the mobility of the subtalar joint [17,19], which differs by individual. Therefore, we speculated that the compensation of the ankle joint has its limits. Exceeding the limits may result in an unacceptable rate of high-degree knee joint obliquity even after a small correction. This study aimed to investigate the correlation of the knee-ankle joint line angle (KAJA) and the knee joint line obliquity (KJLO). The hypothesis of our study was that higher postoperative KAJA would result in higher KJLO, and there would be a critical value of KAJA above which the KJLO could not be well compensated.

Methods
We conducted a retrospective observational study at the corresponding author's hospital. The study was approved by the Ethics Committee of the hospital with a waiver of informed consent for the retrospective use of patient data (approval number: 201910040RIND). We investigated 111 consecutive patients who underwent MOWHTO between January 2016 and April 2019. The indications for MOWHTO were as follows: symptomatic medial unicompartmental osteoarthritis with a mechanical tibio-femoral angle of at least 5°, and exion contracture of less than 10°. Fourteen patients were excluded owing to lack of preoperative or postoperative standing anteroposterior radiographs of the lower extremities. One patient was excluded because proximal bulectomy was performed concurrently with MOWHTO. Finally, 96 patients were included.

Surgical technique
We performed arthroscopic examination before MOWHTO. Arthroscopic drilling was performed if there was subchondral bone exposure. After arthroscopy, we made a skin incision on the anteromedial aspect of the tibia. Generally, the detailed steps were the same as those described previously [20,21]. We used a locking plate designed for the medial proximal tibia for xation of the biplanar MOWHTO. The procedure aimed to align the weight-bearing line through the Fujisawa point [1]. We used "the alignment rod" method [22] to check the correction intraoperatively.

Radiological evaluations
On the standing radiographs, we measured the preoperative and postoperative MPTA, KAJA, weight bearing line (WBL) ratio, KJLO. The measurement methods are described as follows: (1) MPTA: It was the medial angle between the mechanical axis of the tibia and the articular surface of the proximal tibia. (Fig. 1a) (2) KAJA: It was the angle between the lines tangent to the articular surfaces of the proximal tibia and distal tibia. A positive value represented a valgus relationship of these two surfaces, and a negative value represented a varus relationship. (Fig. 1b) (3) mHKA: It was the angle between the mechanical axes of the femur and tibia. (Fig. 1c) (4) Mechanical lateral distal femoral angle (mLDFA): It was the lateral angle formed between the femoral mechanical axis and the joint line of the distal femur. (Fig. 2a) (5) WBL ratio: It was determined by the intersection of the articular surface of the proximal tibia and a line from the center of the femoral head to the center of the talus. The ratio was obtained by dividing the distance measured from the edge of the medial tibial plateau by the total width of the articular surface.
The discrepancy between preoperative and postoperative WBL ratios was de ned as the correction amount.  Table 2. Table 1 The   The postoperative KJLO signi cantly differed between ≦ 5 , 5 -10 And ≧ 10 KAJA groups. The mean KJLO value were 1.6 , 1.9 and 5.6 , respectively. The rates of high-degree KJLO were 8.5%, 14.8% and 68%, respectively. The results of different KJLO groups were shown in Table 3. Intraclass correlation coe cients of intraobserver and interobserver agreement of radiologic evaluations were all acceptable, > 0.87 (range 0.87-0.98). The most important nding of this study was that the postoperative KAJA signi cantly correlated with the KJLO. The contribution of KAJA was stronger than that of previously reported factors. The rate of ≧ 5 KJLO achieved 68% when the postoperative KAJA exceeded 10 . Therefore, double osteotomy at the distal femur and proximal tibia should be considered if ideal alignment cannot be achieved even when the KAJA reach 10 .
Non-anatomic knee joint obliquity can cause excessive shear force on the articular cartilage. When the obliquity angle is 5 , the shear force in the medial compartment elevate to almost twice as high as the normal knee [10]. The force value becomes even higher as the obliquity angle increases. Although some studies have shown that there was no difference of short-term outcome with high-degree obliquity [8,23,24], it is reasonable there would be a long-term adverse effect on the articular cartilage [12]. Lee et al found that the change of KJLO was signi cantly less than that of anatomical geometry of the proximal tibia. This phenomenon can be explained by the compensation of the ankle joint [8,18]. However, the capacity of compensation differs among individuals [9,15,25], and currently there is no information about the maximum of the capacity for most of the patients undergoing MOWHTO. Understanding when the knee obliquity will be "decompensated" is essential to ensure a satisfactory outcome. In this study, although the postoperative KAJA moderately correlated with the KJLO, the difference of the mean KJLO between the ≦ 5 Group and 5 -10 group was as small as 0.3 , which may be lack of clinical relevance. The mean KJLO became substantially higher when the KAJA was larger than 10 . Therefore, these results suggest that the change of the anatomical geometry of the proximal tibia caused by MOWHTO could not be well compensated by the mobility of the subtalar joint once the KAJA exceeded 10 . When performing MOWHTO, the 10 KAJA as a critical value should be kept in mind.
In accordance with previous studies, higher preoperative mHKA and MPTA were also predictors of higher postoperative KJLO [9,14]. It was reasonable that higher mHKA might require a larger correction, which was supposed to result in greater obliquity. However, to our surprise, the correction amount evaluated by the change of WBL ratio was not a signi cant factor. The postoperative mean WBL ratio was 63.3% in this study, which was remarkably close to the classic aim at 62.5% [1]. Therefore, correcting a lower limb from a higher mHKA to the classic surgical goal, adherence the WBL to the Fujisawa point, tended to have greater postoperative obliquity. We speculated that the change of WBL ratio can be affected by multiple factors, such as different anatomical variance of the femur and tibia. Thus, it did not necessarily represent a large angle correction. In contrast, a higher preoperative mHKA could be a more representative factor. Furthermore, a higher preoperative MPTA implied that the proximal tibia attribute less to the varus deformity of the lower extremity. Therefore, MOWHTO on the knees with higher MPTA may result in a less physiologically oriented joint line.
The KAJA of all cases increased after MOWHTO. Although postoperative KAJA contributed the most to the postoperative KJLO, preoperative KAJA did not correlated to it. Similarly, after multiple linear regression, preoperative KJLO did not contribute to it. In other words, only a su cient KAJA increase really resulting in high postoperative KAJA caused high-degree KJLO regardless of the preoperative value of KJLO and KAJA. These results suggest that the KAJA be assessed carefully during the operation to avoid exceeding the critical 10 KAJA. Because a full-tibia image cannot be obtained by uoroscopy, we recommend assessing the intraoperative KAJA by measuring the angles between the tibia articular surfaces and a reference rod at the knee and ankle joints. KJLA can be simply calculated by subtraction of these two angles. The measurements do not add additional tasks because they can be done along with the assessment of the mechanical axis by the "cable method" [26] or "alignment rod" method [22].

Availability of data and materials
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.