The insufficient coverage of lateral trochlear resection is correlated to 1 the tibiofemoral alignment parameters in kinematically aligned TKA: 2 A retrospective clinical study

5 Background: The mismatching of the femoral component and trochlear resection surface is very 6 common in kinematically aligned total knee arthroplasty (KA-TKA) when conventional prostheses 7 are employed. This mismatching is mainly manifested in the insufficient coverage of the bone cut 8 surface of the lateral trochlea. The aim of present study is to explore whether this mismatch is related 9 to the alignment parameters of the tibiofemoral joint. 10 Methods: 45 patients (52 knees) who underwent KA-TKA in our hospital were included. There 11 were 16 patients (16 knees) received surgery using patient specific instrumentations, and 12 conventional instruments with caliper and other special tools were employed in the other 29 patients 13 (36 knees). The widths of exposed resection bone surface at the middle (MIDexposure) and distal 14 (INFexposure) levels on the lateral trochlear were measured as dependent variables, while the hip- 15 knee-ankle (HKA) angle, mechanical lateral distal femoral angle (mLDFA), joint line convergence 16 angle (JLCA), medial proximal tibial angle (MPTA) and transepicondylar axis angle (TEAA) were 17 measured as independent variables. Correlation analysis and subsequent linear regression were 18 conducted among the dependent variables and various alignment parameters of the tibiofemoral 19 joint. 20 Results: The incidence of insufficient coverage of the lateral trochlear cut bone surface was 86.5%, 21 with MIDexposure and INFexposure being 2.3 (0-6mm) and 2.0 (0-5mm), respectively. The widths 22 of two levels of exposed bone resection were significantly correlated to mLDFA and HKA, but were 23 not related to TEAA. Conclusions: the insufficient coverage of trochlear resection surface in KA-TKA is negatively 25 correlated with the degrees of valgus of the distal femoral joint line and the degrees of varus of the 26 knee. The present study suggest that in the development of KA-specific prostheses, attention should 27 be paid to the effects of tibiofemoral alignment parameters on the prosthetic matching of the 28 trochlear resection surface.

3 or Grade IV, ≤ 5° of varus of proximal tibia (medial proximal tibial angle, MPTA ≥ 85°), 76 and ≤ 5° of valgus deformity. Exclusion criteria: inflammatory arthritis, previous knee injury 77 or ligament insufficiency, genu recurvatum, and ≥10° of fixed flexion contracture. Given that 78 trochlear dysplasia may have unknown impacts on this study, patients diagnosed with trochlear 79 dysplasia before KA-TKA were excluded from this study. From May 2018 to August 2020, a total 80 of 55 patients underwent KA-TKA in our institution. After excluding 3 patients diagnosed with 81 trochlear dysplasia and 7 patients with missing intraoperative measurement data of bone cut 82 exposure, 45 patients (52 knees) were finally included in this study. Eleven patients were men and 83 34 patients were women. The mean age was 69.5±6.7 years (56-85 years), and the mean body mass 84 index was 28.6±4.2kg/m 2 (26-34 kg/m 2 ). The preoperative radiographic changes of 1 knee were in 85 accordance with Kellgren-Lawrence Grade Ⅲ, and the other 51 knees were in compliance with 86 grade IV. 50 patients had varus knees and 2 patients had valgus knees. Among them, 16 patients (16 87 knees) were performed KA-TKA with the assistance of patient-specific instrumentation (PSI-KA) , 88 and 29 patients (36 knees) were performed KA-TKA using conventional instruments with 89 measurement tools (Calipered kinematically aligned instrumentation, Calipered-KA). All bilateral 90 KA-TKAs (7 patients) were performed using Calipered-KA technique. 91

Surgical plans 92
The design of PSI was based on full-length computed tomography (CT) of lower extremities, 93 while the surgical technique of Calipered-KA followed the technique recommended by Howell et 94 al [14]. No matter which assistant alignment instrument was used, articular surface-based bone cut 95 approach is adopted [7]. A Vernier caliper was used to measure the thickness of the resected bone 96 pieces of the distal femoral condyles, posterior condyles, and tibial plateau. The general principle is 97 that the sum of the thickness of the resected bone piece, the compensation thickness of the worn 98 cartilage, and the width of saw kerf is equal to the thickness of the component. 99 For the manufacturing of PSI, CT data (Slice thickness, 0.625mm) was collected and imported 100 into Mimics (version 17，Materialise NV, Belgium) for 3D reconstruction, then the solid models 101 was imported into NX 9.0 (Siemens PLM Software, TX, US) for the design of PSI (Figure 1a,b) . 102 Rapid prototyping technology (Formiga P 110, EOS, Krailling, Germany) was used for 3D printing 103 of the PSI. The printing material is medical nylon (PA2200 Polymer powder, EOS, Krailling, 104 Germany), which can be sterilized using autoclaving. 105 For the intraoperative application of PSI, all the residual articular cartilage should be removed 106 using a curette before the PSI was secured to its unique position. Different from the management of 107 articular cartilage in PSI-KA, Calipered-KA only removed the residual cartilage on the severely 108 worn side. If the contralateral articular cartilage is intact, then keep it in place. Stacked neodymium 109 magnets (1mm of thickness each) were used to compensate the cartilage thickness on the severely 110 worn side (Figure 2a, b). prostheses (Gemini MK Ⅱ, Link, Hamburger, Germany) were used in current study. 118

Parameters measurement 119
Before KA-TKA, the full-length weight bearing radiographs of lower limbs were obtained from 120 all patients. hip-knee-ankle angle (HKA) was measured from the full-length radiograph. HKA was 121 defined as the angle between the mechanical axes of the femur and the tibia, the value of varus HKA 122 was defined as a positive, and the value of valgus HKA was defined as a negative. Other alignment 123 parameters' measurements followed the methods described by Paley [15].  Figure 4). The widths 136 of the exposed trochlear resection at these two levels were used as the dependent variable for 137 subsequent analysis. 138

Statistical analysis 139
All measurement parameters are subjected to Shapiro-Wilk test. Parameters that conformed to the 140 normal distribution were presented as mean±standard deviation, and those that did not conform to 141 the normal distribution were presented as Median (Interquartile range). In patients receiving PSI-142 KA, the reliability test of the two sets of TEAA data from CT measurement and intraoperative 143 measurement was conducted. Reliability was determined by calculating the intraclass correlation The data of mLDFA, MPTA and TEAA conformed to the normal distribution, while other 152 parameters did not conform to the normal distribution. The exposed width of the distal and middle 153 5 level of lateral trochlear bone cut surface were 2.0(2.4)mm (0-5mm) and 3.0(2.9)mm (0-6mm), 154 respectively. The other measurement results of the respective variables are shown in Table 1

159
The reliability test (ICC, Two-way mixed effects, random, fixed effects) of the two sets of TEAA The results of Spearman correlation analysis showed that the two levels of bone cut exposure 165 were significantly correlated to mLDFA and HKA, but not correlated to other independent variables 166 (Table 2)  Current study confirmed that the extent of the exposed bone cut surface of the lateral trochlear 175 during KA-TKA is indeed negatively correlated with mLDFA: the more valgus the joint line of the 176 distal femur, the larger the exposed bone cut surface on the lateral trochlea. In addition, the extent 177 of exposed bone cut surface is also weakly correlated with the alignment of the lower extremities: 178 the more severe the varus knee, the smaller the exposed bone cut surface of the trochlea. The results 179 of this study suggested that the design of a new prosthesis that meets the technical requirements of The majoraty of TEAA data were acquired from intraoperative measurement, which might impact 222 the accuracy of TEAA. However, The preoperative CT data were retrospectively collected from 16 223 patients who received PSI-TKA , and the reliability test result of the two sets of TEAAs is "good".

Availability of data and materials 260
The organized dataset used and/or analyzed in the current study has been uploaded as a supplement 261 to the manuscript, and the raw data is available from the corresponding author on reasonable request. 262 263

Competing interests 264
The authors declare that they have no competing interests.  The white and blue arrows represent the width of the exposed bone cut surface at the distal 361 MIDexposure represent the width of the exposed bone resection at the distal and middle level of the 366 trochlea, respectively. 367