Study design
This noncomparative, open-label, single-arm clinical study was carried out after ethics committee approval was obtained. The sample size was calculated using an equality test for a one-sample design (power of 80%, α of 0.05 and standard deviation of 1.7 for the mechanical axis) [26]. Based on these estimates, a total of 235 cases were required. Considering a ~ 5% dropout and/or withdrawal rate, N = 245 knees were included. The inclusion criteria comprised informed consent, coronal plane deformity < 25° varus or valgus, sagittal plane deformity < 15°, and satisfactory LLRs. Knees with > 25° coronal plane deformity, > 15° sagittal plane deformity, or unsatisfactory LLRs were excluded. We screened 239 consecutive patients who underwent TKA between October 2018 and March 2019 and included 191 patients (245 knees) and excluded 48 patients (48 knees). Twenty-three knees with coronal plane deformity > 25°, 13 knees with sagittal plane deformity > 15° and 8 knees with poor-quality LLRs (nonvisualization of the center of the femoral head) were among the excluded knees. Four knees were excluded due to nonavailability of LLRs due to X-ray machine breakdown. No patients were excluded or dropped out postoperatively.
Surgical planning
All surgeries were planned and performed by the first author (DS). As a reliable tool for measuring the mechanical axis and component position, standing LLRs were used to measure all pre- and postoperative radiological parameters and plan proximal tibial and distal femoral bone resection [13]. Standardized digital LLRs were obtained in the standing position by a previously described method [13]. Radiographs were evaluated for acceptable image quality and rotation by assessing the profile of the lesser trochanter, the positioning of the patella over the femur (central position was accepted), and the amount of overlap of the fibular head on the tibia (one-third of fibular head overlap was accepted). The radiographs were considered satisfactory when the above criteria for at least two landmarks were met. Once satisfactory radiographs were obtained, the patients were included in the study. Postoperative standing LLRs were obtained when patients were able to complete a full active straight-leg raise and stand straight on the radiography platform. The following pre- and postoperative coronal plane radiological parameters were measured on a digital radiograph workstation: (i) the mechanical axis (hip-knee-ankle [HKA] axis), i.e., the angle between the femoral mechanical axis (line joining the center of the femoral head and the center of the femoral/femoral implant trochlear notch), and the tibial mechanical axis, i.e., the line joining the point between the tibial spine/center point of the tibial base plate to the central depression on the dome of the talus; (ii) the mechanical lateral distal femoral angle (mLDFA), i.e., the lateral angle between the femoral mechanical axis and the tangent line connecting the distal femoral condyle preoperative and femoral prosthesis postoperative; and (iii) the medial proximal tibial angle (MPTA), i.e., the medial angle between the tibial mechanical axis and the tangent line connecting the proximal tibial condyles preoperative and tibial component postoperative. The aim was to restore the mechanical axis classically within 3° of the neutral mechanical axis (HKA axis 180 ± 3°).
The second author (KCP) measured and recorded all the pre- and postoperative outcome variables. Since he was not the operating surgeon, the use of this method also helped to avoid any operator bias. The outcome variables of 20 randomly selected knees were remeasured after 2 weeks by KCP. These variables in the same set of knees were also measured by DS to test the intraobserver and interobserver variability.
Distal femoral resection planning
The aim of distal femoral resection was to cut the distal femur at 90° to its mechanical axis. The femoral mechanical axis (line AB) joining the center of the femoral head (point A) and the deepest point of the intercondylar notch (point B) on LLRs (Figs. 1a & 2a) was drawn. Then, a second line (CD) was drawn tangent to the distal femoral articular surface. This angle ABC was the mLDFA (Fig. 2a). If the mLDFA was 90°, then bone cuts of equal thickness for both distal femoral condyles were planned. If the mLDFA was > 90° or < 90°, then another line (CD) perpendicular to the mechanical axis of the femur (line AB) and tangent to at least one femoral condyle was drawn. The length of line EF was the difference in bone resection between the two condyles (Fig. 2b). Since intramedullary jigs were used to guide resection, the VCA were calculated by drawing a line from the deepest point of the trochlear groove (portal of entry) to the center of the medullary canal at 23 cm from the portal of entry (line BG). Line BG was the trajectory for the intramedullary rod, and angle GBC was the planned VCA (pVCA) (Fig. 2b). Intraoperatively, bone to be resected from the distal femoral condyle with the pVCA was measured after removing cartilage from the unaffected condyle. If it matched the preoperative plan for bone resection, then the same bone thickness was resected. If it did not match, then the VCA was adjusted to the planned bone resection thickness. The final VCA read on the jig was the executed VCA (eVCA). The femoral component position was confirmed on postoperative LLRs (Fig. 2c, d & e).
Proximal tibial resection planning
Proximal tibial resection was planned at 90° to the tibial mechanical axis. The tibial mechanical axis (line AB) was drawn using the center of the tibial spine (point A) and the deepest point of the superior articular surface of the talus (point B) (Fig. 3a). The second line CD was drawn tangent to both proximal tibial condyles. The angle BAD was the MPTA. Then, a line (CD) tangent to the unaffected proximal condyle and perpendicular to the tibial mechanical axis (AB) was drawn, and the difference in the thickness of the bone to be resected was measured, i.e., the length of line EF (Fig. 3b). Intraoperatively, the mark on the superior surface of the cutting block of the extramedullary tibial jig was aligned to a line drawn on the center of the tibial spine from anterior to posterior ignoring the position of the tibial tubercle. On the unaffected side, the planned depth of resection required for the thinnest tibial component was measured, with a minimum of 9 mm for the ATTUNE Knee System (Johnson and Johnson), including bone and cartilage. Then, cartilage from the unaffected proximal tibial condyle was removed, and the thickness of bone to be resected (9 mm minus the cartilage thickness) was remeasured and matched to the preoperative plan. Similarly, the thickness of bone to be resected on the affected side was measured and matched to that in the preoperative plan. If there was any discrepancy, then the jig was adjusted accordingly. The thickness of the resected bone was measured and matched to that in the preoperative plan and confirmed on postoperative LLRs (Fig. 3c, d & e).
The procedures were performed under combined spinal-epidural or spinal anesthesia. General anesthesia was given when spinal and epidural anesthesia were contraindicated. All procedures were performed using a tourniquet except in patients with heavily calcified arteries in the lower limbs. The ATTUNE Knee System (J&J) was implanted using parapatellar arthrotomy. Routine perioperative antibiotics and thromboprophylaxis (enoxaparin) were administered for 48 h and 5 days, respectively.
Outcome measures
All demographic data and outcome variables were measured and recorded prospectively by KCP. The radiological variables (outcome variables) recorded were the pre- and postoperative mechanical axis (HKA axis), mLDFA, and MPTA. Other radiographic measures recorded were the pVCA and eVCA.
Statistical analysis
The data were analyzed using SPSS statistical software (version 21.0, IBM Corp., Armonk, N.Y., USA). Determination of the change in the outcome parameters between preoperatively and postoperatively was carried out using a paired -sample t-test. Correlations between the HKA axis, mLDFA, and MPTA were assessed using Pearson’s correlation coefficient. The intraobserver and interobserver variability were also evaluated. Fischer’s exact test was applied to compare the percentage of outliers in the radiological parameters between groups stratified by age, sex, side, body mass index (BMI), and type and severity of deformity. P values ≤0.05 were considered to indicate statistical significance.