This study was approved by the review board of our institution. All patients provided written informed consent.
Between November 2017 and March 2021, an image-free navigation system (Precision N; Stryker Orthopedics, Mahwah, NJ, USA was used to perform 75 BCR TKAs (Journey II XR; Smith and Nephew, Memphis, TN, USA) at our institution. The surgical indications for BCR TKAs were knee OA or osteonecrosis (ON) of more than two compartments, intact cruciate and collateral ligaments, preoperative flexion contracture < 15°, and preoperative deformity < 15°. In this study, 59 BCR TKAs met the following inclusion criteria: satisfactory radiographs of the knee in the flexion position in accordance with the method of Kanekasu et al. [6] at 2 weeks postoperatively and at 2 years postoperatively, complete data entry, and adequate follow-up ≥ 2 years.
The patient population comprised 44 women and 15 men (mean age, 71.6 ± 8.2 years; mean body mass index, 24.9 ± 3.2 kg/m2; mean preoperative hip–knee–ankle [HKA] angle, 173.7° ± 5.9°).
Preoperative patient demographics, including age, sex, weight, height, body mass index, HKA angle, and range of motion (ROM), were recorded. Preoperative clinical scores were obtained using the validated version of the knee injury and osteoarthritis outcome score (KOOS) [7, 8].
All procedures were performed by five knee surgeons and a senior surgeon (H.I.) participated in all procedures either as the chief surgeon or first assistant.
Surgical procedure
In all patients, a paramedian approach was used, and the patella was not everted. The surgeon performed aggressive removal of osteophytes and minimal release of medial soft tissues for bone resection. Thereafter, an image-free navigation system was used to perform osteotomy of the femur and tibia. Between November 2017 to March 2019, our target alignment was mechanical alignment (MA), while between March 2019 to March 2021, our target alignment was functional alignment (FA) [9]. The patella was resurfaced for all patients.
Medial and lateral joint component gap evaluation
The joint component gap was measured using a femoral trial implant and force-controlled, compartment-specific ligament tensioner [9, 10], with a distraction force of 80 N for both the medial and lateral compartments at extension and at knee flexion of 45° and 90° (Fig. 3). The patellofemoral joint was reduced during gap measurements. The surgeon performed the measurements twice, and the first assistant surgeon performed them once in 20 randomly selected knees. The intra- and inter-rater reliability values of this evaluation were 0.978 and 0.968, respectively.
Postoperative rehabilitation
The same rehabilitation protocols were applied in all patients. ROM exercise and walking exercise with a crutch and then a walker were started on the first postoperative day. At 2–3 weeks postoperatively, the patient was discharged from our hospital and completed their rehabilitation protocol with physiotherapists [11].
Postoperative evaluation
Radiographs of the knee in the flexion position in accordance with the method of Kanekasu et al. were used to assess the tibial eminence [6]. The radiographs of the tibial eminence taken 2 weeks after the operation were compared with those at the final follow-up (average, 3.2 years; range, 2–5 years). A chronological change ≥ 2 mm was defined as positive (P group) and a change of < 2 mm was defined as negative (N group) in this study. Full-length standing radiography at postoperative 6-month follow-up was performed to measure the HKA. Regarding clinical evaluation, postoperative ROM, anterior–posterior instability, and the KOOS were evaluated at 1 year postoperatively and at the final follow-up.
Statistical analysis
Data were analyzed using the Bell Curve 2016 (SSRI Co., Ltd., Tokyo, Japan) software package for Microsoft Windows. The Mann–Whitney U-test was used to compare the quantitative variables and differences between two groups. Chi-square exact probability test was used to compare the ratio of sex and surgical procedure (MA or FA) between the two groups. All significance tests were two-tailed, and a significance level of P < 0.05 was used for all tests. The power analysis was performed using G*Power 3 (Heinrich Heine Universitat Dusseldorf, FRG) [12]. A post hoc power analysis for joint component gaps was performed, and the power calculated as 0.68.