The most important finding of this study was that the contact stress on the PF joint increased less with the MB UKA design than with the FB UKA design; however, no significant difference in contact stress on the PF joint was found between the medial UKA and intact knee joints. The quadriceps force needed to produce the same flexion angle with the MB UKA design was lower than that with the FB UKA design. UKA can be performed with either an FB or MB design. In a prospective study that involved 48 patients, who were randomly assigned to either FB or MB UKA prostheses, Li et al. observed better knee kinematics and a lower incidence of radiolucency with the MB design, but the Knee Society, WOMAC, and SF-36 scores were equivalent between the two designs (27). In another study, the range of motion, limb alignment, patient-reported outcomes, incidence of aseptic loosening, and reoperation rate were identical between the FB and MB UKA designs (28). However, the time to reoperation and failure mode differed. Early failure due to bearing dislocation occurred with the MB design, whereas late failure due to polyethylene wear occurred with the FB design. A previous study indicated that during a ≥15-year follow-up period, some type of revision arthroplasty was required for 12 (15%) of 77 knees in the case of FB UKA (Miller-Galante; Zimmer) and for 10 (12%) of 79 knees in the case of MB UKA (Oxford; Biomet) (29). No significant differences were observed in the number of knees with progressive lateral OA that required revision arthroplasty between the FB and MB UKA designs [33]. Thus, many arguments have emerged regarding the biomechanical issues of the FB and MB UKA designs. In previous studies on the progression of OA after UKA, the radiological assessment was neither blinded nor randomized (27).
The advantage of FE analysis is that the impact of the UKA design can be determined without external variables [34]. Most in vitro biomechanical studies have involved evaluations using aged cadaveric subjects with loosening between the specimen and the device, and tissue attenuation, which can occur owing to the successive loading in mechanical testing (26). An intact joint model was the foundation of this study and involved FEM validation steps. The results exhibited good agreement with those of previous computational studies (23, 30). Therefore, the UKA models used in the present study and related analyses are considered reliable.
Kozinn and Scott proposed that UKA should not be offered to patients with PF joint arthritis for optimal results (31). This sparked a contentious debate on PF joint disease because other authors demonstrated only a weak correlation between PF degenerative changes and anterior knee pain (11, 32).
In addition, owing to the differences in the design and biomechanics of the FB UKA model, damage to the PF joint has traditionally been a contraindication. Lim et al. recently showed that the presence of significant preoperative radiological PF disease does not affect long-term implant success, and patients had excellent postoperative functional outcomes for 10 years (33). In the present study, the MB UKA design produced a smaller increase in contact stress on the PF joint than the intact joint and FB UKA design models. The results of previous studies on MB UKA indicated that the presence of PF degeneration does not compromise clinical outcomes because the implant is believed to be more patella friendly owing to better kinematics, which supports our results (33, 34). Moreover, although the contact stress on the PF joint increased with both the FB and MB UKA designs, the increase was not statistically significant. Biomechanical studies have indicated that the progression of arthritis of the PF joint typically does not necessitate revision.
The quadriceps force needed to produce a squatting motion was greater for the FB design than for the MB design (by as much as 120 N for knee flexion angles >100°). Thus, increased quadriceps strength leads to improved functional performance (35). As patients who have undergone OA and knee arthroplasty experience significant quadriceps weakness, the FB UKA design, which increases the required quadriceps force, can result in more difficulty for patients to walk, kneel, or perform a deep knee bend (36). This agrees with the results of a previous in vitro study in which a UKA model required less quadriceps force at a mid-flexion angle than an intact model (27).
From a biomechanical viewpoint, our results indicate that the risk of progressive OA of the PF joint can be reduced with the MB UKA design because it preserves the normal biomechanical effect, in contrast to the FB UKA design. In addition, the MB UKA design requires a lower quadriceps force and makes it easier for recipients to kneel, squat, or rise from a chair.
The three strengths of our study should be highlighted. First, a well-validated setup that accounted for numerous previous results was used. Second, in contrast to previous UKA studies, the present study included the tibia, femur, and related soft tissues in the FE model. Third, in contrast to the current biomechanical UKA models, the model used in this study included the deep-knee-bend and squat loading conditions, rather than the simple vertical static loading condition.
Despite these strengths, this study had certain limitations. First, the results did not predict clinical results or patient satisfaction. Second, the computational model was developed using data from four male subjects and one female subject. Using data from subjects of various ages would improve the validity of the results, as it would increase the diversity of the knee joint geometry. However, in this study, our objective was to evaluate the biomechanical effect of UKA in young individuals. Third, the bony structures were assumed to be rigid. In reality, the bone is composed of cortical and cancellous tissues. However, the main purpose of the study was not to evaluate the effects of different prostheses on bone. In addition, this assumption had a minimal influence on the results of the study because the bone stiffness exceeds that of the relevant soft tissues (23). Finally, the simulation only involved a deep knee bend; thus, simulations involving rising from or sitting on chairs, climbing/descending stairs, and squatting should be performed in future investigations.
In conclusion, this study provides biomechanical evidence that degenerative changes in the PF joint should not be considered an absolute contraindication to treatment with medial UKA. In addition, UKA is not problematic even in PF joints with OA, unless accompanied by anterior knee pain, because no significant difference in contact stress was observed.