Summary

In patients with preoperative varus deformity of ≧10˚, estimated HKA (eHKA) was a simple indicator that could be calculated with full-length weightbearing radiograph, and the cutoff value of eHKA that could be corrected to postoperative HKA ≧-7˚ was -7.4˚.

Abstract

Introduction

According to the classical indication proposed by Kozinn et al, preoperative varus deformity of ≦10˚ is considered an indication for medial unicompartmental knee arthroplasty (UKA). However, the acceptable range of postoperative alignment recommended for medial UKA has recently been expanded, and if the postoperative hip-knee-ankle angle (HKA) can be corrected to less than 7˚ of varus alignment, then patients with large preoperative varus deformity are also eligible for UKA. However, there are few reports examining factors that can predict the postoperative alignment of UKA, and it is difficult to determine if preoperative correction is possible in patients with large varus deformity. In this study, we investigated factors that predict postoperative alignment in knees with preoperative varus deformity of ≧10˚.

Subjects and Methods
Among medial UKAs performed at our institution from 2009 to 2014, 110 knees with preoperative varus deformity of ≧10˚ were retrospectively investigated (male/female: 23/87, mean age: 74.8±6.8). The surgical goal was to establish a relative undercorrection within the range of 1˚-7˚ of varus, in order to avoid degenerative progression on the lateral compartment. Valgus stress radiography and MRI were performed preoperatively. Full-length weightbearing radiograph was obtained pre-and postoperatively. To investigate the associations, HKA, medial proximal tibial angle (MPTA), lateral distal femoral angle (LDFA), and joint line convergence angle (JLCA) were measured, and also the osteophyte area at the medial femur and tibia were evaluated . The arithmetic HKA (aHKA; MPTA - LDFA) and estimated HKA (eHKA; preoperative HKA + JLCA) were calculated according to MacDessi et al and Kleeblad et al, respectively. HKA was defined as negative for varus and positive for valgus. JLCA was defined as positive for a medial JLCA convergence and negative for a lateral JLCA convergence. From these parameters, factors influencing postoperative HKA were identified by correlation analysis, multiple regression analysis was performed using the stepwise method to construct a prediction model for postoperative HKA, and cutoff values for the predictors were calculated by ROC analysis.

Results

When the correction target was postoperative HKA ≧-7˚, 71 of 110 knees (64.5%) could be corrected. The postoperative HKA was significantly correlated with preoperative HKA (r=0.336, p<0.001), LDFA (r=-0.319, p<0.001), aHKA (r=0.264, p=0.005), and eHKA (r=0.466, p<0.001). The prediction model for postoperative HKA was obtained by multiple regression analysis: postoperative HKA = -1.92 + 0.5447 × eHKA (adjusted R2 = 0.21, p<0.001). The eHKA was the only predictor extracted, and the cutoff value of eHKA that could be corrected to postoperative HKA ≧-7˚ was -7.4˚ (area under the curve: 0.701, 95% CI 0.602 to 0.8).

Conclusion

In patients with preoperative varus deformity of ≧10˚, eHKA was a simple indicator that could be calculated with full-length weightbearing radiograph, and the cutoff value of eHKA that could be corrected to postoperative HKA ≧-7˚ was -7.4˚.