Modification of the coronal plane alignment of the knee classification considering the joint line convergence angle in high tibial osteotomy: a geometric approach.

BACKGROUND

Excessive joint line obliquity (JLO) is associated with poor outcomes in open wedge high tibial osteotomy (OWHTO). The Coronal Plane Alignment of the Knee (CPAK) classification categorizes knee phenotypes by alignment and JLO but overlooks the joint line convergence angle (JLCA), which impacts JLO pre- and postoperatively in OWHTO. This study aims to modify JLO calculation by incorporating JLCA through a geometric approach, with the goal of better reflecting the knee joint orientation under weight-bearing conditions.

METHODS

323 knees that received OWHTO were retrospectively reviewed. The study included OWHTO patients who had pre- and 1-year postoperative long-standing hip-to-ankle radiographs for CPAK classification and JLCA measurement. JLOs were calculated using two methods based on a geometric approach: conventional JLO (cJLO), defined as (lateral distal femoral angle [LDFA] + medial proximal tibial angle [MPTA])/2-90°, and modified JLO (mJLO), which incorporates the JLCA, defined as (LDFA + MPTA)/2 + JLCA/2° - 90°. These were compared to the knee joint orientation using intraclass correlation coefficients (ICCs) and Bland-Altman plots. Clinical outcomes were evaluated using the Hospital for Special Surgery (HSS) score.

RESULTS

Mechanical hip-knee-ankle angle (mHKA) was corrected from - 7.1 ± 2.7° preoperatively to 2.1 ± 2.3° at 1 year postoperatively, and JLCA decreased from 3.1 ± 1.9° to 2.4 ± 1.7°. The HSS scores showed significant improvement, rising from 68.6 ± 11.7 preoperatively to 91.3 ± 7.4 postoperatively (p < 0.001). The knee joint orientation was measured at - 1.0 ± 1.9° preoperatively and 1.9 ± 2.3° postoperatively at 1 year. Comparatively, cJLO values were - 3.3 ± 1.5° and 0.9 ± 2.0°, while mJLO values were - 1.8 ± 1.8° and 2.1 ± 2.3°. ICCs showed preoperative and 1-year postoperative cJLO at 0.842 (95% CI, 0.804-0.873) and 0.927 (95% CI, 0.909-0.941), respectively, while mJLO demonstrated 0.881 (95% CI, 0.851-0.904) and 0.949 (95% CI, 0.936-0.959). Bland-Altman plots showed that both preoperative and 1-year postoperative mJLO had smaller differences from the knee joint orientation than cJLO, indicating that mJLO more closely reflects functional joint orientation.

CONCLUSION

The mJLO, derived through a geometric approach that incorporates JLCA, was found to more closely reflect the knee joint orientation compared to the cJLO described in the CPAK classification. To effectively apply the CPAK classification in OWHTO, it is essential to account for JLCAs before and after surgery. Accordingly, mJLO can be represented as (LDFA + MPTA)/2 + JLCA/2° - 90°.