The pre-diseased coronal alignment can be predicted from conventional radiographs taken of the varus arthritic knee.

Abstract

The concept of restoring the constitutional, pre-diseased alignment has gained lots of interest among knee surgeons. Previous attempts to use the contralateral limb to bridge the gap between the arthritic and the constitutional alignment were unsuccessful. We investigated the usability of a mathematical formula to predict the constitutional (pre-diseased) coronal alignment once arthritis has occurred. It is our hypothesis that by using the KL grade, CPAK classification and four radiographic measurements of the arthritic knee, the pre-diseased coronal alignment could be predicted. Hundred arthritic patients with consecutive X-rays were used to determine a mathematical formula. Five alignment parameters were determined on full-length X-rays: HKA angle, mechanical lateral distal femoral angle (mLDFA), medial proximal tibial angle (MPTA), joint line convergence angle (JLCA) and the tibial joint line angle (TJLA). A refinement of the algorithm was implemented based on a reference cohort, consisted of 250 young adults aged between 20 and 27years. Moreover, all knees were subdivided based on their CPAK-phenotype and the Kellgren-Lawrence scale (KL scale). An independent arthritic cohort of 289 patients scheduled for primary total knee arthroplasty was included to verify the accuracy of the predicted HKAs (HKAPRED). In CPAK type 1, the HKAPRED was 3.86° varus (STD 1.39) and the HKAYHA was 4.0° varus. In CPAK type 2, the HKAPRED was 1.68° varus (STD 1.95) compared to a HKAYHA of 1.34° (STD 0.81). The average constitutional HKA is not different in both CPAK 1 (p = 0.61) and CPAK 2 (p = 0.25), and the difference in the mean is estimated to be equal to - 0.14 (95CI - 0.68 to 0.40) in CPAK 1 and 0.35 (95CI - 0.06 to 0.75) in CPAK 2. Using the KL grade, CPAK classification and four radiographic measurements of the arthritic knee, the pre-diseased coronal alignment can be predicted in 80% of the varus knees with an accuracy of ≤ 0.5°. The predicted HKA (HKAPRED) can be very useful in the current and future clinical practice.