• BACKGROUND
    • The Coronal Plane Alignment of the Knee (CPAK) classification has been proposed to guide the alignment strategy in total knee arthroplasty (TKA). However, the classification does not take into account the variability of the soft-tissue envelope unique to each knee. The purpose of our study was to determine the variability and similarity of extension balance across each CPAK phenotype.
  • METHODS
    • Using the anatomic and ligament data acquired from 4,362 robotic TKAs, the distraction of the medial and lateral joint spaces was simulated and classified across CPAK phenotypes I to VI. The joint balance was calculated for both the arthritic and nonarthritic states, accounting for cartilage wear. The extension joint balance for each CPAK phenotype was compared to another to determine the degree of variability and overlap across the various CPAK classes. In addition, a sensitivity analysis was performed by widening the boundaries of CPAK II and V from the anatomic hip-knee-ankle axis up to five degrees.
  • RESULTS
    • Although there were differences in extension balance across the various CPAK phenotypes in the arthritic knee, there was also major variability and overlap. When accounting for cartilage loss, the similarities further increased. The greatest extension balance overlap was observed between CPAK II and IV (74%), whereas CPAK I and VI were expectedly dissimilar (6%). Accounting for cartilage wear, the similarity between CPAK II and IV and I and VI increased to 84 and 40%, respectively. Increasing the anatomic hip-knee-ankle boundaries for CPAK II and V to ± 3 degrees saw the greatest increase in reduction of extension balance overlap between neutrally aligned and varus knees.
  • CONCLUSIONS
    • The knee is highly complex and variable. While the CPAK classification provides a description of commonly encountered knee morphologies undergoing TKA in a single plane, it does not account for soft-tissue characteristics. Refinement of this classification system is required to improve its predictive value both intraoperatively and postoperatively.