• 
  
• Rocking or torsion ot" the posterior stabilized femoral, component
  during removal may result in a fracture of the femur through the area of
  the jntercondylar notch (87).                          
  In a revision setting, where infection has been ruled out, it is better to leave remaining well-fixed cement in the canal rather than riskin<> excessive bone loss or perforating the canal trying to remove it (93).
  4-Reconstniction :
  When performing revisions, Brassard et al.. prefer to use a three-step method. The three steps are recreating the femur, rebuilding the flexion space, and rebuilding the extension space. The femur is prepared first because of the availability of more reliable anatomical landmarks (93).
  I-Recreate the femur:
  a-Size the femur:
  It is helpful preoperatively to procure the operative notes from the previous procedure. Another useful preoperative step is to template the opposite side in order to obtain a relative idea of the sizes (93).
  There usually posterior bone loss, so templating intraoperatively runs the risk of undersizing the federal component. The epicondylar width of the femur can also be helpful in selecting the appropriate size (93).
  b-Fcmoral component rotation:
  With 90 degree flexion, the rotation of the femoral component and the stability in flexion should be determined. Rotational landmarks are identified. In revision surgery, the posterior aspect of the femoral condyles and Whiteside lines are not available. The epicondylar axis may be used and is considered by many authors as the best landmark but it is