IN-VIVO DETERMINATION OF PATELLOFEMORAL SEPARATION AND LINEAR IMPULSEFORCES

Problems associated with the patellofemoral joint account for nearly h alf of all total knee arthroplasty (TKA) revisions. Under in vivo cond itions, we previously determined that TKA subjects experience patellof emoral separation while performing dynamic, weightbearing activities. This study investigates the impulse loading conditions that may exist at the time the patella impacts the femur during knee flexion. Fifty-s even subjects (68 knees) performed three successive deep knee bends un der fluoroscopic surveillance. Eleven subjects (14 knees) had a poster ior cruciate retaining (PCR) TKA, 19 subjects (25 knees) had a posteri or cruciate substituting (PS) TKA, 15 subjects(17 knees) had a normal knee, and 12 subjects (12 knees) had an anterior cruciate ligament def icient (ACLD) knee. Velocities of each subjects' patella relative to a fixed point on the tibia were used as input to a mathematical model i ncorporating the impulse-momentum equation. At full extension,12 of 14 PCR knees, 11 of 25 PS knees, 1 of 12 ACLD knees, and none of the 17 normal knees exhibited patellofemoral joint separation. The maximum se paration, detected in a PCR knee, was 12 mm. The relative force determ ined upon patellofemoral impact was minimal (1.0 N). Simulated walking conditions for each subject were then entered into the mathematical m odel at a rate of 100 Hz and the calculated patellofemoral impact forc es ranged from 78 N to 213 N. Since impulse loading conditions occur o ver a very small period of time, it was concluded that capturing fluor oscopy images at a rate of 30 Hz was too slow. Under simulated walking conditions, the impact forces due to impulse loading could contribute to polyethylene failure if these conditions induce fatigue of the pol yethylene.