Osteotomy is a surgical technique used in the treatment of patients with compartmental osteoarthritis. This surgery attempts to change the stress distribution in the knee to unload the affected compartment of the knee and place the joint load primarily on the unaffected compartment. While the efficacy of this technique has been well documented, predictions of the post-surgical alignment have been shown to display a high degree of variability. This variability can be attributed primarily to changes in joint surface alignment following surgery. This thesis attempts to identify the factors that are most influential in estimating the change in joint alignment following osteotomy surgery.

The hypothesis was that soft tissue structures of the knee are the primary controls of the joint surface alignment. The objectives were to obtain quantifiable measures of joint stiffness (using the MEDmetric KT1000® and a custom built knee analyzer), clinical measures of pain and function (using the WOMAC osteoarthritis index), lower limb alignment and geometry (through radiographic analysis), and measure of gait performance (using principal component analysis).

The principal component models were developed for six knee kinematic and kinetic gait waveforms using the waveforms of 58 normal subjects collected by Kevin Deluzio. These models were used to characterize the normality of a small cohort of 8 patients scheduled to undergo osteotomy surgery. Waveforms were analyzed over the entire gait cycle for knee angles and over the stance phase of the gait cycle for knee moments. O f these 8 patients, 5 patients were tested following surgery to ascertain which of the pre-operative measures displayed the greatest predictive power in determining the post-surgical alignment.

It was determined that measures of knee joint initial stiffness in the valgus direction, as examined pre-operatively, were highly correlated with the change in the frontal alignment of the joint bony surfaces following surgery (r²=0.923 p<0.05). This correlation indicates that knees with low initial stiffness are more susceptible to changes in relative joint alignment following the redistribution of stress imposed by the surgery. In addition, measures of patient questionnaires (WOMAC) and dynamic frontal alignment of the knee, as measured by the first principal component scores of the adduction/abduction knee angle waveforms showed a similarly strong correlation with the joint surface realignment following surgery (r²=0.923 p<0.01 and r²=0.927 p<0.01 respectively). These findings further indicate that patients with large varus or valgus deformities as measured during gait and patients with high incidence of pain are more likely to be affected by change in the joint surface alignment following surgery.

It was further noted in the course of analysis that the first principal component scores of knee adduction/abduction moment and internal/external rotation moment were correlated among both the normal and patient groups (r²=0.665 p<0.01), indicating that these moments may be coupled.