Abstract
Introduction
Knee arthroplasty is an effective intervention for painful arthritis when conservative measures have failed. Despite recent advances in component design and implantation techniques, a significant proportion of patients experience problems relating to the patella-femoral joint (PFJ).
Detailed knowledge of the shape and orientation of the normal and replaced femoral trochlea groove is critical when considering potential causes of anterior knee pain. Furthermore, to date it has proved difficult to establish a diagnosis due to shortcomings in current imaging techniques for obtaining satisfactory coronal plane motion data of the patella in the replaced knee.
The aim of this study was to correlate the trochlea shape of normal and replaced knees with corresponding coronal plane PFJ kinematic data.
Method
Bony and cartilagenous trochlea geometries from 3T MRI scans of 20 normal healthy volunteers were compared with both anatomical and standard total knee replacements (TKR) and patellofemoral joint replacement (PFJR) geometries. Following segmentation and standardized alignment, the path of the apex of the trochlea groove was measured using customized Matlab software. (Fig1).
Next, kinematic data of the 20 normal healthy volunteers (Normal) was compared with that of 20 TKR, and 20 PFJR patients using the validated MAUSTM system (Motion Analysis and UltraSound) comprising a 12-camera, motion capture system used to capture images of reflective markers mounted on subjects lower limbs and an ultrasound probe. A mapping between the ultrasound image and the motion capture system allows the ultrasound probe to be used to determine the locations of the patella relative to bony landmarks on the femur during a squat exercise.
Results
In normal knees the arc of the trochlear groove apex was orientated progressively laterally for both cartilage and. Neither of these trends were reproduced by any of the knee prostheses. Indeed far from being a laterally directed trochlea groove, both the anatomic TKR and PFJR have a medially orientated trochlea, whilst the TKR showed a neutral straight path (Figure 2). The direction of displacement in the replaced knee is significantly different (opposite) to that of the native knee (p<0.05).
The accuracy of the MAUS technique registering the ultrasound images within the motion capture system is 1.84 mm (2 × SD).
The three groups showed very different patella tracking patterns which matched the orientation of the underlying trochlea (Figure 3). The sine wave pattern of coronal plane patella motion displayed by the Normal group was not recreated in the TKR or PFJR groups. Movements of the Normal group were significantly different from the TKR group (p=0.03) and the PFJR group (p<0.01), whilst there was no significant difference between the TKR and PFJR groups (p=0.27).
Discussion
We present a new, accurate, reliable in vivo technique for measuring 3D patellofemoral kinematics in native and replaced knees. Our data suggest that many aspects of patellofemoral kinematics are absent following TKR and PFJR. This can be explained by the differences in shape of the underlying femoral component. Anterior knee pain problems might be addressed by alterations to the patellofemoral joint in future designs of knee arthroplasty.
