Summary

Failure to accurately restore the trochlea anatomy in total knee arthroplasty leads to poorer functional outcomes following total knee arthroplasty.

Abstract

The variation in recreation of the trochlear when a standardised implant design is applied to variable patient anatomy and its impact on patient outcomes is not fully understood. The aim of this study was to analyse if a relationship existed between recreation of the trochlear and patient outcomes following total knee arthroplasty (TKA).

Methods

This was a retrospective cohort study of consecutive patients presenting for primary TKA who had pre and post-operative CT knee scans performed.

Patients with minimum 2-year follow-up were included for analysis. Patient reported outcomes (PROMs) were evaluated.

3D implant and bone models from the preoperative CT scans were registered to the post-operative CT scan. Cross-sectional slices at increments of 10° from 0° to 30° of flexion on the femoral component were used to compare offset differences between the pre-operative bone and post-operative implant position at the medial and lateral peaks and the trough of the trochlear groove . Trochlea offset wasgraded into 5 groups(<-2.5mm under-stuffing, -2.5 - 0.mm under-stuffing, 0mm - 2.5mm over-stuffing, 2.5 - 5.0mm over-stuffing and >5.0mm over-stuffing). Outcomes were compared between grades.

Results

Following exclusions and loss to follow-up a total of 232 were analysed. Mean BMI was 29+/- 5, mean age was 74+/- 7 and mean follow-up time was 41+/- 13 months. 12.5% of patients were dissatisfied.

Implant-bone offset differences were higher medial to lateral and superior (0° to 30°) to inferior at all 3 points. At the trochlea apex (°0 flexion), mean implant-bone offset differences were 5.12mm (medial condyle apex), 1.43mm (sulcus apex) and 1.42mm (upper lateral condyle apex). The mean offset differences at 30° flexion were -0.01mm medial condylar apex, -1.99mm sulcus base and -3.11mm lateral condylar apex. Pearson’s correlation indicated a significant correlation between medialisation of the trochlea with KOOS pain scores ( -1.62 p = 0.030) and the mean lateral apex offset difference and several PROM scores (FJS -.158 p = 0.016, KOOS pain -.190 p = 0.004, Kujala -.176 (p = 0.007) and KSS satisfaction -.132 (p=0.045). Grade 5 (n= 20) lateral condyle offset (mean 6.2mm of over-stuffing) was significantly correlated with a lower FJS, Kujala and KOOS scores .


Patients below the KOOS patient acceptable symptom state (PASS) score (<84.5) had a groove medialized by a mean of 2.4 +/- 1.0 mm compared to those with better pain scores had groove medialized less, 1.8 +/- 1.0mm (p = 0.012).

Analysis of implant position and patient characteristics between grades of lateral apex offset grades found no statistically significant differences between groups for all 3 planes of femoral tibial component positioning or tibio-femoral rotation, BMI, age, gender, time from surgery, pre or post-operative alignment or the size of coronal or sagittal plane deformity correction.

Conclusion

Patient outcomes 2 years after TKA correlated with trochlea recreation. The effect of medializing the trochlea groove and increasing the height of the lateral trochlea apex by more than 5mm from the native bone was associated with significantly poorer outcomes. These findings have implications for implant design and alignment philosophy.