Introduction

Tibial tubercle osteotomy (TTO) is a commonly used surgical treatment for patients with recurrent patellofemoral instability. The presence of significant patella alta is an indication for distalisation of the tibial tubercle and this necessitates a cut in the anterior cortex of the tibia. This causes a weakness in the tibia which may result in the rare complication of tibial shaft fracture during the perioperative period. A step cut is thought to cause the greatest stress riser. Our hypothesis was that a step cut at the distal end of a TTO would cause a greater stress riser than a bevelled cut at time zero.

Methods

Composite saw bones with cortical and cancellous components which are mechanically validated against human bones were used. Three types of osteotomy were performed using 3D-printed cutting jigs (step cut, bevelled cut and “V” cut; n = 4 per group) and compared to a control group with no osteotomy (n = 4). Tibia were stressed to the point of fracture using a 4-point bending technique in a 10 kiloNewton Instron machine. Mechanical data was then entered into an FEA model for further analysis.

Results

Mean differances between groups were as follows: control vs step cut 2935 N (95% CI 1113 – 4758; p = 0.0022), control vs V cut 3752 (95% CI 1929 – 5574, p = 0.0003), control vs taper cut 4050 N (95% CI 2228 – 5873; p = 0.0001), V cut vs step cut -816 (95% CI -2639 – 1006; p = 0.56), V cut vs taper cut 298 N (95% CI - 1524 – 2121; p = 0.96), step cut vs taper cut 1115 N (-707 – 2937; p = 0.31).

Conclusion

All types of TTO were associated with increased tibial shaft fracture risk at time zero (mean 39.6% reduction in tibia strength versus control group). Contrary to our hypothesis, TTO with distal step cut was not associated with an increased fracture risk. Consequently, bevelled and “v-shaped” cuts were not associated with reduced fracture risk. All data is tested at time zero and does not account for duration of healing or fixation used in vivo. Despite using symmetrical lab-tested saw bone models there was variation in loads to failure in each group.