Treatment of both simple and complex patella fractures is a challenging clinical problem. The aim of this study was to investigate the biomechanical performance of recently developed lateral rim variable angle locking plates versus tension band wiring used for fixation of simple and complex patella fractures.

Twelve pairs of human anatomical knees were used to simulate either two-part transverse simple AO/OTA 34C1 or five-part complex AO/OTA 34C3 patella fractures by means of osteotomies, with each fracture model created in six pairs. The complex fracture pattern was characterized by a medial and a lateral proximal fragment, together with an inferomedial, an inferolateral, and an inferior fragment mimicking comminution around the distal patellar pole. The specimens with simple fractures were pairwise assigned for fixation with either tension band wiring through two parallel cannulated screws, or a lateral rim variable angle locking plate. The knees with complex fractures were pairwise treated with either tension band wiring through two parallel cannulated screws plus circumferential cerclage wiring, or a lateral rim variable angle locking plate.

Each specimen was tested over 5000 cycles by pulling on the quadriceps tendon, simulating active knee extension and passive knee flexion within the range of 90° flexion to full knee extension. Interfragmentary movements were captured via motion tracking.

For both fracture types, the longitudinal and shear articular displacements measured between the proximal and distal fragments at the central patella aspect between 1000 and 5000 cycles, together with the relative rotations of these fragments around the mediolateral axis were all significantly smaller following the lateral rim variable angle locked plating compared with tension band wiring, p<0.01.

Lateral rim locked plating of both simple and complex patella fractures provides superior construct stability versus tension band wiring under dynamic loading.

In spite of extensive accounts describing the blood supply to the femoral head, the prediction of avascular necrosis is elusive. Current opinion emphasises the contributions of the superior retinacular artery but may not explain the clinical outcome in many situations, including intramedullary nailing of the femur and resurfacing of the hip. We considered that significant additional contribution to the vascularity of the femoral head may exist. A total of 14 fresh-frozen hips were dissected and the medial circumflex femoral artery was cannulated in the femoral triangle. On the test side, this vessel was ligated, with the femoral head receiving its blood supply from the inferior vincular artery alone. Gadolinium contrast-enhanced MRI was then performed simultaneously on both control and test specimens. Polyurethane was injected, and gross dissection of the specimens was performed to confirm the extraosseous anatomy and the injection of contrast. The inferior vincular artery was found in every specimen and had a significant contribution to the vascularity of the femoral head. The head was divided into four quadrants: medial (0), superior (1), lateral (2) and inferior (3). In our study specimens the inferior vincular artery contributed a mean of 56% (25% to 90%) of blood flow in quadrant 0, 34% (14% to 80%) of quadrant 1, 37% (18% to 48%) of quadrant 2 and 68% (20% to 98%) in quadrant 3. Extensive intra-osseous anastomoses existed between the superior retinacular arteries, the inferior vincular artery and the subfoveal plexus.