Periprosthetic tibial fracture in robotic-assisted versus conventional medial UKA: in vitro testing and finite element analysis of baseplate footprint

Background Periprosthetic tibial fracture after unicompartmental knee arthroplasty (UKA) is uncommon but clinically important. Although robotic-assisted UKA may improve the precision of bone preparation and component positioning, whether these technical differences influence fracture resistance or fracture morphology remains unclear. This in vitro study examined whether robotic-assisted medial UKA alters axial load-to-failure compared with conventional UKA, whether fracture morphology differs between the two techniques, and whether implant base footprint influences fracture behavior in complementary finite element (FE) simulations. Hypothesis Robotic-assisted medial UKA would alter tibial fracture behavior compared with conventional UKA, as reflected by differences in fracture resistance and fracture morphology. Methods Fourteen fresh-frozen cadaveric specimens underwent medial UKA using either conventional instrumentation (n = 7) or a robotic-assisted workflow (n = 7). After implantation, the tibiae were isolated and subjected to monotonic axial compression at 0.5 mm/s until macroscopic periprosthetic fracture occurred. Load-to-failure, failure load normalized to body weight, and fracture morphology were evaluated. Complementary FE models incorporating implant baseplate sizes 8 and 10 were used to compare load-displacement behavior and crack propagation. Results Axial load-to-failure did not differ significantly between conventional and robotic-assisted UKA (3712.4 ± 1746.2 N vs. 3988.3 ± 1534.5 N, p = 0.90). Failure load normalized to body weight was also not significantly different between groups (524.42 ± 262.60% vs. 647.24 ± 212.54%, p = 0.71). Different predominant fracture patterns were observed despite similar fracture resistance. Robotic-assisted specimens primarily demonstrated metaphyseal depression-type fractures, whereas fractures in conventional specimens more frequently propagated toward the meta-diaphyseal region. In the FE simulations, peak loads were 3755.36 N for the size-8 baseplate model and 4005.36 N for the size-10 baseplate model, both within the variability observed experimentally, with differences in crack propagation between implant baseplate sizes. Discussion Robotic-assisted medial UKA did not significantly alter axial fracture resistance compared with conventional UKA, although different predominant fracture morphologies were observed between techniques. Complementary FE analysis showed that implant base footprint influenced simulated load-displacement behavior and crack propagation, suggesting that implant footprint may contribute mechanically to differences in fracture behavior after medial UKA. Level of evidence V; biomechanical cadaveric and in silico investigation.