Fallacies of CT based component size prediction in total knee arthroplasty - Are patient specific instruments the answer?

Background: The purpose of the study was to assess the accuracy of tibial and femoral component size prediction using computerised tomography (CT) based patient specific instruments in total knee arthroplasty.

Methods: Eighty-eight knees in 58 patients underwent total knee arthroplasty (TKA) using CT based patient specific instruments between March 2015 to April 2016. All patients were assessed for the pre operative femoral and tibial component sizes predicted by the CT-based pre-operative plan. These sizes were compared with the actually implanted sizes during surgery, and the results were assessed. The data were evaluated using Wilcoxon signed rank tests, and p value set at <0.05 for significance.

Results: Approximately 72% predicted tibia sizes matched the final implanted sizes whereas 66% femoral implants matched their pre-operative predicted sizes. The difference in the tibial implant size was not statistically significant (p-value > 0.05). However, the difference in the femoral size was statistically significant (p-value 0.009). The downsizing of the tibial component was needed in 14.8% knees whereas upsizing was required in 13.6% of the knees. At the femoral side, 22.7% components required downsizing at the time of implantation as compared to 11.4% components wherein a bigger component was used.

Conclusions: We conclude that size prediction using CT-based technology for patient specific instrumentation is not fool proof. The size prediction accuracy for femoral and tibial components at 66% and 72% are low and cannot be relied upon at present. The patient specific technology using CT scan based jigs holds promise for the future, needs refining and fine tuning.