A comparative study between two commonly used robotic systems on the accuracy of the intraoperative cuts in total knee arthroplasty.

Abstract

Purpose: We aimed to compare the accuracy of robotic Total Knee Arthroplasty (TKA) intraoperative cuts between Computed Tomography image-based MAKO and imageless ROSA.

Methods: We retrospectively analysed registry data for patients who underwent imaged-based MAKO and imageless ROSA robotic TKA. Two reviewers independently radiographically measured the coronal alignment of the implants in relation to the mechanical axis at 3 months post-operation. These were compared to the planned intraoperative robotic cuts.

Results: One hundred sixty-one patients underwent MAKO and 110 ROSA TKA. Intraclass correlation for radiographic readings was 0.88 to 0.96. When comparing the variability between intraoperative planned alignment and 3 months postoperative, MAKO femur showed no statistically significant difference (- 0.1° ± 0.8° vs. - 0.1° ± 1.2°, p = 0.12) and likewise no difference for tibia (- 0.9° ± 1.2° vs. - 0.6° ± 1.3°, p = 0.60). However, for ROSA, its femur showed statistically significant difference (- 1.2° ± 1.0° vs. - 0.7° ± 1.3°, p < 0.01), likewise for tibia (- 0.4° ± 0.9° vs. - 0.2° ± 1.4°, p < 0.01). The variability from intraoperation to 3 months postoperative was significantly different between MAKO and ROSA for femur (0.0° ± 1.0° vs. 0.4° ± 1.1°, p < 0.01), but was not statistically significant for tibia (0.3° ± 0.1° and 0.1° ± 1.1°, p = 0.26). There was a significant difference with a larger proportion of femur implant that deviated < 1º from intraoperative plan in MAKO compared to ROSA (62.1% and 37.2% respectively, p < 0.01). At the tibia component, there was no statistically significant difference in the proportion of the extent of deviation (p = 0.73).

Conclusion: There was statistically significant variability in coronal alignment between intraoperative plan and postoperative 3 months in the imageless robotic system at both femur and tibia components. The extent of variability between the imageless and CT-image based systems in femur was also statistically significant. There was also a larger proportion of implant deviation of > 1º in the femur component in imageless robotic system. Nevertheless, the overall extent of the deviation between the intraoperative plan and the postoperative position was < 0.5º. Future studies may be performed to determine whether this would translate into any clinical significance.