Page 343 - Read Online

Maalouly et al. Mini-invasive Surg 2021;5:35 https://dx.doi.org/10.20517/2574-1225.2021.57 Page 9 of 12

[4]
improve the accuracy of screw placement. Huntsman et al. also reported a similar high accuracy of 99%
using the ExcelsiusGPS system in a study of 100 cases.

The single position lateral surgery has not gained widespread approval due to technical difficulty of
cannulating a pedicle screw in the lateral position. Recent studies have reported a reduction in operative
time in single position lateral pedicle screw fixation following lateral interbody fusion as compared to dual
positioning without an increase in complication rates or compromised perioperative outcomes [18-20] . The
authors believe the use of the rigid robot arm of the ExcelsiusGPS system will provide a stable and accurate
insertion of pedicle screws in the lateral decubitus position. The accuracy of screw placement in the subset
of patients operated in a single lateral position was 97.87%, where two screws were revised out of 94 pedicle
[4]
screws. A similar high accuracy of 98% was reported by Huntsman et al. using the ExcelsiusGPS robot
system. Surgeons operating in a single lateral position surgery face the technical challenge of cannulating
pedicle screws in this position and some technique-related complications; however, recent studies showed
comparable results to the flip position [18-20] . Furthermore, as the placement of S2AI screws can be challenging
by the freehand technique, the robot can be used to place these screws [21,22] .

The learning curve for accurate planning and execution of screw placement was reached in Group C in this
study. The time taken to plan screw trajectories saw a significant reduction in the initial period of this study.
The mean time taken to plan in Group A was 5.35 min and in Group B was 3.15 min. The robot arm could
not reach planned trajectories in six instances in Group A and once in Group B. The authors believe this
reduction in missed trajectories was due to improved planning and increasing familiarity of the 3D
interface. However, the planning time increased in the following case groups as the planning interface was
utilized as an academic tool to train fellows and visiting surgeons.

Similarly, this study demonstrates a gradual reduction in screw insertion time which is comparable to other
studies. Urakov et al. studied the mean insertion times for trainee surgeons using the older generation
[23]
robotic guidance system and reported a mean time of 5.7 min per screw for percutaneous screw placement
and 3.6 min per screw for freehand insertion. In our study, the initial case (Groups A and B) showed a
declining trend in screw insertion time from 5.23 min per screw to 4.36 min per screw. The least amount of
average time per screw was reached around the Group C mark with a mean time of 3.78 min per screw.

Radiation exposure in image-guided surgery is a significant concern to the surgical team. The use of
robotic-assisted navigation has been postulated to reduce radiation exposure as compared to traditional
fluoroscopy-based surgery. A recent study reported significantly lower radiation times in the robot-assisted
[24]
group than in the fluoroscopy-guided group . However, the cumulative radiation exposure reported
during surgery in the robot-assisted group was 93.5 ± 37.9 s, which was significantly higher than what was
[25]
found in our study (16.32 ± 13.22 s). Similarly, Pennington et al. conducted a systematic review which
found that robot-assisted surgery with preoperative CT imaging had significantly less radiation exposure
than in fluoroscopy-based surgery. The researchers estimated a dose of 14-16 mSv in fluoroscopy-navigated
surgeries as compared to 4.8 mSv exposure dose of intraoperative O-arm scan for CT navigation-based
surgeries. This estimation, although less than the one encountered in this study (8.2 ± 3.74 mSv), can be
explained by the need for a high-dose wide-view CT scan for validation of the ExcelsiusGPS system. Even
though the CT radiation dose in this study was higher than that of the O-arm scan utilized in navigation-
only systems (Costa et al. ), it is significantly lower than that encountered in a fluoroscopy-based
[26]
procedure. The mean fluoroscopy time reported by these researchers was 20.1 ± 17.2 s per screw, which was
[2]
significantly higher than the one reported in this study (16.32 ± 13.22 s per case). Benech et al. , in a study
on the initial experience and radiation exposure using a similar robotic guidance system, reported a