Letter: Placement of Stereotactic Electroencephalography Depth Electrodes Using the Stealth Autoguide Robotic System: Technical Methods and Initial Results

摘要

To the Editor: The referenced article by Tay et al1 describes a technique for implantation of depth electrodes for stereoelectroencephalography (sEEG) using the Stealth Autoguide system (Medtronic). This approach is a frameless technique, which uses a robotic arm for assisted placement of electrodes. Although we value the reporting of such surgical data, we believe there are significant safety and efficacy concerns with this study that warrant clear delineation to ensure the safety of patients and clarify the known scientific literature reported on this subject. First, we are obliged to draw attention to safety concerns with the reported method. Cardinale et al2 presented one of the earliest and largest studies regarding the safety associated with modern sEEG implantation, with a reported 0.7% symptomatic hemorrhage rate per procedure and 0.02% per electrode (none of which occurred with the use of robot assistance). A representative single-institutional experience reported a comparable 0.66% rate of clinically significant hemorrhage per procedure and 0.06% per electrode.3 Specific to robot-assisted implantation, González-Martínez et al4 cited a 1% hemorrhage rate per procedure and 0.08% per electrode. This study, by comparison, reports a symptomatic hemorrhage rate of 11.1% per procedure and corresponding 0.98% complication rate per electrode. Furthermore, it is important to clarify the author's discussion of their safety results. The authors report equivalency for safety based on their 0.98% hemorrhage rate and previous reports of 1% and 1.5% rates of hemorrhage.5,6 However, the cited reports from Mullin et al5 and Ollivier et al6 are per procedure hemorrhage rates while this study is the per electrode hemorrhage rate; by comparison, the per electrode hemorrhage rate as reported by Ollivier et al6 was 0.1%. As such, although the referenced studies again report hemorrhage rates comparable with the aforementioned publications, this study reports a hemorrhage rate that is an order of magnitude greater than previous sEEG studies. Beyond the safety measures, the authors note accuracy issues in their discussion of study limitations. Their reported accuracy—defined by the Euclidean target point (TP) error—for 77 trajectories among 9 patients was measured as 4.67 ± 0.27 mm, which is 2.5 to 3.0× worse than any other robot/frame-based technique reported in the known literature. In a large meta-analysis, Vakharia et al7 reported a mean TP error of 2.89 mm (2.34-3.44 95% CI) for frameless systems and mean TP error of 1.93 mm (1.05-2.81 95% CI) for frame-based systems, synthesized from 8 studies reporting Euclidean error and 5 radial error. With respect to robot-assisted methods, they cite a TP error of 1.71 mm (1.66-1.75 95% CI). Furthermore, a recent meta-analysis/meta-regression conducted by our own research group determined a Euclidean TP error of 2.34 mm (95% CI: 2.1-2.6) for sEEG as a whole.8 This study is not without merits, and it is not our intention to rebuke the efforts of the research team. We agree that innovation in stereotactic and functional neurosurgery is an admirable and worthwhile endeavor and are in full support of the open publication of studies reporting both the positive and negative conclusions. It is clear that this study is victim to its relatively small sample size, and previous work has noted the disproportionate rate of complications in association with smaller studies.5 Our concerns are specifically regarding the interpretation of the reported data and the consequent conclusions. In the context of the accuracy and safety concerns, we present in this letter that we believe the described technique is neither “equivalent” to existing methods nor “safe” on the basis of the limited data they present.