Complications in stereoelectroencephalography: are we too severe?

Abstract

We read with interest the article recently published by Restrepo et al. on Acta Neurochirurgica, validating the use of a mobile cone-beam CT (CBCT) scanner (O-arm O2TM, Medtronic Ltd.) for image-guidance registration in stereoelectroencephalography (SEEG) implantation [9]. To compare the use of conventional fan-beam CT or CBCT for patient registration is the main aim of their study. The authors reported no significant difference between O-arm and CT groups neither in terms of application accuracy nor for procedural time. Differently, O-arm-based workflow guaranteed lower radiation dose. The authors reported 4 complications in 27 analyzed patients (14.8%), with no difference between the two groups. Therefore, the complication rate seems definitely higher than usual for SEEG implantations [2, 8]. To us, it is important to reconsider the definition of complication. Three of the four reported complications were indeed asymptomatic bleedings, while only the remaining patient experienced a transient mild headache with no neurological deficit. We actually think that such events should not be classified as a complication. In 2019, our group published a retrospective analysis of 742 SEEG procedures and proposed a new classification. “We defined ‘complication’ as any deviation from the normal postoperative course when: (i) a surgical treatment was adopted; (ii) an unexpected new neurological deficit lasting > 3 months was observed; (iii) the event led to the abortion of the clinical program; (iv) a potentially life-threatening event occurred; or (v) the patient died. ... We defined ‘notable event’ as any deviation from the normal postoperative course when none of the above listed criteria were matched.” [4] McGovern et al. have also proposed a dedicated scoring system for SEEG hemorrhagic complications [7]. The four events reported by Restrepo et al. could therefore be classified just as notable events, thus avoiding a misleading message about SEEG safety. Nonetheless, even if the four reported events should not be considered as true complications, they could be a red flag for lower accuracy. In fact, the entry point localization error (EPLE) and the target point localization error (TPLE) are higher than in other studies with similar procedural workflow (Table 1). The authors could have described more deeply their workflow in order to hypothesize which is the reason of such inaccuracy. Which are the image parameters? What about the rigidity of the frame holder? Which is the used tool holder for the twist drill? For example, one of the possible causes could be the use of optical tracking for image-to-patient registration. More details are needed. In conclusion, we think that small and clinically irrelevant intracranial bleedings are reported too often as SEEG complications, differently from most other neurosurgical areas of interest.