P156. Good long-term error rate retest reliability of 10 Hz neuronavigated repetitive TMS speech/language mapping in healthy volunteers

Abstract

Introduction Maximizing the extent of resection whilst preserving important brain functions is a major goal in the surgical treatment of brain tumours. For primary motor functions, the presurgical diagnostic techniques such as functional MRI (fMRI) and navigated transcranial magnetic stimulation (TMS) have improved a lot over the last decades and are extremely useful for preoperative risk evaluation and planning of the surgical approach. By contrast, presurgical mapping techniques for language-relevant areas are regarded poorly reliable (fMRI) and/or rather unspecific (repetitive TMS, rTMS). Methods We investigated the retest reliability of neuronavigated rTMS for speech/language mapping in 11 right-handed healthy volunteers in 3 consecutive sessions, spaced by 2–5 (short term) and 21–40 (long term) days. Navigated 10 Hz rTMS (Nexstim eXimia 4.2) was applied over the dominant hemisphere after determination of the individual inhibition threshold over the primary motor representation of the face/tongue. The bursts were triggered to picture presentation (naming task) without delay (picture-to-trigger interval = 0). Errors were categorised as follows: arrest, delay, anomia, dysarthria, semantic and phonematic paraphasia. Results As shown in Table 1 , a good feasibility in terms of the evocation of errors was observed. Per session, speech delays occurred most frequently (9.2 ± 2.1), followed by dysarthria (5.1 ± 2.4) and speech arrests (3.8 ± 1.8) whilst paraphasias were rather rare. There was a high variability in the distribution and the frequency of speech errors between the subjects. The reliability of the error frequency was rather good regarding long-term comparisons (Cohen’s weighted Kappa k = 0.41 [session 1–3]/k = 0.42 [session 2–3]) and still fair for the short-term (k = 0.29 [session 1–2]). Analyses of the spatial reliability of error representation are ongoing. Conclusions Presurgical speech/language mapping by neuronavigated rTMS is a promising tool to improve the risk evaluation and the planning of the surgical approach. In this preclinical study a rather satisfactory retest reliability of rTMS mapping at 10 Hz could be shown. However, since the method is still young, optimisation of the stimulation protocol and further intraoperative validation in clinical trials are required.