Transcranial ultrasound neuromodulation for epilepsy: A pilot safety trial

Abstract

Drug-resistant epilepsy (DRE) affects millions worldwide with significant morbidity and mortality [[1]Kwan P. Brodie M.J. Early identification of refractory epilepsy.N Engl J Med. 2000; 342: 314-319https://doi.org/10.1056/NEJM200002033420503Crossref PubMed Scopus (3996) Google Scholar]. Surgery and implanted devices benefit only a minority [[2]Bubrick E.J. McDannold N.J. White P.J. Low intensity focused ultrasound for epilepsy- A New approach to neuromodulation.Epilepsy Current. 2022; 22: 156-160https://doi.org/10.1177/15357597221086111Crossref PubMed Scopus (9) Google Scholar,[3]Ryvlin P. Rheims S. Epilepsy surgery: eligibility criteria and presurgical evaluation.Dialogues Clin Neurosci. 2008; 10: 91-103https://doi.org/10.31887/DCNS.2008.10.1/pryvlinCrossref PubMed Google Scholar]. Transcranial ultrasound (TUS) neuromodulation may be a promising noninvasive treatment for DRE. We report our experience, a first-in-human pilot safety trial using serial TUS treatments to the hippocampus in mesial temporal lobe epilepsy (mTLE). Treatments were planned and guided based on T1W and T2W MR excluding lesions in the ultrasound path; SWI excluded blood products at the target; UTE excluded bone anomalies at the temporal acoustic window. MRI was acquired at 3T using a 20-channel head coil and registration fiducials (PinPoint 128, Beekley Medical, Bristol, Connecticut) on the forehead, face, and near ears. A surgical navigation system (Polaris Vicra/Spectra, NDI, Waterloo, Ontario) was then used to register the fiducials with the MRI dataset. Four anatomical sites—inion, nasion, and pre-auricular points—were registered to the MRI dataset and TUS system using tracked probes. Four targets (4mm apart) in the anterior hippocampus were landmarked for sequential sonication beginning anteriorly and moving posteriorly. A previous publication details the TUS device [[4]Brinker S.T. Preiswerk F. White P.J. Mariano T.Y. McDannold N.J. Bubrick E.J. Focused ultrasound platform for investigating therapeutic neuromodulation across the human Hippocampus.Ultrasound Med Biol. 2020; 46: 1270-1274https://doi.org/10.1016/j.ultrasmedbio.2020.01.007Abstract Full Text Full Text PDF PubMed Scopus (34) Google Scholar]. Six adults (supplementary Table) with drug-resistant mTLE completed the treatments. The treatment period included 6 sessions (2 per week for 3 weeks). Ultrasound was delivered transcranially to the hippocampus perpendicular to its long axis through the ipsilateral acoustic temporal window. For each target, 140-s sonications were performed with the following specifications: f0 = 548 kHz, PRF = 500 Hz, Duty Cycle = 18–50 %, BRP = 7 s, ISPTA = 0.50–1.1 W/cm2, PNP = 0.14–0.42 MPa. All reported intensities are free-field values based on measurements in degassed deionized water at room temperature and ambient atmospheric pressure. In addition to pre-treatment structural MRI, a 7-min resting-state functional MRI (rs-fMRI) was obtained in a subset of patients, and imaging was repeated 1 month post-treatment in all subjects. rs-fMRI was analyzed using a data-driven approach, independent component analysis (ICA), to reveal spatially independent components that represent large-scale resting-state networks [[5]McKeown M.J. Makeig S. Brown G.G. Jung T.P. Kindermann S.S. Bell A.J. Sejnowski T.J. Analysis of fMRI data by blind separation into independent spatial components.Hum Brain Mapp. 1998; 6: 160-188https://doi.org/10.1002/(SICI)1097-0193(1998)6:3<160::AID-HBM5>3.0.CO;2-1Crossref PubMed Scopus (1631) Google Scholar]. All fMRI data were pre-processed (realignment, spatial normalization to the MNI (Montreal Neurological Institute) space, and spatial smoothing) using Statistical Parametric Mapping (SPM12, https://www.fil.ion.ucl.ac.uk/spm). Pre-processed rs-fMRI data was analyzed using the Group ICA Of fMRI Toolbox (GIFT, https://trendscenter.org/software/gift) [[6]Calhoun V.D. Adali T. Pearlson G.D. Pekar J.J. A method for making group inferences from functional MRI data using independent component analysis.Hum Brain Mapp. 2001; 14: 140-151https://doi.org/10.1002/hbm.1048Crossref PubMed Scopus (2295) Google Scholar], and 40 group-level independent components were separated from the data. Individual subjects’ components were back-reconstruct from the group results, and the default mode network (DMN) component was identified based on visual inspection. For Subject 1, the spatial-peak temporal-average intensity (ISPTA) was scaled per session by adjusting the sonication amplitude. ISPTA = 0.75 W/cm2 was delivered for each target in session 1, ISPTA = 0.18 W/cm2 for each target in session 2, ISPTA = 0.27 W/cm2 for each target in session 3, ISPTA = 0.55 W/cm2 for each target in session 4, and ISPTA = 0.82 W/cm2 for each target in the last sessions. For this subject, the duty cycle for session 1 was 50 %, this was reduced to 18.3 % for her remaining sessions by changing the number of wavelengths per burst (see reasoning below). The burst repetition period remained 2 ms (f = 500 Hz) and the pulse repetition period remained 7s (f = 0.14 Hz). For Subjects 2–6, the (ISPTA) was scaled per session such that ISPTA = 0.27 W/cm2 was delivered for each target in session 1, ISPTA = 0.55 W/cm2 for each target in session 2, ISPTA = 0.82 W/cm2 for each target in session 3, and ISPTA = 1.1 W/cm2 for each target in sessions 4–6. For these subjects, the duty cycle for all sonications was 18.3 %. The burst repetition period was 2 ms (f = 500 Hz) and the pulse repetition period was 7s (f = 0.14 Hz). Seizure frequencies were tracked 3 months prior to treatment and for 6 months after. Mood and memory evaluations [[7]Jones J.E. Hermann B.P. Woodard J.L. Barry J.J. Gilliam F. Kanner A.M. Meador K.J. Screening for major depression in epilepsy with common self-report depression inventories.Epilepsia. 2005; 46: 731-735https://doi.org/10.1111/j.1528-1167.2005.49704.xCrossref PubMed Scopus (165) Google Scholar,[8]Rossetti H.C. Munro Cullum C. Hynan L.S. Lacritz L.H. The CERAD neuropsychologic battery total score and the progression of alzheimer disease.Alzheimer Dis Assoc Disord. 2010; 24: 138-142https://doi.org/10.1097/WAD.0b013e3181b76415Crossref PubMed Scopus (68) Google Scholar] were completed prior to, during, and 1 month post-treatment. There were no adverse events, and no side effects reported during or after treatment. The first subject had one of her typical seizures during the first treatment session. Though any association with the TUS was unclear, the duty cycle was lowered from 50 % to 18.3 % for the remainder of her treatments and for all subsequent subjects. The lowest intensity exposure was used for her second treatment then scaled up again per above. No further seizures occurred with any participants during treatments. All six subjects experienced seizure reduction (average ∼ 50 %) post-treatment, five showed significant reductions (Fig. 1a). Duration of improvement varied from weeks to several months, with the exception of one subject who has remained seizure free for over a year. There was no significant worsening of mood or memory. Some improvements in mood were seen, one subject reported slightly worsened mood. Minimal improvements in memory scores were observed. There were no structural changes on any of the subjects’ post-treatment MRI scans. Pre- and post-treatment rs-fMRI was available for 3 subjects. Subjects 5 and 6 showed significant improvement in their DMN organization 1 month after treatment (Fig. 1b), while Subject 3 showed none. Additionally, single subject ICA results indicated that post-FUS rs-fMRI signal had a significantly smaller number of distinct sources of variation (i.e., estimated number of components) compared with pre-FUS data, in Subjects 5 and 6, but not Subject 3. The DMN component of Subjects 5 and 6 also showed improvement in stability.Fig. 1bComparison of pre- and post-TUS DMN maps derived from rs-fMRI ICA. White arrows indicate improvement in network organization in Subjects 5 and 6.View Large Image Figure ViewerDownload Hi-res image Download (PPT) Stern et al. demonstrated safety of a single TUS session in DRE [[9]Stern J.M. Spivak N.M. Becerra S.A. Kuhn T.P. Korb A.S. Kronemyer D. Khanlou N. Reyes S.D. Monti M.M. Schnakers C. Walshaw P. Keselman I. Cohen M.S. Yong W. Fried I. Jordan S.E. Schafer M.E. Engel Jr., J. Bystritsky A. Safety of focused ultrasound neuromodulation in humans with temporal lobe epilepsy.Brain Stimul. 2021; 14: 1022-1031https://doi.org/10.1016/j.brs.2021.06.003Abstract Full Text Full Text PDF PubMed Scopus (33) Google Scholar]. Our findings suggest TUS can be safely delivered in serial treatments to the hippocampus in adults with mTLE. Though a small cohort, our preliminary data indicates TUS may also be effective for seizure control, as most (5 out of 6) subjects experienced significant seizure reduction that was sustained over several months. The rs-fMRI data was also promising even at these low intensities. Of the subjects with rs-fMRI data, two showed improvement in DMN organization, while Subject 3 did not. Interestingly, Subject 3 also had very little improvement in seizures. This correlation suggests rs-fMRI may be a means to confirm target engagement and assess TUS’ neuromodulatory effects on the hippocampus. Further, the degree of functional connectivity enhancement with TUS may serve as a biomarker for effectiveness of TUS in seizure reduction. Further study is warranted for epilepsy and other neuropsychiatric disorders. This work was supported by grants (EJB) from the Brigham Research Institute (Bright Futures Prize) at Brigham and Women's Hospital of Harvard Medical School and the Epilepsy Foundation of New England.

Topics & Concepts