Transcranial ultrasound pulse stimulation reduces cortical atrophy in Alzheimer's patients: A follow‐up study

Abstract Introduction Ultrasound for the brain is a revolutionary therapeutic concept. The first clinical data indicate that 2–4 weeks of therapy with transcranial pulse stimulation (TPS) improve functional networks and cognitive performance of Alzheimer's disease (AD) patients for up to 3 months. No data currently exist on possible benefits concerning brain morphology, namely the cortical atrophy characteristic of AD. Methods We performed a pre‐/post‐therapy analysis of cortical thickness in a group of N = 17 AD patients. Results We found a significant correlation between neuropsychological improvement and cortical thickness increase in AD‐critical brain areas. Discussion AD patients who benefit from TPS appear to manifest reduced cortical atrophy within the default mode network in particular, whose memory‐related subsystems are believed to be disrupted in AD. TPS may therefore hold promise as a new add‐on therapy for AD.


INTRODUCTION
Ultrasound for the brain is a revolutionary new therapeutic concept.
Transcranial techniques such as high-intensity focused ultrasound are already in clinical use for non-invasive focal surgery, and clinical feasibility studies exist for blood-brain barrier opening for targeted local drug applications or local gene therapy. 1,4 Very recently, a first clinical study has presented a third major modality of ultrasound application: therapeutic brain activation in Alzheimer's disease (AD). 2 This study applied a novel pulsed ultrasound technique, transcranial pulse stimulation (TPS). 3 There were two major outcomes: (1)  functional magnetic resonance imaging (fMRI) and electroencephalogram (EEG) evidence for improved brain activity and memory network connectivity. TPS is particularly promising, because it represents an add-on therapy that may be applied concurrently with established therapies and therefore increase patients' chance of improvement. In pathological brains, in which tissue conductivity can be profoundly altered, electromagnetic methods of transcranial noninvasive brain stimulation (such as transcranial magnetic-and direct current stimulation, TMS and tDCS) are difficult to target. 6  areas. 5 As known from a large body of literature, neuroplastic rehabilitation involves not only functional 7 but also cortico-morphological changes, in terms of, for instance, MR-derived cortical thickness or voxel-based morphometry. 8,9 However, currently no human or animal data exist on whether therapy based on ultrasound brain stimulation can also lead to morphological (namely, cortical thickness) changes.
Here, we aim to fill this gap. Based on the recent evidence of functional improvement in AD patients, we investigate a possible effect of ultrasound stimulation (in the form of TPS) on cortical thickness in those same patients.

Participants
We analyzed data from the 20 AD patients of our previous TPS study, 2 for whom structural MR data (T1-weighted, magnetization-prepared rapid acquisition with gradient echo [MPRAGE] sequence) were available. The post-scans of two of the patients, and the pre-scan of one patient, were not available, leaving a sample size of N = 17 with complete pre-post MR data. The quality of cohort was generally "mild AD," with clinical AD classification according to the National Institute on Aging-Alzheimer's Association criteria.
The study inclusion/exclusion criteria were rather broad, to reflect the real out-patients' situation. Namely, inclusion criteria were: clinically stable patients with probable AD, at least 3 months of stable antidementia therapy (if any), signed informed consent, age ≥18 years. Exclusion criteria were: noncompliance with the protocol, relevant intracerebral pathology unrelated to the AD (eg, brain tumor), hemophilia or other blood clotting disorders or thrombosis, corticosteroid treatment within the last 6 weeks before first treatment.
The protocol for the study has received prior approval by the appropriate institutional review board (MUV 1227/2015). Informed consent was obtained from each patient.

Procedure
Most patients received TPS therapy for 4 weeks (three patients for only 2 weeks, one for 3 weeks). Functional and morphological MR data were recorded before and after TPS therapy.

TPS therapy and region of interest selection
Individual regions of interest (ROIs) were defined by a neurologist (R.B.) to target AD-relevant brain areas, that is, the AD network, consisting of: the classical AD stimulation target dorsolateral prefrontal cortex, areas of the memory (including default mode) and language networks. Specifically, ROIs comprised: bilateral frontal cortex (dorsolateral prefrontal cortex and inferior frontal cortex extending to Broca's area, ROI volume 136/164 cmş -2 × 800 pulses per hemisphere), bilat-

Highlights
• Ultrasound for the brain provides revolutionary and effective therapeutic concepts. • Transcranial pulse stimulation (TPS) can precisely target cortical and subcortical areas.
• TPS has been shown to improve functional networks and cognitive performance in Alzheimer's disease (AD) • We report such pre-post TPS changes that are predictive of cortical thickness increase.
• TPS might reduce cortical atrophy within AD-critical brain areas such as the default mode network.

Research in context
1. Systematic review: Our previous study indicates that therapy with ultrasound brain stimulation improves functional networks and cognitive performance of Alzheimer's disease (AD) patients, long term. These prior results informed our selection of regions of interest for the present study. We found no other published clinical studies on the brain-functional and -morphological correlates of ultrasound brain stimulation; this was ascertained by searching PubMed using the terms ("ultrasound" AND "brain stimulation") AND ("magnetic resonance" OR "neuroimaging") AND ("clinical" OR "Alzheimer's").

Interpretation: As patients improve cognitively, cortical
thickness-in AD-relevant brain regions-increases. We demonstrate, for the first time, that therapeutic ultrasound may change brain morphology, possibly by reducing cortical atrophy.
3. Future directions: These and other (preclinical) results provide ample evidence that ultrasound for the brain is an effective therapeutic concept for precisely targeted noninvasive brain stimulation. Ultrasound may therefore act as an add-on therapy for AD by inducing therapeutically relevant functional and morphological changes.

Cortical thickness derivation and analysis
We used a surface-based morphometry approach to analyze our anatomical images, as implemented in FreeSurfer (Massachusetts General Hospital, Harvard University, Cambridge, Massachusetts, USA).
This involves a standard pre-processing stream to derive cortical surface models (meshes), followed by a longitudinal processing stream to produce cortical thickness values at both time points; details of these procedures can be found elsewhere. [13][14][15][16][17] For the group analysis of cortical thickness, a simple two-stage model was used, which performs a linear fit in each subject independently, to reduce the repeated measures to a single slope, and then performs a regular cross-sectional analysis across subjects, to look for a pre-post effect. 15,16

Regions of interest and correlational approach
Our previous fMRI data showed that the entorhinal, parahippocampal, lateral parietal (inferior and superior), and precuneus areas responded most significantly to TPS therapy, in the sense that CTS increase was predicted by activation and connectivity of these areas. 2 Therefore, to examine the case of morphology, we computed correlations between pre-to-post changes in CERAD CTS and pre-to-post changes in the cortical thickness of these areas, as defined in the Desikan-Killiany atlas. 14 Both robust correlations, 18 that is, skipped Pearson's r that accounts for bivariate outliers, and multiple testing correction 19 were used, ensuring valid inference. 20

Data availability
The datasets generated during and/or analyzed during the current study are available in the TPS_CorticalThickness_Alzheimers repository, at osf.io/4gsj5.

DISCUSSION
Our study provides the very first data concerning ultrasound brain stimulation and brain morphology in humans. Results indicate that and human studies-but may also reduce atrophy of brain tissue that is treated with TPS. It is important to note that the principle of therapeutic neuromodulation with TPS that we describe here need not only apply to AD but, plausibly, to all neuropsychiatric diseases for which neuroplastic reorganization may be helpful. If this is true, then in the case of AD and its specific neuropathology, our results could be incipient evidence that TPS also acts by reducing cortical atrophy.
The exact cellular mechanisms behind ultrasound-based stimulation, and the direction of the overall effects, that is, whether they cause an increase or decrease in excitability at a systems level, is currently under intense research. Most likely, ultrasound neuromodulation rests on an initial change of cell membrane permeability. 21 This causes a cascade of transmitter, humoral factor, and cell activity changes with longterm neuroplastic effects. In our previous clinical AD study with TPS, the procedure was well tolerated and after treatment, memory performance improved for up to 3 months. These results parallel those of Notes: Robust Pearson correlations between pre-to-post-TPS change in neuropsychological score (CTS) and pre-to-post-TPS change in cortical thickness of all the functionally relevant ROIs considered. The P-value in bold (left superior parietal) is significant also after correcting for multiple comparisons using a projection-type outlier detection method. 19 Abbreviations: CI, confidence interval; CTS, corrected total score; ROI, region of interest; TPS, transcranial pulse stimulation. another brain stimulation study investigating the long-term effects of TMS as a treatment for AD. 22 These are the very first in vivo data on ultrasound related brainmorphological changes, however several limitations of the study have to be considered. First, TPS stimulation has not been controlled with a sham stimulation. Notwithstanding, our previous neuropsychological and functional data demonstrated that only stimulated brain networks improved, while non-stimulated networks worsened, in that the corresponding neuropsychological performance deteriorated-according to the expected course of the disease. 2 In addition, the long-term course of neuropsychological improvements clearly differed from the known placebo responses. 23 This corresponds well to our present finding of no global cortical thickness effect, but of specific effects in functionally relevant areas. Second, with this pilot study MR data are only available for pre-/post-therapy time points spanning 2-4 weeks. Whether morphological changes may be induced for longer time periodscomparable to the neuropsychological 3-months-on effects-remains to be investigated.
In conclusion, available evidence indicates that therapeutic ultrasound with TPS may indeed be effective as a new add-on therapy for AD and possibly for other neurodegenerative diseases. Besides inducing memory improvements, here an associated reduction of cortical atrophy is suggested. This result remains to be confirmed or refuted in future clinical trials that can provide higher quality of evidence.

CONFLICTS OF INTEREST
The authors declare no competing interests.

AUTHOR CONTRIBUTIONS
Tudor Popescu analyzed the data, generated the figures, and wrote the manuscript. Cyril Pernet helped with study design and data analyses. Roland Beisteiner designed the study, recruited patients, gave input in data collection, and wrote parts of the manuscript. All authors approved the final contents of the manuscript.