Excitatory rTMS on Pre-symptomatic High-risk AD Individuals to Delay Cognitive Decline

This five-year hypothetical study using repetitive transcranial magnetic stimulation (rTMS), a noninvasive way to stimulate brain cells, was proposed to help people at high risk for Alzheimer’s. Based on past research with spinal fluid biomarkers and memory testing, I predict that those receiving rTMS will maintain stronger cognitive function than those with sham treatment. This project explores whether early rTMS could delay symptoms and improve quality of life before Alzheimer’s fully develops.

Abstract

Alzheimer's Disease (AD) is a progressive neurodegenerative disorder which can be biomarked preceding cognitive decline. While repetitive transcranial magnetic stimulation (rTMS) has shown to be promising in mitigating symptoms in diagnosed individuals, its preventive effects remain unexplored. This study proposes a five-year longitudinal study of high-risk individuals to assess whether excitatory rTMS targeting the Dorsolateral Prefrontal cortex (DLPFC) can delay the onset of cognitive decline in participants predisposed to AD characterized by cerebrospinal fluid (CSF) biomarkers and genotype. Participants will undergo regular rTMS and cognitive decline will be assessed using the Montreal Cognitive Assessment (MoCA). This study hypothesizes that the group receiving rTMS will exhibit reduced cognitive decline over the five-year period when compared to controls. This study aims to pioneer preventative rTMS AD therapies. 

Introduction

AD is a neurodegenerative disease characterized by cognitive decline, memory loss, neuronal death, and fatal outcomes. Current drug therapies show some effectiveness, but are accompanied by adverse side effects¹. Many studies have researched the effects of rTMS on those living with AD and the recovery of their cognitive ability with positive results, however, there is little to no research on the preventative effects of rTMS on those predisposed to this disease. This study is grounded in a biomarker-based framework with the idea that preclinical AD progression can be detected with measurable risk indicators of Aβ42 and APOE4, and potentially delayed through early rTMS intervention.

AD can be predicted using different genes and biomarkers, such as the APOE4 gene and levels of Aβ42 CSF². Aβ42 is a small, soluble protein fragment in the CSF and brain interstitial fluid, produced as the brain metabolizes amyloid precursor protein. In healthy brains, Aβ42 is cleared away. In Alzheimer's brains, Aβ42 clumps, forming plaques mainly in the hippocampus³. When this happens, CSF levels of Aβ42 drop, indicating more amyloid plaques have accumulated in the brain. Low CSF Aβ42 is an early biomarker of AD progression⁴. This study hypothesizes that, in a group of participants who are equally predisposed to AD based on biomarkers, participants who receive rTMS will show less cognitive decline compared to those who do not receive it.

This research topic is very relevant to my personal life. I play violin and facilitate art therapy for those living with AD. These firsthand experiences of comforting memory care individuals have driven me to find a cure. This summer, I am researching Niemann-Pick Type C, a metabolic disease causing pediatric neurodegeneration. I hope to continue advancing the field of neurodegenerative diseases throughout my career.

Methods

Some biomarkers for AD include the APOE4 gene and CSF Aβ42 levels. Based on a pilot study of Cotelli et al. 2008⁵, this study requires 210 participants aged 55-70 who score 27-30 on the Montreal Cognitive Assessment (MoCA) with Aβ42 600–1000 pg/mL and both copies of the APOE4 gene. 

In normal individuals, CSF contains Aβ42: >1000 pg/mL, at-risk/preclinical AD CSF contains 600–1000 pg/mL, and symptomatic AD (mild cognitive impairment or dementia) CSF contains <600 pg/mL⁶. This studies’ inclusion criteria targets individuals with intermediate amyloid burden (600–1000 pg/mL), corresponding to a preclinical disease stage approximately 5–10 years before symptom onset⁶. Genotypically, carrying 2 copies of APOE4 is a strong indicator that AD will eventually onset⁷. This criteria will ensure participants have the genotype indicative of AD and enough Aβ42 to predict that AD will onset in the next 5 years, but not too little Aβ42 CSF to affect cognitive ability at the start of the study. 

To further justify our CSF concentration criteria, longitudinal biochemical analyses have shown that amyloid-β accumulates at an estimated rate of 4.5 to 6.5 milligrams over a ~19-year preclinical period before the onset of clinical dementia⁶. This study population is already positioned along this pathologic trajectory. At this constant rate, all participants are expected to build an additional ~1.2 to 1.5 mg of amyloid-β over 5 years. Other fluid biomarker studies demonstrate that CSF Aβ42 levels decline at an approximate rate of 2.5% per year once this plaque deposition begins⁸. Therefore, we expect a ~10–15% decrease in CSF Aβ42 concentrations in all participants independent of rTMS intervention, projecting CSF Aβ42 levels to decline to ~530–880 pg/mL after 5 years. This means that some participants are predicted to progress into the symptomatic AD CSF concentration category. This information will aid tracking AD progression and help in determining effect of the rTMS against biological AD pathology. Choosing to utilize CSF over PET scans or plasma assays ensures higher specificity and sensitivity in detecting early changes in Aβ42 concentrations^4,8, offering a more cost-effective and accessible method for biomarker tracking.

To assess our dependent variable of cognitive ability, this study utilizes the MoCA. The MoCA was chosen because it detects early cognitive changes more accurately than other tests, such as the Mini-Mental State Examination (MMSE), and detects more subtle preclinical changes over time⁹. MMSE might miss mild executive function or memory impairments that MoCA registers, which is integral to what this study aims to measure. Participants will need scores in the normal cognitive functioning range of 27-30 points. 

Based on pilot data from Krishnan et al., 2016¹⁰ showing MoCA declines of d = 0.29–0.64 over 3.5 years in high-risk older adults, we anticipate a small effect size (d = 0.30) when comparing rTMS-treated versus untreated groups. Using G*Power calculations with α = 0.05 and power = 0.80, approximately 176 participants (88 per group) are needed. Allowing for 15% drop-out over the study period, we will recruit 210 participants. 

Using specific MoCA data from this prior work, cognitively intact individuals’ scores declined by an average of 0.17 MoCA points per year, whereas individuals who developed mild cognitive impairment declined by approximately 0.52 points per year. Given the high-risk nature of our study population, we predict that the sham group will experience an intermediate decline of approximately 0.35 points per year, resulting in an estimated 1.7-point decline over the five-year study period. In contrast, participants in the active rTMS group are predicted to show minimal decline of 0 to 0.1 points per year, resulting in a 0-0.5 point decline over the study period due to the therapeutic nature of rTMS, supporting an expected effect size of approximately d = 0.30. 

Typically, AD works from the front of the brain to the back, first affecting cognition in the prefrontal cortex and memory in the hippocampus, eventually leading to neuronal death in vital regions like the medulla. Our study utilizes rTMS on the DLPFC. The DLPFC is important for executive functions such as planning, decision-making, and working memory. AD commonly affects these cognitive processes, and it is related in part to DLPFC dysfunction¹¹. Since we are measuring individuals in whom AD has not onset, the DLPFC is of interest to measure early onset of AD in the next 5 years.

Regarding the method of rTMS, it has been shown that high-frequency pulsed magnetic field stimulation can directly induce persistent changes in synaptic activity¹². It is widely believed that changes in synaptic strength contribute to learning and memory. Synaptic enhancement has been reported in cortical structures following 10 to 20-Hz stimulation using rTMS¹¹. Therefore, using excitatory rTMS on the DLPFC is expected to strengthen synapses at risk in those with AD.

Based on this pilot study of Moussavi et al.¹¹, this study will be a randomized, double-blind, sham-controlled trial investigating these effects of excitatory rTMS on cognitive outcomes. Participants will be randomized into two groups: active rTMS or sham stimulation. Randomization assures minimization of selection bias, distributes potential confounding factors, and ensures internal validity. Given the difference of AD progression rates, baseline cognitive ability, and individual responses to rTMS, random assortment will help balance variables and make this study applicable to wider populations. However, certain exclusion criteria will apply, including personal or familial history of seizures, traumatic brain injuries, alcohol abuse, and metal in the body. Ethical considerations include the potential burden of CSF lumbar punctures and the time commitment required for this long-term study. All participants will provide informed consent prior to enrollment and will be informed of their right to withdraw from the study at any point.

In the initial phase of the study, participants in the rTMS group will receive daily rTMS sessions of 20 Hz five days per week for four consecutive weeks. The sham group will receive the same treatment with 0 Hz stimulation. The stimulation will be delivered bilaterally to the DLPFC located using MRI. Previous brain MRIs will be used to conserve resources. If participants do not have one, MRI will be provided or a reference head model will be used to approximate. Each session will deliver 1,500 pulses using 25 trains of rTMS per day, following the protocol exemplified in Moussavi et al.¹¹. Sessions will last approximately 30-45 minutes.

After the initial treatment phase, participants will receive maintenance sessions weekly for three months, biweekly for the next nine months, and monthly for years two through five. The sham group will follow the same stimulation schedule, but with a sham coil that mimics the sound and sensation of real rTMS without delivering magnetic pulses. During the study, the MoCA will be given at baseline and annually, with a final MoCA at completion. Although the difference per year may be small, this data will be useful for tracking progress and for potential future work. If a participant misses a session, it will be made up as soon as possible. Absence for more than 10% of sessions will bar participants from the remainder of the study.

Predictions/Results

Based on prior longitudinal studies of at-risk adults, we predict that participants receiving active rTMS will demonstrate stabilization or very minimal decline in cognitive function over the study period measured by annual MoCA assessments. In contrast, participants in the sham stimulation group are expected to show a gradual decline in MoCA scores from natural aging and amyloid accumulation in preclinical AD.

Specifically, using data from Krishnan et al.¹⁰ and similar cohorts, we anticipate that participants in the sham group will show an average MoCA decrease of ~1.7 points over the study. Conversely, we expect that participants in the active rTMS group will exhibit negligible change or a decline of less than 0.3 points over the same timeframe. At the end of 5 years, a possible statistically significant difference in MoCA scores is expected between the two groups, supporting the hypothesis that high-frequency rTMS applied to the DLPFC can delay cognitive decline in high-risk, presymptomatic individuals. While exact changes will vary among individuals, we expect standard deviations around ±1.0 in MoCA scores based on past longitudinal AD biomarker studies¹⁰.

Figure 1. Predicted average MoCA scores at baseline and after 5 years in participants receiving active rTMS versus sham stimulation

Discussion 

Previous studies have shown the beneficial effects of rTMS on individuals already experiencing cognitive decline from AD ^5,11; this study aims to bridge the gap between current and preventative treatments, determining if early intervention could extend healthy years and maximize the highest quality of life. 

This study design has limitations to consider. Individual response to rTMS may vary due to genetic, structural, or neurochemical factors affecting baseline neuroplasticity, even in biomarker-controlled cohorts. The strict inclusion criteria of APOE4 homozygotes with specific CSF Aβ42 levels may limit generalizability to broader populations at risk for AD. Additionally, while rTMS may delay the cognitive symptoms of AD, it probably will not stop the progression of amyloid accumulation, which could have effects after study conclusion. Lastly, prolonged participation over the five-year period may be difficult to achieve. 

Future research could examine the effects of excitatory rTMS in more diverse genetic populations, broadening applicability. Studies could combine rTMS with pharmaceuticals to maximize effects. Additionally, rTMS could also be applied to other parts of the brain, and neuroimaging techniques like fMRI could be added to track the structural or functional strengthening of brain regions. Finally, future work could explore shorter, more intense rTMS therapy to fit more practical timeframes.

References

1. Weiler, M., Stieger, K. C., Long, J. M. & Rapp, P. R. Transcranial magnetic stimulation in Alzheimer’s disease: Are we ready? eNeuro 7, ENEURO.0235-19.2019 (2020). https://pmc.ncbi.nlm.nih.gov/articles/PMC6948923/

2. Anoop, A., Singh, P. K., Jacob, R. S. & Maji, S. K. CSF biomarkers for Alzheimer’s disease diagnosis. Int. J. Alzheimers Dis. 2010, Article ID 606802. https://pmc.ncbi.nlm.nih.gov/articles/PMC2915796/

3. Harten, A., Wiste, H. & Weigand, S. Detection of Alzheimer’s disease amyloid beta 1-42, p-tau, and t-tau assays. Alzheimers Dement. 17(S6), e051572 (2021). https://alz-journals.onlinelibrary.wiley.com/doi/10.1002/alz.12406

4. Teunissen, C. E., Chiu, M.-J. & others. Plasma amyloid-β (Aβ42) correlates with cerebrospinal fluid Aβ42 in Alzheimer’s disease. J. Alzheimers Dis. 62, 1523–1532 (2018). https://pubmed.ncbi.nlm.nih.gov/29614646/

5. Cotelli, M., Miniussi, C., Zanetti, O., Cappa, S. F. & Manenti, R. Transcranial magnetic stimulation improves naming in Alzheimer disease patients at different stages of cognitive decline. Eur. J. Neurol. 15, 1286–1292 (2008). http://onlinelibrary.wiley.com/doi/10.1111/j.1468-1331.2008.02309.x/full

6. Roberts, B. R., Lind, M. & Wagen, A. Biochemically-defined pools of amyloid-β in sporadic Alzheimer’s disease: Correlation with amyloid PET. Brain 140, 1486–1498 (2017). https://academic.oup.com/brain/article/140/5/1486/3098577

7. Hampel, H., Dubois, B. & Feldman, H. H. The enigma of mixed dementia. Alzheimers Dement. 12, 701–709 (2016). https://alz-journals.onlinelibrary.wiley.com/doi/10.1016/j.jalz.2006.09.002

8. Jack, C. R. et al. Tracking pathophysiological processes in Alzheimer’s disease: An updated hypothetical model of dynamic biomarkers. Lancet Neurol. 12, 207–216 (2013). https://www.sciencedirect.com/science/article/pii/S1474442212702910

9. Pinto, T., Bul, T. & Machado, L. Is the Montreal Cognitive Assessment (MoCA) superior to the MMSE in detecting mild cognitive impairment and Alzheimer’s disease in the elderly? Int. Psychogeriatr. 27, 1781–1791 (2014). https://pubmed.ncbi.nlm.nih.gov/30426911/

10. Krishnan, K. et al. Changes in Montreal Cognitive Assessment scores over time. Assessment 24, 772–777 (2017). https://pmc.ncbi.nlm.nih.gov/articles/PMC6757333/

11. Moussavi, Z. et al. Repeated transcranial magnetic stimulation for improving cognition in patients with Alzheimer disease: Protocol for a randomized, double-blind, placebo-controlled trial. JMIR Res. Protoc. 10, e25144 (2021). https://www.researchprotocols.org/2021/1/e25144/

12. Tokay, T. et al. High-frequency magnetic stimulation induces long-term potentiation in rat hippocampal slices. Neurosci. Lett. 461, 123–126 (2009). https://www.sciencedirect.com/science/article/pii/S0304394009007848