A New Treatment for Alzheimer's: Rewiring the Brain's Default Mode Network
Ken Marash, CEO of Synaptica, describes a non-invasive approach to Alzheimer's that combines TMS with EEG to personalize stimulation of the brain's default mode network. In two sham-controlled phase 2 studies, the therapy slowed cognitive and functional decline by striking margins, reframing Alzheimer's as a disorder of network dysfunction rather than amyloid alone.
Overview
Matt Kaeberlein sits down with Ken Marash, a longtime neuromodulation executive from Boston Scientific who now leads Synaptica. Rather than focusing on removing amyloid plaques, Synaptica personalizes non-invasive transcranial magnetic stimulation (TMS) to restore the function of the default mode network (DMN), the brain network most disrupted in Alzheimer's. The key innovation is pairing TMS with EEG so the team can confirm, in real time, that they are engaging the right circuits inside the highly variable precuneus.
Ken walks through two published, sham-controlled phase 2 studies in which the therapy slowed decline on cognitive and functional endpoints by roughly 82% at 24 weeks and preserved everyday function even better. He explains the mechanistic case, including rising BDNF, reduced neuroinflammation, and amyloid reductions in animal models, while stressing that these results are still investigational.
The conversation reframes Alzheimer's as a disease of network dysfunction that may begin decades before symptoms. They also explore an intriguing adjacent finding: short, personalized stimulation appeared to boost memory recall in healthy volunteers for up to a week. Ken closes with a vision of widely available, affordable therapy and a plea to think in terms of systems biology rather than a single culprit molecule.
Key quotes
5They published a phase 2 trial and the data showed 82% slowing at 24 weeks on a global measure of dementia.
It's like Manhattan. If you're off by a block, you can be in a completely different neighborhood.
What if Alzheimer's was a disease of network dysfunction?
On functionality, we slowed the disease by 109%, meaning the active arm actually got better.
One course of stimulation can boost recall, and the durability of that boost is up to 7 days.
Key ideas
9Neuromodulation comes of age
Ken spent about a decade at Boston Scientific watching electrical stimulation grow into a roughly $7 billion field. It moved from a fringe idea into mainstream medicine, complementing drugs in pain and Parkinson's care.
The default mode network
The DMN is the brain's episodic memory and self-reference network, most active when the mind is at rest and wandering. It is the network most affected by Alzheimer's disease.
How TMS reaches the brain
Butterfly-shaped coils generate a brief, intense magnetic field that induces electricity two to three centimeters deep. Millions of neurons depolarize non-invasively, and the patient feels almost nothing but a faint click.
TMS plus EEG confirms the target
Combining stimulation with 64-channel EEG lets the team watch where energy travels through the brain. For the first time they can confirm they are engaging the DMN rather than shooting in the dark.
Personalization and dosing
The precuneus can vary by centimeters between individuals, and moving the coil a centimeter can change connectivity dramatically. Dose is calibrated per patient to stay both safe and effective.
Alzheimer's as network dysfunction
Ken argues amyloid is a product, not the primary cause of the disease. DMN dysfunction may appear 20 years before symptoms and drive a vicious cycle of inflammation and protein deposition.
Phase 2 results
In 50 sham-controlled patients, the therapy slowed CDR-SB decline by 82% at 24 weeks and preserved daily function by 109%. Compliance reached 94% with only mild, non-limiting side effects.
Mechanism of action
Published work shows increased BDNF in human plasma, upregulated dopamine sensitivity, and lower IL-6 in animal models. Amyloid was reduced across five different animal models.
Memory enhancement in healthy people
Small blinded crossover studies suggest three minutes of personalized stimulation can boost recall by 40 to 60% in healthy volunteers. The benefit appeared to last up to seven days.
Practical takeaways
6- 1
Rest fuels memory 10:30
The default mode network integrates experiences into memory when the mind is at rest. Giving yourself unhurried, screen-free downtime may support how the day is consolidated.
- 2
Stay critically curious 1:30
Promising early results still need larger trials to confirm them. Weigh investigational therapies with healthy skepticism and qualified professional guidance.
- 3
Think in systems, not single causes 34:00
Complex conditions rarely trace to one molecule. A systems-biology mindset invites multimodal thinking about long-term brain health.
- 4
Function is what families feel 47:00
Everyday abilities such as cooking, dressing, and using a phone are what preserve independence. Protecting function eases the burden on caregivers.
- 5
Lifestyle still counts 59:00
Social connection, enriched environments, and mentally demanding activities show modest but real benefits. They may complement other brain-health approaches rather than compete with them.
- 6
Personalization over guesswork 59:40
When it comes to the brain, precise and calibrated approaches are valued over one-size-fits-all. It is fair to ask how any therapy is targeted to the individual.
Topics & chapters
15Why rethink Alzheimer's
Matt frames the limits of amyloid-first treatments and introduces Ken Marash and Synaptica's network-based approach.
From Boston Scientific to Synaptica
Ken's decade in neuromodulation and how the field grew into a multibillion-dollar complement to drug therapy.
The result that changed his mind
The phase 2 headline of 82% slowing that convinced a self-described skeptic to come aboard.
Meet the default mode network
What the DMN is, why it governs episodic memory and sense of self, and how it fails in Alzheimer's.
How brain networks are mapped
Using resting-state and task fMRI to reveal which brain regions communicate with one another.
How TMS works
Non-invasive magnetic stimulation, butterfly coils, and what a treatment session actually feels like.
Adding EEG to see the target
Pairing TMS with EEG to confirm that the right circuits are being engaged.
The precuneus problem
Individual variability in size, location, and connectivity, and why dose and placement must be personalized.
Alzheimer's as a network disease
The case that amyloid is downstream of an earlier, self-reinforcing network dysfunction.
A systems-biology mindset
Moving beyond linear, single-target thinking toward personalized, multimodal combinations.
Inside the phase 2 trial
Design, endpoints, and the striking cognitive and functional results in 50 patients.
The 52-week extension
Durable benefits and improvements in behavioral symptoms like agitation and apathy.
Mechanism of action
BDNF, dopamine sensitivity, reduced inflammation, and amyloid reduction in animal models.
The road to approval
The data pipeline, the pivotal trial, and a possible market arrival by 2030.
Memory beyond Alzheimer's
Boosting recall in healthy people and a vision for accessible, affordable therapy.