Scientists have found a way to distinguish between two nearly identical forms of dementia using just a drop of spinal fluid — and it could change how doctors fight these diseases forever.
A new study published in Nature Medicine reveals that a biomarker called AcTau174 (acetylated tau at lysine 174) can accurately separate two major subtypes of frontotemporal lobar degeneration (FTLD) — one driven by a protein called TDP-43, and the other by tau — simply by measuring its concentration in cerebrospinal fluid (CSF).
In a cohort of 513 patients, AcTau174 levels were elevated across all dementia groups, but the largest spike was seen specifically in FTLD-TDP patients, particularly those with semantic variant primary progressive aphasia and GRN mutation carriers.
The biomarker distinguished FTLD-TDP from FTLD-Tau with an area under the curve (AUC) of 0.83, and separated FTLD-TDP from healthy controls with a striking AUC of 0.95 — meaning it correctly identified the disease in approximately 95 out of 100 cases when compared to controls.
That is not a minor refinement in a crowded field.
That is a potential game-changer for one of neurology’s most stubborn diagnostic problems.
What Is Frontotemporal Lobar Degeneration — And Why Does This Matter?
Most people have heard of Alzheimer’s disease.
Fewer have heard of frontotemporal dementia (FTD), even though it is just as devastating — and in some ways, far more insidious.
Research in Cambridge found that the prevalence of early-onset FTD and Alzheimer’s disease was identical in the 45 to 64 age group, with a mean age of onset for FTD of just 52.8 years.
That means FTD often strikes people who are still working.
People who are raising children.
People who are nowhere near what most of us think of as “dementia age.”
FTD is recognized as one of the most common presenile dementias, affecting an estimated 50,000 to 60,000 Americans, and representing between 10% and 20% of all dementia cases.
Because of the nature of its symptoms, and because patients are often “too young” for dementia to be considered, FTD is frequently misdiagnosed as a psychiatric problem or a movement disorder such as Parkinson’s disease.
And here is where the complexity multiplies.
FTD is not a single disease — the most common underlying forms involve either the deposition of tau protein (FTLD-tau) or TAR DNA-binding protein 43, known as TDP-43 (FTLD-TDP).
These two forms can look nearly identical from the outside.
A patient losing the ability to recognize familiar faces, or gradually losing the ability to form coherent sentences, may have either one.
Doctors have long been unable to tell which, without waiting for an autopsy.
Biomarkers to differentiate these subtypes have not been widely available, yet they are essential requirements for studying the natural course of disease and for including the right patients in clinical trials.
That gap has been sitting quietly inside clinical practice for decades.
This new research addresses it directly.
The Two Proteins at the Center of the Mystery
To understand why this discovery matters, you first need to understand the two proteins it separates.
Tau is a protein that helps stabilize the structural scaffolding inside neurons.
When it misfolds and clumps together, it forms the tangles associated with both Alzheimer’s and certain forms of FTD.
TDP-43 is a different kind of problem entirely.
In 2006, TDP-43 was identified as the cardinal protein in the most common subtypes of frontotemporal dementia and amyotrophic lateral sclerosis — and alongside the simultaneous discovery of progranulin mutations, this represented a major breakthrough in the understanding of FTD.
Under normal conditions, TDP-43 is primarily found in the nucleus of neurons, but when it misfolds and migrates to the cytoplasm, it forms harmful aggregates that contribute to neurodegeneration — a mechanism central to both ALS and FTD.
Together, FTLD-tau and FTLD-TDP-43 account for nearly 90% of behavioral variant FTD cases.
So nearly all behavioral FTD cases fall into one of two molecular camps.
And until now, the only reliable way to find out which camp a living patient belonged to was clinical judgment, imaging, and guesswork.
The Spinal Fluid Signal That Changes Everything
To develop this new test, researchers created an ultrasensitive immunoassay specifically designed to measure acetylated tau at the lysine 174 position in cerebrospinal fluid.
Think of acetylation as a chemical tag attached to a specific spot on a protein.
When tau gets tagged at this particular location, lysine 174, it behaves differently — and as it turns out, that tag tells a very different story depending on which disease is driving it.
Tau is known to be aberrantly acetylated in various neurodegenerative conditions, including Alzheimer’s disease, frontotemporal lobar degeneration, and traumatic brain injury.
Prior research had already shown that reducing acetylated tau by pharmacologically blocking p300-mediated acetylation at lysine 174 reduced tau pathology and improved cognitive function in animal models.
So this spot on the tau protein was already on the scientific radar.
What was not yet known was how powerfully its concentration in spinal fluid could serve as a living diagnostic signal in human patients.
The new immunoassay changed that.
AcTau174 concentrations were elevated in all dementia groups compared to controls, spanning FTLD-TDP, FTLD-Tau, Alzheimer’s disease, mild cognitive impairment due to Alzheimer’s, and dementia with Lewy bodies — but the signal was loudest and most distinct in FTLD-TDP patients.
But Here Is What Most People Get Wrong About Tau and Dementia
When the public hears tau discussed in the context of brain disease, the conversation almost always circles back to Alzheimer’s disease.
Tau tangles in Alzheimer’s — that’s the story most people know.
And that assumption leads scientists, clinicians, and even patients to a subtle but important misunderstanding when it comes to FTD.
The conventional wisdom says: more tau signal means more tau disease.
Surprisingly, that logic breaks down here.
In the conventional understanding, FTLD typically presents with either TDP-43, FUS, or tau aggregates, with the expectation of minimal co-occurrence or overlap between pathology types.
AcTau174 flips that intuition on its head.
The largest increase in AcTau174 was not in the FTLD-Tau group — it was in FTLD-TDP patients, particularly those with semantic variant primary progressive aphasia and GRN mutation carriers.
A tau modification is most elevated in patients whose disease is driven by an entirely different protein.
That is the kind of finding that makes researchers stop and rethink what they thought they understood.
What is likely happening is this: AcTau174 is not just a marker of tau pathology.
It is a signal of neuronal stress and damage.
When TDP-43 causes widespread cell disruption, the tau protein responds — and the acetylation at lysine 174 reflects that stress response with surprising precision.
This is what makes the biomarker so uniquely powerful.
It is measuring the brain’s reaction to disease, not just the primary pathology itself.
And that reaction, it turns out, differs dramatically between the two subtypes — enough to tell them apart with clinically meaningful accuracy.
Why Getting the Right Diagnosis Is Now More Urgent Than Ever
There are currently no approved disease-modifying treatments for FTD, and management relies on off-label pharmacotherapy and non-pharmacological approaches targeting symptoms only.
Medications used in the treatment of Alzheimer’s disease have no benefit in FTD patients and can actually worsen neuropsychiatric symptoms.
Let that sink in for a moment.
Giving an FTD patient the wrong drug — even one designed for a related dementia — can actively make things worse.
That is not a hypothetical concern.
It is a practical reality faced by neurologists managing these patients every day.
As of late 2025, only 12 interventional pharmacological trials targeting disease modification in FTD were identified in a structured review of registered clinical studies — half targeting progranulin gene mutations, and half examining therapies targeting neuroinflammation in sporadic FTD.
The pipeline is still remarkably small.
And one of the biggest reasons it stays small is clinical trial enrollment.
If a trial is testing a drug that targets TDP-43 pathology, but patients with tau pathology get enrolled alongside them, the results become muddied.
The drug can look far less effective than it actually is.
The availability of biomarkers like AcTau174 could be directly valuable in supporting the development of therapies targeting the specific pathologies that characterize FTLD.
AcTau174 could become the filter that makes future trials actually work.
How the Study Was Conducted
The study included individuals who visited the memory clinic of the Alzheimer Center Amsterdam at VU University Medical Center between 2003 and 2024.
All participants underwent comprehensive diagnostic workup including physical, neurological, and neuropsychological evaluations, brain MRI, apolipoprotein E genotyping, and standard CSF Alzheimer’s disease biomarker testing — with spinal fluid samples collected within six months of each diagnostic visit.
The cohort of 513 patients is notably large for this type of biomarker research.
Studies in rare dementias often work with far smaller groups, which limits how broadly any finding can be applied.
The researchers also validated the biomarker’s association with disease severity and progression — not just with the presence or absence of disease.
That is a crucial distinction.
It means AcTau174 does not simply tell you what type of disease is present.
It may also tell you how far along the disease has progressed.
That transforms it from a one-time diagnostic tool into something potentially capable of tracking a patient’s trajectory over time.
What This Means for the Future of Precision Neurology
The phrase precision medicine gets used constantly in oncology.
Cancer treatment has spent two decades moving away from treating “cancer” generically toward targeting the specific molecular subtype a patient actually has.
Brain disease has been far slower to follow.
Partly because the brain is harder to access than a tumor.
Partly because the proteins involved are harder to measure in living patients.
And partly because no one had a good enough biomarker to build around.
The 2025 Alzheimer’s Disease-Related Dementias Summit, hosted by the National Institute of Neurological Disorders and Stroke, set critical research priorities that specifically included developing standardized biomarkers and digital health technologies for FTD.
AcTau174 fits directly into that agenda.
Ongoing trials at UCSF, including studies testing novel PET tracers and drugs specifically targeting the language-predominant form of FTD, are among the frontline efforts trying to push the field forward.
A validated CSF biomarker like AcTau174 could feed directly into all of them.
It could help enroll the right patients, track disease progression, and measure whether an experimental treatment is actually working at the molecular level.
A recent meta-analysis published in JAMA Neurology reinforced the scale of the problem — analyzing data from more than 31 million person-years across 12 world regions to document FTD’s true epidemiological footprint.
The authors of that analysis specifically highlighted the importance of precise epidemiological data in supporting advances in both pharmacological and non-pharmacological treatments, noting that the increasing understanding of FTD’s complex pathophysiology, involving molecular subtypes such as FTLD-tau and FTLD-TDP, underscores the need for exactly this kind of precision tool.
The Broader Landscape of Biomarker Research in FTD
AcTau174 does not exist in isolation.
Other recent work published in Nature Medicine has explored whether plasma-based biomarkers — specifically extracellular vesicles carrying tau and TDP-43 — could serve a similar purpose without requiring a spinal tap.
That earlier research showed that plasma extracellular vesicles contain quantifiable amounts of TDP-43 and full-length tau, allowing reasonably accurate assessment of pathology in frontotemporal dementia and related disorders.
That approach has obvious appeal — a blood test is far less invasive than a lumbar puncture.
But the AcTau174 findings, published in the same journal and validated in a large real-world clinical cohort, offer a level of discriminatory accuracy that blood-based approaches have not yet matched.
Earlier studies using conventional tau and amyloid-beta ratios in CSF could distinguish FTLD broadly from Alzheimer’s, but could not reliably separate FTLD-TDP from FTLD-Tau.
That was the missing piece.
The new study provides evidence that CSF AcTau174 concentrations tend to be higher in individuals with FTD symptoms associated with TDP-43 pathology, and that this marker may help discriminate FTLD-TDP from FTLD-Tau with potential prognostic relevance for FTLD-TDP patients.
What Comes Next
No biomarker reaches clinical practice overnight.
The path from a validated research finding to a test a neurologist can order for a patient involves replication across independent cohorts, regulatory review, development of standardized assay protocols, and eventually, insurance and reimbursement decisions.
Researchers at the ALLFTD consortium — a network of academic medical centers across the United States and Canada conducting long-term studies of both sporadic and familial FTLD — are actively working to create disease progression models and validate novel biomarkers exactly like this one.
The FTD Disorders Registry continues to support enrollment in clinical trials across all forms of frontotemporal degeneration, including behavioral variant FTD, semantic variant primary progressive aphasia, and FTD linked to ALS.
The infrastructure for translation is already in place.
What it has been waiting for is a biomarker precise enough to build around.
A Molecule That Tells a Story Medicine Has Been Waiting to Hear
There is something quietly profound about this discovery.
For decades, families have watched loved ones lose themselves to a disease that doctors could not fully name while the patient was still alive.
Was it tau?
Was it TDP-43?
The answer often came only at autopsy — too late to guide treatment, too late to select the right clinical trial, too late to give families the clarity they deserved.
AcTau174 was not only elevated in FTLD-TDP patients — it was associated with disease severity and progression, suggesting it reflects the ongoing biological process of neurodegeneration in real time.
That is the detail that elevates this finding above a simple diagnostic test.
It means this molecule, floating in a few milliliters of spinal fluid, may be quietly narrating the story of a patient’s disease as it unfolds.
And that kind of real-time biological storytelling is exactly what precision neurology has needed.
The Bigger Picture
One small protein.
One acetyl group attached at one specific position on a chain of amino acids.
Measured in a few milliliters of fluid drawn from the lower spine.
And suddenly, a disease that looked like one thing to clinicians reveals itself as something more specific, more targetable, and potentially more treatable.
We appear to be on the cusp of a new era of treatment and management for frontotemporal dementia spectrum disorders — one in which rapid advances in understanding genetics, pathophysiology, and neuroimmunology are finally opening doors that stayed closed for a long time.
Research like this is what that era is built on.
Not dramatic headlines about a single cure.
Not a blockbuster drug that fixes everything at once.
But the slow, careful, precise work of understanding exactly what is wrong in a given patient’s brain — so that when the right treatment arrives, it can find exactly the right person.
That is the kind of science that changes outcomes.
And a molecule called AcTau174, hiding in plain sight in spinal fluid all along, may be one of the keys that finally unlocks the door.
