ARO-MAPT: The Last Open Question
What TIDES revealed, what it doesn’t, and what Phase 1/2a still has to show
Robert Toczycki, JD, MBA
bioboyscout@gmail.com
847.227.7909
X: @BioBoyScout
Arrowhead presented updated ARO-MAPT data at TIDES USA 2026 yesterday. Most of what was shown has been public for months. The mechanism of how the drug crosses into the brain has been disclosed since September 2025. The competitive case against the existing intrathecal-delivered tau drugs has been on slides since September. The 65% reduction in tau biomarkers in monkeys has been public since September. Sell-side will write up the deck this week and most of what they cover will be material that has been sitting in Arrowhead’s investor relations page for the better part of a year.
Two things in the deck were genuinely new, and they matter more than the rest combined. This note explains what they are, what they mean, and what they do to the case heading into the Phase 1/2a readout.
Before going further, a quick primer for readers who don’t live in this corner of biology. Alzheimer’s disease (AD) involves two abnormal proteins building up in the brain: amyloid (which forms plaques) and tau (which forms tangles inside neurons). The current FDA-approved AD drugs, Leqembi and Kisunla, go after amyloid. They produce modest cognitive benefit at the cost of brain bleeding and swelling side effects, and they don’t reduce tau. The next frontier in AD is reducing tau, because tau tangle burden, not amyloid, is what correlates with how fast patients decline cognitively.
ARO-MAPT is Arrowhead’s drug that aims to reduce tau. MAPT is the name of the gene that produces tau protein. ARO-MAPT is a small RNA molecule that goes into the brain, finds the MAPT gene’s instructions for making tau, destroys those instructions, and stops new tau from being produced. Less tau being made means less tau available to misfold into tangles. That’s the thesis.
1. The pathology data: why “AT8” matters
Every tau-lowering drug has faced the same bear case for years. The argument runs like this: tau exists in healthy brains too. It is a normal protein that stabilizes the internal structure of neurons. The problem in Alzheimer’s is not that tau exists, the problem is that a subset of tau gets chemically modified, phosphorylated, in ways that cause it to misfold, clump up, and form the toxic tangles that kill neurons. The bear case asks: if you lower all tau indiscriminately, are you actually reducing the toxic tangles, or are you just lowering the healthy pool while the bad version keeps accumulating?
Several companies tried to solve this by developing antibodies that target only the bad version of tau. Roche, AbbVie, Biogen, UCB all ran trials with anti-tau antibodies. Most have failed. The implicit message from those failures was that maybe you cannot get to the toxic version by going at the upstream gene. Maybe the species-selectivity problem is fatal to the whole approach.
Arrowhead had not, until today, shown that ARO-MAPT actually reduces the toxic version of tau specifically. They had shown they could reduce total tau and the gene’s RNA output, but they had not shown the pathological species itself coming down. That was the last open question on the preclinical case.
The answer is in the TIDES deck. In mice engineered to develop tauopathy (a disease that mimics human Alzheimer’s tau pathology), ARO-MAPT cut the toxic phospho-tau species in the hippocampus by 53%. The hippocampus is the memory center of the brain, it is also the region where Alzheimer’s pathology starts. The toxic species is detected using a research tool called the AT8 antibody, which is the standard way scientists identify pathological tau in tissue samples. In short: in disease-model animals, ARO-MAPT cut the bad tau in the brain region where Alzheimer’s begins by roughly half.
This is the data point the field has been waiting for. It says that going after the gene works for the toxic species, not just total tau. The bear case is answered.
One important caveat. These mice are engineered to produce a mutant form of human tau at high levels. They develop tangle pathology on a fast clock that human disease does not follow. Mouse models of human disease are useful but imperfect, and a successful result in PS19 mice does not guarantee the same result in humans whose tau pathology arises from aging rather than a transgene. The AT8 data is preclinical proof of a mechanism. Whether the mechanism translates is what the Phase 1/2a will test.
The clinical trial is set up to test exactly that. One of the exploratory imaging endpoints is Tau PET, a brain scan that lights up tau tangles in living human patients. Tau PET is the human equivalent of what the AT8 antibody measures in mouse tissue. BIIB080, the Biogen/Ionis tau drug discussed later in this note, has produced Tau PET reductions in its Phase 1, but BIIB080 is an antisense oligonucleotide (ASO), a different class of RNA-targeting drug than ARO-MAPT’s small interfering RNA (siRNA). If ARO-MAPT produces a Tau PET signal in the Phase 1/2a, Arrowhead will have shown the first siRNA drug to reduce tangle burden in human brains, and will have done so using a simple under-the-skin injection rather than the spinal taps that BIIB080 requires.
Tau PET signal is not the cognition readout that takes years to mature, it is the biomarker readout that could show up within the 270-day study window. Late 2026 to 2027 is the realistic catalyst window.
2. The dosing reveal: why quarterly is now the working case
The second new piece of data extends a long-running monkey study that has been the backbone of Arrowhead’s CNS package.
Here is the setup. Arrowhead has been dosing monkeys with ARO-MAPT in a study that involves three weekly loading doses followed by monthly maintenance doses. The September 2025 presentation showed data through day 156 of this study, which is the time point when the last maintenance dose was given. At that point, tau mRNA was down 65-80% from normal and tau protein was down 70-90%.
The TIDES deck extends the same study to day 239, almost four months past the last dose. The mRNA suppression has substantially lifted; the gene is back to producing 55-80% of normal output, but the tau protein in the brain is still down 50-65% from normal.
This is the pharmacology that matters. Tau is a long-lived protein. Even after the drug’s effect on the gene fades and new tau starts being produced again, the existing pool of tau in the brain has been depleted, and it takes a long time to refill. The protein suppression persists long after the mRNA suppression has worn off.
Why does this matter commercially? It means quarterly dosing is realistic, and possibly less frequent than quarterly. Arrowhead’s PK/PD model, which the September deck disclosed, projects that monthly maintenance dosing keeps tau knockdown at roughly 80% indefinitely, and quarterly dosing keeps it at 50-70%. The TIDES data is the experimental confirmation that the model’s duration assumption is correct.
Now compare the patient experience. Leqembi starts with twice-monthly IV infusions at an infusion center for 18 months, then transitions to monthly IV or weekly at-home subcutaneous injection for maintenance. An expanded subcutaneous formulation that would eliminate the IV initiation phase entirely is currently under FDA review. Kisunla requires monthly IV infusions throughout the course of treatment. ARO-MAPT, if the human data tracks the monkey data, would be the first disease-modifying Alzheimer’s therapy whose dosing schedule consists of a quarterly subcutaneous injection from the very first dose; no IV phase, no clinic visit cadence, just an under-the-skin injection at home four times a year.
For a patient population that is overwhelmingly still working, still driving, and still managing their own life, this is not a minor convenience improvement, this is the difference between a drug that fits into a normal life and a drug that organizes a normal life around itself. It is also what makes ARO-MAPT economically viable in the rarer tauopathies, PSP, corticobasal degeneration, frontotemporal dementia, where small patient populations cannot support infusion infrastructure. A single drug, a single dosing paradigm, the entire tauopathy spectrum addressable.
3. What the Phase 1/2a actually has to show
With the pathology data and the dosing extension in hand, the preclinical case for ARO-MAPT is as complete as it gets before human translation. Every reasonable question that could be asked of an animal study has been answered.
Delivery works: the drug crosses from a subcutaneous injection into the bloodstream and then into the brain.
Distribution works: the drug reaches all regions of the brain uniformly, including the deep regions that have historically been impossible to reach with competing approaches.
Knockdown works: 70-80% reduction in tau mRNA across the brain. Protein follows: total tau down, pathological tau down.
Durability is there: protein down 50-65% sixteen weeks after the last dose.
Safety margin is wide: more than ten times higher than the effective dose in animal toxicology studies.
The clinical trial application is filed. The first human subjects were dosed in December 2025.
This brings up the competitive question. Biogen and Ionis have a drug called BIIB080 that also lowers tau. It is in Phase 2 (the CELIA trial) and reads out in 2026. BIIB080 has shown that lowering tau can reduce tau biomarkers and tangle imaging in humans, which is genuine proof-of-concept for the approach. Arrowhead’s own deck acknowledges this and gives BIIB080 credit for it.
But BIIB080 is delivered into the spinal fluid via lumbar puncture; a needle into the lower back that has to be done at a clinic. Arrowhead has data showing that drugs delivered this way cannot reach the deep regions of the brain at therapeutic concentrations. The drug pools near the injection site and distributes mostly to the spinal cord and surface of the brain. The deep regions, caudate, putamen, parts of the brainstem, get essentially no drug. This is not a problem specific to BIIB080, it is a problem of spinal fluid mechanics that affects every drug delivered by lumbar puncture.
Arrowhead’s data shows the spinal-tap route delivers drug 22 times more to surface regions than deep regions. The under-the-skin injection route delivers drug evenly across the whole brain, with less than 2-fold variation. For Alzheimer’s, where the disease starts in deeper structures and spreads outward, this difference matters. For the other tauopathies, PSP affects the substantia nigra and basal ganglia, corticobasal degeneration affects basal ganglia and cortex, it matters even more, because those diseases live entirely in the regions intrathecal delivery cannot reach.
BIIB080 is the better-validated drug today. ARO-MAPT is structurally a better-built version of the same idea. CELIA will tell us whether tau reduction translates to cognitive benefit in patients. ARO-MAPT will tell us whether the same translation happens with a delivery system that can actually reach the disease, dosed quarterly at home rather than via a procedure every few months at a clinic.
The Phase 1/2a study is now the translation event between a complete preclinical package and clinical proof. Three numbers will tell readers whether the platform translates to humans.
First, tau in the spinal fluid. Patients in the trial will have a small amount of spinal fluid drawn before and after dosing, and tau protein levels in that fluid will be measured. The animal data and the predictive model say a 50%+ reduction in spinal fluid tau is the expected number. BIIB080’s Phase 1b showed about 60%. If ARO-MAPT lands in that range, the drug translates as designed and the platform is calibrated. This is the cleanest readout in the study and the most directly comparable to existing data.
Second, p-tau217 in the blood. A newer biomarker that has become the gold-standard blood test for Alzheimer’s pathology. Unlike spinal fluid tau, plasma p-tau217 specifically tracks the disease-relevant tau pool rather than total tau. If this number moves in patients who got the drug compared to placebo, it indicates ARO-MAPT is engaging the specific tau species that drives disease, not just tau in general. A movement here is corroborating evidence on top of the spinal fluid signal.
Third, Tau PET brain imaging. The hardest endpoint and the one that bridges the AT8 mouse data to human disease. Tau PET imaging shows tangle burden in living patients’ brains. In normal Alzheimer’s progression, the signal grows over time as more tangles form. If ARO-MAPT-treated patients show their Tau PET signal flatten or decrease while placebo patients progress normally, that would be the first such data point for a subcutaneously-delivered siRNA in CNS medicine. It does not need to be statistically significant in a Phase 1/2a to be a re-rating event for the stock. Directional change is enough.
Three positives, or two of three with the third pointing in the right direction, and the platform translates. The cognitive testing data from a Phase 1/2a is too small in sample size to be statistically conclusive on its own, but if the biomarkers move, the cognitive direction sets the expectation heading into Phase 2.
This is the read on what TIDES set up. The biology is closed. The dosing case is sharpened. The clinical study is designed to read out on biomarkers that map directly back to the preclinical package. The Phase 1/2a interim biomarker disclosure, whenever Arrowhead chooses to share it, is the catalyst that re-rates the stock. It does not require cognition data. It does not require Phase 2. It requires the spinal fluid number to hit, the blood biomarker to move, and the brain imaging to point in the right direction.
If those three things happen, ARO-MAPT is no longer a preclinical platform story. It is a clinical platform story, and the rest of the TRiM CNS pipeline re-rates with it. That pipeline includes ARO-HTT for Huntington’s disease (preclinical, partnered with Sarepta), ARO-SNCA for Parkinson’s disease and synucleinopathies (preclinical, partnered with Novartis), ARO-ATXN3 for spinocerebellar ataxia 3 (preclinical, Sarepta), ARO-ATXN1 for spinocerebellar ataxia 1 (discovery, Sarepta), and an undisclosed cardiometabolic CNS target showing roughly 90% knockdown in NHP. Five disclosed subcutaneous CNS programs across two partners plus Arrowhead’s own wholly-owned ARO-MAPT, all riding on the same delivery system that the Phase 1/2a is about to validate or invalidate. The platform is what gets re-rated, not the individual bets.
Lastly, management has signaled there is more behind it. On the Q2 FY2026 earnings call, CEO Chris Anzalone said: “If early ARO-MAPT data are encouraging, expect a substantial expansion of our CNS pipeline beginning at the end of 2026.” Substantial pipeline expansion within months of a Phase 1/2a interim readout is not something a company assembles on the fly. It requires programs already in IND-enabling work, manufacturing capacity already reserved, regulatory strategies already drafted, and capital already allocated. Chris’s commitment implies that work has been done, and is sitting queued, waiting for the readout. The five disclosed CNS programs are not the ceiling, they are the floor.
A Note on Supporting Independent Research
If this note has been valuable to you, whether it shaped your thinking, validated your conviction, or simply saved you the time of doing this work yourself, a voluntary contribution is genuinely appreciated and directly funds the next paper.
For individual investors and readers
Any amount you feel reflects the value you received is welcome and meaningful. A contribution in the range of what you might pay for a single premium research report is a thoughtful gesture that makes a real difference.
For family offices, investment funds, hedge funds, and research platforms
This paper is the caliber of work that institutional research desks bill significant retainers to produce. If your team referenced it, distributed it internally, or used it to inform a position, a suggested contribution of $1,000 reflects the professional value of the analysis, though any amount is meaningful. Your support makes it possible to continue publishing at this level without a paywall that limits the reach of the ideas. If your organization requires an invoice to process a payment, please reach out directly at bioboyscout@gmail.com and one will be provided promptly.
There is no obligation and no expectation. This is purely a thank you for work that meant something to you.
Zelle: (847) 227-7909
Thank you for reading, and for being part of a community that takes this thesis seriously.
— Robert Toczycki | BioBoyScout
Important Risks, Disclosures, & Disclaimers
The author, Robert Toczycki (aka BioBoyScout), certifies that:
all views expressed in this note accurately reflect his personal opinions about the topic discussed; and
he was not compensated in any form for producing this note.
This note is published by BioBoyScout and is intended for informational and educational purposes only. It does not constitute investment advice, a solicitation to buy or sell securities, or a guarantee of future results. The author holds a long position in Arrowhead common stock. Arrowhead Pharmaceuticals (ARWR) is a publicly traded company; investments in its shares involve material risks, including the risk of total loss. All financial projections, acquisition price estimates, and valuation analyses herein are hypothetical frameworks for analytical purposes and do not represent predictions of actual outcomes. Readers should conduct their own due diligence and consult a registered investment advisor before making investment decisions. All data cited herein were sourced from publicly available company disclosures, SEC filings, press releases, and peer-reviewed literature as of May 2026.
About the Author
Robert Toczycki is an independent analyst and registered US Patent Attorney with a JD, an Executive MBA completed at the top of his class, and a BS in Mathematics and Computer Science from the University of Illinois at Urbana-Champaign. He has a deep passion for financial analysis, particularly identifying valuation discrepancies and demonstrating them through rigorous, data-driven research and solid analytics.
Comments or questions: bioboyscout@gmail.com.
Copyright © 2026, BioBoyScout. All Rights Reserved.
