BioBoyScout

Robert Toczycki, JD/MBA. Patent attorney & independent biotech analyst. Author of the BioBoyScout RNAi white papers. Builder of biotech valuation tools. Chicago

[BioBoyScout @BioBoyScout]

$ARWR trades at over $10B today.
If ARO-MAPT's Phase 1 reads positive in 2H 2026, the math says it's worth $55-$90B to a strategic acquirer.
Not a guess. Four independent methodologies. Same range.
Full paper: https://t.co/0PnxQiLnzP

[BioBoyScout @BioBoyScout]

$ARWR showed on Oct 14 that a shot under the skin can lower tau in a human brain. New paper is about why nobody else has.

$ALNY was asked about brain shuttles this morning. They're building several, but no disclosed data, and no timeline.

$NVS is running a tau drug into the spinal canal while holding a license to the subQ alternative. $DNLI crossed with an enzyme and is now trying an ASO, by IV.

Convenience is an argument about who takes a medicine, not whether it works.
https://t.co/i79pJHaGxe

[BioBoyScout @BioBoyScout]

A regional evidence map for tau and $ARWR. Where the pathology sits in Alzheimer's, PSP, CBD and MAPT-FTD, against what each delivery route has been shown to reach.

Intrathecal siRNA in monkey: 86% knockdown in thoracic cord, 23% in putamen, none at all in caudate. SubQ reached all twelve regions with no gradient by depth.

ARO-SNCA hit >90% in substantia nigra. ARO-MAPT was never profiled there.

[BioBoyScout @BioBoyScout]

In June I published four papers on why Novartis, Lilly or Roche might be pushed into buying $ARWR. Those papers named two events: a gate the thesis had to survive, and a first bid that starts the cascade.

The gate opened on Oct 14. Nobody has played the check.

New note grades my own work.

Odds through 2027: 50% independent. Roche, Lilly, Novartis 15% each. $RHHBY $LLY $NVS
https://t.co/kzbHznmikC

[BioBoyScout @BioBoyScout]

$ARWR reported the first gene silencing in a human brain from a shot under the skin. The stock closed up 24 cents.

Wave got +74% on two patients. Intellia +55% on six. Sarepta +35% in March on two programs licensed from Arrowhead.

Same category of data. Same market. The hardest organ got four tenths of one percent.

UBS says it plainly: ARO-MAPT is "not included in our valuation."

A blank field is not an estimate of zero. It is a decision not to estimate.
https://t.co/LfXsWElrql

[BioBoyScout @BioBoyScout]

$ARWR silenced a gene in the human brain with a shot under the skin. Mean 73% tau reduction, 85% in one person.

My reaction note is up.

The headline is 73%. The number that matters is 105, the study day. Last dose was Day 15, so that was measured ninety days after the drug stopped.

A primate comparison Arrowhead published in May that nobody has connected to today: spinal canal delivery varied 22x across brain regions. Under the skin, under 4x.

The Bishop Moved
https://t.co/LFmcDOFBl5

[BioBoyScout @BioBoyScout]

WOW! 73%!!!

$ARWR just silenced a gene in the human brain with a shot under the skin. Nobody has ever done that. Not once. Not anywhere.

Here's the number nobody's reading: Day 105 is ninety days after the last dose. That's not a peak. That's a floor that held for three months with no drug on board.

Biogen gets 50-65% through a needle in the spine.

Arrowhead got 73% through a needle in the arm.

Two doses. Not three.

PSP and FTD regulatory talks already underway.

The brain had no book. Arrowhead just wrote the first page.

[ArrowheadPharma @ArrowheadPharma]

Today we announced interim Phase 1/2a clinical data for ARO-MAPT.

The results represent an important step in clinical validation of our TRiM platform for delivering siRNA to the central nervous system.

Join today’s webinar to learn more: https://t.co/enjRuvRB8j

[BioBoyScout @BioBoyScout]

$ARWR nobody takes intrathecal injections for twenty years before symptoms start. Presymptomatic tau lowering requires a subcutaneous drug. That is what gets tested tomorrow.

[Dirk Haussecker @RNAiAnalyst]

$ARWR don't want to sound hyperbolic, but tomorrow could be one of the most important datapoints in drug development of the decade (on par with Glp1):

Can subQ-delivered RNAi substantially knock down genes in the CNS? And can we see the data of the first drug to make a real dent against Alzheimer's (safe and due to subQ could be dosed presymptomatically)?

[BioBoyScout @BioBoyScout]

Roche licensed an ALK7 RNAi obesity drug from Rona today. Up to $100M near-term, $1.06B in milestones behind it. That makes five companies on ALK7: $ARWR, $ALNY, GSK, Novo, Roche. Nobody spends that way on a target they think is wrong.

Arrowhead is the only one to have reported human data, silencing the gene in human fat cells in January. Not mice. Not monkeys. People.

Here's what I'd actually watch.

The target isn't the hard part. ALK7 is published. The human genetics are published. Anyone can look up which gene to silence.

The hard part is getting an siRNA into a fat cell and keeping it there. That's chemistry, delivery, and a decade of failures nobody writes papers about. It's where these programs die.

A hundred million near-term against $1.16B total isn't conviction. It's an option. Roche bought a call option on a program that still has to clear every step Arrowhead already cleared.

Roche never disclosed the stage, but Rona's pipeline shows preclinical. That's two to three years to first human data, if nothing breaks.

Roche just paid a hundred million dollars to start where Arrowhead already finished. That's the going rate for a seat at a table somebody else set.

[BioBoyScout @BioBoyScout]

The DeVos work is the strongest counter to my timing worry and I should have had it. Reversing established pathology is a different claim from slowing new accumulation, and it widens the window. Your tipping point reframe is also better than the question I was asking. One wrinkle on measuring seeding directly: cryo-EM shows Alzheimer's, PSP and CBD each fold tau into a structurally different filament, so a tau seed assay has to catch the right strain. The same problem that makes conformation-selective drugs hard shows up when you try to measure the process. Which maybe argues for your combination idea from the other direction. If you cannot reliably measure the seed, lowering the substrate is the move that does not require you to.

[BioBoyScout @BioBoyScout]

Wednesday, $ARWR reports its tau number. A 60 gets called a triumph, a 45 a disaster, inside an hour.

Two tau drugs already proved that reading wrong. One removed 98% of its target and changed nothing. The other cleared tangles from living brains and still missed its endpoint, with the best result at the lowest dose.

The number answers one question: can a shot under the skin reach the brain. Not whether lowering tau helps anyone.

What to watch instead.

New paper: Ninety-Eight Percent of Nothing
https://t.co/nCYZPKMvig

[BioBoyScout @BioBoyScout]

Thinking about your last question more. The whole supply-reduction case rests on misfolded tau recruiting normal tau. If that's right, timing is everything, since you're starving a process that's already running. If propagation turns out to work some other way, lowering production does nothing and only clearance matters. Nobody's really tested which.

[BioBoyScout @BioBoyScout]

The oncology parallel is the best thing anyone's said to me about this. Molecular response not translating to survival is the same problem, one field further along.

On partial versus maximal, I don't think CELIA can settle it. All three dose arms landed around the same place on tau, so the study never got the separation it needed.

Your "which tau and when" point is what I keep coming back to. Cryo-EM shows each tauopathy folds tau differently, so even misfolded tau isn't one target. And timing might matter more than species. Ricks said last week that tau arrives later, which sounded like him admitting a limit on his own amyloid drug.

One thing on Wednesday: healthy volunteers won't have tau PET or cognition. NfL is just a safety check there. That sequence starts next year.

Your last question is the one nobody can answer, because nobody's run a trial designed to find the optimum instead of the maximum.

[BioBoyScout @BioBoyScout]

Interesting thought, though it all comes down to contract language neither of us can see. These deals usually have improvement clauses that pull in follow-on compounds against the same target. So if Novartis has exclusive rights to RNAi against their target, a dimer with that target in it probably sits inside their grant, not outside it. If the deal names specific compounds instead, then yeah, you're onto something.

What I'd say without seeing the contracts: the dimer is worth the most where nothing is licensed at all.

[BioBoyScout @BioBoyScout]

Ah, yes, digging deeper, that's a fair hit. AZD2693 got 89% knockdown, cut liver fat 12% placebo-corrected, and AstraZeneca still killed it in Phase 2b for inadequate efficacy. That's the experiment I said nobody had run.

Two things I'd still note: it failed at 52-week histology, where most MASH drugs die whatever the mechanism, and it moved liver fat only in homozygotes, not heterozygotes, at any dose. Most carriers are heterozygous. That's a target problem, not a silencing problem.

Wave has to clear the same biopsy endpoint. You've moved me on this one - thanks!

[BioBoyScout @BioBoyScout]

@BiotechTV @EliLillyandCo @DaveARicks Nice work on the Ricks interview, Brad. Twenty-odd minutes and you covered more ground than most people manage in an hour. No fluff, no softballs, and he actually answered. Thanks so much!

More of these, please.

[BioBoyScout @BioBoyScout]

The FDA built a fast lane for drug reviews four years ago.

First three years: six requests, zero accepted.

This week $ARWR walked in.

Why was the lane empty? To use it you hand the FDA your whole application minus the final study reports, months early. Most filings are a scramble that ends the night before.

You cannot buy an increment. You pay for it in advance, with work nobody watches.

New note: The Increment
https://t.co/ZF5gTruk0w

[BioBoyScout @BioBoyScout]

Shocking tech for sure. TfR1, the transferrin receptor, is the hottest door in biotech right now. Roche's Brainshuttle carries a full-size antibody through it by IV, and it works beautifully. Arrowhead's scientists went a step further: they paired a Fab fragment with a tiny siRNA, small enough to be given as a shot under the skin. Turning a brain drug into a simple shot is serious engineering, and that team pulled it off. Same door, lighter key.

[BioBoyScout @BioBoyScout]

On October 14, $ARWR tests whether a shot under the skin can reach the human brain.

The key: an antibody fragment that rides the transferrin receptor across the blood-brain barrier.

The lock: two patents granted this year.

Most investors have never looked at either.

The Fab Choice
https://t.co/B7G4burSVa

[BioBoyScout @BioBoyScout]

@LeftyT75 First brain drug out of the gate, and it may already be the one to beat. Turns out twenty years of practice on other tissues counts more than you know. Best in class? The monkeys gave it an A+.

[BioBoyScout @BioBoyScout]

@RNADilettante Nothing public yet. But this is the company that calls itself an "AND" company, not an "OR" company. Two silencers on one molecule AND a shot that reaches the brain? I'd be more surprised if nobody at Arrowhead has already worked on it.