Does Biogen's messy intrathecal anti-tau ASO result shift Alzheimer's tau optionality to the IV brain-shuttle players (DNLI, ARWR)?
Does Biogen's messy intrathecal anti-tau ASO result shift Alzheimer's tau optionality to the IV brain-shuttle players (DNLI, ARWR)?
The question
Biogen's anti-tau ASO for Alzheimer's missed its primary endpoint but advances to Ph3 on a confusing (possibly inverse) dose response — and it is intrathecal, a commercial handicap. Tau has the strongest disease-modifying rationale in Alzheimer's, so if proof-of-concept emerges, does most of the value accrue to IV-administered brain-shuttle players (Denali, Arrowhead) rather than the intrathecal incumbent?
Why it matters
A call-option on the strongest DMT target in Alzheimer's with a delivery-mode edge — DNLI/ARWR can dose IV with a plausibly lower efficacy bar given ease of use. Cheap optionality if tau proof-of-concept lands.
What we currently believe
Speculative. Buy-side is "nine to one negatively inclined" on Biogen's decision to advance, and phase-two Alzheimer's programs carry deep skepticism — so this is a low-priority watch-item, not a thesis. The data now exists (see below) and it is genuinely messy — the drug technically failed its primary (high dose vs placebo) with an inverted dose response (low dose looked better), yet the low dose showed a clinically-meaningful 0.54-point CDR-SB slowdown, so Biogen is advancing to Ph3. The panel's net read leans "the target (tau) may still be real; the mess is ASO-delivery, not tau" — which, if right, is exactly the argument for the IV brain-shuttle expressions (DNLI/ARWR) plus alnylam's intrathecal tau program.
Evidence we have
- paul-matteis in 2026-07-10-podcast-biotech-hangout-episode-188-july-10-2026: "The more interesting play are the brain shuttle approaches like Denali and Arrowhead, which can be given IV, and maybe a much lower efficacy bar given ease of use."
- eric-schmidt in 2026-07-10-podcast-biotech-hangout-episode-188-july-10-2026: "In terms of my buy side interactions, it's like nine to one sort of negatively inclined toward Biogen's decision to move this forward."
2026-07-17 update — the diranersen Ph2 data landed (resolves the "is the data real" gate)
The Biogen/Ionis diranersen (intrathecal anti-tau ASO) Ph2 readout, discussed on 2026-07-17-podcast-biotech-hangout-episode-189-july-17-2026:
- The result, as summarized on-air (yaron-werber): randomized placebo vs low vs high dose, primary endpoint at 76 weeks was high-dose ADAS-cog/CDR-SB vs placebo. "Technically the study failed because the primary endpoint... the high dose did not do as well against placebo. The lower dose... looked better, which is actually consistent with the phase 1b data" — an inverted dose response — with tau-imaging reduction and "a slowdown of deterioration of ADAS cog and also CDR summer boxes 0.54 points which is considered clinically meaningful by FDA. The bogey was 0.5. So based on that Biogen is going to go into phase three." Biogen stock fell on the print.
- The tolerability/ASO read (why this points to alternative modalities) — john-maraganore in 2026-07-17-podcast-biotech-hangout-episode-189-july-17-2026: "the nature of this inverted dose response is due once again to the poor tolerability that one sees with ASOs versus other technologies. And that's why Arrowhead with what they're doing. And also Al Nylam has a program targeting Tau which is intrathecal. Arrowhead's program is a Trans BBB shuttle." He is "net... encouraged by this as it relates to Tau... a lot of the complexity around the results could be mostly tolerability and ASO related." The higher dose had more AEs — consistent with a tolerability-driven inversion rather than a target failure.
- Bear counter (kept, not reconciled): the Street reaction was mixed because the efficacy achieved was "really no better than current anti-amyloid drugs like Leqembi and Kisunla" — so even if tau is real, the delivered benefit isn't yet differentiated. sam-facelli flags the on-target risk of merely lowering tau production (tau has normal function), though he concedes the biomarker dose-response data don't clearly support that concern.
Evidence we need
- Denali / Arrowhead brain-shuttle tau clinical data with a delivery (IV, better tolerability) advantage confirmed — the inversion supports the thesis mechanistically but no shuttle efficacy data is cited yet.
- Whether diranersen's Ph3 (on the lower dose) reproduces the 0.54-point CDR-SB slowdown — the proof-of-concept that the whole IV-shuttle read-across leans on.
How to resolve
The "is tau real at all" gate is now partially cleared (a clinically-meaningful low-dose signal exists, muddied by the inversion). Next: if ARWR/DNLI trans-BBB shuttle tau data shows efficacy with clean tolerability, re-rate them as the IV expressions; if diranersen's Ph3 fails to reproduce the low-dose signal, the whole tau-in-Alzheimer's category discounts again.