Provenance Record

Well-structured News piece (896 words, within the 400-1200 News range) using a clear Overview / What We Know / What We Don't Know / Analysis format. The nuanced result — a missed dose-response primary endpoint paired with positive secondary cognitive signals and uniform biomarker reductions — is handled with care and precision.

Both source snapshots read from disk and integrity-verified (uncompressed sha256 matched manifest exactly: source-0.html.gz = 47805ef0..., source-1.html.gz = 9e515610...). source-0 (BioPharma Dive, 'Biogen data highlight promise, questions surrounding tau-targeting Alzheimer's drugs', July 14 2026) and source-1 (Globenewswire/Biogen press release, 'Biogen Presents Phase 2 CELIA Data at AAIC...'). Both domains are on config/source_allowlist.txt. Every claim attributed to each outlet was located in that outlet's snapshot and confirmed. All body source URLs appear in article.sources (no orphan URLs; no bidirectional gap).

All clinical statistics verified verbatim against the Biogen release: 416 enrolled; mean age 68; 51% female; ~69% ApoE4 carriers; 60% MCI / 40% mild dementia; primary endpoint = dose response on CDR-SB at Week 76, NOT met; 60 mg every-6-months (24-week) arm slowed CDR-SB decline by 0.54 points (26%), with 42% ADAS-Cog13 / 50% MMSE / 30% modified iADRS / 23% ADCOMS; higher-exposure 115 mg arms at 14% and 9% on CDR-SB (0.28 and 0.18 points); CSF total tau reductions 50-65% across all doses; 'first tau-directed therapy to demonstrate reductions in both CSF total tau and brain tau pathology, as measured by PET, across all studied doses in a Phase 2 study' (exact quote). Every direct quote checked verbatim: BioPharma Dive's 'gumming up the genetic instructions...', 'Since the opposite occurred, the trial technically failed', 'most pronounced in the lowest dose arm, where clinicians reported a 26% slowing of decline', 'declined at least 23% and, by one measure, as much as 50% slower', 'generally well tolerated, with most adverse events being mild to moderate', 'why higher or more frequent dosing didn't spur greater effects', disclosed-in-mid-May sentence, Gallagher's 'looks comparable to amyloid lowering', Mamtani's 'de-risking event', Schmidt/Cantor's 'reopen its past can of worms' and 'making a very large bet on a neurodegenerative disease program...'; and the Biogen release's Priya Singhal quote ('If confirmed in Phase 3...') and the ARIA 'are not anticipated with diranersen based on its tau-targeting mechanism of action'. No hallucinated numbers or quotes detected. Ionis as originator, intrathecal administration, and antisense-oligonucleotide descriptor all sourced.

High-quality, accurately sourced medical submission. The nuanced clinical result is reported precisely and without hype; every specific traces to a cited allowlisted source and every direct quote is verbatim. Approved for publication.