Biogen's diranersen (BIIB080) produced reductions in tau biomarkers and a signal of slowed cognitive decline in early Alzheimer's disease, but the Phase II CELIA study failed its primary endpoint, and a puzzling inversion — where the lowest dose showed the strongest clinical effect — leaves the optimal dosing strategy unresolved ahead of any Phase III program. Data from the randomized, placebo-controlled study were presented by Biogen at the Alzheimer's Association International Conference (AAIC) on July 14, 2026, marking what the Alzheimer's Drug Discovery Foundation (ADDF) described as the first evidence from a randomized trial that a tau-targeting drug can produce both a robust biomarker effect and a signal of clinical benefit.
What the CELIA data show
The CELIA trial (NCT05399888) enrolled patients with mild cognitive impairment due to Alzheimer's disease or mild AD dementia and evaluated three intrathecal dosing regimens of diranersen: 60 mg every 24 weeks, 115 mg every 12 weeks, and 115 mg every 24 weeks, over a 76-week placebo-controlled treatment period. The primary endpoint was a dose response-related effect on change from baseline in the Clinical Dementia Rating-Sum of Boxes (CDR-SB).
The study did not meet that primary endpoint. Across all three doses, CDR-SB results favored diranersen over placebo, but the pattern was not consistent with a dose-response relationship. The lowest dose — 60 mg every 24 weeks — showed a 26% slowing of decline on CDR-SB, a 42% slowing on the ADAS-Cog13 cognitive scale, and a 50% slowing on the MMSE. Exploratory analyses also showed directional slowing of decline on the modified iADRS (30%) and ADCOMS (23%), providing additional support for the observed clinical signal. Biogen reported that tau biomarker reductions were maintained throughout the dosing period across all dose levels studied. Meanwhile, no cases of ARIA were observed, consistent with diranersen's tau-targeting mechanism rather than amyloid removal.
Dose-response discordance complicates interpretation
The central interpretive challenge from CELIA is the divergence between the biomarker and clinical dose-response curves. Higher doses reduced tau more effectively by biomarker measures, yet the lowest dose produced the most pronounced clinical signal. This pattern does not follow the assumptions embedded in the trial's primary endpoint design, and it raises unresolved questions about about whether there is a threshold beyond which additional tau reduction does not translate into greater clinical benefit, or whether the relationship between biomarker change and clinical outcome is nonlinear.
Laura Nisenbaum, interim chief science officer at the ADDF, acknowledged the complexity directly: "The dose-response discordance between the biomarker and clinical endpoints raises important questions about the optimal level of tau reduction and the right dose of diranersen to test in subsequent studies." Biogen has said the results support advancing diranersen into further trials and that analysis of the full dataset is ongoing.
The context
Diranersen is an antisense oligonucleotide (ASO) administered intrathecally that targets MAPT mRNA, the transcript encoding tau protein. Through an RNase H-mediated cleavage mechanism, it reduces tau production at the transcriptional level, with the goal of slowing tau-driven neurodegeneration. The approach is mechanistically distinct from approved anti-amyloid therapies and is intended to address tau pathology, the second of the two defining hallmarks of Alzheimer's disease.