API ADAD Colombia
(2026)Objective
To evaluate the efficacy and safety of crenezumab in cognitively unimpaired PSEN1 Glu280Ala mutation carriers at near-certain risk for autosomal-dominant Alzheimer's disease, assessing whether treatment could delay or prevent the onset of cognitive impairment.
Study Summary
• No significant effect on amyloid plaque removal (florbetapir PET), CDR-SB, RBANS, or time to MCI/dementia; however, unadjusted analyses showed significantly less CSF Aβ42 decline (p=0.005) and greater increases in CSF Aβ40, CSF oligomeric Aβ, and plasma Aβ40/Aβ42 (all p<0.0001, not multiplicity-corrected) with crenezumab
• Serious adverse events were comparable (27% crenezumab vs 25% placebo) with no fatalities; ARIA-H 5% vs 2%; no ARIA-E
• Robust fibrillar amyloid removal, not achieved by crenezumab, appears necessary for clinical efficacy based on these and other anti-amyloid trial results
Intervention
Subcutaneous crenezumab 300 mg every 2 weeks (escalated to 720 mg SC every 2 weeks), with optional further escalation to 60 mg/kg IV every 4 weeks; treatment duration 5 to nearly 8 years (common-close design)
Inclusion Criteria
PSEN1 Glu280Ala kindred membership; age 30–60 years; cognitively unimpaired and medically stable; having a study partner; no history of Alzheimer's disease treatment; body weight 45–120 kg
Study Design
Arms: Crenezumab–mutation carrier (n=85) vs Placebo–mutation carrier (n=84); Placebo–non-carrier (n=83) as genetic kindred control
Patients per Arm: Crenezumab: 85; Placebo-carrier: 84; Placebo-non-carrier: 83
Outcome
• FCSRT-CI: annualized rate of change −0.03 (crenezumab) vs −0.04 (placebo); difference 0.01 (95% CI 0.00 to 0.02), p=0.16
• Serious adverse events: 23/84 (27%) crenezumab vs 21/84 (25%) placebo; no fatalities
• No significant effect on amyloid PET (relative reduction 3.6%, 95% CI −14.3 to 19.1) or downstream clinical biomarkers; unadjusted CSF/plasma Aβ species differences favored crenezumab-related target engagement without clinical benefit
Bottom Line
Crenezumab administered for 5 to nearly 8 years did not prevent or delay cognitive decline in preclinical autosomal-dominant Alzheimer's disease; robust fibrillar amyloid removal — not achieved by crenezumab — appears necessary for clinical efficacy, and future prevention trials should prioritize agents with demonstrated potent amyloid-clearing capacity.
Major Points
- Neither co-primary endpoint reached statistical significance: API ADAD composite test showed a between-group difference of 0.33 (95% CI −0.48 to 1.13, p=0.43) and FCSRT-CI showed a difference of 0.01 (95% CI 0.00 to 0.02, p=0.16)
- No significant treatment effect on amyloid PET (key secondary endpoint; relative reduction 3.6%, 95% CI −14.3 to 19.1), CDR-SB, RBANS, or time to MCI/dementia; however, unadjusted analyses showed less CSF Aβ42 decline (p=0.005) and greater increases in CSF Aβ40, CSF oligomeric Aβ, plasma Aβ42, and plasma Aβ40 (all p<0.0001, not multiplicity-corrected) suggesting target engagement without downstream benefit
- Crenezumab was generally well tolerated; serious adverse events occurred in 27% (23/84) of the crenezumab group vs 25% (21/84) of the placebo-carrier group, with no fatalities; ARIA-H 5% vs 2%; no ARIA-E
- The trial enrolled participants as young as age 30, up to approximately 14 years before the median symptom onset age of 44 years for this kindred; with treatment durations of 5–8 years, this represents one of the longest anti-amyloid prevention exposures ever studied
- Amyloid PET positivity was not required for enrollment; approximately 45% of carriers were amyloid PET-negative at baseline, potentially diluting any detectable treatment effect
- The dose was escalated progressively during the trial (300 mg SC → 720 mg SC → optional 60 mg/kg IV), suggesting early participants received subtherapeutic exposure
- Combined with other anti-amyloid β trial results, these findings indicate that robust fibrillar amyloid plaque clearance is necessary for clinical benefit — an endpoint crenezumab did not achieve
- Longitudinal biomarker and cognitive data from this trial will inform trajectories of preclinical ADAD and the design of future secondary and primary prevention trials
Study Design
- Study Type
- Phase 2 randomised double-blind placebo-controlled trial
- Randomization
- Yes
- Blinding
- Double-blind; investigators and participants masked to treatment allocation; study drug preparers and the Independent Data Monitoring Committee were unmasked
- Sample Size
- 252
- Follow-up
- 5 to nearly 8 years (common-close design at week 260 of last randomised participant); final data collection March 22, 2022
- Centers
- 1
- Countries
- Colombia
Primary Outcome
Definition: Dual co-primary outcomes: (1) Annualized rate of change in API ADAD composite test total score; (2) Annualized rate of change in FCSRT-CI (Free and Cued Selective Reminding Test–Cueing Index); both assessed via random coefficient regression in randomised mutation carriers who received at least one dose (mITT n=168)
| Control | Intervention | HR/OR | P-value |
|---|---|---|---|
| API ADAD composite: −1.43 (SE 0.29); FCSRT-CI: −0.04 (SE 0.00) | API ADAD composite: −1.10 (SE 0.29); FCSRT-CI: −0.03 (SE 0.00) | - (API ADAD composite: −0.48 to 1.13; FCSRT-CI: 0.00 to 0.02) | API ADAD composite: p=0.43; FCSRT-CI: p=0.16 |
Limitations & Criticisms
- Crenezumab likely did not achieve sufficient fibrillar amyloid plaque removal to produce clinical benefit; insufficient target engagement at the doses used may fully explain the null result
- Amyloid PET positivity was not required for enrollment, and approximately 45% of carriers were amyloid PET-negative at baseline and therefore not candidates for anti-amyloid benefit
- Progressive dose escalation during the trial (300 mg SC → 720 mg SC → optional 60 mg/kg IV) means early enrollees received subtherapeutic exposure for extended periods
- The common-close design produced heterogeneous treatment durations (5 to nearly 8 years) across participants, complicating interpretation of time-dependent effects
- Findings apply only to the PSEN1 Glu280Ala mutation in a single Colombian kindred, limiting generalizability to other ADAD mutations or sporadic Alzheimer's disease
- Success of blinding procedures was not formally assessed
- For one of the primary endpoints, the API ADAD composite change from baseline at year 5, the signal-to-noise ratio was approximately eight times worse than expected, indicating lower-than-expected statistical power
Citation
Tariot PN, Lopera FS, Ríos-Romenets S, Sink KM, et al. Safety and efficacy of crenezumab in cognitively unimpaired carriers of the PSEN1 Glu280Ala mutation at risk for autosomal-dominant Alzheimer's disease in Colombia. Lancet Neurol. 2026;25(2):147-159.