Reduction of Aβ production by inhibition of Beta-amyloid cleaving enzyme (BACE) BACE1 has been proposed as a promising treatment in Alzheimer’s disease (AD), especially when treating early in the disease process. Subjects with prodromal AD (pAD, CDR 0.5) and subjects asymptomatic at risk for developing Alzheimer’s dementia (ARAD, CDR 0) can be identified by positive biomarker patterns and are referenced here as early Alzheimer’s disease subjects. JNJ-54861911 has been described as a potent oral BACE inhibitor1. We report the results from the first clinical trial in biomarker-identified subjects with early AD, which show strong and dose-dependent reductions of Aβ in plasma and CSF.


      After a screening funnel, including general health, cognition, brain MRI and CSF biomarkers, pAD and ARAD aged 50 to 90 were identified. Eligible subjects were randomized (1:1:1;) to receive Placebo, 10mg or 50mg JNJ-54861911 once daily for a 4-week treatment period, tolerability, plasma and CSF pharmacokinetics (PK) and pharmacodynamics (PD) (including Aβ1-37, 1-38, 1-40 and 1-42, as well as sAPPα and sAPPβ) were assessed. To account for inter-subject variability of baseline Aβ and sAPP levels, reductions were expressed as percentage change from predose levels.


      424 subjects were screened and for 112 of these subjects biomarker information was collected. 30% of subjects with CDR 0 and 70% of subjects with CDR 0.5 were biomarker-positive, identifying 45 eligible subjects in total. Overall, treatment was safe and well tolerated. The 4 week treatment resulted in strong Aβ reductions in plasma (10mg: 83%; 50mg: 93%) and CSF (10mg: 67%; 50mg: 90%). The changes observed in CSF for all 4 Aβ peptides were consistent (Fig 1). sAPPβ peptides decreased while sAPPα increased up to 2-fold and no change was observed in total sAPP.


      JNJ-54861911 is a potent, brain-penetrant BACE inhibitor, yielding strong central Aβ reduction in the target population. Four weeks dosing was safe and well tolerated. The Pharmacodynamic effects observed in this study confirmed earlier observations in healthy volunteers. This is based on comparative modeling analyses including data from healthy elderly subjects and early AD subjects.
      Figure thumbnail fx1