Gerald Novak, MD1, H. Robert Brashear, MD 2, Luc Bracoud, MSc3, Enchi Liu, PhD4, Steve Einstein, MS1, Florent Roche, MSc3, Joel Schaerer, PhD3, Chahin Pachai, PhD3, Nzeera Ketter, MD2
1 Janssen R&D, Titusville, NJ; 2 Janssen R&D, South San Francisco, CA; 3 BioClinica, Lyon, FR; 4 Janssen R&D, LaJolla, CA
Brain atrophy and ventricular enlargement are well described phenomena in Alzheimer's disease (AD), reflecting the extent of neurofibrillary pathology and cumulative neuronal and synaptic loss. The rate of change in these volumes, as quantified on MRI, has been used as a biomarker of progression of AD in clinical trials.
ACC-001 (vanutide cridificar), a vaccine consisting of N-terminal peptides of Aβ42 combined with the adjuvant Quillaja saponaria (QS-21), induces antibodies that bind to the N-terminus of the Aβ peptide.
In a recently-completed phase 2 study, the effect of ACC-001 on rate of change of whole brain, ventricular, and hippocampal volume in subjects with mild to moderate AD was assessed with volumetric magnetic resonance imaging (vMRI).
Assessment of change from baseline in brain volume, an indicator of disease progression and potential effect of treatment, was an exploratory objective in this Phase 2 study.
The primary objective was the evaluation of effect of ACC-001+QS-21 compared to placebo on removing cerebral fibrillar amyloid in subjects with mild to moderate AD, as evidenced by florbetapir PET.
Other objectives included
Evaluation of the safety of ACC-001+QS-21 compared to placebo
Assessment of clinical efficacy
Assessment of effects on CSF Aβ, tau, and p-tau and on plasma Aβ
Assessment of Immunogenicity (serum IgG/IgM titer)
Randomized, double-blind, placebo-controlled
Patients aged 50-89 year, with mild to moderate AD
Baseline mini-mental status examination scores 18-26
Positive florbetapir PET based on visual interpretation.
Randomized 1:1:1 to 3 µg or 10 µg of ACC-001/QS-21 or placebo, stratified by ApoE4 status.
Treatment given by intramuscular injections over 18 months at Weeks 0, 4, 12, 26, 52 and 78, with follow up through Week 104.
Changes from baseline in whole brain, ventricular, and right and left hippocampal volume (indexed by brain, ventricular and R and L hippocampal boundary shift integrals; BBSI, VBSI, and L and RHBSI, respectively), assessed at screening and Weeks 24, 50, 76 and 102
Change in florbetapir PET GCA SUVr at Week 104
Clinical efficacy - change from baseline to Week 104 in the following:
11-item Alzheimer's Disease Assessment Scale – cognition (ADAS-Cog/11)
Clinical Dementia Rating Scale – Sum of Boxes (CDR-SB)
Disability Assessment in Dementia (DAD)
Mini-Mental Status Examination (MMSE)
All subjects underwent MRI examinations at screening and Weeks 2, 10, 24, 50, 76 and 102, composed of a high-resolution 3DT1 sequence in compliance with the ADNI-1 imaging protocol, FLAIR, T2, T2* and DWI.
All MRI data were sent for central processing and eligibility/safety monitoring to BioClinica; only datasets with good quality scans (no significant imaging artifacts, same MRI scanner used across timepoints) were considered for volumetric analysis.
The volumes of whole brain, lateral ventricles and hippocampus were assessed at screening using a fully-automated atlas-based segmentation algorithms (Figure 1), and volume changes at follow-up timepoints were assessed using BBSI, VBSI, LHiBSI, and RHiBSI.
BSI measures were converted to a rate of change per year by taking the change in baseline volume (in mL), dividing by the number of days since the baseline scan, and multiplying by 365.25.
Treatment differences (ACC-001/QS21 minus placebo) were analyzed by a mixed model for repeated measures (MMRM) with annualized xBSI as the response variable; the fixed effect model terms included:
treatment (3 μg, 10 μg or placebo)
visit (scheduled week)
baseline whole brain volume
APOE*E4 status (carrier or noncarrier) baseline age.
Treatment effects were estimated using least-squares (LS) means with factor levels weighted according to overall baseline sample proportions.
Demographics were generally balanced across treatment groups, though slight differences in several baseline characteristics suggest that subjects randomized into the placebo arm had slightly less severe AD (Table 1).
At Week 102, the LS mean annualized BBSI (±SE) was significantly greater for the 10 µg ACC-001/QS21 group than for placebo (Table 2 and Figure 2). BBSI was intermediate for the 3 µg group, but not significantly different from placebo.
The difference relative to placebo was greatest in the subset of patients receiving 10 µg ACC-001/QS21 with moderate AD (MMSE < 20) than for mild AD (MMSE > 21) (Table 3 and Figure 3).
A similar pattern of volume change (10 µg > 3 µg > placebo) was seen for VBSI, LHBSI and RHBSI, but neither of the differences relative to placebo were statistically significant (Table 2 and Figures 2 and 4).
There were no statistically significant differences between either dosage and placebo for ADAS-Cog/11, CDR-SB, MMSE, or DAD.
A total of 5 subjects (2 receiving 3 µg ACC-001/QS21 and 3 receiving 10 µg ACC-001/QS21) developed amyloid-related imaging abnormalities (ARIA-E); all were asymptomatic and resolved.
Immunization with ACC-001, at a dosage of 10 µg in combination with QS21, was associated with a greater rate of whole brain atrophy, as indexed by BBSI, compared to placebo.
These results are consistent with other anti-amyloid immunotherapeutic agents, including the vaccine AN-1792 and the monoclonal antibody, bapineuzumab. The explanation for this incremental volume reduction is unknown; hypotheses have included clearance of amyloid plaques, reduction of inflammatory components and shifts of interstitial fluid from amyloid-laden perivascular pathways.