N. Ketter¹, E. Liu¹, J. Di1, L.S. Honig², M. Lu1, G. Novak¹, J. Werth³, G. LePrince Leterme⁴, A. Shadman¹, H.R. Brashear¹

1. Janssen Alzheimer’s Immunotherapy Research & Development, LLC, South San Francisco, CA, USA; 2. Columbia University College of Physicians & Surgeons, Taub Institute for Research on Alzheimer’s Disease and the Aging Brain, New York, NY, USA; 3. Pfizer R&D, Collegeville, PA, USA; 4. Pfizer Global Research and Development, Paris, France

Corresponding Author: Dr. Nzeera Ketter, Janssen Alzheimer Immunotherapy Research & Development, 700 Gateway Boulevard, South San Francisco, California 94080, Tel: 510 248 2708, Fax: 510 248 2560, Email: nketter@janimm.com

J Prev Alz Dis 2016;3(4):192-201

Published online October 24, 2016, http://dx.doi.org/10.14283/jpad.2016.118

Abstract

Background: Vanutide Cridificar (ACC-001), a novel investigational immunotherapeutic vaccine designed to elicit antibodies against the N-terminal peptide 1-7 of the amyloid-beta peptide, believed to be important in the pathogenesis of Alzheimer’s disease (AD).
Objectives: To evaluate the immunogenicity, safety and impact of ACC-001 with Quillaja saponaria (QS-21) adjuvant on the reduction of brain fibrillar amyloid burden, assayed by positron emission tomography (PET) imaging, in patients with mild to moderate AD.
Design: Randomized, phase 2, interventional study. Trial registration: ClinicalTrials.gov Identifier: NCT01284387.
Participants: Individuals with mild to moderate Alzheimer’s disease (Mini-Mental State Examination scores 18-26; measurable amyloid burden in the expected range, on the screening 18F-florbetapir PET scan; and a Rosen modified Hachinski ischemic score ≤4).
Intervention: Participants were randomized to 3 μg or 10 μg ACC-001 (each in combination with 50 μg QS-21) or placebo (without QS-21).
Measurements: Primary endpoint was the change from baseline to week 104 in cerebral amyloid burden as measured by the global cortical average (GCA) standard value uptake ratio (SUVR) based on the brain 18F-florbetapir PET composite cortical SUVR between each ACC-001+QS-21 dose compared with placebo. Secondary endpoints included safety, immunogenicity and pharmacodynamics. Exploratory endpoints included cognitive and functional efficacy, and health outcome measures.
Results: Of 126 randomized patients (placebo: 40; ACC-001 3 µg+QS-21: 43; and ACC-001 10 µg+QS-21: 43), 125 received study treatment; 92 (73%) completed the study. Change in 18F-florbetapir PET GCA SUVR, was not significantly different between either of the two ACC-001+QS-21 treatment groups and placebo (3 μg +QS-21 vs. placebo diff=-0.03, p=0.54; 10 μg +QS-21 vs. placebo diff=-0.08, p=0.07), but the trend was numerically consistent with a dose response. The geometric mean peak anti-Aβ IgG titers were slightly higher in the 10 μg than the 3 μg group. The proportion of responders was similar in both dose groups of ACC-001+QS-21. The cerebrospinal fluid (CSF) p-tau changes from baseline in both active treatment groups were not statistically different from placebo, but were numerically consistent with a dose response (3 μg +QS-21 vs. placebo diff=-3.2, p=0.57; 10 μg +QS-21 vs. placebo diff=-7.0, p=0.19). The vMRI showed statistically significant faster treatment-related decrease in brain volume in the 10 μg group but was not significant in the 3 μg group, compared with placebo (3 μg diff =-1.3 mL/year, p=0.50; 10 μg diff=-4.2 mL/year, p=0.02). Measured plasma Aβ levels increased in parallel with peak anti-Aβ titers after each injection. Amyloid-related imaging abnormalities with edema/effusion (ARIA-E) were more frequent in patients who received ACC-001+QS-21 than placebo (6% vs. 0%) but none were symptomatic. The most common treatment-emergent adverse events in the active groups were injection reactions, and occurred more frequently in the ACC-001+QS-21 groups than the placebo (48% vs 8%), the majority of which were mild and transient.
Conclusions: Primary biomarker efficacy endpoints were not statistically significant in either dose group. The numerical decreases in 18F-florbetapir PET GCA SUVR suggests a dose-related trend for greater reductions in fibrillar amyloid burden in the ACC-001+QS-21 10 μg group compared with placebo. Likewise, while not significant, there was a numerical trend of decreased CSF p-tau levels with ACC-001, possibly consistent with a downstream effect in the ACC-001+QS-21 group. Insufficient antibody titers or quality, insufficient power to detect a difference, or too short duration of follow up may be reasons why a statistically significant response was not observed. Brain volume measures showed faster volume loss in the 10 µg treatment group, similar to the effect seen in few earlier AD immunotherapy trials which may suggest removal of amyloid and resultant decrease in inflammation. No new, unexpected safety signals were detected.

Key words: Alzheimer’s disease, amyloid, mild to moderate AD, PET, vaccine, vanutide cridificar.

Introduction

Alzheimer’s disease (AD) is a progressive neurodegenerative disorder, characterized by specific neuropathological changes, including intraneuronal (neurofibrillary tangles) and extracellular protein aggregates (neuritic plaques). The predominant components of neuritic plaques are amyloid beta (Aβ) peptides, particularly the 42 amino acid isoform (Aβ42) that is derived from a larger amyloid precursor protein (1). The “Aβ-cascade” hypothesis of AD pathogenesis addresses early events in the pathological process leading to amyloid deposition, which drives subsequent abnormal tau phosphorylation, neurofibrillary tangle formation, and ultimately neuronal death (2, 3). Multiple lines of evidence support aberrant Aβ42 production or clearance in the pathogenesis of AD, with a likely progression in neuronal dysfunction and neurodegeneration, and ensuing decline in cognition and function (4, 5).          
Currently available therapies for AD are symptomatic in nature, and do not show evidence of slowing disease progression or neurodegeneration (6, 7). Hence, there is an unmet need to develop therapeutics that can alter the underlying progressive disease pathology and pathophysiology of AD, and potentially slow clinical decline.
Vanutide cridificar (ACC-001) is an investigational therapeutic vaccine comprised of the N-terminal 1-7 amino acids of the Aβ peptide conjugated to a mutant diphtheria toxin carrier protein, CRM197, which is designed to stimulate the systemic production of anti-Aβ antibodies against the N-terminus 1-7 of Aβ. ACC-001 elicited anti-Aβ titers in non-human primates without evidence of a cytotoxic T-cell response when administered with or without the adjuvant Quillaja saponaria (QS-21). QS-21 is a highly potent stimulator of antigen-specific immune responses, and higher antibody titers are elicited when the ACC-001 vaccine is combined with QS-21 in individuals with mild to moderate AD (8-10). Preclinical studies demonstrated that ACC-001 with adjuvant immunostimulatory agent QS-21 is capable of inducing significant anti-Aβ antibody responses (8, 10, 11).
An earlier study with active Aβ immunotherapy, AN1792 with adjuvant QS-21, elicited anti-Aβ antibodies in 20% of patients with AD, but was associated with meningoencephalitis (12). An Aβ-specific T-cell response has been strongly suspected as the cause of this adverse event (AE) because the vaccine contained T-cell epitopes (7, 13, 14). Based on these data, ACC-001 was developed to restrict the immunologic response to a more defined portion of the Aβ molecule (N-terminal amino acids 1-7), which has been shown to have plaque-clearing activity in mice, and to avoid potential stimulation cytotoxic T-cells (10).
This phase 2, randomized, double-blind, placebo-controlled study was designed to evaluate the effect of ACC-001 with the adjuvant QS-21 compared with placebo in reducing cerebral amyloid burden, as well as safety, in patients with mild to moderate AD. Exploratory objectives were to evaluate the effect of ACC-001 on other biomarkers, immunogenicity, pharmacodynamic measures, efficacy, and health outcome measures.

Methods

Participants

Eligible participants were men and women, aged 50−89 years (inclusive), with a diagnosis of probable AD according to the National Institute of Neurological and Communicative Disorders and Stroke-Alzheimer’s Disease and Related Disorders Association (NINCDS-ADRDA) criteria; with a magnetic resonance imaging (MRI) scan not consistent with AD; a Mini–Mental State Examination (MMSE) score of 18 to 26 (inclusive); measurable amyloid burden on the screening 18F-florbetapir positron emission tomography (PET) in the range expected for AD patients (visual analysis); and a Rosen modified Hachinski ischemic score of ≤4.
Exclusion criteria included significant neurological or medical disease other than AD that could affect cognition; evidence of any abnormality on central read MRI including but not limited to ≥2 microhemorrhages, history or evidence of a prior hemorrhage >1 cm3, ≥2 lacunar infarcts, a prior infarct >1 cm3, cerebral contusion, encephalomalacia, aneurysms, vascular malformations, subdural hematoma, or space-occupying lesions; major psychiatric disorder; history of seizures or stroke; and history of clinically significant carotid or vertebrobasilar stenosis and other risk factors for thromboembolic stroke.
The Independent Ethics Committee or Institutional Review Board at each study site reviewed and approved the protocol, and the study was conducted in accordance with the ethical principles that have their origin in the Declaration of Helsinki and that are consistent with International Conference on Harmonization for Good Clinical Practices guidelines and applicable local and regulatory requirements. Written informed consent was obtained from each patient or their legally acceptable representatives before enrollment.

Study design and treatment

This phase 2, double-blind, randomized, placebo-controlled, parallel-group, study (ClinicalTrials.gov Identifier: NCT01284387) was conducted across 22 sites in the United States from February 2011 to January 2014. The study duration was approximately 26 months comprising a 7-week screening phase, followed by a 104-week (about 24-months) double-blind treatment phase (including 78 weeks [about 18 months] of dosing followed by an additional 6 months safety follow up after the last injection). At the beginning of the double-blind phase, eligible patients were randomized (1:1:1) to the three treatment arms: 3 μg or 10 μg of ACC-001 (each in combination with 50 μg QS-21 adjuvant) or placebo (without QS-21), based on a computer-generated randomization schedule stratified by APOE status (APOE*E4 carrier vs. non carrier). The investigational product or placebo was administered intramuscularly on day 1 and at weeks 4, 12, 26, 52, and 78 (total of 6 injections per patient) (Figure 1). The investigator, all study staff (except the dispensing pharmacists) and the patients were blinded to the treatment allocation.

Figure 1. Study Design and Patient Disposition

ADAS-Cog/11, Alzheimer’s Disease Assessment Scale – Cognitive subscale (11-item version); AD MACQ, AD Medication Administration Concerns Questionnaire; CDR-SB, Clinical Dementia Rating Sum of Boxes; DAD, Disability Assessment for Dementia; DS, Dependence Scale; FAQ, Functional Activities Questionnaire; MMSE, Mini-Mental State Examination; NPI, Neuropsychiatric Inventory; NTB, Neuropsychological Test Battery; PET, Positron emission tomography; RUD-Lite V.2.4, Resource Utilization in Dementia version 2.4; vMRI, Volumetric magnetic resonance imaging; w, week. Anti- Aβ titers were collected at the dosing visit and then 2 weeks

Other experimental medications and devices, immunosuppressants, chemotherapeutic agents, anticonvulsants, anticoagulants, and opioid pain relievers were not allowed during the study. Routine vaccinations with commercially available products could be given only 2 weeks after the administration of the investigational product/placebo.

Evaluations

Primary

The primary evaluation was the brain amyloid accumulation as measured by change from baseline to week 104 in global cerebral amyloid burden as measured by the global cortical average (GCA) which was calculated as the average standardized uptake value ratio (SUVR) based on the brain 18F-florbetapir PET scan using cerebellar gray matter as the reference region.

Safety

Safety evaluations included evaluation of AEs, deaths, prespecified events of special circumstance (ESCs) which included amyloid-related imaging abnormalities associated with edema or effusion (ARIA E), vasculitis, and intracranial hemorrhage, adverse drug reactions (ADRs), vital signs, clinical laboratory tests, electrocardiograms (ECGs), brain MRIs, suicidality assessments, physical and neurological examinations.

Exploratory

Exploratory evaluations included assessment of the effect of ACC-001+QS-21 on the following: a). other biomarkers (PET amyloid signal in individual regions of interest [ROIs], whole brain, ventricular, and hippocampal volume as measured by brain boundary shift integral [BBSI], ventricular boundary shift integral [VBSI], hippocampal boundary shift integral [HBSI], and cerebrospinal fluid [CSF] biomarkers [Aβx-42 and Aβx-40 concentration, phosphorylated tau (p-tau) and total tau (T-tau)]); b). immunogenicity (anti-Aβ [IgG and IgM] titer, CSF anti-Aβ titer); c). pharmacodynamic measures (plasma Aβx-40 concentrations); d). clinical efficacy (Alzheimer’s Disease Assessment Scale-Cognitive subscale [ADAS-Cog 11 items], Clinical Dementia Rating sum of Boxes [CDR SB], Disability Assessment for Dementia [DAD], Functional Activities Questionnaire [FAQ], MMSE, neuropsychological test battery [NTB], and Neuropsychiatric Inventory [NPI]); e). health outcome measures (Dependence Scale [DS], the Resource Utilization in Dementia Lite version 2.4 [RUD-Lite V2.4], and the AD Medication Administration Concerns Questionnaire [AD MACQ]).

Assessments

Blood samples for all analysis (efficacy, safety and exploratory) were collected at specified time intervals from baseline to the end of 104 weeks. 18F-florbetapir PET scans were planned to be obtained for all the patients at 4 time points: baseline and at 50, 78, and 104 weeks (or early termination, ET). Amyloid-PET images were acquired 50 minutes after the intravenous injection of 10 (370 MBq) ±1 mCi of florbetapir during screening phase and at week 50. Visual assessment of the PET scan was performed by expert readers at a central imaging core laboratory. The scans were also analyzed quantitatively by the same core laboratory. The primary amyloid PET endpoint was the GCA which was calculated as the average SUVR of five cortical ROIs: anterior cingulate, frontal cortex, lateral temporal cortex, parietal cortex, and posterior cingulate/precuneus. Exploratory amyloid PET endpoints consisted of SUVRs for specific ROIs. Brain MRI scans was acquired at screening and at weeks 2, 10, 24, 50, 76, 102 or at ET of the study in all patients for safety. The volumetric analysis of MRIs from screening and weeks 24, 50, 76 and 102 was performed.
An optional CSF substudy was included such that in consenting patients, lumbar puncture was performed at screening visit and at week 76, and CSF samples were analyzed for p-tau, T-tau and Aβ by enzyme-linked immunosorbent assay (ELISA). Plasma Aβ concentration were measured at screening and at weeks 2, 4, 6, 12, 14, 28, 54, 78, 80, 104 or at ET visit. Plasma and CSF Aβx-40 and Aβx-42 concentrations were determined using validated ELISA methodologies developed in compliance with Janssen AI Bioanalytical Development operating procedures and documented in method validation reports on Meso Scale Discovery (Meso Scale Diagnostics, Rockville, MD) technology and plate reader. Serum anti-Aβ IgG (total) and anti-Aβ IgM titers were evaluated by an ELISA assay at baseline, prior to and 2 weeks after each immunization at weeks 4, 6, 12, 14, 26, 28, 52, 54, 78, 80, 104 or at ET. The proportion of serological responders (patients with serum anti-Aβ IgG titer ≥300 U/mL no later than 30 days after the administration of the third injection of active investigational product) at each visit was analyzed at study end. The efficacy and health outcome measures were evaluated at baseline and at specified visits.

Statistical methods

Sample size determination

A total of 108 patients were planned to be enrolled in the study. The sample size calculation assumed a GCA SUVR standard deviation of 0.154 based on a prior study of bapineuzumab, using 11C-PIB as the tracer (15). Assuming maximum dropout rate of 30% in any treatment group, the planned sample size provided 94% power (2-sided alpha=0.05) to detect a difference of 0.134 between the pooled ACC-001+QS-21 group and placebo in the mean change from baseline at week 104 PET GCA SUVR. For comparing individual ACC-001+QS-21 dose and placebo, the power was 85% under same assumptions.

Analysis populations

The 18F-florbetapir PET analysis population included all randomized and dosed patients who had a baseline and at least 1 postbaseline assessment of cerebral amyloid burden (GCA SUVR) using 18F-florbetapir PET. The safety analysis set included all randomized patients who received at least one dose of the study medication. Immunogenicity analysis was based on the safety population.

Statistical analysis

All endpoints were summarized by treatment group using descriptive statistics. Primary analysis used a mixed-effect model for repeated measures (MMRM) to compare between treatment groups change from baseline in the GCA SUVR at week 104. This model included fixed effects of treatment, baseline GCA SUVR, baseline age, APOE*E4 status stratum, visit, treatment by visit interaction. An unstructured variance-covariance structure was used to model the within-patient errors.
Similar MMRM models were used for the analysis of exploratory clinical endpoints. For vMRI biomarkers, the MMRM for BBSI rate used baseline WBV as the baseline assessment, the MMRM for left/right HBSI rate used baseline left/right HCV as the baseline assessment, but the MMRM for VBSI rate did not include any baseline assessment. The change from baseline in p-tau, T-tau, and Aβ (x-40 and x-42) concentrations at week 76 were analyzed using an analysis of covariance (ANCOVA) model that included treatment, corresponding baseline assessment, baseline age, and APOE*E4 status stratum as covariates. Anti-Aβ titer was measurable in serum, and the proportion of serological responders (defined by a ≥300 U/ml serum anti-Aβ IgG titer no later than 30 days after the administration of the 3rd injection of active study medication) was summarized by treatment group with 95% confidence intervals calculated using the exact Binomial method. Anti-Aβ titer could not be successfully measured in the CSF due to lack of sensitivity. In addition to the comparisons of individual dose groups with placebo, the two active dose groups were combined and compared with placebo. All statistical tests were conducted with a two-sided type I error rate of 0.05 without adjustment for multiple comparisons.

Results

Unless otherwise specified, analysis results were presented in the form of mean (standard deviation [SD]) for numerical variables and count (%) for categorical variables. Out of 126 patients enrolled and randomized in three treatment groups (placebo: n=40; ACC-001 3 µg+QS-21: n=43; and ACC-001 10 µg+QS-21: n=43), 125 patients received study treatment (placebo: n=39; ACC-001 3 µg+QS-21: n=43; and ACC-001 10 µg+QS-21: n=43). A total of 92 (73%) patients completed the study. There were 115 screen failures excluded because of clinical and concomitant medication (n=65 [56%]) MMSE out of range (n=33 [28%]), amyloid below the threshold (n=18 [15%]) and miscellaneous. In all 3 treatment groups, the most common reasons for withdrawal were AEs and withdrawal due to patients’ request. The percentages of patients who were withdrawn were higher in the ACC-001 10 µg +QS-21 group (n=14 [33%]) compared with the 3 µg+QS-21 (n=12 [28%]) and placebo group (n=8 [20%]). The mean duration of AD was slightly shorter in the placebo group (1.9 [1.52]) than in the two ACC-001 groups (ACC-001 3 µg+QS-21: 2.8 [2.79]; ACC-001 10 µg+QS-21: 2.4 [2.07]). Otherwise, the overall demographic characteristics were similar across all treatment groups (Table 1).

Table 1. Demographic and Baseline Characteristics (All Randomized Analysis Population)

† Use of baseline treatment of cholinesterase inhibitors or memantine;  ‡  placebo group: n=39; total: n=125; Aβ42, Amyloid beta peptide 42-amino acid isoform;  AD, Alzheimer’s disease; ADAS-Cog/11, Alzheimer’s Disease Assessment Scale – Cognitive subscale (11-item version); APOE*E4, apolipoprotein E E4 allele; BMI, Body mass index; CDR-SB, Clinical Dementia Rating Sum of Boxes; CSF, Cerebrospinal fluid; MMSE, Mini-Mental State Examination; N, Number of patients in each treatment group, which was used as the denominators for percentage; n, number of patients in group with test result for that visit; PET GCA SUVR, Positron emission tomography global cortical average standard uptake value ratio;  p-tau, phosphorylated tau; SD, Standard deviation. 

A large proportion of subjects received concomitant medications for the symptomatic treatment of AD, including cholinesterase inhibitors (placebo: n=37 [95%]; ACC-001 3 µg+QS-21: n=37 [86%]; and ACC-001 10 µg+QS-21: n=34 [79%]) and memantine (placebo: n=22 [56%]; ACC-001 3 µg+QS-21: n=30 [70%]; and ACC-001 10 µg+QS-21: n=23 [54%]).
The mean MMSE total score was 22.0 (range: 17-27), and 85 (68%) patients had a baseline MMSE total score of ≥21 (ie, mild AD). Compared with the placebo group, the ACC-001+QS-21 groups appeared to have slightly higher mean values for ADAS-Cog/11 and CDR-SB total scores and slightly higher PET GCA SUVR, indicating greater impairment. However, baseline conditions were similar across treatment groups based on all randomized and Florbetapir PET analysis population (placebo: n=1.8 [0.32]; ACC-001 3 µg+QS-21: n=1.8 [0.29]; and ACC-001 10 µg+QS-21: n=1.9 [0.26]).

Primary endpoint

The MMRM-based least square mean (LSM) change from baseline in PET GCA SUVR  indicated a reduction in fibrillar amyloid signal in each treatment group at week 104, and the change from baseline was statistically significant for the ACC-001 10 µg +QS-21 group and the pooled active-treatment group. Although none of the differences compared with placebo were statistically significant (3 μg +QS-21 vs. placebo diff=-0.03, p=0.54; 10 μg +QS-21 vs. placebo diff=-0.08, p=0.07, figure 2), the numerical differences appeared to be consistent with a dose response.

Figure 2. Florbetapir PET GCA SUVR (CGM-Ref): Change from Baseline over Time (Mixed Model for Repeated Measures Analysis) (Florbetapir PET Analysis Population)

 GCA , global cortical average; PET, Positron emission tomography; SUVR, standard value uptake ratio ; PET SUVR calculated using cerebellar gray matter as reference region.

The vMRI showed incrementally faster treatment-related decrease in brain volume. The rate of change in whole brain volume as measured by annualized BBSI rate was statistically significant in the 10 μg group at week 102 (3 μg diff =-1.3 mL/year p=0.50; 10 μg diff=-4.2 mL/year p=0.02) compared with placebo, while the change in ventricular volume and left/right hippocampal volume was not statistically significant albeit showing numerically consistent dose-dependent trends (Figure 3). The CSF p-tau changes from baseline in both the active treatment groups were not statistically different from placebo, but were numerically consistent with a dose response (3 μg diff=-3.24 p=0.57; 10 μg diff=-7.02 p=0.19) (Table 2). There were no statistically significant differences from placebo for either of the ACC-001 groups in the changes from baseline to week 76 in CSF T-tau, Aβx-40, or Aβx-42 levels.

Table 2. Summary of Biomarker and Clinical Endpoints

ADAS-Cog/11, Alzheimer’s Disease Assessment Scale – Cognitive subscale (11-item version); BBSI, brain boundary shift integral; CDR-SB, Clinical Dementia Rating Sum of Boxes; CSF, Cerebrospinal fluid; DAD, Disability Assessment for Dementia; DS, Dependence Scale; FAQ, Functional Activities Questionnaire; MMSE, Mini-Mental State Examination; NPI, Neuropsychiatric Inventory; NTB, Neuropsychological Test Battery; p-tau, phosphorylated tau; VBSI, Ventricular Boundary Shift Integral; SD, Standard Deviation; t-tau, total tau.

Figure 3. vMRI Annualized BBSI Rate (mL/Year): Values over Time (Mixed Model for Repeated Measures Analysis) (vMRI Analysis Population)

 BBSI, brain boundary shift integral; VMRI, volumetric brain MRI

Mean serum anti-Aβ antibody titers indicated that immunization with ACC-001 in combination with QS-21 produced both IgG and IgM antibody responses (Figure 4A & 4B). After 1 year of treatment, geometric mean serum anti-Aβ IgG titers were slightly higher in the 10 µg group than in the 3 µg group of ACC-001+QS-21. None of the patients in the placebo group had measurable anti-Aβ IgG titers except for 2 patients each at week 52 and 54 when levels were slightly above the lower limit of quantitation. More than 90% of the patients in each ACC-001 groups were considered serological responders (anti-Aβ IgG titer ≥ 300 U/mL no later than 30 days after administration of the 3rd dose). At week 6, the number of patients with anti-Aβ IgG titer ≥ 300 U/mL were 31 (74%) patients in the ACC-001 3 µg + QS-21 group and 29 (67%) patients in the ACC-001 10 µg + QS-21 group, whereas at week 14, the counts became 36 (90%) in the ACC-001 3 µg+QS-21 group and 38 (95%) in the ACC-001 10 µg+QS-21 group.

Figure 4A. Serum Anti-Aβ IgG ELISA Titer (U/mL): Geometric Mean over Time

 Aβ, beta amyloid protein; ELISA, enzyme-linked immunosorbent assay; IgG, immunoglobulin G

Figure 4B. Serum Anti-Aβ IgM ELISA Titer (U/mL): Geometric Mean over Time (Safety Analysis Population)

Aβ, beta amyloid protein; ELISA, enzyme-linked immunosorbent assay; IgM, immunoglobulin M

Plasma Aβx-40 levels increased with the changes in anti-Aβ IgG titers in the active treatment groups, were much higher with ACC-001 than placebo, and were similar for the two ACC-001 dose levels (Figure 5).

Figure 5. Plasma Aβx-40 (pg/mL): Geometric Mean over Time (Pharmacodynamic Analysis Population)

Aβ, beta amyloid protein

The exploratory efficacy evaluations (cognitive [ADAS Cog/11 score, NTB score, and MMSE score]; global [CDR-SB score]; functional [DAD score and FAQ score]; and behavioral [NPI score]) showed no statistical differences between ACC-001 (either dose) and placebo groups (Table 2). There were no statistically significant or clinically meaningful effects of ACC-001 on health outcome measures.

Table 3. Treatment-Emergent Adverse Events in ≥10% of Patients in any Treatment Group

TEAE, treatment-emergent adverse event.

Safety

Overall, ≥93% of patients in each treatment group experienced at least 1 treatment emergent adverse event (TEAE), most of which were mild to moderate in severity. The most common TEAEs (≥10% in pooled active-treated group) and those at a rate that was ≥5% higher than in the placebo group were injection site pain, injection site reaction, urinary tract infection, behavioral and psychiatric symptoms of dementia, depression and fall. Except for injection site reactions, there were no apparent dose-related trends in AE incidence (Table 3). For injection site pain, although the ACC-001 3 µg and 10 µg groups were similar in terms of frequency (ACC-001 3 µg+QS-21: n=12 [28%]; ACC-001 10 µg+ QS-21: n=12 [28%]), the severity of the events was higher in the ACC-001 10 µg+ QS-21 group. Two deaths were reported in ACC-001 10 µg+QS-21 group, which were due to serious TEAEs (sepsis and progression of AD) and not considered to be related to the study medication. The TEAEs leading to early discontinuation of study medication or the study occurred in higher percentages in the ACC-001 groups (ACC-001 3 µg+QS-21: n=4 [9%]; ACC-001 10 µg+QS-21: n=7 [16%]) compared to the placebo group (n=2 [5%]). The only TEAE leading to early discontinuation of the study medication in >1 patient was ARIA-E (n=4). Pooled ACC-001 treatment group showed a 6% (n=5) incidence of asymptomatic ARIA-E, not seen with placebo (ACC-001 3 µg+QS-21: n=2 [5%]; ACC-001 10 µg+QS-21: n=3 [7%]). These events were considered to be mild or moderate in severity in all patients as reported by the investigator. None of these patients received additional doses of investigational product after ARIA-E had been confirmed. Because ARIA-E was confirmed retrospectively in 1 case, a single patient received a single dose of ACC-001 3 µg after the onset of ARIA-E.

The frequency of treatment-emergent serious AEs occurring during study treatment was similar across all the treatment groups. The most common ADR noted was injection site reaction, which showed an increase in the pooled ACC-001 treatment group as compared to placebo (48% vs 8%), the majority of which were mild and transient. No clinically relevant changes were observed in hematology, urinalysis and vital parameters in any of the treatment groups.

Discussion

This was the first study designed to assess the efficacy of ACC-001+QS-21 vaccine in reducing cerebral fibrillar amyloid burden in patients with mild to moderate AD. The safety profile of the vaccine (antigen with adjuvant) was assessed against saline for this early study to assess local and systemic reactogenicity against saline. Should the vaccine have been developed further the placebo would have included adjuvant to ensure that the appearance of active and placebo and the differential reactogencity did not unblind patients or the staff.   Both doses of ACC-001 (3 and 10 µg) revealed small numerically but not statistically significant, progressive reductions in amyloid-PET signal compared with the placebo, suggestive of target engagement.
The earlier studies in patients with AD had reported an increased 18F-florbetapir PET GCA SUVR over time in the placebo group (15, 16). The numerical decline in 18F-florbetapir PET GCA SUVR in the placebo group reported in this study was not statistically significant or clinically meaningful. This could be attributed to an increase in the amyloid burden in the cerebellar grey reference region resulting in a decrease in the ratio as a true reduction of cortical amyloid in untreated patients with AD does not appear plausible. The results may also represent a regression to the mean, but such a change would be unlikely to fully offset disease progression. Nonetheless, this decline in SUVR in the placebo group raises questions about the interpretability of the results and raises the possibility that the treatment effect of ACC-001+QS-21 may have been underestimated.
The rate of brain atrophy was consistent with typical AD progression. Brain volume measurements showed greater volume loss in the treatment groups, consistent with prior AD immunotherapy trials (16, 17); the etiology of this volume loss is unknown but possible explanations include amyloid removal, fluid shifts, reduced inflammation or neuronal loss (18).
There were larger decreases in CSF p-tau levels from baseline to week 76 in the ACC-001 groups than in the placebo group, consistent with potential downstream effect of ACC-001. The effects were small but showed a trend indicative of dose response, while no reductions in CSF T-tau levels were observed. However, it should be emphasized that T-tau is a marker of general neurodegeneration that can occur in other dementias and neurodegenerative diseases, whereas p-tau is a marker of the hyperphosphorylation state of tau proteins that is more specific to AD. No difference in the CSF T-tau levels were observed between placebo and the anti-amyloid antibody bapineuzumab in a phase 3 study in patients with mild-to-moderate AD (16), whereas a dose-related trend was observed in another phase 2 bapineuzumab study in a similar patient population (19). These differences in results may be due to the smaller sample size and hence need replication with larger sample sets.
Immunization with ACC-001+QS-21 stimulated the production of both IgG and IgM antibody responses, with moderately higher serum anti-Aβ IgG and IgM titers with the 10 µg dose than with the 3 µg dose. A greater peak dose response was reported after the 12 month boost. An earlier phase 2 ACC-001 study conducted in Japan without an adjuvant showed a much greater dose response (8). A dose response was also reported in another study of CAD106 administered without an adjuvant (20). The addition of QS-21 adjuvant has shown to increase the immunogenicity of the ACC-001 so that a significant dose response was not observed in geometric mean titer (8, 11).
In this study, more than 90% of the patients in each ACC-001 group were considered serological responders. The plasma Aβx-40 levels paralleled the changes in anti-Aβ IgG titers in the active treatment groups, were consistently higher and similar in the ACC-001+QS-21 groups than in the placebo group from week 14 through the end of the study. The decrease in titers at week 104 was largely attributed to the fact that the last injection of ACC-001+QS-21 was given 6 months prior to that evaluation. No apparent, consistent correlations between plasma Aβx-40 AUC and 18F-florbetapir PET GCA SUVR were reported for any of the treatment groups.
The demonstration of a prime response with IgM and prolonged elevation of IgM compared to that achieved with standard prophylactic vaccines is unusual but the mechanism or significance is unknown. The epitopes present in the vaccine are also present in endogenous Aβ in patients with AD. The inclusion of CRM in the ACC-001 vaccine construct may have resulted in a novel presentation of these vaccine antigens, resulting in the production of IgM usually observed when an antigen is presented for the first time.
The effect on biomarkers seen with ACC-001+QS-21 is small and may not represent sufficient alteration of the underlying disease processes to translate to clinical efficacy, which is consistent with the study outcomes. These results are consistent with the earlier studies in subjects with mild-moderate AD and with early AD (8, 11, 21). The lack of clear response differences in exploratory clinical efficacy measures in these studies could be due to the small sample size or the short duration of follow up.
No new safety signals for ACC-001 were detected in this study. The TEAEs that occurred were typical for a long-duration vaccine study in elderly patients with mild to moderate AD. The TEAEs were more frequent in the ACC-001 groups than in the placebo group, largely reflecting a difference in the incidence of injection site reactions, which were transient and mostly mild in severity. No meningoencephalitis or autoimmune diseases were reported in this study. The incidence of ARIA-E was low and only occurred in the ACC-001+QS-21 treatment groups, suggesting that the antibody had crossed the blood brain barrier. A relationship to APOE*E4 genotype has been seen in studies of monoclonal anti-amyloid antibodies (22), but among these 5 cases, 4 out of 5 were APOE*E4 non-carriers (1 in the 3 μg group and 3 in the 10 μg group).   The frequency of ARIA-E was too low and the study too small to draw firm  conclusions.

Conclusions

Amyloid PET imaging results did not show significant change in GCA SUVR. There was a numerical dose-response trend suggestive of target engagement. These results were consistent with CSF p-tau levels but not T-tau levels. Brain volume measurementss showed a faster volume loss in the ACC-001+QS-21 treatment groups, consistent with prior AD immunotherapy trials. The rate of brain volume decrease was consistent with typical AD progression. An acceptable safety and tolerability profile was reported. The incidence of ARIA-E was very low and therefore no relationship with risk factors such as APOE*E4 status was possible. All ARIA-E cases were asymptomatic and radiologically mild, despite all patients having detectable amyloid burden at study entry.

 
Acknowledgments: We acknowledge Rishabh Pandey (SIRO Clinpharm Pvt. Ltd.) for providing writing assistance (funded by Janssen Research & Development, LLC) and Dr. Bradford Challis (Janssen Research & Development, LLC) for additional editorial support for the development of this manuscript. Authors thank Bioclinica for the interpretation of the MRI scans and MNI for the interpretation of the PET scans. Authors also thank the study participants and the investigators for their participation in this study.

Funding: This study was sponsored by Janssen Research & Development, LLC and Pfizer, Inc. Writing assistance for this manuscript was funded by Janssen Research & Development, LLC.

Trial Registration: ClinicalTrials.gov Identifier: NCT01284387

Declaration of Conflicting Interests: Dr. Honig was an investigator for this study. He received research funding, but no personal compensation for this study. He has received personal compensation from Janssen in his role as a member of the Steering Committee of a different Janssen study. Drs. Ketter, Lu, Di, Novak and Brashear are employees of Janssen Research & Development, LLC and hold stocks in the company. Dr. Liu and Ms. Shadman were at Janssen Research & Development, LLC during the time the study was conducted, but now are affiliated with Prothena and Amgen respectively. Drs. Werth and LePrince Leterme are employees of Pfizer Clinical Sciences. All authors meet ICMJE criteria and all those who fulfilled those criteria are listed as authors. All authors had access to the study data and made the final decision on where to publish these data.

Author Contributions:  Drs. Ketter, Liu, Novak, Brashear and Ms. Shadman participated in the study design, data collection, and conduct of the trial, data analysis plan, interpretation, and review of the manuscript. Dr. Honig was an investigator in this study and involved in patient recruitment, study operation, study management, and data collection, and was the primary scientific reviewer involved in review and finalization of study report. Drs. Werth, LePrince, participated in the data analysis plan, interpretation and review of the manuscript. Dr. Ketter was the study responsible physician and was involved in the execution, data analysis plan, interpretation of data, and review of the manuscript. Dr. Liu was the biomarker lead and had a primary role in the study design and data interpretation. Dr. Brashear was the medical lead for the program and had a primary role in the study design and data interpretation. Dr. Novak was involved in data analysis plan and interpretation. Dr. Lu was the clinical pharmacologist who contributed to the data analysis and review. Dr. Di was the project statistician who made contributions to the study design, data analysis plan, and interpretation of the results.

References

1.    Hardy J, Selkoe DJ. The amyloid hypothesis of Alzheimer’s disease: progress and problems on the road to therapeutics. Science 2002; 297:353-356.
2.    Armstrong RA. The pathogenesis of Alzheimer’s disease: A reevaluation of the «amyloid cascade hypothesis». Int J Alzheimers Dis 2011; 2011:630865.
3.    Pimplikar S.W. Reassessing the amyloid cascade hypothesis of Alzheimer’s disease. Int J Biochem Cell Biol 2009; 41:1261-1268.
4.    Haass C, Selkoe DJ. Soluble protein oligomers in neurodegeneration: lessons from the Alzheimer’s amyloid beta-peptide. Nat Rev Mol Cell Biol 2007; 8:101- 112.
5.    Thal DR, Griffin WS, de Vos RA, Ghebremedhin E. Cerebral amyloid angiopathy and its relationship to Alzheimer’s disease. Acta Neuropathol 2008; 115:599-609.
6.    Citron M. Alzheimer’s disease: strategies for disease modification. Nat Rev Drug Discov 2010; 9:387-398.
7.    Winblad B, Graf A, Riviere ME, Andreasen N, Ryan JM. Active immunotherapy options for Alzheimer’s disease. Alzheimers Res Ther 2014; 6:7.8.    
8.    Arai H, Suzuki H, Yoshiyama T. Vanutide cridificar and the QS-21 adjuvant in Japanese subjects with mild to moderate Alzheimer’s disease: results from two phase 2 studies. Curr Alzheimer Res 2015;12:242-254.
9.    Gin DY, Slovin SF. Enhancing Immunogenicity of Cancer Vaccines: QS-21 as an Immune Adjuvant. Curr Drug ther 2011; 6:207-212.
10.    Hagen M, Seubert P, Jacobsen S, et al. The Aβ peptide conjugate vaccine, ACC-001, generates N-terminal anti-Aβ antibodies in the absence of Aβ directed T-cell responses [abstract]. Alzheimers Dement 2011;7:S460-S461.
11.    Arai H, Suzuki H, Yoshiyama T, et al. Safety, tolerability, and immunogenicity of an immunotherapeutic vaccine (vanutide cridificar [ACC-001]) and the QS-21 adjuvant in Japanese subjects with mild to moderate Alzheimer’s disease: a phase 2a, multicenter, randomized, adjuvant- and placebo-controlled, multiple-ascending-dose study [abstract P1-338]. Alzheimers Dement 2013;.9:282.
12.    Orgogozo JM, Gilman S, Dartigues JF, et al. Subacute meningoencephalitis in a subset of patients with AD after Aβ42 immunization. Neurology 2003; 61:46-54.
13.    Cribbs DH, Ghochikyan A, Vasilevko V, et al. Adjuvant-dependent modulation of Th1 and Th2 responses to immunization with beta-amyloid. Int Immunol 2003;15:505-514.
14.    Nicoll JA, Wilkinson D, Holmes C, Steart P, Markham H, Weller RO. Neuropathology of human Alzheimer disease after immunization with amyloid-beta peptide: a case report. Nat Med 2003; 9:448-452.
15.    Rinne JO, Brooks DJ, Rossor MN, et al. 11C-PiB PET assessment of change in fibrillar amyloid-beta load in patients with Alzheimer’s disease treated with bapineuzumab: a phase 2, double-blind, placebo-controlled, ascending-dose study. Lancet Neurol 2010; 9: 363-372.
16.    Salloway S, Sperling R, Fox NC, et al. Two phase 3 trials of bapineuzumab in mild-to-moderate Alzheimer’s disease. N Engl J Med 2014; 370:322-33.
17.    Gilman S, Koller M, Black RS, et al. Clinical effects of Aβ immunization (AN1792) in patients with AD in an interrupted trial. Neurology 2005; 64: 1553-1562.
18.    Lemere CA, Masliah E. Can Alzheimer disease be prevented by amyloid-beta immunotherapy? Nat Rev Neurol 2010; 6:108-119.
19.    Blennow K, Zetterberg H, Rinne JO, et al. Effect of immunotherapy with bapineuzumab on cerebrospinal fluid biomarker levels in patients with mild to moderate Alzheimer disease. Arch Neurol 2012; 69:1002-1010.
20.    Winblad B, Andreasen N, Minthon L, et al. Safety, tolerability, and antibody response of active Abeta immunotherapy with CAD106 in patients with Alzheimer’s disease: randomised, double-blind, placebo-controlled, first-in-human study. Lancet Neurol 2012; 11:597-604.
21.    Christopher H, van Dyck CS, Leterme GLP, et al.Vanutide cridificar (ACC-001) and QS-21 adjuvant in individuals with early Alzheimer’s disease: amyloid imaging positron emission tomography and safety results from a phase 2 study. J Prev Alz Dis 2016; Accepted.
22.    Salloway S, Sperling R, Gilman S, et al. A phase 2 multiple ascending dose trial of bapineuzumab in mild to moderate Alzheimer disease. Neurology 2009; 73:2061-2070.