FUNDAMANT: an interventional 72-week phase 1 follow-up study of AADvac1, an active immunotherapy against tau protein pathology in Alzheimer's disease

Abstract

Background: Neurofibrillary pathology composed of tau protein is closely correlated with severity and phenotype of cognitive impairment in patients with Alzheimer's disease and non-Alzheimer's tauopathies. Targeting pathological tau proteins via immunotherapy is a promising strategy for disease-modifying treatment of Alzheimer's disease. Previously, we reported a 24-week phase 1 trial on the active vaccine AADvac1 against pathological tau protein; here, we present the results of a further 72 weeks of follow-up on those patients.

Methods: We did a phase 1, 72-week, open-label study of AADvac1 in patients with mild to moderate Alzheimer's disease who had completed the preceding phase 1 study. Patients who were previously treated with six doses of AADvac1 at monthly intervals received two booster doses at 24-week intervals. Patients who were previously treated with only three doses received another three doses at monthly intervals, and subsequently two boosters at 24-week intervals. The primary objective was the assessment of long-term safety of AADvac1 treatment. Secondary objectives included assessment of antibody titres, antibody isotype profile, capacity of the antibodies to bind to AD tau and AADvac1, development of titres of AADvac1-induced antibodies over time, and effect of booster doses; cognitive assessment via 11-item Alzheimer's Disease Assessment Scale cognitive assessment (ADAS-Cog), Category Fluency Test and Controlled Oral Word Association Test; assessment of brain atrophy via magnetic resonance imaging (MRI) volumetry; and assessment of lymphocyte populations via flow cytometry.

Results: The study was conducted between 18 March 2014 and 10 August 2016. Twenty-six patients who completed the previous study were enrolled. Five patients withdrew because of adverse events. One patient was withdrawn owing to noncompliance. The most common adverse events were injection site reactions (reported in 13 [50%] of vaccinated patients). No cases of meningoencephalitis or vasogenic oedema were observed. New micro-haemorrhages were observed only in one ApoE4 homozygote. All responders retained an immunoglobulin G (IgG) antibody response against the tau peptide component of AADvac1 over 6 months without administration, with titres regressing to a median 15.8% of titres attained after the initial six-dose vaccination regimen. Booster doses restored previous IgG levels. Hippocampal atrophy rate was lower in patients with high IgG levels; a similar relationship was observed in cognitive assessment.

Conclusions: AADvac1 displayed a benign safety profile. The evolution of IgG titres over vaccination-free periods warrants a more frequent booster dose regimen. The tendency towards slower atrophy in MRI evaluation and less of a decline in cognitive assessment in patients with high titres is encouraging. Further trials are required to expand the safety database and to establish proof of clinical efficacy of AADvac1.

Trial registration: The studies are registered with the EU Clinical Trials Register and ClinicalTrials.gov : the preceding first-in-human study under EudraCT 2012-003916-29 and NCT01850238 (registered on 9 May 2013) and the follow-up study under EudraCT 2013-004499-36 and NCT02031198 (registered 9 Jan 2014), respectively.

Keywords: Active immunotherapy; Alzheimer’s disease; Clinical trial; Immunotherapy; Neurofibrillary pathology; Tau; Vaccine.

Fig. 1

Antibody response to AADvac1 over 96 weeks of treatment. Values shown indicate geometric mean titres and 95% CI of the geometric mean. Time points of AADvac1 administration indicated by bold treatment weeks. Light grey area indicates values obtained in the first-in-human study

Fig. 2

The antibody response to AADvac1 is immunoglobulin G1-dominated. Values were obtained after the sixth dose of AADvac1

Fig. 3

AADvac1-induced antibodies label tau extracts from corticobasal degeneration (CBD) and progressive supranuclear palsy (PSP) brains. Left: Patient R17, pre-treatment serum. Right: Serum of the same patient after six doses of AADvac1. The serum labels both high-molecular-weight aggregates and low-molecular-weight fragments of tau protein. Staining of Alzheimer’s disease (AD) brain extract is shown as a positive control

Fig. 4

Sera of patients treated with AADvac1 recognise tau pathology in Alzheimer’s disease and Pick’s disease. a–f Alzheimer’s disease. g–l: Pick’s disease. Sera of three different patients with different strengths of antibody responses (patient R17 with an anti-AD-tau titre of 1:30,999; patient R25 with 1:18,185; and patient R10 with 1:12,800) were used for staining. Staining with pre-treatment sera is shown as a negative control (d–f, j–l)

Fig. 5

Sera of patients treated with AADvac1 recognise tau pathology in progressive supranuclear palsy (PSP) and corticobasal degeneration (CBD). a–d PSP. g–l CBD. Sera of three different patients with different strengths of antibody response (patient R17 with an anti-AD-tau titre of 1:30,999; patient R25 with 1:18,185; and patient R10 with 1:12,800) were used for staining. Staining with pre-treatment sera is shown as a negative control (d–f, j–l)

Fig. 6

Change in cognition (Alzheimer‘s Disease Assessment Scale 11-item cognitive assessment [ADAS-Cog11], Controlled Oral Word Association Test [COWAT]) over 96 weeks, displayed as correlation with the immunoglobulin G (IgG) titre AUC. Results are shown for completers (left) and for patients with a positive biomarker profile (right). Results are shown with raw AUC values (a–d) and with AUC values corrected for disease severity (e–h). Category Fluency Test results were inconclusive (not shown)

Fig. 7

Hippocampal atrophy over 96 weeks, displayed as correlation with the immunoglobulin G (IgG) titre AUC. Results are shown for completers (a, b) (n = 18) and for patients with a positive biomarker profile (c, d) (n = 9). Results are also shown with raw AUC values (a, c) and with AUC values corrected for disease severity (b, d). One patient with frontotemporal dementia and one patient with poor magnetic resonance imaging scan quality were excluded. HCV Hippocampal volume