Background

BIIB037 is a high-affinity, fully human IgG1 monoclonal antibody against a conformational epitope found on Aβ. It was originally derived by the biotech company Neurimmune in Schlieren, Switzerland, from healthy, aged donors who were cognitively normal. The rationale was that these donors' immune systems had successfully resisted Alzheimer’s disease, and that the operative antibodies could be turned into therapeutics by a process called "reverse translational medicine." BIIB037 binds aggregated forms of Aβ, not monomer. In the brain, BIIB037 preferentially binds parenchymal over vascular amyloid. Thirteen-week chronic dosing in old APP-transgenic mice reduced plaques of all sizes; vascular amyloid stayed unchanged. Brain-exposure studies in mice suggest that microhemorrhages start at 500 mg/kg, more than 100 times the minimally effective dose for plaque clearance, 3 mg/kg (Apr 2013 conference news). 

Findings

A Phase 1 safety and pharmacokinetics study began with a single-ascending-dose trial of 0.3 mg/kg to 30 mg/kg intravenous BIIB037 in 56 people with mild to moderate AD. Participants were assessed at 10 time points up to two years after dosing. Amyloid-related imaging abnormalities (ARIAs) were monitored with four MRI scans, read both locally and by a central reader at an imaging CRO. Side effects included headache, diarrhea, and dizziness. They were mild or moderate, some possibly related to the study drug but not to the dose. No new ARIA developed during the trial beyond age- and AD-related baseline ARIA. In dose-ranging studies, a new dose is given once the safety-monitoring board has deemed the previous dose safe. Because BIIB037 seemed safe at 30 mg/kg, 60 mg/kg was added. This amounts to 4 grams in a 150-pound person. (For comparison, the multiple sclerosis antibody natalizumab [Tysabri®] is administered as a 300-mg infusion.) BIIB037 exposure changed in a linear fashion across doses, with little variability from person to person. Consistent with preclinical work, the antibody generated no plasma spike.

In summer 2012, Biogen Idec started PRIME, a multicenter, multiple-dose study in 166 people with prodromal or mild AD. Besides aducanumab, this trial evaluates the performance of proposed research diagnostic criteria (see Dubois et al., 2010; Albert et al., 2011). Prospective participants must score more than 19 on the MMSE, between 0.5 and 1 on the Clinical Dementia Rating, and 27 or lower on the Free and Cued Selective Reminding Test (FCSRT). They then undergo amyloid PET and MRI scans; a positive amyloid scan is required for entry.

Nineteen imaging centers using 12 different scanner models participated in the aducanumab prodromal trial and sent the scan data to a central reading CRO, where a binary classification method and quantitative standardized uptake value ratios (SUVRs) were compared to establish best methods of interpreting an amyloid scan. Of the first 80 patients scanned, 44 had been diagnosed as prodromal, 36 as mild AD. Of those, the visual read confirmed the clinical diagnosis in 60 percent and the quantitative analysis confirmed the diagnosis in 55 percent of the subjects. The visual read produced no false positives. The false negatives were prodromal cases with a mean MMSE of 27 and an SUVR near the cutoff (Apr 2013 conference news).

In December 2014, Biogen announced it would move into Phase 3 based on interim data suggesting target engagement and a cognitive benefit.

In March 2015, an interim analysis of the first data cut prespecified in the protocol was reported at the AD/PD conference (Mar 2015 conference news). Aducanumab dose-dependently reduced amyloid deposition in six cortical regions of the brain. The effect was large: After one year, the highest dose appeared to have reduced cortical amyloid close to the cut point of positivity. ARIA-E occurred with increasing dose and ApoE4 genotype; about a third of cases were symptomatic, with mild headache and confusion reported in retrospect. Two exploratory signals were reported, as well: Aducanumab appeared to reduce decline on the MMSE and CDR-SB in dose-dependent fashion, although only the 1, 3, and 10 mg/kg doses were reported at AD/PD. The PRIME study will continue collecting data until 2016.

In July 2015, one-year data for the 6 mg/kg dose were presented to have reduced brain amyloid levels and, in exploratory analyses, to have slowed decline on the CDR-SB in accordance with a dose-dependent effect. On the MMSE, the 6 mg/kg group numerically was closer to the 1 mg/kg than the 3 or 10 mg/kg group, but overall dose-dependence remained significant. Instances of ARIA-E increased with dose and ApoE4 carriage, up to 55 percent among ApoE4 homozygotes on 10 mg/kg. Most instances occurred early in the trial, about a third were symptomatic; all resolved (Aug 2015 conference news). In one case later reported after autopsy, a woman who had received 32 monthly doses of antibody up to 6 mg/kg each had little plaque and widespread tangles of low density at autopsy (for paper and expert commentary, see Plowey et al., 2022).

In May 2015, development began in Japan with a Phase 1 study of increasing doses up to 6 mg/kg in 25 patients with mild to moderate AD.

In August 2015, Phase 3 began with two efficacy trials. The 221AD301 ENGAGE study planned to enroll 1,350 people with MCI due to AD or mild AD as ascertained by a positive amyloid PET scan. It aimed to compare monthly infusions of one of three doses of aducanumab or placebo over an 18-month treatment course; the primary outcome measures were cognitive and functional decline per the CDR-SB; secondary outcomes include the MMSE, ADAS-cog13, and the ADCS-ADL mild cognitive impairment version. The trial was set to run until 2022, in 150 centers in North America, Europe, Australia, and Asia. The 221AD302 EMERGE study is identical to ENGAGE; it was conducted in 1,350 additional patients at 131 other sites in North America and Europe. 

In 2016, Biogen formally published the PRIME data (Sevigny et al., 2016), presented data suggesting that a dose titration schedule mitigated ARIA-E, and announced it would be used in Phase 3 (Dec 2016 conference news).

In 2017 and spring 2018, two- and three-year data from the long-term open-label extension phase of the PRIME study were being reported at conferences as continuing to show dose-dependent amyloid removal by up to 70 centiloids, and also as slowing cognitive decline as per exploratory analysis. ARIA in the open-label extension was reported to be similar to ARIA in the placebo-controlled phase of the trial (Dec 2017 and May 2018 conference news).

On March 21, 2019, Biogen and Eisai announced they would terminate all currently ongoing aducanumab trials, following an interim analysis that predicted EMERGE and ENGAGE would miss their primary endpoints (see Mar 2019 news). On April 24, 2019, Biogen announced it would not initiate an anticipated Phase 3 secondary prevention program with aducanumab (Biogen Q1 Update), and removed it from its pipeline (May 2019 conference news).

On October 22, 2019, Biogen announced that the interim futility analysis was wrong, and that subsequent analysis of a larger dataset instead showed EMERGE had met its primary endpoint. People on the highest dose, 10 mg/kg, had declined significantly more slowly on the primary endpoint, the CDR-SB. This group also declined less on secondary endpoints MMSE, ADAS-Cog, and ADCS-ADL-MCI. The low-dose group had some slowing of progression, but the differences were not statistically significant from placebo. 

The ENGAGE trial missed its primary endpoint. An exploratory analysis suggested that a subgroup of people who had received 10 or more 10 mg/kg doses declined more slowly, similar to comparable EMERGE participants. Results of both trials were published after peer review (Budd Haeberlein et al., 2022). A post-hoc analysis attributed the failure of ENGAGE to an imbalance in rapid progressors between treatment groups, and sub-optimal drug exposure (Mallinckrodt et al., 2023).

In substudies to these trials, aducanumab caused a dose-dependent reduction in brain amyloid and some CSF phospho-Tau reduction. As in prior trials, the most common adverse events were ARIA-E and headache. A subsequent analysis of the CSF biomarkers Aβ40, Aβ42, pTau-181, and total tau found robust concordance with amyloid PET status, particularly when biomarker ratios were used (Nisenbaum et al., 2023).

Based on interactions with the FDA, Biogen announced plans to apply in early 2020 for regulatory approval for aducanumab in the U.S. The company also planned to ask eligible patients from EMERGE, ENGAGE, and PRIME to return for renewed dosing and observation (see Oct 2019 news and commentary).

At CTAD in December 2019, Biogen attempted to link cumulative exposure to efficacy with a subgroup analysis of participants who enrolled after a protocol amendment that allowed higher doses for ApoE4 carriers. Post-amendment enrollees received a median cumulative dose of 153 mg/kg, versus 116 mg/kg for pre-amendment enrollees. Later enrollees in both trials were reported to have declined more slowly on the CDR-SB, by 30 percent in EMERGE and by 27 percent in ENGAGE. Secondary outcomes were said to have followed a similar pattern, but data were not shown. Results from a tau PET imaging substudy in 36 patients were shown to suggest a dose-dependent decline in MK6240 tracer uptake in medial temporal brain regions after 14 months of treatment, suggesting aducanumab reduced tangle pathology. The incidence of ARIA was high. One-quarter of participants in the 6 mg/kg dose group and one-third of those receiving 10 mg/kg developed ARIA-E, compared with 10 percent in the placebo group. Seventeen percent of people on drug developed ARIA-H microhemorrhages, compared with 6 percent of controls (Dec 2019 conference news and commentary; see also Muralidharan et al., 2023). 

On January 27, 2020, Biogen listed a Phase 3b open-label study for 2,400 previous aducanumab trial participants, who will receive monthly injections of 10 mg/kg for two years. The trial lists only safety and tolerability parameters as primary outcomes. Called EMBARK, the study was projected to run through September 2023. At the November 2020 CTAD virtual meeting, the company presented more detail on secondary and exploratory endpoints related to long-term efficacy and biomarkers. Endpoints for efficacy would be the same as in EMERGE and ENGAGE, with the addition of the CGIC every six months. Biomarker endpoints will include amyloid-PET, tau-PET and CSF in a subset of participants, and volumetric MRI in all. As of November 2021, the study had enrolled 1,770 participants (Nov 2021 conference news).

On July 8, 2020, Biogen submitted a biologics license application to the U.S. Food and Drug Administration for aducanumab, requesting priority review (Jul 2020 news), and later in the year, filed for approval in the European Union and Japan.

On November 6, 2020, an FDA advisory committee voted against approval, citing weaknesses in the efficacy data, and recommended a confirmatory trial (see Nov 2020 news and Nov 2020 news for commentary). On November 9, Public Citizen lodged a complaint. In February, the FDA delayed its licensing decision until June; in April, the FDA advisory committee renewed its argument against approval (Feb 2021 news; Apr 2021 news).

On June 7, the FDA approved aducanumab under the agency's accelerated approval pathway, which requires substantial evidence of effect on an intermediate marker (amyloid removal), reasonable likelihood of a meaningful clinical benefit, and Phase 4 evidence for such a benefit to be gathered in a subsequent trial, after the marketing license has been granted (FDA statement; Jun 2021 news and commentary). The approval decision has sparked controversy among scientists (e.g. Endpoint News; In The Pipeline blog). 

In June 2021, Biogen began a Phase 1 study to compare Aduhelm’s bioavailability after subcutaneous injection versus intravenous infusion in 30 healthy volunteers. The trial was completed in October 2021.

In July at AAIC, appropriate-use recommendations were issued. They urged selection of patients resembling the RCT population and exclusion of patients with CAA and other cerebrovascular risk factors; they also recommend frequent MRI monitoring (Aug 2021 conference news).

Also at AAIC, an observational study to obtain real-world effectiveness and safety measures was introduced. Called ICARE-AD, the study will follow 6,000 people on Aduhelm for up to five years, aiming for at least 1,000 African-American and Latino participants. Unlike the efficacy RCTs, ICARE-AD will include people with comorbid health conditions. Data will be collected on cognition, function, neuropsychiatric symptoms, quality of life, caregiver burden, cost of dementia care, and safety, particularly ARIA (Aug 2021 conference news). There is no placebo, or other control group of any kind. The study is sponsored by Biogen but will rely on insurers and public funds to cover Aduhelm and other costs; it will likely take up to 10 years.

At the 2021 CTAD conference, scientists reported that plasma analysis from the Phase 3 trials showed a reduction in pTau with Aduhelm (Nov 2021 conference news). In December 2021, safety data from the Phase 3 EMERGE and ENGAGE trials was published (December 2021 news; Salloway et al., 2021). Approximately one-third of people taking Aduhelm developed ARIA; a quarter of those were symptomatic; 3 percent were serious. Participants with brain microhemorrhages at baseline or APOE4 were at greater risk.

On December 16, the European Medicines Agency rejected a marketing application for Aduhelm (December 2021 news); later that month, regulators at Japan's Health Ministry indicated they would likely do the same. 

In April 2022, the U.S. Centers for Medicare and Medicaid Services (CMS) issued a national coverage decision restricting Aduhelm coverage to clinical trials (see Apr 2022 news; Feb 2022 news; Jan 2022 news and commentary). Soon after, Biogen withdrew its marketing application before the EMA (press release) and announced cutbacks in its Aduhelm marketing efforts but a continuation of some ongoing Aduhelm clinical studies (May 2022 news). On May 20, the company terminated the ICARE-AD real-world use study.

In June 2022, the Phase 4 confirmatory trial required by the FDA began. Called ENVISION, it was to enroll 1,500 people with early AD and randomize them to drug or placebo for 18 months (press release), with a long-term extension of up to four years. The primary endpoint was going to be the CDR-SB; additional standard cognitive and functional measures, and amyloid and tau PET were planned as secondary outcomes. The trial aimed to recruit at least 18 percent of participants in the U.S. from black and Hispanic communities. Results were expected by 2026.

In January 2023, a U.S. Congressional investigation criticized the FDA and Biogen over their approval process for Aduhelm (Jan 2023 news; Tanne 2023).

On January 31, 2024, Biogen announced it would stop marketing Aduhelm, and terminate the ENVISION trial (press release; Feb 2024 news). The rights to the drug will revert to Neurimmune.

Research using aducanumab to understand its treatment effects continues, for example to investigate therapeutic blood-brain barrier opening by way of ultrasound (Rezai et al., 2024; Nov 2023 conference news). Other work explores the duration of, and microglial response to, amyloid removal after drug discontinuation, or the potential for co-pathologies such as Lewy bodies to interfere with amyloid removal (Cadiz et al., 2024, Millet et al., 2024; VandeVrede et al., 2023). One center reported its experience of ARIA management after aducanumab approval (Howe et al., 2023). A genome-wide association study in ENGAGE and EMERGE participants aimed to identify genetic risk factors for ARIA, and found the only significant association to be with ApoE4. In this population, E4/E4 homozygotes were at a 4.28-fold increased risk for ARIA, and a greater risk of radiographically severe ARIA, but no increased risk for symptomatic ARIA (Loomis et al., 2024). A case report describes an ApoE4 carrier who developed ARIA in the aducanumab trial, and who subsequently showed an adverse inflammatory reaction to COVID19 and flu vaccines (Restifo and Erikson, 2023).

To view registered clinical trials on aducanumab, see clinicaltrials.gov.

Clinical Trial Timeline