Part 1 of 3
While much of the Alzheimer’s world was enthralled by the drama of the FDA advisory committee’s shellacking of aducanumab’s application for marketing approval (Nov 2020 news) the Clinical Trials in Alzheimer’s Disease (CTAD) conference, held November 4-7, featured encouraging data and ambitious plans for a different monoclonal anti-Aβ antibody—BAN2401.
- In an ongoing open-label extension, plaques plummeted in newly dosed volunteers; ARIA-E was manageable.
- A Phase 3 trial for people with early AD is nearly fully enrolled.
- AHEAD 3-45, a pair of trials for preclinical AD, is recruiting.
As seen at the virtual meeting, participants in an open-label extension of BAN2401’s large Phase 2 trial, who had previously taken placebo, saw large drops in brain amyloid, while those who had previously received high doses maintained low amyloid levels. Meanwhile, Eisai has enrolled more than 1,200 people into CLARITY-AD, a Phase 3 trial evaluating the antibody’s efficacy in people with MCI or mild AD. For its part, the public-private Alzheimer’s Clinical Trial Consortium started two studies—a Phase 2 and a Phase 3—aimed to tease out effects of this biologic drug in two stages of preclinical AD, designated by amyloid load (see Part 2 of this story). And separately, after years of slow development, plus some interference from COVID-19, a trial-ready cohort is finally enrolling amyloid-positive participants for secondary prevention trials (see Part 3).
Trained against large, soluble aggregates of Aβ called protofibrils, BAN2401 was developed by BioArctic, subsequently licensed by Eisai for clinical development, and later partnered with Biogen. Eisai used an ambitious, adaptive design for its 856-participant Phase 2 study, which shifted new enrollees into more effective-looking dose arms as the trial progressed. In that trial, BAN2401 missed its primary cognitive endpoint at 12 months but, by 18 months, had dramatically reduced amyloid and also slowed cognitive decline in the highest dose group (Dec 2017 news and Jul 2018 news).
Complicating the interpretation was that this highest dose group had few ApoE4 carriers, because regulators had ordered the sponsor midway through the trial to stop randomizing them to that dose. The concern was their incidence of amyloid-related imaging abnormalities due to edema (ARIA-E). Subsequent analyses reaffirmed the treatment benefit despite this snafu (Nov 2018 news).
Spurred by this result, Eisai invited participants who had completed the placebo-controlled portion of the trial back to join a 24-month open-label extension. All participants in this OLE received the highest dose of BAN2401—10 mg/kg twice monthly. They had been off treatment for an average of two years, with the gap ranging from nine to 60 months. At last year’s CTAD, Eisai had reported that despite this interruption, amyloid plaques remained at bay for those who had received the two highest doses of BAN2401—10 mg/kg either monthly or twice-monthly (Dec 2019 conference news). Alas, while amyloid plaques stayed away, cognition once again slipped during the gap period, suggesting, perhaps, that the antibody’s continued activity against protofibrillar forms of Aβ was needed for a cognitive benefit.
At this year’s CTAD, Eisai’s Chad Swanson showed 12-month brain-imaging data from the OLE. Of the 180 people who are getting BAN2401 in the extension, 38 had previously been on placebo, 60 on 10 mg/kg biweekly, 45 on 10 mg/kg monthly, and 37 had received one of the lower doses of BAN2401. Swanson presented longitudinal amyloid-PET imaging data for a subset of 76 participants who had gotten either placebo or one of the two highest doses in the core trial.
For previous placebo recipients, their longitudinal scans indicated a steep drop in amyloid between baseline and the first three months of the OLE, with a continued lowering at 12 months. In these treatment-naïve participants, amyloid fell by 0.33 SUVR over 12 months, from 1.38 to 1.05 SUVR, dipping below the threshold for brain amyloid positivity of 1.1 SUVR.
People who had received monthly infusions of 10mg/kg BAN2401 during the core study started the OLE with an average SUVR of 1.22. Over the 12-month extension, it dropped by a further 0.12 SUVR to 1.1. Those who had received 10 mg/kg twice monthly in the core study started the OLE with 1.08 SUVR, and inched down further to 1.04 12 months later.
What about ARIA-E in the OLE? The incidence of this side effect mirrored data from the placebo-controlled portion of the study. ARIA-E cropped up in 14 people, or 7.8 percent of all participants. As expected, it was more frequent, at 8.9 percent, among people who had previously received placebo. Thirteen of the 14 cases occurred in ApoE4 carriers, most within the first three months of the OLE. Three of the 14 cases experienced symptoms of ARIA-E, and all resolved within four to 12 weeks.
While the core study had automatically stopped treatment for all ARIA-E cases, the researchers took a less conservative approach in the OLE. They continued to dose six participants who had mild to moderate ARIA-E, as gauged by radiographic intensity on MRI, and this caused no ill effects.
Larisa Reyderman of Eisai presented an ARIA-E interpretation from both the trial’s core and OLE portions. Based on pooled ARIA-E incidence statistics from Phase 1 and Phase 2 studies of BAN2401, Reyderman developed a pharmacokinetic/pharmacodynamic model that predicted ARIA-E incidence in the OLE. Essentially, Reyderman reported that exposure to BAN2401—as gauged by Cmax, its maximum serum concentration—was the primary predictor of ARIA-E. In addition, ApoE4 carriers had a 26 percent higher risk of ARIA-E than noncarriers. Overall, the Eisai researchers concluded that incidence of ARIA-E was low, and that milder cases could stay on the antibody rather than leave the trial.
Swanson told Alzforum that 22 of out of the 180 participants in the OLE, or 12.2 percent, had ARIA-H, another imaging abnormality caused by microhemorrhages. In the core study, ARIA-H did not appear to track with BAN2401 dose, as 5.3 percent of people in the placebo group and 6.8 percent of those in the 10mg/kg biweekly group had ARIA-H. For those in one of the lower BAN2401 dose groups, ARIA-H incidence ranged from 11 to 18 percent.
This data informed the design of subsequent Phase 3 studies of BAN2401, which will allow continued dosing of people with mild ARIA-E. For more on that, see Part 2 of this story. —Jessica Shugart
- FDA Advisory Committee Throws Cold Water on Aducanumab Filing
- BAN2401 Forges AHEAD into Phase 3, Preclinical AD
- TRC-PAD Funnel Finally Touches Down
- No Man’s Land: Neither Early Success nor Failure for BAN2401
- BAN2401 Removes Brain Amyloid, Possibly Slows Cognitive Decline
- Second Look at BAN2401 Data Still Positive, Despite Snafu
- Amyloid Clearance: Check. Cognitive Benefit: Um … Maybe.
No Available Further Reading