Full BAN2401 reveal gives Biogen a reality check
Data on Alzheimer's disease project spoiled by patient exclusions.
Holding shares going into a full data presentation on an Alzheimer’s disease candidate continues to be a losing investment strategy. Biogen’s backers were punished this morning after the 18-month analysis of BAN2401’s phase II trial revealed a mixed dose response that appeared to be driven by the most rapidly declining patients being excluded from the highest-dose group – the one that generated the best results.
Biogen’s valuation shed $7bn today, erasing nearly half of the gains it had made since the group announced a positive finding from the 18-month analysis on July 6. Anybody making the long-shot bet that these data would be sufficient for regulatory filing was disappointed; at best, the results only point the way to phase III design, and regulators will be asking many questions about that.
Seemingly forgotten in the hype of the past three weeks over the Eisai/Biogen anti-amyloid beta antibody is that Study 201 had already failed in the primary analysis, and that yesterday’s results could therefore only be said to be exploratory, even when the traditional statistical significance threshold of 0.05 was met.
The most encouraging finding was that the highest dose of 10mg/kg every other week slowed decline on the as-yet-unvalidated ADCOMS composite endpoint by 30% compared with placebo. This result exceeded expectations – but several red flags popped up during the presentation at the Alzheimer’s Association International Conference in Chicago.
For one, three doses, 5mg/kg and 2.5mg/kg every other week and 5mg/kg monthly, performed worse than placebo, a finding that the companies had apparently withheld from an embargoed press briefing before the full data were presented at the AAIC last night.
Furthermore, the result in the composite ADCOMS endpoint appears to be driven by one component, ADAS-COG, where the 10mg/kg dose showed a 47% reduction. A second component, CDR-SB, missed statistical significance.
And further still, the positive findings for the 10mg/kg fortnightly dose have to be considered in the context of the exclusion of ApoE4 carriers specifically from this dose group. This was done because EU regulators were concerned about amyloid-related imaging abnormalities, a sign of cerebral oedema.
ApoE4 carriers are thought to show cognitive decline more quickly than non-carriers, meaning that the cohort getting the highest BAN2401 dose was loaded with patients less likely to show a decline.
Placebo progresses faster
“You have a placebo group that is primed to progress faster,” Matthew Barrett, associate professor and clinical director of the aging and memory care clinic at the University of Virginia, told EP Vantage.
In a call with analysts, Biogen executives said subgroup analyses excluding ApoE4 carriers, which would enable a proper comparison, had not been completed.
Despite all of this, Professor Barrett believes that the data are sufficient to justify advancing BAN2401 into phase III – although a victory here is far from certain. “This is certainly a therapeutic strategy that has not yet failed or been proven to be effective,” he said. “This latest result gives reason for optimism.”
It is almost understandable why investors get caught up in the frenzy surrounding an Alzheimer’s disease candidate, given the likely mega-blockbuster status that would be earned by any product that managed to slow patient decline. So far, however, the record favours the sceptics; the amyloid hypothesis should have been dead and buried by now, and persists only because few other explanations exist.