Roche set to benefit from FDA effort to tackle Alzheimer’s paradox

Whatever doubts still remain around the relevance of the secondary analysis of Lilly’s phase III study of solanezumab, it is clear that the trial highlighted the value of treating Alzheimer’s patients early in the course of their disease.

Now the US FDA has put out for comment a draft document suggesting ways to incentivise pharma to target these early forms of dementia. Although patients with early Alzheimer’s might be easier to treat, they are hard to identify, and standard cognitive and functional endpoints are of little use here – a paradox that the agency attempts to tackle. An early beneficiary of this stance might be Roche.

Yesterday’s draft guidance makes four key concessions, proposing the use of a sole cognition endpoint, a composite scale and/or biomarkers, and suggesting a novel study design. Only a handful of novel agents are in the clinic specifically in early stages of the disease, and the upper hand of Roche’s gantenerumab is now clear: not only is it one of the most advanced, but the design of its phase III study seems to have pre-empted the FDA document.

There are also lessons for companies working on more advanced Alzheimer’s, where the agency reiterates its requirement for an effect on cognitive and functional domains. This bears out Lilly’s decision not to file solanezumab on a secondary analysis in cognition alone (Confirmatory sola study plan takes Lilly down a peg, December 12, 2012).

Co-primary impractical

However, the co-primary cognitive/functional measure might prove impractical, and the FDA concedes that delaying cognitive impairment alone might be enough to prove effectiveness in so-called pre-clinical Alzheimer’s – a very early form of the condition.

Meanwhile, patients with prodromal Alzheimer’s or mild cognitive impairment – conditions closer to overt dementia – often do feature functional impairment, and here the FDA suggests the use of a composite scale like CDR-SB sum-of-boxes score. This is surprising, since CDR-SB is a tool by which doctors make an initial assessment, and has not been viewed as having specific relevance as a study endpoint. The agency says CDR-SB allows for smaller and shorter studies, and downplays the use of measuring time to diagnosis of dementia.

It suggests that a meaningful effect on biomarkers could make out a disease modification claim, but cautions that approval on the basis of a surrogate endpoint will not be possible until there is widespread academic agreement that a single biomarker is likely to predict a clinical benefit. However, it says a biomarker-based secondary endpoint might be enough to support a single primary outcome measure.  

The agency, however, cannot help find the right patients for a study; the recruitment of those with mild cognitive impairment in early Alzheimer’s trials is fraught with difficulty as many will not in fact develop the full disease. While the FDA cannot endorse any specific diagnostic approaches, in principle it supports trials being enriched with patients most likely to progress to full Alzheimer’s on the basis of clinical criteria and biomarkers.

The FDA also suggests that a randomised-start study design can be considered in some cases, even though such studies have previously played only a supportive role.

Early beneficiary

Yesterday’s guidance supports the design of the 770-patient phase III trial of gantenerumab in prodromal Alzheimer’s, since its co-primary endpoints comprise change in CDR-SB and change in brain amyloid – a biomarker.

Biomarkers feature exclusively in the primary endpoints of a trial recently initiated by the Alzheimer’s Association, investigating the possibility of preventing dementia with gantenerumab or solanezumab. In contrast, studies of Baxter’s Gammagard in mild to moderate Alzheimer’s have tended to focus on traditional functional and cognitive endpoints.

As the FDA’s 2007 guideline on obesity drug development showed, these documents do not form an unbreakable rule. Nevertheless, developers of several novel agents in phase II in early Alzheimer's, including Novartis (CAD106 beta amyloid vaccine) and GlaxoSmithKline (AC02), will pay close attention to the FDA's developing stance.

News on Lilly’s design of the confirmatory solanezumab trial will also be keenly awaited.

To contact the writer of this story email Jacob Plieth in London at[email protected]or follow@JacobEPVantageon Twitter