Eli Lilly & Co. is indicating there may be a path forward for its foundering Alzheimer’s drug solanezumab through FDA programs aimed at speedy approval of drugs with potential for dramatic therapeutic impact, after announcing Oct. 8 that a pooled analysis from two otherwise failed Phase III trials showed patients with mild disease had a statistically significant 34% decrease in the rate of cognitive decline.

In August, the company announced that solanezumab, sola for short, had failed to meet cognitive and functional primary endpoints in the EXPEDITION I and EXPEDITION II trials, but attempted a positive spin by suggesting that a pre-specified analysis of pooled data had revealed statistically significant results in mildly affected patients and that numbers would be forthcoming ([A#14120824002]).

With more specifics now public, the company will be taking the data to regulators in the U.S and globally to discuss next steps, Dave Ricks, president of Lilly Biomedicines, said during a webcast to go over the results. While conducting a confirmatory Phase III trial in the mildly affected population would ostensibly be the chief topic in those conversations, Ricks hinted at the possibility of regulatory filings based on the current data.

“Legislators have given these regulators tools with which to approve drugs in these kinds of circumstances,” Ricks said. For that reason, Lilly will “not remove any option from the table.”

“The data will have to speak for itself, and [regulators will] have to make the judgment, not us, in terms of whether that’s the appropriate benefit to society,” Ricks said.

Legislators have given FDA “tools with which to approve drugs in these kinds of circumstances,” Lilly’s Rick said.

A Test Of FDA’s New “Breakthrough” Designation?

Consensus in the investor community is that the company will conduct a new Phase III trial for sure, but few believe Lilly could get an approval based on the data it has in hand.

“With what looks like hypothesis-generating data following two studies that failed to hit their primary endpoints and no predefined path toward accelerated approval, we believe it is most likely that Lilly will have to conduct at least one additional Phase III trial in 1,500-2,000 mild patients over the course of two to three years,” Leerink Swann analyst Seamus Fernandez said in an Oct. 9 note to investors.

Still, the unmet need is frightening, and sola could easily develop the cachet that catapulted the novel prostate cancer therapy Provenge onto the public stage with patient demonstrations and congressional interest after FDA asked for more data to support approval of the immunotherapeutic. FDA stood its ground, Dendreon Corp. completed an additional trial and Provenge was eventually approved three years after the initial rebuff Also see "Provenge Nod Lifts Spirits And Stocks, But Also Sets High Bar For Success" - Pink Sheet, 17 May, 2010.. The Provenge odyssey became a reminder of the high bar for clinical and regulatory success in uncharted territory.

A testament to the pent up hope in the Alzheimer’s community is the fact that 95% of people who completed the EXPEDITION trials (the dropout rate for the 2,000-patient program was around 25%) rolled over into the open-label extension study, with no patients lost to follow-up.

The FDA Safety and Innovation Act, signed into law in July, incorporated a number of provisions aimed at speeding drug approvals for novel medicines. In part, it codified FDA’s existing accelerated approval and urged broadened use of the mechanism. In addition, it introduced the “breakthrough” designation, which promises expedited development and review for drugs that make substantial improvement in addressing a serious or life-threatening disease Also see "PDUFA V: Accelerated Approval Expansion May Outshine Rare Disease Improvements" - Pink Sheet, 17 September, 2012..

Rich commented only that legislator-designed pathways for approval could include a conditional approval “based on just making an exception with commitment for follow-up or based on interim looks at future studies,” adding that “no one has had those conversations with a disease modifier in Alzheimer’s.”

Pointing to precedents, notably in oncology, Credit Suisse analyst Catherine Arnold said in Oct. 9 comments that “given the strength of the data and the unmet need of the condition, it is possible that regulatory authorities could show flexibility and create a pathway for the agent to come to market sooner than expected.”

The Rationale For Pooling

The pooled analysis presented by Lilly showed a 34% reduction in the slope of incline (a graphic representation of cognitive decline as the disease progresses) for mildly affected patients after 18 months of monthly intravenous infusions. Interestingly, while the data didn’t reach significance on the functional measure, activities of daily living, it did show significance on a subset of the more difficult tasks on that scale.

The data come from a prespecified secondary analysis that teased mild patients out of the mild-to-moderate population enrolled in the two pivotal trials that failed to demonstrate overall efficacy on both cognitive and functional endpoints.

Pooling was justified because the study designs were very similar, with a great deal of consistency in their findings, said Lilly’s Siemers.

EXPEDITION I, which Lilly said was identical to EXPEDITION II except for some geographical differences, finished first. The subgroup analysis showed a statistically significant 42% decline in cognitive decline for mild patients compared to placebo, but the effect on function was not statistically significant.

Based on those results, Lilly modified the statistical analysis plan for EXPEDITION II to specify a single primary endpoint of cognition in mild patients and upgraded the test measuring cognition from ADAS-Cog 11 to ADAS-Cog 14, adding three questions to be more sensitive for that group. In the end, the trial showed a 20% reduction in cognitive decline that was not statistically significant, as well as a lack of significant difference compared to placebo on function.

Both studies were statistically powered for the combined population, about a third of which was patients with moderate disease. Once those patients were dropped from the analysis, the individual studies were underpowered to show an effect, Eric Siemers, senior medical director of Lilly’s Alzheimer’s disease team, said during the webcast. However, because the studies were so similar, with a great deal of consistency in the findings, Lilly decided it would be “quite justified and valid” to pool the data and regain statistical powering, he said.

At 18 months, the combined mild patients taking placebo had a 6.21 point average change on the ADAS-Cog scale from baseline, while patients taking sola had a 4.08 point average change, a statistically significant difference of 2.13 points or 34% (p=.001).

On the more demanding subset of instrumental activities of daily living (the iADL), at 18 months the placebo group had a 6.77 point loss, while the sola group had a 5.56 point loss, a statistically significant difference of 17.8% (p=.045).

The Lilly webcast followed on a session earlier that day at the American Neurological Association meeting in Boston, where a post-hoc independent third-party analysis of the sola data was presented by the Alzheimer’s Disease Cooperative Study. Lilly agreed to provide ADCS with the full raw data set from the EXPEDITION program for independent analysis before the company was aware that both trials had failed in the overall population, and ADCS has control over its publication. The two presentations used different statistical methods but did not differ meaningfully.

ADCS will be reporting on biomarker data from the EXPEDITION program at the Clinical Trials on Alzheimer’s Disease meeting Oct. 29 in Monte Carlo, Monaco, as well as at subsequent medical meetings. Among the results to be reported in Monte Carlo are data from the Clinical Dementia Rating Scale Sum of Boxes score, which combines cognitive and functional measures to provide a global assessment of the severity of the disease.

Alzheimer’s disease is incorrectly diagnosed approximately 20% of the time because the only way to identify whether a patient’s illness is associated with amyloid plaque buildup has been by autopsy. Lilly has been at the forefront of companies developing new imaging agents to solve that problem. In April FDA approved Lilly’s Amyvid (florbetapir F 18), an imaging agent from the company’s Avid Radiopharmaceuticals Inc. unit, for use with Positron Emission Tomography to estimate β amyloid plaque density. The company has had an uphill fight for reimbursement, and CMS recently opened a national coverage analysis to consider Medicare reimbursement for the agent (see sidebar).

[Sidebar#00121015018]

Solanezumab is a humanized anti-β amyloid monoclonal antibody that binds soluble Aβ and promotes its clearance from the brain before it can become part of an amyloid plaque. Siemers said Lilly was able to do Amyvid-enabled PET imaging on only about 13% of patients in the global trials because the program was started a few years ago, but that those results mirrored the anticipated outcome that about 20% of the patients would have negative scans. If anything, the percentage of people who probably did not have Alzheimer’s “would tend to mute or attenuate the signal we saw,” he said. “There’s always that in the background.”

Banking On A Failed Drug

The new analyses were met with a mixture of hope and doubt, as the value of the company’s stock shot up 5.29%, – adding roughly $3 billion to Lilly’s market cap in one day – and stakeholders combed available data for clues to support optimism.

An approval based on the current data is a Hail Mary play for Lilly. Most observers believe at least one new trial is called for and some point out that reimbursement for what is likely to be a very expensive drug may depend on sola being backed by a more substantial positive data package. A trial with a treatment period of more than 18 months, for example, might disclose a more convincing change in the slope of cognitive decline over time.

But, even if the company goes the confirmatory trial route, with success sola could be on the market in 2017, and it will be a multi-billion dollar drug. As BernsteinResearch analyst Tim Anderson pointed out in an Oct. 9 report, even Eisai Co. Ltd.’s Aricept (donepezil) an older, barely effective drug, managed to sell around $4 billion a year despite only improving symptoms and having no effect on disease progression.

Efficacy Signal For Lilly’s Solanezumab Stirs The Alzheimer’s Pot

More from Clinical Trials

More from R&D