乌鸦传媒

LOS ANGELES -- Intranasal insulin to treat Alzheimer's disease showed puzzling results in the federally funded, phase II/III study presented .

While the drug had no effect on cognition in the study's primary intention-to-treat analyses, a mid-trial change in the device that delivered insulin may have affected results.

"Treatment with intranasal insulin to deliver insulin directly to the brain showed a clinically significant benefit for participants who were treated with an effective device," reported Suzanne Craft, PhD, of the Wake Forest School of Medicine, in a plenary session at the 2019 Alzheimer's Association International Conference (AAIC).

At the end of 18 months, these participants had a 6-point advantage in ADAS-Cog 12 scores compared with people originally assigned to placebo, which translates to a slowing of Alzheimer's progression of 1 to 2 years, she added.

In the brain, insulin carries out a number of different functions: it enhances communication between neurons, increases levels of chemicals in the brain, improves blood flow, and protects against abnormal amyloid-beta (Aβ) and tau, Craft explained.

"Previous work has shown that there are either low levels of insulin in Alzheimer's disease or it does not work effectively," she added. "We asked the question of whether restoring normal brain insulin function would improve the symptoms and the pathology of Alzheimer's disease."

In SNIFF, participants started out using the same device that had been used in earlier Alzheimer's studies (device 1, Kurve ViaNase), but to make it easier for people with cognitive impairment to use, the manufacturer added an electronic timer. "The timer kept malfunctioning," Craft told 乌鸦传媒. "It was tricky to get the device to start and participants became frustrated."

Partway through the trial, the investigators switched to another type of device (device 2, Impel NeuroPharma), one that had never been used for Alzheimer's research. A total of 49 participants used device 1 and 240 participants used device 2.

Participants in SNIFF all had mild cognitive impairment or Alzheimer's disease. They were treated daily with 40 IU of intranasal insulin or placebo in a 12-month blinded trial and continued similar dosing in a 6-month open-label extension. The study's primary endpoint was the ADAS-Cog 12 score, measured quarterly. Device 2 was pre-specified for the primary analyses and device 1 for secondary analyses.

At the end of the initial 12-month study, device 2 showed no benefit for insulin. Device 1 participants who received insulin, however, trended toward improved ADAS-Cog 12 scores (P=0.091), in line with an pilot study.

Most SNIFF participants, including 44 of 49 people in the device 1 group, entered the open-label extension. Here, insulin-treated participants in the device 1 group who continued in the trial showed better ADAS-Cog 12 scores at months 15 and 18 (-5.70 and -5.78 points; P=0.004 and 0.018) than participants originally assigned to placebo, and better ADL-MCI scores at month 18 (4.85 points; P=0.047). Device 2 results did not differ between treatment and placebo on ADAS-Cog 12 or other outcomes.

"This is a large effect in the device 1 group, particularly as it is on top of background treatment," Craft said. "At 18 months, the results show a prolonged, statistically significant benefit for the insulin-treated group who used device 1 that strengthens over time, with a pattern consistent with a disease-modifying effect."

Cerebrospinal fluid analysis showed the Aβ42/40 ratio (P=0.01) and Aβ42/tau ratio (P=0.03) also improved for people treated with device 1. (Aβ42 and Aβ40 are the 42- and 40-amino acid forms of the amyloid-beta protein.)

"It's encouraging to see that if you do get insulin to the brain, you see changes in biomarkers," said Laurie Ryan, PhD, Chief of the Dementias of Aging Branch of the National Institute on Aging, which funded the study. "That's important, but it needs to be tested," she told 乌鸦传媒. "It's got to go to a larger phase III study."

The two intranasal devices may differ in their ability to deliver insulin to the brain, Craft noted. The researchers tested acute delivery by dosing then measuring CSF insulin for device 1, but not for device 2.

"It's hard to know about device 2, because we have not done the proof-of-concept study that we did with the first device," Craft said. "They work through different mechanisms: the first device is a nebulizer-like system; the second device is not. What needs to happen with the second device is the kind of validation study we did with the first."

"We understand now what due diligence has to occur if you're going to deliver something intranasally," she continued. "For the phase III study, we will have that nailed down." That trial should start in the spring of 2020, she said.

Comment