Still No Clear Answer on Intranasal Insulin for MCI, Alzheimer's

Michael Vlessides

July 10, 2020

A new multicenter trial has yielded conflicting results regarding intranasal insulin's ability to deliver cognitive and functional benefit for patients with mild cognitive impairment (MCI) and Alzheimer's disease (AD).

The randomized trial of nearly 300 patients showed that although one insulin administration device produced marked benefit in terms of change in mean score on the Alzheimer Disease Assessment Scale–Cognitive Subscale 12 (ADAS-cog-12) over 12 months, reliability was inconsistent.

A second device, used on the majority of patients in the study's intention-to-treat population, showed no difference in these measures between patients who did and those who did not receive intranasal insulin.

"The primary analysis of the study showed no benefit of intranasal insulin on any measures of cognition or cerebrospinal fluid Alzheimer's disease biomarkers when using the new device," principal investigator Suzanne Craft, PhD, told Medscape Medical News.

"But when we looked at our planned secondary analysis with the original device ― which has been successful in previous studies ― we saw quite a different picture," added Craft, director of the Alzheimer's Disease Research Center at Wake Forest School of Medicine, Winston-Salem, North Carolina.

"We found a pronounced benefit with that device, such that after 18 months of administration, participants who had been receiving insulin from the beginning of the study had a large and clinically significant advantage in the primary outcome measure."

Craft described the findings as complex. "The primary results were negative," she added. "But the secondary results replicated those of several earlier studies when we used the same device that was used in those."

The study was published online June 22 in JAMA Neurology.

Important for Brain Function

Insulin has been shown to play several important roles in brain function. The hormone is associated with a variety of cognitive functions, including memory. Through its association with vasoreactivity, lipid metabolism, and inflammation, insulin also plays an important role in vascular function.

"In the normal brain in healthy individuals, insulin is very important for synaptic function and viability. Insulin also promotes dendritic growth and facilitates synaptic health. Through this role, it plays an important part in memory," said Craft.

Given these connections, it is not surprising that reduced insulin levels or activity in brain and cerebrospinal fluid have been documented in some, but not all, studies of AD.

Markers of insulin resistance have also been detected in both neuronally derived exosomes and brain tissue from adults with AD.

In light of the several important roles that insulin plays in the brain ― coupled with the evidence connecting dysregulation of brain insulin and AD pathology ― restoring brain insulin function may offer therapeutic benefit for adults suffering either AD or MCI.

"There are a number of ways to do this," said Craft. "But one of the approaches that we've focused on is providing insulin directly to the brain through intranasal administration.

"By doing this, you circumvent potential issues if you administered insulin systemically."

Previous research has shown that through this mode of administration, insulin can bypass the blood-brain barrier and reach the brain through olfactory and trigeminal perivascular channels, with little effect on peripheral insulin or blood glucose levels.

As previously reported by Medscape Medical News, an earlier pilot study, also conducted by Craft and her team, showed that 4 months of daily intranasal administration of 20 IU or 40 IU of insulin preserved cognitive performance in individuals with AD or MCI.

Deeper Dive

In the current investigation, the researchers wanted to broaden these findings in a larger, longer, randomized double-blinded clinical trial.

The investigators assessed the efficacy of intranasal insulin on cognition, function, and biomarkers of AD, as well as the safety and feasibility of the delivery method. The multicenter trial was conducted from 2014 to 2018 and included 27 sites.

Study participants were between the ages of 55 and 85 and had been diagnosed with amnestic MCI or AD on the basis of National Institute on Aging–Alzheimer Association criteria, a score of 20 or higher on the Mini–Mental State Examination, a clinical dementia rating of 0.5 or 1.0, or a delayed logical memory score within a specified range.

In total, 289 participants were randomly assigned to receive 40 IU of insulin or placebo for 12 months, followed by a 6-month open-label extension phase.

The first 49 participants (32 men; mean age, 71.9 years) underwent insulin administration with the same device the investigators used in previous trials.

Of these, 45 completed the blinded phase, and 42 completed the open-label extension. When this device, which uses an electronic nebulizer-like delivery system, proved unreliable, the researchers switched to a second device, which uses a liquid hydrofluoroalkane propellant to deliver a metered dose of insulin through a nose tip without electronic assistance.

Device 2 was used for the remaining 240 participants (123 men; mean age, 70.8 years). These patients became the study's primary intention-to-treat population.

The study's primary outcome was the mean change in score on the Alzheimer Disease Assessment Scale–Cognitive Subscale 12 (ADAS-cog-12), which was evaluated at 3-month intervals.

Secondary clinical outcomes were assessed at 6-month intervals. These included the mean change in scores for the Alzheimer Disease Cooperative Study Activities of Daily Living Scale for Mild Cognitive Impairment and the Clinical Dementia Rating Scale Sum of Boxes.

Safety and adherence were also assessed during each study visit. Physical and neurologic examinations were performed at baseline and at months 6, 12, and 18.

Of the primary intention-to-treat population of 240 patients, 121 were randomly assigned to receive intranasal insulin. The remaining 119 received placebo and served as controls. The two groups were demographically comparable.

Better Cognitive Performance

A total of 215 participants completed the blinded phase; 198 participants completed the open-label extension. Discontinuation rates were comparable in both arms.

The researchers found no differences between groups with respect to mean change in ADAS-cog-12 score from baseline to month 12 (0.0258 points; 95% CI: –1.771 to 1.822 points; P = .98). The two groups also proved comparable in terms of performance on all other cognitive tests.

The open-label portion yielded similar results. Participants originally assigned to the insulin arm and their counterparts in the placebo arm did not differ with respect to mean score change on the ADAS-cog-12 test (or any other outcome) at either month 15 or 18.

Cerebrospinal fluid insulin levels were unchanged between groups, as were blood glucose and A1c values. Indeed, levels of Aβ42, Aβ40, total tau protein, and tau p-181 were comparable for the patients who received intranasal insulin and those who received placebo.

The most common adverse events were infections, injuries, respiratory disorders, and nervous system disorders, though these did not differ between groups. In addition, there were no differences between groups with respect to severity of adverse events; most were rated as mild.

In contrast with the intention-to-treat population, the study's secondary analysis ― using data from the original administration device ― yielded markedly different results.

In the blinded phase, patients who received insulin had better ADAS-cog-12 performance at 12 months (−2.81 points; 95% CI: −6.09 to 0.45 points; P = .09) and nominally significant effects at 6 months (−3.78 points; 95% CI: −6.79 to −0.78 points; P = .01).

Device Type Critical

These effects persisted in the open-label analyses. Patients who received intranasal insulin had superior ADAS-cog-12 scores at month 15 (−5.70 points; 95% CI: −9.62 to −1.79 points; P = .004) and month 18 (−5.78 points; 95% CI: −10.55 to −1.01 points; P = .02) compared with their counterparts who received insulin via the second device.

Interestingly, this part of the study also showed that although individual biomarkers did not differ significantly between the two arms, the ratios of Aβ42 to Aβ40 (P = .01) and Aβ42 to total tau (P = .03) increased with use of the first device.

The number, type, and severity of adverse events were comparable between the insulin and placebo groups in this arm of the study.

The mixed results revealed by the trial demonstrate that the device used for intranasal insulin administration is paramount in determining the therapy's potential efficacy.

"Our take-home message is that the device is a very important factor for these studies and that one needs to validate their ability to effectively deliver insulin to the CNS," said Craft.

"We were quite confident that the first device was able to do that. On the other hand, the second device has never been tested in that way, and we still don't know whether or not that device was able to successfully deliver insulin," she said.

Perhaps not surprisingly, the investigators recognize the need for more research in the field. Such studies, Craft noted, will utilize administration devices that have been previously verified to have the ability to deliver insulin to the central nervous system.

"We're currently testing several devices," she noted. "We're using a protocol where we administer insulin with the devices and then conduct a lumbar puncture about 30 minutes later to verify that it is actually raising insulin levels in the cerebrospinal fluid."

Not a Failure

Commenting on the findings for Medscape Medical News, Samuel E. Gandy, MD, PhD, who was not involved in the study, said the research illustrates the challenge when a new therapy, a new delivery device, and a cohort of cognitively impaired patients collide.

"The result is not quite a slam dunk but is also by no means a failure," commented Gandy, Mount Sinai Chair in Alzheimer's Research at Mount Sinai Medical Center, New York, New York.

"One looks forward to future iterations of the Craft et al approach, wherein the trialists tweak the ligand and/or the delivery schedule and/or the device and/or the disease and/or the disease stage," Gandy added. "Another ligand, VGF, also holds promise for intranasal delivery, based on work from Steve Salton, Michelle Ehrlich, and Eric Schadt, all from Mount Sinai. Perhaps the nose knows!"

For Craft, the potential upside of intranasal insulin for these patients is significant and warrants further investigation. "I understand why people who are not familiar with prior research in this area might be skeptical of our enthusiasm, given the results in the intention-to-treat population," she said.

"But those of us who have been working along with this for a while now, we feel like we've got to do the next study. But we need to have a device that we know works," Craft added.

"If this is real, then there may be a very large clinical benefit in symptomatic patients, and there's nothing so far that has really improved symptomatic disease."

The study was supported by the National Institute on Aging. Eli Lilly provided diluent placebo for the blinded phase and insulin for the open-label phase of the clinical trial at no cost. Craft received grants from the National Institute on Aging and nonfinancial support from Eli Lilly during the conduct of the study and personal fees from T3D Therapeutics and vTv Therapeutics outside the submitted work.

JAMA Neurol. Published online June 22, 2020. Abstract

For more Medscape Psychiatry news, join us on Facebook and Twitter.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.