BOSTON – Long-acting intranasal insulin imparted a significant memory boost to patients with mild cognitive impairment or Alzheimer’s disease who carried a high-risk allele for the disease in a randomized, placebo-controlled trial.
"It’s a very interesting result," principal study investigator Suzanne Craft, Ph.D., of Wake Forest University, Winston-Salem, N.C., said at the Alzheimer’s Association International Conference 2013. "It’s not what we have previously observed with regular insulin, and so far it’s very difficult to find a treatment that benefits these patients."
Intranasal insulin travels quickly up the trigeminal and olfactory nerves and penetrates the blood-brain barrier within 15-30 minutes. Dr. Craft has previously examined the effect of regular insulin delivered intranasally. In those studies, patients had improved cognition and cerebral glucose metabolism. The treatment also modulated levels of biomarkers in cerebrospinal fluid. But because regular insulin’s effect peaks shortly after administration and falls off over 8 hours, she and her colleagues reasoned that long-acting insulin might confer improved cognitive results. "The longer-acting insulins mimic basal insulin, which remains mildly elevated even postprandially," she said.
The trial, dubbed SNIFF-LONG (Study of Nasal Insulin to Fight Forgetfulness), gave 20 or 40 IU of long-acting insulin or placebo to 60 patients (mean age, 72 years) with mild cognitive impairment (MCI) or Alzheimer’s disease. Most (39) had a diagnosis of MCI, and the rest had Alzheimer’s. Patients had a mean Mini-Mental State Examination score of 26. Thirty carried the apolipoprotein E (apo E) epsilon 4 allele, putting them at a very high risk for developing Alzheimer’s.
The patients had an average insulin resistance index of 2.8. At baseline, 14 patients were taking an acetylcholinesterase inhibitor or memantine. The primary endpoints were scores on a test of composite memory; secondary endpoints included scores on pattern separation and visual memory counting tests. There was also a planned subgroup analysis that examined results by apo E epsilon 4 status.
Intranasal insulin and placebo were administered once a day for 3 weeks via a nasal delivery device that has been specially developed for these trials. The group underwent memory testing at baseline and after the treatment period.
Overall, there were no significant improvements on the primary endpoint measuring a composite of scores on verbal memory tests (the Story Recall Test and the Buschke Selective Reminding Test) in patients taking either the 20- or 40-IU dose.
However, in a planned subgroup analysis, patients carrying the apo E epsilon 4 allele performed significantly better than did those without it. Apo E epsilon 4 noncarriers declined about 1 point relative to their baseline score, while carriers improved about 1 point, said Dr. Craft, who conducted the trial when she was professor of psychiatry and behavioral sciences at the University of Washington, Seattle, and director of the Memory Disorders Clinic at the American Lakes division of the VA Puget Sound Health Care System. At Wake Forest, she is a professor of gerontology and geriatric medicine.
"This was a very robust result that characterized the vast majority of patients in each group," she noted.
The results were also examined by the level of insulin resistance. "In those with the highest insulin resistance, the 40-IU dose showed a significant benefit [of about 1 point]; and those without insulin resistance did not, and, in fact, showed memory decline" of about 1 point, she said.
Both patient groups responded well to the 40-IU dose in the secondary endpoints of scores on tests of pattern separation (Benton Visual Retention Rest-Recognition) and visual memory counting (Dot Counting Test). "There were no interactions with [apo E epsilon 4] status or insulin resistance. Instead we saw a nice, robust improvement on both the pattern separation and dot counting tests" for patients taking that dose.
Dr. Craft will soon be recruiting patients for a 1-year study called SNIFF. This will include 240 subjects randomized to 20 IU of insulin (Humilin) inhaled twice daily (total dose, 40 IU) or placebo. At the end of the first 12 months, insulin will be given open-label to all participants for an additional 6 months. Tests of cognition are planned at baseline and at 3, 6, 9, 12, and 15 months. Memory and changes in daily functioning will be assessed at 6, 12, and 18 months.
The study was sponsored by the University of Washington and the National Institute on Aging. Dr. Craft had no financial disclosures.
On Twitter @Alz_Gal