In a first human imaging study researchers found that intranasal insulin reached brain regions involved in memory and cognition, with uptake and clearance patterns differing between cognitively normal adults and those with mild cognitive impairment, supporting further research into intranasal delivery for Alzheimer’s disease.
Researchers used a radiolabeled form of insulin, [68Ga]Ga-NOTA-insulin, to trace its path through the brain using positron emission tomography scanning. Sixteen older adults—7 cognitively normal (CN) and 9 with mild cognitive impairment (MCI)—were enrolled. The participants ranged in age from 60.6 to 84.6 years (mean, 72.4 years) and included 6 women and 10 men.
Each participant received 0.055 gigabecquerels of [68Ga]Ga-NOTA-insulin via a nasal spray device called the Aptar Cartridge Pump System. The device delivered six insulin sprays—three to each nostril—while the participant sat upright. Brain scans began within 5 minutes after the final spray and continued for 40 minutes, followed by a 15-minute whole-body scan.
The radiotracer was detected in 11 brain regions with significantly elevated uptake (P < .01), including the hippocampus, parahippocampus, amygdala, olfactory cortex, thalamus, superior and middle temporal poles, insula, caudate, putamen, and anterior cingulum. The mean whole-brain standard uptake value over the scan period was 0.68 ± 0.01 g/mL.
CN participants had significantly higher insulin uptake in the hippocampus (P = .0254), parahippocampus (P = .0048), amygdala (P = .0185), olfactory cortex (P = .0068), and thalamus (P = .0068) compared with participants with MCI. Time–activity curves showed that the CN group had peak uptake around 750 seconds after dosing, while the MCI group peaked earlier at 150 seconds and showed faster clearance from the brain.
Aea under the curve, was higher in all 11 brain regions in the CN group (all P < .0001). The researchers also examined how clinical measures related to brain uptake.
In the CN group, baseline plasma insulin was positively correlated with insulin uptake in multiple regions (Spearman r = 0.86–0.96; P = .0137–.0005). Pulse pressure was also positively associated with uptake in these regions (r = 0.78–0.94; P = .0408–.0019). In contrast, participants with MCI showed fewer and weaker correlations.
Sex-specific analysis highlighted that female participants had positive correlations between pulse pressure and brain insulin uptake (r = 0.80–1.00; P < .0001) and negative associations with diastolic blood pressure (r = –0.83 to –1.00; P = .0416–.0001) and phosphorylated tau217 levels (r = –0.84 to –0.93; P = .0360–.0077).
Plasma insulin levels remained stable throughout the procedure. Blood glucose levels decreased slightly from 99.31 mg/dL at baseline to 92.69 mg/dL at 50 minutes (P < .05), but no participants experienced hypoglycemia. Two participants reported mild headaches that resolved within 24 hours.
The radiotracer demonstrated high purity (>99%) and molar activity of 112 GBq/μmol, with the administered dose well below the 0.185 GBq safety threshold. Researchers concluded the approach was safe and could support future clinical trials using intranasal delivery for brain-targeted therapies.
“Our results show that intranasal administration of [68Ga]Ga-NOTA-insulin is safe and delivers insulin to multiple brain regions that support cognition and are affected by AD pathology,” said Kiran K. Solingapuram Sai of the Department of Radiology, Wake Forest University School of Medicine, Winston-Salem, North Carolina, and colleagues.
These findings confirm that intranasal insulin reaches brain regions involved in cognition and support its use in future clinical research on Alzheimer’s disease and MCI.
The authors reported no conflicts of interest.
Source: Alzheimer's & Dementia