Effects of Sedative Doses of Propofol, Dexmedetomidine, and Fentanyl on Memory and Pain in Healthy Young Adults: A Randomized, Controlled, Single-blind Crossover Study Using Functional Magnetic Resonance Imaging at 7 Tesla.

BACKGROUND

Anesthetic agents are well known for their effects on memory and pain; however, previous studies quantifying anesthetic modulation of memory have not included experimental noxious stimulation. This study used functional magnetic resonance imaging to determine how low doses of propofol, dexmedetomidine, and fentanyl affect the brain systems for memory encoding and pain perception.

METHODS

This was a single-blind, 1:1:1, randomized, placebo-controlled crossover study of 92 healthy volunteers ages 18 to 40 yr. Effect-site concentrations were targeted for propofol (1.0 mcg/ml), dexmedetomidine (0.15 ng/ml), or fentanyl (0.9 ng/ml). Participants listened to a series of 80 words creating a mental picture. Thirty were accompanied by a 2-s painful shock. Blood oxygen-weighted images were obtained at 7 T using a custom head coil. The primary outcome was next-day memory performance, measured by d', a normalized measure of correct identifications versus false positives. Mixed models were fit to test outcome differences between drug groups. Only statistically significant ( P < 0.05) changes are reported, after adjustment for multiple comparisons.

RESULTS

Recollection, reported as mean d' (95% CI), was 1.16 (95% CI, 0.97 to 1.34) under no drug. This was reduced under propofol (0.51; 95% CI, 0.182 to 0.842; P = 0.006) but not under dexmedetomidine (1.04; 95% CI, 0.73 to 1.35; P = 0.99) or fentanyl (0.98; 95% CI, 0.68 to 1.28; P = 0.99). Propofol decreased memory-encoding activation of the hippocampus and amygdala. Propofol reduced pain-related activation in the insula, anterior cingulate, hippocampus, and amygdala. Dexmedetomidine showed decreased memory-related activation in the hippocampus but did not change pain ratings or show activation differences in pain-processing areas. Fentanyl showed decreased memory activation in the hippocampus and amygdala. During painful stimulation, fentanyl decreased activation in the primary somatosensory cortex and insula and increased activation in the anterior cingulate, hippocampus, and amygdala.

CONCLUSIONS

These findings add important details to the complex framework of how these distinct anesthetics affect different aspects of cognition through diverse neuroanatomic targets in the brain.