Hot Off the Press – October 14, 2026

Published in Proceedings of the National Academy of Sciences by Ginevra D’Ottavio, Alana Sullivan et al. of the NIDA IRP Neurobiology of Relapse Section.
Summary
Researchers used a rat model to examine whether differences in heroin addiction-like behavior are linked to distinct patterns of brain metabolism during abstinence. Rats could self-administer heroin and were also given opportunities to choose between heroin and social interaction. The researchers combined measures of heroin intake, binge-like use, and preference for heroin over social interaction into a measure of addiction severity.
Brain imaging during 3 to 5 weeks of abstinence showed widespread metabolic differences between rats that had self-administered heroin and control rats. More importantly, individual differences in addiction severity were associated with activity in specific brain regions. At rest, greater addiction severity was linked to lower metabolism in the piriform cortex and higher metabolism in the ventral hippocampus. During heroin seeking, greater severity was associated with lower metabolism in the postsubiculum and cerebellum. These findings suggest that vulnerability to addiction may be reflected especially strongly in brain activity when animals are at rest, rather than only when they encounter drug-related cues.
Heroin exposure and addiction severity did not reduce the rats’ motivation for social interaction, and addiction severity was not associated with meaningful metabolic differences during social seeking. The findings identify brain metabolic patterns that could help researchers investigate vulnerability to opioid addiction, but the study was conducted in rats and did not establish whether these patterns can identify addiction risk or guide treatment in people.
Publication Information
Brain metabolic correlates of heroin addiction severity during abstinence and heroin versus social seeking: A PET study in rats Journal Article
In: Proc Natl Acad Sci U S A, vol. 123, no. 36, pp. e2610087123, 2026, ISSN: 1091-6490.
