From a systems view to spotting a hidden island: A narrative review implicating insula function in alcoholism

Excessive use of alcohol promotes the development of alcohol addiction, but the understanding of how alcohol induced brain alterations lead to addiction remains limited. To further this understanding, we adopted an unbiased discovery strategy based on the principles of systems medicine. We used functional magnetic resonance imaging data from patients and animal models of alcohol addiction-like behaviors, and developed mathematical models of the 'relapse-prone' network states to identify brain sites and functional networks that can be selectively targeted by therapeutic interventions. Our systems level, non-local, and largely unbiased analyses converged on a few well-defined brain regions, with the insula emerging as one of the most consistent findings across studies. In proof-of-concept experiments we were able to demonstrate that it is possible to guide network dynamics towards increased resilience in animals but an initial translation into a clinical trial targeting the insula failed. Here, in a narrative review, we summarize the key experiments, methodological developments and knowledge gained from this complete round of a discovery cycle moving from identification of 'relapse-prone' network states in humans and animals to target validation and intervention trial. Future concerted efforts are necessary to gain a deeper understanding of insula function a in a state-dependent, circuit-specific and cell population perspective, and to develop the means for insula-directed interventions, before therapeutic targeting of this structure may become possible.

Automated summary

Excessive alcohol use can lead to alcohol addiction, but the understanding of how alcohol-induced brain changes lead to addiction is limited. An unbiased discovery strategy based on systems medicine identified specific brain regions, with the insula emerging as a consistent finding. While initial translation failed, proof-of-concept experiments have shown the potential to guide network dynamics in animal models.