I am delighted to give a lecture at the Academy of Medicine at the invitation of Professors Jean-Louis Guéant and Bernard Bioulac on April 14. The title of my lecture will be “The Neuroarchaeology of Neurodevelopmental Disorders: A New Approach to Predicting and Treating Autism.” I will try to cover the areas addressed by my work on autism in terms of prediction—the Pelargos Project—and treatment—the Bumetanide Project.
As a reminder, the Pelargos Project, which we launched five years ago, aims to identify at birth the infants who will later be diagnosed with autism. We used a proprietary algorithm to retrospectively analyze and compare all data from the maternity ward regarding autistic and neurotypical children. This work shows that it is possible to identify, from birth, nearly all neurotypical children and half of those who will later receive an autism diagnosis. This work provides solid evidence that autism originates in utero and that the development of the brain is impacted by the pathological event that occurred. We have recently signed agreements with five university hospitals to validate this study and expand its conclusions. It also provides interesting insights into the in utero alterations that cause these sequelae. Above all, it will enable early diagnosis, which is crucial for starting psychoeducational treatments as soon as possible—treatments whose effectiveness increases when initiated early.
The concept of neuroarcheology is based on the idea that these initial pathological events result in clusters of neurons that retain immature properties—that, in a sense, fail to mature. They generate activity that disrupts the functioning of neural networks, particularly the oscillations essential for proper brain function. Pathogenesis would be a sort of return to the ante state; it recapitulates ontogenesis.
Based on this principle, we have undertaken clinical trials with Dr. Lemonnier using an agent—Bumetanide—which specifically blocks these immature activities, notably by restoring inhibition, a consequence of blocking an ion co-transporter (NKCC1). The trials previously mentioned were successful in seven Phase 2 trials conducted in several countries but failed in the final phase—Phase 3. However, by again using a “machine learning” algorithm, we were able to demonstrate that 30 to 40% of children respond to treatment within this Phase 3 trial that failed. In other words, if we validate this result, we will have the first effective treatment for a significant percentage of children with autism.
In our view, there is a very promising aspect to this approach, namely the use of AI to divide a syndrome as heterogeneous as autism into subpopulations—a challenge for any large-scale trial, as it precludes the idea that a single treatment could be effective for all autistic children.
Neuroarcheology opens the door to treatments for numerous neurological and psychiatric disorders that actually originate in utero. This reminds us once again just how much motherhood is—and must be—at the heart of our understanding and treatment of brain disorders.
Photo : https://www.academie-medecine.fr
