Spinal cord injury (SCI) is a life-changing event. People who suffer from it lose their ability
to move normally and are subject to life-threatening symptoms. When the insult is not too
severe, partial recovery is possible. Otherwise, for severely affected people, the perspectives to
ever retrieve normal functions is almost non-existent. A critically important window to promote
plasticity is in the acute phase after the trauma (weeks to months). However, at this stage,
people cannot take active part in rehabilitation programs due to their incapacity to activate their
limbs and to the apparition of severe autonomic symptoms such as orthostatic hypotension.
Hence, current activity-based treatments are of little help for this population. Consequently, SCI
dramatically affects their autonomy and quality of life on the life-long term.
Continuous Epidural Electrical Stimulation (EES) of the spinal cord has been proposed as a
method to reactivate dormant neural networks innervated below the site of injury. It enables
robust control over locomotion and autonomic functions. In animal models of severe lesion,
continuous EES promoted long lasting recovery of motor abilities when coupled with an intensive
training. In human, its recent application indeed supported overground walking in chronically
and severely injured people, but only showed limited plastic effects when compared to animals.
Why? Recent advances support that continuous EES abolishes proprioceptive feedback that
is essential to promote recovery. Moreover, late treatment of severe injuries consistently fail a
promoting recuperation of functions. Alternatively, over the last decade, our laboratory developed
a spatio-temporally patterned approach to EES. In addition to restore fine control over motor
functions, this technique avoids the caveats of continuous EES, and hypothetically opens a window
of opportunity for plasticity to happen. In this thesis, we propose a translation of this concept to
humans. We first demonstrate that spatio-temporal EES restores robust and versatile locomotor
control in people with chronic incomplete, though severe, SCI. Furthermore, the results support
an aptitude of this approach to promote long lasting recovery of functions without stimulation
after 5 months of intensive training. We then report on a patient-tailored technology, that
includes personalized computational models enabling precise pre-operative planning strategies, a
new electrode-lead and control softwares. This technology enables restoration of various motor
functions in people with chronic and motor-complete injuries within few hours after beginning
of therapy. Finally, we transfer this targeted approach to the control of blood pressure. We
demonstrate that closed-loop stimulation protocols support a precise control over arterial blood
pressure in rodents, non-humane primates and humans with severe SCI submitted to challenging
cardiovascular conditions. Combined, the results presented in this manuscript open a clinically
realistic perspective for the application of EES acutely after injury and to potentially promote
robust recovery in severe SCI.