The recent introduction of robotics for locomotor training in paraplegic patients, and in particular the use of anthropomorphic exoskeletons, has opened new frontiers in rehabilitation. Existing literature, though encouraging, is still scarce and studies demonstrating efficacy are highly heterogeneous and have a small sample size. Evidence is also needed about cortical plasticity after SCI, in conjunction with the use of innovative rehabilitation devices, through indicators like neurophysiological and neuroradiological markers, as the knowledge of such mechanisms is crucial to improve clinical outcomes. Cortical circuits controlling prosthetic devices are different from those controlling normal parts of the body and remodeling mechanisms following prosthetic use have been documented, but in conditions other than SCI.

The aims of this randomized controlled trial, with a 2-arm parallel-group design, are:

1. to evaluate and quantify the efficacy of locomotor rehabilitation with a robotic anthropomorphic exoskeleton (EKSO-GT) in terms of clinical and functional outcomes, and the persistence of such efficacy;
2. to investigate the presence and persistence of brain neuronal plasticity and cortical remodeling mechanisms underlying the robotic rehabilitation approach.

Fifty patients will be recruited and randomly assigned to 2 treatment arms. Both groups will follow a program of standard locomotor rehabilitation for 8 weeks. One group will also undergo an overground locomotor training with the EKSO-GT during the first 4 weeks.

The increasing incidence of incomplete Spinal Cord Injury (SCI) has raised new rehabilitation challenges. Recovery of walking is one of the top priorities in SCI persons and growing efforts have been pursued aimed at identifying effective alternative techniques for improving gait performance.

Standard rehabilitation approach has been so far the most widely used, but the recent introduction of anthropomorphic exoskeletons may open new frontiers in the field. Anthropomorphic exoskeletons have been developed to assist SCI patients with mobility, but there is also a certain optimism that they may have potentialities to improve walking patterns of incomplete SCI persons after a rehabilitation period with such devices is terminated. So far, however, while different systematic reviews and meta-analyses have reported on the safety of the training with such exoskeletons, there are no significant Fiftystudies on its efficacy. Along with this, central mechanisms underlying the anatomical and functional changes induced by these approaches have never been investigated in SCI.

This longitudinal randomized controlled trial, with a 2-arm parallel group design, aims at evaluating the efficacy of the training with an anthropomorphic, robotized exoskeleton (EKSO-GT, by Ekso Bionics), as "add-on" to the standard locomotor rehabilitation, in improving walking performance, when compared to the standard locomotor rehabilitation alone, in a population of patients with non-acute motor incomplete SCI. Along with this and other clinical outcomes, neurophysiological and structural markers of Central Nervous System (CNS) plasticity will be explored, aimed at capturing mechanisms underlying how anthropomorphic exoskeletons affect CNS plasticity.

Fifty patients will be recruited in 3 Italian rehabilitation hospitals setting and assigned to 2 groups, with an allocation ratio of 1:1, through a block randomization approach. One group will perform a 4-week standard locomotor training (sLT) alone, while the other will perform a 4-week period sLT plus a training with the EKSO-GT (sLT + EX-T). Afterwards, both groups will undergo a further 4-week sLT alone.

Patients will be evaluated at several time points (always when the exoskeleton is not worn): clinical outcomes will be assessed by means of clinical examinations, standardized tests and validated scales; neurophysiological modulations will be evaluated by means of paired Motor and Sensory Evoked Potentials and a study of Electroencephalographic (EEG) slow waves oscillations and signal coherence during sleep; anatomical and structural cortical modifications will be studied with brain functional Magnetic Resonance Imaging (fMRI).

It is expected that the overground locomotor training with a new-generation exoskeleton, as "add-on" to standard locomotor training, can further improve clinical outcomes (especially walking performance) in the studied population, and that such clinical improvements are underlined by mechanisms modulating synaptic plasticity occurring also at the CNS level.

Inclusion Criteria:

• SCI due to traumatic or vascular etiology;
• Incomplete motor SCI (C or D in ASIA Impairment Scale);
• T1-L1 (included) neurological level;
• 1-5 years since injury;
• Functional gait ability (also with braces or orthoses);
• Sufficient Range of Motion (ROM) of lower limbs joints to achieve a reciprocal gait pattern and allow transition from sitting to vertical position;
• Stable clinical conditions;
• Minimum height of 157 cm;
• Maximum height of 188 cm;
• Maximum weight of 100 Kg;
• Maximum intertrochanteric distance of 46 cm;
• Cognitive integrity and full collaboration of the subject.
• Specific research informed consent signed.

Exclusion Criteria:

• Intensive walking rehabilitation training undergone in the last 3 months;
• Previous use of a robotic exoskeleton;
• Instability or major deformity of the spine;
• Lower limbs joints instability;
• Indication to spinal orthosis;
• Uncontrolled spasticity (score > 3 of the Modified Ashworth Scale) in the majority of the muscle groups of the lower limbs;
• History of traumatic brain injury;
• Recent significant bone fractures, traumatic and/or pathological for the required training;
• Presence of neurogenic paraosteoarthropathies (POAN) at the onset or phlogistic phase;
• Discrepancy in femurs length (> 1.3 cm) and legs length (> 1.9 cm);
• Symptomatic orthostatic hypotension;
• Severe and recurrent uncontrolled autonomic dysreflexia;
• Cardiopulmonary comorbidities limiting physical effort;
• Skin lesions that can interfere with the study rehabilitation trainings;
• Documented psychiatric pathology;
• Contraindications to fMRI and polygraphic EEG execution;
• Contraindications to TMS;
• Ongoing pregnancy.

Because of the use of an evident Medical Device (exoskeleton), enrolled subjects cannot be blind about their assignment group, however objective and standardized clinical measure scales will be employed; moreover, assessments (also neurophysiological and neuroimaging ones) will be conducted by blind experts.