Safety and feasibility of fat injection therapy with adipose-derived stem cells in a rabbit hypoglossal nerve paralysis model: A pilot study.

OBJECTIVE: The aim of this study is to establish a unilateral tongue atrophy model by cutting the hypoglossal nerve and to evaluate the safety and feasibility of a fat injection of adipose-derived stem cells (ADSCs) to restore swallowing function. METHODS: A total of 12 rabbits were randomized to three groups; the ADSCs+fat group (n=4), the fat group (n=4) and the control group (n=4). All rabbits were treated with denervation of the left hypoglossal nerve and their conditions including body weight and food intake were checked during follow-up periods (8 weeks). At 4 weeks after the transection of the nerve, rabbits received the injection therapy into the denervated side of the tongue with 1.0mL fat tissue premixed with 0.5mL ADSCs in the ADSCs+fat group, 1.0mL fat tissue premixed with 0.5mL PBS in the fat group and 1.5mL PBS in the control group. Rabbits were euthanized 8 weeks post-treatment and resected tongues were collected, formalin-fixed and paraffin embedded. To evaluate the change of the intrinsic muscles of the tongue, muscle fibers around the treatment area was analyzed by evaluating 5 consecutive hematoxylin-eosin slides per rabbit. RESULTS: Food intake did not decrease upon nerve denervation, and none of the rabbits displayed adverse effect such as aspiration, surgical wound dehiscence or infection. No significant body weight changes were found between the three groups at 4 and 8 weeks after nerve transection (p>0.05). In the control group, the denervated side of tongue had significantly smaller muscle fiber areas and diameters compared to the non-denervated side (p<0.05). The ADSCs+fat group demonstrated a larger area of inferior longitudinal muscle fibers compared to the control and the fat groups (582±312µm2 vs. 405±220µm2 and 413±226µm2; p<0.05). A significant thicker lesser diameter of inferior longitudinal muscle fibers was found in the ADSCs+fat group compared to the control and the fat groups (24±8µm vs. 20±6µm and 20±7µm; p<0.05). CONCLUSION: The rabbit tongue atrophy model was found suitable for the assessment of muscle change after nerve transection. Fat injection therapy with ADSCs demonstrated great potential to prevent the muscle atrophy after denervation and to promote the muscle regeneration around the injection area.