The most severe sequelae after rehabilitation from SARS are femoral head necrosis and pulmonary fibrosis. We performed a 15-year follow-up on the lung and bone conditions of SARS patients. We evaluated the recovery from lung damage and femoral head necrosis in an observational cohort study of SARS patients using pulmonary CT scans, hip joint MRI examinations, pulmonary function tests and hip joint function questionnaires. Eighty medical staff contracted SARS in 2003. Two patients died of SARS, and 78 were enrolled in this study from August 2003 to March 2018. Seventy-one patients completed the 15-year follow-up. The percentage of pulmonary lesions on CT scans diminished from 2003 (9.40 ± 7.83)% to 2004 (3.20 ± 4.78)% (P < 0.001) and remained stable thereafter until 2018 (4.60 ± 6.37)%. Between 2006 and 2018, the proportion of patients with interstitial changes who had improved pulmonary function was lower than that of patients without lesions, as demonstrated by the one-second ratio (FEV₁/FVC%, t = 2.21, P = 0.04) and mid-flow of maximum expiration (FEF_25%–75%, t = 2.76, P = 0.01). The volume of femoral head necrosis decreased significantly from 2003 (38.83 ± 21.01)% to 2005 (30.38 ± 20.23)% (P = 0.000 2), then declined slowly from 2005 to 2013 (28.99 ± 20.59)% and plateaued until 2018 (25.52 ± 15.51)%. Pulmonary interstitial damage and functional decline caused by SARS mostly recovered, with a greater extent of recovery within 2 years after rehabilitation. Femoral head necrosis induced by large doses of steroid pulse therapy in SARS patients was not progressive and was partially reversible.

At present, repair methods for peripheral nerve injury often fail to get satisfactory result. Although various strategies have been adopted to investigate the microenvironment after peripheral nerve injury, the underlying molecular mechanisms of neurite outgrowth remain unclear. In this study, we evaluate the effects of exosomes from gingival mesenchymal stem cells (GMSCs) combined with biodegradable chitin conduits on peripheral nerve regeneration. GMSCs were isolated from human gingival tissue and characterized by surface antigen analysis and in vitro multipotent differentiation. The cell supernatant was collected to isolate the exosomes. The exosomes were characterized by transmission electron microscopy, Western blot, and size distribution analysis. The effects of exosomes on peripheral nerve regeneration in vitro were evaluated by coculture with Schwann cells and DRGs. The chitin conduit was prepared and combined with the exosomes to repair rat sciatic nerve defect. Histology, electrophysiology, and gait analysis were used to test the effects of exosomes on sciatic nerve function recovery in vivo. We have successfully cultured GMSCs and isolated exosomes. The exosomes from GMSCs could significantly promote Schwann cell proliferation and DRG axon growth. The in vivo studies showed that chitin conduit combined with exosomes from GMSCs could significantly increase the number and diameter of nerve fibers and promote myelin formation. In addition, muscle function, nerve conduction function, and motor function were also obviously recovered. In summary, this study suggests that GMSC-derived exosomes combined with biodegradable chitin conduits are a useful and novel therapeutic intervention in peripheral nerve repair.