Abstract: Objective To investigate the effect of recombinant human keratinocyte growth factor 2 (rhKGF-2) on lung tissue of rabbits with severe smoke inhalation injury. Methods A total of 120 New Zealand rabbits were divided into 5 groups by random number table after being inflicted with severe smoke inhalation injury, with 24 rats in each group. Rabbits in the simple injury group inhaled air, while rabbits in the injury+phosphate buffer solution (PBS) group inhaled 5 mL PBS once daily for 7 d. Rabbits in injury+1 mg/kg rhKGF-2 group, injury+2 mg/kg rhKGF-2 group, and injury+5 mg/kg rhKGF-2 group received aerosol inhalation of 1 mg/kg, 2 mg/kg, and 5 mg/kg rhKGF-2 (all dissolved in 5 mL PBS) once daily for 7 d, respectively. On treatment day 1, 3, 5, and 7, blood samples were taken from the ear central artery of 6 rabbits in each group. After the blood was taken, the rabbits were sacrificed, and the tracheal carina tissue and lung were collected. Blood pH value, arterial oxygen partial pressure (PaO₂), arterial blood carbon dioxide pressure (PaCO₂), and bicarbonate ion were detected by handheld blood analyzer. The expressions of pulmonary surfactant-associated protein A (SP-A) and vascular endothelial growth factor (VEGF) in lung tissue were detected by Western blotting. Pathomorphology of lung tissue and trachea was observed by hematoxylin-eosin staining. Data were processed with analysis of variance of two-way factorial design and Tukey test. Results (1) Compared with those in simple injury group, the blood pH values of rabbits in the latter groups on treatment day 1-7 had no obvious change (P>0.05). The PaO₂ of rabbits in injury+2 mg/kg rhKGF-2 group on treatment day 5 and 7 were (75.0±2.4) and (71.0±4.5) mmHg (1 mmHg=0.133 kPa), respectively, which were significantly higher than (62.0±6.8) and (63.0±3.0) mmHg in simple injury group (q=4.265, 8.202, P<0.05 or P<0.01). The PaO₂ of rabbits in injury+5 mg/kg rhKGF-2 group on treatment day 7 was (82.0±4.9) mmHg, which was significantly higher than that in simple injury group (q=6.234, P<0.01). Compared with that in simple injury group, the PaCO₂ of rabbits in injury+2 mg/kg rhKGF-2 group on treatment day 3 was significantly decreased (q=4.876, P<0.01) and significantly increased on treatment day 5 (q=5.562, P<0.01); the PaCO₂ of rabbits in injury+5 mg/kg rhKGF-2 group was significantly increased on treatment day 5 and 7 (q=5.013, 4.601, P<0.05 or P<0.01). Compared with that in simple injury group, the serum bicarbonate ion of rabbits in injury+1 mg/kg rhKGF-2 group on treatment day 7 was significantly increased (q=5.142, P<0.01); the serum bicarbonate ion of rabbits in injury+2 mg/kg rhKGF-2 group on treatment day 5 and 7 were significantly increased (q=4.830, 6.934, P<0.01); the serum bicarbonate ion of rabbits in injury+5 mg/kg rhKGF-2 group on treatment day 5 were significantly increased (q=3.973, P<0.05). (2) The expressions of SP-A in lung tissue of rabbits in simple injury group and injury+PBS group in each treatment time point were close (P>0.05). The expressions of SP-A in lung tissue of rabbits in injury+2 mg/kg rhKGF-2 group and injury+5 mg/kg rhKGF-2 group on treatment day 3 were 0.091±0.007 and 0.101±0.009, respectively, significantly higher than 0.069±0.009 in simple injury group (q=10.800, 13.580, P<0.01). The expressions of SP-A in lung tissue of rabbits in injury+1 mg/kg rhKGF-2 group, injury+2 mg/kg rhKGF-2 group, and injury+5 mg/kg rhKGF-2 group on treatment day 5 and 7 were 0.127±0.008, 0.132±0.006, 0.194±0.006, 0.152±0.017, 0.166±0.004, 0.240±0.008, significantly higher than 0.092±0.003 and 0.108±0.005 in simple injury group (q=6.789, 12.340, 17.900, 9.875, 31.480, 40.740, P<0.01). (3) On treatment day 1 and 5, there was no significant difference in the expression of VEGF in lung tissue of rabbits among the 5 groups (P>0.05). Compared with those in simple injury group, the expressions of VEGF in lung tissue of rabbits in injury+2 mg/kg rhKGF-2 group on treatment day 3 and 7 were significantly increased (q=4.243, 8.000, P<0.05 or P<0.01), and the expression of VEGF in lung tissue of rabbits in injury+5 mg/kg rhKGF-2 group on treatment day 7 was significantly increased (q=20.720, P<0.01). (4) On treatment day 1, the injury of rabbits in each group was similar, with a large number of neutrophils infiltrated and abscess formed in the alveolar and interstitial tissue, thickened alveolar septum, some collapsed alveolar and atelectasis; large area of tracheal mucosa was degenerated and necrotic, with a large amount of inflammatory exudates blocking in the cavity. On treatment day 3, the inflammation of lung tissue and trachea in each group were improved, but the inflammation in simple injury group and injury+PBS group was also serious. On treatment day 5, the inflammation in lung tissue and trachea of rabbits in injury+2 mg/kg rhKGF-2 group and injury+5 mg/kg rhKGF-2 group were improved much obviously than those in the other groups. On treatment day 7, the inflammation in lung tissue of rabbits in injury+5 mg/kg rhKGF-2 group alleviated obviously than those in the other groups, most alveoli had no obvious exudative fluid, the alveolar cavity was intact and clear, the local alveolar dilated like a cyst, and the alveolar septum thinning; the improvement of inflammation of trachea was more obvious than the other groups, the tracheal mucosa tended to be more complete, and few neutrophils were infiltrated in the endotracheal cavity. Conclusions Atomization inhalation of rhKGF-2 can improve the PaO₂ level of rabbits with severe smoke inhalation injury, reduce airway inflammation, increase the expression of SP-A and VEGF in lung tissue, thus promoting the repair of lung tissue.