OBJECTIVETo observe the changes in plasma gelsolin (pGSN) level of patients with severe burn and to explore its relationship with sepsis and death of patients.

METHODSOne hundred and two patients with total burn area equal to or larger than 30% TBSA hospitalized from May 2010 to May 2012 were included as burn group. Twenty-five healthy volunteers were recruited as healthy control group. Peripheral venous blood of patients was harvested on post burn day (PBD) 1, 3, 7, 14, and 21 to determine the pGSN level with double antibody sandwich ELISA kits, and the same maneuver was carried out in healthy volunteers. (1) Patients in burn group were divided into three groups by burn size: small burn area group (30% - 49% TBSA, n = 39), medium burn area group (larger than 49% and smaller than or equal to 69% TBSA, n = 33), and large burn area group (larger than 69% and smaller than or equal to 99% TBSA, n = 30). (2) According to diagnostic criteria of burn sepsis, patients in burn group were divided into sepsis group (n = 43) and non-sepsis group (n = 59). (3) According to the prognosis of patients with sepsis, patients in sepsis group were further divided into non-survival sepsis group (n = 14) and survival sepsis group (n = 29). The levels of pGSN in above groups were compared, and their relationship with sepsis and death of patients was analyzed. Data were analyzed with analysis of variance, LSD test and one-way Logistic regressions.

RESULTS(1) Levels of pGSN in burn group were obviously lower than those of healthy control group on PBD 1, 3, 7, 14, and 21 (with F values respectively 140.01, 369.52, 702.15, 360.14, 84.16, P values all below 0.01). (2) The mean levels of pGSN in large, medium, and small burn area groups at five time points were (43 ± 11), (85 ± 23), (124 ± 38) mg/L, showing statistically significant differences among them (F = 367.76, P < 0.01), and they were all lower than that of healthy control group [(326 ± 51) mg/L, P values all below 0.01]. (3) The mean levels of pGSN in sepsis group and non-sepsis group at the five time points were (77 ± 12), (122 ± 38) mg/L. Levels of pGSN in sepsis group were lower than those in non-sepsis group on PBD 3, 7, 14, and 21 (with F values respectively 30.35, 111.59, 209.36, 422.76, P values all below 0.01). (4) The mean levels of pGSN in non-survival sepsis group and survival sepsis group at the five time points were (53 ± 8) and (103 ± 25) mg/L. Levels of pGSN in non-survival sepsis group were lower than those in survival sepsis group on PBD 1, 3, 7, 14, and 21 (with F values respectively 9.05, 18.48, 41.34, 107.11, 180.48, P values all below 0.01). (5) Logistic regression analysis showed that the level of pGSN is the independent risk factor related to the complication of sepsis (odds ratio: 5.44, 95% confidence interval: 2.35 - 12.74, P < 0.01) and death (odds ratio: 5.52, 95% confidence interval: 2.34 - 12.19, P < 0.01) in burn patients.

CONCLUSIONSSevere burn injury could down-regulate the pGSN level of patients, and it decreases along with the increase in the area and severity of burn trauma. pGSN level appears to be an early prognostic marker for patients suffering from severe burns.