A metabonomic approach to early prognostic evaluation of experimental sepsis.

Objectives: Early prognostic evaluation of sepsis is an attractive strategy to decrease the mortality of septic patients, but presently there are no satisfactory approaches. Our goal is to establish an early, rapid and efficient approach for prognostic evaluation of sepsis.

Methods: Forty-five septic rats, induced by cecal ligation and puncture, were divided into surviving (n=23) and nonsurviving group (n=22) on day 6. Serum samples were obtained from septic and sham-operated rats (n=25) at 12h after surgery. HPLC/MS assays were performed to acquire serum metabolic profiles, and radial basis function neural network (RBFNN) was employed to build predictive model for prognostic evaluation of sepsis.

Results: Principle component analysis allows a clear discrimination of the pathologic characteristics among rats from surviving, nonsurviving and sham-operated groups. Six metabolites related to the outcome of septic rats were then structurally identified, which included linolenic acid, linoleic acid, oleic acid, stearic acid, docosahexaenoic acid and docosapentaenoic acid. A RBFNN model was built upon the metabolic profile data from rat serum, and a high predictive accuracy over 94% was achieved.

Conclusions: HPLC/MS-based metabonomic approach combined with pattern recognition permits accurate outcome prediction of septic rats in the early stage. The proposed approach has advantages of rapid, low-cost and efficiency, and is expected to be applied in clinical prognostic evaluation of septic patients.