Mass spectometry-based protein patterns in the diagnosis of sepsis/systemic inflammatory response syndrome.

Early differential diagnosis of systemic inflammatory reactions in critically ill patients is essential for timely implementation of lifesaving therapies. Despite many efforts made, reliable biomarkers to discriminate between infectious and noninfectious causes of systemic inflammatory response syndrome (SIRS) are currently not available. Recent advances in mass spectrometry-based methods have raised hopes that identification of spectral patterns from serum/plasma samples can be instrumental in this context. We compared protein expression patterns from patients with SIRS of infectious and noninfectious origin. Plasma samples from 166 patients obtained under rigorously standardized preanalytical conditions were applied to Q10 and CM10 ProteinChips. Protein profiles were used to train and develop decision tree classification algorithms. Discriminatory peaks were isolated and identified. Classification trees distinguished patients with noninfectious SIRS with organ dysfunction following open heart surgery using cardiopulmonary bypass from those with severe sepsis or septic shock with distinct sensitivities and specificities. Results were validated in a blinded test set in two independent experiments and in a second independently collected test set. Discriminatory peaks at 13.8 and 55.7 kd were identified as transthyretin and α1-antitrypsin; the third protein at m/z 4,798 was assigned to a proteolytic fragment of α1-antitrypsin. Taken together, our data demonstrate that plasma protein profiling allows reproducible discrimination between patients with infectious and noninfectious SIRS with high sensitivity and specificity. However, rigorous standardization as well as considering drug-related interferences is essential when interpreting protein profiling studies. Identification of discriminatory proteins suggests a direct link between infectious-related protease activity and a sepsis-specific diagnostic pattern for discrimination of patients with SIRS.