Evolution of surfactant protein-D levels in children with ventilator-associated pneumonia.

RATIONALE: The pathobiology of ventilator-associated pneumonia (VAP) in children is poorly understood; investigation has been limited by lack of universally applied diagnostic criteria and reliable biomarkers for this condition.
OBJECTIVES: We evaluated the clinical pulmonary infection score (CPIS) in diagnosing VAP and prospectively characterized the relationship between surfactant protein-D (SP-D) metabolism and VAP.
METHODS: Children admitted to an Egyptian PICU requiring intubation were screened for the absence of primary pulmonary pathology. Thirty-nine children underwent two evaluations: during the first 36 hr following intubation and after 4 days of mechanical ventilation. During both, bronchoalveolar lavage fluid (BALF) was obtained for culture and SP-D assay. CPIS was computed during the second evaluation.
RESULTS: Optimum performance of the CPIS against BALF culture occurred at a cutoff value of 6, (ROC AUC of 0.89 ± 0.05). Children who developed VAP had significantly higher SP-D levels, both preceding (129.9 ± 33.5 ng/ml at the 1st BAL)-and following positive BALF culture (249.5 ± 51.2 ng/ml at the 2nd BAL), compared to children whose BALF remained sterile (62.6 ± 18.1 ng/ml and 64.9 ± 9.4 ng/ml; P < 0.001). This increase in SP-D levels was most evident in children infected with Pseudomonas aeruginosa compared to children with Klebsiella pneumonia or S. aureus.
CONCLUSIONS: The CPIS performed well against BALF culture. We observed a bacterial species-specific difference in SP-D levels in children who developed VAP; this change preceded detection of infection by CPIS or BALF culture.