Preoperative and postoperative endotoxemia in children with congenital heart disease.

Recent data indicate that increases in inflammatory cytokines are seen in patients with diverse cardiac diseases. However, the primary stimulus for cytokine secretion during cardiac illness remains unknown. Since bacterial endotoxin is a potent inducer of cytokines, we determined the incidence, magnitude, and clinical relevance of endotoxemia in children with congenital heart disease before and after surgical repair. A prospective, observational study. A large, urban, university-affiliated, tertiary-care children's hospital. Thirty children with a variety of congenital heart defects (median age, 59 days; median weight, 4.0 kg) were sequentially enrolled. Blood was sampled prior to surgery, and at 1, 8, 24, 48, and 72 h following cardiopulmonary bypass. Assays included plasma endotoxin, lipopolysaccharide-binding protein (LBP), and interleukin-6 (IL-6). Twenty-nine of 30 patients (96%) had evidence of endotoxemia during the study period. Twelve of the 30 patients (40%) were significantly endotoxemic prior to surgery. LBP, a plasma marker that responds to bacteria and endotoxin, rose significantly following cardiopulmonary bypass, as did the plasma levels of IL-6. Fifteen of 30 patients met prospectively defined criteria for experiencing a severe hemodynamic disturbance in their postoperative course. These patients had significantly higher preoperative plasma LBP (p < 0.02) and plasma endotoxin levels (p < 0.05), compared to patients with less-severely disturbed hemodynamics. Mortality was 25% in patients with preoperative endotoxemia, compared with no mortality in patients who were not endotoxemic before surgery (p = 0.05). These data demonstrate that endotoxemia in children with congenital heart disease is more common than previously suspected, and is associated with clinical outcomes. We conclude that clinical trials targeting endotoxin will be necessary to determine if endotoxin is a causal, etiologic agent in the disease process.