2240. Can Antibiotic Duration Be Reduced by the Sequential Use of Procalcitonin and Endotoxin in Patients with Sepsis? A Prospective Double-Blind Clinical Trial

BACKGROUND: Patients hospitalized with sepsis are commonly treated with antibiotics for 10–21 days. Evidence is lacking to support this duration. We hypothesize that the sequential measurement of procalcitonin [PCT] (host biomarker) and endotoxin activity assay [EAA] (microorganism biomarker) could provide evidence to safely reduce antibiotic exposure. METHODS: Consented patients with bacteremic sepsis on antibiotics for <48 hours had PCT and EAA measured every other day until discharge or 14 days. Investigators and care providers were blinded to results. Median (interquartile), days to biomarker clearance (reduction from baseline PCT> 80% and EAA> 0.1), and univariate/multivariate regression done for all analyses. Clinical response was evaluated by blinded investigators. RESULTS: We enrolled 215 patients. Median age: 60 years old (18); Female 45%; Baseline: Temperature: 38.5°C (0.7); WBC: 11.4 (7.7), SIRS: 3 (1); APACHE II: 19.2 (8); 60% had severe sepsis (61% Gram-negative; 39% Gram-positive). Sepsis source: 33% abdominal, 22% line, 19% urinary, 12% cutaneous; 8% pulmonary; 6% other. Findings: Day 0: PCT 4.62 (19), EAA 0.53 (0.24); Days to Clearance: PCT: 4.0 (4), EAA: 4.0 (4), and PCT/EAA: 4.0 (4). At the day of biomarker clearance patients also demonstrated clinical response. The median duration of antibiotics was 16 days [7 inpatient plus 9 outpatient-days]. Compared with both total and in-hospital antibiotic duration, time to clearance was significantly shorter: PCT (P < 0.0001), EAA (P < 0.0001), PCT/EAA (P < 0.0001). After multivariate adjustment for disease severity with APACHE II and organ failure, time-to-clearance for each or both biomarkers remained significantly shorter by 12 and 3 days compared with total and in-hospital antibiotic days, respectively. Additionally, a faster time to PCT/EAA clearance was associated with a 75% mortality reduction at 28 days (OR 0.25 (0.09–0.68); p = 0.007). CONCLUSION: The median time-to-clearance of procalcitonin and endotoxin was 4 days, and faster clearance was associated with significant mortality reduction. However, patients received additional 12 days of antibiotics. Our new findings support the use of shorter antibiotic courses for patients with bacteremic sepsis. DISCLOSURES: All authors: No reported disclosures.