Role of C-reactive protein in response-guided therapy of pyogenic liver abscess.

Background: Protocols for antibiotic treatment of pyogenic liver abscess (PLA) are usually based on clinicians' own experience without any validation. Our study was to evaluate the clinical implication of C-reactive protein (CRP) in predicting treatment outcome and adequacy of antibiotic therapy of PLA. Patients and methods: Patients with PLA in whom white blood cell (WBC) count, erythrocyte sedimentation rate (ESR), and CRP were checked regularly during the clinical course were included. The prolife of CRP during the clinical course was compared with that of ESR and WBC. The usefulness of CRP in predicting the chance of recovery and adequacy of antibiotic therapy was examined. Results: From 2000 to 2011, 109 patients with PLA underwent regular monitoring of WBC, ESR, and CRP. Except for ESR, both WBC and CRP showed an initial rapid reduction in first 3 weeks, followed by a relatively slow decrease. From week 3 to week 6, the CRP ratio (relative to CRP at week 1) of patients with and without adverse events (i.e. including mortality) was compared; a significant difference was found at week 3 (P=0.001), week 4 (P=0.004), week 5 (P=0.011), and week 6 (P=0.018), whereas no statistically significant difference was found in the WBC ratio over the same period. By week 3, a CRP ratio of 0.423 or less was a marker of good outcome (sensitivity 0.846; specificity 0.667) and was also a marker of adequacy of antibiotic therapy of 5 weeks or less (sensitivity 0.786; specificity 0.714) if the ratio was 0.278 or less. Conclusion: Weekly CRP measurement was useful in the identification of patients with PLA with good outcome and adequacy of antibiotic therapy of 5 weeks or less.