A Multi-centre Randomised Open-label Active Comparator-controlled Non-inferiority Trial Comparing Oral to Intravenous Antibiotics in the Early Management of Klebsiella Pneumoniae Liver Abscess

Background: Klebsiella pneumoniae liver abscess is the most common etiology of liver abscess in Singapore and much of Asia, and its incidence is increasing. Current management includes prolonged intravenous antibiotic therapy, but there is limited evidence to guide oral conversion. The implicated K1/K2 capsule strain of Klebsiella pneumoniae is almost universally susceptible to ciprofloxacin, an antibiotic with high oral bioavailability. Our primary aim is to compare the efficacy of early (<1 week) step-down to oral antibiotics, to continuing 4 weeks of intravenous antibiotics, in patients with Klebsiella liver abscess.

Methods/Design: The study is designed as a multi-centre randomised open-label active comparator-controlled non-inferiority trial, with a non-inferiority margin of 12%. Eligible participants will be inpatients over the age of 21 with a CT or ultrasound scan suggestive of a liver abscess, and Klebsiella pneumoniae isolated from abscess fluid or blood. Randomisation into intervention or active control arms will be performed with a 1:1 allocation ratio. Participants randomised to the active control arm will receive IV ceftriaxone 2 grams daily to complete a total of 4 weeks of IV antibiotics. Participants randomised to the intervention arm will be immediately converted to oral ciprofloxacin 750mg twice daily. At week 4, all participants will have abdominal imaging and be assessed for clinical response (CRP <20 mg/l, absence of fever, plus scan showing that the maximal diameter of the abscess has reduced). If criteria are met, antibiotics are stopped; if not, oral antibiotics are continued, with reassessment for clinical response fortnightly. If criteria for clinical response are met by week 12, the primary endpoint of clinical cure is met. A cost analysis will be performed to assess the cost saving of early conversion to oral antibiotics, and a quality-of-life analysis will be performed to assess if treatment with oral antibiotics is less burdensome than prolonged IV antibiotics.

Discussion: Our results would help inform local and international practice guidelines regarding the optimal antibiotic management of Klebsiella liver abscess. A finding of non-inferiority may translate to the wider adoption of a more cost-effective strategy that reduces hospital length of stay and improves patient-centered outcomes and satisfaction.