Clinical features and mortality of patients on renal replacement therapy receiving polymyxin B

• 30-day mortality of patients on renal replacement therapy (RRT) treated with polymyxin B (PMB) was high (51.1%).
• Mortality increased with higher Charlson co-morbidity index and APACHE II score, continuous venovenous haemodialysis and Pseudomonas aeruginosa infection.
• PMB dose ≥200 mg/day was independently associated with lower 30-day mortality.
• This fixed dose must not be taken as definitive since higher total daily doses were significantly correlated with dose normalised by weight.

There are no clinical data for polymyxin B (PMB) in patients on renal replacement therapy (RRT). The aim of this study was to evaluate the characteristics of patients on RRT receiving PMB and to identify predictors of 30-day mortality, with special focus on dosage. A multicentre prospective cohort study including patients aged ≥18 years treated with PMB for ≥48 h while on any type of RRT was performed. In total, 88 patients were evaluated, including 34 (38.6%) on continuous venovenous haemodialysis (CVVH) and 54 (61.4%) on intermittent haemodialysis. Most patients (81.8%) received recommended doses between 1.5 mg/kg/day and 3.0 mg/kg/day. The 30-day mortality was 51.1% (45/88 patients). There was no significant association of dose (in mg/kg) with mortality. A PMB average daily dose ≥200 mg was predictive of decreased 30-day mortality in the multivariate model (hazard ratio = 0.35, 95% confidence interval 0.14–0.90; P = 0.03), whilst CVVH (P = 0.04), higher Charlson co-morbidity index (P = 0.02) and Acute Physiology and Chronic Health Evaluation (APACHE) II score (P = 0.04), and Pseudomonas aeruginosa infection (P = 0.001) were independent risk factors for mortality. The results were not changed by the inclusion of patient weight or dose (in mg/kg) in the model, although the latter was significantly correlated with total daily dose. This is the first clinical study to show that higher doses of PMB are associated with lower mortality in patients on RRT.