Systematic review on estimated rates of nephrotoxicity and neurotoxicity in patients treated with polymyxins

BACKGROUND: Nephrotoxicity and neurotoxicity are commonly associated with polymyxin treatment; however, the emergence of multidrug-resistant Gram-negative bacteria with limited therapeutic options has resulted in increased use of polymyxins. OBJECTIVES: To determine the rates of nephrotoxicity and neurotoxicity during polymyxin treatment and whether any factors influence these. DATA SOURCES: Medline, Embase and Cochrane Library databases were searched on 2 January 2020. STUDY ELIGIBILITY CRITERIA: Studies reporting nephrotoxicity and/or neurotoxicity rates in patients with infections treated with polymyxins were included. Reviews, meta-analyses and reports not in English were excluded. PARTICIPANTS: Patients hospitalized with infections treated with systemic or inhaled polymyxins were included. For comparative analyses, patients treated with non-polymyxin-based regimens were also included. METHODS: Meta-analyses were performed using a random-effects model; subgroup meta-analyses were conducted where data permitted using a mixed-effects model. RESULTS: In total, 237 reports of randomized controlled trials, cohort and case-control studies were eligible for inclusion; most were single-arm observational studies. Nephrotoxic events in 35,569 patients receiving polymyxins were analysed. Overall nephrotoxicity rate was 0.282 (95% confidence interval (CI) 0.259-0.307). When excluding studies where >50% of patients received inhaled-only polymyxin treatment or nephrotoxicity assessment was by methods other than internationally recognized criteria (RIFLE, KDIGO or AKIN), the nephrotoxicity rate was 0.391 (95% CI 0.364-0.419). The odds of nephrotoxicity were greater with polymyxin therapies compared to non-polymyxin-based regimens (odds ratio 2.23 (95% CI 1.58-3.15); p < 0.001). Meta-analyses showed a significant effect of polymyxin type, dose, patient age, number of concomitant nephrotoxins and use of diuretics, glycopeptides or vasopressors on the rate of nephrotoxicity. Polymyxin therapies were not associated with a significantly different rate of neurotoxicity than non-polymyxin-based regimens (p 0.051). The overall rate of neurotoxicity during polymyxin therapy was 0.030 (95% CI 0.020-0.043). CONCLUSIONS: Polymyxins are associated with a higher risk of nephrotoxicity than non-polymyxin-based regimens.