Is colistin-associated acute kidney injury clinically important in adults? A systematic review and meta-analysis.

Colistin is the last-resort antimicrobial agent against infections caused by multidrug-resistance Gram-negative bacteria (MDR-GNB). However, a differing risk of colistin-associated acute kidney injury (CA-AKI) has been demonstrated without affecting mortality, thus the association and its importance needs to be questioned. To assess the impact of this adverse effect, a meta-analysis comparing colistin with other antibiotics in treating MDR-GNB infections was conducted. The PubMed, Embase and Cochrane Library electronic databases were searched up to 31 December 2018 for cohort studies and randomised controlled trials with at least two arms with one arm containing colistin-based treatment. The primary endpoint was the incidence of AKI. The secondary endpoint was 30-day all-cause mortality. A total of 34 studies, including 26 regarding colistin-based therapy versus other antibiotics and 9 regarding colistin monotherapy versus combination therapy, were included. The incidence of CA-AKI was 32.3%. Colistin was associated with an 82% higher incidence of AKI than other antibiotics [odd ratio (OR) = 1.82, 95% confidence interval (CI) 1.13-2.92; P = 0.01]. Most CA-AKI events were mild and reversible without a higher rate of mortality or the requirement for renal replacement therapy (RRT). Only 1.0% of patients required RRT for > 4 weeks. Compared with colistin monotherapy, combination therapy was associated with a significantly lower incidence of AKI (OR = 1.46, 95% CI 1.10-1.94; P = 0.009), particularly in combination with a carbapenem (OR = 1.97, 95% CI 1.30-2.99; P = 0.001). In conclusion, CA-AKI might not be an important limitation of colistin in MDR-GNB therapy.