Comparative Neurotoxicity of Polymyxin B and Colistin in Patients With Cancer.
The polymyxins, polymyxin B and colistin, are last-line options for the treatment of infections caused by drug-resistant gram-negative organisms. The comparative neurotoxicity of these agents using contemporary dosing is poorly understood. We sought to compare the incidence of neurotoxicity among recipients of polymyxin B or colistin.
This was a single-center, retrospective cohort study of all patients who received colistin or polymyxin B for the treatment of a serious gram-negative infection between March 2016 and April 2020. The primary outcome of interest was treatment-emergent neurotoxicity, defined as new-onset perioral paresthesia, non-oral paresthesia, peripheral neuropathy, seizure, diaphragmatic paralysis, or other neurotoxicity. Bivariate analyses were performed using Fisher's exact test and the Wilcoxon rank-sum test; no multivariable analyses were performed.
A total of 82 patients receiving polymyxin B (n = 39) or colistin (n = 43) were identified. Neurotoxicity, nearly exclusively paresthesia, was more commonly seen in recipients of polymyxin B (10/39 evaluable patients, 26%) versus colistin (1/42 evaluable patients, 2%; p < 0.01). Neurotoxicity most commonly developed on the first day of therapy and resolved with discontinuation of therapy or prolongation of the infusion duration.
Polymyxin B is associated with significantly more neurotoxicity than colistin in unadjusted analyses. This potential risk should be considered when choosing between these two agents.
This was a single-center, retrospective cohort study of all patients who received colistin or polymyxin B for the treatment of a serious gram-negative infection between March 2016 and April 2020. The primary outcome of interest was treatment-emergent neurotoxicity, defined as new-onset perioral paresthesia, non-oral paresthesia, peripheral neuropathy, seizure, diaphragmatic paralysis, or other neurotoxicity. Bivariate analyses were performed using Fisher's exact test and the Wilcoxon rank-sum test; no multivariable analyses were performed.
A total of 82 patients receiving polymyxin B (n = 39) or colistin (n = 43) were identified. Neurotoxicity, nearly exclusively paresthesia, was more commonly seen in recipients of polymyxin B (10/39 evaluable patients, 26%) versus colistin (1/42 evaluable patients, 2%; p < 0.01). Neurotoxicity most commonly developed on the first day of therapy and resolved with discontinuation of therapy or prolongation of the infusion duration.
Polymyxin B is associated with significantly more neurotoxicity than colistin in unadjusted analyses. This potential risk should be considered when choosing between these two agents.