Extensive Therapeutic Drug Monitoring of Colistin in Critically Ill Patients Reveals Undetected Risks

Abstract

(1) Background: With the rise of multi-/pan-drug resistant (MDR/PDR) pathogens, the less utilized antibiotic Colistin has made a comeback. Colistin fell out of favor due to its small therapeutic range and high potential for toxicity. Today, it is used again as a last resort substance in treating MDR/PDR pathogens. Although new guidelines with detailed recommendations for Colistin dosing are available, finding the right dose in critically ill patients with renal failure remains difficult. Here, we evaluate the efficiency of the current guidelines’ recommendations by using high resolution therapeutic drug monitoring of Colistin. (2) Methods: We analyzed plasma levels of Colistin and its prodrug colisthimethate sodium (CMS) in 779 samples, drawn from eight PDR-infected ICU patients, using a HPLC-MS/MS approach. The impact of renal function on proper Colistin target levels was assessed. (3) Results: CMS levels did not correlate with Colistin levels. Over-/Underdosing occurred regardless of renal function and mode of renal replacement therapy. Colistin elimination half-time appeared to be longer than previously reported. (4) Conclusion: Following dose recommendations from the most current guidelines does not necessarily lead to adequate Colistin plasma levels. Use of Colistin without therapeutic drug monitoring might be unsafe and guideline adherence does not warrant efficient target levels in critically ill patients.

Steady‐state pharmacokinetic and pharmacodynamic profiling of colistin in critically ill patients with multi‐drug–resistant gram‐negative bacterial infections, along with differences in clinical, microbiological and safety outcome

“…4,30 Previous pharmacokinetic studies have reported large variations in the t ½ of colistin in critically ill patients with mean values ranging from 1.6 to 22 hours. 9,25,33 The t ½ of colistin found in our study, that is 2.99 ± 0.43 hours, is consistent with previous reports. 21,28 Variability in the V d and CL was determined to be accountable for the heterogenicity in the t ½ of colistin.…”

“…The terminal half-life of colistin varies widely and values between 1.6 and 22 hours. 33 The differences in the terminal half-lives observed in previous studies primarily arise from the differences in the V d , which is widely altered by pathophysiological changes occurring in critically ill patients. 28 Also, studies describing a longer terminal half-life of colistin, that is >6 hours, have a rapid workup procedure following sampling, thereby reducing ex vivo conversion of CMS to colistin.…”

“…Colistin concentrations measured just before administration of the next dose would minimize ex vivo conversion of CMS to colistin, minimize artificially elevated plasma colistin levels and therefore help in clinical practice to enhance colistin dosing. 24,33 Our study had a few limitations, which need to be addressed. We did not use a formal method to determine the sample size for our study.…”

“…Other mechanisms of colistin elimination include enzymatic degradation in various organs rich with proteases/peptidases, with the exact fraction remaining unclear. 33 There has been an inconsistency in the available pharmacokinetic evidence regarding the elimination half-life of colistin. The terminal half-life of colistin varies widely and values between 1.6 and 22 hours.…”

Editorial for Special Issue “Multidrug-Resistant Pathogens”

“…The authors evaluated the efficiency of the current guidelines’ recommendations by using high resolution therapeutic drug monitoring of colistin. The authors analyzed plasma levels of colistin and its pro-drug colisthimethate sodium (CMS) in 779 samples, drawn from eight intensive care unit (ICU) patients with pan-drug resistant (PDR) infections [ 6 ]. This study found that CMS levels did not correlate with colistin levels and over- or under-dosing occurred regardless of renal function and the mode of renal replacement therapy, while colistin elimination half-time appeared to be longer than previously reported [ 6 ].…”