A “Time-Kill” Study to Assess the Effect of Colistin on Gram-Negative Strains: An Experimental Study for Developing an Indigenous Microfluidics-Based Antimicrobial Susceptibility Testing Device

TL;DR

This study explores how quickly a microfluidics-based device can detect antibiotic resistance in gram-negative bacteria using colistin.

Contribution

The study demonstrates that AST results for colistin can be obtained as early as 2 hours using imaging techniques in a microfluidics-based device.

Findings

Log count values were significantly lower in colistin-exposed groups at 2, 4, and 8 hours.

AST results for colistin can be obtained as early as 2 hours using imaging techniques.

Time-kill and spread-plate assays showed consistent results for colistin susceptibility.

Abstract

The emergence of infections caused by multidrug-resistant gram-negative strains has necessitated the reuse of polymyxins, which are repurposed antimicrobials due to their initial toxicity profile. However, the present antimicrobial susceptibility testing (AST) methods against them are challenging to perform or need validation. Through this study, we sought to determine how early we can differentiate between the susceptible and resistant gram-negative strains when exposed to a single breakpoint concentration of antibiotic, as expected from a microfluidics-based AST. We used time-kill assay and cell-viability assay to image the organisms and draw a parallel to the principle of microfluidics-based assay.

Reference and clinical strains of Enterobacterales and Pseudomonas aeruginosa of a particular minimum inhibitory concentration (MIC) value were selected for the study after initially screening by the colistin broth disk elution method and confirmed by colistin broth microdilution (BMD). For Acinetobacter baumannii complex, only a BMD test was used to determine the MIC. Ten strains of varying MICs, ranging from ≤0.5 to ≥16 μg/mL, were selected for time-kill assay against colistin using acridine orange dye to obtain each strain's total cell count of antibiotic-exposed and non-exposed populations at different time points. These strains were then subjected to the spread-plate method for viability counts at various time points.

We obtained 192 unique measurements from 48 colistin-sensitive replicates belonging to 8 strains with MICs ranging from ≤0.5 to 2 μg/mL at different time points. We found that the log count values were statistically lower in the colistin-exposed group at 2, 4, and 8 hours by both time-kill and spread-plate assays. We concluded that it is possible to get AST results against colistin as early as 2 hours if imaging techniques are used for visualizing the bacteria in a microfluidics-based AST device.

Figures