Amikacin Combined with Fosfomycin for Treatment of Neonatal Sepsis in the Setting of Highly Prevalent Antimicrobial Resistance

Abstract

Antimicrobial resistance (particularly through extended-spectrum β-lactamase and aminoglycoside-modifying enzyme production) in neonatal sepsis is a global problem, particularly in low- and middle-income countries, with significant mortality rates. High rates of resistance are reported for the current WHO-recommended first-line antibiotic regimen for neonatal sepsis, i.e., ampicillin and gentamicin. We assessed the utility of fosfomycin and amikacin as a potential alternative regimen to be used in settings of increasingly prevalent antimicrobial resistance. The combination was studied in a 16-arm dose-ranged hollow-fiber infection model (HFIM) experiment. The combination of amikacin and fosfomycin enhanced bactericidal activity and prevented the emergence of resistance, compared to monotherapy with either antibiotic. Modeling of the experimental quantitative outputs and data from checkerboard assays indicated synergy. We further assessed the combination regimen at clinically relevant doses in the HFIM with nine Enterobacterales strains with high fosfomycin and amikacin MICs and demonstrated successful kill to sterilization for 6/9 strains. From these data, we propose a novel combination breakpoint threshold for microbiological success for this antimicrobial combination against Enterobacterales strains, i.e., MIC_F × MIC_A < 256 (where MIC_F and MIC_A are the fosfomycin and amikacin MICs, respectively). Monte Carlo simulations predict that a standard fosfomycin-amikacin neonatal regimen would achieve >99% probability of pharmacodynamic success for strains with MICs below this threshold. We conclude that the combination of fosfomycin with amikacin is a viable regimen for the empirical treatment of neonatal sepsis and is suitable for further clinical assessment in a randomized controlled trial.

Keywords: amikacin; aminoglycosides; antimicrobial resistance; combination antibiotics; fosfomycin; hollow fiber; mathematical modelling; neonatal sepsis; pharmacodynamics; synergy.

FIG 1

Modeled output for checkerboard assays with three antimicrobial combinations against 16 isolates, with a combined total statistic for each combination. α is the interaction parameter in the Greco model indicating the level of synergy. A CI of >0 indicates the presence of synergy, a CI of <0 indicates antagonism, and a CI containing 0 indicates no interaction, with additive effects only. α and P values for combined statistics are given below the graph. I² represents the heterogeneity in effect between individual strains.

FIG 2

PD output of the 16-arm fosfomycin-amikacin combination HFIM experiment, with the indicated fAUC_0–24 values for each arm. The gray symbol in arm 15 was a real data point in the initial experiment but was not reproducible in repeat experiments. It is presented here for completeness but was not included in the modeling.

FIG 3

PD relationships for the emergence of resistance in relation to modeled fAUC/MIC ratios for each agent. (A) Increasing fosfomycin fAUC/MIC on a background of fixed amikacin fAUC/MIC. (B) Increasing amikacin fAUC/MIC on a background of fixed fosfomycin fAUC/MIC.

FIG 4

Summary of PD outputs of fosfomycin-amikacin combination and monotherapy regimens in the HFIM, shown by pathogen fosfomycin and amikacin MICs (A) and fosfomycin and amikacin fAUC/MIC ratios (B). Success is defined by bacterial kill to sterility at the end of the experiment.