A new technology for rapid antimicrobial susceptibility testing – DAMOCLES
With DAMOCLES, a rapid antibiotic susceptibility testing (AST) in less than 3 hours
The DAMOCLES project confirmed the value of a new technology for performing rapid ASTin less than 3 hours from Gram-negative bacilli colonies. Through two articles published in 2023 and 2024, the project demonstrated the robust and universal nature of the technology for performing rapid AST, while complying with an ISO standard that follows the reference methodology for performing AST.
Conventional AST require 16 to 24 hours, and automated methods require approximately 8-12 hours. The DAMOCLES project aimed to develop rapid AST in 3 hours.
The SdFFF (Sedimentation Field Flow Fractionation) technique has emerged as a potential solution for determining AST. SdFFF is a cell sorting technique related to liquid chromatography. It is distinguished by its ability to differentiate cell species without prior labeling, based solely on their biophysical characteristics, including size, density, shape, deformability, and rigidity. The concept underlying the use of SdFFF as a rapid AST method is that bacteria susceptible to an antibiotic will exhibit phenotypic variations after exposure to that antibiotic, while resistant bacteria will not exhibit biophysical changes. Thus, by comparing the elution profile of bacteria treated with an antibiotic to that of the same untreated bacteria, it is possible to determine whether the bacterial strain is susceptible or resistant to that antibiotic. The objectives of the DAMOCLES project were: 1) To optimize culture conditions to reduce the time it takes for bacteria to be available for SdFFF analysis The objective of this component was to optimize culture parameters based on the bacteria/antibiotic combination in order to obtain a bacterial suspension "ready to be injected" into the SdFFF machine within two hours: higher bacterial inoculum, richer culture medium, higher agitation speed, greater oxygenation, etc. 2) To establish standardized protocols for SdFFF AST The objective of this component was to define the optimal elution conditions for SdFFF for each bacterial species/antibiotic combination to measure susceptibility or resistance. 3) To establish clinical proof of the feasibility of rapid AST using SdFFF on a panel of clinical strains. Once a bacterial species/antibiotic pair was optimized and validated, a panel of several hundred clinical strains was analyzed to achieve a certain degree of completeness.
The bacterial strains were obtained from the Bacteriology-Virology-Hygiene Laboratory of the Limoges University Hospital. The antibiotics tested (amoxicillin, amoxicillin-clavulanic acid, cefotaxime, gentamicin, ciprofloxacin, and trimethoprim-sulfamethoxazole) were used at critical concentrations determined by EUCAST/CA-SFM. A calibrated bacterial suspension was incubated in the absence or presence of antibiotics for 2 hours at 37°C. A fraction of the bacterial suspension was injected into the SdFFF machine, and fractograms were obtained.
The fractogram represents the instantaneous variation in the quantity of bacterial species aaccording to the retention time. It consists of two peaks: the first corresponds to the dead volume, the second corresponds to the peak of the microbial population. A retention factor was used to determine the average velocities of the "treated" and "untreated" bacterial populations. Two parameters were evaluated to quantify signal variation: the percentage change in the retention factor (PΔR) and the percentage change in the area under the curve (PΔAUC). Comparative analysis of the "treated" and "untreated" fractograms was performed using the significance threshold, determined by a ROC curve. If the PΔR and/or PΔAUC value was below the significance threshold, the bacterium was considered resistant. If the PΔR and/or PΔAUC value was above the threshold, the bacterium was considered susceptible.
For all strains, a clinical categorization (susceptible or resistant) was performed. These results were compared to those obtained with the reference liquid microdilution method (Sensititre, Thermo Fisher Scientific), and errors were defined as follows: major errors (ME), strains were resistant by SdFFF and susceptible by the reference method; very major errors (VME), strains were susceptible by SdFFF and resistant by the reference method. Categorical agreement (CA), sensitivity (Se) and specificity (Sp) were calculated according to ISO 20776-2. Any discordant result, regardless of the associated error type, was repeated and if the results of the initial test and the repeated test were not the same, the result of the repeated test was used for data analysis.
SdFFF technology was performed to assess the repeatability and reproducibility of the results. In this study, four isolates from three bacterial species commonly encountered in infections and identified as priority pathogens by the WHO in terms of antibiotic resistance (Escherichia coli, Klebsiella pneumoniae and Pseudomonas aeruginosa) were repeatedly tested with two panels of antibiotics. The results of this study showed a perfect concordance of 100% between the AST produced by the SdFFF method and AST obtained by the reference method, demonstrating the ability of SdFFF to accurately detect morphological changes caused by various antibiotics in different bacteria, in less than 3 hours and to ensure the robustness of the method. These results were published in the journal Analytical Chemistry in 2023 (Sedimentation Field-Flow Fractionation: a diagnostic tool for rapid antimicrobial susceptibility testing; DOI: 10.1021/acs.analchem.3c03134).
In addition, clinical performance validation was performed on a broad panel of clinical strains of E. coli (100 strains) and K. pneumoniae (50 strains), with seven antibiotics (amoxicillin, amoxicillin-clavulanic acid, piperacillin-tazobactam, cefotaxime, gentamicin, ciprofloxacin, trimethoprim-sulfamethoxazole). This study enabled the development of decision-making algorithms for the clinical categorization of strains analyzed in SdFFF as susceptible or resistant. The analysis of 149 Enterobacterial strains showed very good results with the reference method, with a categorical agreement (CA) of 98.6%, a sensitivity (Se) of 98.0%, and a specificity (Sp) of 99.0%. Some of these results were published in the Journal of Antimicrobial Chemotherapy in 2024 (Sedimentation field-flow fractionation for rapid phenotypic antimicrobial susceptibility testing: a pilot study; DOI: 10.1093/jac/dkae132).
A second performance study was conducted with a panel of six antibiotics specific to urinary tract infections (amoxicillin, amoxicillin-clavulanic acid, mecillinam, cefixime, temocillin, fosfomycin, and nitrofurantoin). The preliminary results are very encouraging with a CA of 97.3%, a Se of 96.2% and a Sp of 97.6%. We also developed and implemented a protocol to perform ASTdirectly from positive blood cultures. For this study, 50 strains of E. coli and 5 antibiotics of clinical interest (amoxicillin, cefotaxime, gentamicin, ciprofloxacin, trimethoprim-sulfamethoxazole) were tested. The results of this study are promising with an Se of 98.2%, a specificity of 100% and a CA of 99.6%.
The results obtained open up a wide range of perspectives. First of all, it will be important to evaluate the performance of the SdFFF in performing rapid AST using other bacterial species and different families of antibiotics, this will allow testing the robustness and universality of the SdFFF technology. In addition, it will be interesting to continue the work initiated for positive blood cultures and develop rapid AST directly from urine samples. Finally, it will be interesting to test several types of human or animal samples (e.g.: mastitis milk). To add of course the development perspective of the start-up DAMOCLES Diagnostics.
Increase of antimicrobial resistance is a major public health problem. This problem, which mainly concerns Gram-negative bacteria (GNB), occurs in humans as well as in animals and in the environment. Faced with this overall increase in antibiotic resistance levels, the clinician who suspects a bacterial infection must institute a so-called probabilistic antibiotic therapy pending the microbiological results obtained within 36-48 hours. He must therefore choose between a broad-spectrum antibiotic therapy which is likely to be effective in treating the infection but with the side effect of exerting a strong antibiotic selection pressure, and a narrower antibiotic therapy but at the risk of being ineffective in treating the infection of his/her patient. Faced with this dilemma, research is currently being carried out with the aim of reducing the time required to obtain the antimicrobial susceptibility testing (AST) and thus allow the clinician to institute targeted and appropriate antibiotic therapy more quickly. It is in this context that our project is situated.
The goal of the DAMOCLES project is to use an innovative technology to reduce the time it takes to get an AST to adapt antibiotic therapy in less than 24 hours. A preliminary study conducted on strains of Escherichia coli with different antibiotics has shown that this technology can detect the effect of the antibiotic after only 2 hours of incubation. This preliminary study enabled the deposit of a patent in early April 2020.
The DAMOCLES project aims at extending this preliminary study to a large panel of bacterial species and antibiotics, this will define the scope of the test. After a period of research laboratory validation, it is planned to implement the technology in a medical bacteriology laboratory and to carry out tests on routine bacterial strains.
Due to the universal nature of the AST, this project has a great potential for industrial development. Note that at the same time, the DAMOCLES project will be based on the creation of a spin-off dedicated to the development of new devices specifically designed for medical diagnosis in bacteriology (CE marking, medical device).
Project coordination
Olivier BARRAUD (ANTI-INFECTIEUX : SUPPORTS MOLÉCULAIRES DES RÉSISTANCES ET INNOVATIONS THÉRAPEUTIQUES)
The author of this summary is the project coordinator, who is responsible for the content of this summary. The ANR declines any responsibility as for its contents.
Partnership
RESINFIT ANTI-INFECTIEUX : SUPPORTS MOLÉCULAIRES DES RÉSISTANCES ET INNOVATIONS THÉRAPEUTIQUES
Help of the ANR 179,280 euros
Beginning and duration of the scientific project:
January 2021
- 36 Months
