Analytical Performance and Potential Clinical Utility of EUCAST Rapid Antimicrobial Susceptibility Testing in Blood Cultures after Four Hours of Incubation

ABSTRACT EUCAST rapid antimicrobial susceptibility testing (RAST) provides antibiotic susceptibility results after 4 to 8 h of incubation. This study assessed the diagnostic performance and clinical usefulness of EUCAST RAST after 4 h. This was a retrospective clinical study performed on blood cultures with Escherichia coli and Klebsiella pneumoniae complex (K. pneumoniae and Klebsiella variicola) at Karolinska University Laboratory (Stockholm, Sweden). The rate of categorized RAST results and the categorical agreement (CA) of RAST with the standard EUCAST 16-to-20-h disk diffusion (DD) method for piperacillin-tazobactam, cefotaxime, ceftazidime, meropenem, and ciprofloxacin were analyzed, as well as the utility of RAST for adjusting the empirical antibiotic therapy (EAT) and the combination of RAST with a lateral flow assay (LFA) for extended-spectrum β-lactamase (ESBL) detection. A total of 530 E. coli and 112 K. pneumoniae complex strains were analyzed, generating 2,641 and 558 readable RAST zones, respectively. RAST results categorized according to antimicrobial sensitivity/resistance (S/R) were obtained for 83.1% (2,194/2,641) and 87.5% (488/558) of E. coli and K. pneumoniae complex strains, respectively. The RAST result categorization to S/R for piperacillin-tazobactam was poor (37.2% for E. coli and 66.1% for K. pneumoniae complex). CA with the standard DD method was over 97% for all tested antibiotics. Using RAST, we detected 15/26 and 1/10 of the E. coli and K. pneumoniae complex strains that were resistant to the EAT. For patients treated with cefotaxime, RAST was used to detect 13/14 cefotaxime-resistant E. coli strains and 1/1 cefotaxime-resistant K. pneumoniae complex strain. ESBL positivity was reported the same day as blood culture positivity with RAST and LFA. EUCAST RAST provides accurate and clinically relevant susceptibility results after 4 h of incubation and can accelerate the assessment of resistance patterns. IMPORTANCE Early effective antimicrobial treatment has been shown to be crucial for improving the outcome of bloodstream infections (BSI) and sepsis. In combination with the rise of antibiotic resistance, this calls for accelerated methods for antibiotic susceptibility testing (AST) for effective treatment of BSI. This study assesses EUCAST RAST, an AST method that yields results in 4, 6, or 8 h after blood culture positivity. We analyzed a high number of clinical samples of Escherichia coli and Klebsiella pneumoniae complex strains and confirm that the method delivers reliable results after 4 h of incubation for the relevant antibiotics for treating E. coli and K. pneumoniae complex bacteremia. Furthermore, we conclude that it is an important tool for antibiotic treatment decision-making and early detection of ESBL-producing isolates.

was just theoretical as potential valuable information for treatment escalation or de-escalation but not through clinical chart review. These are limitations of the study, and should be mentioned as well the single center evaluation. The title should modify accordingly. A suggestion: "The analytical performance and potential clinical utility of EUCAST RAST in blood cultures" -Why do the authors did not include in the study a lateral flow for carbapenemases? They had the meropenem results for the implementation. Specific comments -Lines 161. Why lateral flow was not performed directly from positive blood culture once the identification was performed with Malditof? -Line 223-224. Although in M&M section clarify that these isolates were just flagged. positive until 11 pm.
Reviewer #2 (Comments for the Author): The authors present a study on the performance of the EUCAST RAST (rapid AST from blood cultures). Additionally, they analyze the clinically relevant outcomes such as treatment appropriateness. Notably, the study explicitly investigates the EUCAST RAST after four hours of incubation. This is an interesting aspect of the study as this time point is seen as critical for the reading of disk diffusion. While reading after 6 h and especially 8 h according to EUCAST RAST breakpoints is often regarded as useful, reading after 4 h has been frequently described as practically not feasible due to insufficient growth. Therefore, this study significantly contributes to the knowledge and is of practical relevance.
The methods are technically sound and the results are well described and discussed.
Abstract: "S/R categorized RAST-results were obtained for 83.1 % 38 (2194/2641) and 87.5 % (488/558) for E. coli and K. pneumoniae complex respectively". What was the reason for not-obtaining S/R categorization. ATU? Please the reason in the abstract. This is important because frequent ATU results seem to considerably limit the application of EUCAST RAST worldwide.
Lines 148-149: "The EUCAST standard DD was used as reference (15) and performed in parallel using the same media and disks as described above". Please state whether the same plates as for RAST were further incubated for reading after 16-20 h or other plates were inoculated for the standard disk diffusion?
Lines 233-235: The highest rates of results in the ATU were seen in piperacillin-tazobactam in both E. coli and K. pneumoniae (62.8 % and 33.9 % respectively)" This is a very important data from the study -please include it into the abstract. In many areas, piperacillin-tazobactam is the most commonly used antibiotics in sepsis. High number of ATUs with pip/tazo considerably limits the application of the method.
In the discussion section, please explain the choice of bacterial species included into the study. Why Acinetobacter, enterococci, pneumococci (EUCAST RAST breakpoints are also available for 4 h incubation time) were excluded?
Reviewer #4 (Comments for the Author): This is a real-world evaluation of the implementation of a rapid antimicrobial susceptibility test from EUCAST (RAST) applied directly to positive blood culture bottles for E. coli and Klebsiella pneumoniae at a single microbiology laboratory in Sweden. 530 blood cultures positive for E. coli and 112 positive for K. pneumoniae were included. RAST was read after 4 hours of incubation for five antibiotics and results were compared to EUCAST reference disk diffusion. They also performed a rapid CTX-M lateral flow assay if RAST demonstrated cefotaxime resistance. They found {greater than or equal to}89% of the results were interpretable as S or R at 4 hours for cefotaxime, ceftazidime, meropenem, and ciprofloxacin, but only 37% of the results were interpretable for pip-tazo. Diagnostic performance of interpretable results was generally high. RAST only detected resistance to the empiric antibiotic that was used in 15/26 E. coli and 1/10 K. pneumoniae. The ESBL LFA provided results much earlier than the reference method for ESBL production.
I have the following comments for the authors to respond to: Major comments: 1) The RAST method seems unreliable (at least at 4-hr read) for pip-tazo. There were a high proportion of ATU results with piptazo and of the 15 patients treated empirically with PTZ for PTZ-resistant isolates, PTZ resistance was never detected by RAST. In contrast, performance for cefotaxime was much better. The poor results with pip-tazo should be noted in the Abstract, as well as the antibiotics evaluated in the study. 2) Table 1: The denominators do not appear correct for the calculation of %VME and %ME. For example, for VME for CTX/E. coli, the denominator should be no greater than 53 (the # of resistant isolates). I recommend recalculating these %s with the appropriate denominators and reporting the denominators (e.g., denominator for VME the # of isolates resistant by reference method that had RAST S or R) 3) It is important to also report the diagnostic accuracy of the ESBL LFA test compared to the reference ESBL double disk synergy test. Minor comments: 1) EUCAST RAST can be performed at multiple times (4, 6, 8, and 16-20 hours). This study evaluated a 4-hour RAST. I think the use of 4-hour reads should be in the Title so that a reader can clearly identify this and put the results into context.
2) It is unclear why blood cultures form children were excluded from this analysis 3) Please clarify whether the reference disk diffusion method was performed on isolated colonies or on the positive blood culture broth. 4) For Table 2: do the authors mean "deceased" instead of "diseased"? 5) The rationale for only including patients treated empirically with pip-tazo, meropenem, and ciprofloxacin as empiric antibiotic therapy (and not the entire dataset), should be more clearly explained. I assume it is because these are patients with opportunities for antibiotic de-escalation, but this is not well explained.

Preparing Revision Guidelines
To submit your modified manuscript, log onto the eJP submission site at https://spectrum.msubmit.net/cgi-bin/main.plex. Go to Author Tasks and click the appropriate manuscript title to begin the revision process. The information that you entered when you first submitted the paper will be displayed. Please update the information as necessary. Here are a few examples of required updates that authors must address: For complete guidelines on revision requirements, please see the journal Submission and Review Process requirements at https://journals.asm.org/journal/Spectrum/submission-review-process. Submissions of a paper that does not conform to Microbiology Spectrum guidelines will delay acceptance of your manuscript. " Please return the manuscript within 60 days; if you cannot complete the modification within this time period, please contact me. If you do not wish to modify the manuscript and prefer to submit it to another journal, please notify me of your decision immediately so that the manuscript may be formally withdrawn from consideration by Microbiology Spectrum.
If your manuscript is accepted for publication, you will be contacted separately about payment when the proofs are issued; please follow the instructions in that e-mail. Arrangements for payment must be made before your article is published. For a complete list of Publication Fees, including supplemental material costs, please visit our website.
Corresponding authors may join or renew ASM membership to obtain discounts on publication fees. Need to upgrade your membership level? Please contact Customer Service at Service@asmusa.org.
Thank you for submitting your paper to Microbiology Spectrum.

Reviewer #1 (Comments for the Author):
This is a retrospective single center evaluation, including analytical performance and clinical utility, of EUCAST RAST guidelines. Evaluation was performed with E. coli and Klebsiella pneumoniae complex isolates once identified in blood cultures flagged positives an identified by MALDI TOF. The article is interesting but need to clarify some issues.
General comment 1. In the clinical evaluation, only appropriateness of empiric antibiotic therapy was analyzed but no other issues. This evaluation was just theoretical as potential valuable information for treatment escalation or de-escalation but not through clinical chart review. These are limitations of the study, and should be mentioned as well the single center evaluation. The title should modify accordingly. A suggestion: "The analytical performance and potential clinical utility of EUCAST RAST in blood cultures" We thank the reviewer for this comment on the external validity of the study. On lines 386-389 we describe the limitations of not studying patient records for actual antibiotic decisionmaking and evaluation of clinical outcome. We have now added the single-center study design as a limitation as the reviewer suggested. We also adjusted the title in accordance with the reviewer's suggestion and added the word "potential" in the Introduction as well.
Title: "The analytical performance and potential clinical utility of EUCAST RAST in blood cultures after four hours of incubation" Lines 104-106: "To evaluate the potential clinical usefulness of RAST in antibiotic treatment decision-making, the cefotaxime RAST results for patients treated with piperacillintazobactam, meropenem and ciprofloxacin were analyzed." Lines 388-390: "Furthermore, this was a single center study and similar multi-center studies analyzing the performance of RAST are warranted." 2. Why do the authors did not include in the study a lateral flow for carbapenemases? They had the meropenem results for the implementation. As the reviewer suggested the following sentence was included in the results section of the revised version: Line 224-225 "Blood cultures with gram-negative bacteria that signaled positive before 11:00 a.m. were included in the study."

Reviewer #2 (Comments for the Author):
The authors present a study on the performance of the EUCAST RAST (rapid AST from blood cultures). Additionally, they analyze the clinically relevant outcomes such as treatment appropriateness.
Notably, the study explicitly investigates the EUCAST RAST after four hours of incubation. This is an interesting aspect of the study as this time point is seen as critical for the reading of disk diffusion. While reading after 6 h and especially 8 h according to EUCAST RAST breakpoints is often regarded as useful, reading after 4 h has been frequently described as practically not feasible due to insufficient growth. Therefore, this study significantly contributes to the knowledge and is of practical relevance.
The methods are technically sound and the results are well described and discussed. 3. Lines 148-149: "The EUCAST standard DD was used as reference (15) and performed in parallel using the same media and disks as described above". Please state whether the same plates as for RAST were further incubated for reading after 16-20 h or other plates were inoculated for the standard disk diffusion?
We thank the reviewer for this comment, as the sentence could be interpreted as we later used the RAST plates for standard DD, which was not the case. The intention of the sentence was that the same manufacturer and type of media and disks were used for both EUCAST RAST and standard DD.
As we have written in the manuscript, the EUCAST standard DD was performed in parallel with EUCAST RAST. Thus, a parallel set of plates were incubated and read after 16-20 hours, and we did not re-incubate the EUCAST RAST plates.
We have clarified this in the revised manuscript, and also that the MH agar used for EUCAST RAST was from Oxoid/Thermo Fisher Scientific,Basingstoke, UK.
Lines 148-150: "The EUCAST standard DD was performed in parallel (i.e. from subcultured colonies and zone readings after 16-20 hours of incubation) with media and disks from Oxoid/Thermo Fisher Scientific,Basingstoke, UK. The results from standard DD were used as reference (20)." Lines 137-138: "Both agar plates and disks were from Oxoid/Thermo Fisher Scientific,Basingstoke, UK." 4. Lines 233-235: The highest rates of results in the ATU were seen in piperacillintazobactam in both E. coli and K. pneumoniae (62.8 % and 33.9 % respectively)" This is a very important data from the study -please include it into the abstract. In many areas, piperacillin-tazobactam is the most commonly used antibiotics in sepsis. High number of ATUs with pip/tazo considerably limits the application of the method.
We agree with the reviewer that the low rate of categorized S/R results for piperacillintazobactam is an important result that should be presented in the abstract. We have included a description of this result in the revised version.
Line 38-39: "S/R categorization for piperacillin-tazobactam was poor (37.2 % and 66.1 % respectively)." 5. In the discussion section, please explain the choice of bacterial species included into the study. Why Acinetobacter, enterococci, pneumococci (EUCAST RAST breakpoints are also available for 4 h incubation time) were excluded?

When introducing the EUCAST RAST method in our clinical microbiology laboratory, the lab chose to only perform RAST on gram-negative isolates and to only report results for E. coli and K. pneumoniae complex. The reason for this was that these isolates are the most common gram-negatives isolated from blood cultures in our lab. We have included the following explanation in the Discussion section:
Lines 374-380: "In the present study only E. coli and K. pneumoniae complex were analyzed, even though EUCAST RAST has been validated for several other bacteria (e.g., Acinetobacter baumanii, Pseudomonas aeruginosa, Staphylococcus aureus among others). The reason was that our laboratory chose to implement EUCAST RAST only for E. coli and K. pneumoniae complex, as these bacteria are the most common gram-negatives isolated from BCs in the hospitals served by our laboratory. Studies of diagnostic performance and clinical utility of EUCAST RAST for different species will be useful for further assessment of the method."

Reviewer #4 (Comments for the Author):
This is a real-world evaluation of the implementation of a rapid antimicrobial susceptibility test from EUCAST (RAST) applied directly to positive blood culture bottles for E. coli and Klebsiella pneumoniae at a single microbiology laboratory in Sweden. 530 blood cultures positive for E. coli and 112 positive for K. pneumoniae were included. RAST was read after 4 hours of incubation for five antibiotics and results were compared to EUCAST reference disk diffusion. They also performed a rapid CTX-M lateral flow assay if RAST demonstrated cefotaxime resistance. They found {greater than or equal to}89% of the results were interpretable as S or R at 4 hours for cefotaxime, ceftazidime, meropenem, and ciprofloxacin, but only 37% of the results were interpretable for pip-tazo. Diagnostic performance of interpretable results was generally high. RAST only detected resistance to the empiric antibiotic that was used in 15/26 E. coli and 1/10 K. pneumoniae. The ESBL LFA provided results much earlier than the reference method for ESBL production.
I have the following comments for the authors to respond to: Major comments: 1. The RAST method seems unreliable (at least at 4-hr read) for pip-tazo. There were a high proportion of ATU results with pip-tazo and of the 15 patients treated empirically with PTZ for PTZ-resistant isolates, PTZ resistance was never detected by RAST. In contrast, performance for cefotaxime was much better. The poor results with pip-tazo should be noted in the Abstract, as well as the antibiotics evaluated in the study.
We agree with the reviewer that the EUCAST RAST method's performance with piperacillintazobactam is important and should be presented in the abstract. We also agree that specification of the antibiotics tested should be included in the abstract. We have included both in the revised version of the manuscript.
Line 38-39: S/R categorization for piperacillin-tazobactam was poor (37.2 % and 66.1 % respectively)." Lines 31-35: "The rate of categorized RAST results and the categorical agreement (CA) of RAST with standard EUCAST 16-20 h disk diffusion (DD) for piperacillin-tazobactam, cefotaxime, ceftazidime, meropenem and ciprofloxacin were analyzed, as well as the utility of RAST for adjusting empiric antibiotic therapy (EAT) and combining RAST with a lateral flow assay (LFA) for ESBL detection.
2. Table 1: The denominators do not appear correct for the calculation of %VME and %ME. For example, for VME for CTX/E. coli, the denominator should be no greater than 53 (the # of resistant isolates). I recommend recalculating these %s with the appropriate denominators and reporting the denominators (e.g., denominator for VME the # of isolates resistant by reference method that had RAST S or R) Minor comments: 1. EUCAST RAST can be performed at multiple times (4, 6, 8, and 16-20 hours). This study evaluated a 4-hour RAST. I think the use of 4-hour reads should be in the Title so that a reader can clearly identify this and put the results into context.
We agree with the reviewer and have adjusted the title accordingly.
Title: "The analytical performance and potential clinical utility of EUCAST RAST in blood cultures after four hours of incubation."

It is unclear why blood cultures form children were excluded from this analysis
We wanted to keep the study material as homogeneous as possible and therefore we chose to exclude children.
1. Please clarify whether the reference disk diffusion method was performed on isolated colonies or on the positive blood culture broth.
The reference disk diffusion method was performed on isolated colonies after subculturing of positive BC bottles.