Evaluation of Early Reading of Broth Microdilution Technique for Polymyxin B

Abstract

Delay in the results of standard phenotypic susceptibility tests is the main obstacle to adequate antibiotic treatment. For this reason, the European Committee for Antimicrobial Susceptibility Testing has proposed the Rapid Antimicrobial Susceptibility Testing for the disk diffusion method directly from blood culture. However, to date, there are no studies evaluating early readings of polymyxin B broth microdilution (BMD), the only standardized methodology for assessing susceptibility to polymyxins. This study aimed to evaluate modifications in the BMD technique for polymyxin B using fewer antibiotic dilutions and reading after an incubation time of 8–9 hr (early reading) in comparison to 16–20 hr of incubation (standard reading) for isolates of Enterobacterales, Acinetobacter baumannii complex, and Pseudomonas aeruginosa. A total of 192 isolates of gram-negative bacteria were evaluated and the minimum inhibitory concentrations were read after early and standard incubations. The early reading presented 93.2% of essential agreement and 97.9% of categorical agreement with the standard reading of BMD. Only three isolates (2.2%) presented major errors and only one (1.7%) presented a very major error. These results indicate a high agreement between the early and the standard reading times of BMD of polymyxin B.

Introduction

Polymyxins are old polycationic peptide antibiotics that have been revitalized to combat emergence of carbapenem-resistant gram-negative rods¹ and are considered last line drugs for the treatment of infections caused by multidrug-resistant (MDR) gram-negative bacteria.² Infections caused by MDR bacteria, in particular when associated to polymyxin B resistance³ lead to treatment complications and significant morbi/mortality. High rates of antimicrobial resistance (AMR) are reported in national⁴ and international setting,⁵ being one of the most serious global public health threats in this century.⁶

In the past resistance to polymyxins was mostly due to intrinsic mechanism in few species of gram-negative bacteria and acquired resistance was only occasional. More recently, reports of resistance to polymyxins were due to intrinsic and acquired mechanism in a similar extend.^7,8 This fact emphasizes the need for rapid polymyxin B susceptibility tests to better treat patients with infections by MDR gram-negative rods.⁹ Although molecular methods are considered gold standards to detect antibiotic resistance,¹⁰ they usually present limitations to detect polymyxin B resistance, which is usually due to a variety of genotypes in gram-negative species.¹¹

As the polymyxins do not diffuse properly in agar, the disc diffusion method is not suitable to evaluate the susceptibility of polymyxin B. Moreover, the minimum inhibitory concentrations (MIC) determination by semiautomated equipment such as BD Phoenix^® (Becton and Dickson, New Jersey, USA) or Vitek^® (Biomérieux, Marcy-l'Étoile, France) also do not provide reliable results due to very major errors (VMEs) for polymyxins.^10–13 The broth microdilution (BMD) is the only traditional method recommended by the Clinical and Laboratory Standards Institute (CLSI)¹⁴ and the European Antimicrobial Susceptibility Testing Committee (EUCAST)¹⁵ to evaluate polymyxins susceptibility. However, in 2020, CLSI established two new MIC-based alternative methods to test the susceptibility to polymyxins: colistin broth disk elution and colistin agar test.^14,16 These are simpler methods than the BMD, but still require overnight incubation.

There is another simple method, the drop test, which detects polymyxin B resistance using a drop of polymyxin B on a Mueller–Hinton agar previously inoculated with a 0.5 MacFarland of the bacterial suspension. The latter is an assay methodologically user-friendly but also requires 16 to 18 hr of incubation.¹⁷ There is yet another polymyxin susceptibility assay available: the rapid polymyxin Nordmann and Poirel (NP) test that detects bacterial growth by pH change. The polymyxin NP test is a phenotypic assay that provides results in the same day but does not provide a MIC result.^12,18,19

As most of the methods to test the susceptibility to polymyxins require overnight incubation, there is a need to develop and validate a rapid phenotypic test to evaluate the minimum inhibitory concentration to polymyxin B.

Considering the need of a shorter turnaround time to access polymyxins susceptibility, this study aimed to evaluate the reading of the polymyxin B BMD after an incubation time of 8–9 hr (early reading) in comparison to 16–20 hr of incubation (standard reading) for isolates of Enterobacterales, Acinetobacter baumannii complex, and Pseudomonas aeruginosa. In addition, we have proposed the use of fewer polymyxin B dilutions (1–8 μg/mL).

Materials and Methods

Clinical isolates

A total of 192 nonduplicate gram-negative clinical isolates collected between 2018 and 2021, comprising nine different genera, were selected by convenience sampling considering MIC values (131 Enterobacterales, 50 A. baumannii complex, and 11 P. aeruginosa), collected from a variety of sources (sputum, urine, blood, etc.) were included in this study. The Enterobacterales isolates belonged to the repository collection of “Laboratório de Pesquisa em Resistência Bacteriana—LABRESIS” and the nonfermentative bacteria were obtained from the routine microbiology laboratory in our institution—“Hospital de Clínicas de Porto Alegre.” Identification of isolates was confirmed by the matrix-assisted laser desorption ionization–time of flight mass spectrometry.

The isolates were evaluated by BMD technique after an incubation time of 8–9 hr (early reading) in comparison to 16–20 hr (standard reading). According to the reference method (standard reading), 100% (n = 11/11) of P. aeruginosa, 86% (n = 43/50) of A. baumannii, and 61% (n = 81/131) of Enterobacterales were susceptible to polymyxin B. In total, we found 70.3% (n = 135/192) of isolates susceptible and 29.7% (n = 57/192) resistant to polymyxin B (Fig. 1).

FIG. 1.

Distribution of polymyxin B MIC (μg/mL) values of 192 gram-negative rods according to early and standard readings.

Polymyxin B susceptibility test

The susceptibility tests were carried out from fresh (18–24 hr of incubation) monomicrobial cultures of the isolates, which were grown on MacConkey Agar plates. The MIC values for polymyxin B were obtained by BMD in duplicate according to CLSI/EUCAST^14,15 using a 96-well sterile polystyrene microplate. Cation-adjusted Mueller Hinton broth (CA-MH; BioMerieux, Heidelberg, France) and polymyxin B solution (Sigma-Aldrich, St Louis, MO, USA) were used in all BMD experiments. P. aeruginosa ATCC 27853 and Escherichia coli ATCC 25922 were used as quality control strains. A stock solution of polymyxin B was prepared and serial macrodilution of polymyxin B with CA-MH broth was performed in a 15 mL falcon tube to achieve concentrations from 64 μg/mL to 0.125 μg/mL for both quality control strains (10 concentrations were tested) and 8 μg/mL to 1 μg/mL for the clinical isolates (4 concentrations were tested).

The lines A and B (1–12 wells) of the 96-microplate were assigned to the quality controls (QCs). A total of 50 μL of polymyxin B solution were added in lines A and B as the following: 0.25 μg/mL to wells 11, 0.50 μg/mL to wells 10, 1 μg/mL to wells 9, 2 μg/mL to wells 8, 4 μg/mL to wells 7, 8 μg/mL to wells 6, 16 μg/mL to wells 5, 32 μg/mL to wells 4, 64 μg/mL to wells 3, and 128 μg/mL to wells 2. Wells 1 were used for growth control to which 100 μL of the standardized bacterial inoculum were added; wells 12 were sterility control to which 100 μL of CA-MH were added.

Lines C to H from the wells 1 to 12 were destined for clinical samples and the plate was divided into two parts: (1) lines C to H from 1 to 6; (2) lines C to H from 7 to 12. For clinical samples, a total of 50 μL of polymyxin B solution was added in lines C to H, as the following: 2 μg/mL to wells 5 and 11, 4 μg/mL to wells 4 and 10, 8 μg/mL to wells 3 and 9, and 16 μg/mL to wells 2 and 8. Wells 1 and 7 were growth control to which 100 μL of standardized bacterial inoculum were added; wells 6 and 12 were sterility control to which 100 μL of CA-MH were added.

For the QCs, 50 μL of the standardized bacterial inoculum were added to the wells 11 to 2 (lines A and B); for clinical samples, 50 μL of the standardized bacterial inoculum were added to the wells 5 to 2 and 11 to 8 (lines C to H), in duplicate. The microtiter plates were incubated at 35 ± 1°C in ambient air. The first reading was carried out after 8–9 hr of incubation (early reading) and the plate was incubated again up to 16–20 hr and a new reading was made (standard reading).

After incubation, the presence or absence of bacterial growth was visually evaluated. Any turbidity (even slight turbidity) was considered growth. The MIC was established as the lowest polymyxin B concentration, which did not present bacterial growth. The Brazilian Committee on Antimicrobial Susceptibility Testing (BrCAST) breakpoints for polymyxin B susceptibility of ≤2 μg/mL and >2 μg/mL were used to classify isolates into susceptible and resistant, respectively.²⁰

Data analysis

Essential agreement (EA) represented the rate of isolates with the same MIC in early and standard readings. Categorical agreement (CA) indicates the number of isolates grouped in the same susceptibility category (susceptible or resistant). Categorical errors (CE) were defined as VMEs when isolates were susceptible (S) to polymyxin B in early incubation and resistant (R) in standard incubation; major errors (MEs) were defined when isolates were categorized as R in early incubation and S in standard incubation. As polymyxin B does not have intermediate category, minor errors were not analyzed. Results of EA, CA, and CE were evaluated according to the regulatory parameters recommended by Food and Drug Administration.²¹ The test was repeated twice for the isolates, which presented different results between the early and standard readings.

Ethical approval

Approval was obtained from the Ethics Committee of Hospital de Clínicas de Porto Alegre (CAAE: 41738520.0.0000.5327).

Results

After 8–9 hr of incubation, all isolates presented growth which allowed the MIC reading. In all, 64.5% (124/192) of isolates presented MIC ≤1 μg/mL in early incubation and 61.4% (118/192) in standard incubation; 4.68% (9/192) present MIC of 2 μg/mL in early incubation and 8.85% (17/192) in standard incubation; 10.9% (21/192) present MIC of 4 μg/mL in early incubation and 8.3% (16/192) in standard incubation; 5.2% (10/192) present MIC of 8 μg/mL in early and standard incubation; and 14.6% (28/192) present MIC >8 μg/mL in early incubation and 16.1% (31/192) in standard incubation.

The early reading presents 93.2% (179/192) of EA and 97.9% (188/192) of CA in comparison to standard reading. False susceptibility (VMEs) at 8–9 hr of incubation was 1.7% (one isolate of Serratia marcescens with MIC of ≤1 μg/mL in early reading and >8 μg/mL in standard reading) and the corresponding percentage for false resistance (MEs) was 2.2% (three isolates of E. coli with MIC of 4 μg/mL in early reading and 2 μg/mL in standard reading) (Table 1). All of the MEs were related to twofold polymyxin B MIC value increase. Early reading failed to detect only one polymyxin B-resistant isolate (Table 2 and Fig. 1).

Table 1.

Results of Agreements and Errors of Early Reading in Comparison to Standard Reading According to Different Groups of Gram-Negative Rods

	VME	ME	CA	EA	Total of isolates
Enterobacterales	2.0%	3.7%	96.9%	93.1%	131
A. baumannii complex	0	0	100%	94.0%	50
P. aeruginosa	0	0	100%	90.9%	11
Total	1.7%	2.2%	97.9%	93.2%	192

CA, categorical agreement; EA, essential agreement; ME, major error; VME, very major error.

Table 2.

Species Distribution and Comparison Between Polymyxin B MIC Values of Early (8–9 hr) and Standard Incubations (16–20 hr)

Number of isolates	Species	MIC (μg/mL)
Number of isolates	Species	Early incubation	Standard incubation
3	A. baumannii complex	>8	>8
38	A. baumannii complex	≤1	≤1
3	A. baumannii complex	≤1	2
2	A. baumannii complex	2	2
3	A. baumannii complex	8	8
1	A. baumannii complex	4	4
10	P. aeruginosa	≤1	≤1
1	P. aeruginosa	≤1	2
1	K. aerogenes	4	8
20	E. cloacae complex	≤1	≤1
1	E. bugandensis	>8	>8
1	E. bugandensis	≤1	≤1
3	E. coli	4	2
7	E. coli	≤1	≤1
11	E. coli	4	4
1	E. coli	≤1	2
7	E. coli	2	2
1	K. oxytoca	≤1	≤1
1	K. oxytoca	>8	>8
40	K. pneumoniae	≤1	≤1
7	K. pneumoniae	>8	>8
5	K. pneumoniae	8	8
1	K. pneumoniae	4	8
1	K. pneumoniae	8	>8
1	M. morganii	>8	>8
1	P. mirabilis	>8	>8
4	S. enterica	4	4
1	S. enterica	≤1	≤1
1	S. marcescens	≤1	>8
14	S. marcescens	>8	>8
1	S. marcescens	8	>8

MIC, minimum inhibitory concentrations.

Considering results related to each individual group, there were no differences in the susceptibility patterns using rapid and standard readings for the nonfermenters group. On the contrary, only three isolates of Enterobacterales were classified as resistant and susceptible according to the early and standard readings, respectively. Only one isolate was classified as susceptible according to the early reading and resistant by the standard reading (Table 1).

Discussion

Our study included 192 clinical isolates from a tertiary care hospital in southern Brazil, which represented gram-negative species (Enterobacterales, A. baumannii, and P. aeruginosa) with a variety of MICs to polymyxin B.

From the 192 isolates, a total of 188 isolates were categorically correctly classified by early reading of BMD technique for polymyxin B, indicating a high agreement with standard incubation times (93.2% and 97.9% of EA and CA, respectively). Only four discrepancies were observed when the early readings were compared to the standard readings: one isolate presenting an underestimated MIC value - VME (the standard reading resulted in a high MIC to polymyxin B, whereas the early reading resulted in a low MIC to polymyxin B) and three isolates presenting an overestimated MIC value – ME (the standard reading resulted in low MIC, whereas the early reading indicated high MIC). VME was observed in an isolate of Serratia spp. (MIC ≤1 in the early reading and >8 in the standard reading).

As the genus Serratia spp. presents intrinsic resistance to polymyxin B, we assumed that the short time of incubation was not enough for the isolate to express the resistance mechanism. However, we have tested other 15 isolates of Serratia spp. and all of them proved to present MIC ≥8 μg/mL, which indicates that the discrepancy of the VME was probably due to a characteristic peculiar only to that isolate. The three isolates which presented MEs had MIC value of 4 μg/mL and 2 μg/mL according to early and standard readings, respectively. Noteworthy, polymyxin B does not have the “intermediate” category and this could explain the false-resistance results (ME), which are due to the difference of only one log2 dilution.

BrCAST 2022 has currently added polymyxin B breakpoints into brackets. According to EUCAST, it means that these breakpoints, based on epidemiological cut-off, can distinguish isolates with and without phenotypically detectable resistance mechanisms. Therefore, when reporting a susceptible result, a comment should be added suggesting that this antimicrobial may have clinical evidence for monotherapy lacking and recommending it to be used in combination with another active agent.²⁰ Nonetheless, methods for determining polymyxin B susceptibility continue to be necessary and performed as recommended in the literature. Commercial panels only allow a categorical classification, as the MICs obtained with this test are superior to those obtained with BMD. Even more, commercial panels for nonfermenters are less comparable to the BMD benchmark test.²²

In the Rapid Antimicrobial Susceptibility Testing study developed by the EUCAST, 4, 6, and 8 hr of incubation were evaluated and specific breakpoints were established for each microorganism and time for the disk diffusion method directly from blood culture.²³ In the present study, we compared early reading of BMD for polymyxin B with the standard reading, and we found no need to change the standard MIC breakpoints as the results indicated a high concordance between both readings.

We have reduced the number of dilutions of polymyxin B in the BMD technique as the dilutions of 1 to 8 μg/mL of polymyxin B are enough to classify the isolates into susceptible and resistant categories. The reduced number of polymyxin B dilutions in a 96-well microtiter plate allowed the evaluation of six isolates in duplicate instead of only three isolates using in the original number or dilutions and this approach would reduce time and costs for the routine of clinical microbiology laboratories.

The early reading of BMD technique for polymyxin B provides reliable results for relevant MDR gram-negative rods in 8–9 hr of incubation and may reduce the “time to report” of antimicrobial susceptibility in 8 to 12 hr considering EA and CA when compared with standard reading. Therefore, the early reading enables the clinical microbiology to report the MIC of polymyxin B in the same day that the bacteria is identified and this would favor the proper treatment of the patient.²

Conclusions

Our results demonstrated an overall CA of 97.9% and they can be obtained after 8–9 hr of incubation for isolates of Enterobacterales, A. baumannii, and P. aeruginosa, enabling early evaluation of BMD to polymyxin B. The evaluation of polymyxin B activity through susceptibility testing with less antimicrobial dilutions becomes more practical and cheaper for busy clinical laboratories.

Footnotes

Acknowledgments

The authors thank laboratory of microbiology of Hospital de Clínicas de Porto Alegre, “Fundo de Incentivo a Pesquisa e Eventos do Hospital de Clínicas de Porto Alegre (FIPE/HCPA)” as well as of “INPRA—Instituto Nacional de Pesquisa em Resistência Antimicrobiana—Brazil.”

Authors' Contributions

All authors comprise the qualification for authorship and approved the article before submission, including the names and order of authors.

Disclosure Statement

No conflict of interest declared.

Funding Information

This study received a grant of “Fundo de Incentivo a Pesquisa e Eventos do Hospital de Clínicas de Porto Alegre (FIPE/HCPA)—(No. 2020-0737)” as well as of “INPRA—Instituto Nacional de Pesquisa em Resistência Antimicrobiana—Brazil (INCT/CNPq: 465718/2014-0 and INCT/FAPERGS: 17/2551-0000514-7).”

References

Olaitan

, Morand

, Rolain

. Mechanisms of polymyxin resistance: Acquired and intrinsic resistance in bacteria. Front Microbiol, 2014; 5:1–18; doi: 10.3389/fmicb.2014.00643

Yang

, Pogue

, Li

, et al. Agents of last resort: An update on polymyxin resistance. Infect Dis Clin North Am, 2020; 34:723–750; doi: 10.1016/j.idc.2020.08.003

Cerceo

, Deitelzweig

, Sherman

, et al. Multidrug-resistant gram-negative bacterial infections in the hospital setting: Overview, implications for clinical practice, and emerging treatment options. Microb Drug Resist, 2016; 22:412–431; doi: 10.1089/mdr.2015.0220

Sampaio

JLM

, Gales

. Antimicrobial resistance in Enterobacteriaceae in Brazil: Focus on β-lactams and polymyxins. Braz J Microbiol, 2016; 47:31–37; doi: 10.1016/j.bjm.2016.10.002

Jeannot

, Bolard

, Plésiat

. Resistance to polymyxins in Gram-negative organisms. Int J Antimicrob Agents, 2017; 49:526–535; doi: 10.1016/j.ijantimicag.2016.11.029

Prestinaci

, Pezzotti

, Pantosti

. Antimicrobial resistance: A global multifaceted phenomenon. Pathog Glob Health, 2015; 109:309–318; doi: 10.1179/2047773215Y.0000000030

Wink

, Caierão

, Nunes

AGA

, et al. Evaluation of EDTA and dipicolinic acid in broth microdilution with polymyxin B as a phenotypic test to detect the mcr-1 gene. Microb Drug Resist, 2020; 26:329–333; doi: 10.1089/mdr.2019.0172

Poirel

, Jayol

, Nordmann

. Polymyxins: Antibacterial activity, susceptibility testing, and resistance mechanisms encoded by plasmids or chromosomes. Clin Microbiol Rev, 2017; 30:557–596; doi: 10.1128/CMR.00064-16

Deris

, Kumar S. Rational use of intravenous polymyxin B and colistin: A review. Med J Malaysia, 2018; 73:351–359.

10.

Bardet

, Rolain

. Development of new tools to detect colistin-resistance among Enterobacteriaceae strains. Can J Infect Dis Med Microbiol, 2018; 2018:1–25; doi: 10.1155/2018/3095249

11.

Stefaniuk

, Tyski

. Colistin resistance in Enterobacterales strains: A current view. Pol J Microbiol, 2019; 68:417–427; doi: 10.33073/pjm-2019-055

12.

Jayol

, Nordmann

, Lehours

, et al. Comparison of methods for detection of plasmid-mediated and chromosomally encoded colistin resistance in Enterobacteriaceae. Clin Microbiol Infect, 2018; 24:175–179; doi: 10.1016/j.cmi.2017.06.002

13.

Food and Drug Administration. No Title. Class 2 Device Recall VITEK 2 Gram Negative test kit containing colistin (cs01n); 2017. Available from: https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfRes/res.cfm?id=155857 [Last accessed: December 15, 2021].

14.

CLSI. Performance Standards for Antimicrobial Susceptibility Testing. Clinical and Laboratory Standards Institute: Wayne, PA, USA; 2020.

15.

CLSI-EUCAST. Recommendations for MIC determination of colistin (polymyxin E) as recommended by the joint CLSI-EUCAST Polymyxin Breakpoints Working Group; 2016. Available from: https://www.eucast.org/fileadmin/src/media/PDFs/EUCAST_files/General_documents/Recommendations_for_MIC_determination_of_colistin_March_2016.pdf [Last accessed: December 22, 2021].

16.

Pasteran

, Danze

, Menocal

, et al. Simple phenotypic tests to improve accuracy in screening chromosomal and plasmid-mediated colistin resistance in Gram-negative bacilli. J Clin Microbiol, 2021; 59:1–9; doi: 10.1128/JCM.01701-01720

17.

Conceição-Neto

, Costa

, Pontes

, et al. Difficulty in detecting low levels of polymyxin resistance in clinical Klebsiella pneumoniae isolates: Evaluation of rapid polymyxin NP test, colispot test and superpolymyxin medium. New Microbes New Infect, 2020; 36:1–7; doi: 10.1016/j.nmni.2020.100722

18.

Simar

, Sibley

, Ashcraft

, et al. Evaluation of the rapid polymyxin NP test for polymyxin B resistance detection using Enterobacter cloacae and Enterobacter aerogenes isolates. J Clin Microbiol, 2017; 55:3016–3020; doi: 10.1128/JCM.00934-17

19.

Sadek

, Tinguely

, Poirel

, et al. Rapid polymyxin/pseudomonas NP test for rapid detection of polymyxin susceptibility/resistance in Pseudomonas aeruginosa. Eur J Clin Microbiol Infect Dis, 2020; 39:1657–1662; doi: 10.1007/s10096-020-03884-x

20.

BrCAST. Brazilian Committee on Antimicrobial Susceptibility Testing. Breakpoints tables for interpretation of MICs and zone diameters. BrCAST: Rio de Janeiro; 2021.

21.

Food and Drug Administration. Guidance for Industry and FDA Class II Special Controls Guidance Document: Antimicrobial Susceptibility Test (AST) Systems. 2009:1–42.

22.

Yusuf

, Van Westreenen

, Goessens

, et al. The accuracy of four commercial broth microdilution tests in the determination of the minimum inhibitory concentration of colistin. Ann Clin Microbiol Antimicrob, 2020; 19(1):1–8; doi: 10.1186/s12941-020-00383-x

23.

Åkerlund

, Jonasson

, Matuschek

, et al. EUCAST rapid antimicrobial susceptibility testing (RAST) in blood cultures: Validation in 55 European laboratories. J Antimicrob Chemother, 2020; 75:3230–3238; doi: 10.1093/jac/dkaa333