Evaluation of Polymyxin NP Test as a Rapid Method for Detection of Polymyxins Resistance in Enterobacteriaceae

(35.7%). For both colistin and polymyxin MIC, the polymyxin NP test exhibited an overall sensitivity, specificity, PPV, and NPV of (77.8%, 91.7%, 87.5%, 84.6%) and (93.3%, 92.6%, 87.5%, 96.2%) respectively. Seventeen isolates were positive for mcr-1 gene (40.5%) and five isolates were positive for mcr-2 gene (11.9%). Conclusion: The rapid polymyxin NP test is an easy and quick that can reliably detect both polymyxin-resistant and susceptible Enterobacteriaceae isolates. This test can be implemented as a screening tool in the outbreak management and active surveillance for presence of Polymyxins resistance. Therefore, combined with PCR, it can play a substantial role to contain antimicrobial resistance and prevent transfer of resistance genes among patients.


INTRODUCTION
The carbapenemase-producing Enterobacteriaceae (CPE) are among the most clinically significant multidrug-resistant bacteria (MDR).Owing to the increase in infection with MDR bacteria, along with sparsely available antibiotic options, polymyxin B and colistin usage has risen again after being halted in the 1970s.Despite their potential toxicity, polymyxins gained worldwide attention as CPE are still susceptible to this old class of antimicrobials [1][2][3].
This mounting colistin usage can be the origin of the acquisition of Enterobacteriaceae of colistin resistance in addition to the carbapenem resistance [4].In research from the USA, 13% of carbapenemresistant Klebsiella pneumoniae isolates were resistant to colistin [5].
Lipopolysaccharide modification by adding either or both phosphoethanolamine and 4-amino-l-arabinose cationic groups to lipopolysaccharide can give rise to colistin acquired resistance in Enterobacteriaceae through reducing polymyxin binding to the bacterial outer membrane.This can be associated with chromosome encoded mechanisms [6].Other researchers reported that resistance might be plasmid-mediated through the mcr-1 gene [7-10].
Though broth microdilution (BMD) is the standard method for testing the susceptibility to polymyxins, it is meticulous and needs a long time (24 hours).Other techniques such as disk diffusion and E-test have been proposed for performing polymyxin antimicrobial susceptibility.They can also give results in 18-24 hours.But due to polymyxin molecules' poor diffusion in agar, these tests yielded high false susceptibility results (up to 32%) [11,12].
To overcome these drawbacks, rapid two-hour assays were developed using the principle of detecting acidic products of bacterial metabolism in the existence of colistin or polymyxin B [ [13][14][15][16].
Rapid Polymyxin NP test was introduced for quickly detecting colistin-resistant Enterobacteriaceae.The basis of this test is detecting the colour change caused by rapid glucose metabolization by the growing bacteria in certain concertation of colistin or Polymyxin B [14].Despite being sensitive and specific, these assays do not supply data on the resistance mechanisms [17].
So, the current study aimed to assess the diagnostic efficacy of the rapid polymyxin NP test using Polymyxin B for the diagnosis of colistin-and polymyxin resistant Enterobacteriaceae and determine the frequency of mcr-1 and mcr-2 genes among the colistinresistant Enterobacteriaceae at Ain Shams University hospitals (ASUH).

Bacterial strains:
This prospective study included 42 isolates of MDR-Enterobacteriaceae collected from different clinical samples of 42 patients submitted for routine culture and sensitivity in main microbiology laboratory, ASUH.This work was conducted during the period from April 2020 to August 2020.

Study procedures:
All Enterobacteriaceae isolates were preliminary identified by conventional microbiological techniques including colonial morphology, gram stain characteristics and biochemical reactions.Final identification was confirmed by Vitek2C system (Biomerieux, France).
Multidrug-resistant Enterobacteriaceae isolates were chosen to be resistant to one or more agents in at least three groups of antibiotics by disc diffusion method as per CLSI breakpoints [18].Broth microdilution method (BMD) was performed to detect the minimum inhibitory concentration (MIC) for polymyxin B and colistin according to EUCAST breakpoints [19].Subculture was performed for all isolates on Mueller Hinton agar.Then, Rapid polymyxin NP test was done [14].

Antibiotic susceptibility testing by broth microdilution method (BMD):
The standard BMD was performed and interpreted as per EUCAST [19] to determine the polymyxin B and colistin MICs using cationadjusted Mueller-Hinton broth (MHB-CA), Colistin sulphate and polymyxin B powder (Sigma-Aldrich Chemical Co.).Stock solutions for both Colistin sulphate and Polymyxin B sulphate were prepared at a concentration of (50mg/ml) from which a fresh working solution was prepared to reach a final concentration of 64 ug/ml.Escherichia coli ATCC 25922 was used as a negative control for susceptibility testing.

Polymyxin NP test:
We carried out Polymyxin NP test using Polymyxin B antibiotic according to the experimental procedure explained by Nordmann et al. [14].To prepare solutions for the polymyxin NP test, we used the following reagents: Polymyxin B powders, Phenol red powder, MHB-CA, 10% HCl solution, and anhydrous glucose.We purchased all reagents from (Sigma-Aldrich Chemical Co.).After preparation of the reagents, we inoculated the 96well polystyrene micro-test plates with a freshly prepared 3.0 McFarland bacterial suspension.Then, we incubated the inoculated tray for up to 4 hours at 35 ± 2°C in ambient air without sealing and inspected it visually after 10 min and then every hour for 4 hours.We used E. coli ATCC 25922 as a Polymyxin-susceptible control and Proteus mirabilis ATCC 12453 as Polymyxin-resistant isolates.

Conventional Polymerase chain reaction (conventional PCR):
We examined all the 42 Enterobacteriaceae isolates enrolled in the current study for the presence of mcr-1 and mcr-2 genes using conventional PCR [8].

DNA extraction, purification, and detection
DNA extraction and purification using GeneJET PCR Purification Kit (Qiagen, USA) was performed as per the manufacturer's instructions.For DNA amplification and detection, we used a master mix consisting of Thermo Scientific DreamTaq Hot Start Green DNA Polymerase (Qiagen, USA), primers of mcr-1 (forward primer: AGTCCGTTTGTTCTTGTGGC, reverse primer: AGATCCTTGGTCTCGGCTTG) and mcr-2 (forward primer: CAAGTGTGTTGGTCGC AGTT, reverse primer: TCTAGCCCGACAAGCA TACC) genes.PCR was done using a thermal cycler with the following thermal cycling conditions: 95°C for 3 minutes, then 40 cycles of 95°C for 30 sec, 55°C for 30 sec, 72°C for 1 minute, and finally 72°C for 15 minutes.
We used E. coli ATCC 25922 as a negative control and Proteus mirabilis ATCC 12453 as a positive control.The quality control reference strains were supplied by (Remmel, UK).
We used gel electrophoresis and DNA-binding dye to visualize the results of PCR reaction.A standard, or DNA ladder, was included so that the size of the fragments in the PCR sample can be determined.For mcr-1 gene, the PCR reaction produced a 320-base pair (bp) and for mcr-2 gene produced a 715-base pair (bp) [20].

Statistical Analysis:
The collected data were processed using Statistical Package for Social Science (SPSS 20).For descriptive statistics, we calculated frequency and percentage of non-numerical data.As for analytical statistics, Chi-Square test was used to examine the relationship between two qualitative variables.Fisher's exact test was used when the expected count is less than 5 in more than 20% of cells.We calculated Sensitivity, Specificity, Positive predictive value (PPV), Negative predictive value (NPV), and accuracy to detect the diagnostic performance.Major errors and very major errors were calculated.P value was used to indicate levels of significance where P0.05 = non-significant (NS), P<0.05 = significant (S), and P<0.01 = highly significant (HS).

RESULTS
In this study, 42 well identified MDR-Enterobacteriaceae isolates (36 isolates of Klebsiella (85.7%) and 6 isolates of E.coli (14.3%)) were randomly obtained from patients' specimens during the period from April 2020 to the end of August 2020.

Antibiotics intake:
Among the 42 patients included in our study, we found that 30 (71.4%) patients received antibiotics.Meropenem rated the first among all the used antibiotics where 10/30 (33.3%) patients received it as shown in table (1).

Antimicrobial susceptibility testing (AST):
As for the antibiotic susceptibility testing, although not all isolates included in our study were tested for all antibiotics, yet, all of those tested recorded 100% resistance rate against almost all β-lactams with exception of ceftazidime (97.6%) and imipenem (90.0%).The results of the rest of the antibiotics are shown in table (2).

Correlation between Phenotypic methods and MIC for both Colistin and Polymyxin B:
The correlation between MIC method for colistin versus polymyxin B MIC, Polymyxin NP test and PCR for mcr-1 and mcr-2 genes, was highly statistically significant as shown in table (8).
Similarly, the correlation between the MIC method for polymyxin B versus colistin MIC, Polymyxin NP test and PCR for mcr-1 and mcr-2 genes, was highly statistically significant (P<0.01) as shown in table (9).

The diagnostic performance and evaluation of Polymyxin NP test:
The diagnostic performance of Polymyxin NP test compared to MIC for colistin and polymyxin B was summarized in tables ( 10) and ( 11) respectively.

Table (1):
The percentage of antibiotic intake among the 30 studied patients who received antibiotics in the current study.

DISCUSSION
The carbapenemase-producing Enterobacteriaceae are among the most clinically significant MDR bacteria.Such a rise drove an increase in the usage of polymyxins because of limited effective alternative antibiotic treatment.So, this global concern of polymyxins and the mounting use of colistin explicates the addition of acquired colistin resistance to the resistant traits in Enterobacteriaceae [3].
We aimed to assess the diagnostic efficacy of the rapid polymyxin NP test using Polymyxin B for the diagnosis of colistin-and polymyxin resistant Enterobacteriaceae and determine the frequency of mcr-1 and mcr-2 genes among the colistinresistant Enterobacteriaceae at ASUH.
In the present study, the age of the patients ranged from 9 days to 77 years.The number of female and male patients included were equal (21/42, 50%).Concerning the sample type, it was observed that majority of isolates were retrieved from blood cultures (12/42, 28.6%).
In a similar pattern to our study, research conducted in Croatia reported that their patients harboring colistin-resistant isolates were equally distributed between both sexes (50% each), and their ages ranged from 6 to 84 years.The predominant specimen were blood cultures (9/24), followed by urine samples (5/24) [21].
As for the antibiotic susceptibility testing, all the tested isolates recorded a 100% resistance rate against almost all β-lactams with exception of ceftazidime (97.6%) and imipenem (90.0%).These findings came in concordance with the internationally high level of antimicrobial resistance among all classes of antimicrobial, especially in Egypt which caused surging in colistin use as the last resort antimicrobial.
In Egypt, a one-year retrospective study showed that Gram-negative isolates displayed a high rate of resistance to all used antibiotic classes.This study reported that resistance towards the Betalactam group ranged from 70% to more than 98% against different antibiotic members of this group [22].Such a high resistance was demonstrated by other researchers as well.In Croatia, the sixteen enterobacterial isolates were resistant to almost all antibiotics; β-lactams, ciprofloxacin, and gentamicin.Except for colistin, the MIC90 of all antibiotics exceeded 128 mg/L [21].
Regarding antimicrobial consumption among the 42 patients included in our study, we found that 30 (71.4%) patients received antibiotics.Meropenem rated the first where 10/30 (33.3%) of our patients received it.
Other researchers reported a high overall antibiotic consumption among different hospitals.They found that Cephalosporins were the most usually prescribed antibiotics, followed by quinolones, penicillins, and carbapenems.They reported high significant pearson's correlation coefficient factors (r) between carbapenem consumption and its resistance rate of 0.271 and 0.427 for E. coli and Klebsiella pneumoniae respectively [23].
Also, other investigators demonstrated that the rate of antibiotic usage among their patients was very high ranging from 32.9% to 91.7% and 59% of them were receiving one or more antibiotics.Third-generation cephalosporins ranked first among the prescribed antibiotics and accounted for 28.7%.Penicillins with beta-lactamase inhibitors and metronidazole derivatives accounted for 19.7% and 15.2% respectively [24].
Unlike our study, other researches in China and Egypt proved that cephalosporins are the most received drug [23,24].Whereas in the current study, Meropenem was the most frequently used one.This backs to the high level of multidrug resistance that was detected in the last years and obliged the physicians to prescribe carbapenem as the last option before resorting to polymyxins.
Similarly, a study conducted in Thailand reported that colistin resistance was identified in 13 (35%) E. coli and 213 (76%) Klebsiella pneumonia isolates, with MIC ranging from 0.5 to 32 mg/L and 0.25 to >128 mg/L respectively [25].
For polymyxin MIC done for our tested isolates, the results ranged between 0.5 to 16 ug/ml.Fifteen isolates out of 42 were resistant to polymyxin (35.7%).The majority of the resistant isolates were Klebsiella isolates (86.7%) and E. coli was (13.3%).
In light of results stated worldwide, we found that they agreed to a great extent with those reported in our study and this affirms the fact of increasing resistance against colistin and polymyxin B which are considered the last hope for treatment of MDR organisms, this foreshadows an upcoming worldwide disaster.
Our results are in concordance with those reported by many authors where their studies showed excellent sensitivity and specificity [14,[29][30][31][32].Other authors from Greece found that the Rapid Polymyxin NP test yielded a high sensitivity (99%) and a lower specificity of 82%.Although their results were different from ours, yet, these findings still point out that the Rapid Polymyxin NP test can be used as a screening test for early diagnosis of colistin-resistant isolates [33].Thus, this test can represent a reliable alternative to BMD, especially in resource-limited settings.
Although the sensitivity of the test obtained in our study was lower for colistin MIC compared with the polymyxin MIC, many worldwide studies stated that the Rapid Polymyxins NP test exhibited an outstanding performance for detecting resistant and susceptible Enterobacteriaceae for both polymyxin B and colistin.We proved in our study that this test is easy to perform and could be accomplished within 4 hours compared to the MIC by the BMD method, which consumes around 24 hours.
A Brazilian study asserted this fact and reported that the required time to obtain the result was 2 hours for the Rapid Polymyxins NP test versus 24 hours for the BMD.In addition, this test is more advantageous, being less burdensome.This fast identification of polymyxin-resistant bacteria might aid in precisely identifying the optimum therapy choices, and rapidly implementing contact precaution measures, thus halting further outbreaks with MDR-isolates [29].
In our study, we carried out conventional PCR for confirmation of the presence of polymyxins resistant genes (mcr-1, 2 genes).In correlation with Colistin MIC, we found that 17/18 colistinresistant isolates had the mcr-1 gene (94.4%) and five (27.8%) isolates harbored the mcr-2 gene.Whilst, correlating it with polymyxin MIC revealed that all the 15 (100%) polymyxinresistant isolates were positive for the mcr-1 gene and 5/15 (33.3%) positive for the mcr-2 gene.We observed that the mcr-1 gene was found in all of the 14 Rapid Polymyxins NP positive isolates, while, the mcr-2 gene was detected solely in five isolates.
Although in our study mcr-1 and mcr-2 genes were present in most of our isolates, this represented a hindrance to finding a statistical relationship between the types of genes and the results of the rapid polymyxin NP test.
Research conducted in Switzerland found that the Rapid Polymyxins NP displayed 100% sensitivity when performed using polymyxinresistant isolates of environmental, animal, and human origin and mediated by the plasmid carrying genes mcr-1 or mcr-2 [16].
On the other hand, investigators from Brazil reported that Polymyxins NP gave false-positive results in two polymyxins-susceptible isolates of their 19 selected isolates (11 polymyxinsresistant and 8 polymyxins-susceptible) carrying the mcr-1 gene [29].
A study in Thailand detected the mcr-1 gene in 11/37 E. coli isolates and their colistin MIC range was 4-32 mg/L, while, they found the mcr-1 gene in 4/280 Klebsiella pneumoniae isolates with a colistin MIC range of 4-64 mg/L [34].
The world has been suffering for a long time from the era that we fear the most; the era of antimicrobial resistance, this dreadful social challenge.Nowadays, the globe is attempting to find a way out of this trouble.But, since we are unable to add new therapeutic agents to our inventory, so, the only possible solution we have is to try decreasing the rate of colonization and infections with resistant strains, thereby reducing mortality and morbidity.To do so, we need to apply preemptive measures before we are left handcuffed without any antimicrobial choices.
In the current study, the rapid Polymyxin NP test proved to be a reliable, easy, affordable, and speedy assay to perform.Moreover, it exhibited promising performance compared to our reference technique.Hence, we can deduce that this test can be implemented as a screening tool in the outbreak management and active surveillance for the presence of Polymyxins resistance.Therefore, combined with PCR, it can play a substantial role as a part of infection control measures to contain antimicrobial resistance and prevent the transfer of resistance genes among patients.

CONCLUSION
The rapid polymyxin NP test is an easy, quick, sensitive, and specific test that can reliably detect both polymyxin-resistant and susceptible Enterobacteriaceae isolates, whatever the molecular mechanism of resistance.Even more, it is less sophisticated and can provide results at least 16 hours before the BMD method .
The performance of the polymyxin NP test in the current study was lower for colistin MIC in comparison with the polymyxin B MIC .Therefore, we recommend using the polymyxin NP test for early detection of polymyxin-resistant Enterobacteriaceae, especially in low resource health care settings.Further studies on broad scales with large sample sizes are recommended to study the effect of each mcr gene separately on the sensitivity and specificity of this test.

Figure ( 1 )
Figure (1):Polymyxin NP test in a Microtitre plate show first and 3 rd row contain polymyxin free solution, the 2 nd row contains polymyxin NP solution.first well in the first and 2 nd row were inoculated with saline, the 2 nd well in the first and 2 nd rows were inoculated by positive control strain, first and 2 nd rows were inoculated with saline, the 3 rd wells in the first and 2 nd rows were inoculated by negative control strain, the 4th wells in the first and 2 nd rows were inoculated by tested isolate and so on.color change from red to yellow or red to orange indicate resistant bacterial growth and positive polymyxin NP test.

Figure ( 2 )
Figure (2): mcr-1 PCR results on gel electrophoresis.DNA ladder in lane 1, positive control for mcr-1 gene at 320 bp in lane 2, positive isolate in lane 3, negative control in lane 4 and negative isolates in lanes 5 and 6.

Table ( 2
): Antimicrobial susceptibility testing by Disc Diffusion Method and Vitek2 C for 42 studied isolates in the current study.

Table ( 3
): Colistin MIC results among the 42 studied isolates in the current study.

Table ( 4
): Polymyxin B MIC results among the 42 studied isolates in the current study.

Table ( 5
): Results of Polymyxin NP test for the 42 isolates included in the current study.

Table (
6): Distribution of mcr-1 gene among 42 tested isolates in the current study.

Table ( 7
): Distribution of mcr-2 gene among different tested isolates in the current study.

Table (
8): Correlation between colistin MIC versus polymyxin B MIC, polymyxin NP test and PCR for mcr-1, 2 genes among 42 tested isolates in the current study

Table ( 10 ) :
Diagnostic performance of Polymyxin NP test compared to the reference method MIC for colistin among 42 tested isolates in the current study.

positive, TN: True negative, FP: False positive, FN: False negative, PPV: Positive predictive value, NPV: Negative predictive value. Table (11):
Diagnostic performance of Polymyxin NP test compared to the reference method MIC for polymyxin B among 42 tested isolates in the current study.

:
CPE: carbapenemaseproducing Enterobacteriaceae, MDR: multidrugresistant bacteria, BMD: broth microdilution, MIC: minimum inhibitory concentration, PCR: Polymerase chain reaction, E.coli: Escherichia coli, All study procedures were as per ethical guidelines of the 1975 Declaration of Helsinki and approved by the ethical committee of Faculty of Medicine, Ain Shams University.(Ethical approval number: FWA 000017585).The rapid polymyxin NP test exhibited high specificity and sensitivity for diagnosing polymyxin resistance.42.9% of the isolates were resistant to colistin, and 35.7% were resistant to polymyxin.The mcr-1 gene was more prevalent (40.5%) than the mcr-2 gene (11.9%). In conclusion, the rapid polymyxin NP test proved to be an easy and quick technique that can reliably detect both polymyxin-resistant and susceptible Enterobacteriaceae isolates.This test can be used for diagnostic purposes in laboratories with limited resources or infection control.