Abstract
Introduction Polymyxins, the cationic lipopeptide antibiotics, are the last line of therapeutics against the multi-drug resistant (MDR) pathogenic Gram-negative bacteria. Unfortunately, the rising cases of polymyxin-resistant strains from across the globe have adversely impacted the available treatment options. While the molecular mechanisms responsible for developing polymyxin resistance (PolR) are largely understood, the prevalence of PolR strains in India is not investigated systematically.
Method For this study, a systematic search for relevant articles was performed in the applicable databases for the period January 2015 to December 2023. A meta-analysis was performed using the comprehensive meta-analysis software (V.4) to determine the overall prevalence of PolR bacterial strains in India. Publication biases were assessed using funnel plots and Egger’s regression analysis.
Result Considering a total of 43 studies selected based on the relevant inclusion and exclusion criteria, the present meta-analysis found the rate of PolR bacteria in India to be at 15.7%, which is higher than the global average. Among the Indian States, Tamil Nadu topped with the highest prevalence of PolR at 32.4%. Investigating the contribution of the mobile colistin resistance (mcr) genes, it was observed that among the PolR strains, 8.6 % were mcr positive.
Conclusion The mcr-positive strains reported from India could be an underestimation of the actual numbers due to the non-inclusion of mcr screening in many previous studies. Our study provides insight into the state of the PolR situation in India, which may be useful to develop a monitoring strategy to contain the spread of such strains and preserve the potency of the polymyxins.
INTRODUCTION
The significant use of antibiotics in healthcare, and agriculture, and their indiscriminate prophylactic consumption has driven the rise of antibiotic-resistant bacterial pathogens. Infections caused by antibiotic-resistant bacteria are increasingly becoming fatal due to the non-availability of potent antibiotics against them. As per an estimate, drug-resistant pathogens caused 4.95 million deaths in 2019 1, predicted to rise to 10 million deaths annually by 2050 2. Considering the significant economic consequences of rising drug-resistant infections, urgent mitigation measures are warranted.
Antibiotic-resistant bacteria belonging to the Gram-negative type pose a significant challenge to treatment as increasing reports of resistance to third and fourth generations of cephalosporins are reported globally 3. In this unprecedented situation, polymyxins, a group of cationic peptide antibiotics, remain the last option for treatment 4,5. Polymyxin was initially isolated in 1947 from a soil bacterium Paenibacillus polymyxa 6 and approved for clinical use in 1959 against Gram-negative bacteria. Due to their nephrotoxic and neurotoxic properties, polymyxins were prohibited for clinical use in the 1970s 7,8 and were consequently substituted by aminoglycosides, quinolones, and β-lactams. However, with the rising infections caused by multi-drug resistant (MDR) Gram-negative bacteria (GNB) such as Klebsiella pneumoniae, Acinetobacter baumannii, and Pseudomonas aeruginosa, and failure of the frontline antibiotics, polymyxins have re-emerged albeit with some modifications in the chemical structure as the last-resort antibiotic option against GNB 4,9.
Polymyxins are of five types (polymyxin A to polymyxin E), of which colistin (polymyxin E) and polymyxin B are approved for clinical use 10. Structurally, polymyxins are cationic decapeptides consisting of a central cyclic heptapeptide with a tripeptide side-chain acylated at the N-terminus by a fatty acid tail 11,12. The primary target of cationic polymyxin is the negatively charged outer membrane (OM) of GNB. Polymyxin destabilizes the OM lipopolysaccharide (LPS) by displacing the divalent cations Mg2+ and Ca2+ resulting in enhanced permeability of the membrane and leading to leakage of cytoplasmic content that ultimately causes cell death 5,13,14.
The extensive use of polymyxins in the preceding decade in both healthcare and animal feeds has triggered the emergence of polymyxin-resistant (PolR) bacterial strains across the globe. PolR bacterial species such as K. pneumoniae 15–17, P. aeruginosa 4,18, A. baumannii 19, E. coli 20, Shigella sonnei 21, Salmonella enterica 22, Enterobacter aerogenes and Enterobacter cloacae 23 have been reported from different parts of the world that have adversely affected the healthcare sectors. Moreover, the detection of PolR GNB in both clinical and environmental samples has made the situation alarming.
The primary mechanism responsible for PolR development is the remodelling of the OM in GNB making it increasingly positive [by the addition of 4-amino-4-deoxy L-arabinose (L-Ara4N) or phosphoethanolamine (pEtN)], leading to the repulsion of cationic polymyxins 4,13. The OM remodelling is generally carried out by bacterial two-component signal transduction systems (TCS). TCS generally sense the environmental cues and signal appropriately to modulate the expression of genes involved in the response mechanisms. PhoPQ and PmrAB are the two primary TCSs in the GNB involved in the OM remodelling in response to various stresses including polymyxin treatment 5,13. Apart from PhoPQ and PmrAB TCSs, there are several other such systems reported in various bacterial species 14,24,25. Though organisms can develop resistance to polymyxins via TCS, there are many GNBs like Serratia marcescens, Proteus mirabilis, Proteus penneri, Proteus vulgaris, Providencia rettgeri, Providencia stuartii, Vibrio cholerae, Campylobacter spp., Legionella spp., Morganella morganii, Burkholderia spp., Edwardsiella spp., Aeromonas jandaei, Aeromonas hydrophila, Acinetobacter junii, Neisseria spp. that are naturally resistant to polymyxins due to the constitutive expression of the arnBCADTEF operon and/or the eptB gene, which increase the charge of LPS by addition of pEtN and/or L-Ara4N 4,26. Moreover, all Gram-positive and anaerobic bacteria are also inherently resistant to polymyxins.
Though chromosome-encoded polymyxin resistance is the prevalent mechanism, the detection of the plasmid-borne mobile colistin resistance (mcr) gene from the bacteria E. coli and K. pneumoniae in China has made the situation alarming 20. The mcr gene encodes a pEtN transferase enzyme that catalyses the addition of a pEtN residue to lipid-A moiety which alters the net charge of the LPS reducing the binding of polymyxin. To date, ninety-eight mcr alleles have been discovered belonging to ten different classes of mcr genes (mcr-1 to mcr-10) among many GNB pathogens. Bacterial plasmids bearing the mcr genes have been reported from across the globe. Moreover, the different mcr classes have shown distinct geographical distribution 27,28. As the potential of rapid dissemination of plasmid-borne mcr genes is significantly high, it is pertinent to conduct periodic surveillance to monitor their distribution in the environment and clinic.
In the Indian context, mcr genes have been reported in several studies conducted using clinical samples 29, hospital sewage 30, urban sewage water 31, and food samples 32. However, a comprehensive analysis of the distribution of mcr genes across India is lacking. It is important to highlight that the increasing number of carbapenem-resistant pathogenic bacteria has made India one of the top users of polymyxins in the clinic. Furthermore, the use of colistin as a growth promoter in poultry farms has triggered the spread of PolR GNB in the environment. Though the Govt. of India has banned the use of colistin in farm animals since 2019, the impact of their extensive use before 2019 is yet to be assessed. Very few investigations have been carried out in the Indian context, especially using environmental samples to understand the dynamics of PolR bacterial strains in the environment and the incidence of mcr-positive bacterial strains from across India.
The current meta-analysis was performed to determine the pattern and spread of PolR bacterial strains in India and to find a correlation between PolR and the presence of the mcr genes.
MATERIALS & METHODS
Using published articles, the present study was conducted to determine the prevalence and spread of PolR Gram-negative bacteria (GNB) from clinical and environmental sources in India and the contribution of mcr genes in gaining the resistance phenotype. The study was undertaken as per the guidelines of Systematic Reviews and Meta-analyses (PRISMA) 33 (Fig. 1).
Data collection
Databases like PubMed, Google Scholar, Science Direct, and Scopus were extensively searched using the keywords “polymyxin resistance”, “colistin resistance”, “mobile colistin resistance gene” and “India”. Original research articles with reports of PolR bacteria from India, published between January 2015 to December 2023 were included in this study. The last search was performed on January 06, 2024.
Inclusion criteria
For the meta-analysis, we only included articles that met the following requirements: 1) the samples used in the studies must hail from India, 2) data on resistance to polymyxin B and/or colistin must be available, 3) broth-microdilution (BMD) method must have been followed to determine PolR of the bacterial isolates, 4) the article must mention the total sample size and the number of PolR strains isolated or the prevalence frequency, and 5) the article must have been published in English.
Exclusion criteria
The following exclusion criteria were strictly followed: 1) studies with no PolR isolate were excluded from this analysis; 2) studies conducted outside India were not considered; 3) studies that employed resistance detection methods other than BMD were disregarded; 4) studies that were not published in English, were not peer-reviewed, and were published before 2015 were excluded.
Data extraction
Data were extracted from relevant articles that were qualified based on the study’s inclusion and exclusion criteria (Fig. 1). Each report was individually screened and investigated for eligibility by two different authors (SKD and IP). Following a thorough review, the following information was extracted and summarised for each study: first author name, publication year, State, or UT (Union Territories) of occurrence, total sample size, the total number of PolR isolates, reported PolR bacteria, resistance detection method, genes involved in resistance development, and mcr gene screening status. For further analysis, the data were tabulated in a Microsoft Excel sheet.
Subgrouping of data
The research articles with PolR reports were classified into two groups based on criteria such as sample size (more than one) and the involvement of mcr genes in the development of PolR. Studies reporting a specific number of samples containing both PolR and susceptible isolates and a particular number of PolR bacteria as events were placed in the first group. This group was further subdivided based on the States or UTs from where they were reported. In this classification system, those States were included that have at least four different reports on PolR isolates. In the second group, those reports were included where mcr gene screening status was mentioned clearly in the study.
Quality assessment
The quality of all eligible reports was determined by the Newcastle–Ottawa Scale (NOS) and only the high-quality reports were considered for the present meta-analysis. According to the NOS guidelines, qualitative scores are given as stars, and a total score of five stars or more for a study is considered a high-quality report.
Meta-analysis
To measure relative risk, we included articles reporting the prevalence of PolR bacteria and the contribution of mcr genes in this meta-analysis. All statistical analyses were performed using Comprehensive Meta-Analysis (CMA) software (version 4.0; Biostat Inc. USA). The generated data were visualized using forest plots. Publication bias in the included studies was assessed using funnel plots with visual analysis 34 and Egger regression analysis 35. The heterogeneity of the studies was examined using the Cochrane Q statistics 36 and I2 statistics37. Based on the outcome of the heterogeneity analysis, a fixed-effect model (homogenous) or random-effect model (heterogeneous) was used for the meta-analysis.
Overall, 43 articles reporting the prevalence of PolR isolates in India were subjected to meta-analysis, whereas 58 reports were included to decipher the role of mcr gene in the polymyxin resistance phenotype. Of the 58 reports of mcr contribution to PolR, 41 were eligible for the meta-analysis. The combined event rate and 95% confidence interval were calculated to determine the frequency of PolR isolates in India and the contribution of mcr to polymyxin resistance. In addition, sensitivity analysis was performed to explore the robustness of the meta-analysis by excluding one study at a time, performing the meta-analysis, and comparing it with the results of the parent analysis.
RESULT
Literature review and screened results
Overall, 1408 articles were obtained by searching different databases with suitable keywords (PubMed = 802, Google Scholar = 431, Science Direct = 109, Scopus = 66). The article selection approach is presented in a PRISMA flow chart (Fig. 1). A total of 43 eligible articles reporting PolR (Supplementary Table 1) and 41 articles reporting the contribution of mcr (Supplementary Table 2) were included in the meta-analysis after careful screening according to the stipulated inclusion and exclusion criteria.
Publication bias
Visual analysis of the funnel plot and Egger’s regression analysis showed the absence of publication bias in the analysis of the incidence of PolR bacteria in India (intercept: 0.63, 95% CI = -2.38 to 3.65, p = 0.67), as well as in the analysis of the role of the mcr gene in PolR (intercept: -1.3, 95% CI = -2.69 to 0.08, p = 0.065) (Table 1) (Supplementary Fig. 1).
Furthermore, after the categorization of all studies according to the states from which they were reported, the funnel plot and Egger’s regression analysis revealed no significant biases while analysing the incidence of PolR bacteria in states such as Tamil Nadu, Uttar Pradesh, Odisha, West Bengal, and Delhi (Table 1) (Supplementary Fig. 2).
Heterogeneity analysis
Cochrane Q and I2 measurements demonstrated significant heterogeneity among the included studies. Hence, a random-effects model was used for all meta-analyses (Table 1).
Prevalence of PolR bacteria in India
After investigating 43 reports, the present meta-analysis found that the rate of PolR bacteria in India was 15.7% (Fig. 2). Out of 28 states and eight UTs in India, 12 states and 3 UTs have reported the presence of PolR bacterial strains. Among all the states that have reported PolR, Tamil Nadu, Uttar Pradesh, and Odisha are the top three, where the rate of PolR is higher than the Indian average at 32.4%, 16.3%, and 15.9%, respectively. Upon State-wise analysis, Delhi and West Bengal showed significant cases of PolR at 13.5%, and 10.9%, respectively, though less than the national average (Fig. 3).
Role of mcr gene in the development of PolR
Furthermore, to understand the prevalence of mcr genes in PolR bacterial isolates in comparison with other possible mechanisms in India, 41 reports with positive mcr gene screening status and a sample size of more than one (PolR isolate) criterion were considered for the analysis (Supplementary Table 2). The meta-analysis results showed that the contribution of the mcr gene to the rise in polymyxin resistance was 8.6% in India (Fig. 4).
Sensitivity analysis
Sensitivity analysis was performed to measure the robustness of the meta-analysis by excluding a single study each time and relating the results to the parental meta-analysis. As illustrated in Supplementary Fig. 3, no significant deviation was observed in the sensitivity analysis compared with the original meta-analysis results for studying the incidence of PolR bacteria in India. Furthermore, a meta-analysis on the contribution of the mcr gene to PolR also revealed a robust observation, as shown in Supplementary Fig. 4.
DISCUSSION
The current meta-analysis systematically evaluated the incidence of PolR bacteria in India, which was higher than the global average. In this study, we found that 15.7% of GNB were PolR, whereas the global average remained at 10% 38. A previous study has put the rate of PolR among clinical Enterobacteriaceae isolates in India at 13.8% 39.
The present study revealed that GNB resistant to the last line of therapeutics is spread all over India, and States like Tamil Nadu, Uttar Pradesh, and Odisha have a higher incidence of PolR strains than the national average. The rising incidence of polymyxin resistance in India could be attributed to their indiscriminate use in the healthcare sector, in poultry farms as growth promoters, and the rapid dissemination of mcr genes via horizontal gene transfer. Even though India has banned the use of colistin in animal feeds since 2019, its implications on the spread of PolR may be revealed a few years down the line. Interestingly, there are no reports of PolR in 16 states and 5 UTs in India, which may be due to the lack of surveillance in those States and UTs. The overall spread of PolR GNB in India is shown in Fig. 5A.
Although most of the reports used in this study indicated PolR rates under 20%, there are a few that have reported significantly higher PolR rates 40–45. These discrepancies are due to the different methodologies adopted for determining polymyxin resistance. In India, various methods such as the Kirby-Bauer disk diffusion test, broth microdilution (BMD), agar dilution, E-test, VITEK-2, gradient diffusion, colistin broth disc elution, rapid polymyxin Nordmann Poirel Test, Mikrolatest kit, BD Phoenix M50, MicroScan WalkAway 96 Plus, Micronaut-S and Modified Stokes DDT, have been used. However, as per the recommendation of the CLSI and EUCAST, BMD is the only approved method for determining colistin susceptibility 46,47, and therefore in this study, the colistin resistance reports determined using the BMD method have only been included.
Meta-analysis revealed that in 8.6% of cases, the mcr gene was responsible for spreading the PolR phenotype in India. In India, after the first detection of the mcr gene in a clinical E. coli isolate 48, several reports of mcr occurrence have been published 30,32,49–51. To date, six classes of mcr have been reported in India: mcr-1, mcr-2, mcr-3, mcr-4, mcr-5, and mcr-9. Of the 115 mcr variants detected till date, mcr-1 was the predominant variant responsible for driving polymyxin resistance in India. In the current study, the frequency of mcr-1 was 72%, whereas the frequencies of mcr-2, mcr-3, mcr-4, mcr-5, and mcr-9 were 3%, 19%, 3%, 1%, and 2%, respectively (Fig. 5C). Considering that mcr-1 was the first plasmid-borne colistin resistance gene to be detected 20 it has spread globally in the last 6-7 years.
Unlike the distribution of PolR GNB, the spread of mcr has been limited to nine states and two UTs in India (Fig. 5B). Among these states, Uttar Pradesh was found to be the hotspot for the spread of mcr variants, with 48 reports, followed by Tamil Nadu and Assam with 31 and 15 mcr variant reports, respectively. However, several reports on PolR have not provided sufficient information on mcr gene screening status. The limited screening for mcr genes among these studies may be responsible for the low rate of detection and reporting of mcr genes in India.
As mentioned previously, mcr genes are largely plasmid-borne, and various types of plasmids are involved in carrying mcr genes in GNB 27,52. In India, different mcr variants were found to be linked with plasmids including IncHI2, InCHI2A, IncX1, IncX4, and FIA Inc 32,41,53. The association of these plasmids with other antibiotic-resistance genes conferring resistance to carbapenems, cephalosporins, and fosfomycin has also been previously established 54,55. Although plasmids are the primary genetic vehicle through which mcr genes disseminate among bacterial species, their chromosomal integration has also been reported in different bacterial species in India 56–59. The major advantage of chromosomal integration of mcr genes could be their reduced chance of loss from the genome, in the absence of polymyxin selection. Interestingly, some naturally resistant bacteria also harbour the mcr genes and can potentially spread them to the polymyxin-susceptible species in the environment 60. Moreover, few studies have reported the presence of the mcr-1, mcr-3, and mcr-9 genes in bacterial species that are susceptible to polymyxin 51,61. This is primarily due to the insertional inactivation of mcr-1 and mcr-3 genes 51,61. However, in the case of mcr-9 positive Enterobacter hormaechei strain, qseBC TCS genes were absent downstream 62 which has been previously described to induce the colistin resistance phenotype by upregulating the mcr-9 gene expression in response to the subinhibitory concentration of colistin 63.
This meta-analysis has certain limitations. First, the current study included articles available in the English language only, as a result, publications in other languages have been excluded. Second, we searched reports from databases including PubMed, Google Scholar, Scopus, and Science Direct, so may have missed publications indexed in other databases. Third, as certain Gram-negative bacterial species are inherently resistant to polymyxins, reports on polymyxin resistance from the environmental strains may have been biased toward such species.
CONCLUSION
The present study found that the rate of PolR bacterial strains in India was 15.7%, which is higher than the global average. While PolR strains have been reported from most of the Indian States and UTs, there are still regions where adequate data are not available, warranting monitoring and surveillance. This study further found that in 8.6% of PolR strains, the phenotype was determined by mcr genes carried on plasmids. This may be grossly underestimated due to the lack of mcr surveillance data in many published articles. As polymyxins are one of the last option antibiotics available against Gram-negative bacterial pathogens, preservation of their antibiotic properties is paramount. However, the increasing incidence of polymyxin resistance in clinics and the environment has the potential to limit antibiotic treatment options. Therefore, along with developing better versions of polymyxins and/or their derivatives, it is also important to devise strategies for monitoring the frequency of PolR bacterial species and, the prevalence of mcr genes in them. Sustained measures need to be in place to contain the further spread of the polymyxin resistance phenotype in the near future.
Data Availability
All data produced in the present study are available upon reasonable request to the authors.
Author Contributions
Conceptualization: SKD, AKP, and SSM; Data extraction and analysis: SKD, IP, AKP, and SSM; Data validation: SKD and IP; Manuscript writing and reviewing: SKD, AKP, and SSM. All authors agreed to the submitted version.
Funding
This work is supported by the funds received from the Science and Technology Department, Govt. of Odisha (Grant no. ST-BT-MISC-0005-2023-2463/ST, dt. 23-05-2023). Infrastructure support from the “Centre of Excellence on Bioprospecting of Ethno-pharmaceuticals of Southern Odisha (CoE-BESO)” to the Dept. of Biotechnology, Berhampur University is gratefully acknowledged. Indira Padhy is a recipient of the “Biju Patnaik Research Fellowship (BPRF)” from the Science and Technology Department, Govt. of Odisha.
Conflict of interest
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Footnotes
The manuscript has been revised to update the inclusion/exclusion criteria used for performing the meta-analysis. Specifically, the use of broth microdilution method to determine polymyxin resistance has been used as the inclusion criterion to select studies for meta-analysis. Therefore, with the revision, the figures 1, 2,3, and 5 have been updated. Moreover, the supplementary file is also updated.