In the present report we have investigated the β-lactam resistance pattern amongst Enterobacteriaceae in gut flora of neonates (1–60 days) by enrolling babies using various selection criteria so as to avoid any possible source of antibiotic selection pressure. Acquisition of resistance through food and water was also ruled out as neonates were exclusively breast fed. Compliance was ensured through household follow up by trained field workers upto D60 of life.
The present study shows that majority of the babies were colonized by D1. With the acquisition of mother’s flora the babies are equally likely to get the antibiotic resistance strains. Our data revealed that overall there was nearly 87% (232/267) resistance to the ampicillin by D60 in Enterobacteriaceae. The overall rate of ESBL was 20.6% which may be just a glimpse of bigger picture as in the present study only dominant population was studied. Selective media were not used for screening ESBL gut carriage which would reflect the true representation of ESBL carriage in the community. The low isolation of ESBL producers on D1 may be due to the short duration of exposure to the maternal flora during delivery (Table 2). Various factors could have contributed to the increase in the resistance by day 60. After delivery, exposure related to mothers environment, oral and skin flora provide the major sources of bacteria which may transfer to the neonates by several ways including suckling, kissing and caressing. In addition, breast milk is also a source of bacteria, which contains up to 109microbes/L in healthy mothers . Other sources may be household contact with siblings, pets , as well as horizontal transfer of gene within the commensal flora . In our study acquisition of resistance via supplementary food has been ruled out as babies were completely breast fed. Several studies have shown the prevalence of antibiotic resistance in absence of direct use of antibiotic. Presence of tetracycline resistance bacteria in breastfed infants  and commensal ESBL producers in pre-school healthy children  suggest contamination in the family environment rather than direct exposure to antibiotic. The limitation of our study is that we have not studied the environmental flora and compared it with that of neonatal gut flora.
Besides ESBL, AmpC producing Enterobacteriaceae were also isolated. AmpC producing isolates were approximately 20% and co-production with ESBL was seen in 11.2% throughout the study period (Table 2). AmpC β-lactamases producers are of major concern as they are resistant to β-lactam and β-lactam inhibitor combination as well as cefoxitin which further narrows down the treatment options. As carbapenems are drug of choice for ESBL and or AmpC producing bacteria, coexistence of these enzymes can pose a threat to the community acquired pathogens as MIC of such strains are 10 fold higher for various carbapenems .
The ampC gene showed diverse profile, in contrast CTX-M-15 was predominant ESBL gene in gut flora. Previous studies from India have also shown CTX-M-15 as predominant ESBL from clinical isolate . Approximately, 50% of neonates admitted to neonatal unit in our hospital with early onset sepsis had ESBL producing Enterobacteriaceae which is strongly supported by early colonization with ESBL producing Enterobacteriaceae in the neonates in the present study.
Recent report of isolation of CRE (NDM-1) from environmental samples  and community acquired infections  indicate that CRE producing NDM-1 enzyme may be widely distributed in India. However, there is paucity of data regarding fecal carriage of CRE in the community in absence of antibiotic pressure. Different studies have used different culture based techniques like MacConkey agar plates supplemented with 1 μg/ml imipenem, Chrom Agar KPC, Mac Conkey Agar with imipenem, meropenem and ertapenem disc (10 μg) and two step selective broth enrichment method using 10 μg carbapenem disc to evaluate gut colonization with CRE with good performance . Most of these techniques are validated for KPC detection in organisms with MIC range 0.5 - >32 μg/ml for various carbapenems . Nordmann et.al. screened 27 NDM-1 positive isolates and reported that the MIC of these isolates vary from 0.5 - >32 μg/ml, 1.5 - 231 >32 μg/ml and 1.5 - >32 μg/ml for ertapenem, meropenem and imipenem respectively. However, only one isolate i.e. P Providencia rettgeri A showed MIC of 0.5 μg/ml for ertapenem . In present study with 2 step broth enrichment method using meropenem disc only one strain of Enterobacter sp was positive by MHT and PCR confirmed presence of kpc-2 gene. MIC of other 28 suspected CRE isolates were ≤ 0.5 μg/ml for all carbapenems. Two isolates were positive for ESBL and AmpC, having MIC of 0.5 μg/ml for ertapenem but were negative for carbapenem genes.
In the present study widespread resistance to Ampicillin and 3rd generation cephalosporin (3GC) was observed but carbapenem resistance was rare. This can be explained by indiscriminate use of 3GC in human and animals due to availability of oral formulations and over the counter unrestricted access. Ampicillin and 3GC are used as an empirical therapy in India for the management of neonatal sepsis and other heath related complications like UTI, meningitis, bacterial sepsis (6, 1). The high prevalence of resistance to these drugs as indicated in our study raises the question regarding the efficacy of these antibiotics as an empirical therapy.
Carbapenems on the other hand are used sparingly as they are available as parentral formulation for which a patient have to visit the health care facility and in addition there is no reports of their use in animals from India. It is noteworthy that the presence of kpc-2 gene in antibiotic naive neonates may be an alarming finding as carbapenem resistance genes are on plasmids and have a potential for rapid dissemination in future. Commensal flora can colonize the human gut without causing any symptoms, but most of the infections are endogenous and come from patient’s own gut flora .
The present study estimate of β-lactam resistance may be biased due to following reasons. Babies were supplemented with probiotics which have beneficial effect on gut by producing organic acids, bacteriocins, peptides and in turn decreasing pH of gut leading to inhibition of colonization of Enterobacteriaceae. In addition, only the subdominant population was screened for ESBL carriage resulting in an under estimate of ESBL in the community. However, this data could not be an over-estimate as there are no reports of presence of ESBL genes in probiotic bacteria or transfer of antibiotic resistant genes from gram positive (Probiotic) bacteria to gram negative bacteria.