Multi-ESBL carriage among K. pneumoniae and E. coli from Neonatal units in Mwanza, Tanzania

Please read the article at:

https://doi.org/10.3390/antibiotics10050476

Disclaimer: Below is an interpretation of the article thus do not necessarily represent the views of the authors.

Introduction

ESBLs are enzymes which hydrolyse 3rd Generation Cephalosporins such as cefotaxime, and aztreonam but sensitive to clavulanic acid. These are encoded for by genes such as bla-CTX-M and certain variants of bla-TEM and bla-SHV. However, other mutations can result in comparable phenotypic resistance. Thus, "ESBL bacteria" is occasionally used to describe both genotypic carriers of ESBL genes and those phenotypically similar (p-ESBL).

Methods

Silago et. al. investigated 75 p-ESBL E. coli (n = 20) and K. pneumoniae (n = 55) from January 2020 at CUHAS in Mwanza. These were isolated primarily from neonate faecal and blood samples, however the collection included a few cot swabs, and mother and healthcare-worker hand swabs. Species were determined via biochemical tests and ESBL gene carriage was confirmed via multiplex PCR against CTX-M, TEM and SHV primers.

Phenotypic testing was done through three variations of disk diffusion assays. All isolates were p-ESBL on all tests except for two which failed the modified double disk synergy assay. However, only 93.3% were PCR positive for the genes tested.

Results

In Tanzania, CTX-M variants are most commonly reported ESBL genes which was echoed in the dataset with 98.6% of ESBL isolates positive for CTX-M. This was followed by TEM (85.7%) then SHV (71.4%). Co-carriage of CTX-M, TEM and SHV was also high at 62.9%.

All triple carriers were resistant to 25µg trimethoprim-sulfamethoxazole though were less commonly resistant to 30µg gentamicin (79.5%) and ciprofloxacin (34.1%). Four isolates were resistant to 10µg meropenem which were also triple carriers. Based on phenotypic resistances and gene presence, the authors identified pairs of isolates which may be closely related, or acquired the same conjugative plasmids.

They concluded that CTX-M, TEM and SHV are the most common ESBL genes within their dataset and the similar phenotypes and ESBL gene carriage may indicate the spread of conjugative plasmids carrying them. Five isolates were negative for the three genes, though they may contain other ESBL genes not included in the multiplex. They also hypothesised that the two isolates which failed the modified double disk synergy assay may produce other beta-lactamases which can lower synergism-based assays' sensitivity. Finally, they believe their pairings suggest potential gut to blood translocation, and pathogen shedding from healthcare workers. Thus, improved sanitation practices were recommended.

Thoughts

Tanzania is an LMIC, where there are financial, logistical and infrastructure-related obstacles that have hindered researchers interested in exploring AMR in the region. Thus, emerging research should be appreciated and the context considered when evaluating. Unfortunately, this study was limited in some aspects which prevent it from supporting key conclusions and hypotheses sufficiently.

The authors stated that CTX-M, TEM and SHV were the most common ESBLs within their study, however those were the only genes examined. More importantly, only some TEM and SHV variants are ESBLs. non-ESBL TEM variants are occasionally found in composite transposons with CTX-M while non-ESBL SHV are commonly found on self-conjugating plasmids. Indeed, the authors of the multiplex method stated that direct genotyping of ESBL variants using their primers was not possible. Therefore, it cannot be confirmed via PCR amplification alone as to whether the TEM or SHV genes identified contribute to an ESBL phenotype.

However, the phenotypic tests show differences between strains that are unlikely to be accountable through CTX-M alone. The most interesting aspect of this study, in my view, is what has not been shown. The five p-ESBL strains could carry different ESBL genes or manifest the phenotype through mutations such as gene amplification, hyper-transcription or other mechanisms. Four isolates were meropenem resistant, which is extremely concerning and warrants sequencing, especially because at least two colonised blood.

These two were among the "pairs" identified by the authors. The basis of these pairings were phenotypic resistance and gene carriages, which is insufficient to support a pathogen shedding from the child to a healthcare worker to the care unit surfaces hypothesis. Especially when the sample sizes for non-faecal/blood isolates are so low and the lack of genomic analysis. However, reminders to maintain sanitation standards is always prudent and wise.

Considering the uncertainty of TEM or SHV contributing to p-ESBL and the limited genes tested, pairing based on gene carriage may have dismissed strains which do share common genomic features that confer p-ESBL. From which, a larger subset of isolates which potentially translocated from the gut to the blood could be formed for a future study.

As my project involves isolates from Tanzanian neonates, this article piqued my interest and helped develop my understanding of the setting and AMR context. Though this study has limitations, I believe there is potential yet to be uncovered.