Transmissible colistin resistance encoded by mcr-1 detected in clinical Enterobacteriaceae isolates in Singapore

Dear Editor, Following the recent description of transmissible plasmid-mediated colistin resistance encoded by mcr-1 in clinical and veterinary Escherichia coli and Klebsiella pneumoniae isolates in China,1 several groups have reported mcr-1 in colistin-resistant isolates from humans and food animals across Asia2–4 to Europe.5–8 It is evident that mcr-1 has disseminated globally. We performed a prospective study in January 2016 using 306 consecutive clinical Enterobacteriaceae isolates collected from blood, urine and miscellaneous samples (swabs and pus). The species investigated were E. coli (n= 166), K. pneumoniae (n= 87), Klebsiella oxytoca (n= 4), Enterobacter spp. (n= 22), Proteus spp. (n= 10), Citrobacter spp. (n= 9), Morganella morganii (n= 5), Providencia rettgeri (n= 1), Salmonella enteritidis (n= 1) and Serratia marcescens (n= 1). Isolates were PCR-screened for mcr-11 without prior knowledge of their antibiograms or colistin-resistance status. Three of these isolates (two E. coli and one K. pneumoniae) were mcr-1 positive, with their full-length mcr-1 gene matching the nucleotide identity of the first-described isolate exactly.1 Multilocus sequence typing (MLST) was performed (http://bigsdb.web.pasteur.fr/index.html). There was no evidence of nosocomial transmission, as the two E. coli isolates were of different sequence types (STs; Table 1). The isolates were urinary specimens (Table 1). Our detection rate was estimated to be 0.98% (95% confidence interval of 0.3%–2.8%, Wilson score interval). This was very close to 1% mcr-1 prevalence in China.1 Sensitivity testing was performed via E-test for polymyxin B and colistin and with the Vitek2 GNR257 card for all other antimicrobials. Phenotypically, the isolates were resistant to polymyxin B (minimum inhibitory concentration (MIC) 4 mg/L–24 mg/L) and colistin (MIC 4 mg/L–8 mg/L) but were sensitive to carbapenems.