Changes in colistin resistance and mcr-1 abundance in Escherichia coli of animal and human origins following the ban of colistin-positive additives in China: an epidemiological comparative study

Background Following the discovery and emergence of the plasmid-mediated colistin resistance gene, mcr-1, the Chinese government formally banned colistin as an animal growth promoter on April 30, 2017. Herein, we report patterns in colistin resistance and mcr-1 abundance in Escherichia coli from animals and humans between 2015 and 2019, to evaluate the effects of the colistin withdrawal. Methods We did an epidemiology comparative study to investigate: annual production and sales of colistin in agriculture across mainland China according to data from the China Veterinary Drug Association from 2015 to 2018; the prevalence of colistin-resistant E coli (CREC) in pigs and chickens in 23 Chinese provinces and municipalities as reported in the China Surveillance on Antimicrobial Resistance of Animal Origin database from Jan 1, 2015, to Dec 31, 2016, and Jan 1, 2017, to Dec 31, 2018; the presence of residual colistin and mcr-1 in faeces from 118 animal farms (60 pig, 29 chicken, and 29 cattle) across four provinces over July 1, 2017, to August 31, 2017, and July 1, 2018 to August 31, 2018; the prevalence of mcr-1-positive E coli (MCRPEC) carriage in healthy individuals attending routine hospital examinations across 24 provinces and municipalities from June 1 to July 30, 2019, comparing with equivalent 2016 data (June 1 to September 30) from our previous study in the same hospitals; and the patterns in CREC prevalence among hospital E coli infections across 26 provinces and municipalities from Jan 1, 2015, to Dec 31, 2016, and Jan 1, 2018, to Dec 31, 2019, reported on the China Antimicrobial Surveillance Network. Findings After the ban on colistin as a growth promoter, marked reductions were observed in the production (27 170 tonnes in 2015 vs 2497 tonnes in 2018) and sale (US$71·5 million in 2015 vs US$8·0 million in 2018) of colistin sulfate premix. Across 118 farms in four provinces, mean colistin residue concentration was 191·1 μg/kg (SD 934·1) in 2017 versus 7·5 μg/kg (50·0) in 2018 (p<0·0001), and the median relative abundance of mcr-1 per 16S RNA was 0·0009 [IQR 0·0001–0·0059] in 2017 versus 0·0002 [0·0000–0·0020] in 2018 (p=0·0001). Across 23 provinces and municipalities, CREC was identified in pig faeces in 1153 (34·0%) of 3396 samples in 2015–16 versus 142 (5·1%) of 2781 in 2017–18 (p<0·0001); and in chickens in 474 (18·1%) of 2614 samples in 2015–16 versus 143 (5·0%) of 2887 in 2017–18 (p<0·0001). In hospitals across 24 provincial capital cities and municipalities, human carriage of MCRPEC was identified in 644 (14·3%) of 4498 samples in 2016 versus 357 (6·3%) of 5657 in 2019 (p<0·0001). Clinical CREC infections in 26 provinces and municipalities comprised 1059 (1·7%) of 62 737 E coli infections in 2015–16 versus 794 (1·3%) of 59 385 in 2018–19 (p<0·0001). Interpretation The colistin withdrawal policy and the decreasing use of colistin in agriculture have had a significant effect on reducing colistin resistance in both animals and humans in China. However, continuous colistin monitoring is essential, in particular to act as an early warning system for colistin stewardship in Chinese hospitals. Funding National Key Research and Development Program of China, National Natural Science Foundation of China, and UK Medical Research Council.