Invited review: Antimicrobial resistance genes in milk-A 10-year systematic review and critical comment

The occurrence of antibiotic resistance genes (ARG) in milk is eagerly discussed as a public health risk, and frequently investigated. Here, we perform a systematic review on the abundance of antimicrobial resistance genes in milk from primary production over a 10-year period. We aim to provide a comprehensive dataset on known and emerging antimicrobial resistance genes in major mastitis pathogens, occurring worldwide in milk at primary production, and to critically discuss the relevance and constraints of these findings. We searched PubMed for peer-reviewed studies published between 2012 and 2022 that fit fixed combinations of key words and did not meet exclusion criteria such as "mixed with other sources." For synthesis, data on occurrence was extracted from studies and supplements. To address plausibility issues, we performed an National Center of Biotechnology Information Basic Local Alignment Search Tool (BLAST) search. Our search revealed 2,222 publications in total. Of them, 500 studies were eligible for full-text reads and 306 publications were included in data compilation. An overwhelming majority of studies dealt with mecA in Staphylococcus aureus, followed by extended-spectrum β-lactamase-encoding genes such as blaCTXM in Escherichia coli, while other mastitis pathogens, such as Streptococcus spp., were scarcely investigated. In most cases, <5% of milk samples were positive for major pathogens bearing the antimicrobial resistance gene of interest. However, huge study-to-study differences were found between regions, but also on a national level. For instance, the estimate prevalence of Escherichia coli-borne blaCTXM in mastitis milk samples ranged from 0.0% to 55%, with a median value of 7.3%, but in healthy individuals and bulk milk, the prevalence ranged from 0.0% to 20.0%, with a median value of 0.8%. Several studies reported antimicrobial resistance genes for the very first time in a species, but did not stand up to scrutiny. As an example, frequent detection of blaTEM-genes in streptococci is most likely attributed to contamination of molecular reagents, as reported elsewhere. Despite the large amount of data, there is a need for more quality control, more representative sampling of milk, more quantitative research, and deeper insights into bacterial genomics, to identify relevant or emerging antimicrobial resistance genes in milk. Considering a low percentage of contaminated milk samples, unknown ARG concentrations, and an unproven role in human disease, the risk attributed to ARG in milk seems to be exaggerated by far. However, the risk of ARG selection on farm, resulting in low treatment success in cattle, is a real one and should be met by prudent use of antibiotics.